-------
the greater insolubility of B(a)P compared to pyrene. With a decrease
in the amount of radiolabel in the methylene chloride extracts, there
was an increase in the radiolabel found associated with the sediment.
Further evaluation of the sediment with acid/base extraction pro-
cedures indicated that approximately 75% of the label contained in the
sediment was associated with the inorganic fraction while 25% was
found associated with the organic fraction.
DISCUSSION
The conclusions from this study can only be interpreted for the par-
ticular soil used. Soil used in this study was McLauren sandy loam
from a Mississippi hazardous waste site and was acidic (ph * 5.4) with
very little organic carbon content (<0.5%). Soils with more organic
carbon (e.g., humus) may show larger increases of radiolabel concen-
tration in acid/base extractions. The vermiculite clay in this soil does
not hydrate easily, nor does it have the absorption properties of other
clays, such as montmorillinite. Soils with hydratable clays should show
a greater affinity for chemicals which would influence the rate of dis-
appearance of the parent compound.
Through capture of evolved radiolabeled carbon dioxide, minerali-
zation of B(a)P and pyrene was demonstrated to occur albeit at extremely
slow rates.
Increase toxicity from the water extract together with the greater
amounts of radiolabel in the methanol eluate compared to methylene
chloride eluate suggests formation of oxidized intermediates or by-
products. Identification of the polar intermediates is being evaluated
through HPLC fractionation and GC/MS spectrometry.
The increase in partitioning of radiolabel into the acid/base-extracted
fractions suggests adsorption of the parent compound and/or transfor-
mation products on inorganic and organic soil fractions. This result
also implies the possible increase in number of ionizable functional
groups causing chemical coupling or polymerization (e.g., humifica-
tion) which renders the radiolabel immobile. Increased association of
the radiolabel with the organic fraction over time suggests that the
kinetics of humification and/or absorption of the organic contaminant
may be an important factor to consider when developing detoxification
measures for polluted soils. Results from the comparison of sterile vs.
nonsterile microcosms suggest that immobilization as well as humifi-
cation may be mediated abiotically and biotically.
An important observation from this study is that half-lives of chemi-
cals in soil determined by chemical extractions should not be confused
with biodegradation or mineralization. Without further investigation
into the actual fate of the chemical in the extraction sediments, including
a complete mass balance analysis, complete detoxification cannot be
properly evaluated.
Further investigation of abiotic and biotic reactions of soil with chemi-
cal components describing the nature of the adsorption mechanisms
needs to be performed. This information can then be correlated with
the chemical or physical characteristics of soils.
Site-specific, soil characteristic studies must be conducted if we are
to successfully detoxify the many hazardous waste sites located in the
United States. Due to the different localities, types of soil, weather con-
ditions, etc., site-specific information should be considered when evalua-
ting detoxification approaches.
CONCLUSIONS
Association of the chemical with the inorganic fraction was a dominant
partitioning process. This reaction could account for the largest decrease
in concentration of label to the methylene chloride extract. Despite the
large amounts of inorganically associated label, there were significant
amounts of the radiolabel associated with the organic fraction. The
primary mechanisms of disappearance of the parent compound appeared
to include partial chemical oxidations together with adsorption of com-
pounds to spoil components.
ACKNOWLEDGEMENT
Thanks to Dr. Ingeborg Bossert for her generosity, Lula Hassan
Abusalih for her dedication and consistent work in the laboratory and
Dr. Michael McFarland for his assistance in preparing the final
manuscript.
REFERENCES
1. Herbes, S.E., Southworth, G.R. and Gehrs, C.W., "Organic Contamina-
tion in Aqueous Coal Conversion Effluents: Environmental Consequences
and Research Priorities," in Tract Substances in the Environment, ed. D.
D. Hemphill. Univ. of Missouri, Columbia, MO, 1976.
2. Mahmood, R. J., Enhanced Mobility of Polynuclear Aromatic Hydrocar-
bons in Unsaturaied Soil, PhD Dissertation, Department of Civil and
Environmental Engineering, Utah State University, Logan, UT, 1989.
3. Sims, R. C., "On-sile Bioremediation of Wood Preserving Contaminants
in Soils." Proc. of the Forum: Technical Assistance to U.S. EPA Region IX;
forum on Remediation of Wood Preserving Sites, ed. E. F. Barth and J. E.
Matthews. U.S. EPA, Oct. 24. San Francisco, CA, 1989.
4. U.S. EPA. Bioremediation of Hazardous Ubsle Sites Workshop, CER1-89-11,
U.S. EPA Center for Environmental Research Information, Cincinnati, OH,
1989.
5. Bossert. L., Kachel. W. M. and Banha, R., "Fate of hydrocarbons during
oily sludge disposal in soil," AppL Environ, Microbiot,, 47, pp. 764-767,1984.
6. Bulman, T.S., Lesage, S., Fowlie, P. J. A. and Weber, M. D., The Persis-
tence ofPolynuclear Aromatic Hydrocarbons in Soil, A report prepared for
the Petroleum Association for Conservation of the Canadian Environment
(PACE) Report No. 85-2, by Environment Canada, Wastewaler Technology
Centre, Burlington, Ont.. 1985.
7. Coover. M. P. and Sims. R. C. "The rate of benzo(a)pyrene degradation
in a manure amended sandy loam soil." Haz. Wistr/Haz. Mat., 4(2), pp.
151-158, 1987.
8. Sims. R. C andOvercash, M. R.. "Fate of polynuclear aromatic compounds
in soil-plant systems." Residue Reviews. 88, pp. 1-68, 1983.
9. Sims. R. C. Sims. J. L.. Sorenscn, D. L. and Hastings, L. L., Waste/Soil
Treotobility Studies for four Complex Industrial Hbstes: Methodologies and
Results, EPA-600/6-86-003a,b, Office of Research and Development, Robert
S. Kerr Environmental Research Laboratory. U.S. EPA, Ada, OK, 1986.
10. Schoeny, R., Cody. T., Wirshawsky, D. and Radike. M., "Metabolismof
mutagenic polycyclic aromatic hydrocarbons by photosynthetic algal spedes."
Mutation Research, 197 pp. 289-302. 1988,
11. Miller, W. E.. Greene. J. C and Shiroya, T., "The Selenastnun capricor-
nutum Printz algal assay bottle test experimental design, application, and
data interpretation protocol." U.S. EPA. EPA-600/9-78-018, 1978.
12. Beckman Instruments, Inc., Beckman Microtox Systems Operating Manual,
Microbial Operations, Carlsbad, Ca, 1982.
13. Symons. & D. and Sims. R. C. "Assessing detoxification of a complex
hazardous waste, using the Microtox™ Bioassay," Arch. Environ. Contain.
Toxicol., 17. pp. 497-505, 1988.
14 Coover. M. P., Sims, R. C. and Doucene, W. J., "Extractionof polycyclic
aromatic hydrocarbons from spiked soil," J. Assoc. Off. Anal Chem., 70,
pp. 1018-1020, 1987.
15. Schoeny, R., Cody, T.. Radike. M. and Warshawsky, D., "Mutageniciry
of algal metabolites of benzo(a)pyrene for Salmonella typhimurium. Environ-
mental Mutagenesis," 7. pp. 839-855, 1985.
16. Skujins. J. and Richardson, B. Z., "Humic matter enrichment in reclaimed
soils under semi-arid conditions" GermicrobiologyJ. ,"< pp. 299-311,1985.
26 MONITORING &«6AMPLING
-------
Comparison of Shallow Electromagnetic and the Proton
Precession Magnetometer Surface Geophysical Techniques to
Effectively Delineate Buried Wastes
Todd Struttmann, P.E.
Tom Anderson
Metcalf & Eddy, Inc.
Columbus, Ohio
ABSTRACT
During the ground preparation for the construction of a building,
buried drums were encountered and the construction was halted. Geo-
physical techniques were used to rapidly delineate the location and extent
of the hurried waste. The results of the geophysical survey were used
immediately to direct waste excavation.
The primary objective of this paper is to compare two surface geo-
physical techniques, an electromagnetic conductivity meter (Geonics
EM-31) and a proton precession magnetometer (Geometries 856A), and
their use at a hazardous waste site. The instrument responses were com-
pared using statistics. Combined responses lead to an improved interpre-
tation of the subsurface. The EM-31 was less affected by man-made
surface features.
This paper compares the anomalies seen in the geophysical data with
field-verified physical and chemical features. The intent is to help the
reader determine which instrument to use, given knowledge of the field
conditions, and to aid in interpretation of anomalies.
INTRODUCTION
This paper presents the results of a surface geophysical investigation
at a hazardous waste site. All geophysical anomalies suspected of corres-
ponding to buried waste were investigated directly with excavation equip-
ment. Verification trenching was conducted in non-anomalous areas.
This direct investigation of the anomalous and non-anomalous areas,
along with a thorough analysis of the anomalies created by man-made
and geologic features, provided the authors a unique opportunity to
evaluate fully the effectiveness of the geophysical techniques used.
The surface geophysical techniques used to delineate buried paint
and paint process waste were an electromagnetic (EM) terrain conduc-
tivity meter (Geonics EM-31) and a proton precession magnetometer
(Geometries G856). Preceding the survey, historic aerial photos were
gathered and employee interviews were conducted to determine the
approximate locations of the waste pits and gain an understanding of
the burial practices employed on-site. This information was used to iden-
tify patterns in the former disposal practices that could be used to define
the investigation boundaries and estimate the probable size and depth
of the waste disposal area(s).
It is a common practice to use several geophysical techniques to inves-
tigate the subsurface. Neev1 used electromagnetics and a magne-
tometer to do preliminary site investigations looking for buried ordnance.
McGuinness2 used ground penetrating radar and magnetometer tech-
niques to locate buried drums. Benson3 suggests that a combination of
techniques will aid in the interpretation of the data.
The objective of this paper is to compare the EM and magnetometer
responses to buried waste and other anomalous cultural and geologic
features at the site. Although both instruments are used widely in geo-
physical investigations, their result have not often been compared.
It became very important to show that the geophysical data were useful
in delineating areas where no waste burial had occurred because the
area being investigated was the future site of a 250,000-ft2 building.
Backhoe trenching operations generated information to verify these data.
The verification trenching provided a high level of confidence in the
geophysical data, suggesting that all of the buried waste pits had been
found and removed. In light of this confidence, the state environmental
agency involved approved the placement of the building over the site.
BACKGROUND
When a manufacturing company began ground-clearing for its new
warehouse, a bulldozer doing surface leveling encountered some shallow
buried waste. The company's environmental manager halted the con-
struction work and hired Metcalf & Eddy, Inc. to locate and charac-
terize the waste and coordinate its removal and proper disposal.
Through initial soil sampling, employee interviews and the review
of historical aerial photographs, it was determined that the waste was
the result of former disposal practices used by the company from 1961
through 1979 (just prior to the enactment of RCRA).
Paint process wastes generated prior to 1980 were disposed of on-
site. It is unclear whether these paint wastes, which include ashes,
solvents, charred timbers, scrap metal and buffing pads, were disposed
of in burial pits after being burned in an on-site incinerator, or whether
Fig. 1
1973 Aerial Photograph
MONITORING & SAMPLING 27
-------
the waste was set on fire after being placed in the pits. A 1973 aerial
photograph (Fig. 1) shows the rough outlines of three small pond-like
areas with a road leading from the manufacturing facility. The photo-
graph gave an indication of the probable dimensions and locations of
the waste pits in 1973. Although this was just one point in time, and
there was no way of knowing exactly how representative this photo-
graph was, it does suggest a systematic process of waste disposal. This
wa« valuable information for the design of the investigation.
The topography is extremely flat, with less than 2 ft. elevation change
across the site. An initial review of the geological literature suggests
that the site is underlain by approximately 10 ft. of glacial till over more
than 300 ft. of shale. Groundwater occurs in fractures and other perme-
able zones in the shale.
Site leveling and landscaping associated with a building expansion
in 1980 had erased any surface evidence of the waste pits, as seen in
the 1973 aerial photograph. Because the bedrock was so near the surface
(approximately 10 ft.) and because of comments about the disposal prac-
tices from long-time employees, it was believed that the waste was buried
in very shallow pits. The aerial photographs suggested that the size of
the waste pits was on the order of 10 by 20 ft. to as much as 50 by 60 ft.
Employee interviews and a review of the manufacturing and painting
operations prior to 1980 (when on-site waste disposal was active) were
used to estimate the physical and chemical characteristics of the former
wastes.
Description of Equipment
Benson4 presents a methodology for deciding which geophysical
technique is appropriate based on site conditions. Based on the evidence
presented above, a decision was made to use a shallow electromagnetic
technique and a magnetometer to locate the buried waste.
A brief description of the basic theory and use of the Geonics EM-31
electromagnetic conductivity meter and the Geometries 856A proton
precession magnetometer follows.
Electromagnetic techniques measure the terrain conductivity by
imparting an alternating current to a transmitter coil placed on or near
the earth's surface5. The current passing through the transmitter coil
produces a magnetic field, which in turn induces small currents in the
underlying strata. Currents within the strata produce a secondary mag-
netic field, which is sensed by a receiver coil. It has been shown that
the ratio of the electromagnetic field detected by the receiver coil to
the electromagnetic field produced by the transmitter coil is directly
proportional to terrain conductivity. This allows terrain conductivity
to be read directly from the instrument in millimhos per meter
(mmhos/m).
The terrain conductivity value read by the instrument is an average
conductivity over the effective depth of the survey. The effective depth
is determined by the intercoil spacing (i.e., distance between the trans-
mitting and receiving coils) used in the survey. The Geonics EM-31
electromagnetic terrain conductivity meter was used at this site. It has
an intercoil spacing (distance between receiver coil and transmitter coil)
of 3.66 m and an effective depth of approximately 6m.
A proton precession magnetometer, such as the Geometries G-856A
portable proton magnetometer used at this site, utilizes the precession
of spinning protons or nuclei of the hydrogen atom in a hydrocarbon
fluid to measure the total magnetic intensity*. The spinning protons in
the fluid behave as small, spinning magnetic dipoles. These magnets
are aligned or polarized temporarily by the application of a uniform
magnetic field generated by a current in a coil of wire in the G-856A.
The processing protons then generate a small signal in the same coil
used to polarize them. This signal has a frequency that is precisely
proportional to the total magnetic field intensity and independent of
the orientation of the magnetometer. The proportionality constant, which
relates frequency-to-field intensity, is the atomic constant, the gyromag-
netic ratio of the proton. The precession frequency is measured by digital
counters as the absolute value of the total magnetic field intensity with
an accuracy of 1 gamma in the earth's field of approximately 50,000
gammas.
The total magnetic field intensity, as measured by the proton
magnetometer, can be looked upon as the magnitude of the earth's field
vector, independent of its direction. Local disturbances, as might oc-
cur near buried metal objects, will add to the earth's magnetic field
in the usual manner of vector addition. The local disturbances leave
signatures (anomalies) in the data that can be very useful for locating
buried metal objects.
Limitations in the use of the instrument come about when local large
magnetic anomalies that are not the target of the investigation distort
the signal greater than the targeted anomaly, effectively masking the
target anomaly. This problem is common when magnetometer surveys
are conducted near buildings, power lines, underground pipelines, etc.
Multiple techniques can be used with the EM, including vertical vs.
horizontal dipole, profiling and phasing.7 Magnetic gradient measure-
ments can be taken with the magnetometer.6 None of these techniques
were used since the additional information they would have provided
was not necessary for this particular site. However, these applications
may be very useful elsewhere.
COLLECTION OF DATA
The aerial photograph (Fig. 1) indicates three areas of possible waste
disposal. The first geophysical survey was made over a 200 x 400 ft.
area covering the locations of the suspected areas and conducted on
evenly spaced 20 ft. grids running north and south parallel to the existing
building.
After the grids were set up. the instruments were calibrated and data
were collected on every node. Terrain conductivity readings using the
EM-31 were made in a north-south and east-west orientation over each
node. The differences in readings at a node were later used as an addi-
tional criterion for determining anomalies. The readings were observed
when walking between nodes, and any changes of over 5 mmhos/m
dictated collecting mid-node readings Total field magnetic intensity
readings were made at each node using the Geometries 856A proton
precession magnetometer.
Following the initial survey, the grid was expanded to 400 x 700 ft.
to collect data over the entire area on which the new building was to
be built.
PRESENTATION OF DATA
This section presents the data collected with the EM and magne-
tometer. Figures 2 and 3 show topographic contour maps of the data
from the EM and the magnetometer (entire 400 x TOO ft. survey). Note
that magnetometer data in Figure 3 show a decrease in intensity on
the eastern edge. This decrease is due to the magnetic properties of
the steel-frame building that is 50 ft east of the survey area.
Building Anomaly correction
The magnetometer anomaly caused by the steel-framed warehouse
50 ft. from the eastern edge of the survey area can be modeled. Because
the southern half of the survey was relatively free of anomalies (except
for the building), the effect of the building can be calculated. Taking
the average of each north-south line in the southern half of the survey
(0 to 300 ft north) and comparing these values to the average values
for the western pan of the survey, the change due to the building is
determined.
Figure 4 is an x-y graph showing the deviation in gammas versus
the distance from the building. The actual deviation as well as the
modeled theoretical deviation are shown. The modeled deviation treats
the building as a series of monopoles and varies as the inverse square
of the distance from the source (i.e., 1/r)6. Now the building anomaly
(900 gammas at the eastern-most edge) can be subtracted from the raw
data and the data re-analyzed. The replotted magnetometer data are
shown in Figure 5.
Defining Anomalies Using Statistics
The geophysical anomalies were defined using a simple statistical
approach. The data were resolved by taking the average and standard
deviation of the population and then considering as anomalous any data
points outside plus or minus one standard deviation from the mean.
This approach was intended to remove the subjective nature of deciding
what is anomalous and assumes that the background readings, including
28 MONITORING & SAMPLING
-------
700
600 -
500 -
400
300 -
200
100 -
100
200
300
400
700
• 600
500
400
- 300
200
100
0 100 200 300 400
CONTOUR INTERVAL 4 mmohs/m
Fig. 2
Electromagnetic Conductivity Date.
N-S Orientation.
noise, fell inside one standard deviation and that the anomalous readings
fall outside that range. The success of this approach is discussed later.
Analyzing the 762 data readings collected on the magnetometer survey,
the mean was 55,408 gammas, with a range from 53,380 to 56,361 and
a standard deviation of 254 gammas. Using the approach mentioned
above, an anomaly would have to be more than 250 gammas to be dis-
tinguished.
Statistical analysis was performed on the modified readings that
accounted for the building anomaly (values plotted in Figure 5). Using
this approach, the mean is 55,529 with a standard deviation of 148
gammas. Based on this approach, any variation over 150 gammas was
considered an anomaly. This second set of statistics was used in this
paper.
EM Statistics
The same statistical approach can be used with the EM data as the
magnetometer. Using all the data, the mean was 24.6 mmhos/m with
a range from 0 to 69 and a standard deviation of 12 mmhos/m.
As with the magnetometer data, the desire is to determine the changes
from the surroundings rather than the absolute readings. As seen in
Figure 2, the values of conductivity are of two distinct levels; in the
30 range in the northern section and in the teens in the southern section.
The transition between these two levels occurs at between 350 and 450 ft
north on the grid.
Upon further investigation of the geotechnical borings gathered during
the building design, a change in depth to bedrock was noted. Figure 6
shows the depth to bedrock along a north-south section of the site and
700
600 -
500
400 -
300 -
200 -
100 -
0 t*
100 200 300 400
I I I I I I I I I I I r-K4;
700
600
- 500
- 400
^ 300
200
- 100
0 100 200 300 400
CONTOUR INTERVAL 100 GAMMAS
Fig. 3
Total Magnetic Field Contour Map
ASSUMES T-k/r*2
WHERE k-2.5ŁB AND r-DlSTANCE TO BUILDING
400 390 360 340 320 300 280 260 240 220 200
EAST - WEST GRID LOCATION
Fig. 4
Building Anomaly.
Actual Deviation vs. Calculated
MONITORING & SAMPLING 29
-------
0 100 200
700 I I I I I I I I I I I I'
600 -
500 -
400 -
300
200 -
100 -
300
400
700
600
- 500
400
100 200 300
CONTOUR INTERVAL 100 GAMMAS
Fig. 5
Total Magnetic Field Contour Map
After Removal of Cultural Anomalies
•J 300
200
- 100
400
NORTH - SOUTH GRID LOCATION
D - AVERAGE CONDUCTMTV VALUES ALONG CAST-tCST UNC
A - DCPTH TO BEDROCK M BORINGS
REPRESENTS THE Entcnvt DEPTH or THE EW-JI
Fig. 6
Conductivity Change Due to Geological Structure
the corresponding average EM readings. The northern section is 7 to
10 ft to bedrock; the middle section is approximately 15 ft and the
southern section did not encounter bedrock in the first 30 ft.
Recall that the depth of investigation of the EM-31 is approximately
18 ft. The EM-31 in the northern section was responding to a two-layer
system of glacial till and shale, and the southern section was responding
to only the till. Because the fractured shale has a higher conductivity,
a higher reading in the north resulted.
With this observation in mind, the EM data were separated into two
sections prior to statistical analysis (as northern and southern segments).
Area
Southern
Northern
Statistical Analysis of EM Data (mmhos/m)
Mean Range Standard Deviation
15.3 0-27 3.3
35.6 0-69 96
Using this approach, two criteria arise for determining an
anomaly: greater than 4 mmhos/m in the southern section and
greater than 10 mmhos/m in the northern.
Absolute Difference Approach With the EM-31
It is very likely that the waste in the burial pits was placed
randomly and of varying conductance. Therefore, electrical con-
ductivity readings from varying directional orientations of the
EM-31 could have different values.
Figure 7 is a plot of the absolute value of the difference in
EM conductivity from the north-south and east-west orientations.
This method identified all of the anomalies related to buried
waste, plus additional responses to subsurface features.
100
200
400
300
200
100
I
I
l
l i i i
400
300
200
100
0 100 200
CONTOUR INTERVAL 2 mmhos/m
Fig. 7
Electromagnetic Conductivity Data.
Absolute Vdue of the Difference Between the
N-S and E-W Orientation.
30 MONITORING & SAMPLING
-------
Applying statistics to this set of data yields an average value of 1.2
mmhos/m, with conductivity differences ranging from 0 to 30 mmhos/m
and a standard deviation of 2.1 mmhos/m.
Man-Made Anomalous Features
There were three man-made surface features on this site that created
anomalies in the geophysical data (Figs 8 and 9): the building, a trailer
on-site for personnel decontamination and a paved road.
700
600 -
500
400 -
300
200 -
100 -
100 200 300
CONTOUR INTERVAL 100 GAMMAS
400
Fig. 8
Total Magnetic Field Contour Map
After Removal of Cultural Anomalies.
Anonamous Areas Delineated.
The building had a large impact on the magnetometer data, even
though the survey never got closer than 50 ft to the building. The
building had no significant anomalous effect on the EM data.
The decontamination trailer was located on the northern edge of the
site. The trailer had a steel frame, aluminum body and dimensions of
10 x 25 x 10 ft (high). Anomalous values were seen in both the EM
and magnetometer data due to this feature, although the effect on the
EM was only at one point (this point, 700 north - 360 east, was not
used in the analysis of the EM data). The trailer created an anomalous
feature in the magnetometer data that extended up to 60 ft radially (Fig.
11), but did not affect the resolution of other anomalies.
The blacktop road (Fig. 11), given the accuracy of the magnetometer
data and its close proximity to another anomalous feature (the building),
had no distinguishable impact on the magnetic data. The effect of the
road on the EM data was fairly uniform and consisted of elevated con-
ductivity values of 6 to 8 mmhos/m.
Several other anomalies identified by the data are shown and labeled
A through K on Figures 8, 9 and 10. Since all the anomalies related
700
600 -
200 300 / AOO
100 200 300
CONTOUR INTERVAL 4 mmhos/m
Fig 9
Electromagnetic Conductivity Data.
N-S Orientation.
Anomalous Areas Delineated.
to known man-made surface features have been identified and delineated,
these labeled anomalies represent subsurface features with electro-
magnetic or magnetic properties significantly different from ambient
conditions.
INTERPRETATION AND VERIFICATION OF ANOMALIES
All of these anomalies (A through K on Figures 8, 9 and 10) were
then investigated directly by excavating the waste and digging additional
confirmation trenches with a backhoe. Buried waste was removed to
a temporary storage pad away from the proposed building site. Borings,
both from a geotecrinical investigation for the building site and from
monitoring well installations, supplied additional correlative data.
The waste excavation, verification trenches and soil borings created
a unique opportunity for the authors to compare each geophysical
instrument's response to anomalous features at a hazardous waste site.
Figure 11 outlines excavated waste pits, verification trenches and soil
borings with respect to the geophysical anomalies. The waste pit exca-
vations shown in Figure 11 include the excavation of natural soils under
and around the waste pits that had been contaminated by an organic
leachate from the pits. The actual dimensions of the waste pits are some-
what less than shown. This organic leachate was not detected by the
geophysical instruments.
Table 1 summarizes each geophysical anomaly detected. Included
in the table are a physical description of what was found during exca-
vation and trenching, the maximum instrument responses of each tech-
nique in relation to surrounding values and whether these responses
were sufficient to be called an anomaly (> 1 standard deviation above
the surrounding values).
MONITORING & SAMPLING 31
-------
400
400
100
200
300
700
100 200
CONTOUR INTERVAL 2 mmhos/m
Fig. 10
Electromagnetic Conductivity Data.
Absolute Value of the Difference Between the
N-S and E-W Orientation.
Anomalous Areas Delineated.
All these techniques detected the waste pits (Anomalies A to E), but
each instrument also responded to additional surface and subsurface
features. The EM readings also detected two natural subsurface features
(bedrock slope and subsurface tree roots). Taking the difference between
EM orientations at a node eliminated these two, but detected two addi-
tional features (concrete slab and surface material). The magnetome-
ter was not affected by the natural subsurface features and the smaller
man-made features, but it was highly affected by the nearby building.
Each instrument was "fooled" by extraneous geologic or man-made
features, but had a high correlation in detecting waste when all three
techniques detected an anomaly.
Figure 12 shows the geophysical properties of anomalies detected and
also highlights the anomalies where buried waste was found. The wastes
detected were either ferromagnetic (responded to magnetometer),
possessed a significantly different conductivity value than background
(responded to EM reading) or were heterogeneous (responded to EM
orientation). Using the multiple property approach yielded a much
higher success at predicting where the buried waste was located. The
only feature that "fooled" all three techniques was the buried tin
building, which coincidentally possessed the same three properties of
the bnriwl wast**
Each instrument responded to different physical properties of the
buried material. Given a different set of field conditions, one instru-
ment may perform better than another.
600
10
500
400
300
200
100
ij- OB'
?.5'O
- O>is'
O>15'
14-e
^-L
1 1 1
400
300
200
100
100
200
300
400
EXPLANATION
OS' - BORING - DEPTH TO BEDROCK
(D>30' - MONITORING WELL - DEPTH TO BEDROCK
F. :.v:;j - EXCAVATIONS AND TRENCHINGS
Fig. 11
Location of waste Excavation.
Trenching and Soil Borings
The desire is to compare how well each instrument responded to each
anomaly, but the magnetometer reads in gammas and the EM in
mmhos/m. To be able to compare the two instruments, a "signal-to-
noise ratio" was calculated using the maximum response divided by
the standard deviation of the readings for the survey. In this manner,
the relative responses of each instrument can be compared. Table 2
shows the signal/noise ratio of each anomaly using the three techniques.
The EM and magnetometer both responded to all five of the waste
areas (anomalies A to H) Additionally, the EM responded to four more
subsurface features (a buried tin building, shale escarpment, a buried
copper pipe and tree roots) and the magnetometer to one more sub-
surface feature (a buried tin building) and one cultural feature (the
building).
The magnetometer had an average signal-to-noise ratio of nearly four
in the waste areas (anomalies A to E), compared to approximately three
for the EM. This suggests that the magnetometer may have been slightly
32 MONITORING & SAMPLING
-------
Anomaly
Location'
Maximum Instrument Response
Above Background
EM (milliMhos/ml Mag (gamma')
28
40
23
37
11
13
15
12
250
900
500
700
400
200
EM - Difference
N-S less E-W
Orientation
28
12
14
Table 1.
Anomalies Defined
Does Response Meet
Statistical Criteria? EM
EM (10 mmhos/m in North, MAG Difference Physical Description
4 mmhos/m in South')2 (150 gammas') > 3.3 mmhos/m Based on Excavation Work3
900
K
0
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
No
No
Yes Waste pit: 12'x 50', 6'depth,
2' below surface containing two drums.
Yes Waste pit: 15'x 25', 7'depth
2' below surface, no drums.
Yes Waste pit: 50'x 60', 3'- 5'depth,
1' below surface, some drum pieces.
Yes Two waste pits: 20'x 60'and
30' x 40', 3' - 5' depth, more scrap
metal, eastern pit 25 drums, scrap
copper wire and 5-gallon cans.
Yes Waste pit: numerous pits 5' - 10'
diameter, 4' depth, 1' below surface,
some drums, numerous 5-gallon cans.
Yes Tin battery remains: 7' x 15', 4' depth,
2' below surface, sheets of tin, lumber,
nails, concrete; no waste.
No Geologic feature: bedrock; slopes to
the South.
No Tree Root Zone: Aerial photo shows
wooded zone between disposal area and
farm field. 4' depth, 2' below surface.
No Man-made feature: building 50' east of
survey area.
Yes Concrete slab: 3' x 5' x 6", 2' below
surface, contained 1/2" copper pipe;
used as a ground in prior field office.
Yes Surface material:nosubsurfacefeaturefound.
1 Location as shown on Figures 10, 11, and 12.
2 Statistical Criterion: > 1 standard deviation response.
3 Waste pits contained an assortment of waste, including ash, paint sludge, miscellaneous scrap metal, and occasionally partially intact 55-gallon drums.
Table 2
Relative Response to Anomalies
FERROMAGNETIC
(MAGNETOMETER)
Anomaly
A
B
C
D
E
F
G
H
I
J
K
Relative Response1
EM EM-DIF MAG
- BURIED WASTE FOUND
2.9
4.2
2.4
3.9
1.2
3.9
1.0
1.0
0.0
0.0
0.0
7.8
1.1
3.3
3.9
2.2
2.2
0.0
0.0
0.0
1.1
1.1
1.7
6.1
3.4
4.7
2.7
1.3
0.0
0.0
1.1
0.0
0.0
Fig. 12
Property of the Geological Anomalies
Comparing the Response of the Instruments
1 Relative Response is the maximum instrument response divided by the standard deviation of the data.
Analogous to a signal-to-noise ratio.
MONITORING & SAMPLING 33
-------
more responsive to the waste than the EM.
The magnetometer is much more sensitive to cultural features such
as buildings. Although the authors later accounted for the response due
to the building, other less intense anomalous features near the building
would have been difficult to distinguish from the building anomaly,
making the analysis more complex. The EM, on the other hand, was
not affected by the building and at 50 ft from the building did not show
a measurable instrument response.
CONCLUSIONS
The hazardous waste disposal practices used at this manufacturing
facility are most likely similar to many of those that can be found at
hundreds of other small- to medium-size manufacturing facilities that
were operating prior to the enactment of RCRA. It is likely that the
majority of these small waste sites will have to be dealt with in the
future. As with this site, there usually is very limited information avail-
able. The surface geophysical techniques discussed in this paper can
very effectively locate and delineate suspected waste disposal sites. The
information presented here should provide valuable help to others faced
with similar problems relative to the effective use and interpretation
of the surface geophysical data at a hazardous waste site.
The use of multiple geophysical techniques is much more diagnostic
than the use of only one instrument. By using two techniques, different
properties of a buried material are being tested. The presence of a posi-
tive response for both properties, conductivity and ferromagnetism, con-
sistently detected the buried waste and yielded a false positive less
frequently.
Each individual technique was "fooled" due to geologic or man-made
features. When two out of three techniques responded positively, buried
waste was found.
Considering all the information derived from the geophysical survey,
the EM-31 proved to be a more versatile investigative tool. It identi-
fied all the waste pits and provided information about the subsurface,
the location of a buried bedrock escarpment and the extensive root zone,
which was valuable to the hazardous waste site investigation.
REFERENCES
1. Neev, D., "Application of Geophysical Methods for Subsurface Metal
Screening: A Case Study," Proc. of the Symposium on the Application of
Geophysics to Engineering and Environmental Problems, SAGEEP, Mar.
1988.
2. McGinnis, L.D., Winter, R.C, Miller, S.E and Tome, C, "Decision Makiig
on geophysical techniques and results of a study at a hazardous waste silt,"
Proc. of the Symposium on the Application of Geophysics to Engineering
and Environmental Problems. SAGEEP, Mar 1988.
3. Benson, R.C, Glaccum, R.A. and Noel, M.R., Geophysical Techniques
for Sensing Buried Wales and Htute Migration, U.S. EPA. Contract No.
68-03-3050: Technos, Inc., Miami. FL, 1983.
4. Benson, R.C., Surface aruidawnhoU geophysical techniques for hazardous
waste site investigation, HML pp 9-18, 53-60, April/May 1988.
5. Geonics Limited, Overating Manual for EM-31-D Non-contacting Terrain
Conductivity Meier, June 1984.
6. Breiner. S.. Applications Manual for Portable Magnetometers: Geometries,
Sunnyvale, CA, 1973.
7 McNeil, J.D., "Advances in Electromagnetic Methods for Groundwaler
Studies," Proc. of the Symposium on the Application of Geophysics to
Engineering and Environmental Problems, SAGEEP. Mar. 1988.
34 MONITORING & SAMPLING
-------
Detection and Location of Leaks in Geomembrane
Liners Using an Electrical Method:
Case Histories
Daren L. Laine
Michael P. Miklas, Jr.
Southwest Research Institute
San Antonio, Texas
ABSTRACT
A field-proven electrical technique, developed at Southwest
Research Institute, San Antonio, Texas, is commercially avail-
able to detect and locate leaks in geomembrane liners. The elec-
trical technique is used to inspect 100% of the geomembrane ma-
terial that is covered by a conducting liquid. A voltage applied
across the liner produces a uniform electrical potential distribu-
tion in the liquid or soil above the liner when no leaks are present
in the geomembrane. If leaks are present, they are detected and
located by searching for localized anomalies in the potential dis-
tribution caused by current flowing through the leak in the geo-
membrane liner. Sixty-one new or in-service geomembrane-lined
waste storage facilities were investigated using the electrical leak
location method. An average of 3.2 leaks per 10,000 ft? were
located with a range of 0.3 to 5 leaks per 10,000 ft2 of liner sur-
veyed. Many leaks were located in new installations that had been
tested using conventional inspection tests.
INTRODUCTION
Survey Method
Figure 1 shows a diagram of the Southwest Research Institute
electrical leak location method which illustrates the technique
described in this paper. When no leaks are present, the high elec-
trical resistivity of the geomembrane liner material will prevent
electrical current flow from the liquid in an impoundment to the
earth ground or leak collection zone beneath the geomembrane
liner. When a voltage is impressed across a geomembrane liner
with no leaks, a relatively uniform potential voltage distribution
is found in the liquid or soil cover above the liner. If a leak exists
in the liner, conductive fluid will flow through the leak establish-
ing a path for electrical current. An anomaly in the measured
electrical potential is generated in the immediate vicinity of the
leak through which electrical current is flowing. Leaks can be
accurately located to less than 1 in. by searching for the point of
highest electrical potential.
Survey Equipment
The equipment used in a manual leak location survey consists
of a DC power source, lightweight man-portable electronic de-
tector, scanning probe and associated instrumentation as shown
in Figure 2. The probe is most conveniently used while wading in
the liquid. However, with an extension, it can be used from a
floating platform in deeper liquid applications.
CURRENT SOURCE
ELECTRODE
/4.yx yx y
f EARTH
Figure 1
Diagram of the Electrical Leak Location Method
Figure 2
Manual Leak Location Equipment Consisting of an Electrode
Probe and Electronics Unit
MANUAL LEAK LOCATION SURVEY IN
LIQUID IMPOUNDMENT
To conduct a manual leak location survey, a minimum of 12 in.
of a conducting liquid and a maximum of 30 in. of conducting
MONITORING & SAMPLING 35
-------
liquid (preferably fresh water) must cover the liner. Filling the
impoundment to the operating depth with fresh water is recom-
mended to hydrostatically load the liner prior to the leak location
survey. Testing the liner after hydrostatically loading it is a valid
method to determine if the liner will perform satisfactorily under
the intended operating conditions. The water is then lowered in
stages as the side slopes of the impoundment are electrically
tested. After the water has been lowered to 30 in. in depth, the
bottom floor area is surveyed.
In surveying a double liner impoundment, provisions must be
made to ensure that the material between the geomembrane lin-
ers provides electrical conduction to a return electrode placed in
the leak collection zone. The test is best accomplished by flood-
ing the leak collection zone with fresh water. To provide electrical
contact to the leak collection zone, a stainless steel return elec-
trode with connecting wire is placed in the zone prior to the in-
stallation of the primary liner. The return electrode also can be
temporarily placed in the leak collection drain pipe if access is
available. In both cases, the return electrode must be covered with
water.
Air vents should be provided along the perimeter edges of the
primary liner near the top of the berm to vent air trapped be-
tween the liners. This procedure will help prevent damage to the
liner caused by trapped air floating the liner during flooding of
the leak collection system. Impoundments that use sand as the
material in the drainage layer usually do not require water flood-
ing of the leak collection zone. This is because the sand contains
sufficient residual moisture to allow electrical current flow in the
sand drainage layer. However, a permanent stainless steel elec-
trode placed in the sand drainage layer prior to the placement of
the primary liner will greatly facilitate electrical leak location
surveys.
Electrical conduction paths, other than leaks, such as steel pip-
ing, piers, fasteners and battens must be electrically isolated for
best leak location results. Certain preparations such as rubber
packers in inlet and discharge pipes will prepare most geomem-
brane lined impoundments for a successful leak location survey.
The electrical leak location survey method can be most effective-
ly and economically applied if the impoundment or landfill is de-
signed such that electrical conduction paths between the liquid in
the impoundment and the earth ground are eliminated or can be
electrically insulated.
SURVEYS OF SOIL-COVERED GEOMEMBRANES
A protective soil cover often is placed over the primary geo-
membrane liner of landfills to protect the liner from mechanical
damage when placing the waste material in the landfill. In addi-
tion, a sand drainage layer often is used as the drainage medium
in the leak detector zone of double liner installations. However,
during the placement of the protective soil cover or the sand
drainage layer, the liner can be damaged by the equipment used
to place the soil cover, tools used to spread the material, sharp
rocks in the soil or by a variety of other mechanical mechanisms.
Often the mechanical damage to the liner is undetected and cov-
ered by the placing of the protective soil cover. The electrical leak
location survey technique has been successfully adapted to locate
leaks in geomembranes covered with up to 2 ft of a protective soil
cover or sand drainage layer. Leaks were located and later veri-
fied beneath protective soil cover, sand drainage layers and thin
sediment layers at several sites surveyed.
A protective soil cover or sludge cover over a geomembrane
can decrease the effectiveness of a leak survey in three ways:
(1) The strength of the signal received may be reduced because
of inhomogeneities in the soil cover or sand drainage layer
(2) The ability of the electrodes to detect leak signals is decreased
because of the dissimilarity of the soil and water medium
contacting the electrode, resulting in undesirable transient
signals caused by polarization of the electrodes
(3) The scanning probe cannot be scanned close to the geomem-
brane liner
The first condition is solved by systematically conducting the
survey on an established survey grid and recording the current
signature every 24 in. The acquired data are analyzed in the field
and a plot of anomalies is produced which allows for a resolu-
tion of the leak locations. The dissimilarity or polarization prob-
lem is overcome by using specially designed electrodes to elim-
inate electrode polarization.
TYPES OF FACILITIES AND MATERIALS SURVEYED
FacUlry Types Surveyed
The electrical leak location survey method was used to survey
geomembrane lined facilities ranging in size from 970 to 584,800
ft*. The facilities tested include:
Primary and secondary liners at landfills
Concrete vaults for solid waste storage
Wastewater storage ponds for sewage treatment facilities
Above ground steel tanks for storage of hazardous materials
Brine storage impoundments
Descaling ponds for natural gas transmission companies
Cooling water ponds
Materials Surveyed
Approximately 927* of all materials by area surveyed were high
density polyethylene (HOPE). At installations lined with HOPE,
the predominant material thickness was 60 mil. The remainder of
the HDPE material had a thickness of 80 or 100 mils. The other
liner materials were polyvinyl chloride (PVQ, oil-resistant poly-
vinyl chloride (XR-5) and oil-resistant chlorosulfonated poly-
ethylene (OR-CSPE). Generally, the seams at a given facility had
been inspected using conventional inspection techniques such as
visual inspection, air-lance, spark testing or vacuum box prior to
the electrical leak location survey. After the electrical leak loca-
tion survey was completed, the presence of the leaks detected and
located by the electrical method was verified at several of the
facilities using the vacuum box technique.
DISCUSSION OF LEAKS DETECTED AND LOCATED
Leak Statistics
Sixty-one sites with an approximate total area of 4,368,785 ft2
of liner material have been commercially surveyed. Tables 1, 2
and 3 present a summary of all the commercial leak surveys con-
ducted to date using the electrical method developed at Southwest
Research Institute. A total of 1409 leaks were located at the 61
sites surveyed which equates to an average of 3.2 leaks/10,000 ft2
of liner material inspected.
Figures 3 through 7 are plots of the data as a function of the
area surveyed and the leak location on seams or sheet, total num-
ber of leaks or area ratio of the leaks located. Figure 7 is a plot of
the number of sites surveyed vs. the area ratio of the leaks located
which indicates that there may be between 0.3 and 0.5 leaks/
10,000 ft2 of geomembrane liner.
Leaks on Side Slopes
The side slopes were surveyed at approximately 25% of the
liners surveyed. The majority of leaks on the side slopes occurred
on the seams. At the facilities where the side slopes were tested,
leaks on the side slopes comprised approximately 20% of the total
leaks located.
Leaks In the Bottom of the Liner
Leaks on the bottom of liquid impoundments were found in
the parent material, field seams and factory seams. Eighty-seven
36 MONITORING & SAMPLING
-------
Table 1
Leak Detection and Location Survey Data for Impoundment Where the
Bottom Floor Area was Surveyed.
SURVEY
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
TOTALS
SIZE
SQ. FEET
958
958
958
1,000
1,798
2,625
3,000
3,000
3,200
4,951
4,951
4,951
5,175
7,007
12 , 600
18 , 346
26,016
26,016
27,297
32,292
43,560
45,345
50,000
50,400
54,500
55,025
58,900
62,500
64,583
65,340
65,369
65,369
65,369
65,500
65,500
74,088
82,500
87,120
87,120
99,050
135,036
150,781
152,460
152,460
157,584
164,085
362,690
2,769,336
TOTAL
LEAKS
2
3
3
4
0
6
21
4
0
0
17
2
2
4
7
50
7
4
8
25
2
4
6
193
29
12
8
21
29
56
6
7
5
7
5
20
18
8
17
18
17
64
2
7
12
18
51
811
LEAKS
BOTTOM
2
3
3
4
0
6
21
4
0
0
17
2
2
4
7
50
7
4
8
25
2
4
6
193
29
12
8
21
29
56
6
7
5
7
5
20
18
8
17
18
17
64
2
7
12
18
51
811
LOCATED
SEAM
2
3
3
3
0
6
21
4
0
0
17
2
1
4
7
35
7
4
6
25
2
4
6
188
18
12
6
19
21
55
6
5
3
5
3
19
15
7
17
14
16
46
2
7
10
16
37
709
IN
SHEET
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
15
0
0
2
0
0
0
0
5
11
0
2
2
8
1
0
2
2
2
2
1
3
1
0
4
1
18
0
0
2
2
14
102
LEAKS PER
10,000
SQ. FEET
20.9
31.3
31.3
40.0
0.0
22.9
70.0
13.3
0.0
0.0
34.3
4.0
3.9
5.7
5.6
27.3
2.7
1.5
2.9
7.7
0.5
0.9
1.2
38.3
5.3
2.2
1.4
3.4
4.5
8.6
0.9
1.1
0.8
1.1
0.8
2.7
2.2
0.9
2.0
1.8
1.3
4.2
0.1
0.5
0.8
1.1
1.4
2.9
Table!
Leak Detection Data for Impoundment with the Side Slopes and
Bottom Floor Area Surveyed.
SURVEY SIZE TOTAL
NO. SQ. FEET LEAKS
LEAKS PER
LEAKS LOCATED IN SIDE 10,000
BOTTOM SEAM SHEET SLOPE SQ. FEET
1
2
3
4
5
6
7
8
9
10
11
12
13
14
TOTALS
9,620
12 , 540
24,000
24,272
25,000
25,000
35,291
42,022
50 , 000
51; 000
62,500
130,680
522,720
584,804
1,599,449
16
16
40
47
22
15
42
14
4
20
50
192
41
79
598
12
12
33
31
10
7
31
7
4
13
26
183
31
54
454
14
12
33
46
15
10
33
12
3
19
44
183
31
61
516
2
4
7
1
7
5
9
2
1
1
6
9
10
18
82
4
4
7
16
12
8
11
7
0
7
24
9
10
25
144
16.6
12.8
16.7
19.4
8.8
6.0
11.9
3.3
0.8
3.9
8.0
14.7
0.8
1.4
3.7
Table 3
Survey Data for All Impoundments Inspected.
LEAKS PER
TOTAL LEAKS LOCATED IS SIDE 10,000
SITES TOTAL AREA LEAKS BOTTOM SEAM SHEET SLOPE SQ. FEET
BOTTOM AREA OHLY
BOTTOM AND SIDE AREA
47 2,769.336
14 1,599.449
811
598
811
454
709
516 •
102
82
H/A
144
2.9
3.7
in
X
10
0 5 10 15
Impoundment Size (Sq. Feet x 1000)
Figure 3
Histogram of Total Leaks Located vs. Bottom Floor Area Surveyed
O1
w
5 10 15
Impoundment Size (Sq. Feet x 1000)
Figure 4
Histogram of Leaks per 10,000 ft2 of Liner Surveyed
0)
.C
c
-H
tn
X
TOTAL
61 4,368,785 1,409 1,265 1,225
iiTimiiiiiiiiiiiiiiiiiiii
05 10 15 20
Impoundment Size (Sq. Feet x 1000)
Figure 5
Histogram of Leaks in the Parent Material vs. Impoundment Size
MONITORING & SAMPLING 37
-------
I 5 10 IJ
IapoundB«nt Six* (Sq. F««t x 1000)
Figure 6
Histogram of Leaks in Seam vs. Impoundment Size
a
w
Leaks par 10,000 Sq. Feet
Figure?
Histogram of Number of Sites Surveyed vs. Number of Leaks
Located per 10,000 ft* of Oeomembrane Liner
percent of the leaks were in seams, and the remaining 13% were
in the parent material. Figures 8 and 9 show examples of seam
leaks detected with the Southwest Research Institute electrical
leak location system. Leak sizes and shapes ranged from rela-
tively circular holes from less than 0.025 to 1 in. in diameter, to
slits from 0.25 to 12 in. long, to gashes and gouges up to 6 by 8
in., to evidently tortuous paths through seam welds.
Leaki la Parent Material
The leaks in the parent material generally can be attributed to
accidental damage from equipment or tools, crescent-shaped
cracks due to equipment being dropped, slits due to razor-edged
tools cutting the liner, burns from cigarettes, gashes and gouges.
Figures 10 and 1 1 show typical leaks in the parent material. Some
of the leaks in the parent material probably were caused by im-
proper material handling or wind buffeting. Many leaks in the
parent material of installations with a protective soil cover
appeared to have been created during the application of soil cover
over the liner.
The observed ratio of parent material leaks to seam leaks may
be slightly less than actual because the seams are double-checked
during the leak location survey process. While rechecking the
seams, the search probe tip is scanned within 1 in. from the leaks
in the seams. However, during the general survey of the geomem-
brane, the parent material is swept at 12 in, intervals placing the
Figure!
Leak in HOPE Seam. Approximate Leak SbeiUBS la.
(Note: Leak not apparent in reproduced photograph.)
Figure 9
Leak in Parent Material
Figure 10
Large Leak in HOPE Parent Material
electrical probe as much as 6 in. from a potential leak,
cause the probe tip is approximately six times closer to,_
leaks when surveying the liner seams, it is probable that
small leaks found in the seams are not detected in the parent
terial.
38 MONITORING & SAMPLING
-------
Figure 11
Cut in HDPE Parent Material
Leaks In Seams
Inadequate field seaming appears to be the primary cause of
leaks in geomembrane lined impoundments. Eighty-seven percent
of the total number of leaks were in field welded or bonded
seams. Many of the leaks occurred at T-joints, patches and at
seams in highly-stressed areas such as at the base of the sideslope.
Some leaks were found in seams which previously had been re-
paired and tested. Figures 12,13 and 14 show typical leaks located
in seams. Leaks may not develop in the seams until a hydrostatic
load is placed upon the liner. Cases were documented where ob-
viously poor seaming techniques resulted in seams failing indis-
criminately after repair and hydrostatic loading. In such cases, it
is suggested that the entire liner installation be redone.
Figure 12
Leak in Seam
Leaks Associated with Penetrations and Structures
In some facilities, numerous leaks were found around penetra-
tions or structures in an otherwise excellent field installation.
Many designs incorporate complex seam requirements when
attempting to isolate drainage cribs, separation walls, concrete
sumps, concrete pads and other structures. Where such structures
are necessary, the electrical method may be the only method
which can be applied to test for leaks.
Leaks Associated with Material Types
Because of the limited use of materials other than HDPE in the
Figure 13
Leak in Seam After Grinding, Just Prior to Repair
Figure 14
Leak in Seam Where Seaming Material Did Not Bond to Sheet
facilities tested by the Southwest Research Institute electrical
method, it is not possible to formulate any valid conclusions on
the relationship of material type to numbers and types of leaks.
Leaks Beneath Soil Covers and Sludge
The Institute has successfully located leaks beneath installed
soil cover up to 2 ft thick. Leaks have been found beneath chem-
ical precipitate sludges, but the application of the electrical
method in the sludge environment is extremely tedious and de-
manding. The leaks found beneath soil covers have included seam
leaks and leaks in the parent material apparently caused by the
heavy equipment which was placing the protective soil cover ma-
terial. Figures 15,16 and 17 show leaks located under 2 ft of sand
place over the primary geomembrane liner. No significant numer-
ical relationships between leaks, leak occurrence and types of
leaks can be developed on leaks discovered beneath soil covers
because of the limited field testing experience in such environ-
ments.
CONCLUSIONS
The electrical leak location method is a very sensitive, accu-
rate and valid method for locating leaks in geomembrane liners.
Leaks were found in every liner surveyed except for three liners
that were less than 500 ft2 in area. Leaks were located in liners
that had been rigorously tested using one or more of the conven-
tional methods for testing geomembrane liners.
MONITORING & SAMPLING 39
-------
Figure 15
Leak Under 2 ft of Protective Sand Cover
Figure 17
Tear in Liner Covered with 2 ft of Sand
Figure 16
Mechanical Damage to Liner Under 2 ft of Sand Cover
The number of leaks per 10,000 ft* of surveyed area typically
ranged from 0.3 to 5 with an average density of 3.2 leaks/10,000
ft2 of geomembrane liner. Several liners had greater than 20
leaks/10,000 ft* of area surveyed.
The density of leaks generally decreases as the liner size in-
creases. Possible explanations for this are:
• Smaller installations have proportionally more complex
features such as corners, sumps and penetrations
• Small installations tend to have higher proportions of hand
seaming
• Larger installations tend to have better QA/QC programs
• Larger installations generally receive proportionally less traffic
From our experience, and knowledge of the history of some of
the liners surveyed, the major factors for minimizing the number
of leaks in geomembrane liners in the general order of importance
are: the professionalism and skill of the seaming machine oper-
ator; environmental factors such as moisture, temperature and
wind; simplicity of the liner design; thickness and weldability of
the liner material; and liner care and handling procedures.
The electrical leak location method has demonstrated that
geomembrane installations can benefit from an electrical method
leak location survey as a pan of the construction quality assur-
ance program. Pre-service testing of new installations using the
electrical leak location method will enhance the overall perfor-
mance of the containment facility.
40 MONITORING & SAMPLING
-------
Field Analytical Screening of Soil for
Preremedial Hazardous Waste Site Investigations
by Thermal Chromatography/Mass Spectrometry
Pamela D. Greenlaw
Raymond J. Bath, Ph.D.
David J. Grupp
Richard P. Hubner
Roberta Riccio
NUS Corporation
Edison, New Jersey
Richard D. Spear, Ph.D.
U.S. EPA Region 2
Edison, New Jersey
Raymond Worden
Chris Sutton, Ph.D.
John Zumberge, Ph.D.
Ruska Laboratories
Houston, Texas
Robert J. White
George Collins, Ph.D.
Eric Johnson
Finnigan MAT
Livingston, New Jersey
ABSTRACT
The Field Analytical Screening Project (FASP) for the U.S. EPA
preremedial program requires rapid and chemically specific analyses
of samples for hazardous substances. The preremedial U.S. EPA Region
2 FASP program is also an interactive program that requires the field
project manager and the FASP analytical manager to make field deci-
sions on the data generated in the screening process.
For soil organic analyses, the U.S. EPA's Contract Laboratory Pro-
gram (CLP) requires extensive wet chemical extraction and cleanup
before mass spectral analysis. These time-consuming methods can only
be done effectively in conventional fixed-base laboratories. Therefore,
samples from a preremedial site investigation are transported to CLP
labs for extraction and analysis. This procedure can cause delays of
weeks or even months between sample collection and return of the
results. This delay hinders efficient site evaluation efforts and can result
in repetition of work. The development of analyte-specific alternative
methods for use by the FASP program can complement the CLP program
while decreasing the sample turn-around time.
In an effort to obtain fast organic results to guide screening and cleanup
work, in-field portable gas chromatographs (GCs) have been utilized.
Unfortunately, the low specificity of these instruments and the broad
gap between in-field protocols and CLP methods can lead to poor coore-
lation with CLP results. Laboratory tests done in the last few years
indicate that a new technique known as thermal extraction/gas chro-
matography (TC) can give results comparable to conventional wet chemi-
cal extraction of soils. TC is fast and since no sample preparation is
necessary, it can speed up considerably the time from sample receipt
to analytical data.
Coupling the TC to a mass spectrometer (MS) leads to a new era
for organic analysis. Analytical equipment with excellent data systems
and small, rugged thermal extractors and mass spectrometers have been
improved and downsized to the extent that they are easily transporta-
ble. In light of these developments, transportable equipment of this
nature has been added to the FASP organic protocols in U.S. EPA
Region 2. This paper reports the results of a site investigation using
a transportable TC/MS system for the FASP organics investigation.
INTRODUCTION
The preremedial program of the U.S. EPA involves the investigation
of suspected hazardous waste sites for inclusion on the NPL. The
investigation includes a preliminary assessment (PA), a screening site
inspection (SSI) and a listing site inspection (LSI). These investiga-
tions assess the relative threat associated with actual and/or potential
releases of hazardous substances from the site's soil, surface water,
groundwater or air. At the end of the investigation phase, the site is
ranked by using the Hazard Ranking System (HRS) model1", which
evaluates and assigns a numerical score to each potential pathway of
exposure. This numerical score depends to a great extent on the evalua-
tion of the analytical data from the investigations of the contamination
of the existing air, soil, groundwater, and surface water from the site.
In U.S. EPA Region 2, NUS Corporation, the Field Investigation Team
(FIT) contractor for the U.S. EPA, introduced an interactiyeJEield Ana-
lytical Screening Project (FASP) program for the LSI preremedial stage
of the investigation. The FASP program provides the field project
manager with on-site unambiguous analytical data of high quality in
a timely manner and complements the existing fixed-base Contract
Laboratory Program (CLP) by prioritizing and screening the samples
sent for analysis.
Since the biggest advantage of on-site analysis is the ablity to pro-
vide the project manager with immediate results, the U.S. EPA Region
2 FASP program performs analysis for target chemicals only, rather
than general unknowns. It also utilizes methods and instrumentation
that require minimal sample preparation and provides unambiguous
high-quality data. Small portable nonspecific instrumentation is not used
MONITORING & SAMPLING 41
-------
in the U.S. EPA Region 2 FASP program because the lack of analyte
specificity coupled with the lack of continuing quality assurance can-
not provide data of sufficient quality for scoring purposes. The FASP
program instituted by U.S. EPA Region 2 FIT utilizes open-path
FTIR/UV remote sensing techniques for air investigations2 )0,
secondary target X-Ray Fluoresence6 (XRF) for soil inorganic analy-
sis investigations, and thermal extraction/gas chromatography/mass spec-
trometry (TC/MS)78 for soil semi-volatile organic investigations.
The TC/MS was chosen for the U.S. EPA Region 2 FASP program
because it requires no sample preparation and produces unambiguous
data in a limited amount of time. Thermal extraction is a relatively new
technique which is very simple in principle. Basically, a sample is placed
in a sealed chamber where it is heated, with the resulting gases being
passed through to a detector (in this case a mass spectrometer) for iden-
tification and quantification. Recent instrument advances have seen the
development of thermal extraction systems with flow-through extrac-
tion cells and fused quartz systems. Fused quartz systems allow for quick
heating and cooling of the instrument with no loss of instrument in-
tegrity. Mass spectrometers have also undergone a revolution in pumping
capacities, total glass systems and software simplification that has moved
MS out of the specialty laboratory to the routine analytical services.
These advances made TC/MS even more applicable to the U.S. EPA
Region 2 FASP preremedial program.
This paper discusses the utilization of a thermal chromatograph/mass
spectrometer (TC/MS) for the interactive FASP soil semi-volatile
organics program at a site in U.S. EPA Region 2. CLP analyses of sam-
ples from previous sites allowed for the selection of four target semi-
volatile organic compounds. Based upon these determinations, a mass
spectral library with corresponding gas chromatographic retention times
was established prior to site arrival. The equipment is set up in a trans-
portable mode; i.e., the equipment is not mounted permanently in a
vehicle but is cart-mounted and is moved into a vehicle prior to travel.
The vehicles used to transport the analytical equipment are standard
vans equipped with generators and air conditioning. Processing of
samples can be started within 1 hr of site arrival.
SYSTEM DESCRIPTION
The U.S. EPA's Region 2 FASP program's transportable TC/MS con-
sists of a Ruska thermal extractor and gas chromatograph coupled to
a Finnigan INCOS 50 mass spectrometer. The TC/MS system (Fig.
1) is permanently mounted on two specifically designed and constructed
carts, which enable the system to be easily loaded onto the vehicle.
Each cart consists of a shock-mounted table on an aluminum frame
with heavy-duty wheels for ease of maneuverability. One cart carries
the TC/MS and the other cart is used for the computer systems neces-
sary for control of the instruments. The carts can be loaded onto or
off the vehicle in less than 1 hr to allow the flexibility of use either
in the field or in a fixed-base application.
RUIKA THIRUAL
DATA ITITfMt
The vehicle used to transport and house the TC/MS system for field
analysis is a Chevrolet UtilMastcr stepvan. Vehicle modifications
include: a lift gate on the rear for ease of loading and unloading; two
6,500-W undercarriage generators to provide electrical power; two air
conditioners and heaters to provide a stable environment; a ceiling vent;
cabinetry and shelving for storage of necessary equipment; and a bench
top for work space. A portable hood can also be used in the vehicle
for samples, standards and solvents. The vehicle has two separate elec-
trical systems: one generator provides both 220V and 110V power for
the instruments; the other generator provides 110V power for the air
conditioners, heaters, lights and additional outlets. The quality of the
power supplied to the instruments is ensured by the use power condi-
tioner transformers that eliminate voltage fluctuations, sags, surges and
transients. Figure 2 presents an illustration of the TC/MS system in-
stalled in the vehicle for Held analysis.
Figure 1
Region 2 TC/MS Organic Soil Analyzer
Figure 2
Transportable TC/MS
The thermal chromatograph consists of a fused quartz thermal
extractor coupled with a capillary column gas chromatograph. The ther-
mal extractor uses temperature programmed heating, cooling and
isothermal methods to thermally extract a sample. The thermally ex-
tracted organic compounds are then further separated in the integrated
capillary gas chromatograph system prior to mass spectrometric identi-
fication. The 1C system (Fig. 3) consists of four controlled thermal
zones in a vertical stack: the pyrocell, the trap, the splitter and the
column. Three of these zones, the pyrocell, trap or column, can be
controlled via linear temperature programming (LTP), held isothermally
or cooled with liquid CO,. The fourth zone, the splitter, can only be
controlled isothermally. The pyrocell is the portion of the thermal chro-
matograph where thermal extraction takes place and can be programmed
for temperatures from 0° to 625 °C.
The trap can be operated from -70° to 625 °C, but for our applica-
tions it is used not as a trap but rather as a hot, pass-through zone.
The splitter is also maintained at a high temperature, the maximum
being 350 °C. to ensure column flow. The temperature of the column
thermal zone can range from -60° to 400 °C, depending on the upper
limit of the stationary phase of the column being used.
The sample is placed in a porous fused quartz cup which is inserted
into the pyrocell where it is heated while helium flows through the
pyrocell. In the splitter zone, a portion of the sample is passed onto
the column while the remainder is vented into a carbon filter and
released outside the instrument. The column is initially maintained at
a cryogenic temperature to trap the sample on the head of the column.
The column is then heated via LTP for further separation prior to iden-
tification in the mass spectmmeter
The column is inserted into the Finnigan INCOS 50 mass spectro-
meter (Fig. 4) through a heated transfer line into an evacuated analyser
assembly. The Finnigan uses a quadnipole positive ion mass analyser
with a corresponding vacuum system consisting of a high-speed 170-Us
turbomolecular pump, rotary vacuum forepump and a glass vacuum
manifold. The extra capacity pumps are required to enable the system
to pump down and be ready for analyses in less than an hour.
42 MONITORING & SAMPLING
-------
Scale
10cm
COLUMN EXIT —
LEFT OR WCHT
SIDE
PYROCELL
H. SPUTTER PURGE
SPUTTER' EXHAUST
Uq C02 COOLANT
Liq C02 COOLANT
H* CARRIER INLŁT
TC ANALYZER
Figure 3
Ruska Thermal Chromatograph
EXPERIMENTAL
This section of our paper discusses the project design, project- and
instrument-specific quality assurance/quality control (QA/QC), sample
collection and preparation, and instrument operating conditions for the
TC/MS semi-volatile organic FASP analysis.
Project Design
The FASP semi-volatile organic analysis concentrates on target com-
pound analyses. Following the SSI, the CLP data were reviewed by
the FASP manager, project manager and U.S. EPA project manager to
select target chemicals for field analysis. These chemicals, usually four
to six in number, are selected based on their toxicity, abundance and
instrument detection limitations. Chemicals on the Target Compound
List (TCL) are selected first because more information about their analy-
sis is known and confirmation of positive results is easily obtained.
As a general rule, tentatively identified compounds (TICs) are not
selected as target chemicals unless a high priority is placed on their
toxicity. For the site in this study, the target chemicals selected include
diphenylamine, mercaptobenzothiazole, benzothiazole and aniline. The
final selection of these chemicals was based on their toxicity and abun-
dance, even though three of the four target chemicals are TICs.
The initial SSI found high quantities of the target analytes in subsur-
face soil and waste samples. Surface soil samples from the site were
then collected and analyzed by CLP laboratories to determine the ex-
tent and degree of contamination. These samples and others, assumed
to be of high concentration, were also analysed with the U.S. EPA Region
2 FASP TC/MS. A majority of the samples analysed contained no
detectable levels of the target chemicals, although the non-CLP samples
did produce some positive hits. None of the CLP samples were found
to be positive for target chemicals by CLP or FASP TC/MS analyses.
The TC/MS was then transported to the site, which is located within
U.S. EPA Region 2, to help determine whether there were measurable
quantities of the target chemicals in the dust from the homes at the
site. Due to the emergency nature of the program at this site, a deci-
sion was made to analyze samples around the clock, thereby enabling
the analytical results obtained from the TC/MS to match the collection
team's sampling efforts.
Standards and Reagents
The target compouds, diphenylamine, mercaptobenzothiazole, ben-
zothiazole and aniline, were purchased as pure reagents from Aldrich
Chemical Co., and 200 ug/mL stock solutions were prepared with
HPLC-grade methylene chloride. A mix of the stock solutions was then
prepared, and a standard quanitation curve was developed to determine
detection limits for these target analytes. The detection limits were found
to be as follows: aniline-1.0 mg/L, diphenylamine-0.1 mg/L,
benzothiazole-0.1 mg/L, and mercaptobenzothiazole-2.0 mg/L.
An internal standard, base neutral (B/N) mix was prepared from the
Supelpreme standard consisting of l,4-dichlorobenzene-d4,
acenaphthene-d|0, chrysene-d]2, naphthalene-dg, perylene-dp and
phenanthrene-d]0.
INSTRUMENT AND PROJECT QA/QC
The QA/QC applied to this project was derived from the QA/QC
requirements for CLP analysis of semi-volatiles. The mass spectrome-
ter was tuned manually using FC43, adjusting the parameters for proper
peak shape and ion ratios. DFTPP was then analyzed and the CLP
abundance criteria achieved. The internal standard mix was added to
every sample, and it was found that the area counts and retention time
variability were within that required by the CLP Statement of Work
(SOW). Duplicate analyses were performed every 10 samples to ensure
result integrity, and blanks were analysed every 12 hrs to confirm system
cleanliness. Calibration response and minimum detection limits were
established for the target analytes and retention times determined to
provide a clear indication of compound presence. The instruments were
cleaned and reconditioned as deemed necessary by the performance
of the QA/QC samples (duplicates and blanks).
A comprehensive FASP quality assurance program has been institued
for U.S. EPA Region 2 FIT that ensures the integrity and validity of
all aspects of the TC/MS and the generated data. The minimum
requirements include a full standard operating procedure (SOP), main-
tenance plan, written documentation for all activities and an initial and
ongoing monitoring program that demonstrates the consistency of the
generated data. A minimum of 20% of all samples are normally spent
for CLP confirmatory analysis. Due to the low sample volume at this
site, a decision was made not to utilize CLP verification QA/QC.
Data reduction, validation and reporting procedures are performed
by trained personnel after a full review by the FASP manager.
SAMPLE COLLECTION AND PREPARATION
Soil samples were collected from the site using sampling techniques
as required by work plan and QA procedures. The sampling technique
chosen for the house dust was to collect dust samples obtained by
sweeping the kitchen areas of the homes. No further homogenization
was performed on the samples prior to analysis. For the house dust
samples, the analysis was performed on the actual dust portions of the
sweep samples. A portion of each sample was placed in the porous fused
quartz sample cup and weighed on an analytical balance; sample quan-
tities ranged from 20 to 140 mg, depending on sample type (soil or
dust) and density-the cup was filled with a loosely packed sample. Five
MONITORING & SAMPLING 43
-------
Figure 4
Finnigan MAT INCOS 50 Mass Spectrometer
uL of the base/neutral internal standard mix at a concentration of
200 ug/mL were added to the top of each sample before capping with
a porous fused quartz cap.
INSTRUMENT OPERATIONAL CONDITIONS
The capped sample, with internal standard added, was placed in the
pyrocell of the TC for heating and analysis. TC/MS operating condi-
tions were as follows:
Throughout the analysis the trap was maintained at 360°C while the
splitter was held to 310 °C. The pyrocell was heated from 30 to 260 °C
at a rate of 34°/min while the column was held at 5°C. At the end of
the pyrocell LTP cycle, the column was heated from 5 to 285 °C at
15°/min. Helium flow through the pyrocell was 30cc/min, but because
of a 30:1 split performed on the extracted sample, helium flow through
the column into the mass spectometer was Icc/min. The capillary column
(HP-5 12M x 0.2 MM ID with 0.33-m film thickness) was run through
a transfer line at 280°C to the mass spectrometer with an ion source
temperature of 180 °C. The total analysis time, including sample heating.
was 37 mins. Figure 5 illustrates the temperature plots for a typical run.
Results and Discussion
This site presented a considerable analytical challenge in the selec-
tion of target compounds due to the limitations and requirements of
the samples used. In the initial CLP analyses, the TCL compound de-
tected was n-nitrosodiphenylamine with a large number of tentively iden-
tified compounds (TICs). TC/MS analysis of spike soil samples showed
that only diphenylamine could be detected within the protocols deve-
loped. Current CLP SOW indicates, however, that the n-
nitrosodiphenylamine cannot be distinguished from diphenylamine.
Diphenylamine was selected as one of the target analytes for this FASP
program based on the assumption that negative results were expected
and that if the sample did not contain diphenylamine, then n-
44 MONITORING & SAMPLING
-------
PYROCELL
TRAP
COLUMN
INITIALIZATION
OF MS DATA
ACQUISITION
189.0-1
RIC
!• 2* 22 24 24 2* 3* 32 34 3*
IIM
Figure 5
Thermal Chromatograph Temperature
Plot Programmed vs Actual
IS
IS
IS - INTERNAL STANDARD
IS
see
5:00
leee
10:00
1500
15:00
2000
20:00
2500 SCAN
25:ee TIME
Figure 6
RIC of Soil With Internal Standards
MONITORING & SAMPLING 45
-------
ioe.o-1
RIC
SCAN
TII€
IS - INTERNAL STANDARD
Figure 7
RIC of Household Dust
nitrosodiphenylamine would not be present. Two target analytes, benzo-
thiazole and mercaptobenzothiazole, both sulfur-containing TICs, were
selected based on their prevalence in certain site areas and their toxici-
ty. The fourth target analyte, aniline, was selected based on its toxicity.
Figure 6 shows the reconstructed ion chromatogram (RIC) of soil
with base/neutral internal standard mix. Figure 7 shows the RIC of a
household dust sample. The dust samples exhibited a large number of
peaks, most of which were fatty acids, hydrocarbons and other normal
household contaminants as determined by library spectral identifica-
tion. The large quantities of organic material in these dust samples neces-
sitated a change in the experimental design. Whereas in the original
design the plan was to analyze ten samples, a random duplicate and
a blank, blank analysis was required after three samples just to con-
firm that the system was clean.
Figure 8 illustrates positive household dust analysis. Figure 9 presents
the RIC chromatogram from 800 to 900 scan numbers showing the
region where benzothiazole is found; the upper portion of the chro-
matogram indicates which peaks have mass 135 (benzothiazole) as a
base peak. Figure 10 shows the mass spectrum of the peak at scan 841,
and Figures 11 and 12 show the library matches confirming the presence
of benzothiazole.
Duplicate analyses were performed on all positive samples to con-
firm the presence of target analytes. The results of all the original positive
samples wer confirmed, demonstrating the reliability of the original
procedure. The only target analyte found in the household dust samples
was benzothiazole.
CONCLUSION
The US EPA Region 2 preremedial FASP program has been sig-
nificantly enchanced with the additional of the TC/MS system for target
organics analyses in soil. This unique analytical system provided the
field project manager with unambiguous data and rapid turn-around.
This instrumentation was utilized in the transportable mode that needed
only generator power and a constant temperature environment; the sys-
tem was fully operational within 1 hr of site arrival and ran continuously
for 4 days. The quartz inlet system of the Ruska thermal extractor-gas
chromatograph ensured constant temperature control with fast cool-down
capabilities. The Finnigan INCOS 50 mass spectrometer equipped with
a high-speed pump and all-glass vacuum manifold ensured rapid start-
up and very stable operation in the vehicle during the field operations.
By producing mass spectral confirmed data, this interactive FASP
program allowed rapid decisions to be made in the field. In our
experience, the only minor limitations are residue contamination
problems with the thermal extractor trap, especially after analyses of
high concentrations of organic materials; operator fatigue, especially
on rotating field shifts; and the necessity for a full QA/QC data reduc-
tion system.
DISCLAIMER:
Trade names and company names are used for identification only
and do not imply endorsement by NUS Corporation or the U.S. EPA.
46 MONITORING & SAMPLING
-------
180.6
RIC _
see
5:09
* - BENZOTHIAZOLE
188.8-1
962.8-1
JL
^VLxi^ r
1090
10: 00
1500
15:00
2008
20: ee
2500 SCAN
25:00 TIME
Figure 8
RIC of Household Dust Containing Benzothiazole
135 MAJOR ION
8:48
388 SCAN
9:88 TIME
Figure 9
RIC and 135 Major Ion Mass Chromatogram
MONITORING & SAMPLING 47
-------
BENZOTHIAZOLE
MW 135
Figure 10
Mass Spectrum of Benzothiazole in Household Dust
REFERENCES
1. US EPA Hazard Ranking System (MRS) for Uncontrolled Hazardous Sub-
stance Releases; 40 CFR Part 300, Appendix A of the National Oil and
Hazardous Substances Contingency Plan, U.S. EPA, Washington, DC.
2. Balh. R.J., Minnich, T.R., Naman, R.M., Spear, R.D., SimpsonO., Faust,
J., Stedman, D.H., McLaren S.E., Herget, W.F., and \fcughn W.M., "Remote
Sensing of Air Toxics Using Slate-of-the-Art Techniques," Proc. of the
EPA/AW MA International Symposium on Measurement of Toxic and Related
Air Pollutants. Raleigh. N.C., May 1989.
3 Grupp, D., Rojek, G , Balh, R.J., Minnich, T.R., Naman, R.M., Breda,
A J and Spear, R. D., "The Pre-Remedial Air TOXJCS Program: A Case Study
using Remote Sensing." Proc. of the EPA/AWMA International Symposium
on Measurement of Toxic and Related Air Pollutants, Raleigh, NŁ., May 1989.
4. Minnich, T.R . Bath. R J . Spear, R.D., Simpson, O.A., Faust, J., Herga,
WF, Stedman. D.H.. McLaren. S.E. and Vaughn. W.M., "RemoteSensing
of Air toxics for Pre-Remedial Hazardous Waste Investigations," Proc. of
the 82nd Annual AWMA Meeting and Exposition, Anaheim, CA, June 1989.
5 Scotio, R L., Bulich, J , Greenlaw. P , Vaughn. W.M. and Ennis, R., "Re-
mote Sensing Data Quality Objectives and Quality Assurance for a Pre-
Remedial Hazardous Waste Site Program," Proc. of the 82nd Annual AWHA
Meeting and Exposition. Anaheim, CA, June 1989.
6. Grupp, D.J.. Eventi. D.A.. Bath. RJ. and Spear, R.D., "The use of a Trans-
portable X-ray Fluorcscense Spectrometer for On-Site Analysis of Mercury
1828
SAMPLE
C7.H5.N.S
1628]
IKIS
RANK 1
1 4939
FIT 984
C8.H3.02.
1628
M HT 179
B PK 133
RAW 2
1 12966
FIT 888
C7.HS.N.S
1828 i
81$ ii
RANK 3
1 4943
FIT 881
M/Z
r
lii. i
ll .
1 1 1
BENZOTHIAZOLE
r
H.
1 ill
,1 .
5 l,2-BENZISOTHIAZOLE-3-CARBOXYLIC ACID
r
1 ill
i 1 i
1,2-BENZISOTHIAZOLE
r
| (...) ,, . 1 . ,.. 1
58 68
il i
|l
||
h
78 80 98 186 118 128 138
Figure II
Mass Spectral Library Search of SCAN 841
48 MONITORING & SAMPLING
-------
1016
SAMPLE
C7.H5.N.S
1916 i
IKffi
RANK 1
1 4939
FIT 994
r
1 1 1 1 1 1 I! ill , i 1 i
1
BENZOTHIAZOLE
r
I i 1 1 1 -1 • i
i,
SAMPLE MINUS LIBRARY
1816
8
-1816
M/2
r
, , i i ,i i, il ui ,i .1 i ,,, ...,
58 68 70 88 98 188 118 128 138
Figure 12
Library Match of Benzothiazole in Sample
in Soils," Am. Environ. Lab. VI, Nov. 1989.
I Overton, E.B., Henry, C.B., Shane, B.S., Junk, T., Irvin, T.R., Nocerino,
J.M., Butler, L.C., Petty, J.D. and Pritchett, T.R., "Application of Thermal
Extraction GC/MS Technologies for Rapid Chemical Analysis of Contami-
nated Environmental Samples." Fifth Annual Waste Testing and Quality As-
surance Symposium Proceedings, Washington, D.C., July 24-28, 1989.
8. Henry, C.B., Overton E.B. and Sutton C, "Applications of the PYRAN Ther-
mal Extraction GC/MS for the Rapid Characterization and Monitoring of
Hazardous Waste Sites." Proc. of the First International Symposium for
Hazardous Waste Site Investigations, Las Vegas, NV, Oct., 1988.
MONITORING & SAMPLING 49
-------
Performance of GC/MS Analysis Quality Indicators
Bruce Peterson
CH2M HILL
Bellevue, Washington
ABSTRACT
The United States currently spends over a billion dollars a year to
find, characterize and remediate sites contaminated with hazardous
chemicals. The primary method of locating con,!laminated areas is the
collection and chemical analysis of environ,!mental samples. The results
of these analyses are used to make decisions about site remediation
activities that can cost millions of dollars.
The quality of the data used in making these decisions is crucial in
the decision-making process. One measure of the quality of the ana-
lytical data is the precision and ac.curacy of the analytical methods.
Precision is estimated by analyzing replicate samples, while accuracy
is estimated by analyzing matrix spikes.
However, the cost of analyzing environmental samples is high, and
other indirect measures of analytical quality are often used. These in-
direct measures include surrogate spikes and matrix spike/matrix spike
duplicates. The indirect measurements attempt to estimate the accuracy
and precision of analytical methods for target compounds with measures
of accuracy and precision on surrogate compounds.
The Love Canal Habitability Study provided an opportunity to
examine the performance of these surrogate measures of analytical
performance. During the course of this study, two batches of soil were
analyzed repeatedly for the Love Canal indicator chemicals. For these
samples, both direct and indirect measures of analytical performance
are available.
Both direct and indirect measures of data quality are presented here.
The value of current data quality measures of laboratory performance
is discussed, and the future course of data quality measures is explored
in the context of electronic data transfer.
Comparing the two measures of analytical performance allows the
hazardous waste community to evaluate the efficacy of surrogate
measures of data quality. A better understanding of the limitations of
surrogate performance measures allows remediation decisions to be more
defensible.
INTRODUCTION
With the passage of the SARA, the United States embarked on a multi-
billion dollar effort to clean up toxic chemical contamination of the
environment. This effort involves both the federal Superfund program
and industrial programs outside of the Superfund process.
Efforts to clean up the environment begin with the collection of
samples and the chemical analysis of the samples to identify and quan-
tify levels of contamination. The anlysis of environmental samples has
become a $300 million-a-year business influencing decisions as to
appropriate site remediation. Clearly, the quality of the data used to
make these decisions is of paramount importance.
The purpose of this discussion is to introduce the regulatory com-
munity to one of the most common analytical techniques, used for
analyzing environmental samples, gas chromatography-mass spectro-
scopy (GC/MS), and the methods used for estimating the precision and
accuracy of this method.
The precision and accuracy of concentration estimates are of par-
ticular concern in a decision framework and are not as straightforward
as might be thought. Four methods are commonly available for esti-
mating the precision and accuracy of GC/MS analyses. Two of these,
blind quality control samples and sample replicates, are external esti-
mators created from samples prepared outside of the laboratory. The
other two, the addition of surrogate compounds to a sample (surrogate
spikes) and the addition of target compounds to selected samples (matrix
spikes), are internal estimators from samples prepared within a
laboratory.
Ideally, a number of external and internal estimates of precision and
accuracy should be available for each study site to allow a comparison
of each estimator's performance and a contrast of what the samples
represent. However, because of the high cost of chemical analyses, the
number of samples analyzed from a site is minimized. This limited
number of analyses places greater emphasis on the use of the internal
estimators of precision than on the external estimators; little informa-
tion is available on how the two estimators compare.
The Love Canal Habitability Study, completed in 1988, provided a
unique opportunity to compare the usual GC/MS internal quality
assurance measurements with the results of replicate analyses of soil
samples. Exceptional quality control measures were employed in the
Love Canal study, ranging from providing partici,!pating laboratories
with identical glassware from the same manufacturing lot to developing
analysis protocols with un,usually strict operating constraints. Many
aspects of the chemical analyses were tracked and stored in a data base
for later analysis. This data base of analytical results provides a good
basis for comparing the internal and external estimators of precision
and accuracy.
STUDY BACKGROUND
In the Love Canal study, the concentration levels of eight indicator
chemicals found in neighborhoods near the canal were compared with
concentration levels of these indicator chemicals found in control areas
in Niagara Falls and Buffalo. Although the main purpose of the study
was a statistical comparison of the Love Canal neighborhoods with con-
trol areas, a number of other investigations were undertaken because
of the unique aspects of the study.
One investigation conducted as part of the Love Canal study was to
collect two samples of soil from two neighborhoods near the canal.
Each sample was homogenized and aliquots of each were sent to each
of two laboratories. These aliquots were analyzed in duplicate at 5-day
intervals for 65 days over the duration of the sample analysis. This scale
of replicate analysis, unusual in environmental studies, allowed a com-
50 MONITORING St SAMPLING
-------
parison of the four measures of precision and accuracy.
The two external measures of precision and accuracy are obtained
from the analysis of blind quality control samples or sample replicates.
Sample replicates were prepared at Love Canal by extruding soil from
the sampling tool (a hollow tube pushed into the soil), quickly mixing
it by hand and then placing aliquots of the soil into two or more jars
for shipment to the laboratories.
Blind quality control (BQC) samples (samples of soil similar to that
found near the Love Canal) were spiked with known amounts of indi-
cator chemical by a U.S. EPA laboratory. These spiked soil samples
were placed in sample jars for shipment to the laboratories. The iden-
tity of the BQC samples was known to the laboratories, although the
spiking concentrations were not. Each laboratory was responsible for
analyzing BQC samples at a set frequency during the course of the study.
The two internal measures of precision and accuracy consisted of
the analysis of matrix spike samples and the use of surrogate compounds.
Matrix spike samples were created by the laboratory by splitting a sample
into three aliquots. Known quantities of indicator chemicals were added
to two of these aliquots. The third aliquot was analyzed for background
concentrations of the indicator compounds. The spiking concentrations
of indicator chemicals were standards known to the laboratories. All
three samples were analyzed, with the two spiked aliquots becoming
known as the matrix spike and matrix spike duplicate.
A surrogate compound is a chemical that is similar to the target com-
pound yet is not normally found in environmental samples. For the Love
Canal indicator chemicals that were chlorinated compounds, the
surrogates were similar compounds that contained bromine rather than
chlorine. These surrogate compounds were added to all samples before
the start of the extraction process.
The GC/MS analysis of a soil sample for the Love Canal study con-
sisted of several steps. These steps and the quality control measures
associated with them are shown in Figure 1. Analysis consisted of
weighing an aliquot of the sample and adding the surrogate compounds.
The soil was then mixed with other chemicals that removed compounds
not of interest to the study, primarily hydrocarbons. The hydrocarbons
were removed to reduce the interference that these compounds create
in identifying the target indicator compounds. Solvent was added to
the mixture to extract the indicator chemicals and surrogate chemicals
from the soil.
The extract obtained was then stored until GC/MS analysis. Before
analysis, compounds used as standards for quantification were added
to the extract and a small portion of the extract was removed for further
concentration. The concentrated aliquot was injected into the GC/MS.
Data obtained from the GC/MS for Love Canal consisted of chroma-
tograms, which are the time traces of ion detection intensity, for three
ions of each of the target compounds. These chromatograms were used
to identify and quantify the target compounds.
The Love Canal samples typically had concentrations below 10 mg/L.
At this extremely low concentration it was often difficult for compound
identification to pass all quality assurance criteria. It was also very easy
for other compounds to mask or otherwise interfere with the identifi-
cation of the Love Canal indicator compounds. Thus, it was possible
for samples with known concentrations of indicator chemicals, such
as the BQC samples, to be reported as having nondetectable concen-
trations of indicator chemicals.
STUDY RESULTS
Because of the complexity of the GC/MS analytical technique, a
number of factors influence the precision and accuracy of the method.
These include the frequency of calibration of the instrument, the labora-
tory performing the measurement and the soil matrix being analyzed.
Further, each of these factors can have a different influence on each
compound being analyzed. Each estimator of either accuracy or
precision reflects the influence of confounding factors differently.
Three of the measures discussed (BQC samples, matrix spikes and
surrogates) can be used to estimate accuracy. Soil samples do not have
a known concentration, so accuracy cannot be calculated for replicate
analyses.
Box plots are one method of comparing the different measures of
qualities. Notched box plots are a method of presenting and comparing
distributions of values without making assumptions about the form of
the distribution. Figure 2 illustrates the attributes of a box plot. Each
box plot presents six statistics about a distribution in graphical form.
These are:
• The 25th percentile of values, represented by the bottom of the box
• The median or 50th percentile of values, represented by the line within
the box
• The 75th percentile of values, represented by the top of the box
• The range of the data value, represented by the lines extending from
the ends of the box
• Outlier values, represented by asterisks or circles
• Approximate 95 % confidence limits for the median represented by
an indentation or notch in the box (If the confidence limits are wider
than the box, the box will be folded at the notch, resulting in a some-
what peculiar figure.)
Distributions of values can be compared across categories by
examining the notches on each box. When notches do not overlap, the
median values are significantly different. When the notches do overlap,
there is no significant difference between the medians.
QUANTITATIVE OA/QC
QUALITATIVE OA/QC
Logbooki
Sample Tnclun0 (In-HouM)
-------
O Far-out
valuaa
Outtlda
valua*
75th
50th
p«rc*ntila
— 25th
pOTCffltliA
-T- T
T '
Notch
_I-
1
Data
Subset
Figure 2
Sample Box Plot
OBtaBHC
QmraftC
Sioo
061 kCO LA82 hGO L>fll BQC U«2 BOC
LA61 MBO 1>B2 kOO U«1 BOC LAB2 BOC
LA82l>eO UABt BOC L>e2BOC
U>61 tOD L>B2 hGO mat BOC LAB2 BOC
Figure 3b
Distribution of Recoveries for Indicator Compounds
Tricftorctenzere
TstracrtactHnzene
LAB1 fc6D LAB2 fcGD LAB1 BQC L>S2 8OC
L>61 hGO L>B2 h60 LABI BOC LAB2 BOC
LAB1 MBO LAB2 M8O 1>BI BOC L>B2 BOG
LA81 MBO LAB2 hfiO LABI BOC LABS BOC
Figure 3a
Distribution of Recoveries for Indicator Compounds
OtranAereeiB
loo
0
.ISO
100
60
LAS1 SUR
LABI SUR
LAB28UR
Figure 3c
Distribution of Recoveries for Indicator Compounds
Figure 3a illustrates some of the factors influencing the accuracy of
GC/MS measurements. In this graph, the recovery of dichlorobenzene
52 MONITORING & SAMPLING
-------
is broken out into two factors that may influence the recovery: the labora-
tory doing the analysis and the source of the compound. These two
factors combine to create four categories of recovery estimate.
In Figure 3a there is a significant difference between the two labora-
tories in their ability to recover dichlorobenzene from the matrix spike
samples; Laboratory 2 is better than Laboratory 1. This trend is
consistent for all compounds spiked in the laboratory (i.e., both matrix
spikes and surrogates). However, there is no consistent difference
between the two laboratories for the blind quality control samples spiked
at the U.S. EPA laboratories.
The inconsistency of these results illustrates some of the subtle
problems associated with estimating laboratory performance measure-
ments. There are several plausible explanations as to why such
differences exist. Laboratory 2 could have a different technique for
adding spiking compounds and then extracting them. This method might
differ from that used by Laboratory 1 and allow Laboratory 2 to retrieve
newly added compounds more effectively.
Another possibility is the difference in exposure time for compounds
added to soil in the laboratory and immediately extracted, and
compounds added to soil at a U.S. EPA laboratory and then stored for
some time before extraction and analysis. Although Figure 3a shows
a more efficient extraction of dichlorobenzene for BQC samples, in
general for the other compounds the BQC samples have smaller recov-
eries than matrix spike samples.
Although the recovery of spiked compounds typically is 50 to 75 %,
the recovery observed in any one sample has a broader range. The range
of recoveries observed is one measure of the precision attainable with
the measurement process at a study wide resolution. However, this statis-
tic is not available for the concentrations estimated for replicate samples.
Dchtabercere
samples is the mean recovery of BQC samples for a laboratory. The
nominal value for replicate samples can be either the mean concentra-
tion of replicates or the mean concentration of samples from one area
analyzed at a laboratory. The first estimates short-term precision (within
1 day); the second a longer term precision over the study.
AfchaBHC
B »
MaBHC
BetaBHO
Gamma BHC
Tricttorobenzene
D 10O
I 60
Figure 4b
Estimates of Precision for Indicator Compounds
Tetradtrooanzene
i S
Mxanobenzene
LAB1 REPS LABZFEPS LA61 SBR LAB2 NBR
Figure 4a
Estimates of Precision for Indicator Compounds
Figures 4a through 4c show the distribution of a statistic indicating
the scale of the estimated concentrations. This is the absolute value
of the percent difference between a concentration estimate and its nomi-
nal value. The nominal value for matrix spike samples is the mean
recovery for the two spiked samples. The nominal value for the BQC
Tetratrcmcbenzene
LAB1 REPS LAB2 REPS LA81 tBR LAB2 NBR
l>81 REPS LAB2REPS LAS1 teK LAB2 NBR
Figure 4c
Estimates of Precision for Indicator Compounds
MONITORING & SAMPLING 53
-------
Figures 4a through 4c show each of these four estimates of precision
for both laboratories. Samples labeled "MSD" are the matrix spike
samples; those labeled "BQC" are the blind quality control samples;
those labeled "REPS" are the replicate samples compared to replicate
means; and those labeled "NBR" are The replicate samples compared
to neighborhood means.
In general, the NBR and BQC estimates appear to be more variable
than the other estimators. This is not surprising, as the baseline for
both measures, the mean value over neighborhood for the entire study,
is much broader than the other measures.
Typically the reproducibility of measurements is from 10 to 20% for
most compounds and measures. However, this can vary from compound
to compound depending on the measure used.
Of the four measures of precision and accuracy, only one, the
surrogate recoveries, is available for individual samples. A natural ques-
tion is whether this measure is of sufficient quality to allow concentra-
tion estimates for a sample to be recovery corrected. In other words,
can the surrogate recoveries be used as an estimate of bias to correct
the indicator chemical concentration estimates?
DicWorobenzene
Trichbrobenzene
Tetrachlorobenzene
Chtoronaphtalene
AJphaBHC
DetaBHC
BetaBHC
Gamma BHC
Oibromobenzene
Tetrabromobenzene
Tetrachlorobenzene
Figure 5a
Comparison Sample Versus Surrogate Precision,
Replicate Sample
Dtchtorobenzene
Tilchlorobenzene
Tetrachtorobenzene
Chtoronaphtalene
Alpha BHC
Delia BHC
Beta BHC
Gamma BHC
Dftyomobenzene
Tetrabromobenzene
Tetrachlorobenzene
Figure 5b
Comparison Sample Versus Surrogate Precision,
Study Means
One problem with using surrogate compounds to correct for analyti-
cal bias is that there is no predefined correspondence between a par-
ticular surrogate and an indicator compound. The surrogates were
chosen to span the range of elution times for indicator compounds
through the gas chromatograph. Figures 5a and 5b show scatter plots
of each indicator compound precision against the surrogate compounds
precision. Figure 5a shows the precision calculated on the basis of repli-
cate samples extracted on the same data. Precision in Figure Sb is
calculated on the basis of the neighborhood means.
In these scatter diagrams, if surrogate precision and indicator chemi-
cal precision were perfectly correlated, the data points would align along
the diagonal of each plot. A regression line is drawn in each plot to
illustrate the !actual correlation. As can be seen in the plots, the line
intercepts the indicator variable axis, indicating that target compound
variability is underestimated by surrogate variability. In general, the
regression lines are not parallel with the diagonal, indicating lack of
correlation of the two measures of precision.
Figures 6a and 6b are similar to Figure 4 in showing the absolute
percent difference from nominal values for different estimators. An ad-
ditional pair of estimators has been added to this figure, which shows
the distribution of absolute percent difference for corrected concentra-
tion estimates as compared with the mean value of the replicate. As
can be seen, the variability shown by the corrected concentration esti-
mates is similar to that seen in the uncorrected estimates. The precision
of the corrected estimates varies by compound from the uncorrected
estimates, with some being unproved by the correction and others be-
ing worsened.
DtHcrcbenzene
I:
i A
CntaronettthjfenB
i°
? 1 S
Figure 6a
Estimated Precision for Indicator Compounds Compared with
Surrogate Recovery Corrected Precision
54 MONITORING & SAMPLING
-------
AfchaBHC
A 6
Delta BHC
Beta BHC
a ? 5 g A 5
GamnaBHC
Figure 6b
Estimated Precision for Indicator Compounds Compared with
Surrogate Recovery Corrected Precision
CONCLUSIONS
Good estimators of accuracy and precision are required for an analysis of
environmental data as this analysis begins the chain of events that leads to good
decisions. Good estimators are needed for good designs, which enable good
decisions to be made in an uncertain environment. Reliance on internal labora-
tory estimates of precision and accuracy through the use of matrix spike and
surrogate spike data may overestimate the precision and accuracy achieved by
a study. Replicate analyses and the use of high-quality spiked samples prepared
by another laboratory are the best measures of precision and accuracy.
Precision should be based on repeated measurements over the course of a study.
Precision thus reflects the reproducibility of analyses conducted at different times.
This type of comparison is one of the most frequently used in environmental
data analyses. The usual split sample replication does not measure all sources
of data vulnerabilities.
Finally, correcting concentration estimates from a GC/MS procedure such
as surrogate recovery estimates does not appear to improve the precision of the
estimates.
MONITORING & SAMPLING 55
-------
Understanding Electrical Leak Location Surveys of
Geomembrane Liners and Avoiding Specification Pitfalls
Glenn T. Darilek, RE.
Daren L. Laine
Southwest Research Institute
San Antonio, Texas
ABSTRACT
The electrical leak location method developed under contract for the
U.S. EPA is now being put to use in many commercial applications,
and several contractors are providing electrical leak location services.
The commercial surveys conducted to date have been overwhelming
successes in that many leaks have been efficiently and accurately located
in installations that had been previously tested certified leak-free
environment conventional methods. The results of these surveys lead
to the speculation that a pre-service electrical leak location survey should
be performed on every geomembrane-lined landfill and impoundment
before the installation is considered complete and ready for use.
The electrical method detects areas of localized electrical current flow
through leaks in the otherwise insulating liner. A voltage source is con-
nected to an electrode in water covering the liner and to a grounded
electrode. Leaks are located by searching for the localized areas of rela-
tively high electrical potential in the water caused by current flowing
through a leak. The electrical leak location method can be used in liquid
impoundments and for a pre-service inspection of solid waste landfills.
The testing method will not damage the liner.
As with any new technology, many people in the environmental
industry want a better understanding of the principles, capabilities and
the proper application of the method. Specifiers of electrical leak
location surveys must have this knowledge to specify the most effec-
tive and economical surveys. The objective of this paper is to provide
important up-to-date information to meet this need.
INTRODUCTION
Geomembrane liners, also known as flexible membrane liners
(FMLs), synthetic liners and membrane liners, are sheets of polymeric
materials fabricated in a factory and seamed together at the field site
to form a continuous liner. Installation can result in punctures or sepa-
rated seams, causing loss of the liner's physical integrity. Damage also
can be accidentally caused by heavy machinery used to place protec-
tive bedding material on the liner.
An electrical leak location method was developed and tested under
contract for the U.S. EPA. This method has been demonstrated to be
the most sensitive, reliable and valid method for locating leaks in geo-
membrane liners of waste landfills and impoundments. The electrical
leak location method is now being widely applied and several contrac-
tors are providing electrical leak location services. Several technical
references for the electrical leak location method are listed in the
Bibliography.
Results of Leak Location Surveys
Southwest Research Institute has surveyed 56 geomembrane-lined
storage facilities for leaks using the electrical leak location equipment.
The total liner area surveyed was more than 4,4,000,000 ft2. The sizes
of these installations ranged from less than 1000 ft* to more than
500,000 ft2 and included both double- and single-lined impoundments
and landfills. Almost all of the liners were in new installations. Most
of the liners were constructed of high density polyethylene (HDPE),
but some were chlorosulfonated polyethylene (CSPE) and polyvinyl
chloride (PVC).
Leaks were found at all of the sites except for two sites with small
liners. The average density of leaks was approximately one leak per
3200 ft2 13 leaks per acre. Although most of the leaks occurred in
field seams, a significant number (more than 15%) were found in the
parent material. The high percentage of leaks found in the seams is
partly attributed to the fact that some very small seam leaks are found
when the seams are surveyed a second time with the leak location probe
on the seam.
Typical installations had from four to 12 leaks per acre. Installation
and field seaming problems were experienced on the liners with greater
than 20 leaks per acre. Several of the liners had more than 50 leaks
per acre.
Because some leak location surveys were initiated in response to a
known leakage problem, a significantly higher number of leaks might
be expected for these installations. However, the number of leaks at
the installations with known problems were fewer than installations
where the leak location surveys were performed for construction quality
assurance purposes. The results of these surveys indicate that a pre-
service electrical leak location survey should be performed on every
geomembrane-lined landfill and impoundment.
TECHNICAL DISCUSSION
Theory of Operation
Figure 1 shows the basic electrical leak location method for locating
leaks in a geomembrane liner. The principles of the electrical leak
location method are relatively uncomplicated. A DC voltage is con-
nected to electrodes placed in electrically conductive material above
and below the liner. The impressed voltage produces a very low current
flow and a relatively uniform electrical potential distribution in the water
above the liner in areas with no leaks. If the liner has a leak, water
flows through the leak and establishes an electrical current path through
the liner. Leaks are located by searching for the localized areas of rela-
tively high electrical potential in the water covering the liner. The
increased current density near the leak is indicated as an anomaly in
the measured potential. The electrical leak location method can be used
in liquid impoundments, as a pre-service inspection of solid waste land-
fills and to locate leaks in the final cover for landfills or impoundments.
This testing method does not damage the liner.
If applied properly, the electrical leak location method is very sensi-
tive. To increase the leak detection reliability to a maximum level, leak
56 HEALTH & ENDANGERMENT
-------
CURRENT SOURCE
ELECTRODE
MOVING
MEASUREMENT
ELECTRODES
LIQUID
\LEAK \
PATH \ MEI
EARTH
MEMBRANE
LINER
Figure 1
Diagram of the Electrical Leak Location Method
A source of DC power is used to impress a voltage across the geo-
membrane liner. Figure 4 shows an electrical leak location power supply
with self-contained safety system. The leak detection sensitivity is
proportional to the voltage output of the power supply. Batteries can
be used for a safe low voltage power supply, but leak detection sensi-
tivity will be decreased to a level where smaller leaks can not be detected
and the leak detection reliability is decreased. For best results and sen-
sitivity, a high voltage electronic power supply is used with a safety
circuit. The high voltage power supply has an adjustable output level
of up to 320 V DC. The safety circuit provides a measure of protection
from accidental contact between earth ground and either power supply
output or accidental contact across the power supply output. The safety
circuits disconnect the power when a ground fault current is detected,
or the output current momentarily increases or decreases due to possi-
ble human contact. A bright flashing warning light indicates that the
power supply is energized.
location surveys should be conducted with the maximum practical
impressed voltage and detector sensitivity. Some of the leaks that are
found are very small and may not leak significantly. Nevertheless, all
detected leaks are located and marked for repair. The small leaks can
indicate a weak seam that may fail with time or loading. In almost every
survey, several larger leaks that require repair are found. The small
leaks are repaired at the same time the larger leaks are repaired to in-
crease confidence in the integrity of the liner.
Instrumentation
The manual leak location survey system consists of a lightweight,
portable electrical probe and associated instrumentation. This system
is for inspection of non-hazardous liquid-filled impoundments and for
pre-service inspection of water-filled impoundments and landfills.
Figure 2 illustrates the operation of the equipment.
Figure 3
Leak Location Detector Electronics Assembly
Figure 2
Manual Leak Location Equipment Consisting of an Electrode Probe
and Electronics Unit
Figure 3 shows a typical detector electronics assembly. The battery-
powered detector electronics provides an audio tone that varies in propor-
tion to the measured signal so the operator is not required to continuously
monitor the meter. Controls are provided to adjust the sensitivity,
threshold and audio output level. Test buttons are provided to check
the battery voltage and circuit operation. Connectors are provided to
connect the detector probe outputs and an earphone for the audio
indicator.
Figure 4
Leak Location Power Supply
The detector probe is a long pole with two electrodes. A cable con-
nects the electrodes to the input of the detector electronics. The probe
is most conveniently used while wading in the liquid but, with an
extension, it can be used from a raft in deeper water applications. Sur-
veys of the side slopes are accomplished using a probe with a long handle
HEALTH & ENDANGERMENT 57
-------
and small wheels to support the electrodes. The side slope area is
surveyed by systematically lowering the probe down the slope and then
pulling it up the slope.
EFFECT OF MEASUREMENT PARAMETERS
Computer Model
A mathematical model was developed to investigate the performance
capabilities of the electrical leak location method. The model accom-
modates various electrical and dimensional parameters for a lined
impoundment or landfill. Model studies of the electrical leak location
survey technique were conducted to characterize the performance of
the method with various electrical parameters of the waste materials,
the measurement electrode array geometry, the measurement electrode
depths and proximity to the leak and the size and number of leaks.
Anomaly Effects of a Leak
Figure 5 shows a typical family of leak anomaly responses for horizon-
tal detector electrodes that illustrate the effects of various measurement
depths. The two peaks in the signal occur when the two electrodes pass
within closest proximity of the leak. Figure 6 shows the amplitude of
the leak anomaly for three different electrode spacings as the electrodes
are scanned at various depths. A substantial improvement in detection
sensitivity is obtained when the potential array is scanned closer to the
leak. The computed leak responses and field experience affirm the prac-
tical importance of performing the survey measurements near the bottom
of the impoundment.
-20
HORIZONTAL SCAN DISTANCE, y (m)
Key: s = electrode spacing
h = depth of the water
»w = liquid resistivity
,s = underlying soil
resistivity
= leak radius
= electrode depth
= offset distance
= distance along
along scan line
Figure 5
Plot of the Leak Anomaly Versus Horizontal Electrode Depth
Figure 7 shows the anomaly response of a leak measured with a
vertical electrode pair. The leak is located at the position indicated by
the maximum response. Multiple leaks can be resolved with less
ambiguity when vertical electrodes are used. Again, the computed leak
responses point out the practical importance of performing survey
measurements near the geomembrane liner.
OJ 0.4 OJ Of 0.7
ELECTMCM. SURVEY DEPTH, i,. (ml
Figure 6
Leak Signal Amplitude Versus Survey Depth
a.
<
I
UJ
i
a
4U
18
16
14
12
10
8
6
4
2
o
h 1m
»v 0.3m
pw = ISftm
p, = 30 n m :
, a - 0.0004m
-
-
r* — *m = O-95 m
1
'
i
)
« !„, = 0.9 m
'
.
-
V
i.i. , i - i i_^ ~
0 2 4 6 8 10 12 14 16 18
HORIZONTAL SCAN DISTANCE, y (m)
Figure 7
Leak Anomaly Characteristic for Vertical Electrodes
Effect of Measurement Electrode Spacing
In general, the amplitude of the measured leak signal increases as
the electrode spacing increases. However, the increase is negligible when
the electrode spacing is somewhat larger than the distance to the leak.
This principle can be demonstrated by considering the equation for the
voltage at some distance from the leak. The simplest mathematical model
of a leak is to consider that the leak is a point current source in an
infinite half space. If „__ is the resistivity of the water, I is the current
and the distances from the leak to the two measurement electrodes,
the measured voltage difference will be:
1
27T
58 HEALTH & ENDANGERMENT
-------
Figure 8 shows the amplitude of the leak signal versus electrode sepa-
ration when the electrode closest to the leak is 0.05,0.1 and 0.2 meters
with a current of 5 mamp and a water resistivity of 10 ohm-meters.
The graph shows that little is gained by increasing the electrode spacing
beyond approximately 0.3 meters.
0.15
0.1
0.05
0 0.2 0.4 0.6 0.8 1
DISTANCE (METERS)
D 0.05 TO LEAK + 0.1 TO LEAK A 02 TO LEAK
Figure 8
Leak Signal Amplitude Versus Electrode Separation
Effect of Water Resistivity
Figure 9 shows the amplitude of the leak anomaly for different values
of water resistivity and water depth with the electrodes suspended mid-
way in the water. These curves show that for a given amount of leak
current, the leak detectability is increased essentially linearly with the
resistivity of the water. The injected current must be increased to offset
the effect of lower measured leak anomaly attributed to lower resistivity
of the liquid. For constant current injection, the amplitude of the leak
anomaly is essentially independent of the resistivity of the material under
the liner.
O
1 "
12
h = 0.5 m
'm
s
h/2
= 1 m
p, = 30 flm
a = 0.0004m
0 10 20 30
WASTE MATERIAL RESISTIVITY, pw (ft-m)
Figure 9
Leak Signal Versus Water Resistivity for
Various Water Depths
In practice, a constant voltage power source is used rather than a
constant current source. Therefore, as the water resistivity is decreased,
more current will flow through the leaks. However, the amount of current
increase does not offset the decrease in signal level.
Effect of Offset Distance from Leak
The maximum allowable spacing between the lateral survey lines
depends on the amount of current flowing through the leak and the sen-
sitivity of the leak location equipment. To illustrate this characteristic,
Figure 10 shows the amplitude of the leak anomaly for various elec-
trode offset distances from the leak center as a function of the survey
height above the liner. The amplitude of the anomaly decays rapidly
as the offset distance is increased. These results indicate the impor-
tance of scanning the electrodes close to every point on the liner to
obtain a high level of leak detection sensitivity.
___-=iifcdL
0.1 0.15 0.2
HEIGHT ABOVE LINER, h - zm (m)
Figure 10
Leak Signal Amplitude Versus Height Above Liner for
Various Lateral Offset Distances
Leak Location Accuracy
The leak signal is at a maximum when the leak location electrode
is touching the leak. Therefore, leaks are very accurately located by
decreasing the sensitivity of the leak location electronics to a level where
the point of maximum signal can be observed. The location of the leak
can be essentially pinpointed in this way.
Effect of Leak Size
The size of the leak and the conductivity of the water essentially
determine the amount of current flowing through the leak for a given
impressed voltage. Because the leak signal is proportional to the amount
of electrical current flowing through the leak, larger leaks are much
easier to detect the smaller leaks. Experimental measurements of leak
current versus leak diameter for circular leaks show that the amount
of current flowing through the leak is approximately inversely propor-
tional to the diameter of the leak. Other tests have been conducted to
show that the shape of the leak has little effect upon the shape of the
leak signature.
Effect of Liner Resistivity
Because the liner resistivity is many orders of magnitude greater than
the resistivity of the water, the liner resistivity has no effect on the leak
detection sensitivity. Laboratory tests have been conducted to show that
the change in liner resistivity versus time for exposure to typical levels
of acidity, alkalinity and dissolved salt content have negligible effect
on the resistivity of the liner material.
Effect of Sediment Layer
The electrical leak location method is less sensitive for locating leaks
in geomembrane liners with a sediment layer in the liquid. Physical
model tests and field experience indicates the lower sensitivity and that
the measurements are not as repeatable with sediment layers present.
HEALTH & ENDANGERMENT 59
-------
The lower sensitivity occurs probably because the electrodes cannot
be scanned close to the leak and the liquid shunts the measured poten-
tial field to some degree.
Effect of Soil Cover
Figure 11 is a plot of a measured leak anomaly versus depth of soil
cover for a geomembrane liner when the electrodes are scanned directly
over the leak. The diameter of the leak was 0.3 cm. Although the leak
signal decreases rapidly with increasing soil cover thickness, the leak
anomaly was easily detected for soil depths up to 0.6m. Figure 12 shows
plots of the data with a soil thickness of 0.3m for scan lines offset from
the leak. The leak is barely detectable when the electrodes are scanned
on a line offset 0.6m from the leak. The signal can be improved by
scraping the dry soil off the surface or inserting the electrodes into the
more moist underlying soil. Figure 13 shows the decrease in the
measured noise for these conditions with a soil thickness of 0.6m.
LEAK DIAMETER = 3 MM
01 2345
DISTANCE (METERS)
+ 15.2 CM SOIL 0 25.4 CM SOIL A 30.5 CM SOIL x 61 CM SOIL
Figure 11
Leak Signals for Various Thicknesses of Soil Cover
TYPES OF SURVEYS AND SURVEY TECHNIQUE
Survey of Bottom of Water-Covered
Single Liners or Secondary Liners
When a single liner is in place, the leak location power supply is
connected to a source electrode in the water and a grounded electrode.
Surveys are conducted along survey lanes established across the im-
poundment. The most convenient method of operation is to place the
lines across the shorter dimension of the impoundment and perpendic-
ular to a straight side. Survey lines are spaced approximately 5m apart.
Sufficient accuracy usually is obtained using only a tape measure. Marks
are put on the liner above the water line every 5m on the opposite sides
of the impoundment. Floating polyethylene ropes or non-conducting
survey chains are stretched between opposite marks across the impound-
ment. As an alternative procedure, the panel seams can be used as the
survey lanes. Two or three survey operators can scan the length of a
i
!
Ul
cc
cc
i
o
Soil Cover a 30 cm
Current • 6 mA (320 V)
DISTANCE (METERS)
CENTER * 0.6 M WEST o 13. M WEST A 1.8 M WEST
Figure 12
Leak Signals with 0.3m of Soil Cover for Offset Scan Lines
D UNPREPARED SOIL
DISTANCE (METERS)
* ELECTRODES INSERTED o MOIST TOP SOIL
Figure 13
Improvement in Leak Signal Quality When the Soil is Prepared
panel with overlapping coverage by observing or feeling the seams. This
alternative procedure is more difficult or impractical to implement with
irregular panel layouts.
Horizontal traverse lines are scanned with a coverage of 2.5m on each
side of the traverse lines. The probe is scanned along the bottom in
60 HEALTH & ENDANGERMENT
-------
an arc overlapping under the traverse line and past the midpoint of the
survey lane. After each arc is swept, the operator moves forward ap-
proximately 0.3m and scans a return arc to just beyond the traverse
line. The leak detection probe is thus scanned within no more than ap-
proximately 0.15m of every submerged point on the liner. The threshold
control on the leak location electronics is adjusted frequently to main-
tain maximum leak detection sensitivity.
Leaks are indicated by a sudden increase in the frequency of the tone
in the earphone as the electrode is scanned near the leak. When a leak
is detected, the threshold and sensitivity controls are adjusted to obtain
a peak on-scale meter reading both laterally and longitudinally when
the tip of the probe is scanned. This procedure determines the exact
location of the leak. The probe tip is held on the leak while the probe
is swung to vertical. The leak is then marked with lead sinkers con-
nected to a small float with a length of string.
The locations of the leaks also are measured relative to a temporary
survey grid for a permanent record. Where practical, the location and
type of leak also is noted (i.e., on a seam or patch, or in the panel).
In addition to covering every square meter of the liner, all liner field
seams and patches are double checked.
Survey of Bottom of Water-Covered Primary Liners
By placing the current return electrode in electrical contact with the
liquid-saturated drainage layer located between the two liners, the elec-
trical leak location method can be used to locate leaks in the upper
liner. The survey procedures for a single liner are then followed. Simple
electrical continuity tests between the drainage layer and the earth also
can determine the existence of leaks in the bottom liner but not their
location.
Survey of Side Slopes
Surveys of water-covered side slopes are accomplished using the probe
with a long handle and small wheels to support the electrodes. The
side slope area is surveyed by systematically lowering the probe down
the slope and then pulling the probe up the slope. The operator moves
forward approximately 0.3m between sweeps. Each survey sweep covers
an area approximately 0.3m wide down the flooded sidewall. Any leaks
found are accurately located, and the locations are referenced to a tem-
porary survey grid established on the berm.
When more than approximately 7m of the side slope are immersed,
the manual survey of the side slopes is conducted in stages. The water
level is raised or lowered in stages that allow approximately 7m of the
immersed side slope to be surveyed at a time. The surveys should pro-
vide overlapping coverage between the stages.
The side slopes can be surveyed by raising or lowering the water level
in stages either before or after the bottom of the liner is tested. If the
side slopes are tested first, from the top down, the cell will be filled
with water to the working level prior to the leak location survey. This
procedure exposes the liner to loads representative of actual in-service
loading. Usually the level of the water can be lowered faster than it
can be raised, therefore, the survey can be completed with less standby
time as the water level is adjusted.
The advantage of surveying the side slopes after the bottom of the
liner is surveyed is that washout or settling of the subgrade under the
liner caused by possible large leaks in the bottom of the liner might
be avoided if leaks in the bottom are located and repaired prior to full
hydrostatic loading. However, there is no assurance that additional leaks
will not occur because of the increased hydrostatic loading during the
side slope survey. Therefore, additional testing of the bottom of the liner
may be required after the side slopes are surveyed.
Survey of Soil-Covered Liners
Often a layer of sand or soil is placed on the liner to serve as a pro-
tective layer or drainage layer. Geomembrane liner material is also
covered with soil when used for landfill final cover systems. Because
of the high probability of damaging the geomembrane liner in the process
of emplacing the soil, a leak location survey of the soil-covered geo-
membrane is a highly effective method of ensuring the integrity of the
liner. The electrical leak location method is the only method capable
of locating leaks in a geomembrane covered with protective soil. The
method is particularly valid because the liner is tested under load and
after the liner has been exposed to possible damage incurred in the
process of emplacing the protective soil cover.
The electrical leak location method was modified to make surface
soil potential measurements to locate leaks in geomembranes covered
by a protective or cap soil layer. The soil is dampened with water to
allow good electrical contact and allow the water to percolate through
the leaks. Completely flooding the liner is not necessary. Surface poten-
tial measurements are made using a portable digital data acquisition
system. Surveys are conducted by making potential measurements on
closely spaced survey lines. Point-by-point potential readings are made
along the survey lines with a fixed measurement electrode separation.
The data are downloaded to a computer for storage and plotting. When
a suspect area is located, manual measurements are made to further
isolate the leak. When the surface of the soil is dry, the dry soil is scraped
away so that accurate measurements can be made on the uncovered moist
soil.
The data are examined for leak signatures. The characteristic leak
signal is a bipolar signal with the initial signal deflecting opposite to
the polarity of the current injection electrode. Signals caused by other
features such as drainage laterals can be recognized and rejected.
The leak location sensitivity increases as the thickness of the soil
decreases. Typically, leaks with a diameter greater than 0.3cm can be
located in a geomembrane covered with 0.3m of soil. Testing for leaks
with only a portion of the soil cover in place is recommended if the
thickness of the soil cover will be greater than approximately 0.3m.
Any possible damage to the liner will most likely occur during the in-
stallation of the first layer of soil.
The leak location accuracy for surveys conducted with soil cover
depends upon several factors including the closeness of the spacing of
the point-by-point measurements and the homogeneity of the soil cover.
A practical accuracy guideline for leak location surveys with soil cover
is approximately one half of the soil thickness. After the soil has been
removed, followup measurements can be made to locate the leak within
1.5 cm.
The survey parameters (survey line spacing, spacing of measurements
and spacing of measurement electrodes) must be designed for proper
coverage and leak detection sensitivity. The design of the surveys must
be based on the physics of the electrical leak location method.
Another survey methodology can be successful in some cases, par-
ticularly when an electrical leak location was previously conducted with
the liner flooded with water and only a few major leaks are suspected.
Rather than performing a systematic survey on closely spaced survey
lines to locate smaller leaks, the reconnaissance measurements are
intended to attempt to isolate a few large leaks in the hope that no smaller
leaks are present. The measurement sequence is to locate a leak, remove
the soil from over the leak, insulate the leak and then measure the power
supply current. This sequence is repeated until the current level
decreases to a low level indicating that all of the major leaks are found.
Multi-Channel Leak Location Surveys
Southwest Research Institute has developed a multi-channel leak
location system for locating leaks in impoundments with hazardous
wastes, for locating leaks in the side slopes of deep impoundments in
one stage and for surveying in deep water. The system is particularly
cost-effective for large impoundments and landfills. The new system
has 12 weighted electrodes suspended from a nonconducting horizon-
tal axle between two large plastic wheels. Twelve data acquisition chan-
nels, a serial data telemetry system and a portable computer or
multi-channel chart recorder are used to acquire, display and record
the leak location data.
The sensor assembly is systematically pulled across the bottom of
the impoundment using a power winch. Each survey sweep covers an
area approximately 4m wide. If feasible, the sweeps are referenced to
liner seams to provide overlapping coverage of the seams as well as
complete coverage of the water-covered liner panels. The locations of
HEALTH & ENDANGERMENT 61
-------
the leaks are referenced to a temporary grid system established on the
berm of the impoundment.
The leak location data acquisition system has been applied at one
large impoundment to survey the 18-m-long side slopes. The sensor
and electronics subsystems operated properly and located several leaks.
Mechanical modifications are needed to make the assembly more
rugged.
Remote-Controlled Leak Location Survey System
A small remotely-controlled boat equipped with potential measure-
ment electrodes and electronics, servo-controlled steering and data
telemetry has been developed to locate leaks in hazardous waste im-
poundments. In one mode, the measured potentials are used with the
servo-controlled steering to automatically seek leaks. The system has
been constructed and tested in a geomembrane test impoundment. The
method is described in U.S. Patent 4,719,407 for Automated Search Ap-
paratus for Locating Leaks in Geomembrane Liners.
SITE PREPARATIONS
Water Covering the Liner
To conduct a leak location survey of the bottom of the liners, a mini-
mum of 0.15m and a maximum of 0.75m (0.6m preferred) of water con-
taining no hazardous or foul substances must cover the liner. Because
hydrostatic loading produces mechanical stress in both the seams and
the material, leaks may occur only after the liner is subjected to these
loads. Therefore, testing the liner after the impoundment has been filled
with water is a valid method for determining if leaks will occur under
realistic loading conditions.
The depth of the water for the survey (within the specified range)
can be determined on a case-by-case basis. Surveying with a shallow
water level requires less water and pumping, but limits the hydrostatic
loading. The survey covers only the submerged liner area when the
cell is filled with water to the depth specified for the survey. There-
fore, surveying with shallow water decreases the amount of the side
slope that is covered by water and thereby limits the area of survey cover-
age for the cases where all of the side slopes are not surveyed.
Flooding the Leak Collection Zone
To survey the primary liner of a double liner system, an electrical
conduction path through any leaks to the leak collection zone must be
established. This process can be accomplished by pumping water in
the leak collection system while the primary liner is being filled with
water. Water can be pumped into the discharge side of the leak collec-
tion system. In some cases, air vents must be provided in the perimeter
edges of the primary liner near the top of the berm to allow air trapped
between the two liners to be vented. The water also can be pumped
into the air vents. The water level in the leak collection zone must be
slighdy below the level of the water in the primary liner to prevent the
primary liner from being lifted.
In some cases when moist sand is used in the leak collection zone,
an alternative method can be used to establish the electrical conduc-
tion path without flooding the leak collection zone. The reliability of
this alternative method depends on the type and moisture content of
the sand. The alternative method is to allow the water from the leaks
to percolate through the leak collection zone. This method is most
effective when the water on top of the liner has been allowed to stand
at least 3 days and good electrical contact can be established with the
current electrode in the leak collection zone.
Current Electrode in Leak Collection Zone
Provisions should be made to allow the placement of a metal elec-
trode into the leak collection zone of a double liner system. In some
cases, a slit is cut in the liner above the water level to allow the inser-
tion of the metal electrode. This slit must be repaired when the leaks
are being repaired. In some installations, the electrode can be inserted
through a straight plastic pipe that extends down into the leak collec-
tion sump.
A third method for providing the electrode is to install a permanent
electrode constructed of approximately O.lm2 of thin stainless steel
sheet in the drainage layer near the lowest point of the leak collection
system. The corners and edges of the electrode should be rounded to
prevent damage to the liner In addition, the electrode can be wrapped
with geotextile or geonet to further protect the liners. An insulated wire
(16 AWG to 12 AWG) must be connected between the electrode and
a test terminal located at a convenient, accessible site near the im-
poundment. The connections should be insulated with a suitable coating.
Isolate Electrical Paths Through the Liner
The electrical leak location method locates leaks by detecting elec-
trical conduction paths through leaks in the liner. If feasible, any other
electrical conduction paths through or around the liner must be elimi-
nated or insulated. All penetrations, such as fill lines, drain pipes, batten
anchors, penetration flanges, footings, pump lines, pump wiring,
instrumentation wiring, instrumentation conduits and access ramps mak-
ing contact with the water in the liner should be insulated from ground
or constructed of an insulating material. Electrical paths also can be
established through the liquid in plastic pipes if the pipes connect to
a grounded metal valve or metal pipe.
Rubber packers can be placed in plastic drain and fill pipes to insure
that the fluid in the pipes does not act as an electrical path to ground.
In some cases a temporary geomembrane cover can be seamed over
pipes and batten anchor bolts Metal pipes penetrating the liner can
be insulated using large plastic garbage bags or caps constructed of
insulating foam rubber, geomembrane and plywood.
For the electrical paths to be a factor, the paths must form a conduc-
tion path through or around the liner being surveyed. The presence
of such electrical conduction paths does not preclude the application
of the method. However, if these paths can not be eliminated, isolated
or insulated, the paths will be indicated as leaks that may mask the
signal from other smaller leaks in their immediate vicinity. In addi-
tion, if the conduction paths are substantial!) lower in resistance than
the electrical paths through the leaks, the amount of current flowing
through the leaks may be too small to detect small leaks. The design
and construction of the impoundment can be reviewed to determine
the best methods to eliminate or minimize the effect of these conduc-
tion paths on the survey.
Remove Debris
For safety and better leak location reliability, debris such as unneces-
sary sand bags and non-floating liner material must be cleared from
the liner.
Conducting Structures
A leak is indicated as an electrical potential anomaly in an other-
wise relatively uniform potential distribution. Conducting structures
such as concrete footings, metal supports and sand bags can distort the
potential distribution, making leaks more difficult to locate. Small leaks
that are substantially covered by structures such as a concrete footing
probably cannot be detected. Moderate-size leaks at the perimeter of
such structures can usually be detected.
Power Requirements
Electric power of single phase 95 to 125 V AC, 45 to 70 hertz, at
approximately 5 amp must be provided at the site for operation of the
leak location power supply. The power outlet should be located at the
top of the berm.
SAFETY
A potential for injury is present in any work at a construction site.
Specific hazards include electroctuion, slipping and felling on the geo-
membrane material, falling in the water, hypothermia and drowning.
Job safety is the most important aspect of doing a complete and through
leak location survey. Proper safety precautions must be followed.
In addition to the standard construction site safety rules, specific safety
procedures must be used to safely conduct an electrical leak location
survey using a high voltage power supply. The survey operators wading
in the water are exposed to an electrocution hazard if they come in con-
tact with a grounded electrical conductor. Precautions must be taken
62 HEALTH & ENDANGERMENT
-------
to avoid this possibility. Some precautions include using only dry
electrically-insulating hand lines for entering or exiting the basin and
being sure that wire rope, wet rope, metal cables, electrically-conducting
poles, electrically-conducting ladders, or any other electrically con-
ducting objects are not available or used for rescue or used to aid
personnel in the water.
A safety circuit for the high voltage power supply provides a measure
of protection in case of accidental contact of personnel with the high
voltage. Because making the power supply inherently safe and making
the safety circuit completely reliable are not possible, survey procedures
and procedures should be such that personnel can never make electri-
cal contact across the power supply. The safety circuit must absolutely
never be tested by human contact. The safety interlocks must not be
bypassed to allow operation of the power supply without the flashing
red safety strobe.
Other elements of an effective safety plan include proper training
of survey personnel, safety briefing for visitors to the site, high-voltage
warning signs and employing personal flotation devices for operations
near deep water. The water in the impoundment must be non-hazardous
if an operator is to be completely immersed. Surveys must never be
performed when there is a threat of lightning or under adverse weather
conditions such as cold weather, rain, or snow or where the operator
has difficulty concentrating on safety.
On some work sites, the survey operators must be qualified to meet
OSHA 29 CFR 1910.120 safety requirements. This OSHA regulation
requires 40 hr of instruction, on-the-job-training, a medical surveil-
lance program and annual 8 hr training refresher courses.
Operators should be trained in first aid and cardiopulmonary resus-
citation. Additional safety procedures must be followed depending on
the hazards and conditions present at each site.
SPECIFYING ELECTRICAL LEAK LOCATION SURVEYS
The Appendix is a guide for specifying electrical leak location sur-
veys. The guide offers suggestions for typical surveys as well as assigning
responsibilities for preparations for the surveys.
BIBLIOGRAPHY
1. Peters, W.R., Shultz, D.W. and Duff, B.M., Electrical Resistivity Tech-
niques for Locating Liner Leaks, Proc. EPA Eighth Annual Research Sym-
posium Land Disposal, Incineration and Treatment of Hazardous Waste, Ft.
Mitchell, KY, Mar., 1982.
2. Peters, W.R., Shultz, D.W. and Duff, B.M., Electrical Resistivity Techniques
for Locating Liner Leaks, Technical Program Abstracts, Society of Explo-
ration Geophysicists 52nd Annual International Meeting, Dallas, TX, Oct.,
1982.
3. Shultz, D.W., Duff, B.M. and Peters, W.R., Performance of an Electrical
Resistivity Technique for Detecting and Locating Geomembrane Failures,
Proc. International Conference on Geomembranes, Denver, CO, June, 1984.
4. Shultz, D.W., Duff, B.M. and Peters, W.R., Electrical Resistivity Technique
to Assess the Integrity of Geomembrane Liners, EPA Report No.
EPA-600/S2-84-180, U.S. EPA, Cincinnati, OH, Jan., 1985.
5. Boryta, D.A. and Nabighian, M.N., "Method for Determining a Leak in
a Pond Liner of Electrically Insulating Sheet Material," U.S. Patent No.
4,543,525, Sept. 24, 1985.
6. Fountain, L.S. and Shultz, D.W, Liquid Waste Impoundment Leak Detec-
tion and Location Using Electrical Techniques, Proc. SME Annual Meeting,
New Orleans, LA, Mar., 1986, Preprint No. 86-95.
7. Converse, M.E. and Shultz, D.W, "Automated Search Apparatus for Locating
Leaks in Geomembrane Liners," U.S. Patent No. 4,719,407, Jan. 12, 1988.
8. Owen, T.E., "Geomembrane Leak Assessment Shell Shaped Probe," U.S.
Patent No. 4,720,669, Jan. 19, 1988.
9. Converse, M.E., Glass, K.B. and Owen, I.E., "Directional Potential Analyz-
er Method and Apparatus for Detecting and Locating Leaks in Geomembrane
Liners," U.S. Patent No. 4,725,785, Feb. 16, 1988.
10. Darilek, G.T. and Parra, J.O., The Electrical Leak Location Method for Geo-
membrane Liners, U.S. EPA Report No. U.S. EPA/600/S2-88/035, U.S. EPA,
Cincinnati, OH, Mar., 1988.
11. Darilek, G.T. and Parra, J.O., The Electrical Leak Location Method for
Geomembrane Liners, Proc. U.S. EPA Fourteenth Annual Research Sym-
posium, Land Disposal, Remedial Action, Incineration and Treatment of
Hazardous Waste, Cincinnati, OH, May, 1988.
12. Cooper, J.W., "System for Determining Liquid Flow Rate Through Leaks
in Impermeable Membrane Liners," U.S. Patent No. 4,751,467, June 14, 1988.
13. Parra, J.O. and Owen, T.E., Model Studies of Electrical Leak Detection
Surveys in Geomembrane-Lined Impoundments, Geophysics, 53, p.
1453-1458, 1988.
l4. Parra, J.O., Electrical Response of a Leak in a Geomembrane Liner, Ge-
ophysics, 53, pp. 1445-1452, 1988.
15. Darilek, G.T, Laine, D.L. and Parra, J.O., The Electrical Leak Location
Method for Geomembrane Liners-Development and Applications, Proc. In-
dustrial Fabrics Association International Geosynthetics '89 Conference,
San Diego, CA, Feb., 1989.
16. Laine, D.L., Detection and Location of Leaks in Geomembrane Liners Using
an Electrical Method: Case Histories, Superfund '89, HMCRI 10th National
Conference and Exhibition, Washington, DC, November, 1989.
APPENDIX
Specification Guide for the Electrical Leak Location Method
For a Geomembrane Leak Location Survey
With No Soil Covering the Liner
Introduction
This list of typical specifications is presented with relevant general
discussion to explain the preparations required for surveying primary
or secondary liners for leaks using the electrical leak location method.
Electrical leak location surveys can be contracted for by the owner or
operator of the facility, the general contractor, a third-party quality
assurance contractor, or the liner installer. To best serve the interests
of the facility owner, the electrical leak location surveys should be con-
tracted for by the owner or operator of the facility, or a third-party
quality assurance contractor. The following specifications are written
for this type of contractual arrangement. Separate specifications are
required for the general contractor and for the electrical leak location
contractor.
The specifications are for a manual survey of liners with no soil or
sand covering the liner. The specifications are intended for guidance
and reference only. They are not intended to be all-inclusive, to be neces-
sary in every application, or to recommend any particular practices
or procedures. The specifications for each installation should be written
specifically for the application, using proper engineering practices and
judgement and legal advice and review. Each use of the designations
of Company and Contractor should be reviewed and changed as
applicable to refer to the owner of the facility, the general contractor,
the liner installer, an independent quality assurance consulting firm,
or other subcontractor as applicable. Other terms such as landfill, im-
poundment or pond should be used as appropriate. The specifications
are written to be very comprehensive. They should be abbreviated where-
ever possible. The paragraphs typed bold are provided for explanation
and can be omitted from the specification.
Electrical Leak Location Survey Specifications for
General Contractor
Electrical Leak Location Survey
Under Hydrostatic Load
An electrical leak location survey will be performed by Southwest
Research Institute, 6220 Culebra Road, San Antonio, Texas 78238, (Con-
tact Daren L. Laine, telephone 512-522-3274) or approved equivalent.
The survey will be conducted on the bottom and side slopes of both
the primary and the secondary geomembrane liners of the basin. Con-
tractor will be responsible for preparing the basin for the survey as
described below.
If more than one leak per 2000 ft2 of surveyed area is found in
either liner, the leak location survey will be limited to one man-day
of survey per 20,000 ft2 of liner material. The electrical leak location
survey will be conducted to better categorize the occurrence of leaks
and possible causes of leaks to aid in the specification of corrective
measures. The electrical leak location survey will be curtailed until
the cause of the leaks is determined and corrective measures are taken
by the Contractor. In the case of defective seaming, only patching the
leaks will not be a viable corrective action because additional leaks
will likely form when basin is put in service. If more than one leak
HEALTH & ENDANGERMENT 63
-------
per 2000 ft2 of surveyed area is found in either liner, an electrical leak
location survey of the liner will be performed at the expense of the
Contractor after the corrective actions are taken and the located leaks
are repaired.
The occurrence of greater than approximately one leak per 2000 ff
can indicate defective seaming process or procedures, defective liner
material, or ineffective liner material handling or protection measures.
In these cases, further electrical leak location surveying are not sensi-
ble because the questionable integrity of the installation. In these cases
corrective actions must be taken.
Prior experience indicates that detectable leaks are found in some
repaired leaks when they are tested using the electrical leak location
method. When significantly less than approximately one leak per 2000
ft2 of liner is found, rechecking the leaks with the electrical leak
location method is usually not necessary if the leak is sealed and then
a patch is seamed over the repair. The repair can then be tested using
a vacuum box. When more than approximately one leak per 2000 ff
are found, rechecking the seams and patches using the electrical leak
location method is warranted. The geomembrane installer is responsi-
ble for making the repairs.
Preparing the Basin for Survey
Electrical Paths Through the Liner
Contractor shall electrically insulate electrical conduction paths
through the liner. Such conduction paths can be caused by fill pipes,
drain pipes, batten anchors, penetration flanges, footings, pump lines,
pump wiring, instrumentation wiring, instrumentation conduits and
access ramps. Electrical paths can also be established through the liquid
in plastic pipes if the pipes connect to a grounded metal valve or metal
pipe. Contractor will provide any necessary rubber packers and/or in-
sulated coverings for this purpose. Properly supported temporary geo-
membrane material sealed over the electrical penetrations can also be
used.
The electrical leak location method locates leaks by detecting elec-
trical conduction paths through leaks in the liner. Any other electrical
conduction path which also makes a circuit through or around the liner
will give the same indication as a leak. The presence of such electrical
conduction paths does not preclude the application of the method.
However, if these paths can not be eliminated, isolated, or insulated,
they will be indicated as leaks and they may mask the signal from other
smaller leaks in their immediate vicinity. In addition, if the conduction
paths are substantially lower in resistance than the electrical paths
through the leaks, the amount of current flowing through the leaks may
be too small to allow the detection of small leaks.
Electrode in Leak Collection Zone
Contractor shall make the arrangements for placing a suitable metal
electrode in the leak collection zone prior to installing the primary liner.
The electrode shall be constructed of approximately 1 ft2 of stainless
steel sheet. The corners and edges of the electrode must be rounded
to prevent damage to the liner. In addition, the electrode shall be im-
bedded in the sand or wrapped with geotextile or geonet to further pro-
tect the liners. An insulated wire (16 AWG to 12 AWG) must be connected
between the electrode and a test terminal located at a convenient
accessible location near the basin. The connections must be insulated
with a suitable coating. The electrode shall be buried at a depth
approximately 2 in above the secondary liner near the lowest point of
the collection system.
Some alternative methods include cutting a slit in the liner a few feet
above the water level to allow the insertion of the metal electrode. The
Contractor shall be responsible for having slits cut for inserting the
electrode if necessary and repairing the slits. Where necessary and feasi-
ble, a rod-shaped electrode can be placed in a leak sampling pipe that
extends down into the leak collection sump. However, this last method
is usually not as effective as the other methods because of the danger
of getting the electrode stuck and the increased resistance of the water
in the pipe.
Flooding the Liner
Contractor shall flood the liner to the required depths with water con-
taining no hazardous or foul substances. A source of water will be
provided by the Company. Water disposal facilities will be provided
by the Company. Contractor will be responsible for pumping or other-
wise transferring the water. Contractor will be responsible for damage
to the subgrade or berm caused by water leakage and erosion, or
hydrostatic loading. Provisions must be provided and procedures shall
be followed by the Contractor to minimize the dynamic loading of the
liner and possible damage to the liner, leak collection system and/or
subgrade caused by the water stream or by a rapid change in the water
level. Prior to flooding, Contractor shall clean the basin of debris in-
cluding scraps of liner material, other construction materials and
unneeded sand bags.
The water is needed for the electrical leak location method. The
hydrostatic loading of the liner is also desirable for determining if leaks
will occur under realistic loading conditions.
The basin shall be filled with the water to the working depth. When
more than approximately 20 ft of the side slope is immersed, the manual
survey of the side slopes is conducted in stages. The Contractor shall
lower the water between each survey stage to allow no more than ap-
proximately 20 ft of the immersed side slope to be surveyed at a time.
If the water can not be lowered to the level required for the next stage
of the survey within 16 hr. Contractor shall pay Company for standby
time or additional reduced mobilization costs for the electrical leak
location survey contractor.
In some cases where the basin is large, or the discharge rate far the
water must be limited, standby time or additional reduced mobiliza-
tion costs are inevitable and should be planned and contracted for as
part of the contract with the electrical leak location contractor. In those
cases, the Contractor shall pay Company only for additional standby
time or additional mobilization costs due to delays caused by the Con-
tractor in excess of the planned amount.
The side slopes can be surveyed by raising or lowering the water level
in stages either before or after the bottom of the liner is surveyed. If
the side slopes are surveyed first, from the top down, the basin wUl
be filled with water to the working level prior to the leak location sur-
vey. This exposes the liner to loads representative of actual in-service
loading. In most cases the level of the water can be lowered foster than
it can be wised, therefore, the survey can be completed with less standby
time required while the water level is adjusted.
The advantage of surveying the side slopes after the bottom of the
liner is surveyed is that washout or settling of the subgrade under the
liner caused by leaks in the bottom of the liner might be avoided if leaks
in the bottom are located and repaired prior to full hydrostatic loading.
However, there is no assurance that additional leaks will not occur
because of the increased hydrostatic loading during the side slope survey.
Therefore, additional surveying of the bottom of the liner may be required
after the side slopes are surveyed.
After the side slopes have been surveyed to the toe of the berm at
the most shallow part, the Contractor shall lower the water to the level
where the most shallow portion of the bottom of the basin is covered
with approximately 6 in of water. When the bottom of the liner slopes
more than 30 in, the survey of the bottom shall be conducted in more
than one stage. The Contractor shall lower the water between each survey
stage to allow the bottom of the basin to be surveyed in no more than
30 in of water. The water level is lowered to the level where the most
shallow unsurveyed area is covered with 6 in of water. If the water can
not be lowered to the level required for the next stage of the survey
within 16 hr, Contractor shall pay Company for standby time or addi-
tional reduced mobilization costs for the electrical leak location survey
contractor.
Again, for the cases where the basin is large, or the discharge mte
for the water must be limited, standby time or additional reduced mobili-
zation costs should be planned and contracted for as part of the con-
tract with the electrical leak location contractor. In those cases, the
Contractor shall be liable for paying only for additional standby time
or additional mobilization costs due to delays caused by the Contrac-
tor in excess of the planned amount.
Flooding the Leak Collection Zone for the Survey of
64 HEALTH & ENDAflGERMENT
-------
The Primary Liner
Contractor shall also flood the leak collection zone with water. This
can be done by pumping water in the leak collection system while the
primary liner is being filled with water. To avoid possible damage, the
water level in the leak collection zone must be maintained below the
level of the water in the primary liner to prevent the primary liner from
being lifted. Water can be pumped into the discharge side of the leak
collection system. Air vents must be provided in the perimeter edges
of the primary liner near the top of the berm to allow air trapped between
the two liners to be vented. The water can also be pumped into the
air vents. The Contractor shall be responsible for having slits cut for
flooding and air vents, if necessary and repairing the slits.
To survey the primary liner, an electrical conduction path through
any leak to the leak collection zone must be established. This task is
usually accomplished by flooding the leak collection zone. In some cases
when sand is used in the leak collection zone, an alternative method
can be used to establish the electrical conduction path. The reliability
of this alternative method depends on the type and moisture content
of the sand. The alternative method is to allow the waterfront the leaks
to percolate through the leak collection zone. This method is most ef-
fective when the sand has residual moisture and the water on top of
the liner has been allowed to stand at least three days and good elec-
trical contact can be established with the power supply electrode in
the leak collection zone.
The survey of the secondary liner must be conducted prior to instal-
lation of the primary liner. However, because the secondary liner is
in direct contact with earth ground there is no requirement to flood
the subgrade under the liner.
Electrical Power
Contractor will furnish a source of electrical power of 110-120 V AC
at 10 amp for the electrical leak location equipment. The power outlet
shall be located at the top of the berm.
Safety
Proper safety precautions and safe working practices shall followed.
A written safety plan specifically addressing the electrical leak loca-
tion surveys submitted by the electrical leak location contractor shall
be followed. Contractor will also inform the electrical leak location
survey subcontractor of the specific safety rules, procedures and haz-
ards at the plant site.
Electrical Leak Location Survey Specifications
For Electrical Leak Location Contractor
Electrical Leak Location Survey Under Hydrostatic Load
An electrical leak location survey will be performed by Southwest
Research Institute, 6220 Culebra Road, San Antonio, Texas 78238, (Con-
tact Daren L. Laine, telephone 512-522-3274) or approved equivalent.
The survey will be conducted on the bottom and side slopes of both
the primary and the secondary geomembrane liners of the basin. Con-
tractor will be responsible for preparing the basin for the survey as
described below.
The survey equipment leak detection distance shall be verified prior
to the survey. The results of the verification tests shall be used to
determine the distance between survey scans. The verification test will
be conducted using a simulated leak assembly as shown in Figure 1.
The simulated leak consists of a sealed plastic container with an insu-
lated wire penetrating the container through a sealed hole in the con-
tainer. The insulation at the end of the wire is stripped off for a distance
of approximately 1 in. The opposite end of the wire is connected to
a grounded electrode or a separate electrode in the leak collection zone.
A weight is placed in the container and the container is filled with a
sample of water from the basin being tested. A sample of geomem-
brane liner with the same thickness as the liner being tested is sealed
behind a large hole in the lid of the container. A 0.03 in nominal diameter
circular leak is placed in the center of the geomembrane sample by
penetrating the liner with a heated No. 6 sewing needle (0.030 in nominal
diameter) or a sewing pin (0.034 in nominal diameter).
The simulated leak assembly will be placed in the water in the basin
and survey sweeps will be made as the operator approaches the simu-
lated leak. The distance from the leak locator probe to the leak when
the leak is just detectable is measured. This is the leak detection dis-
tance. Twice this distance will be the maximum distance between survey
scans. The power supply electrode can be put at any position in the
basin, but the survey must be conducted with the power supply elec-
trode no farther from the leak than the distance when the verification
test was conducted.
The leak location sensitivity is proportional to the resistivity of the
water used to flood the liner and the power supply voltage. For rela-
tively high resistivity water such as river or lake water, or water from
a municipal supply, the simulated leak can be usually be detected at
a distance of approximately 18 in. For a saturated brine solution, the
simulated leak can usually be detected from a distance of 6 in. Smaller
leaks will be detected if the leak location probe electrode happens to
pass directly over the leak. Larger leaks can be detected from greater
distances. However, these typical leak detection sensitivities can be
greatly reduced in some instances and some judgement is necessary
for specifying an effective survey for a reasonable cost.
If more than one leak per 2000 ft2 of surveyed area is found in
either liner, the leak location survey will be limited to one man-day
of survey per 20,000 ft2 of liner material. The electrical leak location
survey will be conducted to better categorize the occurrence of leaks
and possible causes of leaks to aid in the specification of corrective
measures. The electrical leak location survey will be curtailed until
the cause of the leaks is determined and corrective measures are taken
by the Contractor. In the case of defective seaming, only patching the
leaks will not be a viable corrective action because additional leaks
will likely form when basin is put in service. If more than one leak
per 2000 ft2 of surveyed area is found in either liner, an electrical leak
location survey of the liner will be performed at the expense of the
Contractor after the corrective actions are taken and the located leaks
are repaired.
The occurrence of greater than approximately one leak per 2000 ft2
can indicate defective seaming process or procedures, defective liner
material or ineffective liner material handling or protection measures.
In these cases, further electrical leak location surveying is not sensi-
ble because the questionable integrity of the installation. In these cases
corrective actions must be taken.
Prior experience indicates that detectable leaks are found in some
repaired leaks when they are tested using the electrical leak location
method. When significantly less than approximately one leak per 2000
ff of liner is found, rechecking the leaks with the electrical leak lo-
cation method is usually not necessary if the leak is sealed and then
a patch is seamed over the repair. The repair can then be tested using
a vacuum box. When more than approximately one leak per 2000 ff
is found, rechecking the seams and patches using the electrical leak
location method is warranted. The geomembrane installer is responsi-
ble for making the repairs.
Preparing The Basin For Survey
The Company is responsible for having the basin prepared for the
electrical leak location survey. These preparations include: electrically
isolating electrical conduction paths; placing a suitable metal electrode
in the leak collection zone prior to installing the primary liner; cleaning
the basin of debris; flooding the liner to the required depths with water;
adjusting the level of the water as necessary; flooding the leak collec-
tion zone with water; and furnishing a source of electrical power.
Leak Location Surveys
Electrical Leak Location Survey of Sidewalls of the
Secondary and Primary Geomembrane Liners of the Basin
The electrical leak location survey contractor shall conduct a leak
location survey of the side slopes of the secondary liner and the primary
liner using the electrical leak location method. The side slope area will
be surveyed by systematically scanning the side slopes. Procedures shall
be followed to assure that the leak detection probe is scanned within
the detection distance for every point on the submerged liner. Twice
the leak detection distance is the maximum distance between survey
HEALTH & ENDANGERMENT 65
-------
GEOMEMBRANE
MATERIAL
UNDER TEST
0.030 INCH
DIAMETER HOLE
WATER FROM
BASIN UNDER
TEST
WEIGHT
PLASTIC LID
PLASTIC
CONTAINER
STRIP WIRE
ONE INCH
INSULATED WIRE
SEALED
CONNECT TO
GROUNDED
ELECTRODE
Figure 1
Simulated Leak Assembly
scans. In addition, all of the seams oriented down the side slopes shall
be surveyed individually by scanning the leak location probe along the
seam.
When more than approximately 20 ft of the side slope is immersed,
the water must be lowered in stages to allow the manual survey of the
side slopes. Any leaks found will be accurately located and the loca-
tions will be referenced to reference marks on liner near the berm of
the basin.
Electrical Leak Location Survey of Bottom of the
Secondary and Primary Geomembrane Liners Basin
The electrical leak location contractor shall conduct a leak location
survey of the bottom of the secondary liner and primary liner using
the electrical leak location method. Procedures shall be followed to
assure that the leak detection probe is scanned within the leak detec-
tion distance of every submerged point on the liner. In addition, all
of the seams shall be surveyed individually by scanning the leak location
probe along the seam.
Detected leaks shall be located to within 0.5 in or less and imme-
diately marked with lead sinkers and floats. The location of the leaks
shall also be measured relative to reference marks on the berm or side
slope of the liner for a permanent record. Where practical, the loca-
tion and type of leak shall be noted (i.e. on a seam or patch, or in the
panel).
Reports, Safety And Other Points
Reports
If requested, the general results of the electrical leak location survey
shall be reported to the designated representative of the Company during
the daily progress of the field work. A list of the locations of the leaks
found shall be submitted to the designated representative of the Com-
pany after completion of the field work and before the electrical leak
location survey crew leaves the site. A letter report documenting the
work, including a brief summary of the survey procedures, results of
the survey and problems encountered shall be prepared and submitted
within 14 days after completion of the field work.
Safety
Proper safety precautions and safe working practices shall be fol-
lowed. A written safety plan specifically addressing the electrical leak
location surveys shall be submitted to the Company for approval by
the electrical leak location contractor prior to the start of the leak loca-
tion field work. The safety plan shah1 be followed. Contractor and Com-
pany will inform the electrical leak location survey subcontractor of
the specific safety rules, procedures and hazards at the plant site.
Confidentiality
Unless agreed to in writing, the name of the facility, the location of
the facility, the identity of the Company, Contractor and the geomem-
brane installer shall be held in strict confidence. Any published rcsulB
of the survey will include only leak statistics. Information shall not be
afforded confidentiality if: such information is publicly available or
rightly obtained without restriction by from a third party; or released
without restriction by the furnishing party to anyone, including the Unit-
ed States Government.
Some facility owners prefer to avoid publicity concerning their opera-
tions. A confidentiality agreement should describe the level of security
desired.
66 HEALTH & ENDANGERMENT
-------
Evaluation of Relative Magnitude of Human Exposure by
Various Routes in a Community with
Multiple PCB-Contaminated Sites
John C. Kissel
Indiana University
Bloomington, Indiana
ABSTRACT
Efficient execution of Superfund activities intended to reduce risks
to humans posed by toxic wastes requires identification of the most sig-
nificant routes of exposure in a given location. Estimates are presented
of the magnitudes of PCB exposure associated with various pathways
in a community with multiple sources of contamination. For many
citizens, designated hazardous waste sites are less significant sources
of exposure to notorious chemicals than are more familiar surroundings.
Integration of hazardous waste management strategies with broader
environmental policies is therefore warranted. Examination of the rela-
tive magnitudes of exposure attributable to diverse routes also facili-
tates realistic assessment of the benefits of incremental cleanup actions.
Substantial mitigation of risk may occur long before formal comple-
tion of site remediation.
INTRODUCTION
Bloomington, Indiana was formerly the location of a capacitor
manufacturing and repair facility operated by Westinghouse Electric
Corporation. Polychlorinated biphenyls (PCBs) were released from the
plant in sewer and air discharges as a result of disposal of retired and
defective capacitors. Discarded capacitors were hauled to several dumps
and landfills. Copper scavengers opened the capacitors and spilled their
contents. In some cases, capacitors were transported to additional
locations before or after scavenging. Discharges from the manufacturing
facility to a city sewer resulted in contamination of a wastewater treat-
ment plant, and contaminated sludge was unknowingly distributed to
citizens as soil conditioner. Westinghouse employees, their families,
copper scavengers, sewage treatment plant employees, sludge users,
persons who frequented the dumpsites, persons who lived in close
proximity to the dumpsites, and the general citizenry have experienced
variable levels of exposure.
The U.S. EPA, Westinghouse, the City of Bloomington, Monroe
County and the State of Indiana agreed in 1985 to a cleanup strategy
involving six sites in or near Bloomington. Four of these sites are NPL.
Remedial measures have been taken at all six sites, but at only one is
no further action anticipated. Two additional sites are undergoing
cleanup outside the terms of the 1985 agreement.
The process of identifying and cleaning sites has become drawn out,
politicized and contentious. To evaluate risks and cleanup strategies,
it is useful to estimate the relative magnitudes of human exposure to
PCBs by various routes in Bloomington. Exposures may be estimated
directly from measured environmental concentrations for those routes
for which such data are available, or back-calculated from observed
body burdens using a pharmacokinetic model. Both methods are utilized
here.
Prior to discussing the presence of PCBs in the environ,ment, two
qualifications must be stated. First, analytical techniques have evolved
concurrently with concern over PCBs. Consequently, data from dis-
parate sources are not always precisely comparable. Second, commer-
cial PCB preparations are mixtures of compounds with variable
physico-chemical and lexicological properties." Ideally, evaluation of
exposure to and risks of PCBs would be approached on a congener-
specific basis. Since most historical data are not congener-specific,
however, total PCB burden serves as an imperfect surrogate measure.
Total PCB trends nevertheless present an illustration of the consequences
of widespread utilization and subsequent abandonment of a particular
class of poorly degraded, lipophilic chemicals.
BACKGROUND EXPOSURE
Exposures to PCBs in Bloomington are of particular interest to the
extent they deviate from exposures typically experienced by the national
population. PCBs were used extensively in a variety of products for
several decades prior to their removal from commerce in the latter half
of the 1970s. As a result, they are widely dispersed in the environment
and routinely identified in human tissue and blood. As shown in
Figure 1, domestic sales of PCBs peaked in 1970;4 sales of PCBs and
were ultimately banned in mid 1979. Since PCBs are lipophilic and rela-
tively resistant to degradation, their appearance in the food chain was
predictable.
Trends in adult dietary exposures estimated from information gathered
in FDA total diet surveys5"8 are presented in Figure 2. Horizontal
scales in Figures 1 and 2 are equivalent to facilitate comparison. An
estimate of a total dietary exposure of about 90 ng/day in 1985 in
Ontario9 is in good agreement with the FDA data. Dietary exposure
in Osaka, Japan,10 however, was estimated as greater than 4 ug/day in
1985. Data from the National Human Adipose Tissue Survey
(NHATS)11 presented in Figure 3 reveal that virtually all United States
residents carried detectable levels of PCBs in the early 1980s. The impact
of removing PCBs from routine commerce is reflected in a decline after
1978 in the fraction of the population having greater than 3 ppm in
adipose tissue and in the increase in those having levels, although
detectable, less than 1 ppm.
Measured blood concentrations12'15 of PCBs in general populations
or groups not known to have occupational exposure are presented in
Figure 4. The suggestion of a peak in blood levels in the mid 1970s
and a subsequent decline is consistent with the NHATS data, but the
scarcity of pre-1978 blood data precludes a firm conclusion to that effect.
Not shown in Figure 4 are (off-scale) blood levels measured in several
non-occupationally exposed populations exhibiting unusual rates of fish
consumption.14
The highest point in Figure 4 represents Bloomington sludge users
and may be an indication of increased exposure or simply an artifact
of the timing of the sampling. An ostensible Bloomington control group
HEALTH & ENDANGERMENT 67
-------
40 1
~ 30-
co
LU
CO
o
I—
CO
LU
2
O
Q
03
TJ
^_
LU
V
DC
LU
Q
O
Q.
O
CL
LL
O
LU
O
o:
LU
Q_
CO
CO
1965
1970
1975 1980
YEAR
1985
1990
Figure 1.
Annual Sales of PCBs in the United Slates
(Data from Reference 4).
6-
4 -
1965
1970
1975 1980
YEAR
1985
1990
100
80-
60-
40
20
Figure 1.
Estimated Adult (70 kg) Dietary Exposure Based on
FDA Figures (References 5-8).
•DL
ppm
1 ppm, <3 ppm
>3 ppm
1965
1970
1975
1980
1985
1990
YEAR
Figure 3.
Distribution of Adipose Tissue PCB Concentrations
In the US. Population (Reference 11).
DL Denotes Detection Limit.
n
a.
Q.
m
o
Q.
Q
O
o
_l
CD
20 -|
15-
10-
5
1965
1970
1975 1980
YEAR
1985
1990
Figure 4
Reported PCB Levels in Blood in General Populations or
Non-occupationally Exposed Control Groups.
All points are Arithmetic Means Except as Noted.
Squares: serum. Reference 13; Diamonds: plasma. Reference 13;
Circles: serum. Reference M; Triangles: serum medians.
Michigan controls (light fish consumers). Reference M;
Pluses: serum. Bloomington sludge-users (1977) and
serum geometric mean, Bloomington controls (1984). Reference 15.
found to have higher serum levels than the sludge users" contained
some health workers who may have been exposed during site
inspections'6 and therefore the data for this group are of questionable
validity for comparison. The slope between the sludge users and the
subsequent (1984) Bloomington controls" is similar to that between
Michigan controls flight fish eaters) sampled in 1973 and 1980.". 1984
Bloomington controls do not appear to have unusual blood levels of
PCB. Other subsets of the Bloomington population surveyed in 1984
had blood levels that ranged from apparently slightly elevated to clearly
elevated.15
Persons in classifications entitled game eating closest residence,
playing, digging, fish eating, and swimming had geometric mean serum
levels 31 to 54% higher than the 5.9 ug/L geometric mean of the controls.
Scavengers had a geometric mean slightly over twice that of the con-
trols and the occupationally exposed group (including Westinghouse
employees and wastewater treatment plant workers) had a level over
four times higher. Only the latter group was statistically distinguish-
able from the controls at the 5% level, but groups were small. Pre-
viously,12 occupationally-exposed persons were found to have an
arithmetic mean serum concentration over four times that of the sludge
users, and their family members' mean was about double that of the
sludge users. Results from more recent sampling of a larger number
of Bloomington residents than were tested in 1984 are not yet
available.*
EXPOSURES CALCULATED FROM
KNOWN CONCENTRATIONS
Environmental measurements from which PCB exposures may be
calculated directly are available from a variety of sources. Estimates
obtained in this manner are presented in Table 1. Commercial food
supply exposures were derived from the same sources as Figure 2.
68 HEALTH & ENDANGERMENT
-------
Table 1
Annual Average Daily Adult PCB Exposures Calculated
From Measured Environmental Concentrations.
Exposed individual
Magnitude Data Ref.
Westinghouse
Bloomington employee
Lake Michigan
sportfisher
Bloomington wastewater
plant operator
U.S. citizenry
U.S. citizenry
Nearest Bloomington
dumpsite neighbor
Occasional Bloomington
workplace air 18-740. 1977 (18)
fish 39-313. 1973-4 (19)
personal air <36. 1976 (17)
commercial food 6.9 1971 (8)
indoor air 0.65-10.9 1979-84 (20-22)
outdoor air
<2.7 1983 (24)
dumpsite visitor
U.S. citizenry
Bloomington residents
near dumpsites
Bloomington citizenry
Bloomington citizenry
outdoor air
commercial food
well water
City water
outdoor air
<2.3
<0 . 1
<0.024
<0.01
<0.008
1983
1985
1986
1989
1986-8
(24)
(8)
(25)
(26)
(27)
Workplace air measurements were made by NIOSH17'18 at a (now
closed) municipal wastewater treatment plant and in the Westinghouse
plant. The range of inhalation exposures to Westinghouse employees
reflects time-weighted average air concentrations associated with the
most (capacitor repairman) and least exposed (boilerhouse operator)
job classifications investigated.
It was assumed the employees worked 250 8-hr work days/yr and
had a 20 nrYday breathing rate. Inhalation exposure for wastewater
plant operators was calculated similarly based on an average of personal
air samples taken in September of 1976. (Results from a previous set
of samples taken in August, 1976, were all below detection, hence the
"less man" designation.)
Dietary exposures for Lake Michigan sportfishers (consumers of more
than 24 to 26 Ib of fish/yr) were taken from the literature" and includ-
ed for comparative purposes. Indoor air exposures to the general pub-
lic were calculated based on measured average air concentrations of
39 to 653 ng/m3 in homes, schools, laboratories and offices21"2 and the
assumption that a typical person is indoors 20 hours per day. A more
recent study23 found no PCBs in residential air, but at a detection limit
(100 ng/m3) that does not exclude the possibility of agreement with
previous results.
Estimated exposure to nearest dumpsite neighbors was based on the
highest 24-hr average concentration measured24 at the boundary of the
Lemon Lane site (before capping) and an assumption of 4 hr/day of
exposure. A somewhat higher boundary concentration was recorded
at another site, but measurements made near adjacent homes were lower.
Inhalation exposure attributable to occasional dumpsite visitation was
based on an assumption of 25 2-hr trips per year and the highest average
summertime daytime concentration recorded (also before interim
remediation at Lemon Lane) at 180 cm in vertical profile meas-
urements.24
The range of possible activities engaged in by persons visiting the
dumpsites is quite broad, and the estimate could be low for a few persons
such as copper scavengers who spent significant time near "hot spots"
(capacitor piles). Estimated exposure from well water was based on
a survey of water quality in wells within 1 mi of major dumpsites25
and a nominal consumption rate of 2 L/day. In most wells tested, PCBs
were not detected. A handful of positive values between 2 and 12 ng/L
were recorded. Two wells that tested higher were no longer in use for
drinking water supply. The city water exposure estimate reflects no
detection of PCBs at 5 ng/L in municipal water.26 Outdoor air inha-
lation exposure to the general Bloomington populace was based on the
highest annual average concentration obtained from three sites monitored
during 1986-1988.27
EXPOSURES BACK-CALCULATED FROM BODY BURDENS
The list of exposures presented in Table 1 clearly is not exhaustive
and is limited to those routes for which estimation is easily undertaken.
For example, potential dermal exposure from contaminated water or
soil or direct contact with PCB oils is not included. In the absence of
adequate knowledge of the PCB concentration in a particular medium
or of the frequency of the pertinent activity, gross exposure may be
back-calculated from measured body burdens given some understanding
of the rate at which PCBs are eliminated from the body.
Half-lives of various PCB congeners and commercial mixtures
reported in or derived from the literature28'32 are presented in Table 2.
Apparent half-lives, calculated from sequential data without consi-
deration of continuing exposure, may be much higher than true half lives.
Table 2
Half-lives of PCBs in Humans Reported in or
Calculated9 from the Literature.
Commercial mixture
or congener"
105
118
Kanechlor 300C
Kanechlor 300 & 500C
Aroclor 1242d
Aroclor 1260d
108/118
138
153
180
Aroclor 1242e
Aroclor 1254e
Half life
(yrs)
0.56
0.82
5.1
>15.
2.0
16.9
0.27-0.82
0.88
0.93
0.34
2.4-3.1f
2.6-6.5f
Sample size
17
20
4
5
1
ii
M
11
58
n
Reference
(28)
(29)
(30)
(31)
(32)
"Assuming no continuing exposure.
"Using numbering system of Ballschmiter and Zell (33).
"rBased on employment history.
^Distinction not specified.
Distinguished as eluting before (1242) or after (1254) DDE.
^Mean values for persons with highest to lowest initial body
burdens.
Rates of elimination are least likely to be distorted by background
exposure in persons with high existing burdens. Increasing half-lives
at lower tissue concentrations of mixtures of compounds such as PCBs
may also result from preferential retention of the least rapidly elimi-
nated congeners. Half-lives of specific congeners presented in Table 2
are relatively short. These data may reflect selection of atypical
congeners, but the particular congeners evaluated are among those which
routinely are found in human samples. The frequency with which they
are identified can be explained partly by their occurrence in commer-
cial mixtures, but, nevertheless, if elimination is rapid, ongoing exposure
must be high to maintain measurable levels in blood and tissue.
Buhler, et al.,32 suggest that typical exposures to each of the in-
dividual congeners they investigated are on the order of 3 to 4 /*/day.
Given the fractional presence of! individual congeners in commercial
PCB preparations, this figure is difficult to reconcile with estimates
of likely total PCB exposure presented here.
It has been demonstrated that elimination of 2,3,7,8-TCDD from
humans can be plausibly simulated using a physiologically based phar-
macokinetic (PBPK) model employing assumptions of simple thermo-
dynamically based partitioning and negligible metabolism.34 With
input of appropriate physical parameters, this model may be applied
to elimination of PCBs. Log octanol water partition coefficients of PCB
congeners of interest range from roughly 4 to 7.5.35'36 Henry's cons-
tants are likely to range from approximately 1 to over 100 Pa m3 mol~'
at physiological temperatures.37'38
PBPK model simulations indicate that (70 kg) adult half-lives attributa-
HEALTH & ENDANGERMENT 69
-------
ble to partitioning alone (excretion/exhalation) should vary between
roughly 0.5 and 5 yr given these physical properties. Shorter half-lives
would result for those congeners that are significantly metabolized.
Longer apparent half-lives could be observed in the presence of con-
current exposure. This effect is illustrated in the data of Phillips, et
al.,32 cited in Table 2. Aggregate PCB half-lives on the order of 2.5 yr
were observed in persons with relatively high body burdens, and longer
half-lives were observed in persons with relatively low body burdens.
Given an appropriate metabolic rate constant, the PBPK model may
be used to compute exposures required to sustain observed body bur-
dens. Selected cases are presented in Table 3. A true half-life of 2.5
yr was assumed.
Some Westinghouse employees had serum PCB levels over 1000 mg/L
in 1977.32 The maximum inhalation exposure shown in Table 1 appears
sufficient to account for only about 500 mg/L in serum. Dermal
exposure is therefore likely to have been very substantial and to have
exceeded inhalation exposure to at least some employees. A similar
conclusion with respect to another group of PCB workers was reached
previously by Lees, et al.,4' The mean serum PCB level in a group of
Westinghouse employees' family members in 1977 was reported as
approximately 34 mg/L as compared to the sludge users' 17 mg/L.12
If the sludge users received negligible non-background exposure, then
about half of the family members' body burden could be attributed to
unusual exposure. If roughly half of the sludge users' burden was the
result of non-background exposures, then about three-quarters of the
family members' burden was unusual. Dermal uptake of a lipophilic
contaminant from soil has been shown to be plausible elsewhere.42
Further application of the PBPK model reveals that likely current
exposures to the general population in the United States are not suffi-
cient to maintain existing serum levels and that continuing decline should
be anticipated.
Table3
Annual Average Daily Supplemental Exposures Estimated to be
Required to Produce Observed Serum Levels.
Exposed individual
Magnitude
(ug/day)
Most exposed
Westinghouse employee dermal
Family member of
Westinghouse employee dermal, inhalation
Sewage sludge user
dermal, inhalation
>740.
25-35.
0-10.
'Assuming aggregate 2.5 year half life.
DISCUSSION
Background exposures to PCBs are declining in the United States.
Given the trend in PCB levels in commercial foods, indoor air may
now be the primary source of PCB exposure for the bulk of the popu-
lation. Temporal trends in indoor air concentrations are poorly defined,
however, and further research is needed in this area.
Indoor air concentrations measured in the United States in the early
1980s are comparable to outdoor air levels measured in the immediate
vicinity of uncontrolled dumpsites in Bloomington at about the same
time. Indoor air exposures certainly impact a greater portion of the
population.
Results from the Total Exposure Assessment Methodology Study
(TEAMS)39 demonstrate that primary exposures to some pollutants of
concern, in particular volatile organics, probably occur indoors. This
also may be the case for the semi-volatile PCBs. Reorientation of U.S.
EPA activities toward a more integrated and consistent assault on en-
vironmental problems, as has been suggested by an internal U.S. EPA
review panel,40 is warranted.
At issue in Bloomington are the adequacy of cleanup efforts under-
way and the risks presented. Despite the fact that the ultimate cleanup
strategy is controversial and a decade or more from completion, the
largest risks listed in Table 1 have been effectively mitigated. Clearly
those at greatest risk of harm by PCBs in Bloomington are persons who
were employed at Westinghouse during the period PCBs were actively
utilized. Cessation of active utilization of PCBs has resulted in a sub-
stantial decline in serum PCB levels in such persons30*11 and also is
likely to have resulted in reduction of exposures to members of their
families. Exposures of the latter type are not adequately addressed within
the current regulatory framework, a situation that reiterates the need
for reorientation of efforts at the U.S. EPA.
Former occupational exposures and any resulting from residual con-
tamination inside the Westinghouse plant fall within the realm of
occupational safety and outside the scope of CERCLA cleanup activi-
ties. Although PCB body burdens of Westinghouse employees con-
tributed to the perception of a hazardous waste problem in Bloomington,
they did not result from waste disposal practice per se and would not
be more effectively remedied if CERCLA-related activities in Blooming-
ton were proceeding more smoothly. These exposures have been mitigat-
ed by eliminating the activity resulting in waste generation rather than
by more effective management of the waste. Occupational exposures
to wastewater treatment plant personnel were eliminated by closing the
facility (as was previously necessitated by city growth). Simple interim
measures (fencing) eliminated routine access to the major dumpsites
and associated exposures.
Remaining concerns, in addition to off-site air transport, include trans-
port in groundwater and access to unremediated sites. Transport in
groundwaler has not yet presented a significant problem in Blooming-
ton, although, in view of local geological characteristics, movement
is inevitable barring complete remediation. Groundwater contamina-
tion, while very expensive to reverse, generally represents potential
rather than immediate risk. The slowness with which groundwater moves
prevents natural flushing from being a viable management strategy, but
it also provides time for implementation of interim mitigation strate-
gies such as provision of alternative water supplies to persons at risk.
Most of the minor sites in Bloomington at which no remediation is
planned are small plots on which contaminated sludge was spread.
Attenuation of PCBs by volatilization and perhaps biodegradation
appears to be occurring at significant rates.43
The exposure estimates presented in Table 1 may be compared to
health criteria. ACGffl lists TLVTWAs of 0.5 mg/m3 for Aroclor 1254
and 1.0 mg/m3 for Aroclor 1242.*4 Assuming 250 8-hr work days per
year, this figure corresponds to acceptable average annual daily
exposures to workers of roughly 2300 and 4600 ug/day, respectively.
On the basis of potential carcinogenic ity, however, NIOSH recommends
a 500 to 1000-fold lower 1 n/m3 standard for each Aroclor.43
The U.S. EPA recently proposed a drinking water MCL of 0.5 yA
for PCBs. This figure corresponds to an approximate acceptable intake
of 1 /i/day and, assuming a carcinogenic potency factor of 7.7 mg~' kg
day,* an excess lifetime cancer risk on the order of K)-4. Interestingly,
the City of Bloomington's NPDES permit for the Dillman Road
wastewater treatment plant requires effluent PCBs to be less than 0.1
/j/1, five times lower than the proposed drinking water MCL.
CONCLUSIONS
Perceived sources of significant exposure to pollutants may differ from
actual sources. External sources, especially identified hazardous waste
sites, are greatly feared. For PCBs (and some other industrial chemi-
cals), more familiar surroundings such as homes and offices appear
to present the greatest exposure to the average citizen under current
conditions in Bloomington and elsewhere. Policies governing cleanup
of hazardous waste sites should therefore be integrated in an overall,
multi-media environmental protection strategy.
Much attention is given to the fact that only a small proportion of
NPL sites have been declared fully remediated. The bulk of the risk
associated with such sites may, however, be eliminated well before
cleanup completion. Progress under CERCLA to date is likely to be
significantly underestimated if measured by cleanup completions alone.
Understanding of the elimination of PCBs from humans is incom-
plete and further congener-specific, human-based investigation is
needed. Nevertheless, existing United States background PCB exposure
70 HEALTH & END^NGERMENT
-------
appears insufficient to maintain typical body burdens. Further decline
therefore is likely.
REFERENCES
1. Safe, S., Bandiera, S., Sawyer, T., Robertson, L., Safe, L., Parkinson, A.,
Thomas, P., Ryan, D., Reik, L., Levin, W., Denomme, M. and Fujita, T.
"PCBs: Structure Function Relationships and Mechanism of Action,"
Environ. Health Perspec., 60 pp 47-56, 1985.
2. Jones, K., "Determination of Polychlorinated Biphenyls in Human Food-
stuffs and Tissues: Suggestions for a Selective Congener Analytical
Approach," Sci. Tot. Environ., 68 pp 141-159, 1988.
3. McFarland, V. and Clarke, J., "Environmental Occurrence, Abundance,
and Potential Toxicity of Polychlorinated Biphenyl Congeners: Considera-
tions for a Congener-Specific Analysis," Environ. Health Perspec., 81 pp
225-239, 1989.
4. Conway, R., Whitmore, F. and Hansen, W., "Entry of Chemicals into the
Environment," Chapter 3 of Environmental Risk Analysis for Chemicals,
R. Conway, ed., Van Nostrand Reinhold, New York, NY, 1982.
5. Gartrell, M., Craun, J., Podrebarac, D. and Gunderson, E, "Pesticides,
Selected Elements, and Other Chemicals in Adult Total Diet Samples, October
1978-September 1979," J. Assoc^ Off. Anal. Chem., 68(5) PP 862-875, 1985.
6. Gartrell, M., Craun, J., Podrebarac, D. and Gunderson, E., "Pesticides,
Selected Elements, and Other Chemicals in Adult Total Diet Samples, October
1980-March 1982," / Assoc. Off. Anal. Chem., 69(1) pp 146-161, 1986.
7. Gunderson, E., "FDA Total Diet Study, April 1982-April 1984, Dietary In-
takes of Pesticides, Selected Elements and Other Chemicals," J. Assoc. Off.
Anal. Chem., 71(6) pp 1200-1209, 1988.
8. Gunderson, E., FDA Division of Contaminants Chemistry, Washington, D.C.,
personal communication, Aug., 1989.
9. Davies, K. "Concentrations and Dietary Intake of Selected Organochlorines,
Including PCBs, PCDDs and PCDFs in Fresh Food Composites Grown in
Ontario, Canada," Chemosphere., 17(2) pp 263-276, 1988.
10. Matsumoto, H., Murakami, Y., Kuwabara, K., Tanaka, R. and Kashimo-
to, T., "Average Daily Intake of Pesticides and Polychlorinated Biphenyls
in Total Diet Samples in Osaka, Japan," Bull. Environ. Com. Toxicol., 38
pp 954-958, 1987.
11. U.S. EPA. NHATS Summary Statistics: 7'/369 through 9'/383 (Unbound),
Office of Toxic Substances, Washington, DC, 1986.
12. Baker, E., Landrigan, P., Glueck, C, Zack, M., Liddle, J., Burse, V., House-
worth, W. and Needham, L., "Metabolic Consequences of Exposure to Poly-
chlorinated Biphenyls (PCB) in Sewage Sludge," Am. J. Epid., 112(4) pp
553-563, 1980.
13. Sahl, J., Crocker, T., Gordon, R. and Faeder, E., "PolychlorinatedBiphenyl
Concentrations in the Blood Plasma of a Selected Sample of Non-
occupationally Exposed Southern California Working Adults," Sci. Total
Environ., 46 pp 9-18, 1985.
14. Kreiss, K., "Studies on Populations Exposed to Polychlorinated Biphenyls,"
Environ. Health Perspec., 60 pp 193-199, 1985.
15. Stehr-Green, P., Ross, D., Liddle, J., Welty, E. and Steele, G., "A Pilot-
Study of Serum Polychlorinated Biphenyl Levels in Persons at High Risk
of Exposure in Residential and Occupational Environments," Arch. Envi-
ron. Health, 41(4) pp 240-244, 1986.
16. Steele, G., Indiana State Board of Health, Indianapolis. Personal commu-
nication, Aug., 1989.
17. NIOSH, Health Hazard Evaluation Determination: Westinghouse Electric
Corporation, Bloomington, Indiana. Report No. 76-52-386, Cincinnati, OH,
1977.
18. NIOSH, Health Hazard Evaluation and Technical Assistance: Winston Tho-
mas Sewage Treatment Plant, Bloomington, Indiana. Report No. TA 76-84,
Cincinnati, OH, 1977.
19. Cordle, E, Locke, R. and Springer, J., "Determination of Risk in Regu-
lating Polychlorinated Biphenyls (PCBs) A Case Study," Chapter 10 of lex-
icological Risk Assessment: Volume II General criteria and Case Studies,
D. Clayson, D. Krewski and I. Munro, eds., CRC Press, Boca Raton, FL,
1985.
20. McLeod, K., "Polychlorinated Biphenyls in Indoor Air," Environ. Sci. Tech.,
15(8) pp 926-928, 1981.
21. Murphy, T., Formanski, L., Brownawell, B. and Meyer, J. "Polychlorinated
Biphenyl Emissions to the Atmosphere in the Great Lakes Region. Municipal
Landfills and Incinerators," Environ. Sci. Tech., 19(10) pp 942-946, 1985.
22. Oatman, L. and Roy, R., "Surface and Indoor Air Levels of Polychlorinated
Biphenyls in Public Buildings," Bull. Environ. Contam. Toxicol., 37 pp
461-466, 1986.
23. Lewis, R., Bond, A., Johnson, D. and Hsu, J. "Measurement of Atmospheric
Concentrations of Common Household Pesticides: A Pilot Study," Envi-
ron. Monitor. Assess., 10:59-73, 1988.
24. Lewis, R., B. Martin, D. Sgontz, and J. Howes. "Measurement of Fugitive
Atmospheric Emissions of Polychlorinated Biphenyls from Hazardous Waste
Landfills," Environ. Sci. Tech., 19(10) pp 986-991, 1985.
25. Hites, R., D. Liebl, W. Jones, and M. Troyer. Collection and Analysis of
Drinking Water Well Samples for PCB Content, SPEA, Indiana University,
1986.
26. City of Bloomington Utilities. Monthly staff reports. Bloomington, IN, 1989.
27. Hermanson, M. and Hites, R., "Long-term Measurements of Atmospheric
Polychlorinated Biphenyls in the Vicinity of Superfund Dumps," Environ.
Sci. Tech., (In press).
28. Chen, P., Luo, M., Wong, C. and Chen, C., "Comparative Rates of Elimi-
nation of Some Individual Polychlorinated Biphenyls from the Blood of PCB-
Poisoned Patients in Taiwan," Fd. Chem. Toxic., 20 pp 417-425, 1982.
29. Hara, I., "Health Status and PCBs in Blood of Workers Exposed to PCBs
and of Their Children," Environ. Health Perspec., 59 pp 85-90, 1985.
30. Steele, G., "Estimates of the Biologic Half-Life of Polychlorinated Biphenyls
in Human Serum," New Eng. J. Med., 314(14) pp 926-927, 1986.
31. Buhler, E, Schmid, P and Schlatter, CH., "Kinetics of PCB Elimination
in Man," Chemosphere., 17(9) pp 1717-1726, 1988.
32. Phillips, D., Smith, A., Burse, V., Steele, G., Needham, L. and Hannon,
W., "Half-Life of Polychlorinated Biphenyls in Occupationally Exposed
Workers," Submitted to Arch. Environ. Health.
33. Ballschmiter, K. and Zell, M., "Analysis of Polychlorinated Biphenyls (PCB)
by Glass Capillary Gas Chromatography," Fresenius Z. Anal. Chem., 302
pp 20-31, 1980.
34. Kissel, J. and Robarge, G., "Assessing the Elimination of 2,3,7,8-TCDD from
Humans with a Physiologically Based Pharmacokinetic Model,"
Chemosphere., 17(10) pp 2017-2027, 1988.
35. Miller, M., Wasik, S., Huang, G., Shlu, W. and Mackay, D. "Relation-
ships between Octanol-Water Partition Coefficient and Aqueous Solubility,"
Environ. Sci. Tech., 19(6) pp 522-529, 1985.
36. Hawker, D. and Connell, D., "Octanol-Water Partition Coefficients of Poly-
chlorinated Biphenyl Congeners," Environ. Sci. Tech., 22(4) pp 382-387,
1988.
37. Burkhard, L., Armstrong, D. and Andren, A., "Henry's Law Constants
for the Polychlorinated Biphenyls," Environ. Sci. Tech., 19(7) pp 590-596,
1985.
38. Dunnivant, F., Coates, J. and Elzerman, A., "Experimentally Determined
Henry's Law Constants for 17 Polychlorinated Biphenyl Congeners," Envi-
ron. Sci. Tech., 22(4) pp 448-453, 1988.
39. U.S. EPA, The Total Exposure Assessment Methodology (TEAM) Study: Sum-
mary and Analysis; Volume 1, EPA/600/6-87/002a, Office of Research and
Development, Washington, DC, 1987.
40. U.S. EPA. Unfinished Business: A Comparative Assessment of Environmental
Problems. Office of Policy Analysis. Washington, DC, 1987.
41. Lees, P., Corn, M. and Breysse, P., "Evidence for Dermal Absorption as
the Major Route of Body Entry During Exposure of Transformer Maintenance
and Repairmen to PCBs," Am. Ind. Hyg. Assoc. J., 48(3) pp 257-264, 1987.
42. Kissel, J. and McAvoy, D., "Reevaluation of the Dermal Bioavailability of
2,3,7,8-TCDD in Soil," Haz. Waste Haz. Mat., (In press).
43. Kissel, J., Bedan, M., Austin, P. and Hermanson, M., "Assessment of Long
Term Loss of PCBs from Soil Amended with Contaminated Sewage Sludge,"
accepted for presentation, 62nd annual Water Pollution Control Federation
conference, San Francisco, CA, 1989.
44. ACGIH, Threshold Limit Values and Biological Exposure Indices for
1987-1988, Cincinnati, OH, 1987.
45. NIOSH, Pocket Guide to Chemical Hazards. U.S. Government Printing
Office, Washington, DC, 1985.
46. U.S. EPA, "Drinking Water Criteria Document for Polychlorinated
Biphenyls," ECAO-CIN-414, Final, Cincinnati, OH, April 1988.
HEALTH & ENDANGERMENT 71
-------
Interim Methodology for Performing Petitioned Health Assessments
Gregory V. Ulirsch
Juan J. Reyes
Mark M. Bashor, Ph.D.
Office of Health Assessment
Agency for Toxic Substances and Disease Registry
Public Health Service
U.S. Department of Health and Human Services
Atlanta, Georgia
ABSTRACT
The Agency for Toxic Substances and Disease Registry may, under
CERCLA, as amended, and RCRA, as amended, perform a Health
Assessment for a facility or release in response to a petition. Such
petition may be offered by individuals (private citizens) or licensed phy-
sicians who supply information that individuals have been exposed to
hazardous substances. In response to this mandate, ATSDR has
developed an interim methodology for performing Petitioned Health
Assessments. This paper describes the methodology developed by
ATSDR for performing an assessment and will include ATSDR interim
procedures and current data on the status of Petitioned Health
Assessments.
INTRODUCTION
The Agency for Toxic Substances and Disease Registry (ATSDR) is
authorized under CERCLA, as amended by SARA, to perform various
Health Assessments. Specifically, the Agency may "perform a Health
Assessment for releases or facilities where individual persons or licensed
physicians provide information that individuals have been exposed to
a hazardous substance, for which the probable source of such exposure
is a release. In addition to other methods (formal and informal) of
providing such information, such individual persons or licensed
physicians may submit a petition to the Administrator of ATSDR
providing such information and requesting a Health Assessment."
In addition to CERCLA, RCRA, as amended, has a provision under
the Exposure Information and Health Assessment section stating that
"any member of the public may submit evidence of releases of or
exposure to hazardous constituents from a facility, or as to the risk or
health effects associated with such releases or exposure, to the Adminis-
trator of ATSDR." Petitions or evidence submitted as defined by the
above acts (i.e., CERCLA and RCRA) are considered Petitioned Health
Assessments. Because these laws, as they pertain to Petitioned Health
Assessments, are broadly defined, it was necessary for ATSDR to
develop an interim methodology for dealing with Petitioned Health
Assessments.
ATSDR recognizes that decisions to perform a Health Assessment
should be based on public health concerns. Determining public health
concerns is an "interpretive" process and such concerns cannot always
be identified from the information received with a petition. Gathering
additional information, analyzing it and thereby identifying the health
concerns is equivalent to performing the Petitioned Health Assessment.
Furthermore, as specified above in the CERCLA and RCRA legisla-
tion, threats to the public, other than those posed by chemical releases
or facilities, although they may be related to Petitioned Health Assess-
ments, may not be the responsibility of ATSDR.
Once a public health concern has been established, even though
ATSDR would like to respond to the needs of the public with a Health
72 HEALTH & ENDANGERMENT
Assessment, performing a Health Assessment on each and every incident
may not be in the best interest of the public. Other, more appropriate,
authorities might better address public health concerns that do not relate
to releases or facilities.
Furthermore, since the CERCLA legislation states that "if [such]
a petition is submitted and the Administrator of ATSDR does not initiate
a Health Assessment, the Administrator of ATSDR shall provide a
written explanation of why a Health Assessment is not appropriate."
Hence, formal procedures for accepting or rejecting petitions are
imperative. The following discussion will outline the interim
methodology that was developed within the ATSDR Office of Health
Assessment (OHA) for addressing Petitioned Health Assessments.
INTERIM METHODOLOGY
Gathering Preliminary Information and
Acknowledging Petitioner
Within a reasonable time period after receiving a petition (i.e., a target
of 10 working days), all appropriate ATSDR personnel provide any first-
hand information they might have on the facility or release. If a Health
Assessment has not been performed on a release or facility (see below),
an acknowledgement letter will be written responding to the specific
information provided by the petitioners) and incorporating any addi-
tional information provided by ATSDR personnel.
If ATSDR has already performed a Health Assessment, die
acknowledgement letter will reflect this and will include a copy of the
document. Furthermore, the acknowledgement letter will state that un-
less the petitioner has additional information not considered in the al-
ready completed Health Assessment, ATSDR will not pursue the petition
any further. If the petitioner sends new information in response to the
ATSDR acknowledgement letter, the request will be considered as a
new petition and dealt with as such.
Collecting Background Information
Once a petition has been acknowledged, background information must
be collected so that ATSDR can determine whether to accept or reject
the petition. As a first step in this interim process, the appropriate
ATSDR Regional Representative develops contacts and collects back-
ground information on the alleged release or facility. At a minimum,
the Regional Representative contacts:
• The Petitioned Health Assessment contact designated by the U.S.
EPA headquarters
• The most appropriate U.S. EPA personnel with knowledge or potential
knowledge of the site or release
• The most appropriate representatives of state health and environmental
agencies
• The most appropriate local health and environmental agencies
• The petitioner(s)
-------
ATSDR uses the information gathered by the Regional Representa-
tive to apply the first interim decision criteria &i.e., Interim Mode A
Decision Criteria (see below)]. In addition to gathering information
and developing contacts, the Regional Representative also obtains a
recommendation from each agency on whether the petition should be
accepted (see Mode A Decision Criteria, below). The information
gathered in this step is not intended as a basis for the Health Assess-
ment but only to provide information for accepting or rejecting the
petition.
Applying the Interim Mode A Decision Criteria
A committee (i.e., Screening Committee), consisting of applicable
ATSDR management personnel, reviews the information gathered by
the Regional Representative and applies the following Interim Mode
A Decision Criteria:
A-l. Has a Health Assessment or its equivalent already been performed
relative to the site, release or population?
if yes, forward the Health Assessment to the petitioner, stating
that ATSDR will take no further action unless the petitioner has
information not considered in the report or its equivalent;
if no, proceed with the evaluation;
A-2. Can a source or release of contaminants alleged by the petition
to exist be identified?
if yes, proceed with the evaluation;
if no, reject for this reason;
A-3. Can a target population exposed or potentially exposed in the past,
present or future alleged by the petition to exist be identified?
if yes, proceed with the evaluation;
if no, reject the petition for this reason;
A-4. Do any of the government agencies recommend that ATSDR accept
the petition?
if yes (one or more), proceed with the evaluation;
if no (all), and reasons appear credible to OHA, reject the peti-
tion for these reasons; in some cases OHA may still proceed
with an evaluation if it believes that there are public health
issues that have not yet been adequately addressed.
If a petition is rejected for any of the above reasons, then a letter
will be sent to the petitioner stating the reasons for this rejection (see
below-Mode B Decision Criteria).
Assigning Site to Scoping Team, Site-Visitation,
Data Collection, and Preparation of Site Summary Report
Once a decision has been made to proceed with the petition, a
member(s) of an appointed evaluation team (i.e., Scoping Team) visits
the site and meets with knowledgeable federal, state and local officials
and the ATSDR Regional Representative. In addition, the Scoping Team
member(s) contacts those individuals previously designated by the
Regional Representative and any other individuals knowledgeable about
the site that were not available for personal communication during the
site trip. Background information and monitoring data, site-visit infor-
mation and any other information or monitoring data collected by the
Scoping Team member(s) are evaluated and used to complete an ATSDR
Site Summary Form.
Presenting Preliminary Information to the
Screening Committee
When the information gathered during the site-visit trip has been con-
solidated and the site-visit report and ATSDR Site Summary Form have
been completed, the information is presented at a Screening Commit-
tee meeting for preliminary feedback. The purpose of this meeting is
to insure that all involved parties have a thorough understanding of the
petition, the release and the implications of the release before the OHA
Scoping Team member(s) meet with the petitioner(s).
Meeting with Petitioners and Preparing Trip Report
After preliminary presentation to the Screening Committee, the ap-
propriate OHA staff travel with the ATSDR Regional Representative
to meet with the petitioners). The trip report prepared after this meeting
becomes part of the official record, and any new information or health
concerns brought to the Agency's attention by the petitioner(s) is con-
sidered when the petition is reviewed under the Interim Mode B Deci-
sion Criteria (see below).
Formal Presentation to Screening Committee and
Applying Interim Mode B Decision Criteria
The Scoping Team member(s) assigned to evaluate the petition
presents the findings and all related information to the Screening
Committee. The Screening Committee will apply the Mode B Deci-
sion Criteria (below) to determine whether the petition is accepted or
rejected.
Mode B Decision Criteria:
B-l. Have individuals been exposed to a hazardous substance for
which the probable source of such exposure is a release?
5-2. Are the location, concentration and toxicity of the hazardous
substances involved significant?
B-3. Is there potential for further human exposure?
B-4. What is the strength of recommendations from other govern-
ment agencies?
5-5. Is the incident applicable to CERCLA or RCRA or to other
more appropriate environmental statutes (can the public best
be served by a more appropriate government agency)?
5-6 Are ATSDR resources available and what other ATSDR
priorities have bearing, such as its responsibilities to conduct
other Health Assessments and health effects studies?
The Above Mode B Decision Criteria require the use of professional
judgment to evaluate the criteria's bearing on the ultimate decision to
accept or reject the petition. After applying these criteria and reaching
a decision to accept or reject the petition, ATSDR drafts a response
letter to inform the petitioner(s) of the decision, the reasons for the
decision (if appropriate) and the nature of any followup action(s) (if
appropriate).
Preparing Draft Health Assessment
If the petition is accepted by applying the Interim Mode B Decision
Criteria, it is then assigned to an appropriate multi-disciplinary Health
Assessment team to develop a Draft Health Assessment. A copy of this
Draft Health Assessment is provided for comment to the U.S. EPA,
State and others in accordance with ATSDR policy.
Preparing Final Draft Health Assessment
Comments on the Draft Health Assessment received from the U.S.
EPA, the State and others are considered, and the document is revised
as necessary to prepare the Final Draft Health Assessment. Once the
Final Draft Health Assessment is completed, a public meeting is con-
ducted to discuss the findings.
Public and Petitioner Comment on
Final Draft Health Assessment
The Final Draft Health Assessment is released for public and
petitioner comment.
Responding to Comment, Preparing and
Distributing (final) Health Assessment and
"closing" the Petition File
All comments on the Final Draft Health Assessment received from
the public and petitioners is considered and the document is revised
as necessary to prepare the (final) Health Assessment. The Health
Assessment is then distributed according to ATSDR policy. It also is
sent to the petitioner with a letter of transmittal closing the petition
file (unless ATSDR is undertaking some followup health action, e.g.,
proceeding with a followup Health Assessment or a health study,
registry, surveillance activity, etc.).
STATUS OF PETITIONED HEALTH ASSESSMENTS
At the time this manuscript was prepared, ATSDR had received 62
requests for Petitioned Health Assessments from private citizens, public
HEALTH & ENDANGERMENT 73
-------
officials, physicians, lawyers and others (Table 1). Approximately 50%
of the requests were received from private citizens. Although physi-
cians are specifically mentioned in the CERCLA regulations, only two
of the 62 requests have come from physicians; however, both physi-
cians were also public officials. Either very few private physicians are
aware of the petition process or they may have elected to have their
patients (private citizens) file the petition requests.
Table 1
Profile of Health Assessment Petitioners
Type of Petitioner Number of Requests Percentage
Table 4
Profile of Petitions by EPA Region
Private Citizen
Public Official
Physician
Lawyer
Other''
Total
30
15
2
13
62
49
24
3
24
100
physicians that petitioned ATSDR were also public officials
2Tribal Council and military officer.
Requests have been received for sites on the NPL, for RCRA sites
and for "other" sites and facilities (Table 2). The site designation of
"other" consisted mostly of active and inactive commercial or indus-
trial facilities (Table 3). As shown in Table 3, some petitions received
by ATSDR do not precisely fall into the category of a release or facility,
as described in CERCLA and RCRA. A profile, by U.S. EPA Region,
of the petitions received is shown in Table 4.
Table 2
Profile of Petitioned Health Assessments by Site Type
Type
Number of
Requests
Percentage
National Prlorltltes List
(NPL)
21
31
Resource, Conservation, and
Recovery Act (RCRA) 11 16
Other
36 53
Notes:
1Inclui
eludes multiple site listings for a single petition received by
ATSDR.
Table 3
Profile of Non-NPL/RCRA Petitioned Health Assessments
By Site Type
Type of Site
Number of
Requests
Commercial/Manufacturing Facilities
Landfills/Abandoned Disposal Areas
Hulitple Source Sites
Contaminated Municipal/Private Water Supplies
Military Base
Federal Penetentlary
Municipal Incinerator
Mining Waste
Smoke from Burning of Timber
Agricultural Pesticide Release
Sewage Contamination of Waterways
Pesticide Test Ponds
Total
14
7
5
2
1
1
1
1
1
1
1
__1
36
At the time this manuscript was prepared, 60 of the 62 petitions
received by ATSDR were being processed in one of the phases of the
74 HEALTH & ENDANGERMENT
Region
1
2
3
4
5
6
7
8
9
10
Total
Number of Requests
5
9
8
15
11
5
2
0
6
1
62
Rank by Number
of Requests
6/7
3
4
1
2
6/7
8
10
5
9
interim methodology for performing Petitioned Health Assessments.
A delineation of the status of the petitions is shown in Table 5. Forty-
four of the 62 petitions received have had the evaluation process (i.e.,
scoping) completed and that 30 Health Assessments currently have been
completed or are in progress.
TaMeS
Status of Petitioned Health Assessments
Status Category
Nunber of Requests
Rejected 14
Assigned (HA in progress) 22
Scoping (In progress) 17
Withdrawn by Petitioner 1
HA Completed J
Total 62
HA- Health Assessment.
Scoping— Evaluating petition request.
CONCLUSIONS
ATSDR, in response to the broadly defined CERCLA and RCRA
legislation, has developed an interim methodology for performing Peti-
tioned Health Assessments. This interim methodology includes a pro-
cedure using two-mode decision criteria for accepting or rejecting
Petitioned Health Assessment requests. Because ATSDR has based the
decision criteria primarily on human health concerns, ATSDR believes
that these procedures are sound from both a legal and, most impor-
tantly, a public health perspective.
Some of the petitions received by ATSDR do not appear to be the
responsibility of ATSDR, as implied in the CERCLA and RCRA legis-
lation, nor would the public health concerns raised by these petitioners
be best served by ATSDR. ATSDR believes that the interim decision
criteria developed to accept or reject these types of requests will best
serve the public interest and ATSDR needs. Wherever applicable,
ATSDR will refer a petition to the appropriate federal, state and local
authorities for follow-up actions to protect public health.
With the experience ATSDR gains through the use of this interim
methodology for performing Petitioned Health Assessments and with
public comments that will be received when these criteria are published
in the Federal Register, ATSDR may modify this methodology in the
future to best serve the public interest and ATSDR needs. It is readily
apparent, however, that public health concerns will remain the primary
basis lor all ATSDR decisions whether to perform Health Assessments
or other appropriate actions at petitioned sites or facilities.
-------
Medical Surveillance/Biological Monitoring
Cleanup at the Rocky Mountain Arsenal
Rupert C. Burtan, M.D., M.P.H., D.P.H.
TRIDEM Services, Ltd.
Denver, Colorado
ABSTRACT
The problems facing the industrial hygienist/safety professionals at
hazardous waste sites during cleanup are numerous.
Those responsible for the safety and health of site personnel must
learn how to make combined use of environmental and medical sur-
veillance data in order to provide adequate protection. Selecting quali-
fied medical monitoring facilities is mandatory. This process includes
accredited laboratory facilities with appropriate QA/QC programs.
The AIHA/ACGIH Hazardous Waste Committee and its Medical
Surveillance Subcommittee have been working with OSHA toward
creating a generic standard for medical surveillance. Such a standard
will give all hazardous waste contractors a starting point. Site-specific
variations can then be added by the site health and safety officer with
assistance from a board'certified occupational/environmental medicine
physician. The AIHA/ACGIH committee and its subcommittee have
created a set of criteria to assist the hazardous waste remedial action
contractors in their selection of appropriate medical facilities to do their
employee medical examinations.
The Rocky Mountain Arsenal cleanup at Basin F is a good example
of a properly operated medical surveillance and environmental
monitoring system.
The remedial action contractor built a state-of-the-art decontamina-
tion facility capable of handling up to 120 people in 30 min. The site
workers of Rocky Mountain Arsenal faced the problems of heat stress
in 90 to 97° F. summer weather as well as frost-bite in sub-zero winter
weather without any serious casualties. The contractor learned how to
keep workers alive and well and still get the job done and make money.
During the early days of Superfund site identification and charac-
terization, the occupational/environmental medicine physician was
monitoring young, health-conscious scientists. As the remedial action
phase got under way, an entirely different group of people came under
medical surveillance. They ranged in age from 18 to 65 and their life
styles in general were in sharp contrast to the scientist group. The
findings on their medical examinations were quite different. These
findings present additional decision'making problems for the hazardous
waste contractor.
There is need for continuous interaction between the contractor
management and the medical monitoring facility. One must meet all
of the requirements of OSHA and the U.S. EPA and still comply with
the Privacy Act relative to confidentiality of medical information. One
must also deal with the mandate of EEOC in terms of non-discrimination
in hiring practices.
Medical surveillance of all persons entering upon a Superfund site
is required by law under OSHA. If it is well done, it can be of great
benefit to all concerned. If it is poorly done, it can create a host of
potentially expensive problems.
Introduction
The history of the Rocky Mountain Arsenal begins in 1942. It was
established by the U.S. Department of the Army as a manufacturing
facility for the production of chemical and incendiary munitions. During
World War II, chemical intermediate munitions, toxic products and
incendiary munitions were manufactured and assembled by the U.S.
Army. From 1945 to 1950, stocks of Levinstein mustard were distilled,
mustard-filled shells were demilitarized and mortar rounds filled with
smoke and high explosives were test fired. Various obsolete ordnance
were also destroyed by detonation or burning during this period.
In the early 1950s, RMA was selected to produce the chemical nerve
agent GB (Sarin) under U.S. Army operations. The North Plants
manufacturing facility was completed in 1953 and was used to produce
agents until 1957. Munitions-filling operations continued until late 1969.
Primary activities between 1969 and 1984 involved the demilitarization
of chemical warfare materials.
Concurrent with military activities, industrial chemicals were
manufactured at RMA by several lessees from 1947 to 1982. The
products included chlorinated benzenes, naphthalene, chlorine, fused
caustic, insecticides (DDT, Aldrin, Dieldrin and Endrin), herbicides,
nematocides, adhesives, anti-icers and lubricating greases.
In May of 1974, di-isopropylmethyl phosphonate and dicyclopen-
tadiene were detected in the surface water at the northern boundary
of the arsenal. Later that year , the Colorado Department of Health
(CDH) detected the same chemical in a well north of the arsenal and
issued three administrative orders against Shell and/or the Army in April
of 1975. Thus began the litigation history of the RMA.
One must remember that in 1942 Denver was a very small town and
the arsenal site was far from civilization. I also doubt that any of the
planners at that time ever dreamed that the arsenal would some day
be surrounded by a large urban area. I seriously doubt that any con-
sideration was given to groundwater and its possible contamination by
arsenal activities. The country was engaged in a global war and it was
"full speed ahead" with little consideration for the environment.
The Rocky Mountain Arsenal encompasses approximately 17,000 ac
of land in Adams County, Colorado. Much of it looks very benign.
There are many species of wildlife roaming the arsenal including herds
of deer, thousands of prairie dogs and rabbits. Games birds abound
and the lakes and ponds are full of fish. Bald eagles have built a nest
on arsenal property.
The full extent of the contamination of soil and groundwater at the
arsenal is not known. It may never be known. The lists of chemicals
fill many pages and are not considered complete. The environmental
problems created by the arsenal have created much political activity
and the public has demanded action.
HEALTH & ENDANGERMENT 75
-------
SITE CHARACTERIZATION MEDICAL MONITORING
The initial stages of cleanup began with site characterization. My
organization has been involved in the medical monitoring for the past
9 years. We have supplied medical surveillance and biological
monitoring to a number of the contractors. In this initial phase of cleanup
activity, the majority of the personnel working on the Rocky Mountain
site were scientists and technicians.
The Activities at this stage of the investigation included drilling test
wells, taking soil and water samples and doing air monitoring. Some
of the locations in which they were working called for level B protec-
tion. This means the use of impervious clothing in multiple layers
including gloves and boots with liners and helmet with supplied air
(either back pack with tanks or airline hose from a compressor). I have
been suited up for level B and have been out on-site carrying equip-
ment and doing physical work in direct sunlight with outside ambient
temperatures in the 70's (°F). Within 20 min. I began to feel the heat
load. I could only imagine how 1 would feel if the outside ambient tem-
peratures were in the high 90's (°F). I was wearing a back pack with
compressed air tank and was put through the experience of a tank
change. I recall that it took me 20 min. to get dressed with the help
of two people from the decontamination unit.
I went through this exercise because I feel that it is important for
the examining physician at hazardous waste projects to experience exactly
what the workers on the project will experience. It gave me a much
broader perspective and enabled me to do a better job for the contractor
personnel.
As each contractor began work in 1980,1 sat down with the industri-
al hygiene and health and safety personnel and reviewed the available
environmental data. Based on this discussion, we jointly developed a
protocol for medical monitoring. We knew that we were dealing with
organic solvents and a variety of chlorinated hydrocarbon pesticides.
These chemicals are known liver toxins and some are kidney toxins.
Some are leukemogens and others can cause aplastic anemia. The
organophosphate pesticides depress the cholinesterase.
All of the baseline physical examinations included: personal and
family medical history, occupational history, hobbies, recreational
activities, use of tobacco, alcohol and medicines; hands-on physical
examination; chest x-ray; lung function testing; resting ecg or tread-
mill ecg (for level A & B protection); complete blood count; biochemical
profile; vision screening; audiometric testing; complete urine analy-
sis; cholinesterase; methemoglobin; heavy metals. Where indicated,
pesticide screens of various kinds were added. All personnel were
screened for drug use as a safety precaution.
Having developed the medical monitoring protocol, the next impor-
tant step was to select laboratories that had good QA/QC programs.
We chose a local reference laboratory that was certified by the Centers
for Disease Control. This laboratory participated in two external qual-
ity control programs. We then found a toxicology laboratory that also
demonstrated proper QA/QC by participating in multiple external pro-
grams. Our trace metals laboratory participates in four national and
two international Q/QC programs and does 40% QC in house. We feel
very comfortable with the results from these laboratories.
Chest x-rays (PA and lateral views) are interpreted by a Board Certi-
fied radiologist. Pulmonary function tests are done on equipment that
is calibrated daily. The tests are conducted by a technician who has
taken a NIOSH-approved course in spirometry and passed the certifi-
cation examination. The ecg's are all interpreted by cardiologists. The
audiometric testing is done on equipment that meets the requirements
of the OSHA standard on hearing conservation.
After the baseline examination, employees are monitored on a periodic
basis. The interval is usually once a year, but in some situations may
be more frequent. They also are examined at the end of a project or
when leaving employment. If there has been a spill or toxic release,
an interim examination usually is conducted.
In the initial phase of the arsenal cleanup, we were examining groups
of scientists and technicians engaged in site identification and charac-
terization. These were almost all young people (aged 20 to 35), in
excellent health and in a good state of physical fitness. Many were com-
petitive athletes. Most were non-smokers and had good dietary habits.
Very few were overweight. Most of them consumed little or no alcohol.
We did not find any drug users.
CLEANUP PROGRAM MEDICAL MONITORING
When the remedial action contractors came for the next phase of the
cleanup, we found ourselves evaluating an entirely different group of
people. They ranged from 18 to 65 years of age. Quite a number were
overweight. Most were in a fair to poor state of physical fitness. Many
were cigarette smokers and heavy users of alcohol. We found a few
drug users. Many had very poor eating habits. Many had elevated serum
lipids (cholesterol and/or triglycerides).
A major remedial action effort took place in 1988-1989 at a location
known as Basin-F. This area was considered one of the most highly
contaminated areas at the RMA. The cleanup plan was drawn up by
the contractor. This plan of course, included a site health and safety
plan which was reviewed by the industrial hygiene group and the medical
monitoring group. The medical surveillance protocol was created and
adopted. This project was unique in that the labor force was union
organized. The "hot zone" was identified and was surrounded by a buffer
zone with a perimeter fence. All personnel working inside the fence
were put in level B protection during the initial phase of the cleanup.
A team of 22 field health and safety personnel was selected and hired
by the contractor. This team included three EMT's (emergency medi-
cal technicians) and seven air monitoring and meteorology technicians.
All of these people were trained by the contractor for this project. The
air monitoring personnel selected seven locations for monitoring stations
at the perimeter. High volume samplers were used with continuous real-
time computerized monitoring. Wind velocity and direction also were
measured.
The contractor designed and built a state-of-the-art decontamination
facility. This facility was equipped with the latest in decontamination
equipment. It was staffed by a supervisor and seven technicians. It was
stocked with large supplies of all of the required protective clothing
and respiratory protective gear. It was capable of handling up to 120
people in 10 min. It also housed the on-site laundry. It was set up to
handle male and female personnel (there were some female heavy equip-
ment operators and laborers).
During the summer months, all field personnel were monitored for
heat stress. Hourly WBGT monitoring was done in the support zone.
The work/rest regimen was based on these readings. As the ambient
temperatures went up, the work periods were shortened and the rest
periods were lengthened. During 90 to 100°F weather, the work day
began at 4:00 a.m. and the project was shut down by 10:00 a.m. Under
these conditions, the labor force would work for 20 min. and rest for
40 min. Some of the work was done on the night shift to avoid the heat.
During the rest periods, personnel were pulled out of the "hot zone"
into the decontamination trailer which was maintained at 60°F. The
EMT's checked vital signs (heart rate, respiration, blood pressure and
body core temperature). The body core temperature was measured by
using a tympanic thermometer. Field medical monitoring was the first
line of defense against heat stress.
At the peak of activity, there were 60 to 70 people in the Basin all
of the time. Approximately half of these were heavy equipment opera-
tors and half were laborers. The heavy equipment operators were in
air-conditioned cabs and therefore had less solar load than the laborers.
The heavy equipment operators were all on supplied air from racks
of tanks on their vehicles. It became necessary in the hot weather to
ice the air tanks and the hoses. Even with all of these measures, the
ambient temperature inside the cabs got into the mid 80's (F). Some
of the laborers used back packs with small 45 min. tanks of compressed
air. Others dragged airline hoses around that supplied air from large
tanks or compressors.
When it came to putting down the black vinyl liner for the waste
pile containment, the contract had to be re-written to allow this work
to be done at night. During the day, thermometers 3 ft. above the liner
registered WOT. At the beginning of the hot weather, I was asked, as
the medical consultant, to go out to the work site and give a lecture
on heat stress to the managemers and supervisors. This lecture was
76 HEALTH & ENDANGERMENT
-------
accomplished by utilizing the charts and slides on heat stroke and heat
stress prepared by ACGIH. The contractor then took the posters and
prepared from the cartoons a booklet that was given to all personnel
in the field.
OVA's (organic vapor analyzers) and HNU equipment were used for
field environmental monitoring. Organic chemical in the ambrent air
were measured to monitor pesticide levels. Ammonia, hydrogen sulfide
and fugitive dust also were measured. Neither ammonia nor hydrogen
sulfide was detected. All of the above measurements were done with
real-time monitoring. The fugitive dust action limit was set at
1 mg/m3. Action limits set for chemicals were fractions of PELs and
TLV's.
During the winter months, insulated coveralls were worn under the
protective clothing. The impermeable suits provided protection against
the wind chill. Hard hats were fitted with insulated liners. Insulated
boot liners were provided and cotton gloves were worn inside the
neoprene gloves. There was a warm up regimen of 15 min/hr of work
when ambient temperatures were at 0°F. During the warm up period,
the EMT's checked fingers, toes and ears for evidence of frostbite. Body
core temperature was measured to be sure that employees were not going
into hypothermia.
As the clean-up progressed, the basin was downgraded to level C
personal protective equipment, based environmental monitoring data.
However, when the OVA reading for toluene (as an example) exceeded
1 ppm, the area would be upgraded to level B and field personnel would
go back on supplied air. Every precaution was taken to prevent exposure
of field personnel to toxic chemical hazards. During me entire project,
there was daily communication between the site health and safety officer
and the medical consultant.
CONCLUSIONS
During the peak phase of activity there were 130 people on-site in-
cluding support personnel outside the perimeter fence. All of these sup-
port personnel, including security guards and office workers, were
included in the medical monitoring program. The contractor was not
taking any chances with the health and safety of his personnel. Some
of the motivation, of course, was not only medical but also legal. All
employees including regular employees staying on with the company
were provided with exit medical examinations as they left the project.
Analysis of the health and safety data did not reveal any evidence of
serious exposure to toxics. There were no heat stroke or frost-bite
casualties.
The contractors and their management and field personnel are to be
congratulated for carrying out a project of this scope and magnitude
with no serious health casualties.
ACKNOWLEDGEMENT
I wish to thank Myron Temchin, Diane Morrell, Mickey Lewis and
Dalene Nicholson of Ebasco/Envirosphere and John Schmerber and
Karen Lewis of Morrison Knudsen Environmental for their assistance
in supplying some of the material for this paper.
REFERENCES
1. Proctor, N.H., Hughes, J.P. and Fischman, M.L., Chemical Hazards of the
Workplace, 2nd Ed., J.B. Lippincott Co., Philadelphia, PA, 1988.
2. Kneip, T.J. and Crable, J.V., Methods for Biological Monitoring, American
Public Health Assoc., Washington, DC, 1988.
3. Williams, PL. and Burson, J.L., Industrial Toxicology, Van Nostrand Rein-
hold Co, New York, NY, 1985.
4. Baselt, R.C., Disposition of Toxic Drugs and Chemicals in Man, 2nd Ed.,
Biomedical Publications, Davis, CA, 1982.
HEALTH & ENDANGERMENT 77
-------
Quantitative Public Health Risk Assessment at a
Low Level Contaminant Superfund Site
John H. Lange
Gerard M. Patelunas
GAI Consultants, Inc.
Monroeville, Pennsylvania
Robert J. Williams
Virginia Electric and Power Company
Glen Allen, Virginia
ABSTRACT
A study of the potential public health risk was undertaken at the Chis-
man Creek Superfund Site. This site is located in south east York County,
Virginia, and consists of four fly ash disposal areas, three man-made
ponds and a freshwater tributary stream that drains into Chisman Creek
and the Chisman Creek estuary. Fly ash was generated from a fuel mix-
ture of bituminous coal and petroleum coke and was disposed at the
site from 1957 to 1974. The site was placed on the NPL in 1983
Contaminants associated with the site are nickel, arsenic, vanadium,
lead and zinc. Potential pathways of exposure reviewed included soil
and surface water. A public health risk assessment was calculated for
chronic intake of carcinogenic and non-carcinogenic contaminants. In-
gestion of contaminants was calculated for members of the general popu-
lation, including sensitive persons. These calculated exposure values
were used to determine the risk associated with this site. Using the
derived data and published information, risks estimated for the local
population were determined. These risk assessment values were deter-
mined not to "exceed" the U.S. EPA's 10-4 to 10-7 level of carcino-
genic risk or unity for the non-carcinogen hazard index. Derived health
assessment information was used as one variable in determining the
necessary remediation criteria.
This paper discusses problems encountered in determining exposure
factors and incremental risks at a site containing low levels of trace
metals. The results of this study indicate that risk interpretation must
be conducted with caution at low level metal sites. The dietary
importance and risk relation ship of trace metals also is noted. Since
the mechanisms of actions for the trace metals studied are different,
no combined effects were calculated. Numerical values for carcino-
genic potencies and acceptable intake concentrations for chronic
exposure were obtained from the U.S. EPA Superfund Public Health
Evaluation Manual. Other factors influencing risk are discussed as
related to the exposed population. The importance of sensitive
individuals in the population is noted. Regulatory evaluation, assump-
tion factors for a sensitive population and risk assessment as a
remediation criteria are discussed.
INTRODUCTION
Numerous federal and state laws have recently been enacted requiring
investigation and remediation of sites contaminated with hazardous sub-
stances including organic, inorganic, pesticide, radionuclide and other
wastes'. The primary factor responsible for site selection, remedy
selection and cleanup levels has been the site's actual or potential impact
affecting human health and the environment, often collectively called
a public health hazard. Several highly publicized incidents resulting
in threat or harm to the public and environment originally triggered
enactment of the initial Superfund legislation (CERCLA) and the re-
authorization legislation (SARA). As a result of these Superfund laws,
environmental engineers have developed new techniques to control,
transport, excavate, stabilize, incinerate, biodegrade and encapsulate
materials considered to be hazardous. Although these technologies can
successfully control and remediate hazardous material at sites, infor-
mation about the public health hazards posed by the sites often was
lacking. To evaluate public health risks, methodology was modified
and/or formulated to quantih risks associated with hazardous waste
sites. These public health epidemiological and statistical methods used
data derived from the fields of toxicology, physiology, industrial hygiene,
biology, chemistry and meteorology. This interdisciplinary approach
resulted in a "new" discipline called risk assessment. With the develop-
ment of any "neu subject," a degree of maturity and growth is necessary
to establish a theoretical and practical basis. This paper will provide
a case example of risk methodology and interpretation used to assist
in the determination of cleanup standards for a low-level Superfund site.
Regulatory Considerations
Since risk assessment is in its early stages of scientific and regulatory
evolution, few, if any, governmental agencies have established procedural
policy to conduct, evaluate, interpret and review this technique. However,
the (U.S. EPA) published five proposed guidelines (carcinogenicity,
mutagenicity, developmental toxicity, chemical mixtures and exposure)
to help risk assessors establish standards for conducting risk
assessment'. Although these guidelines are not regulations, they do
provide a framework in which cleanup risk assessment criteria can be
addressed. In fact, other agencies (e.g., the EPA, Nuclear Regulatory
Commission) have used similar methodological approaches to estab-
lish standards4. As the field of risk assessment develops, better refine-
ment of techniques will allow more governmental agencies to use these
procedures to establish regulatory standards and cleanup criteria.
Routes of Exposure
The routes or pathways of exposure in humans from hazardous waste
site activities include dermal, ingestion and inhalation. Traditionally,
the primary occupational route was inhalation. However, this pathway's
importance is diminished in non-occupational populations. In the general
population, the route and potential exposure can dramatically vary from
one individual to another. Therefore, numerous scenarios must be
evaluated to determine maximum risk. In almost all cases, the worst
practical scenario must be considered when determining the final risk.
This risk assessment process may include a synergism of compounds,
routes and number of exposure events.
Site Background Information
The Chisman Creek Superfund Site is located in southeast York
78 HEALTH & ENDANGERMENT
-------
County, Virginia, approximately 1 mi north of Grafton5. The site con-
sists of four fly ash disposal areas, three man-made ponds, a fresh-
water tributary stream that drains the site and flows into Chisman Creek
and the Chisman Creek estuary. The site has been divided into two
operable units by the U.S. EPA. Operable Unit 1 consists of the four
fly ash disposal areas (designated Pits A, B, C and D) and areal ground-
water. Operable Unit 2 consists of three ponds (designated A, B and
C), the freshwater stream and the Chisman Creek estuary. This paper
addresses the risks associated with operable Unit 2 only.
The site contains fly ash generated from a fuel mixture of bituminous
coal and petroleum coke. The fly ash was produced at Virginia Elec-
tric and Power Company's (Virginia Power) Yorktown Power Station
and was disposed of at the site by R. L. Brandt and Sons, Inc., a local
contractor, from 1957 to 1974. The site was placed on the NPL in 1983.
Previous investigations of the Chisman Creek site include studies by
the Virginia Department of Health and the Virginia Water Control Board
in 1980 and 1981 and by the Virginia Institute of Marine Science in
1983. The RI and FS for Operable Unit 1 were performed by the U.S.
EPA's contractor, CH2M Hill, in November, 1985 and August, 1986,
respectively. The ROD for Operable Unit 1 was issued by the U.S. EPA
in September, 1986. Virginia Power agreed to per form the Operable
Unit 1 remediation, and remedial construction was completed in
December, 1988.
The U.S. Fish and Wildlife Service (FWS) conducted the RI for oper-
able Unit 2. The draft RI report was issued in April, 1987, and the final
RI report was issued in December, 1987. GAI Consultants, Inc. (GAI),
a contractor to Virginia Power, conducted the FS for Operable Unit
2 by agreement with the U.S. EPA. Virginia Power agreed to perform
the Operable Unit 2 remediation, and remedial construction was
completed in December, 1988.
Long-term operation and maintenance activities currently are being
performed on the site by Virginia Power. Additionally, Virginia Power
and the County of York have entered into an agreement to operate the
site as a public park. Softball and soccer fields along with nature areas
have been developed and public use of these facilities is scheduled for
mid-1990.
Methods
Risk assessments were calculated for the chronic intake of carcino-
genic and non-carcinogenic trace elements. No combined effects were
included in the risk assessment calculations since the mechanism of
actions for these metals studied are different. Numerical values for car-
cinogenic and non-carcinogenic potencies and acceptable intakes for
chronic exposure were obtained from the U.S. EPA Superfund Public
Health Evaluation Manual (SPHEM)6.
Chemical concentrations were determined as described in the Chisman
Creek Superfund Site Feasibility Study for Operable Unit 25. Con-
taminated elements evaluated in this study are nickel, arsenic, vanadium,
lead and zinc. The highest concentration of each contaminant was used
for the risk evaluation. Assessment pathways evaluated were ingestion
of sediment (soil) and surface water. All ingestion was considered to
be accidental and does not represent a daily consumption intake. In-
take values for water and sediment are 100 mL (non-carcinogens) and
1000 mL (carcinogens) per day and 10 g per day, respectively. Intesti-
nal absorption was considered 100% for all com pounds. For carcino-
genic calculations, a lifetime was 70 yr, body weight was 35 kg and
exposure duration was 450 days over one's lifetime. For non-carcinogenic
calculations, a body weight of 10 kg was used.
Risk assessment calculations for the carcinogenic elements, nickel
and arsenic, are identical to those described in the Superfund Public
Health Evaluation Manuaf and the "Chisman Creek Superfund Site
Feasibility Study for Operable Unit 2"5 The Hazard Index Value (HI)
for non-carcinogenic elements, nickel, vanadium, lead and zinc, were
determined using identical methods as described in the Superfund Public
Health Evaluation Manual6. All values were considered for chronic
exposure. The Acceptable Intake Concentrations for chronic exposure
(AIC) for vanadium, nickel, lead and zinc were obtained from the
Superfund Public Health Evaluation Manual6 or the Health Effects
Assessment Documents for Arsenic7 or Nickel and Nickel
compounds8. The Chronic Daily Intake values (GDI) were calculated
by dividing the ingestion concentration by the body weight.
RESULTS
The highest concentrations of carcinogenic elements are shown in
Table 1. These values are reported for Ponds A, B, C and the stream
water and sediment. Ingestion was determined from the concentrations
in Table 1 and is represented using in the value of mg/day (Table 2).
The risk assessment based on a lifetime exposure from ingestion of
water or sediment is shown in Tables 3 and 4. These values represent
additional cases of cancer over a lifetime. The carcinogenic potency
values for nickel and arsenic are 0.84 mg/kg-day and 15 mg/kg-day,
respectively.
Non-carcinogenic water and sediment concentrations are shown in
Table 5. Ingestion concentration for water and sediment per day are
shown in Tables 6 and 7, respectively. The hazard index values for
vanadium, nickel, lead and in zinc sediment are shown in Table 8. Values
equal to 1.0 are defined as unity. All methodology is identical to the
Superfund Public Health Evaluation Manual (6) and the "Chisman
Creek Superfund Site Feasibility Study for Operable Unit 2"5.
Table 1
Highest Concentrations Detected of Nickel and Arsenic in Water and
Sediment for Carcinogenic Evaluation
Water Concentration (ppb)
Nickel
Arsenic
Pond A
27
10
Pond B
27
10
Pond C
27
10
Sediment Concentration (ppm)
Pond A Pond B Pond C
Nickel
Arsenic
749
28
79
128
29
15
Stream
83
145
Stream
107
17
Table 2
Ingestion of Nickel and Arsenic from Water and
Sediment (mg/day) for Carcinorganic Evaluation
Water Concentration
Nickel
Arsenic
Pond A
0.027
0.010
Pond B
0.027
0.010
Pond C
0.027
0.010
Sediment Concentration
Pond A Pond B Pond C
Nickel
Arsenic
7.49
0.28
0.79
1 .28
0.29
0.15
Stream
0.083
0.145
Stream
1 .07
0.17
Table 3
Risk Estimate for the Ingestion of Water from the Ponds and
Streams as Additional Cancer Cases Over a Lifetime Period
Nickel
Arsenic
7.5 x 10"
Pond B
3.1 x 10
,-13
7.5 x 10'
,-13
Pond C
3.1 x 10
7.5 x 10~
Stream
2.9 x 10~12
1.6 x 10'
-10
HEALTH & ENDANGERMENT 79
-------
Table 4
Risk Estimate for the Ingestion of Sediment for the Pbnds and
Streams as Additional Cancer Cases Over a Lifetime Period
Pond A
Pond B
Pond C
Stream
Table 8
Hazard Index Values for Non-Carcinogenic Elements
CDI »ie
or ma/l/Omyl l«q/falAlay or »q/l/day)
Nickel 2.
Arsenic 5.
4 x 10~8
9 x ID'10
Water and
2.6 x lO"10 3.5
1 .2 x 10~B 1-7
x ID'11
x ID'10
4.8 x ID'10
2.1 x 10-'°
Table 5
Sediment Concentrations for the
Vanadluv
Hater
8edle»nt
Nlcxel
Hater
SedlMnt
Pond A
0.0008
0.0167
0.0003
0.0075
_ j
2.0 « 10 '
2.0 « 10"'
1.0 * 10"'
1.0 » 10"2
0.0«
0.84
0.0)
0.74
Non-Carcinorganic Evaluation
Nickel
Vanadium
Lead
Zinc
Pond A
27
80
5
4.6
Water Concentration
Pond B
27
19
5
4.2
(ppb)
Pond C
27
19
5
18
St ream
83
70
970
83
Sediment Concentration (ppm)
Nickel
Vanadium
Lead
Zinc
Pond A
749
1,670
26
202
Ingestion of Water
Nickel
Vanadium
Lead
Zinc
Ingestion
Nickel
Vanadium
Lead
zinc
Pond A
0.027
0.080
0.005
0.005
Pond B
79
141
17
38
Table 6
for Non-Carcinorganic
Pond B
0.027
0.019
0.005
0.004
Table 7
Pond C
29
48
13
67
Evaluation
Pond C
0.027
0.019
0.005
0.018
Stream
107
541
62
217
(mg/day)
Stream
0.083
0.097
0.070
0.008
of Sediment for Non-Carcinorganic Evaluation (mg/day)
Pond A
0.749
1 .670
0.280
0.020
Pond B
0.079
0.141
0.017
0.038
Pond C
O.C29
0.048
0.01 3
0.067
Stream
0.011
0.054
0.062
0.022
Lead
Hater
Sedle»nt
Zinc
Hater
Sedltent
VenedlUB
Hater
Sedleant
Klckel
Hater
Svdlaent
Lead
Water
SedlMnt
zinc
«it«r
Sedla»nt
Vanadl UB
Hater
Sedlient
Nickel
Hater
Sedlnnt
Lead
Hater
Sedl»ent
Zinc
Water
SedlMnt
VanAdluei
Hater
SedlMnt
Nickel
Hater
SadlMnt
Lead
•ater
Sediment
tine
Hater
Sediment
0.0001
0.0001
0.0001
0.0020
Pond I
O.OOO2
0.001]
0.000]
0.0079
o.ooot
0.000]
o.ooot
O.OOO4
O.OO01
o.ooos
O.OOO]
0.000)
0.0001
0.0001
O.OO02
O.OO07
Streaai
0.0097
O.OO54
0.0006
0.0011
0.0001
o.ooot
0.0008
0.0022
1.4 * 10" 3
1 .4 x 10"3
_«
2.10 « 10
2.10 « 10"'
2.0 > 10"'
2.0 < 10"2
i.O x 10"2
l .0 « 10"'
1.4 « 10"3
1.4 x 10"3
2.10 « 10"'
2.10 > 10"'
3
2.0 i 10 '
7.0 i 10"'
i.o « 10"'
1.0 x 10"'
1.4 X 10
1.4 X 10"5
2.10 « 10"'
2.10 > 10"'
2.0 « 10"'
2.0 I 10"'
1.0 x 10"'
-\
1.0 x 10
1.4 x 1 O"3
1.4 X 10"3
2.10 X 10"'
2.10 x 10"'
0.04
0.20
0.0002
0.01
0.01
0.07
0.0]
0.08
0.04
0.12
0.0002
0.002
0.01
0.02
0.01
0.0]
0.04
0.09
0.001
0.001
0.41
0.01
0.01
0.11
0.04
0.44
0.004
0.01
DISCUSSION
Health Effects
The concentration values for carcinogenic elements, nickel and
arsenic, and non-carcinogenic elements, vanadium, nickel, lead and
zinc, are represented in Tables 1 and 5. These values are below
concentrations associated with heavy metal diseases'-8-9. Nickel is con-
sidered both a carcinogenic and a non-carcinogenic hazard by U.S. EPA
in the SPHEM6. However, primary association for nickel as a carcino-
genic element is through occupational inhalation8-9 resulting in elevated
nasal and lung cancer. A different, and more important, cancer etiology
is reported for arsenic. Arsenic has been associated with skin cancer
in humans drinking contaminated water' Unlike nickel, this associa-
tion represents a valid concern as related to the public health.
The non-carcinogenic elements have only been associated with disease
in elevated concentrations on a chronic exposure basis or accidental
acute episodes. Since the element concentrations of interest are well
below acute toxicological dose thresholds, no acute non-carcinogenic
hazard index values were calculated. However, recent concerns over
lead have raised the issue of whether low concentrations may be harmful
to sensitive members of the population10. The sensitive groups most
often considered are pregnant women and young children. It is likely,
as with asbestos, that a no observable dose threshold exists with lead.
However, as with numerous trace elements, it is possible that low con-
80 HEALTH &.ENDANGERMENT
-------
centrations are biochemically necessary for normal metabolism. A
detailed discussion of health effects and biology of the elements as-
sociated with this site can be found in references 5, 7, 8 and 9.
Exposure Constants
Values of intake are represented in mg/day for both carcinogenic and
non-carcinogenic elements (Tables 2, 6 and 7). These values were then
divided by the body weight (35 or 10 kg). The low weight values were
used to introduce a highly conservative estimate in the final calcula-
tions. There fore, population members having weights greater than those
used in the calculations exhibit an even larger factor of conservatism.
This process allows risk assessment policies to provide a high degree
of protection to all population members without opting for the "tradi-
tional" average risk estimates". Use of these values can provide
standard guidelines for risk assessment at low level sites without
unrealistic conservatism.
A similar factor was applied to the intake values of water and sediment.
Although these values are large for accidental ingestion in the adult
population, this exposure may be realistic for children who, for exam-
ple, engage in frequent pica ingestion. Most of these conservative
assumptions are balanced by a more liberal exposure time of 450 days.
Providing the values used in this investigations's risk assessment cal-
culations allows other scientists to better judge methodology rather than
use uncertainty factors12. Until more accepted intake values are deter-
mined for sensitive populations, risk assessors should utilize conser-
vative factors to address and satisfy the public's expectation of a high
degree of public protection.
Risk Assessment
Values for additional cancers were determined using the highest
reported water and sediment concentrations (Table 3 and 4). No calcu-
lated risk value approaches the 1.0 x 10~7 additional cancer risk. The
highest risk value reported was 2.4 x 10~8. With risk values so low,
it is impractical to consider synergistic or additive effects of exposure.
If these assumptions were to be incorporated into the risk estimate,
simple addition or multiplication may be a valid mechanism for risk
synergy. Thus, a theoretical mechanism for estimating the synergistic
effect is to multiply the largest risk values of each compound. Inclusion
of factors beyond simple multiplication for com pounds effecting dif-
ferent organ systems is unrealistic.
Hazard index (HI) values were all below unity (Table 8). In fact, some
HI values were below the significant place calculations. With the
exception of vanadium in the sediment of Pond A, no HI value
approached unity. The Superfund Public Health Evaluation Manual6
states: "It is emphasized that the hazard index is not a mathematical
prediction of incidence or severity of effects. It is simply a numerical
index to help identify potential exposure problems. Results for multi-
ple chemicals should not be interpreted too strongly. If some of the
indicator chemicals do not have adequate toxicity information, thus
preventing their inclusion in the hazard index, the hazard index may
not be reflective of actual hazards at the site. Consideration of chemi-
cals that do not have toxicity values could significantly increase the
hazard index to levels of concern. Professional judgement is required
to determine how to interpret the hazard index for a particular site."
The addition of HI values within and between groups may provide
some insight to potential risks. These risks may be future classified
when a cumulative value is determined by addition from the same
source. These risk characteristics, called cumulative values, are defined
as follows:
<1- no hazard or risk
1 to 2-incrementally elevated or an acceptable risk
3 to 5-moderate concern for the sensitive population
5 to 10-moderate concern for the general population
>10-a concern requiring a planned action
>13-immediate concern for the public
Although these cumulative values are arbitrary and have not been
validated in actual population studies, they do provide a range in which
to judge a qualitative HI value risk. However, as with any non-threshold
estimate, judgment based on animal and epidemiological studies must
be considered when making a final determination. This judgmental
process becomes even more evident when evaluating Superfund sites
that contain low levels of contaminants.
Neither the carcinogenic or non-carcinogenic risks at the Chisman
Creek site are at levels of concern as related to the public's safety.
Although some consider lead to be a no-threshold element, most other
elements have dietary importance. Incorporation of an element's essential
dietary requirements usually is not considered in risk assessment.
However, increased levels of some compounds (e.g., selenium) may
be beneficial'3-
Risk Assessment for Evaluating Action Alternatives
Since engineering design methods provide no evidence or guidance
in regard to health effects, the resultant remedial construction usually
is unrelated to risk assessment judgments. However, combining the
engineering design with desired health risks provides a useful selection
mechanism. The action alternatives should be based first on the ability
to achieve the desired risk. This risk must incorporate the surrounding
natural background concentrations, the exposure pathways and accept-
able level of toxicants. At low level sites the aesthetic values, public
pressures and cleanup costs usually are stronger considerations than
elevated public health risks. However, systematic and site-specific risk
determinations provide valuable information for the selection of remedial
actions and the level of cleanup and hazards associated with the actual
remedial construction.
CONCLUSION
This investigation provides additional guidance for determining action
alternatives at low level hazardous waste sites. When carcinogenic risk
assessments are below 10~7 and non-carcinogenic additive values are
below a cumulative value of 2.0, the importance of future site develop-
ment and costs become of greater importance. The selection of a remedy,
the level of cleanup and cost of cleanup should be directly related to
the cumulative non-carcinogenic risk and carcinogenic risk assessments.
REFERENCES
1. Hazardous Sites Cleanup Act of Pennsylvania. Act 1988-108.
2. Hayes, D. J. and MacKerron, C. B., "Superfund II: A New Mandate."
The Bureau of National Affairs Special Report, 1987.
3. Preuss, R. W., Ehrlich, A. M. and Garraham K. G., "U.S. EPA Guide-
lines for Risk Assessment." Proc. Seventh National Conference on Manage-
ment of Uncontrolled Hazardous Waste Sites. HMCRI, Silver Spring, MD
1986.
4. U.S. EPA. National Emission Hazard for Hazardous Air Pollutants: Regu-
lation ofRadionucleotides, 40 CFR 61 (Federal Register Volume 54; page
9612), 1989.
5. Virginia Electric and Power Company. "Chisman Creek Superfund Site.
Feasibility Study for Operable Permit 2, Glen Allen, Virginia," 1988.
6. U.S. EPA. Superfund Public Health Evaluation Manual. EPA/540/1-86/060,
Washington, DC Oct. 1986.
7. U.S. EPA. Health Effects Assessment Document for Arsenic. Publication
Nos. PB 86-134319, Cincinnati, OH.
8. U.S. EPA. Health Effects Assessment Document for Nickel and Nickel Com-
pounds. Publication Nos. PB 86-232212, Research Triangle Park, NC.
9. Rom, W. N., Environmental and Occupational Medicine. Little, Brown and
Company, Boston, MA 1983.
10. Needleman, H. L. and Bellinger, D., "Commentary: Recent Developments,"
Environ. Res. 46, pp 190-191, 1988.
11. Hubbard, A. E., Hubbard, R. J., George, J. A. and Hagel W. A., Quan-
titative Risk Assessment as the Basis for Definition of Extent of Remedial
Action at the Leestown Pesticide Superfund Site, Proc. Seventh National
Conference on Management of Uncontrolled Hazardous Waste Sites on Un-
controlled Hazardous Waste Sites, HMCRI, Silver Spring, MD, 1986, pp
186-192.
12. Hirschhorn, J. S. Oldenburg, K. U. and Doran D., "Using Risk Concepts
in Superfund." Proc. of the Eighth National Conference, Superfund 1987,
HMCRI, Silver Spring, MD.1987, page 166-168.
13. Lange, J. H., Talbott, E. Q, Baffone, K. M., Weyel, D. A., Soboslay, E.
G., Koros, A. M. C. and Sykora, J. L., "Selenium Cancer Activities of
Selenium." Medical Hypothesis, pp 443-447, 1987.
HEALTH & ENDANGERMENT 81
-------
Quantitative Uncertainty Analysis
in Exposure and Dose-Response Assessments
in Public Health Risk Assessments Using Monte Carlo Techniques
David E. Burmaster, Ph.D.
Katherine E. von Stackelberg
Alceon Corporation
Cambridge, Massachusetts
ABSTRACT
Most health risk assessments for Superfund sites combine a series
of high, upperbound or worst-case assumptions to derive a point esti-
mate of risk that is conservative, i.e., protective of public health. By
setting the bias high enough to dominate the uncertainty in each step,
such a risk assessment considers senarios that will rarely, if ever, happen.
In addition, the results from such a risk assessment have an unknown
amount of conservatism built into them. This paper presents a method
for uncertainty analysis using Crystal Ball™ for Monte Carlo simula-
tions. The program combines thousands of realizations for the proba-
bility density functions of each input variable yielding a final probability
distribution rather than a single number.
INTRODUCTION
Following guidance published by the U.S. EPA, most health risk
assessments for hazardous waste sites concatenate a series of high,
upperbound, or worst-case assumptions to derive a point estimate of
risk that is conservative, i.e., protective of public health.23-23 The U.S.
EPA is well aware that risk assessments need to include uncertainty
analyses and sensitivity analyses in every project. Through guidance
documents,22- a handbooks21 and research reports,1 M the Agency
requires uncertainty analyses in Superfund investigations and has
investigated algebraic and computational methods to meet those require-
ments. Unfortunately, the methods proposed to day have been too
cumbersome to accomplish the objective, so most risk assessments
prepared today include only a qualitative discussion of uncertainties.
MONTE CARLO METHODS
Monte Carlo simulations yield numberical estmates of uncertain-
ties.1" w Until the recent arrival of powerful desktop workstations,
Monte Carlo simulations were too computationally expensive to have
practical application in public health risk assessments. Now, as work-
stations become readily available, it is appropriate to find efficient ways
to extend risk assessment methods to estimate point values as well as
distributions of health risk.2-3
In the world of Monte Carlo techniques, most or all input variables
become random variables with known or estimated probability density
functions (called PDFs). [Equivalently, an input variable can be speci-
fied by a cumulative distribution function (CDF)]. Within this frame-
work, one or more variables can take on ranges of values with known
probabilities. For example, one could specify that an adult's weight is
distributed as a normal random variable with a mean of 70 kg and a
standard deviation of 10 kg. In this world view, constants, like pi
(approximately 3.14159), remain fixed values.
Until recently, all Monte Carlo simulations were done using custom
software.4-tt u With the arrival of new forecasting software that works
with a spreadsheet, e.g., Crystal Ball1"7 and ©Risk™19 Monte Carlo
calculations now can be designed and implemented as easily as
spreadsheet calculations.
RISK ESTIMATION USING CRYSTAL BALL"
For a Monte Carlo simulation for steady-state or equilibrium condi-
tions, the analyst uses ordinary algebra to describe the governing equa-
tions for souce strength, flow and fate of the contaminants, exposures
and toxicities—all to make a point estimate of the human health risk
in the Risk Characterization step of the risk assessment.
To illustrate the Monte Carlo method with a simplified example.
Exhibit 1 shows a spreadsheet for estimating the health risks to adults
weighing 70 kg who are exposed to eight carcinogenic and eight non-
carcinogenic polycyclic aromatic hydrocarbons (PAHs) over a 70-yr life-
time via the single pathway of chronic inadvertent ingestion of soil
containing 100 mg/kg of each compound. The spreadsheet uses these
formulae to calculate: (1) the estimated Incremental Lifetime Risk and
(2) the estimated Hazard Index from the exposure:
ILR = Cone • Ing * BAF • CPF
BW • l.E+6
HI = Cone • Ing • BAF
BW • RfD • l.E+6
where:
ILR = Incremental Lifetime Risk of Cancer from Exposure
(O^probability^l)
HI = Hazard Index from Exposure
(0 ^ fraction)
Cone = Concentration of the Compound in the Soil
(mg/kg)
Ing = Mass of Contaminated Soil Ingested per Day
(mg/d)
BAF = BioAvailability Factor (relative to water)
(0^ fractional)
CPF = Cancer Potency Factor of the Compound
(inverse mg/(kg»d))
(1)
82 HEALTH & ENDANOERMENT
-------
RfD = Reference Dose for the Compound
(mg/(kg»d))
BW = Body Weight
(kg)
l.E + 6 = factor to make units commensurable
Using reference doses (RfDs) and relative cancer potency factors
recently developed,11'M using the assumption that each person inadver-
tently ingests 100 mg/d of the contaminated soil and using the assump-
tion that the relative bioavailability of the PAHs from the soils is 0.5,
the spreadsheet calculates that a person has a estimated Incremental
Lifetime Risk of cancer of 2.2-03 (probability point estimate) and an
estimated Hazard Index of 2.4E-01 from this single exposure pathway.
These point estimates are interpreted as protective of public health.
Without information on synergisms or antagonisms, the overall Risk
and Hazard Index are estimated by summing the values for each com-
pound across all pathways. Following a short qualitative discussion of
uncertainties inherent in the different variables, most risk assessments
would stop with these point estimates.
The Monte Carlo method continues with several additional steps, all
keyed into the existing spreadsheet. First, the analyst determines (con-
tinuous or discrete) probability density functions12 to describe each
variable included in the uncertainty analysis. In this step, the analyst
must also determine if any correlations exist among the input varia-
bles and make appropriate calculations if they do. Second, using soft-
ware such as Crystal Ball™ the analyst makes a large number (say, 2,000
to 5,000) of "realizations" of the model. Third, the analyst views the
results to establish: (1) the range of results, (2) the shape of the distri-
bution of results and (3) appropriate statistical summaries of the results,
such as the arithmetic average, the median and various quantiles5-6.
In terms of the spreadsheet in Table 1, the Monte Carlo technique
approximates the PDF for the final estimate after assigning PDFs to
some or all of these input variables: (1) the body weight, (2) the volume
of soil inadvertently ingested each day, (3) the relative bioavailability
of the PAH from the soil and/or (4) the CPFs and RfDs. Because the
input variables enter the formulae by multiplication and division (and
subsequent summation), and because some or all of the input varia-
bles may not have normal distributions, the PDF for the final estimate
is, in general, nonGaussian in shape.
SPECIFICATION OF DISTRIBUTIONS
FOR THE INPUT VARIABLES
To illustrate the method, we have estimated the PDFs for the
Incremental Lifetime Risk and Hazard Index for several scenarios using
these assumptions: First, the weight of an adult is normally distributed
with a mean of 70 kg and a standard deviation of 10 kg.24 Second, the
amount of soil that an adult inadvertently ingests each day is lognor-
mally distributed with a log mean of 3 units and a log standard devia-
tion of 1 unit. In keeping with LaGoy,16 this PDF sets the mean
ingestion at 33 mg/day and sets the 93 percentile of ingestion at 100
mg/day. Third, based on professional judgment, the relative bioavaila-
bility is represented by a triangular distribution with vertices at 0.2,
0.5 and 0.6. Fourth, the CPFs and RfDs are independently distributed
as lognormal variates, as discussed in the appendix.
By assumption, each of these distributions is statistically indepen-
dent of the others. Each of these assumptions is reasonable (or not un-
reasonable) in view of the current knowledge and belief. We do not
offer detailed justifications for each of the assumption here because
PAH
Cone
in Soil
(mg/kg)
ICF
Published
Oral
RfD
(mg/kg-d)
Unit
Normal
(rv)
Random
Variate
Oral
RfD
(mg/kg-d)
EPA
Published
Oral
CPF
(mg/kg-d)-1
ICF
Relative
Potency
(ratio)
Unit
Normal
(rv)
(Random
Variate)
Oral
CPF
(mg/kg-d)- 1
Estimated
Bioavailable
ADD (life)
(mg/kg-d)
Estimated
Hazard
Index
(frac)
Estimated
Incremental
Lifetime
Risk
(frac)
PAH Compounds Considered Potentially Carcinogenic
benzo(a)pyrene 100 1.0E-02 0.0
benzo(a)anthracene 100 1.0E-02 0.0
benzo(b)fluoranthene 100 1.0E-02 0.0
benzojkjfluoranthene 100 1.0E-02 0.0
indeno(1,2.3-cd)pyrene 100 1.0E-02 0.0
chrysene 100 1.0E-02 0.0
dibenzo(a,h)anthracene 100 1.0E-02 0.0
benzo(ghi)perylene 100 1.0E-02 0.0
PAH Compounds Not Considered Potentially Carcinogenic
naphthalene
fluorene
anthracene
phenanthrene
fluoranthene
pyrene
acenaphthylene
acenaphthene
100
100
100
100
100
100
100
100
1.0E-02
1.0E-02
1.0E-02
1 .OE-02
1.0E-02
1.OE-02
1.OE-02
1.OE-02
5.0E-03
5.0E-03
5.6E-04
7.0E-03
2.0E-02
1.5E-02
1. OE-02
2.0E-01
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.0E-03
5.0E-03
56E-04
7.0E-03
2.0E-02
1 .5E-02
1 .OE-02
2.0E-01
1.15E+01
1.00E+00
1.45E-01
1.40E-01
6.60E-02
2.32E-01
4.40E-03
1.11E+00
2.20E-02
0.0
0.0
0.0
0.0
0.0
0.0
0.0
00
1.15E+01
1.67E+00
1.61E+00
7.59E-01
2.67E+00
5.06E-02
1 .28E+01
2.53E-01
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.14E-05
7.1E-03
7.1E-03
7.1E-03
7.1E-03
7.1E-03
7.1E-03
7.1E-03
7.1E-03
1.4E-02
1.4E-02
1.3E-01
1.OE-02
3.6E-03
4.8E-03
7.1E-03
3.6E-04
8.2E-04
1.2E-04
1.2E-04
5.4E-05
1.9E-04
3.6E-06
9.1E-04
1.8E-05
Assumptions:
Adult
Weight
(kg)
Soil
Ingestion
(mg/d)
Bioavail
(ratio)
Toggle NC
(OID
Toggle C
Sums --» 2.4E-01 2.2E-03
70.0
100.0
0.5
0.0
0.0
Table 1
Sample Spreadsheet for Estimating Health
Effects from Ingesting Soil Contaminated
with PAHs
HEALTH & ENDANGERMENT 83
-------
our primary focus is to demonstrate a new computational framework.
At this time, we investigate the effects of these assumptions without
further justification. However, we have additional research underway
to refind, support and document the assumptions.
RESULTS
Printed from Crystal Ball1", Figure 1 shows the histogram of esti-
mated risk with the assumptions that body weight, soil ingestion and
bioavailability are represented by the PDFs in the previous section and
that the CPFs and RfDs are point values. Similarly, Figure 2 shows
the histogram of estimated risk with the assumptions that body weight.
soil ingestion, bioavailability, CPFs, and RfDs are all random variates
described by the PDFs in the previous section.
Forecast: Sum ol Risk
Summary: Confidence Level Is 1 .OOe+2% based on Entire Range
Confidence Range is from —10 *—
Display Range Is from 0 OOe+0 to 1 .OOe-2
Entire Range Is from \ .16e-5 lo 1 31 e-2
Alter 3e«3 Trials, the Std. Error ol the Mean Is 1 72e-5
Statistics
Trials
Percent ol Other
Mean
Median
Mode
Standard Deviation
Variance
Skewness
Kurtosis
Range Width
Range Minimum
Range Maximum
Mean Std. Error
Display Ranoe
3e+3
9.99e+t
685e-4
3.97B-4
8.33e-S
89Se-4
S.OOe-7
3.91e+0
2.496*1
1.OOe-2
O.OOe+0
1 .OOe-2
1 63e-5
Entire Ranoa
3e+3
1 OOe+2
6 938-4
(unavailable)
(unavailable}
9 43e-4
8 89e-7
(unavailable)
(unavailable)
1.30e-2
1 160-5
1.31e-2
1.72e-S
Forecast: Sum of Risk
Frequency Distribution
2,998 Trials
' 176
(t
a
c
(0
2.50e-3
S.OOe-3
7.50B-3
1.OOe-2
Figure 1
Histogram for Estimated Incremental Lifetime Cancer Risk
with Three Exposure Variables as Random Variates
These two graphs show the results of several thousand simulations
to quantify the uncertainties. For the type of health risk calculations
investigated here, Monte Carlo simulations with inputs described by
random variables yield strongly nonGaussian distributions for estimated
health risk. As the number of random inputs increases, the histogram
for the health risk becomes increasingly nonGaussian and the relative
standard deviation increases.
Table 2 presents statistics for the deterministic case and five probabilis-
tic cases, demonstrating the effects of turning one input at a time into
a random variable. For example, the 95-percentile estimate of the overall
cancer risk in the last numerical column is 6.75E-04, less than the con-
servative point estimate of 2.20E-03 in the deterministic case in the
first numerical column.
Based on theoretical considerations, on the practical experience and
on the simulations reported here, we find that the greatest uncertainty
in the shape and location of the PDF for estimated human health risk
comes from the uncertainties in the shapes and positions of the PDFs
for toxicities.
Forecast: Sum ol Risk
Summary Confidence Level Is 1 00e*2% based on Entire Range
Confidence Range Is Irom —to «~
Display Range Is Irom O.OOetO to 1 OOe-2
Entire Range Is Irom 7.76e-7 to 4.87«-3
Alter 5e+3 Trials, the Sid Error ol the Mean Is 4 38e-6
Statistics
Trials
Percent ol Other
Mean
Median
Mode
Standard Deviation
Variance
Skewness
Kurtosis
Rang* Width
Range Minimum
Range Maximum
Mean Std Error
5e+3
1 00e*2
1.83C-4
8840-5
1 67e-S
3 10e-4
9600-8
543e*0
4830.1
1.OOe-2
O.OOe+0
1 OOe-2
4380-6
Entire Rar^m
5o»3
1 00e»2
1.830-4
8.840-5
3100-4
9-600-8
S43o«0
483e»1
4.87e-3
7.76e-7
4.87e-3
4.380-6
Cell rv
2.1408
TJi
a
e
0
tL.
0.0'
o.c
Forecast: Sum of Risk
38 Frequency Distribution 5,000
Oe
L
>-•-•-- 4
.0 2.506-3 S.OOe-3 7.5063 1.00e
Trials
1070
^i
|
e
9
9
n
1C
0
2
Figure 2
Histogram for Estimated Incremental Lifetime Cancer Risk
with Three Exposure Variables and T\ro Toxicities as Random Variates
Table 2
Summary of Scenarios for Cancer Risk from
Ingesting Contaminated Soil
. wi urn wt
NO* MI
U4LWUH
TBJ.UUI
iioe -as
uoc-oa
a* • I3o4-m
i«
<« T TttJ*
IMC4M t«C-M
1QU41 II3C-04
1.000 1000
APPENDIX: GENERIC DISTRIBUTIONS FX)R CPft AND RfDs
Carcinogens: For their CPFs for compounds tested in small mammals,
the U.S. EPA commonly sets the value as the 95% upperbound of the
slope, scaled to human adults, of the linearized multistage model relating
the dose administered in the laboratory and the lexicological response
in test mammals." From Agency publications, it is not possible to
infer the underlying probability distributions from which the published
CPFs represent the 95% upper confidence limits. In the absence of data
or knowledge on the shape of the underlying distribution for CPFs in
humans, it is possible to hypothesize a variety of distributions, one of
which is investigated in this paper.
The lognormal model for a generic CPF distribution is based on
research by Crouch and his colleagues' "•' and on the often spoken
statement that the uncertainty in the variate may be as large as a factor
84 HEALTH & ENDANGERMENT
-------
of 10 above the central measure and as low as a factor of 10 below the
central measure of the distribution. This suggests a lognormal model
for the underlying distribution. By fixing two standard deviations of
the logarithm of the random variate at 10, and by scaling the distribu-
tion so the 95% fractile of the cumulative distribution function falls
at the published CPF value, this function has the appropriate properties:
xl ~(CPF / 6.645) • exp [ 1.1513 • N(0,l) ]
(3)
Non-Carcinogens: Similarly, the U.S. EPA commonly establishes
RfDs for compounds that are one, two, three or four orders of magni-
tude below NOAEL values from animal experiments. One of the factors
of 10 accounts for inter-individual variability in susceptibility. Hattis
and his co-workers13 have found that some inter-individual suscepti-
bilities are distributed lognormalry. On the assumption that four standard
deviations of the logarithm of susceptibility equal a factor of five and
that two standard deviations above the mean of the logarithm of the
susceptibility fall at the factor of 10 used by the U.S. EPA, this func-
tion has the appropriate generic properties:
x2- (2.236 • RfD) / exp [ 0.402 • N(0,l) ]
(4)
ACKNOWLEDGEMENTS
We thank many people for helpful discussions during this research:
E.A.C. Crouch, D.B. Hattis, T.E. McKone, C.A. Menzie, B.W. Schwab,
A.C. Taylor, B.C. Udell and E.W. Wainwright. Any mistakes are our
own. The Gas Research Institute provided partial support for this
research.
REFERENCES
1. Allen, B.C., Shipp, A.M., Crump, K.S., Kilina, B., Hogg, M.L., TUdor,
J. and Keller, B., "Investigation of Cancer Risk Assessment Methods," Final
Report Summary, prepared for the U.S. EPA by Clement Associates, Inc.,
Jan. 1987.
2. Burmaster, D.E. and von Stackelberg, K., "Monte Carlo Simulations of Un-
certainties in Risk Assessments of Superfund Sites Using Crystal Ball,™
Pmc. of the 1989 Nat. Conf. on Environmental Engineering, ASCE, July 1989.
3. Burmaster, D.E. and von Stackelberg, K., "A New Method for Uncertainty
and Sensitivity Analysis in Public Health Risk Assessments at Hazardous
Waste Sites Using Monte Carlo Techniques in a Speadsheet," Superfund
'88, Proc. of the 9th Nat. Conf., Washington, DC, HMCRI, Silver Spring,
MD; Nov. 1988, 550-556.
4. Campbell, J.E. and Cranwell, R.M., "Performance Assessment of Radio-
active Waste Repositories, Science," Mar. 18, 239, 1389-1392
5. Chambers, J.M., Graphical Methods for Data Analysis, Wadsworth Inter-
national Group, Belmont, CA, 1983.
6. Cleveland, W.S., The Elements of Graphing Data, Wadsworth Advanced
Books and Software, Monterey, CA, 1985.
7. Crouch, E.A.C., Wilson, R. and Zeise, L., "The Risks of Drinking Water,"
Water Resources Research, 791983, 1359-1375.
8. Crouch, E.A.C., "Uncertainties in Interspecies Extrapolations of Carcinoge-
nicity," Environ. Health Persp. 50, 1983, 321-327.
9. Crouch, E. and Wilson, R., "Regulation of Carcinogens," Risk Analysis,
1, 1981, 47-57.
10. Eschenroeder, A.Q. and Faeder E.J., "A Monte Carlo Analysis of Health
Risks from PCB-Contaminated Mineral Oil Transformer Fires," Risk
Analysis, 8, 1988, pp 291-297.
11. Environ Corporation, " Potential Health Risks from Former Manufactured
Gas Plans Sites: Toxicity and Chemical Profiles." Prepared for Pacific Gas
and Electric Company, Southern California Edison, and the Southern Califor-
nia Gas Company, 1986.
12. Hastings, N.A.J. and Peacock, J.B., "Statistical Distributions: A Handbook
for Students and Practitioners," Butterworth & Company, London, 1974.
13. Hattis, D., Erdreich, L. and Ballew, M., "Human Variability in Suscepti-
bility to Toxic Chemicals—A Preliminary Analysis of Pharmacokinetic Data
from Normal Volunteers,"S«fc Analysis, 7, 1987, 415-426.
14. ICF-Clement Associates, Inc, "Comparative Potency Approach for Estimation
of the Total Cancer Risk Associated with Exposures to Mixtures of Poly-
cyclic Aromatic Hydrocarbons in the Environment," Final Report, Washing-
ton, DC, 1987.
15. Iman, R.L. and Helton, J.C., An Investigation of Uncertainty and Sensi-
tivity Analysis Techniques for Computer Models," Risk Analysis, 8, 1988,
71-90.
16. LaGoy, P.K., "Estimated Soil Ingestion Rates for Use in Risk Assessment,"
Risk Analysis, 7, 1987, 355-359.
17. Market Engineering Corporation, 1988, Crystal Ball™ User's Guide, Suite
600, 1675 Larimer Street, Denver, CO, 1988.
18. Morgan, B.J.T., Elements of Simulation. Chapman and Hall, London, 1984.
19. Palisade Corporation, Newfield, NY, 1988.
20. Rubenstein, R.Y.,Simulation and the Monte Carlo Method, John Wiley &
Sons, New York, NY, 1981.
21. U.S. EPA, "Exposure Factors Handbook, Final Report," Office of Health
and Environmental Assessment, EPA/660/8-89/043, March 1989.
22. U.S. EPA, Superfund Public Health Evaluation Manual, EPA/540/1-86/060,
OSWER Directive 9285.4-1, Oct. 1986.
23. U.S. EPA, Superfund Exposure Assessment Manual, OSWER Directive
9285.5-1, Dec. 1986.
24. U.S. Department of Health, Education, and Welfare, "1979, Weight and Height
of Adults, 18-74 Years of Age, United States, 1971-1974," Office of Health
Research, Statistics, and Technology, Washington, DC, May 1979.
25. Whitemore, R.W., "Methodology for Characterization of Uncertainty in Ex-
posure Assessments," a report from the Research Triangle Institute to the
Office of Health and Environmental Assessment, U.S. EPA,
EPA/600/8-85/0009, Aug. 1985.
HEALTH & ENDANGERMENT 85
-------
Use of a Retention Index System to Better Identify
Non-Target Compounds
William P. Eckel, M.S.
Thomas A. Jacob, Ph.D.
Viar and Company
Alexandria, Virginia
Joan F. Fisk
U.S. EPA
Washington, D.C.
ABSTRACT
A thorough assessment of the health risks posed by hazardous waste
sites requires that the chemical pollutants present be well characterized.
Because the typical "Target Compound List" analysis provides for the
specific determination of only 126 organic compounds, the identification
of any other chemicals amenable to GC/MS analysis depends upon them
being reported as "Tentatively Identified Compounds" (TIC). Proper
identification of such TICs can be critical to the completeness of site-
specific risk analysis.
TICs are non-target compounds found during a GC/MS run, which
are identified solely by a reverse search of their mass spectra versus
the NIST/EPA/MSDC mass spectral library. Because no use currently
is made of GC retention time data in identifying TICs, the identifi-
cations are less accurate than for target compounds. Given the increasing
interest in using TIC data in risk analysis and other Superfund-related
activities, the Contract Laboratory Program (CLP) has begun to im-
prove the process by which TICs are identified. The first step will be
to make better use of the GC retention time data.
This paper reports the analysis of GC retention time data for TICs
in the CLP Analytical Results Database (CARD). CARD is the com-
puter data base in which organic and inorganic analysis results generated
by CLP Laboratories for Superfund are stored. We have applied the
retention index (RI) system of Lee et al.1 to the semi-volatile TIC data
in CARD in order to validate the TIC data and to test the RI System
for use in TIC data reporting and review. Lee's RI system is based on
polycyclic aromatic hydrocarbons (PAH) where the index values are
naphthalene = 200, phenanthrene = 300, chrysene = 400 and picene
= 500.
CARD has data on the GC retention times of TICs and on those of
naphthalene-d8, phenanthrene-dlO chrysene-d!2 and perylene-d!2 which
are internal standards added to each sample. This database enabled us
to directly calculate RIs for all TICs which eluied between naphthalene-
d8 and perylene-d!2. Preliminary comparison of the RI data from the
TICs to the RI values reported by Lee, et al., shows excellent agree-
ment. Retention indices for non-target PAH were generally within five
points of the expected values. Many non-PAH compounds also showed
statistically well-behaved RIs which agreed with those from U.S. EPA
method 1625C.
INTRODUCTION
The organic chemical analysis methods currently specified by
Superfund's Contract Laboratory Program (CLP) provide for the
analysis of 126 target compounds (the "Target Compound List" or TCL)
by gas chromatography (GC) and GC-mass spectrometry (GC-MS).
Up to 30 "tentatively identified compounds" (TICs) per sample musi
also be reported by comparison of the mass spectra of non-TCL peaks
in the GC-MS chromatograms to the approximately 40JOOO to SOjOOO
mass spectra in the NIST/U.S. EPA/MSDC data base. Because no ac-
tual chemical standards are routinely used to confirm the identity of
TICs and because the amount of time that can be devoted to spectral
interpretation in commercial, production-oriented laboratories is limited,
both the identity and concentrations of reported TICs are far less accurate
than they are for the target compounds.
CLP management recognizes that the proper identification and
reporting of tentatively identified compounds is becoming a more im-
portant issue. For example, thorough risk assessment al hazardous waste
sites depends on the proper identification of potentially toxic compounds.
Also, TIC data are being used to fulfill studies mandated under the
Superfund reauthorization, such as the listing of the 275 most common
toxic substances found at waste sites, which is being conducted by me
Agency for Toxic Substances and Disease Registry. TIC identifications
must be reliable if policy decisions are to be made using them.
To address the need for better TIC data reporting, the CLP estab-
lished the Tentatively Identified Compounds Improvements Workgroup
at its Organics Conference in October, 1988. The Workgroup is respon-
sible for devising methods for more reliable identification of non-target
compounds; it consists of members from the U.S. EPA and other govern-
ment research laboratories and U.S. EPA Regions, the laboratory com-
munity, instrument manufacturers and other interested parties.
One area being explored by the Workgroup is the use of Retention
Index (RI) systems based on the use of GC retention time data. Under
a given set of conditions, the Retention Index is a predictable charac-
teristic of a given compound and can be used to identify it. The use
of Retention Indices is particularly attractive, since GC retention time
data for TICs currently are reported by CLP laboratories in computer-
readable form (diskette) and since no systematic use currently is made
of GC retention time data to identify TICs. Thus the number of pieces
of data used to identify TICs can easily be increased from one (mass
spectrum) to two (mass spectrum and retention index).
This paper will consider the use of a Retention Index system based
on Polycyclic Aromatic Hydrocarbons (PAH) to evaluate data from semi-
volatiles analysis.
METHODS
The PAH Retention Index system was first proposed in 1979 by Lee.
et al.,' and was extended by Willey, et al.,1 and Vassilaros, et al.,'-
Whalen-Pederson and Jurs4 devised a system to predict the Retention
Index of a PAH using molecular structure descriptors. More recently,
the prediction of Retention Indices by multivariate regression analysis
of molecular structure descriptors has been extended to mononitrated
PAHs and polychlorinated biphenyls by Robbat and co-workers54.
The PAH Retention Index system is based on naphthalene (RI=200),
86 HEALTH & ENDANGERMENT
-------
Compound
Table 1
Polycyclic Aromatic Hydrocarbons and Heterocyclics Comparison of
Literature Retention Indices and Tentatively Identified
Compound Retention Indices
CAS No. Number Llter_RI Median Dlfferl Mean
Differ2 STD_DEV Minimum Maximum
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
Benzothiophene
IH-Indole
Isoquinoline
Azulene
Naphthalene, 1-methyl-
Quino 1 ine , 2 - me thy 1 -
l,l'-Biphenyl
Naphthalene, 2- ethyl -
Naphthalene, 1- ethyl -
Naphthalene, 2, 6 -dimethyl -
Naphthalene, 2, 7 -dimethyl -
l,l'-Biphenyl, 2-methyl-
Naphthalene, 1, 3 -dime thy 1-
Naphthalene, 1, 7 -dime thyl -
naphthalene, 1, 6 -dimethyl -
Naphthalene , 2,3- dimethyl -
Naphthalene, 1, 4 -dime thyl -
Naphthalene, 1 , 5 - dimethyl -
Naphthalene, 1, 2 -dime thy 1-
Naphthalene , 1 , 8 - dime thy 1 -
1,1'- Biphenyl , 4 - methyl -
Naphthalene, 2,3,6-trimethyl-
9H-Fluorene, 9-methyl-
5H- Indenopyr idine
Anthracene , 9,10- dihydro -
9H-Fluorene, 2-methyl-
9H-Fluorene, 1-methyl-
9H-Fluoren-9-one
Dibenzothiophene
Acridine
9H-Carbazole
Dibenzothiophene, 4-methyl-
Naphthalene , 1 - phenyl -
Fhenanthrene , 3 - me thy 1-
Fhenanthr ene , 2 - me thyl -
Anthracene, 2-methyl-
Terphenyl
4H-Cyclopentaphenanthrene
Phenanthr ene , 9 - me thyl -
Phenanthrene , 4-methyl-
Anthracene , 1 - me thyl -
Phenanthrene , 1 - me thyl -
9 , 10-Anthracenedione
Naphthalene, 2- phenyl -
Phenanthrene , 3 , 6 -dime thy 1-
Phenanthrene, 2,7-dimethyl-
9-Anthracenecarbonitrile
HH-Benzofluorene
Phenanthrene , 1 - me thyl - 7 -
(1-methylethyl) -
HH-Benzofluorene
Pyrene, 4 -me thyl -
Pyrene, 2 -me thyl -
Pyrene, 1 -me thyl -
Benzonaphtho<2 , 1 - d>thiophene
Benzofluoranthene
Benzophenanthrene
Benzonaphthothiophene
Cyclopentapyrene
Iriphenylene
Benzanthracene , 11-methyl-
Benzanthracene , 1 -methyl -
Benzanthracene , 8 -me thyl -
Chrysene, 3 -me thyl -
Benzanthracene , 12-methyl-
Chrysene, 5 -methyl -
Benzanthracene , 7 -methyl -
2,2'-Binaphthalene
Benzof luoranthene
Benzopyrene
95158
120729
119653
275514
90120
91634
92524
939275
1127760
581420
582161
643583
575417
575371
575439
581408
571584
571619
573988
569415
644086
829265
2523377
244995
613310
1430973
1730376
486259
132650
260946
86748
7372885
605027
832713
2531842
613127
26140603
203645
883205
832644
610480
832699
84651
612942
1576676
1576698
1210124
238846
483658
243174
3353126
3442782
2381217
239350
203123
195197
205436
27208373
217594
6111780
2498773
2381319
3351313
2422799
3697243
2541697
612782
205823
192972
38
3
5
3
155
8
71
15
9
3
4
7
24
19
18
38
14
24
33
40
9
20
5
11
3
7
10
13
91
3
56
3
5
19
29
58
3
58
18
40
30
7
59
26
9
3
4
135
10
39
55
18
42
15
33
47
20
3
3
5
7
3
20
12
3
11
4
149
48
201.47
205.26
215.61
219.95
221.04
224.13
233.96
236.08
236.56
237.58
237.71
238.77
240.25
240.66
240.72
243.55
243.57
244.98
246.49
249.52
254.71
263.31
272.38
279.31
284.89
288.21
289.03
294.79
295.81
304
312.13
312.72
315.19
319.46
320.17
321.57
321.99
322.08
323.06
323.17
323.33
323.9
330.53
332.59
337.83
339.23
350.6
366.74
368.67
369.39
369.54
370.15
373.55
389.21
389.6
391.39
392.92
396.54
400
412.72
414.37
417.56
418.1
419.39
419.68
423.14
423.91
440.92
450.73
202.28
222.28
210.04
200.67
225.42
224.58
237.5
239.67
240.17
241.18
241.67
275.3
244.03
243.65
241.85
244.17
245.9
241.33
244.96
244.43
273.77
262.64
289.77
309
285.09
288.62
288.48
294.4
296.68
309.53
309.36
312.79
312.92
320.35
319.62
320.24
385.9
323.38
321.25
323.22
320.07
320.35
331.2
330.68
337.93
333.93
360.07
367.09
366.71
367.19
370.98
370.8
367.17
390.43
391.97
391.86
390.28
392.12
403.95
413.8
416.67
407
416.07
417.33
414.14
415.37
424.11
451.24
450.96
-0.81
-17.02
5.57
19.28
-4.38
-0.45
-3.54
-3.59
-3.61
-3.6
-3.96
-36.53
-3.78
-2.99
-1.13
-0.62
-2.33
3.65
1.53
5.09
-19.06
0.67
-17.39
-29.69
-0.2
-0.41
0.55
0.39
-0.87
-5.53
2.77
-0.07
2.27
-0.89
0.55
1.33
-63.91
-1.3
1.81
-0.05
3.26
3.55
-0.67
1.91
-0.1
5.3
-9.47
-0.35
1.96
2.2
-1.44
-0.65
6.38
-1.22
-2.37
-0.47
2.64
4.42
-3.95
-1.08
-2.3
10.56
2.03
2.06
5.54
7.77
-0.2
-10.32
-0.23
202.5
221.61
210.88
200.68
225.27
224.6
237.34
242.08
240.24
242.35
242.55
275.15
243.74
243.18
241.65
244.18
246.62
241.47
245.77
245.02
274.15
263.31
289.52
309.16
286.63
288.93
288.41
294.74
296.67
308.76
309.34
312.81
329.73
320.53
320
320.48
385.59
323.62
320.99
322.07
320.52
321.16
330.89
330.74
338.92
336.4
360
366.92
366.2
366.79
371.35
370.24
368.02
390.43
392
395.87
389.88
394.18
403.07
413.64
417.64
406.36
416 . 19
417.96
414.33
414.77
426.39
450.21
450.40
-1.03
-16.35
4.73
19.27
-4.23
-0.47
-3.38
-6
-3.68
-4.77
-4.84
-36.38
-3.49
-2.52
-0.93
-0.63
-3.05
3.51
0.72
4.5
-19.44
0
-17.14
-29.85
-1.74
-0.72
0.62
0.05
-0.86
-4.76
2.79
-0.09
-14.54
-1.07
0.17
1.09
-63.6
-1.54
2.07
1.1
2.81
2.74
-0.36
1.85
-1.09
2.83
-9.4
-0.18
2.47
2.6
-1.81
-0.09
5.53
-1.22
-2.4
-4.48
3.04
2.36
-3.07
-0.92
-3.27
11.2
1.91
1.43
5.35
8.37
-2.48
-9.29
0.33
1.27
1.15
2.28
0.13
1.64
2.09
1.05
4.87
0.61
2.13
1.85
0.95
1.95
1.61
1.52
1.62
2.57
0.85
2.23
2.32
1.14
1.83
0.57
0.99
2.75
1.07
0.32
0.76
0.33
1.33
0.87
0.11
23.47
1.52
0.82
1.66
1.16
1.27
1.49
2.2
1.46
3.63
1.02
0.52
2.23
4.67
0.38
2.15
1.47
2.01
2.85
2.61
2.67
0.41
0.69
5.74
1.29
3.91
1.75
0.98
3.51
1.47
2.39
2.2
0.73
1.39
4.68
3.37
2.26
202.01
220.29
208.66
200.56
219.06
222.28
233.72
234.52
239.54
241.06
241.54
273.67
239.92
240.51
238.73
240.75
242.7
240.35
241.6
241.11
272.89
261.08
288.51
307.78
285
287.38
287.85
293.91
295.34
307.22
306.82
312.72
312.32
318.58
319.16
316.6
384.31
319.05
318.99
316.73
318.34
315.82
327.45
329,91
335.79
333.48
359.51
362.61
364.42
362.17
363.64
366.33
363.21
389.96
390.93
390
386.32
391.72
401.05
412.48
412.22
404.69
412.82
415.45
413.73
412.23
423.93
442.68
444.06
210.04
222.28
214.68
200.82
234.52
228.98
238.82
250.82
241.54
244.8
245.32
276.07
250.5
246.51
244.15
250.84
250.98
244.17
250.4
250.98
275.78
267.2
289.93
310.89
289.81
290.64
288.82
296.32
298.2
309.53
312.09
312.93
356.64
324.64
322.24
324.58
386.57
329.46
324.15
327.32
323.61
326.84
333.98
331.82
342.92
341.79
360.34
372.72
368.23
370.07
375.47
374.64
374.39
391.65
394.83
404.69
391.65
398.69
404.19
414.73
422.15
407.41
421.77
421,33
415.14
416.04
433.41
455. 59
453!79
HEALTH & ENDANGERMENT 87
-------
phenanthrene (RI=300), chrysene (RI=400) and picene (RJ=500). The
Retention Index is calculated by interpolation between the bracketing
standards:
Rl = 100 x
^UNK
+ 100(Z)
(1)
where RTZ and RTZ+| are the retention times of the standards before
and after the unknown and Z is 2 for naphthalene, 3 for phenanthrene
and 4 for chrysene.
The semi-volatile analysis method used by the CLP (based on U.S.
EPA Method 625) employs the perdeuterated analogues of naphthalene,
phenanthrene and chrysene as internal standards; perdeuterated pery-
lene (RI-456.22) is also an internal standard. These internal standards
are added to all semi-volatile samples for quantilation purposes. The
retention times of all of these internal standards, as well as of all TICs
and all other results and QC data, are reported on floppy diskette by
CLP organics laboratories. Prior to uploading the diskette data to the
CLP Analytical Results Database (CARD), the data are stored in SAS
(Statistical Analysis System) files on the U.S. EPA mainframe com-
puter at Cincinnati, Ohio. An extract was made from these files which
contained the Chemical Abstracts Service Registry number (CAS
Number) and retention time of all semi-volatile TICs reported and the
retention times, from the same samples, of the four internal standards
mentioned. Only data for which valid CAS numbers was reported was
retained. Retention Indices were then calculated using Equation 1 and
the four perdeuterated internal standards as retention index markers.
For compounds eluting between chrysene and perylene. Equation 1 was
modified so that the retention time term was multiplied by 56.22, not
100. The calculated Retention Indices were analyzed using the SAS
procedure UNTVARIATE and were compared to the values in references
1 to 3. Table 1 shows the result of this analysis, including the name
and CAS number of reported PAHs and heterocyclic compounds, the
number of times each was reported as a TIC, the trimmed mean (±2
standard deviations) of the Retention Index, the standard deviation of
the RI, the median, minimum and maximum RI, and for comparison,
the literature value of the RI.
RESULTS
PAH and related compounds
Of 878 semi-volatile TICs reported in the CARD data (through Aug. 1,
1989) which were reported to elute between naphthalene and perylene,
69 PAHs had RI values reported in the literature.13 Agreement be-
tween the literature values for RIs and the means and medians from
the CARD data is quite good, especially considering that the RI sys-
tem was not originally used in reporting the data. This result may be
due in part to the (act that condensed aromatic compounds such as these
have mass spectra with strong molecular ions, which would tend to make
library searching more reliable. Inspection of Table 1, which is sorted
by the literature value of the RI, shows that similar compunds (e.g.,
ethylnapththalenes/dimethylnaphthalenes or methylanthra-
cenes/methylphenanthrenes) have not been completely distinguished
from each other by library matching alone. The mean and median RI
values calculated from CARD for compounds in these groups appear
to be average values for the entire groups. If the retention indices for
such groups are plotted, multi-modal distributions indicative of the
presence of several compounds are obtained, as in Figure 1. On the
other hand, 1-phenylnaphthalene (33) and 2-phenylnaphthalene (44) are
nicely distinguished.
There are eight PAH compounds for which agreement between the
literature RI values and those calculated from CARD is poor. These
are 5H-indeno (1,2-b) pyridine, 4-methylbiphenyl, 2-melhylbiphenyl,
9-methylfluorene, indole, 8-methylbenz(a)anthracene, terphenyl and
azulene. The possible reasons for the poor agreement include errors
in library matching, variations in the initial GC oven temperature and
changes in the chemical nature of the stationary phase with extended
use. It also appears that azulene may be confused with napthalene, and
234 235 236 23? 23« 23> 240 241.242 243.244 245 246 247.24« 248 250 251
RETEHDON MDEX
Figure 1
Retention Indices for C2-naphthalenes
indole with methylquinoline.
Regression analyses of the literaure RI values for PAHs versus the
median RIs from CARD were conducted using the SAS procedure PROC
REG. When the regression was performed on the median RI of all 69
compounds versus the literature values, the explained variance was
97.55%. Removel of the eight compounds, mentioned above, for which
the residual was over 10 index points, resulted in an explained variance
of 99.75%. Clearly there is a strong relationship between the literature
RI values and those calculated from CARD. This suggests that most
of the 69 compounds in Table I have been correctly identified, at least
at the structural isomer level.
These results show that systematic application of the PAH Retention
Index system, by contract laboratories, in conjunction with mass spectral
library searching, might result in greatly enhanced qualitative identifi-
cation of non-target compounds. Better quantilation would depend on
methods for better estimating calibration response factors.
Non-PAH Compounds
The utility of this RI system would be greatly improved by extending
it to non-PAH compounds. Retention time information for many com-
pounds which are not on the CLP Target Compound List has been
published in the Office of Water's method 1625C* and in method 525
for drinking water" Retention indices were calculated from the
method 1625 data for 23 compounds and compared to those from the
CARD data; the results are presented in Table 2.
The poor RI matching for some of the normal alkanes is probably
a result of poor library matching due to the similarity of all alkane spec-
tra. In contrast to PAHs, alkanes have a very weak or no molecular
ion, with u characteristic "hydrocarbon" spectrum which does not van
much for alkanes above hexane. Note, however, that the retention Index
from method 1625C is within the reported range from CARD for all
of the alkanes.
Other compounds in Table 2 show better agreement between the
method 1625C RI and the values from CARD. The exceptions are
1,2,4,5-tetrachlorobenzene and squalene, both of which suffer from
small samples, and 7H-benzanthracen-7-one{benzanthrone). The
PAH Retention Index system ought to be as useful for non-PAH com-
pounds as is any other system; example of such a system is one based
on n-alkanes, as is the Kovats index The traditional Kovats index is
not applicable to the CLP semi-volatile method, since the CLP method
is a temperature-programmed GC method, whereas the Kovats ind«
is used for isothermal GC methods. Retention time data from Revision
1 (12/87) of RCRA method 8270" were examined and were found to
be in conflict with method 1625C and the CARD data. The use of
method 8270 as a reference for retention time data is not recommended.
Table 3 gives the RI of other method 1625C compounds which were
not found in the CARD TIC data, and Table 4 gives RI date from the
drinking water method 525. with means and medians from CARD.
Agreement between method 525 and the CARD data is quite good.
HEALTH & BNDANOERMENT
-------
Compound
alpha-terpineol
Dodecane
Benzene, 1,2,3-trichloro-
1,3-Benzodioxole, 5-(2-propen
Benzene, 1,2,4,5-tetrachloro-
1,3-Benzenediamine, 4-methyl-
Tetradecane
Benzene, l.l'-oxybis-
longlfolene
2,6-di-t-butyl-p-benzoquinone
Benzene, pentachloro-
2-Naphchalenamine
Benzothlazole, 2-(methylthio)
Hexadecane
Octadecane
Eicosane
Benzidine
Docosane
Tetracosane
Squalene
7H-Benzanthracen- 7 - one
Hexacosane
Octacosane
Table!
Miscellaneous Compounds from Method 1625C
Comparison of Method 1625C Retention Indices
And Tentatively Identified Compound Retention Indices
CAS No. Number RI-1625C Median
Mean
Maximum
98555
112403
87616
94597
95943
95807
629594
101848
475207
719222
608935
91598
615225
544763
593453
112958
92875
629970
646311
7683649
82053
630013
630024
5
44
6
10
2
3
55
17
3
15
9
2
3
217
65
81
18
87
29
3
53
51
140
201.95
202.93
206.5
220.65
228.94
236.42
239.02
241 . 14
242.6
250.41
261.3
266.34
273.5
282.28
300.4
319.68
354.67
361.83
388.87
391.65
405.24
413.85
440.07
200.95
202.57
206.17
221.27
237.25
237.2
238.28
240.12
242.92
250.77
261.14
264.44
273.65
272.96
297.42
331.89
350.73
362.09
381.45
439.55
386.68
403.82
421.01
1
0.36
0.33
-0.62
-8.31
-0.78
0.74
1.02
-0.32
-0.36
0.16
1.9
-0.15
9.32
2.98
-12.21
3.94
-0.26
7.42
-47.9
18.56
10.03
19.06
202.02
211.02
206.18
221
237.25
237.46
246.27
240.86
244.2
250.89
261
264.44
273.85
293.46
298.06
339.46
350.86
380.24
383.42
438.85
389.18
355.44
391.74
-0.07
-8.09
0.32
-0.35
-8.31
-1.04
-7.25
0.28
-1.6
-0.48
0.3
1.9
-0.35
-11.18
2.34
-19.78
3.81
-18.41
5,45
-47.2
16.06
58.41
48.33
2.56
14.55
0.17
1.16
0.48
0.89
22.71
1.56
2.23
0.78
0.65
1.57
0.48
42.12
5.72
25.54
1.05
37.05
5.57
4.75
3.81
70.91
56.62
200.71
200.58
205.98
219.5
236.9
236.73
235.31
239.3
242.89
249.88
259.98
263.33
273.51
239.98
276.67
303.52
348.96
302.92
375.04
433.8
382.79
216.13
268.1
206.6
258.54
206.4
222.31
237.59
238.45
317.66
244.89
246.78
252.17
262.05
265.56
274.4
413.27
325.14
420.75
352.96
455.11
394.07
443.21
396.82
434.7
454.84
CAS No.
Table3
Retention Indices Of Other Method 1625C Compounds
Compound RIa
Table 4
Retention Indices Of Method 525 Compounds
614-00-6
1888-71-7
121-73-3
700-12-9
108-46-3
137-17-7
120-75-2
95-79-4
634-36-6
608-27-5
3209-22-1
130-15-4
2027-17-0
100-25-4
99-30-9
134-32-7
96-45-7
89-63-4
99-55-8
103-33-3
122-39-4
62-44-2
92-67-1
23950-58-5
882-33-7
92-93-3
86-74-8
2243-62-1
91-80-5
92-84-2
7700-17-6
492-22-8
60-11-7
101-14-4
119-90-4
72-33-3
87-65-0
933-75-5
58-90-2
1689-84-5
N-nicrosomethylphenylaraine
hexachloropropene
3-chloronitrobenzene
pen tame Chylbenzene
1. 3-benzenedlol
2,4.5-trimethylanlline
2-methylbenzothiazole
5-chloro-o- toluldine
1,2,3- crime Choxybenzene
2,3-dichloroaniline
2 , 3-dichloronitrobenzene
1 , 4-naphthoqulnone
2 - isopropylnaphthalene
1 ,4-dinitrobenzene
2,6-dichloro-4-nitroaniline
alpha - naphthy lamine
ethylenethiourea
4-chloro-2-nitroaniline
5 -nitro-o- toluldine
azobenzene
dlphenylamine
phenacetln
4-aminobiphenyl
pronamlde
dlphenyldlsulfide
4-nitrobiphenyl
carbazole
I , 5-naphthalenediamine
meehapyrilene
phenothiazine
crotoxyphos
thioxanthone
p- dime thy laminoazobenzene
4,4' -mechylenebis(2-chloroaniline)
3,3' -dime thoxybenzidine
ethynylestradiol 3-methylether
2,6- dichlorophenol
2 , 3 , 6- trichlorophenol
2,3,4,6- tetrachlorophenol
2,5-dlbromo-4-hydroxybenzonlCrlle
206.88
208.13
208.94
219.51
220.32
220.81
222.11
222.43
226.83
232.03
240.81
242.44
247.32
247.48
248.13
264.23
267.97
274.47
274.63
277.40
277.40
289.27
295.61
300.00
308.95
312.13
314.31
319.48
340.36
343.34
348.51
351.29
368.39
400.42
401.89
426.85
202.93
237.72
266.34
284.23
206.33
207.47
208.28
218.83
219.64
220.13
221.43
221.75
226.14
231.33
240.10
241.72
246.59
246.75
247.40
263.47
267.21
273.70
273.86
276.62
276.62
288.47
294.81
299.19
308.00
311.20
313.40
318.60
339.60
342.60
347.80
350.60
367.80
400.00
401.48
426.63
202.27
237.01
265.58
283.44
2.3-dichlorob ipheny1
siraazine
atrazlne
1Indane
2, A,5 -erlchloroblphenyl
heptachlor
2,2*4,4'-tetrachloroblphenyl
aldrln
heptachlor epoxide
2,2',3',4,6-pentachloroblphenyl
gamma-chlordane
alpha-chlordane
trans-nonachlor
2,2',4.4',5,6-hexachloroblphenyl
endrln
bis(2-ethylhexyl)adipate
2,2',3,3',4,4,6-heptachloroblphenyl
methoxychlor
2,2'3,3',4.5,6,6'-ocEachlorobIpheny 1
CAS Ho.
16605-91-7
122-34-5)
1912-24-9
58-89-9
15862-07-4
76-44-8
2437-79-8
309-00-2
1024-57-3
60233-25-2
5103-74-2
5103-71-9
39765-80-5
60145-22-4
72-20-8
103-23-1
52663-71-5
72-43-5
40186-71-8
Method
A
282.57
288.66
290.96
295.85
308.18
317.82
324.81
328.21
340.43
343.31
347.73
352.48
354.18
360.96
374.20
388.63
400.18
401.77
402.48
525 RIa'b CARI
a Mean
285.14
290.72
292.42 294.66
296.07 296.38
309.35 316.82
318.75 320.50
326.92
329.17
342.24 343.19
346.32
350.00
354.90
356.54
364.71
369.40 368.46
389.83 384.28
400.51
401.65
402 .34
LEI
Median
294.65
296.38
316.82
320.48
343 . 15
--
--
--
368.39
383.70
..
-.
..
referenced to naphthalene-dg, phenanthrene-d^Q, chrysene-d^. perylene
referenced Co naphthalene, phenanthrene, chrysene, perylene
a Referenced to acenapthylene (RI-244.63) phenanthrene, chrysene, benzo(g h i)perylene
(RI-501.32).
Columns A and B refer to the 2 temperature programs In Method 525.
CONCLUSIONS
The PAH Retention Index system is a promising candidate for the
improvement of "Tentatively Identified Compounds" reporting. Because
it uses compounds already present in the calibration mixture as reten-
tion time markers, it will not require major modification to the present
CLP semi-volatile method. The PAH RIs of hundreds of PAH and
heterocyclic compounds, which can be used to confirm the identity of
TICs without further laboratory work, are found in the literature.1'3
Tables 2,3 and 4 in this paper give additional RIs on non-PAH com-
pounds. Thus, data reviewers could begin to use the system immediately.
The ability to predict the RI based on molecular structure
HEALTH & ENDANGERMENT 89
-------
descriptors48 can short-cut the establishment of RIs for the rest of the
compounds in the NIST/EPA/MSDC mass spectral data base, by
eliminating the need to measure the retention time of all the 50,000
compounds in the data base.
ACKNOWLEDGEMENTS
The authors thank Tamara Wexler for her assistance in this work.
While this work was performed under U.S. EPA Contract No.
68-01-7253, this paper does not necessarily reflect the views of the
agency.
REFERENCES
1. Lee, M.L., Vassilaros, D.L., While, C.M. and Novotny, M.. "Retention
Indices for Programmed-Temperature Capillary-Column Gas Chromalography
of Polycyclic Aromatic Hydrocarbons," Anal. Chem. 51, pp. 768-773. 1979.
2. Willey, C, Iwao, M., Castle, R.M. and Lee, M.L., "Determination of Sulfur
Heterocycles in Coal Liquids and Shale Oils," Anal. Chem. 53. pp. 400-407.
1981.
3. Vassilaros, D.L., Kong, R.C., Later, D.W. and Lee, M.L,, "Linear Reten-
tion Index System for Polycyclic Aromatic Compounds. Critical Evalua-
tion and Additional Indices.", J. Chrom. 252, pp. 1-20, 1982
4. Whalen-Pederson, E.K. and Jurs, P.C., "Calculation of Linear Tempera-
ture Programmed Capillary Gas Chromatographic Retention Indices of Poly-
cyclic Aromatic Compounds," Anal. Chem. 53. pp. 2184-2187, 1982.
5. Doherty, P.J., Hoes, R.M., Robbal. A., Jr. and White, CM., "Relation-
ship between Gas Chromatographic Retention Indices and Molecular Con-
nectivities of Nitrated Polycyclic Aromatic Hydrocarbons," Anal, Chem. 56,
pp. 2697-2701, 1984.
6. Robbal, A., Jr., Corso, N.P., Doberty, PJ. and Marshall, D., "Multivariate
Relationships between Gas Chromatographic Retention Index and Molecular
Connectivity of Mononitrated Polycyclic Aromatic Hydrocarbons," Anal
Chem. 58, pp. 2072-2077, 1986.
7. Robbat, A., Jr., Corso, N.P., Dohety. P.J. and Wolf. M.H., "Gas Chro-
matographic Chemiluminescent Detection and Evaluation of Predictive
Models for Identifying Nitrated Polycyclic Aromatic Hydrocarbons in a Diesel
Fuel Paniculate Extract," Anal. Chem. 58, pp. 2078-2084, 1986.
8. Robbat, A. Jr., Xyrafas, G. and Marshall, D., "Prediction of Gas Chro-
matographic Retention Characteristics of Polychlorinaled Bipherryls," Anal.
Chem. 60, pp. 982-985, 1988.
9. U.S. EPA, Office of Water Regulations and Standards. "Analyu'cal Methods
for the National Sewage Sludge Survey," Aug. 1, 1988
10. U.S. EPA, "Methods for the Determination of Organic Compounds in
Drinking Water," EPA-600/4-88/039. Environmental Monitoring Systems
Laboratory, Cincinnati. OH. Dec. 1988.
II. U.S. EPA. "Test Methods for Evaluating Soil Waste Physical/Chemical
Methods (SW 846) 3rd edition. Proposed Update Package, Office of Solid
Waste, Washington, D.C., Dec. 1987.
90 HEALTH & ENDANGERMENT
-------
Toxin-Exposure Medical Surveillance:
A Focus on Cost-Effective Risk Management
R. Hurt Prater, M.D.
David L. Barnes, M.D.
David R. Larimore
Environmental Medicine Resources, Inc.
Atlanta, Georgia
ABSTRACT
Employers of workers exposed to hazardous substances are faced with
a dilemma created, on the one hand, by a Congressional mandate to
"provide a safe workplace for their employees," while, on the other
hand, striving to maintain profitability by cost-containment. The issue
is further complicated by a prevailing litigious climate generating far-
reaching precedents from civil and criminal prosecution. Considering
such external pressures, it is critical that employers incorporate an
effective medical risk-management program in their business plan. The
medical surveillance portion of that program is all too often treated as
just a "physical examination." The long-term liability associated with
toxin-exposure/absorption-related disease dictates that the prudent
employer utilize available professional expertise and biological tech-
nology to design a compliant, cost-effective medical surveillance
program.
INTRODUCTION
This paper will discuss the changing relationships between public
sentiment, current regulations, the recent focus of the legal profession
on toxic torts and advances in today's medical technology. The antiquated
view of medical surveillance as "just a physical examination" will be
compared to the necessity for a medical risk-management program that
embodies a comprehensive, well-designed medical surveillance program
for toxin-exposed employees. Finally, specific guidelines for the
employer will be discussed to facilitate the design and implementation
of a compliant, toxin-exposure medical surveillance program.
DISCUSSION
"For the first time in the history of the world, every human being
is now subjected to contact with dangerous chemicals, from the moment
of conception until death." It may seem strange to start a paper regarding
medical surveillance with a quote from Rachel Carson's Silent Spring.
The publishing of this book in 1962, however, was the catalyst for the
formation of the U.S. EPA in 1970 amidst strong public sentiment
regarding "poisoning of the planet" from chemicals. This land swell
of public concern over the health and environmental impact of chemi-
cals has progressively increased since 1970 and resulted in the passing
of CERCLA among other legislation. Public sentiment, with regard
to human exposure to chemical hazards, has gradually turned from fear
to anger toward those responsible. Deviant generators of chemical waste
and those who carelessly expose their employees to toxic chemicals
will have difficulty finding a sympathetic ear in today's courtroom.
That management has injured employees as a result of either their
ignorance of health issues or their disregard for them is unquestioned.
One has but to look at Gaulley Gap, the radon dial painters of WWII
or, more recently, the Manville asbestos cases. The recent accelera-
tion of toxic torts, prosecution of corporate directors (The Chicago
Magnet Wire case) and even CERCLA itself is a direct result of such
ignorance and disregard.
Many of the current medical regulations regarding toxin-exposure
medical surveillance are in a state of flux and can be confusing. There
are generic guidelines but there also are specific standards for certain
chemicals and action levels that must be taken into consideration for
many others. Uncontrolled hazardous waste sites represent the poten-
tial for a multitude of health and safety concerns. The standards promul-
gated in 29 CFR 1910.120 regarding a health and safety program are
summarized in the Inter Agency Guidance Manual published in 1985
by NIOSH, OSHA, USCG and the U.S. EPA.
This manual assumes that a medical surveillance program will be
used to complement engineering controls, personal protective equip-
ment (PPE) and decontamination procedures. In addition, it assumes
that the average toxic waste site contains many unknown chemicals.
Even though site characterization may identify specific chemicals, it
must be assumed that other chemical hazards may exist. The manual
states, "The program should be designed by an experienced occupa-
tional health physician or other qualified occupational health consul-
tant in conjunction with the Site Safety Officer. The director of a site
medical program should be a physician who is board-certified in
occupational medicine or a medical doctor who has had extensive
experience managing occupational health services. If an occupational
health physician is not available... (the program)... may be performed
by a local physician with assistance from an occupational medicine con-
sultant."
The regulations divide a site medical program into surveillance, treat-
ment, record-keeping and program review. Medical surveillance includes
three types of examinations: (1) pre-employment screening (baseline);
(2) followup examinations (periodic); and (3) termination (exit) exami-
nations. It should be emphasized that the regulations clearly state, "Be-
cause conditions and hazards vary considerably at each site, only general
guidelines are given." In most cases the final decision regarding the
details of the site medical program is left to the physician consultant.
These recommendations should be considered as minimal standards.
Most forward-thinking employers do not feel comfortable with the lia-
bility protection afforded by such standards.
Behind-the-scene changes rapidly are taking place in the regulatory
arena and the U.S. EPA is emerging as a dominant player in the
enforcement of environmental laws. OSHA, as an agency of the govern-
ment, acts slowly both in the promulgation and enforcement of laws,
and as of September 1989, the U.S. EPA is adopting the OSHA
hazardous site worker standards promulgated by 1910.120. Due to the
bureaucratic hierarchy, the U.S. EPA will now have the power to act
swiftly to enforce standards and prosecute violators. In addition, the
RISK ASSESSMENT 91
-------
U.S. EPA will have the ability to promulgate new standards more
effectively than OSHA.
To further complicate the picture, there are numerous other power-
ful and respected groups exerting pressure on Congress. Unions and
consumer advocacy groups are starting to take an active role in lobbying
and enforcement of specific worker protection standards. Professional
associations like the American Conference of Governmental Industrial
Hygienists (ACGIH) and the American Industrial Hygiene Association
(AIHA) are pushing for more stringent worker protection standards.
In particular, these groups are recommending that board-certified
occupational physicians should have the responsibility for designing
medical surveillance programs. In addition, they are insisting on a more
generic industry standard designed to stabilize the current fragmented
approach to biological monitoring.
The High Risk Notification Bill: California Proposition 65, complete
revision of OSHA's permissible exposure limits (PELS), changes in
hazardous waste transportation standards and revisions of the formal-
dehyde and benzene standards are just a few of the upheavals in this
industry. Many states are passing laws that are far more stringent than
their federal counterparts. Compliant health evaluation programs
designed for today could be obsolete by tomorrow. Only health profes-
sionals with a focus and expertise in this regulatory-driven industry
will be capable of designing and maintaining compliance-assured
medical surveillance programs that provide health protection for the
employee and maximum liability protection for an employer.
The escalation of litigation regarding on the job injury and disease
is reflective of the clout from the combination of public sentiment and
rapid changes in worker protection laws. The Chicago Magnet Wire
Case is an example of the extent of personal liability exposure and its
consequences. A recent large settlement against an employer was the
result of the plaintiff showing that benzene exposure can cause chronic
as well as acute leukemia (Skeen v. Monsanto Co. Feb. 21, 1989). This
case is a good example of the necessity for optimal environmental and
biological monitoring and good recordkeeping (the exposure to benzene
occurred in the mid-1970s). Benzene is a toxin that can be monitored
for absorption prior to the onset of disease. Not only is the frequency
of toxic torts increasing, but also the level of awards and punishment
are likewise increasing. At present, there is no reason to believe that
this trend will diminish. On the contrary, there is every indication that
it will accelerate in the foreseeable future.
The employer who is involved in the hazardous waste remediation
business must keep in mind several seldom considered facts when
planning for the health and safety of his employees. In the first place,
the general guidelines noted above were revised in 1985 but were actually
written in 1980. The sole purpose of the Inter-Agency Guidance Manual
was to provide a comprehensive guide for site safety. The Occupational
Health and Safety Act of 1970 clearly states that it is the responsibility
of the employer to provide a safe workplace. There is no qualification
with regard to the limitations of current technology, unforeseen chronic
adverse health effects or specific budgetary restrictions. In fact, some
interpretations hold the employer responsible regardless of the circum-
stances. The design of an effective medical risk management program
depends on the assumption that the employer is ultimately responsible
for worker health and safety, regardless.
MEDICAL EXPERTISE AND BIOMEDICAL TECHNOLOGY
Modern 20th century medicine is generally accepted to be excellent.
American medicine is thought by many to be the best in the world.
Both of these commonly accepted statements may be true, but they must
be evaluated relative to the end-point that would be considered the ul-
timate. That end-point, quite simply stated, is the prevention of disease.
Inarguably, medical authorities and the lay population would agree on
that issue. If our ultimate goal is prevention of disease, we are far from
achieving such a state.
Even the most conscientious employer will be met with frustrations
and limitations represented by the technological inadequacies of modern
medicine. In the hazardous substance exposure business, the primary
concern is absorption of chemicals through the skin, lungs, eyes, ears,
ingestion or penetration. If we had the ultimate diagnostic tool, we could
scan an individual and detect even the most minimal absorption before
the onset of disease. Unfortunately, this tool is not available and in its
place we must rely on a very inexact vast human experiment. This ex-
periment constitutes the inadvertent exposure of large numbers of
humans to chemicals for an unknown period of time and monitoring
for the onset of disease. When disease is diagnosed, we must then
retrospectively look backwards in the hope we can correlate some
exposure to the disease. At this point, it may be too late for the re-
habilitation of the individual to normal health, and it also may be too
late for the employer to convince an agency or jury of his innocence
regarding the employee's injury or disease.
This discussion does not imply that we are unable to detect absorp-
tion with current medical technology. In many cases, we can, but the
ability to monitor a chemical in a biologic system is just now emerg-
ing. NIOSH recently announced that of the over 100X100 chemicals used
in manufacturing today, chronic adverse health effects are known to
occur with less than 20% of that number. The commonality of chemi-
cal pathways for metabolism and excretion is becoming clearly defined
and identified. In fact, we are advancing far more rapidly in the area
of absorption identification than in the area of treatment modalities for
existing absorption disease. (Consider, for example, lung cancer
resulting from absorption of asbestos).
The above description of the limitations of biological monitoring is
not designed to further confuse the reader, but rather to illustrate that
technological overkill in a medical surveillance program, in addition
to excessive cost, can be as fruitless as a "bare minimum" approach.
It also points out the complex and dynamic nature of medical technology,
and the advances we may expect, in the early detection of chemical
absorption prior to the onset of disease. The object is to again empha-
size that only experienced environmental health professionals are capable
of staying abreast of these changes and translating this information into
maximum protection for the employer and employee.
THE MEDICAL SURVEILLANCE PROGRAM
How does this translate into practical information and guidelines for
the modem remediation firm? It is critical that employers in the
hazardous substances business recognize that a health surveillance
evaluation is more than just a physical examination. It is an integral
element in a plan, required by law and demanded by humanitarian prin-
ciples, and designed to protect workers from adverse health effects.. .to
attempt the prevention of disease as opposed to the discovery of dis-
ease. Just as the chain is only as strong as its weakest link, the sound-
ness of such a plan is only as effective as the most poorly planned and
executed element. If we accept the premise that a site health and safety
plan constitutes a risk-management program, then it logically follows
that the health surveillance element of that plan is critical in the over-
all effectiveness of the program.
Many Superfund sites are located in rural areas, and a remediation
firm may have multiple sites scattered around the nation. In all likeli-
hood, many of these sites will be in areas void of occupational physi-
cians trained and experienced in toxin-exposure medical surveillance.
It will be necessary to depend on local physicians for the surveillance
data. Because this situation is the rule rather than the exception, the
examination protocol, in addition to being standardized and thorough,
must be "user friendly" and coordinated with concise clinic operations
manuals. This is the only mechanism by which the company's medical
data can be reliably obtained from multiple medical practitioners.
There are four basic sources of information utilized to design an
examination protocol for a medical surveillance program: (1) the em-
ployees' exposure profile and working conditions, (2) current regula-
tions, (3) the company's philosophy toward worker health and safety
and (4) current technology. (Fig. 1). The examination protocol provides
procedural guidelines for the physician and his staff to complete a health
evaluation of each employee. The protocol should provide the physi-
cian with a completed personal and occupational history which he can
augment at his discretion. With the current technological limitations
in biological monitoring, the history provides, by far, the majority of
the information from the examination. It should be thorough, detailed,
designed with planned redundancy, and, by necessity, it must be long.
92 RISK ASSESSMENT
-------
Because of its length, it must be completed with forethought and without
pressure before the employee presents himself for his medical evaluation.
Employee
Toxin-Exposure
Profile
/
/ Protocol
Design
/ /
/
Employee
Job Description
Current
Technology
& Research
(Advisory Board)
T
MEDICAL EVALUATION
• History
• Laboratory & Physical
Measurements
• Physical Examination
Figure 1
The Anatomy of a Toxin-Exposure, Health Surveillance Evaluation
Likewise, the examination protocol guides the physician in the hands-
on physical examination necessary to determine present state of health
and physical capabilities. In addition to programed responses (Yes or
No), it should encourage the physicians comments and suggestions.
Lastly, the protocol guides the medical staff in physical measurements
and ancillary tests like audiometry (hearing tests) and spirometry
(breathing tests). Clinic laboratory equipment used to measure breathing,
hearing, etc. must meet certain specifications by law. In addition, the
technician who operates the equipment must be certified according to
the standards. If these requirements are not met, the data obtained are
questionable. Likewise, the original test results must be kept in
retrievable fashion for 30 yr after termination.
The protocol should direct the medical team in the obtaining and
packaging of specimens (blood and urine) for mailing to a specific
laboratory. The selection of a reference and specialty laboratory to ana-
lyze blood and urine is critical to a sound medical monitoring program.
Although such services are widely available, there is considerable
variation in quality and reliability of results. A detailed discussion of
the vicissitudes of this industry is not appropriate for this paper. The
recurring theme is, again, the necessary reliance of the employer on
a carefully selected occupational physician for guidance in the selec-
tion of laboratory services.
The end-point of all the medical data (history, physical examination
and laboratory analysis) is review by the physician. It is here that the
axiom "garbage in, garbage out" is appropriate. If these data are not
extracted carefully, accurately and professionally, the results of the phy-
sician's review could be erroneous. Equally as important, the company's
financial outlay for medical surveillance could be wasted. Many
employers select physician consultants with little regard to qualifica-
tions. These consultants are selected to design and implement com-
plex health monitoring programs simply because they work in a clinic
that provides general occupational medicine services. Enormous sums
of money and significant corporate and personal liability are put on
the line when inexperienced and poorly informed health professionals
are selected to manage medical surveillance programs. A working
knowledge of the regulations and standards, experience in toxicology
and training in the pathophysiology of disease from chemical absorp-
tion are essential skills for the physician reviewing biological monitoring
data. Following review of the medical data generated by the examina-
tion process, the physician must generate a written report with recom-
mendations to the employer and employee. This report must reach the
employee within 15 days following the examination. Abnormalities must
be addressed and a disposition made with regard to followup and res-
trictions. It is the employer's responsibility to see that all medical data
collected under 1910.120 are stored in retrievable form for 30 yr after
termination. If a court action regarding a previous employee should
arise 10 yr after a specific examination, OSHA and the court will ex-
pect the employer to provide such records in readable form. Failure
to do so could result in severe penalties and a less than adequate defense.
It should be obvious at this point that medical surveillance is, in truth,
a complex program, not just a physical examination. In addition to as-
suring the ongoing health of a company's work-force, it must be recog-
nized and utilized as an integral part of a corporate risk-management
program. The tendency among less well-informed employers/managers
when attempting to comply with regulations regarding medical surveil-
lance, is to "cut corners" with the cheapest examination available. This
approach may look good on the bottom line of a profit and loss state-
ment, but the penalties from violations and the cost of litigation could
be catastrophic. In many cases, a substandard program is more expen-
sive than a professionally-designed and compliance-assured program.
Cost is always a factor in business, but contracting with the lowest bidder
may be false economy.
GUIDELINES
Every cost-effective medical surveillance program should contain the
following services and benefits:
• A board certified occupational medicine physician with experience
in toxin-exposure disease
• A corporate-wide examination protocol designed by, or in consult
with, that physician
• Review of die examination protocol by someone knowledgeable and
current in the related standards and regulations
• Designated, qualified clinics convenient to each site which have been
trained in the use of the examination protocol
• A nationally recognized laboratory to analyze biological specimens
• Concise, compliant medical results in the form of a report received
within 15 days of the examination
• Reliable storage of all medical records in retrievable form for 30 yr
plus the term of employment
Fortunately for the employer, there are firms today which can pro-
vide these medical services. If difficulty is encountered in locating such
a firm, a board certified occupational medicine physician should be
consulted for advice. A list of such physicians can be obtained by writing
the authors or the American College of Occupational Medicine.
CONCLUSION
The standards regulating the hazardous materials industry require
an employer to provide a medical surveillance program for exposed
workers. The intent of such legislation is to insure the health and safety
of the employee, but, from a business perspective, the standards pro-
vide liability protection for the corporation and its directors. Such legis-
lation is viewed by many employers as an unnecessary burden, but,
in fact, it could be a blessing in disguise for companies which are not
informed in matters of risk management.
It has been demonstrated that a toxin-exposure health surveillance
program is very complex and requires specialized medical expertise.
Only an informed occupational physician who specializes in hazardous
materials exposure can coordinate the appropriate regulations and bio-
RISK ASSESSMENT 93
-------
medical technology to provide a true medical risk-management program. suit with environmental physicians who have experience in the field
The prudent employer designing a site health and safety plan would of absorption disease for guidance. Cutting corners on medical sur-
be well advised to consider his health surveillance program as an integral veillance could jeopardize a company's entire risk-management program
part of a sound risk-management program. To this end, he should con- and, a.s a result, the company's future.
94 RISK ASSESSMENT
-------
.99
Characterization of "Significant Risk5
Under The Massachusetts Contingency Plan
Maria T. Madison, M.S.
Paul W. Locke, M.S.
The Department of Environmental Protection
Office of Research and Standards
Boston, Massachusetts
ABSTRACT
The Massachusetts Contingency Plan (MCP) was promulgated on
Oct. 3,1988 to implement the State Superfund law (the "Massachusetts
Oil and Hazardous Material Release Prevention and Response Act,"
Massachusetts General Laws Chapter 21E, 1983). The assessment and
remediation of state superfund sites are carried out in a phased approach.
Site risk characterization is a critical part of the Comprehensive Site
Assessment (Phase IT) required under the MCP.
M.G.L. Chapter 21E (the statute) requires the achievement of a
"Permanent Solution" at all disposal sites, if feasible. A Permanent
Solution eliminates any "significant or otherwise unacceptable risk"
of harm to health, safety, public welfare or the environment during any
foreseeable period of time. When feasible, a Permanent Solution will
restore the disposal site to background levels. The answer to the ques-
tion of what constitutes a significant risk (and subsequently, "How Clean
Is Clean Enough?") became a major issue in the development of the
regulations implementing the statute.
The MCP approach employs risk assessment processes outlined in
the National Academy of Sciences study "Risk Assessment in the
Federal Government: Managing the Process"2 and is consistent with
the methods adopted by the U.S. EPA for use at Federal Superfund sites.
However, the state approach specifically defines significant risk in a
manner which differs from the risk range approach used by the U.S.
EPA.
This paper will describe the Massachusetts methodology and draw
comparisons with the U.S. EPA approach to evaluating "significant
risk"at hazardous waste disposal sites.
INTRODUCTION
One or more factors may drive the remediation at a disposal site in
Massachusetts. These factors include the risk of harm to human health,
the risk of harm to the environment and the feasibility of restoring the
site to background conditions. This discussion will focus primarily upon
the characterization of the risk of harm to health [310 CMR 40.545 (g)].
Additional consideration should be given to the characterization of risk
of harm to safety, public welfare and the environment [310 CMR 40.545
(h)]. Currently, these concerns are being addressed through, primarily,
qualitative methods.
In 1987, the Massachusetts Department of Environmental Quality
Engineering (now the Department of Environmental Protection, DEP)
contracted Wehran Engineering to survey state and federal environmental
officials and the current scientific literature in an effort to develop a
working definition of "significant risk" for use in the MCP. The result
was "The 'Significant Risk' Project6. The Project surveyed the states
of California, Michigan, New Jersey, New %rk and Wisconsin, as well
as the U.S. EPA concerning methods of standard setting and their use,
risk management policy and approach, definition and use of the term
"significant risk" and their approach to a hypothetical pollution control
scenario.
Concurrent with this process, three goals were identified for the risk
characterization process to be used in the Massachusetts Contingency
Plan:
• Disposal sites in Massachusetts would be remediated to levels which
would be protective of the public health
• Disposal sites would be remediated in a consistent manner through-
out the Commonwealth's four regions
• Disposal sites would be remediated in a manner consistent with
existing state regulatory programs
To achieve these goals, the Massachusetts Contingency Plan outlines
[four methods] for the characterization of risk of harm to human health
at Massachusetts disposal sites. Four methods were developed (as
opposed to one set method) to more closely address the complex idio-
syncrasies of individual sites. Central to this process is the inclusion
of "Total Site Risk Limits" in the regulations. These specific risk limits
contrast sharply with the risk range approach practiced at the federal
level.
The four methods were crafted to satisfy the three principal goals
of the MCP. Since these methods are intended to achieve the specific
requirements of the Massachusetts statute, they often go beyond the
approach developed by the EPA for use at Federal Superfund sites.
However, as the EPA updates the guidance given in the "Superfund
Public Health Evaluation Manual"5, the differences between the two
programs will narrow. The recently completed "Supplemental Risk As-
sessment Guidance for the Superfund Program"4 prepared by the U.S.
EPA Region I Risk Assessment Work Group is in fact consistent with
the Massachusetts DEP's "Guidance for Disposal Site Risk Charac-
terization and Related Phase n Activities - In Support of the MCP"5.
The risk characterization process mandated by the Massachusetts
Contingency Plan is described below, with emphasis on the definition
and use of the term "significant risk."
DEFINING AND EVALUATING "SIGNIFICANT RISK"
UNDER THE MCP
The Massachusetts Contingency Plan (MCP), promulgated by the
Department of Environmental Protection (DEP), became effective on
Oct. 3, 1988. The MCP establishes requirements and procedures for
identifying, evaluating and cleaning up releases of oil or hazardous
materials to the environment. The regulations are based upon the State
"Superfund Law" (The Massachusetts Oil and Hazardous Materials
Release Prevention and Response Act of 1983) and major amendments
passed by voter referendum in 1986.
RISK ASSESSMENT 95
-------
The MCP responds to the public's clear mandate by establishing a
cleanup process that is consistent, strict and highly protective of public
health and the environment. Risk characterization and evaluation are
at the heart of the cleanup process.
As part of the requirements of the Comprehensive Site Assessment
(in Phase II of the six phase process), the regulations provide a frame-
work for determining: (1) whether remediation at a disposal site is
required and (2) the extent of remediation needed to attain a perma-
nent or temporary solution. The Phase II risk characterization directs
consistent and conservative evaluations of human health, safety, public
welfare and environmental risks at all of the disposal sites in the
Commonwealth.
To supplement the language of the regulations and provide more
detailed guidance for their implementation, the Department has
published "Guidance For Disposal Site Risk Characterization And
Related Phase II Activities - In Support of the Massachusetts Contin-
gency Plan" (May 17, 1989). The guidance primarily addresses the
characterization of risk of harm to human health. The evaluation of
the risk of harm to safety, public welfare or the environment relies upon
existing environmental standards and site-by-site considerations. The
Department recognizes that more guidance is needed in this area.
CHARACTERIZING RISK POSED BY DISPOSAL SITES
The Massachusetts Superfund law (M.G.L.Chapter 21E) requires that
cleanups must eliminate "significant or otherwise unacceptable risk"
of harm to human health, safety, public welfare and the environment.
As the MCP was drafted, a great deal of discussion centered on the
questions of: (1) what constitutes a significant risk?, and (2) what
methodology should be used to characterize risk at a disposal site? Risk
assessment and risk management are not unique to the state and federal
Superfund programs. The Commonwealth of Massachusetts has imple-
mented a number of regulatory programs which address environmental
contamination in specific media, including air, drinking water, ground-
water and surface water. One goal of the Contingency Plan was to
preserve that the integrity of these existing programs; remediation at
disposal sites would, at a minimum, meet any applicable or suitably
analogous standards of these programs and the policies of these programs
would be applied when appropriate.
The Department also recognized that many disposal sites are far more
complex than the situations commonly addressed by the medium-specific
programs. In particular, it was felt that the reliance upon standards and
guidelines developed for single contaminant or single-medium situa-
tions might be inadequate to protect the public health at a disposal site
involving multi-media contamination and/or a mixture of contaminants.
Finally, the Department wished to minimize costs to those performing
the risk characterizations by relying upon standards, guidelines and/or
existing sets of cleanup levels whenever possible.
What emerged from these discussions were four risk characterization
methodologies, only one of which would be appropriate at any given
disposal site. The regulations describing these methodologies can be
found in 310 CMR 40.545(3)(g) of the Contingency Plan. Only one
of the four methods involves the classic, full risk assessment. Less
complex sites would use simpler risk characterization methods.
THE FOUR METHODS
As detailed in the regulations and elaborated upon in the Department's
Guidance Document, the characterization of risk of harm to human
health is evaluated using one of four methods. As only one of the four
methods is considered appropriate at any given disposal site, it is ex-
tremely important that the correct risk characterization methodology
be chosen at the beginning of the process. Since the promulgation of
the regulations in 1988, increasing emphasis is being placed on the
proper selection of a method for site evaluations in the risk characteri-
zation process. The first revision of the Department's Guidance Docu-
ment attempted to more fully describe and explain the selection process.
The Methods are meant to be considered in a stepwise fashion, from
the simplest (Method 1) to the most complex (Method 3b).
Method 1
Method 1 applies at sites where, under existing regulations, there
are standards (NOT guidelines, NOT policies) applicable to each oil
and/or hazardous material (OHM) in every medium (air, water or soil)
to which persons might be exposed.
In this Method, the risk characterization compares the OHM exposure
point concentrations to the standards identified. Remediation is required
if any concentration exceeds such a standard, and the standards become
requirements for a permanent solution.
Presently, Method 1 does not apply to a large percentage of disposal
sites under investigation in Massachusetts. Of the 23 contaminants most
commonly found at state sites, ambient air quality standards exist only
for lead, drinking water standards exist only for 12 of the 23 chemicals
and no public health soil standards exist at this tune.
If Method 1 is not appropriate, Method 2 is to be considered.
Method 2
Under Method 2, exposure point concentrations of OHM are com-
pared to specific sets of cleanup levels to be incorporated into the MCP
(310 CMR 40.800). These specific sets of cleanup levels will be
developed by the Department for certain types of disposal sites which
present common problems. For example, a specific set of cleanup levels
may be developed for leaking underground gasoline storage tanks in
a residential area where there are private wells and where no exposure
is thought to occur other than via drinking water. These levels will be
specific for both the contaminants reported at a site and the potential
exposures at such a disposal site.
In this Method, the risk characterization consists of the comparison
of OHM exposure point concentrations to corresponding values con-
tained in the specific set of cleanup levels. Remediation is required if
any concentration exceeds an identified cleanup level, and the set of
cleanup levels becomes a requirement of a permanent solution.
To date, no specific set of cleanup levels has been established; Method
2, therefore, is unavailable for any disposal site. The Department
currently is working on a number of such sets of cleanup levels,
including sets for PCB contaminated soil and Coal Gasification
Waste disposal sites and petroleum contaminated sites.
When neither Method 1 nor Method 2 apply to a site (or when 2
is applicable, but not used), then either Method 3a or 3b is appropriate.
It must be determined at this point whether the site fits the characteris-
tics of a "Single Medium" disposal site (to be evaluated per Method
3a) or a "Multi-Media" disposal site (to be evaluated per Method 3b).
Both Methods may make use of site-specific risk assessment techniques.
Method 3a
Method 3a is appropriate if exposure to the oil or hazardous materials
at or from the disposal site occurs via one contaminated medium. Using
Method 3a, exposure point concentrations are compared to corres-
ponding public health standards, guidelines or Departmental polices.
If no such value is available for a particular chemical, then a site-specific
guideline associated with an excess lifetime cancer risk equal to one
in one million and/or a Hazard Index equal to 0.2 should be proposed
by the primary responsible party.
In this Method, the risk characterization consists of the comparison
of the exposure point concentrations to the identified standards, guide-
lines, policies and/or proposed site specific guidelines. When remedi-
ation is required, these standards, guidelines, policies and/or proposed
site-specific guidelines become requirements for a permanent solution.
Method 3a has been used to characterize the risk at approximately
10 to 25% of the disposal sites assessed to date. (This figure is an esti-
mate as the Department has not tracked the number of sites using each
Method.) One common type of disposal site which would be charac-
terized by this Method involves a contaminated drinking water supply
where no additional exposures are thought to occur. The risk charac-
terization process would employ the drinking water standards and guide-
I ines developed by the Department's Division of Water Supply and Office
of Research and Standards, as well as any applicable Departmental poli-
cies. Any proposed site-specific guideline would be developed in a man-
ner consistent with Departmental policy, using standard risk assessment
96 RISK ASSESSMENT
-------
techniques. This Method differs from Method 1 in that it is limited
to single-medium situations and both guidelines and policies may be
used in addition to standards.
If Methods 1, 2 and 3a are not considered appropriate, then the site
is evaluated using Method 3b.
Method 3b
Method 3b is appropriate if a receptor may potentially experience
exposures to the oil or hazardous materials at or from the disposal site
via more than one contaminated medium, and if Methods 1 and 2 are
not applicable. In Method 3b, exposure point concentrations are com-
pared to applicable or suitably analogous standards, promulgated under
existing regulations. In addition, a site-specific risk assessment is con-
ducted and the "Total Site Risk" estimates are compared to the risk
limits presented in the MCP. For one or more hypothetical receptors,
the estimated "total site risks" reflect potential exposures to all the OHM
via all the exposure pathways. Guidance is given for the development
of these hypothetical receptors for whom the total site risks are esti-
mated. Note that even the chemicals for which standards exist are in-
cluded in the calculation of the total site risk. Under Method 3b, the
most flexible cleanup requirements may be developed while complying
with total site risk requirements and applicable/available public health
standards.
The risk characterization process under Method 3b consists of the
comparison of exposure point concentrations to applicable or suitably
analogous standards, and the comparison of "Total Site Risks" to the
Total Site Risk Limits. Total site cancer risks are compared to a total
site cancer risk limit of one in one hundred thousand (1.0 x 10 ~5- To-
tal site non-cancer risks are compared to a total site non-cancer risk
limit which is a Hazard Index equal to 0.2. When remediation is
required, the identified standards and the Total Site Risk Limits serve
as remediation requirements.
Method 3b has been used to characterize the risk at approximately
75 to 90% of the disposal sites evaluated under the MCP to date. It
is assumed that this level will be reduced in the future as more standards
are set and as the Department develops specific sets of cleanup levels
for use in Method 2. (Further consideration and evaluation of back-
ground contamination levels will affect the implementation of this
process).
OTHER REMEDIATION CRITERIA
As noted earlier, the MCP also requires the characterization of the
risk of harm to public welfare, safety and the environment. In addi-
tion, there must be an evaluation of the feasibility of remediating a site
to background levels.
These additional factors may drive remediation of a site where, based
on assessment of significant risk to human health, no adverse effects
are expected. Such an approach is consistent with the Department's
obligation to protect both public health and the environment.
The Department has attempted to develop a comprehensive means
of characterizing disposal site risks which relies extensively on the iden-
tification of "otherwise unacceptable." Under the MCP, remedial al-
ternatives may be developed that protect public health, while providing
flexibility in the setting of cleanup levels for specific chemicals. In this
manner, the Department has developed a process which methodically
approaches the answer to the question, "How clean is clean enough?"
COEXISTENCE WITH FEDERAL PROGRAMS
The Commonwealth of Massachusetts publishes quarterly lists which
detail the number of state disposal sites and their status. As of July,
1989, there were 1152 Confirmed Disposal Sites which require further
investigation, 1634 Locations To Be Investigated and 270 Sites at which
remedial action has been completed and for which no further actions
are planned (not all of these fall under the requirements of the MCP).
Included among the 1152 confirmed sites are 24 NPL sites. One addi-
tional site has been proposed for the NPL.
Federal Superfund Sites are subject to the requirements of CERC-
LA, and SARA (collectively known as "Superfund") in addition to the
State Superfund law, M.G.L. Chapter 21E. It should be noted here that
most NPL sites would be considered "Multi-media Sites" under the
Massachusetts Contingency Plan, and would thus be subject to the
Method 3b risk characterization process.
While there are many similarities between the two programs, several
important distinctions can be made, particularly in the risk characteri-
zation process. Care must be taken in the development of the Endan-
germent Assessment (EA), the Remedial Investigation (RI), the
Feasibility Study (FS) and the Record of Decision (ROD) to identify
the requirements of the Massachusetts regulations and to explicitly meet
them. It is, of course, most difficult to integrate newly promulgated
regulations into a site remediation process which is already underway.
The most obvious difference between the two programs is the MCP's
Total Site Risk Limits. The U.S. EPA has established an excess life-
time cancer risk range (10~4 to 107 into which the risk based cleanup
goal should fall. Depending upon the site, however, the risk range has
been applied to: (1) the risk associated with a single chemical via a
single exposure route or (2) the risk associated with a mixture of chem-
icals via a single exposure route or (3) the sum of exposure route risks
which could approximate a total site cancer risk. Somewhat more dis-
turbing is an interpretation that all estimations of risk which fall into
the U.S. EPA risk range of 1 x 10~4 to 1 x 10~7 may be considered as
acceptable. This poses particular concern when the estimated risk is
as high as 9 x 10~". In comparison, the MCP Method 3b total site
cancer risk limit is 1 x 10~5 The intention of a risk limit is to guard
that no potential receptor would experience an excess lifetime cancer
risk greater than 1 x 10~5, regardless of how many potential exposure
pathways existed at that disposal site.
An additional difference is the estimation of the total site non-cancer
risk, (i.e., Hazard Index, HI). Massachusetts has adopted a HI of 0.2.
The exposures related to a disposal site are allowed to contribute only
20% of an estimated allowable daily dose. The approach taken by the
Department is similar to that used by the U.S. EPA Office of Drinking
Water to develop drinking water standards and health advisories. In
its Superfund program, the U.S. EPA does not have a risk limit or range
for non-carcinogenic risk. U.S. EPA Region 1 recommends (in the
absence of such guidance) that a HI < 1 is acceptable and that a HI
> 10 may be cause for remediation.
In addition, there are strict requirements to evaluate [all] oil or
hazardous material at or from a disposal site (eliminating the use of
indicator chemicals) and specific consideration is given to the foresee-
able future use of the site and the levels of contaminant which would
exist in the absence of the disposal site (background). These distinc-
tions can influence the choice of remedial alternatives necessary to
achieve a permanent solution at a Federal Superfund site in Mas-
sachusetts.
CONCLUSION
"Significant Risk" is a concept which has no absolute definition.
When circumstances require developing a working significant defini-
tion, many factors must go into that risk management decision. For
the State Superfund program in Massachusetts, such factors included
an explicit mandate to protect human health, public welfare, safety and
the environment, and a need to be consistent with existing state regula-
tory programs. The result is a risk characterization process which would
utilize one of four Methods. Method 3b, which applies to the majority
of disposal sites in Massachusetts, relies heavily upon risk assessment
to determine the need for remediation and the level of remediation
required. A Total Site Cancer Risk Limit of one in one hundred thou-
sand and a Total Site Non-Cancer Risk Limit which is a Hazard Index
equal to 0.2 apply at these Method 3b sites.
The State Superfund program is not inconsistent with the Federal
program, although care must be taken to insure that the requirements
of both are met. The Massachusetts Contingency Plan's reliance on the
total site risk approach rather than chemical- and medium-specific
standards comes at a time when the Massachusetts Department of
Environmental Protection is shifting its structure and focus away from
solely medium-oriented programs.
RISK ASSESSMENT 97
-------
DISCLAIMER
This paper has not been subjected to Department of Environmental
Protection (DEP) review and therefore does not necessarily reflect the
views of the DEP. No official endorsement should be inferred.
REFERENCES
Sections of the previous text have appeared in an article written by the
same authors, and may be referred to for additional information (ref. -1).
1. Madison, M.T., Locke, P.W., and Murphy, M.J., Risk Assessment Proce-
dures favored by MCP Guidance Document, Massachusetts Waste Manage-
ment Report, Hazardous Waste & Related Issues. Editor: Robert A. Parlow.
Esq., 3 (2) 11, May, 1989.
2. National Academy of Sciences, Risk Assessment in the federal Government:
Managing the Process, National Academy Press, Washington, DC, 1983.
3. U.S. EPA - Region I, Supplemental Risk Assessment Guidance far the Su-
perfitnd Program, Draft final, Risk Assessment Work Group, U.S. EPA
901-89-001. June, 1989.
4. Massachusetts DEP, Guidance for Disposal Site Risk Characterization and
Related Phase II Activities • In Support of the Massachusetts Contingency
Plan," Office of Research and Standards, May 17, 1989.
5. U.S. EPA Superfund Public Health Evaluation Manual, Office of emer-
gency and Remedial Response, Washington D.C., EPA 540/1-867060 (OS-
WER Directive 9285.4-1, Oct. 1986 and revised versions).
6. ERT - A Resource Engineering Company, The "Significant Risk" Project,
prepared for Wehran Engineering. Methuen. MA. Document No. P-G359-DOI
Jan . 1988.
98 RISK ASSESSMENT
-------
Defense Priority Model for DoD Site Ranking
Judith M. Hushon, Ph.D.
Roy F. Weston, Inc.
Washington, D.C.
ABSTRACT
The Defense Priority Model (DPM) is intended to permit the use
of site-specific monitoring data from RI/FS and other site reports to
refine priorities for remedial action. This model combines both quanti-
tative and qualitative information on: (1) the hazards posed by pollu-
tant sources, (2) the potential exposure pathways of surface water,
groundwater and air/soil and (3) the potential human and ecological
receptors. The information is combined to reach a final score for each
site that lies between 0 and 100.
The DPM differs from the HRS in that it is used to rank all sites,
not just those for NPL consideration. In addition, it is designed to be
applied later in the data acquisition process when more accurate and
detailed data will be available. There are also some minor differences
in the data that the two models consider relating to pollutant mobility,
food chain exposure and the use of pollutant concentration in the model.
In general, the DPM uses more detailed data and is a more focused
model than the HRS.
This model is being used by DoD to assess relative risk of sites which
are ready for remedial design/remedial action in the fiscal year 1990.
This will give a good indication of its performance and will help to
identify areas where further development can prove beneficial.
INTRODUCTION
Work began on what is now the Defense Priority Model (DPM) in
1984 when the Air Force recognized the need for a defensible metho-
dology for ranking hazardous waste containing sites for cleanup. The
initial work was conducted by Barnthouse and his colleagues at Oak
Ridge National Laboratory and resulted in the development of the
Hazard Assessment Risk Model (HARM)1'2- This model was then
evaluated using comparative testing by a number of reviewers and the
results led to the incorporation of a number of changes and the develop-
ment of HARM IF.
The Air Force determined that the model needed to be computerized
to be maximally useful and decided that expert systems technology would
be preferable to direct computerization using Lotus(r) or dBase(r).
Expert systems technology offered some significant advantages
including:
• Ability to incorporate uncertainty
• Ability to accommodate missing data
• Ability to use alternative pathways to obtain an indication of an answer
• Ability to manage flow through the program so that only appropriate
questions are asked of the user
• Ability to include expert knowledge and make this available to the user
• Ability to include both quantitative and qualitative data in the decision-
making process.
The initial implementation encoded HARM n using the expert systems
shell KES(r) from Software A&E on an IBM compatible PC/AT. This
allowed for a rapid prototyping, but it did not support sufficient power
or screen management. A decision was therefore made to convert the
code to prolog, an AI programming language. Most of the KES code
did not have to be rewritten, but complex definitions that translated
the KES code into prolog were prepared. The prolog chosen was Arity
Prolog version 5.1.
This initial implementation was tested by six professionals ranking
a total of 15 sites with two reviewers per site to determine whether the
model provided a sufficiently broad range of answers, whether the sites
ranked in a logical order and whether the model could be widely used.
The answers to all of these questions were affirmative. Additionally,
some of the reviewers' suggestions for improving the model and the
computerized presentation were incorporated4.
Meanwhile, the U.S. EPA reviewed HARM II in 1987 along with
several other site rating models to determine the best point to start
developing their revised Hazard Ranking System under the NCP5.
Their decision was to continue to develop HRS, adding in those features
felt to be missing, since no existing model met all of their requirements.
This study did lead, however, to an identification of some of the rela-
tive shortcomings in HARM II6. These were:
• There was no soil or air pathway
• The 3 mi. limit on water use was too stringent
• DPM does not consider the quantity of waste at a site
• DPM does not consider pollutant mobility, only groundwater mobility
• There is no provision for including documented evidence of human
exposure
Subsequently, a number of these points as well as those identified
during the comparative testing have been incorporated into the model7.
In November of 1987, the Office of the Secretary of Defense proposed
use of the model (renamed the DPM) for use in ranking DoD sites for
remedial action under the Defense Environmental Restoration Program
(DERP) and solicited comments from interested parties8. Comments
were received from the U.S. EPA and three states; model improvements
have been made in response to these comments.
This paper attempts to provide an overview of the Defense Priority
Model currently being used by DoD to rank sites for remedial action.
OVERVIEW OF THE MODEL STRUCTURE
DPM considers the hazards associated with source materials, path-
ways that may result in exposure and the presence of potential recep-
tors. There are three pathways in DPM:
• Surface Water
• Groundwater
• Air/Soil (considers vaporized compounds and dust).
RISK ASSESSMENT 99
-------
DPM supports both human and environmental receptors, though'the
human receptors are more highly weighted. The environmental recep-
tors include both aquatic and terrestrial populations as appropriate.
Figure 1 demonstrates how the various pathway scores are combined
to yield the six pathway/receptor scores per site. These six scores are
then combined using a root mean square methodology to obtain a sin-
gle site score (Fig. 2). All scores are normalized so that they range
from 0 to 100. This score, by itself, has no meaning and should not
be compared to the MRS ranking number for inclusion on the NPL.
Most sites evaluated to date scored in the 20 to 30 range, but sites have
scored as high as 89 and as low as 3, so a broad range of values can
be expected'.
Coniamlnftlton
ObMfVMl
Yti
NO
AitlQn Pithwiy
Scott ol tOO
Compuli Pwhwiy SCOT*
Bawd on CIWMltrtMkl
P»ihwir
Seora
i
P
\ A
V7
Pl|hw*7 Humin
H»»nrv ico't
Figure 1
How Pathway Scores Are Computed
SwImWMM
•MM HUWl
Scon
u
Surtact w«l«r
Ecotogfctf
Seen
Ground W.l«j
Humill Httllh
Scon
*
Ground W««
Eeotogteal
Scon
X3
Onria Silt Se<
Alirsol
MununttoMh
Scon
«
uiisa
Ecougttti
Scon
Figure 2
How Site Scores Are Computed
PATHWAYS
To more thoroughly understand what is included in the pathway scores,
it is necessary to examine each pathway more closely with regard to
the types of data that are used to obtain a pathway score. Different fac-
tors have different weights. The basic approach is to obtain a score for
each variable and to multiply this score by its weighting factor. The
weighted scores for all factors in a pathway are then added and divided
by the maximum possible score to obtain a normalized value. For each
of the pathways, if a release is observed in that pathway, a maximum
score is assigned. However, this score can be modified by a weighting
based on how well the waste/hazard is contained.
Surface Water Pathway
The surface water pathway of DPM rates the potential for contaminants
from a waste site to enter surface waters via overland flow routes or
from groundwater recharge. If pollutants are not directly observed in
surface water, but are present in sediments or soil, there is a potential
for surface water contamination. The following variables are scored
to provide an indication of this exposure potential:
• Distance to nearest surface water (scores are assigned up to 1 mi.)
• Net Precipitation
• Surface erosion potential (combination of slope and particle size)
• Rainfall intensity
• Surface permeability
• Flooding potential (location within floodplain)
The most important factor by far is flooding potential; net precipita-
tion is the least important. The containment of the waste is also deter-
mined and becomes an important weighting factor.
Groundwater Pathway
The groundwater pathway ranks the potential for pollutant exposure
to occur from contaminated groundwater. If actual groundwater con-
tamination has not been detected but there is contamination in soil or
surface water, there is a potential for future groundwater contamina-
tion. The following factors are scored to obtain a groundwater pathway
score:
• Depth to seasonal high groundwater
• Permeability of the unsaturated zone
• Infiltration potential (measured from net precipitation and the form
of the waste)
• Potential for discrete features in the unsaturated zone to "short circuit"
the pathway to the water table
Waste containment effectiveness is also a weighting factor on the path-
way score. Of the above factors, the depth to the seasonal high water
table is the most important factor.
Air/Soil Pathway
The original HARM model did not have an air/soil pathway. Conse-
quently, it was felt that this model did not account adequately for
exposure resulting from voiatization of organics from the soil or surface
water; neither did it account for exposure to contaminated dust. The
factors that are considered in scoring this pathway are:
Average temperature
Net precipitation
Wind velocity
Soil porosity
Days per year with significant precipitation
Site activity.
All of these factors are weighted evenly. A factor for waste contain-
ment is also used to modify the final score.
CONTAMINANT HAZARDS
The contaminant hazard component of DPM separately rates human
health and ecological hazards of identified or suspected contaminants
in each of the three pathways. Hazard scores are calculated differently
depending on whether environmental contamination has been delected.
For a medium in which contamination has been detected, health hazard
scoring is based on the concept of an acceptable daily intake (ADI).
The highest concentration observed at a site is used. The observed con-
centration is first converted to a daily intake (ug/day) and then is divided
by the appropriate benchmark concentrations (provided in the manual
or on the computer system) which are estimated ADI's. Ecological
hazard scoring for observed contaminants is similar, although an eco-
logical benchmark is used instead. The sum of the ecological hazaid
quotients (concentration divided by the benchmark) is used for all
detected components.
For a medium in which contamination has not been detected, a health
hazard score is based on the ADIs and bioaccumulalion factors of con-
taminants known to be present at the site being rated. In this case, the
score is based on the score for the highest scoring contaminant.
Scoring is similar for all pathways, though the appropriate bench-
marks will vary. For example, if the pathway is surface or goundwater,
aquatic benchmarks will be used as well as terrestrial benchmarks. For
the air/soil pathway, however, only terrestrial factors are employed.
100 RISK ASSESSMENT
-------
RECEPTORS SCORING
The receptors portion of the DPM methodology rates the potential
for human and ecological populations to be exposed to contaminants
from a waste site. The potential receptors are considered separately
for each pathway and for human and ecological targets.
Human Receptors for Surface Water
The following factors are scored to obtain a measure of human ex-
posure to surface water pollution:
• Size of population obtaining drinking water from potentially affected
downslope/downstream surface waters (up to 5 mi.)
• Water use of the nearest surface water
• Population within 1500 ft. of the site
• Distance to the installation boundary
• Land use and zoning within 2 mi. of the site
The first two factors listed above are weighted most heavily.
Human Receptors for Goundwater
The following factors are used as indicators of potential human recep-
tor exposure to contaminants suspected in groundwater:
• Estimated mean groundwater travel time from waste location to
nearest downgradient water supply well(s)
• Estimated mean groundwater travel time from current waste site to
any downgradient surface water body that supplies water for domes-
tic use or for food chain agriculture
• Groundwater use of the uppermost aquifer
• Size of population potentially at risk from groundwater contamination
• Population within 1000 ft. of the site
• Distance to the nearest installation boundary
Of these factors, the estimated groundwater travel time is considered
most important; the water use of the uppermost aquifer also is important.
Human Receptors for Air/Soil
The following factors are used as measures of the potential for human
exposure:
• Size of population near the site (4 mi.)
• Land use in vicinity of the site
• Distance to nearest installation boundary
Land use has the most pronounced impact on the final score.
Ecological Receptors—All Pathways
Exposure of potential ecological receptors is determined by whether
there are sensitive environments (i.e., wetlands or habitats of endan-
gered species) within 2 mi. of the site and whether there are critical
environments (i.e., lands or waters specifically recognized or managed
by federal, state or local government agencies or private organizations
as rare, unique, unusually sensitive or Important natural resources).
COMBINING PATHWAY SCORES TO OBTAIN
A FINAL SITE SCORE
The scores for each pathway are obtained by combining the infor-
mation on the pathway and the hazards for health and ecological recep-
tors. The result are six subscores, one for each receptor/pathway
combination. These scores are then combined using a root mean square
methodology with the human health scores weighted five times heavier.
The final score is then normalized by dividing by the maximum possi-
ble score to obtain a site score ranging from 0 to 100.
AUTOMATION OF THE DPM
The computerized version of the DPM using Prolog has permitted
the introduction of a number of improvements over the paper version.
Some of these are due to the use of expert systems technology while
others are merely due to the greater accuracy and ease of storing and
retrieving data that computers provide10- The new features in the com-
puterized version include:
• Ability to answer a question once even if it is used in several separate
pathways and calculations
Ability to record certainty of answers
Ability to automatically convert units
Ability to use alternate data if information is missing
Range checking of answers
Ability to change responses and to rapidly recalculate a final score
In addition, the automated version can generate a report that includes,
in addition to the scores, full documentation of the final score through
comments and the certainty indication. The automated version also con-
trols the user's passage through the model and only presents those
requests for information that are deemed necessary depending on pre-
viously supplied answers.
FUTURE DIRECTIONS
Work is progressing on DPM and the experience of using it for the
FY-90 scoring will create a large body of data on actual sites. These
data will be analyzed and changes in the weightings used in DPM will
be incorporated where they are felt to be necessary. There is also a
plan to convene a group of experts to determine whether additional data
should be included in the model to facilitate future decisionmaking.
There are also plans to incorporate more expert system features such
as logical checking across related responses, more table look-up features
and increasing the size of the benchmark data base.
ACKNOWLEDGEMENT
Work on DPM was conducted under contracts with the US Air Force's
Occupational and Environmental Health Laboratory located at Brooks
AFB, Texas. The Technical Program Managers for this work were Capt.
Art Kaminski and Phil Hunter.
REFERENCES
1. Barnthouse, L.W., Breck, I.E., Jones, T.D., Kraemer, S.R., Smith, E.D.
and Suter II, G.W., Development and demonstration of a hazard assessment
rating methodology for Phase II of the Installation Restoration Program,
ORNL/TM-9857, Oak Ridge National Laboratory, Oak Ridge, TN, 1986.
2. Barnthouse, L. W., Breck, I.E., Suter II, G. W., Jones, T.D., Easterly, C,
Glass, L., Owen, B.A. and Watson, A.P. Relative toxicity estimates and
bioaccumulation factors in the Defense Priority Model, ORNL-6416, Oak
Ridge National Laboratory, Oak Ridge, TN, 1986.
3. Smith, E.G. and Barnthouse, L. W, User's Manual for the Defense Priority
Model, ORNL-6411, Oak Ridge National Laboratory, Oak Ridge, TN, 1986.
4. Hushon, J.M., Mikroudis, G.M. and Pandit, N., Final Report on Phase
I of DPM, Roy F. Weston, Inc., West Chester, PA, 1988.
5. Industrial Economics, Inc., Analysis of Alternatives to the Superfund Haz-
ard Ranking System, Cambridge, MA, 1988.
6. Parker, W.H., Jr., Deputy Assistant Secretary of Defense letter to J. Cannon,
Acting Assistant Administrator for the Office of Solid Waste and Emergency
Response, U.S. EPA, July 26, 1989.
7. Hushon, J.M., Mikroudis, G.M., and Subramanian, C., Defense Priority
Model (DPM) User's Manual, Version 2.0, Roy F. Weston, Inc., West Chester,
PA, June 1989.
8. Federal Register, 52, no. 222, p. 44304-5, Nov. 1987.
9. Hushon, J.M., Mikroudis, G.M., and Pandit, N., op. cit.
10. Hushon, J.M., Mikroudis, G.M., and Subramanian, C., Automated Defense
Priority Model (ADPM) User's Manual, Version 2.0, Roy F. Weston, Inc.,
West Chester, PA, June 1989.
RISK ASSESSMENT 101
-------
Uncertainty Analysis for Risk Assessment
Chia S."Rocky" Shih, Ph.D., P.E.
University of Texas
San Antonio, Texas
Ren-Yio Cheng
Division of Engineering
Hazardous Materials Control Resources Inc.
INTRODUCTION
Within the context of the prime questions for risk assessment in
hazardous waste management, i.e., "How is the risk estimated?," "What
is the magnitude of the risk?" and "Is this risk acceptable?," the con-
cern for the uncertainties becomes the focal point of all decision-makers.
Mathematically, risk can be defined as a function of the probability
of a negative consequence occurring and the value of that consequence.
Therefore, the uncertainty included in the risk assessment may be com-
posed of the uncertainties associated with: (1) the probability estima-
tion (2) the perception of potential consequences (3)the functional
relationships involved and (4) the acceptable risk limits. Consequently,
the uncertainty analysis addressed herein must deal with all these
uncertainties simultaneously.
For the probability estimation, a method based on event/fault tree
analyses is developed for the convenience of review and revision.
Through the anatomy of risk, the perceptive confusion of the risk can
be precipitated. Applying the concept of revealed preference, the risk
acceptability is analyzed. Sensitivity analysis is utilized to evaluate the
variability of different risks and their acceptabilities while meeting the
prescribed confidence limit. The risk assessment for dioxin analysis
in a laboratory setting is used as an illustrative example for the uncer-
tainty analysis.
RISK FOR TCDD SAMPLE ANALYSIS
The risk problem addressed here is concerned with the determina-
tion of the specific level of safety measures required for the
tetrachlorodibenzo-p-dioxin (TCDD) sample process and analysis at
a U.S. EPA regional laboratory.
To develop a range of potential risk situations, the laboratory proce-
dures followed were those used by U.S. EPA laboratories and their ton-
tractors, beginning with sample packaging in the field and continuing
through the final disposal of the TCDD sample residue. In addition,
it was assumed that the laboratory was located in a populated
office/residential complex. The basic risk elements included the
following key steps of the risk pathways (Figure 1):
Sample packing in the field
Trans-shipment of packaged sample
Pre-analysis storage
Sample cataloguing or inventory
Extraction and cleaning of the sample in the laboratory
Concentration and digestion for 2,3,7,8-TCDD
GC/MS sample preparation
Intra-laboratory transport of the prepared concentrates of dioxin
GC/MS analysis
Data log-in for the computer
Disposal of the residue of sample and wastewater
• Contamination of the air in the building and to the surrounding
community
I / !»-.<• I I «,u-*"' I
• ...«• I I ..KKT> I
V •'••"" V \ \
Figure 1
Elements of Risk for Dioxin Analysis
RISK ESTIMATIONS
To prepare for the risk estimation analysis, a series of alternative risk
scenarios was developed (Fig. 1). In developing the risk scenarios,
alternative risk occurrence pathways and exposure situations for the
laboratory personnel, the co-workers in the building and the surrounding
communities were considered.
To illustrate the risk relationship between potential hazards and events
that may result from contamination to laboratory personnel, co-workers
located in the same building and the surrounding populace, a fault tree
was constructed.
Contamination pathways for the dioxin may include one or more
independent pathways. This is clearly delineated by the separate path-
way columns in the fault tree. The hierarchy of the fault tree structure
is established by the horizontal levels in the fault tree.
102 RISK ASSESSMENT
-------
The mathematical relationship quantifying the probability of dioxin
contamination during its sample analysis was developed based on the
accompanying fault tree and the assigned probability values of specific
events as designated by the alpha-numeric variables printed next to the
event in the fault tree. The fault tree formula is shown in Table 1.
The variables used to define specific events are consistent with the
designation of tree branches in the event tree (Fig. 2). For instance,
A2 designates the event of contamination due to the sample packaging
and A2a designates the event of contamination caused by the contami-
nated vermiculite and/or plastic bags used in the packing. The only
unique events included in the fault tree analysis are those events
associated with the final consequences of the dioxin contamination such
as events A2cl or A2c2.
Tiblel
Probability Model Formulation
P(T) • P(A2)+P(A+)-t-Any coabinaclon of
(A2.B2.C2 . . .02)
WHERE
P(A2) - PCA2a)!1'(A2al/A2a+P(A2a2/A2a»
+ P(A
-------
Figure .
hvcnl Iree for Risk Scenarios
potential consequence to both individuals and the society must be
analyzed in detail. First, the basic characteristics of the risk must be
defined and delineated. Second, the incremental risk acceptance value
for each of the risks identified in terms of risk referent shall be deve-
loped. Finally, the objective risk value computed, based on the poten-
tial consequence, should be compared with the risk referent value to
determine the acceptability of the current practice.
Based on the risk classification as outlined in Table 2, the risks
associated with the dioxin laboratory analysis can be characterized as
immediate siaiisiic.il accidents and categorized as follows:
• Risk for the laboratory personnel specifically assigned to the dioxin
analysis: ordinary voluntary risk
• Risk for the co-workers located in the same building: ordinary volun-
tary regulated risk
• Risk for the surrounding community of the U.S. EPA Regional labora-
tory: ordinary involuntary risk.
The procedures to be followed for the determination of risk accepta-
104 RISK ASSESSMENT
-------
VI
ca
%
m
kie: Su.pl*
&••.•(•
A2cl: Coo-
A2c2: Coot-
••toctln to
Workrr*
Accident
Figure 3
Fault Tree
-------
Table!
Risk References for Immediate Statistical Accidents
Risk Fatalities
Classi- Per
fication Year
Naturally
Occurring:
catastrophic: 1 x 10"6
Ordinary: 7 x 10~s
Man
Originated
Catastrophic-
Involuntary 1 x 10^1
Voluntary 2 x 10~°
Regulated
Voluntary 3 x 10~5
Ordinary-
Involuntary 5 x lO^j
Voluntary 1 x 10"4
Conseqi lumps
Health Affects
Year
5 X 10~6
4 X 10"4
5 X 10~7
2 X 10"6
3 X 10~6
3 X 10"!
6 X 10~2
Prop, rjomage
MS/Per
Year
0.02
3
2 x 10~2
.4
.4
1
200
Reduction of
Life Span
(Years)
3 x 10"2
0.2
3 x 10~4
6 x 10"4
6 X 10"2
—5
1 x 10 z
0.1
Regulated
Voluntary 1 x 10"4
6 x 10'
,-2
30
Man Triggered
Catastrophic-
Involuntary 2 x 10'
Voluntary 4 x 10~
Ordinary
Involuntary 1 x 10
,-7
1 x 10
4 X 10"
,-6
4 X 10
0.8
,-2
0.1
6 X 10~*
6 X 10~3
Voluntary 1 x 10'
Regulated
Voluntary 2 x 10~
r5
,-3
3 X 10
2
0.2
,-2
Source: Rove, W., An Anatomy of Risk, John Wiley & Sons, NY, 1977.
The procedures to be followed for the determination of risk
acceptability for the TCDD processing centers are:
o Develop an appropriate risk classification scheme.
o Determine the risk reference value for each class of risk
encountered in the dioxin analysis procedures.
o Compute risk referents for each class of risk.
o Compare the estimated risk from fault tree analysis with the
risk referent values.
bility for the TCDD processing centers are:
• Develop an appropriate risk classification scheme
• Determine the risk reference value for each class of risk encoun-
tered in the dioxin analysis procedures
• Compute risk referents for each class of risk
• Compare the estimated risk from fault tree analysis'with the risk
referent values
In view of the risk confronted by different sectors of population in
the laboratory and its surroundings, appropriate risk classifications
developed for each sector of the population are summarized in Table 3.
The dioxin analysis is essentially a typical man-originated, ordinary
event. However, since the reliability and statistical validity of existing
data characterizing various consequences of the dioxin exposure acci-
dents are absent, only piecemeal information covering personal inju-
ries and immobility could be collected and reviewed. Thus, the risk
reference value characterizing the personal injury in terms of health
effects per year is the only consequence included in the risk accepta-
bility evaluation as shown in Table 3. In fact, based on limited data,
personal injury seems to be the only visible and pronounced conse-
quence due to dioxin exposure being reported so far.
OBTAINING RISK REFERENT VALUES
These risk reference values are estimated directly from historical and
societal risk data that are analogous to the situations and consequences
involved in dioxin analysis. Transforming the risk reference values into
appropriate risk referents requires the following four steps:
• Determine the appropriate risk proportionality factor (Fl) which
incorporates the societal attitude due to its expectations associated
Table 3
Summary of Risk References for
Dioxin Exposure in Laboratory
Type Risk
Personal injury or
immobility for
laboratory workers
Personal injury or
immobility for
co-workers in the
laboratory
Personal injury
to the population
in local oomunity
Risk Classification
Voluntary, ordinary
ordinary regulated
voluntary man
originated
ordinary involuntary,
nan originated
Risk Reference
(Hlth. Eff./Yr.)
3 x 10"1
6 x 10
3 X 10
,-2
with the degree of voluntarism of the affected population.
• Determine the appropriate risk proportionality derating factor (F2)
which discounts the existing societal risk acceptable level due to the
indirect benefit/cost balance considerations for the dioxin exposure
via laboratory analysis (Table 4).
• Develop and quantify the risk controllability factor (F3) which charac-
terizes the basic control approach, the degree of control, the state
of implementation and the judgment of control effectiveness (Table 5).
• Determine the referent using the factors derived in the above three
steps by the formula:
Risk Referent = (Risk Reference) x Fl x F2 x F2 (1)
These factors are subjective. The first two factors in Equation 1
address the inherent propensity of effected populations to take risks
and also incorporates the additional decision dimension of indirect
benefits/cost balance. This acknowledges the tendency of people to
accept higher levels of risk when the potential benefits far outweigh
the potential costs. On the other hand, people may become increasingly
risk aversive when the potential benefits are likely to be offset by the
costs.
The risk proportionality and its derating factors, as determined for
different sectors of populations, are shown in Table 4. Though not to
the same degree, the controllability factor also varies due to the target
population, as shown in Table 5.
Incorporating all the factors determined above, the appropriate risk
referents for different affected population sectors are derived as shown
in Table 6.
Table 4
Risk Proportionality and Risk Proportionality Derating Factor
Factor Value
Proportionality Factor 1.0
Derating Factor
Laboratory Worker 0.2
Oo-Worker In the Building 0.1
Surrounding Ocraamity 0.1
RESULTS
On the basis of the above judgments and limited data, the risk referent
values were compared to the estimated values the risks for both techni-
cians and their co-workers in U.S. EPA laboratory are considered to
be acceptable. On the other hand, the risks for the surrounding
community may be marginally acceptable.
106 RISK ASSESSMENT
-------
Tables
Controllability Factor
Control Degree of State of Control
Approach Control Inplementn. Effectiveness
laboratory Worker
1.0
1.0
0.5
1.0
Co-Workers in the
Building
Surrounding
Cmunity
0.5 1.0 0.5 1.0
0.3 0.3 0.1 0.5
Table 6
Risk Referents and Estimated Maximum Risks
Population
Sector
USEPA lab
Personnel
USEPA Cb-
Workers in the
Saras Building
Surrounding
Conmmity
Risk
Reference
3 XlO"1
6 x 10~2
3 X 10~5
Risk
Reference
3. x 10~2
1.5 x 10~3
1.4 x 10~8
Est. Max.
Risk
3.2 X
1.2 x
1.1 x
ID'3
ID'3
lO'6
SENSITIVITY ANALYSIS
Sensitivity analysis is a post-solution evaluation technique, intended
to determine the degree of confidence which can be placed on the
selected solution. In this dioxin analysis example, a wide range of proba-
bility values for all events included in the fault tree as shown in Ta-
ble 7 has been evaluated. Monte-Carlo simulation has been used to
analyze the variability of the estimated risk values. The estimated
maximum risks cited in Table 6 have a confidence limit of 95%.
Thus, all the risks shown in Table 6 can be considered as conserva-
tive judgmental values for all potential accidents described in the fault
tree. A difference of magnitude in the order of two to three may still
be within the range of cumulative errors. In reality, the risk estimated
for the community may be too high and it is therefore considered to
be marginally acceptable.
For the facility considered, the sensitivity analysis for specific faulty
events indicates that with minimum modification of sample inventory,
waste disposal procedures and installation of paniculate air filters, the
risk to the community can be significantly reduced to 1.1 x 10"8 which
is well within the acceptable level. In addition, maximum estimated
risk for the co-workers may be lowered from 1.2 x 10'3 to 2.1 x 10"4
which is well below the acceptable level of 1.5 x 10'3 as reflected by
the corresponding risk referent values.
CONCLUSIONS
Uncertainty Analysis for the risk assessment in hazardous waste site
management can be resolved by the three-prong attack:
• Develop a structured risk estimation model based on an integrated
event/fault tree analysis
• Based on a detailed anatomy of the risks involved, conduct the risk
acceptability analysis using the revealed preference concept
• Perform a comprehensive sensitivity analysis for the estimated risks
to determine the confidence limit of the risk values
In the illustrative example of the risk assessment for the potential
hazards to laboratory workers, co-workers and the surrounding com-
munity due to the dioxin sample analysis, we have determined that:
• The risk to both laboratory workers and their co-workers in the
building is acceptable.
• The risk to the surrounding community may be considered marginally
acceptable. However, with minimal modifications to the facility, the
risk can become acceptable.
REFERENCES
1. Rowe, W. An Anatomy of Risk, John Wiley & Sons, New York, NY, 1977.
2. Shin, C.S. and Ess, T., "Perspectives of Risk Assessment for Hazardous Waste
Management,'—Proc., Third National Conference on Uncontrolled Hazardous
Waste Sites, HMCRI, Silver Spring, MD, pp. 408-413, 1982.
3. Veseley, W., et al., Fault Tree Handbook, U.S. Nuclear Regulatory Cora-
mission, Washington, DC, NURGC-0492, 1981.
RISK ASSESSMENT 107
-------
Incorporating Time Varying Parameters
In The Estimation of Human Health Risk
From Superfund Sites
Alison C Taylor
David E. Bui-master, Ph.D.
Alceon Corporation
Cambridge, Massachusetts
ABSTRACT
Risk assessment is a critical step in decision-making in federal and
local governments as well as the private sector. It combines informa-
tion about the frequency, intensity and duration of human exposures
to chemical hazards with data on the toxicity of those compounds to
yield estimates of the risk of carcinogenic and non-carcinogenic effects
associated with those exposures.
Given the complexity and uncertainty inherent in human exposures
there is a need to design great flexibility into the risk estimation process.
Calculations often are repeated with slight variations in particular inputs
to identify the contribution of individual chemical compounds and
exposure pathways to the health risk posed by a complicated scenario.
For this reason most analysts now perform risk assessments on
microcomputer spreadsheets.
To accommodate the need to account for time varying inputs in risk
assessments, we have developed a system on Macintosh computers in
which programmed macros pass data and intermediate results between
Microsoft Excel™ spreadsheets. The system has the capability to
model time varying exposures in the estimation of the average daily
dose of a substance for assessing the risk of both acute and chronic
effects. Both human parameters i.e. skin surface area, breathing rate
and body weight which vary with age and environmental parameters
(i.e. rate of emissions from an incineration facility) may be assigned
different values for each year of exposure.
The linked spreadsheets are designed to calculate exposure doses of
chemical contaminants via inhalation, ingestion and dermal pathways.
The carcinogenic risk for chronic effects and the "hazard index" for
non-carcinogenic effects are automatically estimated for every chemi-
cal compound, pathway and receptor and then summed across all
pathways and compounds to yield an assessed risk for each receptor.
INTRODUCTION
Early in 1989, Alceon Corporation was asked to perform a risk
assessment for a proposed Park & Ride commuter rail station. The
station is proposed to be built on the site of a former municipal solid
waste (MSW) landfill. Portions of the landfill not covered with asphalt
during construction of the Park & Ride facility will continue to receive
deposits of ash for 7 yr from a MSW incinerator also located in close
proximity to the proposed station.
In order to estimate the risk to human receptors at the Park & Ride
facility, it was necessary to calculate the average daily dose (ADD) of
each chemical compound of concern, to each category of receptor,
through all exposure pathways. Many of the parameters used in these
calculations take on a series of values over time due to changes in
incinerator operation, variations in gas generation rates from the MSW
landfill due to the aging of buried waste and the variations in body weight
and inhalation rate of receptors.
108 RISK ASSESSMENT
As an alternative to using average values for these time varying
parameters in the dose calculations, a series of spreadsheets containing
detailed parameter information for each year of exposure and facility
operation has been developed. These spreadsheets are linked by macros
programmed: (1) to calculate the average daily dose of contaminant to
each type of human receptor in each year of exposure: (2) to locale
the year of maximum exposure (the year in which average daily dose
is greatest) to be used in the estimation of potential acute health effects
and (3) to calculate the average exposure across a 70 yr lifetime (the
average of the average daily doses for all 70 yrs of life) for the estima-
tion of potential chronic health effects.
DISCUSSION
The receptors of concern in this case are adult and child (student)
commuters using the Park & Ride facility and a security guard posted
in the facility parking lot. Exposures to the security guard are assumed
to continue for an entire 45 yr career, (i.e., age 20 until retirement at
65 yrs of age), 8 hrs/workday. 5 workdays/wk. It is extremely conser-
vative to assume that one employee would hold this position for 45 yrs.
Exposures to commuters are assumed to occur for 0.5 hr. each
commuting day as the commuters wait for the train and walk to and
from their car or ride. Adults are assumed to commute for an entire
45 yr career (age 20 until 65, again a very conservative assumption)
whereas children are assumed to commute for 6 yrs of secondary school
(between the ages of 12 and 18 yrs). The security guard and adult com-
muter are assumed to experience exposures
50 wk/yr (a 2-wk vacation is assumed). The child commuters are
assumed to commute to school 40 wks/yr.
Sources of contaminants to the air at the commuter rail facility include
gaseous and paniculate stack emissions, gas generated by the buried
waste present in the landfill and paniculate material released during
the transport of ash and its disposal in the landfill. The most toxic
carcinogenic and non-carcinogenic compounds released by these sources
were identified in a sequential ranking exercise and designated as indi-
cator compounds for the risk assessment.
Portions of the landfill not covered with asphalt during construction
of the facility will remain active, receiving ash from the MSW in-
cinerator for 7 yrs. It is estimated that the landfill will reach capacity
after 7 more years of operation and it will be capped at that time. Starting
in year 7, the ash will be transported elsewhere, eliminating the par-
ticulate contributed by fugitive dust from the landfill and landfilling
activities. The gas generated by the landfill will continue to contribute
to contaminant levels in the air at the commuter station since the capping
does not include a collection system for gases. All active sections of
the landfill will be fenced off, preventing direct contact with the soil
by humans. Due to the security provided by the fence, human health
risks may result only from airborne contaminants.
-------
Pathways of exposure considered in the risk assessment include:
(1) inhalation of organic and inorganic compounds in gaseous form,
(2) inhalation of organic and inorganic compounds in particulate form,
(3) (inadvertent) ingestion of organic and inorganic compounds con-
tained in dirt adhered to skin during daily work and (4) dermal penetra-
tion of organic compounds also contained in dirt adhered to skin.
The inhalation exposure model used to estimate the dose of chemi-
cal contaminants to the child commuter by inhalation is examined here
in detail since it incorporates several time varying parameters and
provides an example of the technique used to handle these parameters.
Inhalation Exposure Model
The Inhalation Exposure Model is used to estimate the average daily
dose (ADD) of a specific chemical to an individual exposed to air con-
taining a known concentration of contaminant. All dose estimates are
measured in units of milligrams of bioavailable chemical per kilogram
of body weight per day. All the dose calculations are based on an esti-
mation of the average daily dose on a day of exposure. Averaging factors
are then used to calculate the average daily dose averaged over each
year of exposure. The potential for health effects from compounds with
systemic (non-carcinogenic) potencies, the Hazard Index, is estimated
by dividing the maximum yearly ADD (the largest average daily dose
occurring in any single year) by the reference dose for acute health effects
(RefD) provided by U.S. EPA.5
Hazard Index = ADD^ / RefD
(1)
The estimated carcinogenic risk associated with exposure to the levels
of carcinogenic contaminants present at the facility is calculated by mul-
tiplying the CPF (cancer potency factor provided by U.S. EPA,5 by the
average of the yearly ADDs across a 70-yr lifetime. This calculation
yields a unitless carcinogenic risk estimate.
(2)
Risk = CPFx ADD,.&
lifetime avg
As stated above, all calculations of average daily dose were performed
by macros, written in Microsoft Excel™ for Macintosh computers,
which link a series of spreadsheets, each containing some portion of
the information required for the calculations. Table 1 is an excerpt from
the spreadsheet in which the lexicological properties of the indicator
compounds are stored. Since all the spreadsheets are linked together,
calculations performed in other spreadsheets that require the cancer
potency factors (CPFs) and reference doses (RefDs) draw the values
directly from the lexicological properties table. Since each parameter
value is stored in a single place, rather than in every spreadsheet in
which it is used, updating the values with new information is very easy
and efficienl.
Table 1
lexicological Properties of Indicator Compounds
Indicator
Compound
Organic Compounds
benzene
benzo(a)pyrene
nwthylene chloride
carbon tetrachlorlde
vinyl chloride
1,1-dlchloroethylene
1 ,2-t-dlchloroethylene
1,2-dlchloroethane
trlchloroethylene
tetrachloroathyleno
hexachlorobenzene
2,3,7,8-TCDD
Inorg. Compounds
arsenic
beryllium
cadmium
chromium III
chromium VI
copper
lead
nickel
zinc
hydrogen lulllde
Sourcu:
a US EPA, 1988, IRIS
° US EPA, 1987, PHRED
e US EPA, 1986, SPHEM
Inhalation
Cancer
Potency
Factor
(mg/Kg/d)-1
I
2.90E-02,.
6.10EtOO
1.40E-02
1.30E-01
2.9SE-01
1.20E+00
9.10E-02
1.30E-02
3.30E-03
1.70E+00
1.56E*05
5.00Ł«01
8.40E+00
6.10E«00
4.10E-.01
1.19E+00
Inhalation Inhalation Inhalation
Weight CPF AIC or
of Source Reference
Evidence Dose
(-) (-) (mg/kg/d)
A
B2
BZ
B2
A
C
B2
B2
B2
A
B2
B1
A
D
B2
A
1.00E-02
6.00E-02
7.00E-04
9.00E-03
2.00E-02
1.00E-02'
8.00E-04
5.00E-04
5.10E-03
l.OOE-02
4.30E-04
l.OOE-02
3.00E-03
d GRI, Vol. 3, 1988
a US EPA, 1988, Special Report
0 used other pathway as surrogate
Inhalation
AIC or
RFC
Source
(-)
d
*
*
#
*
«
*
The spreadsheet macros allow the time varying nature of individual
parameters to be accounted for in the averaging of dose across time.
In the case of exposure by inhalation of contaminants, the following
parameters are assigned values in the macros that vary with the age
of the receptor: body weight and inhalation rate. In addition, other
parameters are assigned values in the macros that vary with time such
as the rate of gas generation from the landfill as the MSW ages and
both the concentration of indicator compounds in particulate form and
the fraction of particulate in the air which may be attributed to the site,
before and after the landfilling of ash ceases.
The inhalation dose model follows:
ADD = [(Ca x Ir x Te x (5 day/7 day) x (40 wk/52 wk)] / Bw(3)
Add = average daily dose of a chemical to an individual
(mg/kg/day)
= concentration of contaminant in inhaled air (mg/m3)
= inhalation rate (mVhr)
= time duration of exposure per day of exposure (hr/day)
= body weight of individual (kg)
As stated above, of the parameters used in the above calculation, Ir
and Bw vary over the 6 yrs of child commuter exposure. The average
daily inhalation rate of individuals varies with growth, with the peak
rate occurring in the teenage years. The inhalation rate of commuting
students involved in moderate activity was assigned the following values
in the ADD calculation (data adapted from Snyder, et al.,4 and
Anderson, et al.,1 for the U.S. EPA,4:
Ca
Ir
Te
Bw
Inhalation Rate
(nWhr)
1.7
1.5
Age
(yr)
12 - 14
15 18
The average body weight of an individual increases with age until
about age 17 at which point the average body weight reaches the average
adult body weight, 70 kg (adapted from data presented by Snyder, et
al.,4 and Anderson, et al.1, for the U.S. EPA, 4:
Body Weight
(kg)
45
50
55
60
65
70
Age
(yr)
12-13
13-14
14-15
15-16
16-17
17-18
In addition to the variation in characteristics of the exposed indi-
viduals, the concentration of contaminant in the inhaled air, Ca, was
projected to vary, due to variations over time in the generation of gas
by the landfill. Depending on the length of time since disposal of the
waste, the gas generation rate is known to vary significantly.3 For the
6 yrs of exposure to an individual child commuter, a single value of
annual gas generation in m3 gas generated per kg of waste deposited
was selected; however, across the 45-yr exposure durations of both the
adult commuter and the security guard, the gas generation rate from
the landfill, and thus the concentration of contaminant in air, were
assigned a series of declining values.
The concentration of contaminant in air also varies due to planned
changes in the operation of the landfill, such as the cessation of the
disposal of ash after 7 yrs of Park & Ride facility operation. Since ash
disposal in the landfill would occur only in the first 7 yrs of facility
operation, the fraction of the concentration of particulate material
occurring in the air as fugitive dust from the trucking and disposal of
the ash is not included in the total concentrations of particulate to which
adult commuters and the security guard were exposed after year 7. The
child commuter, however, has an exposure duration of 6 yrs; years;
therefore, the particulate contribution from ash landfilling activities is
included throughout the child commuter's exposure.
RISK ASSESSMENT 109
-------
Table 2 is an excerpt from one of the spreadsheets that performs the
health effects calculations. Airborne concentrations of contaminants and
various exposure factors are called to this spreadsheet by macros. The
macros use the parameter values for each year of interest to calculate
the ADDs. The actual macros are not shown in the excerpt. The ADDs
are then passed to a summary spreadsheet such as the one shown in
Table 3. The summary spreadsheet calculates the hazard index and in-
cremental lifetime carcinogenic risk attributed to each indicator
compound. Finally, the total hazard index and incremental lifetime risk
estimate across all indicator compounds, associated with a single path-
way (such as the inhalation of gaseous contaminants) for a single receptor
(such as the child commuter), are tabulated.
Table 2
Excerpt from Spreadsheet Health Effects Calculation
General
Variables
Hours per
Day
—
0.5
Days per
Week
—
5
Weeks per
Year
—
40
Bioavai lability
1
Age
TCOO
Inorganic Compounds
hydrogen sutnde
OJrv
1.0 prsTcsse
1 jO postfcsse
U6E-04
4-MEX*
2J7E-04
2JtE-M
rxje-oj
4.45E-05
147E-04
1.19E44
7.74E-04
4.7CE44
1MIJ&
4JOE-12
UtE-03
IMtMt
4.77E-11
2.77E-O8
xate-M
7.71 E-07
4.62E-07
J.TlE-Oe
1J4E-OS
S.03E-W
4-ME-OI
240E-11
4J7E-14
U8E-04
Tat
1O1E-07
a.ne-12
1JOE47
241 E4*
&42E-0*
101 E-M
141E-07
L04E-0*
liJŁ-07
122E47
1.C2E-1!
1L04E-1S
230E46
4.77E4*
4J1E-M
U1E4S
S.13&OS
1JSE-04
4.04E-04
X11E4*
1.13E42
_
1.21641
Ut&M
1JOE.11
U2E«
241E-U
i>He«
U1E4C
7J2t«
UOE-Ot
1M&M
17K-12
4.74E-10
_LJ_
7JBB«
REFERENCES
1. Anderson E . et al., "Development of Statistical Distributions or Ranges of
Standard Factors Used in Exposure Assessments," U.S. EPA.. Revised Draft
Final Report, Office of Health and Environmental Assessment, CoMnct
No. 68-02-35K), Sept. 1984.
2. GRI. "Management of Manufactured Gas Plant Sites. Vblume ID, Risk
Assessment." GR1-S7-0260.3, May 1988.
3. Lockman & Associates, "Landfill Gas and Landfill Gas Constituent Emis-
sions Modeling, for a Hypothetical 3000-ton per day Southern California
Landfill." Monterey Park. CA, 1985
4. Snyder W.S.. Cook M J . Nasset E.S.. Kartiausen L.R. Howells G.P.. and
Tipton I.H.. "Report of the Task Group on Reference Man," International
Commission on Radiological Protection Number 23, Permagon Press. Oxford,
1975.
5. U.S. EPA. "Supernmd Public Health Evaluation Manual (SPHEM)," Office
of Emergency and Remedial Response, Washington, DC EPA/540fl-«6rTJ60,
1986.
6. U.S. EPA. "Public Health Risk Evaluation Database (PHRED)." Office of
Solid Waste and Emergency Response. Washington, DC 1978.
7. U.S. EPA. "Integrated Risk Information System (IRIS)," Vblume 1 and Elec-
tronic Information System, Office of Health and Environmental Assessment,
EPA/60GV8-86/032a, Mar. 1987
8. US. EPA. •'Special Report on Ingested Inorganic Arsenic: Skin Cancer. Nutri-
tional Essentiality," EPA/625/3-87/013. Science Advisory Board, Washing-
ton, DC, 1988.
110 RISK ASSESSMENT
-------
Terrestrial Food-Chain Model for Risk Assessment
Jeffrey J. Tasca
Michael F. Saunders
Richard S. Prann
IT Corporation
Edison, New Jersey
ABSTRACT
To assess the potential impact of a proposed Hazardous Waste
Incinerator in Niagara Falls, New York on terrestrial wildlife species,
IT Corporation developed a simplified food web model to predict body
burdens of selected constituents of concern. Appropriate habitat areas
within 5 km of the incinerator were identified and eight species were
selected for a detailed assessment. Only areas capable of supporting
long-term habitation for the selected target species were considered
appropriate for selection. An exposure assessment was performed for
each species at each habitat.
INTRODUCTION
For the purpose of assessing the impact of the stack emissions to
terrestrial species, appropriate habitat areas within 5 km of a proposed
hazardous waste incinerator (HWI) were identified and eight species
were selected for a detailed assessment of the impact of non-carcinogenic
compounds. Five mammalian species were evaluated to assess potential
effects of carcinogens. An exposure assessment was performed for each
species and the risks associated with these exposures were calculated.
To conduct this terrestrial species assessment, assumptions and
adjustments to lexicological data generally available in the literature
were made. The majority of these data are derived using standard
laboratory animals or agricultural crop species. Plants and animals in
the natural environment tend to have longer exposures than those in
the laboratory due to life span, multiple exposure pathways and
differences in subspecies metabolism. The application of uncertainty
factors provides a conservative adjustment for the use of laboratory
derived data. General assumptions used in this assessment regarding
incorporation of the selected indicator constituents into the biological
system are:
• Toxicity is assumed to be independent of dosing schedule
• An average daily food/water consumptions are used for all calculations
which assume no variation
• The food-chain model used in this assessment incorporates emissions
into single trophic levels of the food chain with bio-accumulation
and bio-magnification at subsequently higher species levels
• The HWI produces emissions for an indefinite time
A screening methodology was applied to the waste stream and 11
constituents were selected as indicators for evaluation. Ambient air
concentrations and deposition rates for the selected constituents of
concern were calculated using the U.S. EPA's ISCLT air disperion
model.
ECOLOGICAL ASSESSMENT END-POINTS
Because of the complexity of interactions within a food-chain, it is
difficult to assess the potential impacts to all receptors for all end-points.
Receptors (the selected target species) are the components of the
ecosystem that may or may not be adversely affected by the selected
indicator constituents. End-points are the particular types of impacts
a constituent has on a receptor.
Possible end-points for ecological risk assessments can be divided
into four levels: individual; population; community; and ecosystem.
These levels may be further assessed as:
Individual end-points of biological interest
• Changes in respiration
• Changes in behavior
• Increased susceptibility to illness
• Decreased growth
• Death
Population end-points of biological interest
• Decreased genotypic and phenotypic diversity
• Decreased fecundity
• Decreased growth rate
• Increased frequency of disease
• Increased mortality rate
Community end-points of biological interest
• Decreased species diversity
• Decreased food web diversity
• Decreased productivity
Ecosystem end-points of biological interest
• Decreased diversity of communities
• Altered nutrient cycling
• Decreased resilience
Presently, there are no regulatory standards concerning individual
end-points of biological interest for non-human terrestrial species. There
is, however, a general consensus defining adverse effects at the
population level. For this reason, this level was chosen as the most
appropriate end-point for use in terrestrial species assessments.
HABITAT EVALUATION
Areas inscribed by concentric 400 mi circles radiating out for 5 km
from the HWI were addressed based on the determined ISCLT
depositional pattern. Based on these results, no areas of high deposition
beyond this radius were evident. Areas that could support the selected
target species were delineated. Only areas capable of supporting
long-term habitation of the target species were considered appropriate
for selection. This selection process was based on the following criteria:
habitat must have a sufficient receptor-specific food supply, adequate
area to accommodate the receptor's normal range, sufficient water supply
and a lack of continuous intervention by man. Areas selected are
RISK ASSESSMENT 111
-------
representative of or a composite of the following biomes: Grasslands,
Temperate Forest and Taiga.
SELECTION OF TARGET SPECIES
Not all organisms are suitable for use as target species to evaluate
constituent impacts. General considerations and assumptions must be
applied in selecting target species3. The following criteria were applied
in the selection of target species:
• Target species must be capable of accumulating the selected indicator
constituent to measurable amounts
• Target species should be easily collected or observed and be available
should field calibration or verification studies become necessary
• Relevant information pertaining to interactions between the target
species and the selected indicator constituent(s) should be available
in the scientific literature
• Target species should, as a group, represent all levels of the food web
• Target species should represent various exposure pathways
Target Species
The representative species of wildlife selected for this assessment are:
Avian:
Buteo jamaicensis (Raptor) Red-tailed Hawk
Philohela minor (Non-passerine) Woodcock
Mammalian:
Blarina brevicauda (carnivore) Short-tailed Shrew
Marmota monax (herbivore) Woodchuck
Odocoileus virginianus (herbivore) White-tail Deer
Sylvilagus floridanus (herbivore) Cottontail Rabbit
Wilpes vulpes fulva (omnivore) Red Fox
Reptilian:
Chelydra serpentina (omnivore) Snapping Turtle
Figure 1 shows a general review of the potential routes of exposure
and Figure 2 shows the simplified food web relationship of these target
species. Individual pathway parameters of exposure are presented in
Table 1.
EXPOSURE ASSESSMENT PATHWAY
MEDIA CALCULATIONS
From previously derived emission rates, the 1SCLT model and the
California Air Resources Board deposition algorithm were used to
calculate the total deposition rate and air concentration at the defined
receptor locations.
The deposition rate was used to determine the average soil
concentration which, in turn, was used to estimate the concentration
accumulated in vegetation from root uptake at each of the selected
habitats. Emission rates were used to calculate surface water
concentrations which, in turn, were used to estimate the body burden
of fish in the Niagara River and Gill Creek within a 5 km radius of
the HWI. The air concentration from the initial air modeling effort was
used to calculate the amount of each selected indicator constituent which
would be inhaled directly by the target species.
Exposure Scenario
A food chain or food web was constructed for each target species.
A food web is, by definition, a series of food chains connecting
producers and consumers in an ecosystem. Producers are the plants
which make up the base trophic level. Consumers are the representatives
of all other trophic levels including herbivores, carnivores, omnivorcs
and parasites.
The principal mode of constituent transport is via the atmospheric
pathway with deposition onto soil, surface water and vegetation.
Subsequent fate and transport processes result in the final constituent
concentrations in the selected media as determined below.
The total daily uptake (mg/kg/day) of the target species was calculated
by adding the amount of constituent ingested through: (1) consumption
of vegetation, (2) direct ingestion of soil, (3) surface water, (4) fish
tissue, (5) inhalation and (6) ingestion of other target species.
Methodology for Calculation of
Soil Concentrations
Soil concentrations of the selected indicator constituents were
calculated for each receptor location based on total deposition rates
Average
Body Weight
kg
Water
ml/day
Table 1
Estimated Diets for Selected Iferget Species
Soil
g/day
Inhalation
Air
L/day
Arthropods
Earthworms
g/day
Herbaceous
Plants
g/day
Shrubs
Trees
g/day
Prey
g/day
Fish
g/day
Avian
Red-tailed Hawk
Woodcock
Mammalian
Short-tailed Shrew
Woodchuck
White-tail Deer
Cottontail Rabbit
Red Fox
Reptilian
Snapping Turtle
0.750
0.135
0.022
4.200
50.00
1.450
5.200
45.0"
8.1a
3.2°
87,0a
3000a .
210.0°
312.0s
10.00 600.Ob
0.013C
0.002L
0.420
0.145
360°
65d
10°
1362°
5.000C 22454°]
470"
0.520C 1686d
1248.Oe
20.Or
22.0?
252.0h
1750h u
43.5*
78.Oh
60. Or
112.0T
1750"
43.5"
78. Oh
156.On
540.0"
Average of 0.06 ml/g/day water required for metabolic homeostasls
Average of 0.15 ml/g/day water required for metabolic homeostasis
Soil Ingestion equals 1.0 X 10 of body weight
Average of 0.6755 ml/g/hr oxygen required for metabolic homeostasls
Average of 0.260 ml/g/hr oxygen required for metabolic homeostasls
Average of 0.150 g/g/day food required for metabolic homeostasis
Average of 1.00 g/g/day food required for metabolic homeostasls
Average of 0.06 g/g/day food required for metabolic homeostasis
Newell et al., 1987. Niagara River Biota Contamination Project: Fish Flesh Criteria
for Piscivorous Wildlife.
112 RISK ASSESSMENT
-------
specific for each location.
The basic formula used to determine the concentration of the selected
indicator constituents in soil due to aerial deposition is:
Soil concentration (mg/kg) = DR
MD (m) • BD (kg/m )
Where: (!)
OR Deposition Rate
T Accumulation Time of 30 years
K Conversion Constant 1 mg
1000 ug
R 1.0 Representing no loss of constituents due
to physical or chemical means
MD Mixing Depth of 0.50 meters
BD Bulk Density of soil at receptor site 1250 kg/m
saECTED REMEDIAL WASTE CONSTITUENTS
[DIAL
SION!
ENVIRONMENTAL
EMISSIONS DATA
(CALCULATED EMISSION RATES)
. FACTORS
• MODELING DATA
AIR CONCENTRATION
INHALATION CONCENTRATION
FOR TARGET SPECIES
DAJLY DIET OF FOOD/WATER/AIR •
TARGET SPECIES BODY BURDEN
(TOTAL AMOUNT OF INTAKE)
Figure 1
General Pathway
Methodology for Calculation of
Surface Water Concentrations
The surface water concentrations were derived based on the Tier I
analysis as referenced by the U.S. EPA8. Subsequently these
concentrations will serve as the exposure concentrations for the
calculation of constituent uptake by fish through the surface water
pathway and transfer via the food chain into the target species. The
following equation was used to calculate the surface water
concentrations:
Surface Water Concentration (rag/L) ER (9/s) • EC • R
River flow (ftVsj
(2)
t SIMPLIFIED FOOD WEB INTERACTION OF TARGET SPECIES
Figure 2
Food Web
where:
ER = emission rate
EC = emission constant, which represents the
percentage of stack emissions depoited on the
surface water
R = fraction of selected indicator constituents
in the water column
K = conversion constant
1000
mg 1
ft3
g 28.32L
River Flow = The average annual flow of the body of surface water.
Methodology for Calculation of
Plant Tissue Concentration
Generally, there are four main pathways by which a constituent in
the soil can enter a plant. These are:
• Root uptake and subsequent translocation by the transpiration stream
• Vegetative uptake of vapor from the surrounding air
• Uptake by external contamination of shoots by soil and dust, followed
by retention in the cuticle or penetration through it
• Uptake and transport in oil cells which are found in oil containing
plants like carrots and cress
The amount of an organic constituent found in a plant will be the
sum total of each of these transport routes minus metabolic losses. Their
respective importance will depend upon the nature of the organic
constituent, the nature of the soil and the environmental conditions under
which plant exposure occurs. For the purpose of this risk assessment,
both foliar deposition and root uptake are addressed.
Using the soil concentration and total deposition rates derived
previously, the plant tissue concentration can be determined using the
RISK ASSESSMENT 113
-------
following formula:
Plant Tissue Cone, (mg/kg) = Surface Deposition (mg/kg)
+ Root Uptake (mg/kg)
(3)
where:
Plant Tissue Cone. = Concentration of the indicator constituent in
vegetation as the result of foliar deposition
and root uptake.
Surface Deposition = Concentration of the indicator constituent
in vegetation as the result of foliar
deposition.
Root Uptake = Concentration of the indicator constituent
in vegetation as the result of root uptake.
The following equation was used to calculate foliar surface deposition:
-kt,
Cd (ug/kg) = TDR (ug/m?/yr)
y (kg/m2) • k (1/yr)
(4)
Cd = Concentration of the Indicator constituent In vegetation at tn<-
result of foliar deposition.
TDR • Total Deposition Rate
R • Vegetation Interception fraction as derived from Baes et al., 1981.
k - Rate constant for surface degradation processes as calculated from
Baes et al., 1984 (36.1 yr )
1 - Length of the growing season from Baes et al., 1984 (0.51)
f • Fraction of the year the plant Is In the field (1.0).
V « Blonass of temporal/evergreen forest, Whlttaker and Likens, 1973
(36.0).
The following equation was used to calculate root uptake concentrations:
Root Uptake Concentration (mg/kg) = Soil Concentration x RUF x EP
(5)
where:
Soil Concentration = use site-specific concentrations for
selected indicator constituents.
RUF = Root Uptake Factor
EP = Edible Portion of plant, 50% (Heichel and Hankin, 1976)
which accounts for the percentage from root uptake that
is partitioned to the leaves and growing shoots of the
vegetation.
Root Uptake Factors (RUFs) of organic constituents were derived
based on work by Briggs et al.,2 Briggs studied the uptake of organic
constituents from solution by barley shoots and established the following
relationship between the root concentration factor (RCF) and the
octanol/water partition coefficient (Kow) for the organics tested:
RCF = Antilog [0.77 (logKJ - 1.52] + 0.82
(6)
The RUF for each constituent can be determined from the RCF given
the following relationship:
RUF =
RCF
) (F J
Where: Koc = Soil-organic carbon-water partition coefficient
Foe = percent organic carbon content of soil 0.05
RCF = Root Concentration Factor
Koc values for the selected indicator constituents were obtained from
the U.S. EPA, Superfund Public Health Evaluation Manual* Foe
values were obtained from the USDA Soil Survey. The Root Uptake
Factors for selected inorganic indicator constituents are those published
by Baes, et al1.
Methodology for Calculation of
Fish Tissue Concentration
The accumulation of the constituents in fish tissue involving the
processes of bio-concentration and bio-magnification were calculated
using the following formula:
Fish Tissue Concentration (mg/kg) = Surface Water Concentration
(mg/L) • BCF (Ukg) (7)
where:
Surface Water Concentration = Site-specific values
BCF = Bio-concentration Factor of the selected
indicator constituents
Methodology for Calculation of Target Species
Total Daily Exposure
The methodology used in the calculation of the total daily exposure
of each target species follows the methodology set forth by the U.S.
EPA'. Since the methodology in the US. EPA's Manual has been
developed for human exposure, target species-specific factors were
developed to more accurately describe the exposures to individual
species. These factors are presented in Table 1. This methodology
assumes that the daily concentration of the selected indicator constitueni
bio-accumulated in the target species is assumed to equal the daily dose
ingested and follow a linear additive bio-magnification model.
Total exposure of a target species is defined as the summation of
exposure from each individual pathway. Sources of exposure can be
represented mathematically as:
$oll,,p.<«9/day) • Soli Cone. (>g/tg) • Soil Infested (kg/My) • W • OF (g)
Soil
e«p.
> Constituent exposure of uw target species as the result
Of Ingest Ion of Soil.
Soil Cone. • Soil concentration.
Soil Ingested * The target species specific rate of Ingestion of soil as
defined 1n Table I.
GAF . A Cut Absorption Factor of 100 percent was not used for
this pathway as soil Is not a nomal dietary component.
Constituent associations with foodstuffs and water, are
for the aost part reversible; whereas, particular binding
of constituents to soil Is for the cost part Irreversible.
Therefore, for this eiposure scenario a GAf of 101 for
Inorganics (Osweller et al., 1985) and 301 for organ I Cl
(Oevlto et al., 1988) Is used.
OF • Digestion factor 5SS (0.55) which represents the average
of rualnanl (651) an) non-rva1nant (aU) target species
(Maynard et al.. 1971).
"ater Cg/day) • Surface water cone. («g/l) • water Ingested (I/day) • GAf • If
xater • Constituent eiposure of the target species as the
result of Ingestion of water.
Surface water Cone. • water concentration.
Uiter Ingested • The Target species specific rate of Ingestion of water.
GAF • Gut Absorption factor 1001 (1.0).
Of • Digestion Factor 1001 (1.0).
vegetal lon^.f-g/day) • Plant Tissue Cone, fag/kg) Amount Ingested (kg/day) • GAF • V
on
Vegetation.,.,. • Constituent eiposurt of the target species as the
result of Ingestion of vegetation.
Plant Tliiue Cone. • Plant concentration.
Ajnunt Ingested • The target species specific rate of Ingestton of
"atloi
GAF
vegetation.
• Gut Absorption Factor IOOS (1.0).
OF • Digestion Factor SSI (O.SS) which rtpnunts the
average of nulnant (CSS) and non-runlnant (4B)
target species (Majmard it al., 1979).
Fl»l>,«p. (-g/day) . Flth Tissue Cone, (ig/kg) • FUh Ingested (kg/day) • 6AF • V
114 RISK ASSESSMENT
-------
•tare:
F1sh-x_. - Constituent exposure of the target species is the
v result of Ingestlon of fish.
Fish Tissue Cone." F1sh concentration.
Fish Ingested • The target species specific rate of Ingestlon of fish
as defined in Section E3.0.
GAF - Gut Absorption Factor 100* (1.0).
DF • Digestion Factor 55* (O.SS) which represents the
average of runlnant (65*) and non-ruminant (45X)
target species (Haynard et al., 1979).
*ltexp- ("s/d"y) " P'S- T'SSUe ConC' (>s/l<9) " '•S- InS"ted (kg/day) • GAF • DF (12)
Heat.XD. * Constituent exposure of the target species as the
result of Ingestlon of Heat.
P.S. Tissue Cone.- Prey species tissue concentration
P.S. Ingested « The target species specific rate of Ingestlon of neat
as defined 1n Table 1.
GAF - Gut Absorption Factor 10M (1.0).
DF • Digestion Factor 551 (O.SS) which represents the
average of ralnant (65*) and non-rum1nant (45*)
target species (Haynard et al., 1979).
A1rap. («g/day) - A1r Cone, (ug/n3) • Air Inhaled (L/day) • LAF • K (m3/L)(ng/ug) (13)
!Xp
where:
A1r.KD. • Constituent exposure of the target species as
p result of Inhalation.
A1r Cone. = Air Concentration.
Air Inhaled = The target species specific rate of Inhalation as
defined In Table 1.
LAF - Lung Absorption Factor 100* (1.0).
K = Conversion constants of m^/1000L and mg/100 ug.
Therefore, total dally oral exposure can be defined mathematically as:
TOEor>, (ng/kg/day) - Soilexp. (mg/kg/day) + "aterexp. (mg/kg/day) * D1etexp. (mg/kg/day)
(14)
Risk =
where:
(COI^ x Carcinogenic Potency Factory)
(16)
CDI.J = Chronic Daily Intake for the i Constituent
CPF Carcinogenic Potency Factor from Superfund Exposure
Assessment Manual (USEPA, 1988).
Carcinogenic Risk of Constituents
Eight carcinogenic indicator constituents are associated with stack
emissions. These constituents were assessed to determine daily exposure
by either inhalation or ingestion. Probabilities of additional carcinogenic
risk of the selected indicator constituents were calculated for seven
receptor locations via the following pathways:
• Ingestion Pathway
• Inhalation Pathway
From the data derived for inhalation and ingestion pathways, a total
probability index can be calculated per constituent. This index is the
summation of probability indices for inhalation and ingestion exposures
at each receptor location.
Risk Calculation of Non-carcinogenic Effects
To address the non-carcinogenic effects of the selected indicator
constituents, a hazard index approach has been adopted based on U.S.
EPA Guidelines for Health Risk Assessment of Chemical Mixtures (U.S.
EPA, 1986). The hazard index for a specific constituent is defined as
the ratio of daily intake for that constituent to the constituent specific
RfD. The constituent specific hazard indices were calculated using the
following formula:
HI = GDI • RfD
(17)
TDEQraj = Total dally exposure of the target species as the result
of all oral-associated pathways
So11pxn. = Constituent exposure of target species as the result of
p soil Ingestlon.
Haterexp. = Constituent exposure of target species as the result of
water Ingestlon.
D1etex_. «= Constituent exposure of target species as the result of
vegetation, fish, and meat Ingestlon.
While the Total Dal ly Exposure based on the Inhalation exposure is equal to
the Inhalation pathway alone.
™
lnhalat1on
A1rexp. («g/kg/day)
(15)
^Inhalation " Total Dally Exposure of the target species as the
result of all air-associated pathways.
A1rexD. * Constituent exposure of the target species as the
result of the inhalation.
RISK CHARACTERIZATION
This section of the paper defines the risk characterization for terrestrial
species based on methodologies developed for a human health risk
assessment. Species-specific factors were developed to account for
interspecies differences in uptake, absorption, excretion, etc. and adapt
the models to assess risk to local target species.
A comparison was made between projected intakes and available
reference levels (RFDs) for non-carcinogens and between calculated
risks and target risks for potential carcinogens. For non-carcinogens,
direct comparison is made between estimated intakes and available
reference levels, whereas for carcinogens, estimated intakes are
combined with upper bound carcinogenic potency factors to calculate
risk.
The carcinogenic risk estimate for multiple constituents is represented
by the following equation:
where:
HI.
CDI.
= individual hazard index for exposure to constituent i at
location p
= daily dose for constituent i at location p
= acceptable daily intake, or reference dose (RfD), for
chronic exposure to constituent i
Any single constituent with an exposure level greater than the
reference level will cause the hazard index to exceed unity (1.0), and
when the index exceeds unity, there may be concern for a potential health
risk. For multiple constituent exposures, the hazard index may exceed
unity even if no single constituent exceeds its acceptable level. It is
therefore emphasized that the hazard index is not a mathematical
prediction of incidence or severity of effects.
DISCUSSION
A terrestrial food-chain assessment of a project such as a HWI is
ultimately an integrated evaluation of historical, chemical, analytical,
environmental, demographic and lexicological data that are as
site-specific as possible. Ultimately the precision of an ecological risk
assessment is limited by the size and quality of the data base. This
limitation can be overcome by defining a range of extremes. Specific
areas of uncertainty include:
• Receptor species
• Emissions data bases
• Air modeling
• Fate and transport estimates
• Exposure estimates
• Toxicological data and risk characterization
• Complex interactions of uncertainty elements
To minimize the effect of these uncertainties in the evaluation, each
step should be biased toward conservative estimations. Since each step
RISK ASSESSMENT 115
-------
builds on the previous one, this biased approach should more than
compensate for adjustments made to the human health-based criteria.
REFERENCES
1. Baes, C.F., Sharp, R., Sjoreen, A., and Shor, R., "A Review and Analysis
of Parameters for Assessing Transport of Environmentally Released
Radionuclides through Agriculture," ORNL #5786, 1984.
2. Briggs, G., et al., "Relationships between Lipophilicity and Root Uptake,
and Translocation of Nonionized Chemicals by Barley." Pesticide Science.
13:495, 1982.
3. DeVito, M., Umbreit, T.H. and Gallo, M.Am., "Bio-availability as a Factor
in Human Health Risk Assessment of a Newark, NJ TCDD Contamination
Site." In: Proceedings of the Ninth Annual Meeting of the Society of
Environmental Toxicology and Chemistry. Arlington. VA, 1988.
4, Heichel, G.H. and Hankin, L., In: Biological Monitoring of Heavy Metal
Pollution, Land and Air, Martin and Coughtrcy eds. Applied Science
Publishers Ltd. Essex, England, 1976.
5. Martin, M.H., and Coughtrey, P.J., Biological Monitoring of Heavy Metal
Pollution: Land and Air. Applied Science Publishers LTD. Essex, England
1982.
6. Maynard. L.A., Loosli. J . Hintz, H. and Warner, R., Animal Nutrition.
7th Edition, McGraw-Hill Book Company. New \brk, NY, 1999.
7. Osweiler, G.D., Carson, T, Buck, W. and Van Gelder, G., Clinical and
Diagnostic Veterinary Toxicology. 3rd ed. Kendall/Hum Publishing Co.
Dubuque, LA, 1985.
8 US. EPA, "Superrund Public Health Evaluation Manual" EPA 540/1-86/060.
Office of Emergency and Remedial Response. Washington, DC, 1986.
9. Whitlaker, R.H. and Likens. G.. Carbon in (he biota. In G.M. Vfoodwell
and E.V. Pecan (eds.) pg. 281-300, 1973.
116 RISK ASSESSMENT
-------
Contaminant Fate and Effects in Surface Water and
Groundwater for Site Closure of a
Remediated Dioxin (TCDD) Site
David J. Thorne
Arthur S. Rood
Idaho National Engineering Laboratory
EG&G Idaho, Inc.
Idaho Falls, Idaho
ABSTRACT
A risk analysis was performed for the surface water and groundwater
contaminant transport pathways for a remediated dioxin site. The site
had been used to store barrels of herbicide containing
2,3,7,8-tetrachlorodibenzopdioxin (TCDD). Leaks in the containers and
subsequent TCDD migration had contaminated the soil at the site.
The site had been remediated by removing and incinerating the soil
which was then backfilled into the excavated areas of the site. The hazard
associated with the remaining TCDD levels was determined by modeling
the TCDD leachate from the soil and subsequent transport in the
groundwater to off-site environs. A first order leaching model was used
with semi-analytical solutions to the groundwater contaminant transport
equation for a porous medium. First order kinetic processes were used
to govern the contaminant mass in a recharged surface water body.
Maximum contaminant concentrations were bounded by using an
instantaneous release of the entire TCDD contaminant mass.
TCDD was identified as a probable human carcinogen because of
its classification as a B2 carcinogen through ingestion. Carcinogenic
risks were determined for potable water obtained from the aquifer at
a site boundary well and for the consumption of fish obtained from
the surface water body adjacent to the site.
INTRODUCTION
The site evaluated in this study had been used to store herbicides
containing 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). The soil in the
herbicide storage area had become contaminated as a result of leaks
in the storage containers. This soil was incinerated to reduce the TCDD
contamination levels. After incineration, some residual TCDD remained
in the soil that was backfilled into the excavated areas of the site.
The health hazards associated with the residual soil contamination
were evaluated using simple analytical water and groundwater surface
water models and conservative assumptions. The use of simplified
models frequently is adequate for regulatory purposes if they are
conservative. In addition, a sensitivity analysis was performed in order
to show the bounding, worst case scenario.
The purpose of this paper is to present the methodologies used in
this study and to demonstrate the effective use of simplified models
when the available data do not warrant the use of more sophisticated
models. This methodology is not limited to TCDD-contaminated sites
and may be applied to other contaminants.
CONTAMINANT MIGRATION PATHWAYS
The contaminant migration pathways that were considered in the risk
analysis included leaching of TCDD from the soil to the groundwater
and discharge of TCDD-contaminated groundwater into a surface water
body that was intercepting the groundwater aquifer.
EXPOSURE PATHWAYS
The exposure assessment considered two scenarios: (1) ingestion of
drinking water obtained directly from the aquifer and (2) consumption
of fish obtained from the surface water body. Individuals may be exposed
to the fish which have bioconcentrated the TCDD from the contaminated
water.
The exposure scenarios considered a maximally exposed individual
residing at the site boundary. The hypothetical individual was assumed
to obtain all drinking water from a well located at the site boundary
along the contaminant plume centerline. In addition, fish obtained from
the nearby surface water body were considered to constitute a major
portion of the individual's diet. The surface water body was also assumed
to lie at the site boundary and intersect the groundwater aquifer. U.S.
EPA1'2 values were used for drinking water and fish consumption rates.
CONTAMINANT RELEASE SCENARIO
The release of the TCDD to the groundwater was modeled as two
distinct scenarios: (1) an instantaneous release of the entire residual
TCDD mass in the soil and (2) a time-variant release. The time-variant
release scenario used a first-order kinetic model to predict the TCDD
release from the soil to the groundwater. This procedure allowed the
determination of an upper bound estimate, using the instantaneous
release scenario, and a best estimate, using a time-variant release
scenario. This approach estimated the risk and the uncertainty of the
risk analysis.
CONTAMINANT RELEASE MODEL
The TCDD release model was evaluated using the conceptual model
illustrated in Figure 1.
amount of TCDD In the soil (g)
X d = decay rate constant of TCDD In soil (s )
X L= leaching rate constant of TCDD from soil
to the groundwater (s"1)
Gw= groundwater
Figure 1
TCDD Release Model
The rate of change of TCDD in the soil with respect to time was
described by a first-order loss process as follows:
FATE 117
-------
-(XL + \) Q,
= amount of TCDD in the soil (g)
= decay rate constant of TCDD in soil (sec'1)
= leaching rate constant of TCDD from soil to the
ground water (sec"')
t = time (sec)
The solution to the above equation is:
0,0) = Q* (eA+\)«)
where:
Q^ = amount of TCDD at t = 0
The Release Rate (R, g/s) was determined by:
R = Q,(t) \
Substitution of the solution for Qd(t) yields:
R = [Q
(2)
The total quantity of TCDD released (T, g) was determined
integrating the release rate from zero to infinity:
(4)
by
(5)
Therefore, the total TCDD released (T, g) was given by:
T Q-X.
(\ + \)
The total TCDD mass at time t
Q^= A *D*p
where:
0 in the soil was given by:
A = the area of contamination (mj)
D = depth of contamination (m)
p = bulk density (g/mj)
The leach rate constant used in this study was designed to in evaluate
low level radioactive waste repositories3. Although this work was done
for radionuclides, these same parameters are defined for organic and
non-organic contaminants and may be used to describe their transport
through the soil. The leach rate constant (XL) was given by:
(8)
(XWT 8) + (p K.XWT)
where:
P = percolation rate (cm/sec)
8 = volumetric water content (cm3 of H3O/cm3 of waste)
p = soil density (g/cm3)
XWT = waste thickness (cm)
The distribution coefficient may be related to organic adsorption
phenomena by the organic carbon partitioning coefficient (KK) which
is defined as:
mg of chemical adsorbed/kg of organic carbon ,Q.
oc " - -'— - - -.. _ \y)
K = mg of chemical dissolved/liter of solution
The distribution coefficient was defined in terms of the K by:
where:
f^ = the fraction of organic carbon in the soil
The decay rate constant of TCDD in the soil is given by:
Ln 2
(11)
'1/2
where:
T,: is the half-life of TCDD in the soil (sec) (12)
The volumetric water content 6 was determined using the following
equation:
The retardation factor (R^) was determined as follows (4):
n
P
n.
(13)
where:
ne = effective porosity
n = total porosity
p = soil density (g/cms)
Kd= distribution coefficient (mL/g)
GROUNDWATER AND SURFACE WATER MODELS
A simple groundwater model was used which assumes a constant,
unidirectional flow field in a homogeneous porous medium of infinite
lateral extent and finite thickness. A more sophisticated model was not
warranted due to the lack of data characterizing the aquifer. However,
simpler models frequently give results adequate for regulatory purposes
if they are conservative and provide a worst case scenario. An analytical
solution to the groundwater transport equation was used to calculate
the contaminant concentrations at a well located at the nearest site
boundary. All release of TCDD from the soil was conservatively
assumed to enter the aquifer without any interference or time delays
in the unsaturated zone (Fig. 2).
Well
Figure 2
Cross Section of the TCDD Site
The equation for contaminant transport in groundwater is':
dC n
+ VC - V DVC - XC
at R,
C(0,x,y,z) - 0
aC(t,x,y,z-0,b)
K = K * f
(10)
118 FATE
-------
where:
C
D
b =
V =
t =
If the
and the
solution
where:
TCDD concentration (g/L)
dispersion coefficients (m2/sec) = a/i
dispersivity in x and y directions (m)
retardation factor
distance in the x, y and z directions from the point of
origin (m)
decay constant (sec~')
pore velocity (m/sec)
thickness of the aquifer (m)
the del operator
time (sec)
source is represented by an area with length L and width w
contaminant is released instantaneously at t = 0, then the
to the above equation is4:
C(t,x,y,z) X(x,t) Y(y,t) Z(z,t)
(15)
1 x + L/2 - A x - L/2 - A
X(x,t) - — [erf ———^- - erf „ ] exp (-Xdt)
where:
Qd(t.) = the TCDD mass released during a pulse i (g)
t ' = time of pulse release, i (sec)
k = the number of pulses
Release rates were calculated on a yearly basis. The amount of
contaminant in each pulse was given by:
(20)
R * 6
where:
R = release rate
5 = the incremental time step
The incremental time step was determined by the dimensionless
standard deviation of a pulse given as5:
2 8
a = ( + ) 0.5
Pe Pe
(21)
where:
Pe = the peclet number
The peclet number represents the ratio of advection to dispersion and
is given by:
1 y + w/2 -y t w/2
V(y,t) - [erf — - erf
2w
(«Oyt/Rd)'
(17)
Pe =
(22)
Z(z,t)
(18)
where:
A
Qd
C
Dy =
«L =
<*T =
L =
w =
nt
erf
b
/it/Rd
TCDD instantaneous release mass (g)
dispersion coefficient in the x direction (m2/sec)
dispersion coefficient in the x direction (mVsec)
D, = aL* „.
dispersion coefficient in the y direction (m2sec)
longitudinal dispersivity (m)
transverse dispersivity (m)
the length of the source (m)
the width of the source (m)
retardation factor
decay constant (sec"1)
pore velocity (m/sec)
effective porosity
error function
the thickness of the aquifer (m)
time after release (sec)
This solution assumes complete mixing at the point where the
concentration is calculated. The solution was modified for a continuous
time-variant release by summing over a series of pulse releases. For
a time- variant release, Qd is a function of time described by Qd(t).
When the pulse spacing is kept small relative to the standard deviation
of the pulse at the receptor, a continuous time-variant release can be
simulated by5:
where:
= the longitudinal distance to receptor location
= longitudinal dispersivity
The incremental time step (6) was determined by:
6 = a * GWT
where: GWT = groundwater travel time =(x//i) R
(23)
The contaminant flux entering a surface water body which intersects
the aquifer for the conditions expressed for the groundwater model
described previously is given by:
F(x,t)
where:
Rd
D»
Rd
(4Dxt/Rd)
1/z
(*Dxt/Rd)1/2 [erf(z,) erf(z2)]
(24)
exp(-z?2)]] exp(-Xdt)
(25)
and
C(t,x,y.z) - 2
X(x,t-t,) Y(y,t-t() Z(z,t-t,) (19)
(40xt/Rd)
1/2
(26)
FATE 119
-------
where :
F(x,t) = TCDD flux at distance x and time t (g/sec)
Qd = TCDD impulse release mass (g)
D^ = di spurs ion coefficient in the x direction (mVsec)
D = dispersion coefficient in the y direction (nvVsec)
L = the length of the source (m)
R, = retardation factor
a
Xd = decay constant (sec)
H = pore velocity (m/sec)
erf = error function
t = time after release (sec)
The corresponding flux for the time-variant release scenario is given
by':
F(x,t) -
F(x,t-t,) Qd(t,)
(J7,
where:
F(x,t) = TCDD flux at distance x and time t (g/sec)
Q/t) = the TCDD mass released during a pulse i (g)
t( = time of pulse release, i (sec)
k = the number of pulses
The conceptual model illustrated in Figure 3 was used to model the
contaminant mass in the surface water body.
SWB = Surface water body
R(t) - discharge rate of TCDD to SWB (mg/s)
Q - amount of TCDD in the SWB (mg)
X1 - loss rate constant due to water exchange (s'1)
X j - TCDD surface water decay rate constant (8~1)
Figure 3
Conceptual Model for the Surface Water Body
The conceptual model may be represented mathematically by the
following differential equation:
dQ
dt
where:
= R(t) -(X, +\,) Q
(28)
R(t) = discharge rate of TCDD to the surface water body
(mg/sec)
Q = amount of TCDD in the surface water body (mg)
X, = loss rate constant due to water exchange (sec1)
X2 = TCDD surface water decay rate constant (sec '
t = time (sec)
The solution to the above equation with the conservative assumption
that R(t) was a constant (R) and was equal to the maximum flux to
the surface water body (R = max F(x,t)):
R
Q(t)
(A,
(29)
Ii was assumed that the mass of TCDD in the surface water body
reached steady state, thus the above equation simplified to:
Q =
(30)
where.
R = maximum discharge rate of TCDD to the surface water
body (mg/scc)
The resulting steidy-itite concentration of TCDO In the surface viler
body w»s given by
R
! + xz)
(31)
where:
C = steady state concentration of TCDD (mg/L)
V = volume of the surface water body (L)
The water level in the surface water body was assumed to remain
constant, thus the loss rate constant was given tn:
X. =
_ F
(32)
where:
X( = loss rate constant to due to water exchange (sec1)
F = flow rate out of the surface water body (m'/sec)
V = volume of the surface water body (m')
CARCINOGENIC RISK CALCULATION
TCDD is classified by the U.S. EPA" as a B2 carcinogen through
ingestion, which identifies TCDD as a probable human carcinogen.
The potentiul carcinogenic risk.s from the consumption of potable water
and fish consumption were determined using chronic daily intake
equations and carcinogenic potency factors according to guidance given
by the U.S. HPA'. Potential carcinogenic risks were determined for
concentrations determined in the surface water and groundwater at the
receptor location for both the instantaneous and time-variant release
scenarios.
CONCLUSION
In performing risk assessments, the analyst often is faced with a lack
of the site-specific data needed to define the hydrologic conditions of
the site. In this study, the lack of site data required the use of simplified
models and conservative assumptions. The use of simplified models
and conservative assumptions can provide adequate results for regulatory
purposes. The use of both instantaneous and time-variant release
scenarios allows the analyst to present a simplified quantitative
assessment of the expected uncertainty in the analysis. The models used
in this study are not limited to TCDD- contaminated sites and may be
applied at other hazardous waste sites.
120 FATE
-------
ACKNOWLEDGEMENT
Work performed under the auspices of the U.S. Department of Energy,
DOE Contract No. DE-AC07-76ID01570.
REFERENCES
1. U.S. EPA, Water Quality Criteria Documents, Fed. Reg., 45, No. 231, Nov.
28, 1980.
2. U.S. EPA, Superfund Public Health Evaluation Manual, U.S. EPA Rept. No.
EPA/540/1-86/060, U.S. EPA Office of Remedial Response, Washington, DC,
Oct. 1986.
3. Rodgers and Associates Engineering Corporation, Prototype Safety Analysis
Report Below-Ground Vault Low-Level Radioactive Waste Disposal Facility,
RAE Rept. No. RAE-8716-5, Prepared for EG&G Idaho, Inc., by: Rodgers
and Associates Engineering Corporation, Salt Lake City, UT, May, 1988.
4. Codell R.B. and Duguid J.D., Transport of Radionuclides in Groundwater,
in Radiological Assessment: A Textbook on Environmental Dose Analysis,
ed. I.E. Till and H.R. Meyer, pp. (4) 1-53, NRC Rept. No. NUREG/CR-3332,
Washington, D.C., Sept. 1983.
5. Walton, J.C., Bensky M.S., Mease, M.E., and Bander T.J., Sensitivity of
Radionuclide Transport in Groundwater to Source and Site Characteristics,
Proc. of the Waste Management 88 Conference, pp. 269-276, University of
Arizona, Tuscon, AZ 1988.
6. U.S. EPA, Superfund Exposure Assessment Manual, U.S. EPA Rept. No.
EPA/540/1-88/001, U.S. EPA Office of Remedial Response, Washington, DC,
April, 1988.
FATE 121
-------
Groundwater Source Separation Using Chlorinated Organic
Compound Degradation Series and Inorganic Indicators
Martin J. Hamper
James A. Hill
Warzyn Engineering, Inc.
Chicago, Illinois
ABSTRACT
The investigation was performed to determine if the volatile organic
compounds (VOCs) detected in groundwater adjacent to Winnebago
Reclamation Landfill (WRL) are from leachate releases from the facility
or from a VOC plume reported to be emanating from an upgradient
NPL site, known as Acme Solvents Reclaiming, Inc. (Acme). The
approach undertaken was to first compare the leachate chemistry and
the groundwater chemistry to identify which wells have been affected
by leachate from the landfill. The groundwater chemistry of VOC-
impacted wells was compared with wells impacted by landfill leachate,
indicating that a distinct leachate plume is present and different than
the V(X: plume.
The leachate plume from the landfill is well defined by chloride ion
content and begins in the center of the landfill and extends to just past
the downgradiem edge of the landfill. There are chlorinated ethenes
both in and outside of the leachate plume, indicating that the leachate
plume is mixing in a pre-existing VOC plume. The presence of VOCs
at the east end of the landfill is not attributed to the presence of landfill
leachate since they are present hydraulically upgradient of the landfill,
and the chloride concentrations at that location are not increased as
would be expected if leachate was the source. Groundwater chemistry
which does not show the presence of chlorides in elevated concentra-
tions is not affected by landfill leachate.
The presence of VOCs at the southeast margin of the landfill gives
the appearance of a bimodal distribution of VOCs in the groundwater
in the area. This bimodal distribution may be due to one or more of
the following causes:
• The wells between Acme and WRL do not intersect a flow path
through the fractured dolomite that is responsible for the transport
of VOCs from Acme.
• The appearance of a bimodal distribution could be the result of inter-
mittent and spatially variable recharge.
• Biodegradation may play an important role in explaining the
appearance of the bimodal distribution of VOCs. Biodegradation could
increase the concentration of less chlorinated species which could
give the appearance of a bimodal distribution of VOCs.
Introduction
The primary focus of the investigation performed was to determine
if the volatile organic compounds (VOCs) detected in the groundwater
were the result of a release of leachate from Winnebago Reclamation
Landfill (WRL) or from a VOC plume emanating from an upgradient
NPL site, known as Acme Solvents Reclaiming, Inc. (Acme). The
differentiation of sources of released materials is necessary for deter-
mining responsibility for any required cleanup efforts. Warzyn's
approach was to first compare the leachate chemistry with the ground-
water chemistry and identify which wells have been impacted by leachate
from the landfill. Secondly, the groundwater chemistry of VOC affected
wells was compared with wells impacted by landfill leachate.
A previous study of these two sites as part of the NPL listing process
noted a bimodal distribution of organic constituents in the groundwater
exists in this area. This analysis was interpreted to support the presence
of separate groundwater plumes emanating each site1 Both Acme and
WRL were placed on the NPL and are being studied by consultants
to the PRP groups. WRL was placed on the NPL due to the detection
of arsenic and cadmium in a monitoring well adjacent to the landfill.
The key organic groundwater chemistry difference noted in previous
studies was that the relative amount of trans-l,2-dichloroethene appeared
to be greater in the groundwater under the landfill than under the Acme
site'. Wood et al. stated that vinyl chloride, 1,1-dichloroethene, cis-
and trans-l,2-dichloroethene, 1,1-dichloroethane and chloroetnane are
either not commercially produced or are not in wide use across (he
whole country as are parent compounds such as tetrachloroethene, trich-
loroethene, 1,1.1-trichloroethane and methylene chloride3. It is possi-
ble that the appearance of a bimodal distribution of dichloroethene may
be the result of biodegradation of the VOC plume from the upgradient
Acme site. Cline and Viste reported that U.S. EPA Methods 601 and
624 typically used to analyze water samples for VOCs do not differen-
tiate between the cis- and trans-isomers of 1,2-dichloroethene, although
the data are reported as the trans-isomer since it is the priority
pollutant-' The cis-isomer is predominantly produced as the result bio-
degradation of trichloroethene1.
SETTING
The WRL is located approximately 5 mi south of Rockford. Illinois,
in the Rock River Hill Country of the Till Plains Section of the Central
Lowland Province of Illinois'4. The WRL occupies approximately
60 ac on a topographic high between Killbuck Creek to the west and
unnamed intermittent streams to the north and south. Killbuck Creek,
a perennial stream, flows within 250 ft of the western WRL boundary
and merges with the Kishwaukee River approximately 2 mi to the north.
There are no other surface water bodies within 1 mi mile of the WRL.
This municipal solid waste landfill has been licensed by the State
of Illinois since 1972 and is nearing capacity. The facility has a
bituminous liner and a leachate collection system. The leachate is dis-
posed of off-site. A system of leachate/gas extraction wells is used to
remove landfill gas and leachate. The western half of the landfill addi-
tionally collects leachate through a perforated pipe leachate collection
system on the top of the liner, which gravity drains to central collec-
tion manholes.
Wastes accepted at the landfill are composed primarily of municipal
refuse and sewage sludge. Prior to the startup of the gas collection sys-
122 FATE
-------
terns in 1984, the landfill accepted wet sewage sludge (vacuum filter
cake at approximately 20 to 23% solids). Currently, the landfill gas
is used to power sludge dryers, which dry the sewage sludge prior to
disposal. A very limited quantity of special wastes were disposed of
at the facility prior to December, 1985. Special wastes accepted at the
facility were accepted under approved permits issued by the Illinois
Environmental Protection Agency (IEPA).
Approximately 1000 ft east of the WRL on an approximately 20-ac
parcel is the Acme site. The Acme site was used for disposal of waste
generated by Acme's solvent reprocessing facilities in Rockford, Illinois
from approximately 1960 to 1973. The Acme site has been on the NPL
since 1983. The materials disposed of at Acme are generally un-
documented, but are known to have included solvent still bottom sludges,
non-recoverable solvents, paints and oils. The waste materials were
reported to have been transported to the site in drums which were either
emptied into unlined disposal lagoons or stockpiled. The IEPA indicates
four lagoons were actively used for the disposal of waste materials on-
site. IEPA reported that between 10,000 and 15,000 drums may have
been present on the site when it closed. The total quantity of waste
disposed of at the site during its operation is unknown. IEPA inspec-
tions in late 1972 and early 1973 indicate the waste materials in the Acme
ponds were not removed, but were covered with soil borrowed from
other portions of the site. It is also reported that an unknown number
of drums stored on-site were crushed and buried, rather than
removed1. Clean-up of the Acme site began in August 1986 and con-
sisted of removal of buried drums and contaminated soils.
SITE HYDROGEOLOGY
Unconsolidated Materials
The surficial unconsolidated materials of the area are predominantly
glacial drift deposits. The thickness of the unconsolidated materials
ranges from 8 ft to greater than 70 ft. The body of the deposits thickens
from east to west, forming a relatively thin mantle over the bedrock
upland in the east, and filling the deep bedrock valley to the west. This
transition begins beneath the eastern margin of the landfill where the
bedrock surface slopes downward forming the preglacial bedrock valley
wall. Based on regional information, the thickness of unconsolidated
sediments is expected to be approximately 100 ft under Killbuck Creek
near the WRL.
The soils beneath and east of the site are poorly-sorted sand and gravel
glacial ice-contact deposits. Portions of the sand and gravel were some-
times recognized as weathered bedrock. West of the site in the Kill-
buck Creek Valley, and to the north of the site, the sediments are sand
and gravel outwash deposits. The soil types are predominantly fine to
coarse sands with occasional fine to coarse gravel zones 11 ft to 40 ft
thick. The surficial deposits south of the site are predominantly a silty
clay till up to 24 ft thick.
Bedrock
The unconsolidated sediments in the region are underlain uncon-
formably by the dominantly carbonate rocks of the Galena-Platteville
Groups (Ordovician System). The bedrock surface elevation is highly
variable due to paleo-erosional features. Regional information indicates
the thickness of these groups is expected to range from 250 ft in the
bedrock upland east of the WRL to 100 ft in the adjacent bedrock
valley5. The bedrock near the WRL is composed of dolomite, with
chert layers or nodules commonly noted throughout the dolomite. Shale
partings and coatings were noted only below 695 ft MSL. The dolo-
mite generally is fractured throughout the total depth sampled. The frac-
tures are dominantly horizontal bedding planes, frequently cross-cut
by high angle or vertical fractures. Vugs (void spaces) are consistently
found throughout the dolomite, with their presence ranging from slightly
vuggy to very vuggy. Cavernous zones were not noted. The Rock Quality
Designator (RQD) of dolomite core samples ranged widely from zero
to 100%, averaging 52.5%, with a standard deviation of 28.9% These
data provide an indication of the variably fractured nature of dolomite.
An up to 27 ft thick zone of highly fractured, soft dolomite was en-
countered in the near surface bedrock during exploratory drilling in
the vicinity of the northern intermittent creek on the Acme site, where
the RQD ranged from "too soft to core" to 28%. Highly fractured zones
(low RQD) also were found between rocks containing few fractures (high
RQD), indicating rock competence did not generally improve with
depth.
Groundwater
The uppermost aquifer encountered in the vicinity of the WRL
changes in character due to the abrupt sloping of the bedrock surface
beneath the site. East of the WRL, and below its eastern third, the water
table occurs within the dolomite bedrock. From this boundary to the
west, the water table is present in unconsolidated materials. Regard-
less of the type of matrix material, the uppermost saturated unit in the
immediate vicinity of the WRL is under water table conditions. The
water table also occurs in the silty, clayey till to the south of the site.
The sand and gravel and/or dolomite aquifer beneath the till appears
WINNEBAGO
RECLAMATION
LANDFILL
PREDOMINANTLY SANDY SOILS
PREDOMINANTLY CLAYEY SOILS
Figure 1
Water Table Map for May, 1989
Cross-Section Locations Also Noted
NORTH
0 300 600
1200
SCALE (FEET)
FATE 123
-------
to be under semi-confined conditions.
Ground water generally flows from the uplands east of the WRL to
the Killbuck Creek valley, but precise flow configurations within the
fractured dolomite are likely to be more complex in detail (Fig. 1). East
of the WRL, the water table is a subdued expression of the bedrock
topography; the water table slopes outward to the west, northwest and
to southwest from a generally east-west trending groundwater "high"
in the vicinity of the northeast-southwest trending dolomite bedrock
ridge (Fig. 1). The water table in the unconsolidated sediments gently
slopes towards the Killbuck Creek floodplain to the south and west of
the landfill.
A groundwater mound has been observed seasonally in the vicinity
of the northern intermittent creek, east of the WRL. It is thought the
mounding is due to higher recharge rates localized in this area. As dis-
cussed earlier, sandy sediments are underlain by highly weathered dolo-
mite bedrock perhaps enhancing the potential for recharge there.
RESULTS AND DISCUSSION
Table 1 contains a summary of the leachate inorganic chemistry data
developed in this study. The leachate is dominated by high chloride
content as well as high sodium and potassium content. The leachate
also has high alkalinity and specific conductance. Figure 2 is a trilinear
plot of the major cations: calcium, magnesium and sodium plus
potassium for groundwater and leachate samples as percent milliequiva-
lents per liter. The data plot generally along a line from the endpoints
Leachate Inorganic Analytical
Chloride
Alkalinity
Specific
Condition
(umhos/cm)
PH
Sodium
Potassium
Calcium
Magnesium
17,300
11,200
3,740
9,090
3,630
8,520
>50,000 27,100 26,200
7.27
10,200
1,750
241
812
7.54
1,620
1,220
40.3
136
7.75
1,440
1,300
37
70.8
2,720
7,860
24,200
7.66
1,090
710
29.9
57.1
2,490
6,060
19,900
7.54
968
608
93.1
110
50
CALCIUM (7. meg/I)
Figure 2
Trilinear Plot of Groundwater and Landfill Leachate Data
Showing Trend From Leachate to Leachale-Affected Wells
To Unaffected Wells
of leachate samples and upgradient water samples, indicating that
leachate samples and groundwater samples can be discriminated on this
basis.
The plot of chloride versus sodium plus potassium exhibits a strong
linear relationship (R-squared = 0.998) indicating that chloride, too,
can be used to discriminate between leachate and groundwater samples
(Fig. 3). Chloride is generally considered to be non-reactive in ground-
water systems7' and so is very useful as a groundwater tracer. Alka-
linity often is useful in discriminating between leachate and groundwater,
but a plot of log alkalinity versus sodium plus potassium shows a less
strong positive linear relationship (R-squared = 0.82), indicating the
potential for sources of alkalinity other than leachate (Fig. 4).
t-lw,,.a . O.tl«
Figure 3
Plot of Chloride (mg/L) Versus Sodium Plus Potassium (mg/L) for
Leachate Samples' and Groundwater Samples2
Note Strong Positive Correlation
Figure 4
Plot of Log Alkalinity (mg/L) Versus Sodium Plus Potassium (% meg/1)
for Leachate Samples' and Groundwater Samples2
Note Variation in Groundwater Samples
Chloride concentrations are contoured on cross-sections along tran-
sect A-A' for two separate sampling events (Figs. 5 and 6). Both plots
depict a chloride plume originating from the landfill and extending just
past the downgradient edge of the landfill. Downgradient movement
is evident from the increased chloride concentration in the deeper well
in the distal well nest between the two monitoring periods. It is also
evident that the deeper well (B15P) in the well nest where the plume
appears to originate is unaffected by leachate since it has low chloride
concentrations consistent with upgradient well concentrations.
124 FATE
-------
LANDFILL-
ACME
A G116/G116A P1/MW106
•740
•700
B15/B15P
B15R
B11/B11A
B10 G108 B16/B16A A'
DOLOMITE
-660
Figure 5
Cross-sectional Contour Plot of Chloride Concentrations (mg/L)
Along Transect A-A. Round 1 Data
p* - - - - L
-AINUNLL
/\(
B11/B11A
AG116/G116A P1/MW106 B15/B15P B10 G10B
B15R
740
700 7! V
S
s
^7
// /'BOB
--V ^^:
X ^' t^'tt
n/ x_
;^- i
Vjrf xl
X
IS
/ SANDY
/ SOILS^
X^
v\
(»
-%
11
n
y n
ACME
B16/B16A
DOLOMITE
A'
17
18
-660
•"f CJ^''''
Figure 6
Cross-sectional Contour Plot of Chloride Concentrations (mg/L)
Along Transect A-A'. Round 2 Data
FATE 125
-------
LANDFILL-
-ACME-
B
G115
B14 B13/P6 G114G113/ B9
G110 G113A
GRAVEL
B4
= 9
Figure 7
Cross-sectional Contour Plots of Chloride Concentrations (mg/L)
Along Transect B-B'. Round 1 Data
• 10
B'
-LANDFILL-
•ACME
G115
B
B13/P6 G114G113/ B9
G110 G113A
B4
SAND AND /
GRAVEL /
/ DOLOMFTE
B'
• 10
Figure 8
Cross-sectional Contours of Chloride Concentrations (mg/L)
Along Transect B-B'. Round 2 Data
126 FATE
-------
LANDFILL-
-ACME-
B
G115
B13/P6 G114G113/
G110 G113A
B9
B4
•ZONES OF
ELEVATED
ALKAL1NITIES •
Figure 9
Alkalinities (mg/L) Along Transect B-B'. Round 2 Data
277
LANDFILL-
-ACME-
EB
G115
B13/P6 G114G113/
G110 G113A
B9
/ 1
SAND AND / i?'51 \ 6-84
GRAVEL / V i6.76
I
G109A
/
I
DOLOMITE
B4
B'
-ZONES OF
LOW pH
7.60|
MW105
7.471
/
f
\ 7.05
3 /
7.40
Figure 10
Zones of Low pH Along Transect B-B'. Round 2 Data
FATE 127
-------
Chloride concentrations are contoured on cross-section B-B' for both
rounds of groundwater samples (Figs. 7 and 8). These cross-sections
indicate that chlorides are elevated in the vicinity of only one well (G110).
This indicates that the chlorides at G110 are an anomaly and not charac-
teristic of a plume. This previously was attributed to intermittent surfi-
cial leachate seeps currently under control. More recently it was reported
that this area was used to load trucks for off-site shipment of leachate
for treatment and disposal. Given that chloride is a good indicator of
the presence of leachate, it appears that a well developed landfill leachate
plume is not present at the southern margin of the landfill
Inspection of plots of alkalinities on these same cross-sections indi-
cates there are two zones of elevated alkalinities; one at Acme (B4)
and one at the southeastern margin of the landfill (Fig 9). A similar
pattern exists for pH (Fig. 10). It is evident that landfill leachate is not
responsible for patterns of alkalinities and pH since chlorides, a relia-
ble indicator of landfill leachate, are not increased as would be expected
if landfill leachate were present.
VOCs found at highest concentration in ground water samples col-
lected during this study were chlorinated ethenes, perchloroethene
(PCE), trichloroethene (TCE). cis-I.2-dichloroethene (DCH), vinyl
chloride (VC) and chlorinated ethanes (1,1,1-trichloroethane,
1.1-dichloroethane and chloroethane). Within each grouping, these com-
pounds may biodegrade through loss of a chlorine atom"0 Wood, et
al. and Vogel and McCarty found that PCE degraded to TCE to DCE
to VCX" Wilson, et al., and Barrio-Lage, et al., also found that DCE
degraded to VCXU. Barrio-Lage, et al., additionally determined that
the cis-isomer of DCE degraded to chloroethane as well as to VC12
The degradation product of trichloroethene is dominantly the cis-isomer
of 1.2 dichloroethene'. The degradation process is biologically
mediated and occurs under anaerobic conditions. The potential for
degradation of chlorinated compounds and the less widespread use of
less chlorinated compounds, indicates the presence of less chlorinated
species in groundwater result from the degradation of a more chlori-
nated parent compound.
The percent of PCE and VC relative to the total concentration of
ethenes in selected groundwater samples collected during this study
exhibits a general trend towards decreasing proportion of PCE and
increasing proportion of VC from east to west (Table 2). This finding
and the fact that almost all 1,2-dichloroethene detected in these ground-
water samples was the cis-isomer suggests that degradation is affecting
Table 2
Ptrcenl of Total Ethenes in Groundwater al Various Well Sites
PCE
IŁŁ
DCE
VC
84
B16
G108
G109
B12
G113
Gill
G114
G110
B13
B15R
6115
MW106
P3R
G116
B16A
B11A
G109A
G113A
P6
P4R
PI
G116A
42.3
12
32.5
18.3
5.7
22
16.6
0
2
8.2
8.1
0
3.6
0
0
5
34.1
6.6
14.1
19.4
15.7
0
19.6
18.3
14.3
23.1
6
6.6
4.6
11.6
7.4
7.6
15.3
23.2
0
31.6
13.2
0
5.5
15.9
26.1
30
22.5
15.7
16.7
14.9
39.3
73.7
43.1
55.8
80.2
63.4
71.8
42.4
13
72
52.6
29.2
50.2
71.2
0
89.5
50
59.1
53.4
58.1
66.5
62.9
65.5
0.1
1.3
1.3
19.9
7.5
10
0
50.2
77.4
4.5
16.1
70.8
14.6
15.6
0
0
0
8.3
2.5
0
2
20.4
0
the distribution of ethenes in groundwater. (Note that chloroethane was
not included in these calculations because a specific concentration of
chloroethane as a degradation product of cis-l,2-DCE could not reliably
be assigned since chlorinated ethanes are also present.) The distribution
of total ethenes in the groundwater is illustrated in Figure 11. The highest
concentration of ethenes was observed at location B4 (1912 ug/1), on
the ACME site Concentrations generally decline moving westward in
PREDOMNANTLY SANDY SOLS
22
ACME SOLVENTS |
S^'l _ — rj
PREDCMNANRY CLAYEY SOLS
Figure II
Numbers Adjacent lo Wells refer (o the Sum of (he
Perchloroethene, Trichloroethene, CIS 1.2 Dichloroethene
and Vinyl Chloride Concentrations (/»/!).
Note Thai Highest Concentration is Located East of the
Landfill on the Acme Site
NORTH
00 600
jCALE i!
i.'00
128 FATE
-------
the general direction of groundwater flow (Figure 11). Figure 12 is a
cross-sectional contour plot of total VOCs in wells along cross-section
A-A' which shows that VOCs are present both inside and outside of
the chloride plume.
CONCLUSION
The leachate plume from the landfill is well defined by the chloride
content and begins in the center of the landfill and extends to just past
the downgradient edge of the landfill. Chlorinated ethenes are present
inside and outside of the landfill leachate plume, indicating that the
leachate plume is mixing in a preexisting VOC plume (Figs. 5, 6 and
12). The presence of VOCs at the east end of the landfill is not attributed
to the presence of landfill leachate since the chloride concentrations
are not increased as would be expected if leachate were the source and
the area is hydraulically upgradient of the landfill.
The bimodal distribution of VOCs in this area may be due to one
or more of the following possibilities:
• The wells between Acme and the landfill simply may not intersect
a flow path through the fractured dolomite that is responsible for
the transport of VOCs from Acme. Indeed, the Illinois State Geo-
logical Survey concluded that this dolomite is difficult to monitor
because adjacent wells may be finished in fractures that are not con-
nected to each other13. The southeast margin of the landfill has a
high density of monitoring wells in comparison with other areas on
the bedrock upland increasing the chances of intersecting a VOC flow
path from Acme.
• The bimodal distribution could be the result of intermittent and
spatially variable recharge. This spatially variable intermittent
recharge could dilute the VOC plume resulting in variable VOC
concentrations, perhaps resulting in the appearance of a bimodal
distribution.
• Biodegradation may also play an important role in explaining the
appearance of the bimodal distribution of VOCs. Biodegradation could
increase the concentration of less chlorinated species which could
give the appearance of a bimodal distribution of VOCs.
A leachate plume from the landfill has been identified by the chlo-
ride content and is mixing in a pre-existing VOC plume. Landfill
leachate is not responsible for the groundwater chemistry anomalies
at the southeast margin of the landfill due to the lack of elevated chlo-
ride content and this area is upgradient of the landfill.
REFERENCES
1. Jordan, E.G., Acme Solvents Superfund Site, Winnebago County, Illinois,
Technical Report, Sep., 1984.
2. Cline, P.V., and Viste, D.R. "Migration and Degradation Patterns of Vola-
tile Organic Compounds," Waste Man. Res., 3, pp. 351-360, 1985.
3. Wood, P.R., Lang, R.F. and Payan, I.L. Anaerobic Transformation, Trans-
port and Removal of Volatile Organics in Groundwater, Drinking Water
Research Center, Florida International University, Miami, FL, 1981.
4. Leighton, M.M., Ekblaw, G.E. and Horberg, L. Physiographic Divisions
of Illinois, J. ofGeoi, 56(1), p. 16-33, 1948.
5. Hackett, J.E., Groundwater Geology of Winnebago County, Illinois, Illinois
State Geological Survey Report of Investigations 213, Champaign, IL, 56
pp., 1960
6. Berg, R.C., Kempton, J.P., and Stecyk, A.N., Geology for Planning in Boone
and Winnebago Counties, Illinois State Geological Survey, Champaign, IL,
Circular 531, 69 p., 1984
7. Bentley, H.W., Phillips, P.M., Davis, S.N., Habermehl, M.A., Airey, P.L.,
Calf, G.E., Elmore, D., Gove, H.E. and Torgersen, T. "Chlorine 36 Dating
of Very Old Groundwater 1., The Great Artesian Basin, Australia," Water
Resources Res., 22(13), pp 1991-2001, 1986.
8. Freyberg, D.L., "A Natural Gradient Experiment on Solute Transport in
a Sand Aquifer, 2. Spatial Moments and the'Adnection and Dispersion of
Nonreactive Tracers," Water Resources Res., 22(13), pp 2031-2046, 1986.
9. Vogel, T.M., Criddle, C.S. and McCarty, P.L. "Transformation of Halo-
genated Aliphatic Compounds," Environ. Sci. Tech., 21 pp. 722-736, 1987.
10. Bouwer, E.J. and McCarty, P.L. "Transformers of 1- and 2- Carbon
Halogenated Aliphatic Organic Compounds Under Methanogenic Condi-
tions," Appl. Environ. Microbiol., 45, pp 1286-1294, 1983.
11. Vogel, T.M. and McCarty, P.L. "Biotransformation of Tetrachloroethylene
to Trichloroethylene, Dichloroethylene, Vinyl Chloride and Carbon Dioxide
under Methanogenic Conditions," Appl. Environ. Microbiol., 49, pp
1080-1083, 1985.
12. Barrio-Lage, G., Parsons, F.Z., Nassar, R.S. and Lorenzo, P.A. "Sequen-
tial Dehalogenation of Chlorinated Ethanes," Environ. Sci. Tech., 20, pp
96-99, 1986.
13. Herzog, B.L., Hensel, B.R., Mehnert, E., Miller, J.R. and Johnson, T.M.
Evaluation of Groundwater Monitoring Programs at Hazardous Waste Dis-
posal Facilities in Illinois, Environmental Geology Notes 129, Illinois State
Geological Survey, Champaign, IL, 1988.
LANDFILL-
ACME
A
-740
-700
G116/G116A P1/MW106
B15/B15P
B15R
B11/B11A
B10 G108 B16/B16A A'
35.90
147.32
38.94
18.7
43.16
20.6
150.51
-660
Figure 12
Total Concentration of Volatile Organic Compounds (jt/1) for
Wells Along Transect A-A'. Round 2 Data
FATE 129
-------
Predicting Environmental Effects in a Puget Sound Embayment
David Tetta
U.S. EPA
Seattle, Washington
Don Heinle
Steve Costa
Walt Shields
CH2M Hill
Bellvue, Washington
ABSTRACT
An RI/FS is being performed at the Wyckoff/Eagle Harbor Super-
fund site, which includes Eagle Harbor itself. This is an embayment
located on the east side of Bainbridge Island in central Puget Sound.
The harbor area was first settled in the 1870s. Current and historical
operations are being evaluated as potential sources of contamination.
Several recreationally harvested fish species are found in Eagle
Harbor. National Oceanic and Atmospheric Administration research
has shown the strong relationship between polynuclear aromatic
hydrocarbons (PAH) in sediment and impacts on English sole, including
1 W, ckoff Facility 8 Queen City Yicht Club
1 Biinbridf e IiUnd Boilyird (proposed) 9 City of Wimlow Public Pier
3 Tyee Yacht Club 10 Biinbridge Mirine Service!
4 B*gled«le Moonnji 11 Eigle Hirbor Boil Retwir
5 Bifle Hiibor Mirini (Leiied from Biinbridge Mirine Servicei)
6 Hirbor Mirini l2 DOT Feny Miintenince Ficilily
7 WinilowWhirfMirini 13 DOT Feny Terminil
Figure I
Location of Industrial and Commercial Facilities and
Operable Units Along Eagle Harbor
130 FATE
-------
inhibition of ovarian development. The ecological effects portion of the
RI/FS for Eagle Harbor attempted to better define the areas where
ecological effects were likely, to determine what the sources of con-
tamination were and to estimate whether existing levels of contamina-
tion were likely to continue.
During the RI/FS, a variety of effects were identified. Sediments in
large areas of the harbor have been shown to be toxic to marine
organisms. The use of benthic taxonomic evaluation provided some
additional supportive information.
An evaluation of GC/MS results indicated that the sources of the PAHs
in sediment are mostly of creosote type origin.
Three distinct transport features appear to dominate the overall move-
ment of sediment associated contaminants. These are, in order of
importance: (1) remobilization of bottom sediments by vessel propeller
induced currents; (2) near surface and possibly subsurface flow of dense
non-aqueous phase liquid (DNAPL) from the wood treatment facility
and (3) the potential movement of material from Rockaway Beach to
the north and into the harbor.
Future depositions of sediments are expected to be significantly lower
than in the past because of blocking of natural sediment sources, par-
ticularly shoreline armoring. This implies very little potential for burial
of existing contaminated material in subtidal areas. The potential for
future flow of DNAPL to the harbor is still being evaluated.
INTRODUCTION
Eagle Harbor is a small embayment located in Central Puget Sound
on the eastern border of Bainbridge Island. The harbor area was first
settled in the 1870s. Historical operations along the harbor have included
shipbuilding during World War n as well as wood treatment operations.
Current operations now include ship repair and maintenance facilities,
a wood preserving plant and several marinas. Figure 1 shows Eagle
Harbor along with the major current operations.
Previous investigations by the National Oceanic and Atmospheric Ad-
ministration (NOAA)5*, Washington State Department of Ecology9'"
and the U.S. EPA have shown that sediments and clams in the harbor
are contaminated with poly nuclear aromatic hydrocarbons (PAHs).
NOAA found lesions and PAH accumulation in liver tissue in English
sole collected during trawls of the harbor, as well as impacts on ovarian
development, indicating possible impacts on future populations. In 1985,
the Bremerton-Kitsap County Health Department issued a health
advisory against eating shellfish from Eagle Harbor.
The Wyckoff wood treatment facility and Eagle Harbor were pro-
posed as a Superfund site in 1987. The U.S. EPA has since contracted
CH2M Hill to conduct an RI/FS on the harbor. As part of its focus,
the study has attempted to answer a number of questions including the
following:
• What are the ecological impacts of sediment contamination in Eagle
Harbor?
• How large is the impacted area?
• Where has the contamination originated from?
• What are the major routes by which the contamination is moving
around within the harbor?
ECOLOGICAL ASSESSMENT
General Ecological Characteristics
Eagle Harbor is inhabited by at least 18 species of fish. The harbor
provides nursery and adult habitat for a variety of invertebrate species.
Important fish and invertebrates include rockfish, cod lingcod, cancrid
crabs and pandalid shrimp. Several shellfish species are also present
in the intertidal areas of the harbor.
Most of the subtidal area in Eagle Harbor has sediment that is com-
posed of sandy silt to silty sand. Previous investigations of Eagle Har-
bor have shown elevated abundances of polychaeta—a pollution-sensitive
group. The active biological zone in Eagle Harbor sediments is consi-
dered to be the upper 10 to 20 cm.
Like the subtidal fauna, the nature of the intertidal fauna is deter-
mined in part by substrate (mud, sand or cobble). Intertidal communi-
ties within Eagle Harbor and the surrounding area have not been
extensively evaluated. The absence or near absence of macro-
invertebrates has been noticed in the immediate vicinity of oily seeps
around the Wyckoff facility10
Ecological Study Methods and Results
Previous NOAA research has focused on identifying a variety of
environmental effects in the harbor. The ecological effects portion of
this study focused on identifying those portions of the harbor that were
most likely to be producing those effects. The approach taken was
modeled after the triad method for evaluating environmental effects4.
This approach involves evaluating three components of sediment
quality—chemistry, toxicity and benthic effects. The toxicity of the sedi-
ments is determined through bioassay tests in contaminated and reference
areas. Benthic impacts are determined by evaluating the abundances
of major groups of benthic animals compared with those values found
in reference areas. In this approach to impact analyses, an area is con-
sidered impacted if one or more of the biologic tests shows a signifi-
cant effect.
To evaluate sediment toxicity amphipod and oyster larvae bioassays
were performed at 45 stations in Eagle Harbor as well as 10 reference
stations. Statistical comparisons were then performed between Eagle
Harbor stations and reference stations to determine whether a statis-
tically significant effect was observed in the Eagle harbor station.
One factor that has complicated the statistical comparison of stations
is the high mortalities that were found at some of the reference stations.
An evaluation of the data indicates that a likely reason for higher
mortality at a reference station is the higher level of silt content. Higher
mortality occurred only at reference stations with sediment containing
30% or more silt/clay (primarily silt).
Each station from Eagle Harbor was compared individually with a
group of reference stations with similar silt content using a pairwise
"t test." There were three possible positive end-points for the bioassays:
(1) mortality of amphipods, (2) mortality of oyster larvae and (3) ab-
normality of oyster larvae. Ten stations in Eagle Harbor had mortali-
ties of amphipods that were significantly higher than their respective
reference stations (Fig. 2 - PAH concentrations are shown in Fig. 3).
Eleven stations had mortalities of oyster larvae that were significantly
greater than reference, while nine stations showed significant levels of
live oyster larvae abnormality. All three bioassay responses were sig-
nificant at four stations.
A benthic assessment was performed using counts of total Crustacea,
mollusca, polychaeta and amphipoda, as well as presence or absence
of Phoxocephalid amphipoda. In Puget Sound, degraded areas generally
are characterized by a high proportion of polychaeta and mollusca, a
low proportion of Crustacea and a general absence of certain amphi-
pod families such as Phoxocephalids. However, benthic assessments
are also confounded by the high station to station variability within
impacted or non-impacted areas. Causal connections are also compli-
cated by the variety of factors (biological competition, sediment charac-
teristics and physical disturbances) that can impact community structure.
At the stations sampled in Eagle Harbor, polychaeta and mollusca
comprised more that 75 % of the fauna at 15 of the 42 stations sampled
(Fig. 4). The percentage of benthic fauna that were Crustacea was less
than 25 % at most stations in Eagle Harbor. Statistical analyses were
also performed in comparing benthic results with sediment physical
characteristics and contaminant levels. Polychaeta were found to corre-
late positively and significantly with HPAH and TPAH concentrations.
One additional affect was noted as part of the ecological assessment.
During the collection intertidal shellfish samples to evaluate PAH levels,
an absence of shellfish was seen in areas where intertidal seeps oc-
curred on the shoreline around the Wyckoff facility, and on the north
shore near the ferry maintenance facility and ship repair yard, similar
to the observations noted by Word, et al.10.
FATE AND TRANSPORT MODELING
Potential Sources
Possible scenarios for contaminant transport to Eagle Harbor include:
• Atmospheric deposition
FATE 131
-------
•7SS 717. .„,
728*
,c, 704
•7" 716(5) • •««
-•_ Oyster Larvae Mortality
Response Greater
Sediment Sample Station
Amphipod Bioassay
Response Greater
150 300
T
Scale in Maters
Oyster Larvae Abnormality
Response Greater
NOTE: Includes Station EH-08 Irom the Preliminary Investigation
I I I \
Figure 2
Stations Where Bioassay Responses Wire Significantly
Greater than All Other Stations
• Spills or dumps and subsequent redistribution by bottom currents
• Intertidal and subtidal seeps from the Wyckoff facility
• Longshore processes that would carry contaminated sediments from
the Wyckoff facility
• Seepage of creosote from pilings
An ocean sediment transport model was used to identify possible as
well as unlikely paths for contaminated sediments, identify possible
areas of sediment deposition and erosion for both contaminated and
uncontaminated sediments and provide semiquantitative estimates of
rates for sediment transport and accumulation.
Chemical Fingerprinting
AH sediment chemistry results were evaluated in an effort to deter-
mine whether PAH in a given sediment was due to fuel oil or creosote.
All samples were analyzed for PAH concentration by an HPLC pro-
cedure. Target compounds included the 16 PAH compounds on the
Priority Pollutant list. Samples also were analyzed for nitrogen-
containing aromatic compounds (NCAC) by GC followed by alakali-
tipped flame ionization detection (AFID). Target NCAC compounds
included: carbazole, quinoline, benzothiazole, benzonitrile, isoquino-
line, indole, benzoquinoline, acridine and methylcarbazole. Confirma-
tion of these analyses was performed on 25% of the samples using
GC/MS. The GC/MS analyses also allowed for source identification
and tracing via analyses of tentatively identified compounds.
PAHs constitute a variety of compounds that vary in their physical
and chemical properties. PAH compounds are a major component of
both creosote and fuel oil, which are the suspected sources of contami-
nation in Eagle Harbor. Creosote, which may be approximately 90%
PAHs, is a viscous liquid. Fuel oil typically contains 2 to 20%'
One feature of PAH chemistry that complicates the task of separating
past from present effects is that PAH mixtures in the environment change
over time. As lower molecular weight and more soluble components
dissolve, vaporize or degrade, the remaining resistant components
become relatively more abundant.
The NCAC levels, relative amounts of paraffins and ratios of indi-
vidual PAH compounds are useful indications of sources of contami-
nation. Comparison of samples from Eagle Harbor with suspected
source materials indicates the presence of contaminants from creosote
in the central harbor and on both the north and south shoreline. Other
hydrocarbons are present in greater proportion along the north shore-
line and at greater distance from the Wyckoff facility. Contamination
near the Wyckoff facility closely resembles creosote or wood preser-
vative wastewater and sludges. Figures 5 and 6 show this comparison
for PAH ratios. The ratios for fuel oil and bunker oil are from Neff,
the ratios for the "low napthalene" creosote are from Ingram2 and die
creosote multi-component standard is from Nestler*
Sediment Transport
The model development involved the following elements:
• A numerical model that computes the spatial and temporal distribu-
tion of the velocity field from the known geometric and tidal boundary
conditions
• A numerical model that computes the velocity field generated in
response to surface wind stress
• A calculation technique that uses a number of model applications
to predict alongshore transport
• A computer model that predicts the velocity fields (and critical grain
size for movement) generated by vessel propeller action
132 FATE
-------
ISO 300 600
I
Scale in Meters
NOTE: Includes intertkJal sediment stations and PI data (Tetra Tech, 1986b)
Figure 3
Average Concentrations of TPAH Calculated by Kriging (ug/kg)
• A set of criteria that evaluates the potential for sediment erosion,
transport and deposition
Tidal circulation was simulated with a vertically integrated, two-
dimensional finite difference model; wind driven circulation was simu-
lated with the same type of model. Vessel effects were assessed by using
a far-field velocity prediction routine for propeller induced, jet-like
flows. The overall circulation of the harbor was predicted by taking
a linear superposition of the results of each of the above models.
Oceanographic studies were performed to support this modeling
effort. These include:
• Bathymertric, hydrographic and water level measurements and data
processing
• Compilation and processing of existing wind data
• Wave climate predictions
• Current speed and direction measurements
• An evaluation of the potential for sediment movement and deposition
Examination of velocities in the harbor, based on the model results,
indicates that only a small interval of grain sizes will be subject to depo-
sition. Very fine sand and course silt, if available, can be deposited
throughout most of the harbor. Coarser material cannot be transported
within the harbor, and finer material will not be deposited but will be
flushed out by tidal currents. This phenomenon is shown schematically
in Figure 7.
Depositional areas in Eagle Harbor include deeper parts of the inner
harbor, the shoal northwest of the Wyckoff facility and the immediate
vicinity of streams entering the harbor (Fig. 8). Sediment probably does
not accumulate in the area of the PAH "hot spot" in the central harbor
because of the lack of source of coarse grained materials that could
be deposited there. Ferry propwash prevents deposition of the finer sedi-
ment in the central harbor that has accumulated in other parts of the
harbor.
In summary, three distinct transport features appear to dominate the
overall movement of sediment in the harbor:
• The potential movement of material from Rockaway Beach to the
north and into the harbor
• Remobilization of bottom sediments by vessel propeller induced cur-
rents (Fig. 9)
• Deposition of fine grained material in selected areas of the harbor
Sedimentation rate studies using lead 210 data indicate deposition rates
of 1.0 to 1.7 mm/yr3 at three locations in the harbor. This result is
based on cores dating back to approximately 200 to 300 yr. However,
the sediment load from the watershed and shoreline sources to the har-
bor may have varied significantly from the long-term average in recent
years due to land cover changes and land use practices in the watershed
and the construction of shoreline protective structures.
Although the lead 210 data may be expected to reflect past sedimen-
tation rates, future rates are expected to be much smaller than in the
past. Typically, 75% of the paniculate load to Puget Sound is from river-
line and shoreline sources. There is no reason to expect this deposition
rate to be different for Eagle Harbor. An analysis of watershed processes
and soil loss estimates, and the nearly total armoring of the shoreline
of the harbor and adjacent Rockaway Beach, indicate that the major
FATE 133
-------
• 75-100%
• 50-75%
V 25-50%
O 1-25%
NaphOtatan* AcmapfitfMna FluonM Phananllvana Fkforvn
Figure 4
Relative Percent Polychaeta and Mullusca for
June 1988 SubtidaJ Samples
Naphlhalana AcanipMhana Fluorafta Phananthrana Fluoranthan* Chryaana
FUEL OIL
4-
3-
2-
t-
BUNKER OIL
,»»,»„» Wffif^
"LOW NAPHTHALENE" CREOSOTE
I
8
CREOSOTE
MULTICOMPONENT
REFERENCE STANDARD
Figure 5
PAH Composition in Potential Source Materials
For Eagle Harbor
NOTE: PAH compoiltton baud on GC-MS ratuta
Figure 6
PAH Composition in Sediments from Central Harbor
134 FATE
-------
1000
100 -
d(mm)
o
0.075 mm (Fine Sand)
0 ISO 300
0 500 1000
600 METERS
Figure 8
Areas of Potential Deposition for Fine-Grained Sediment as
Predicted by Transport Model
FATE 135
-------
FERRY
MAINTENANCE
FACILITY
WYCKOFF \ >»
FACtUTY
LEGEND
Bottom Velocities:
5 — •+
6 -*
45 cm/sec
30 cm/sec
20 cm/sec
15 cm/Sec
12 cm/sec
10 cm/sac
Note: 1 through 6 Indicate number ol
boat lengths behind ferry.
300
Scale in Meters
Figure 9
Predicted Area of Ferry Propeller Influence
through near-surface or subsurface flow.
Processes that effect the persistence of PAH in the biological zone
of sediments in the harbor are biodegradation, photo-oxidation and
volatilization. LPAHs generally are more likely to be dispersed from
discharge areas by solution in water and volatilization to the atmosphere.
HPAHs are lost by biodegradation and photo-oxidation, but are gener-
ally very persistent in aquatic sediments. Of these processes, only bio-
degradation is considered to have an impact on PAH persistence.
Another process that can have a significant impact on PAH values
in the biological zone is natural recovery or burial. Based on the sedi-
ment transport assessment presented earlier, it appears that PAHs
associated with bottom sediments: (1) will not be transported out of
the harbor and will not be rapidly dispersed within the harbor and (2)
will not be rapidly buried by clean sediment. In addition, reduced
sedimentation may enhance concentrations of PAH arriving from the
watershed.
CONCLUSIONS
Available data do not permit us to determine whether the PAH present
in percent amounts in the central harbor are there as a result of past
discharges or spills are as a result of possible continuing discharges
of DNAPL. Some suspension or resuspension of PAH-contaminated
sediment may occur in areas affected by ferry propeller wash, but trans-
port away from the areas of higher level contamination may be inhibited
by the low bottom velocities and the grain size of the affected sedi-
ment. Separate phase flow of DNAPL contaminants may occur over
short distances in areas of the harbor where hydraulic gradients are
present. Additional studies are being planned to evaluate the potential
for DNAPL transport to the harbor.
REFERENCES
1. Gilchrist, C.A., Lynes, A., Steel G. and Whitham, B.T., "The Determina-
tion of Polycyclic Aromatic Hydrocarbons in Mineral Oil by Thin-layer
Chromatography and Mass Speclrometry," Analyst. 97 pp. 880-888, 1972.
2. Ingram, L.L., el at., "Migration of Creosote and Its Components from
Treated Piling Sections in a Marine Environment," Proc. American Bborf-
Preservers' Association. 78, pp. 120-128, 1982.
3. HartCrowser. Conuwunani Deposition and Sediment Recovery, Eagle Harbaf
Site, Kitsap County, Hbshington. Draft Report prepared for Washington Sate
Department of Transportation, Mar., 1989.
4. Long, E.R., and Chapman, P.M., A Sediment Quality Triad; Measurement
of Sediment Contamination, Toxicity and Infaunal Community Composi-
tion in Pugel Sound. Mar. Poll. Bull, 16, pp.405-415, 1985.
5. Malms D.C., Summary Report on Chemical and Biological Data from Eagle
Harbor, National Marine Fisheries Service, Seattle, WA, 1984.
6. Malins D.C. et al., "Toxic Chemicals in Sediments and Biota from*
Creosote-polluted Harbor: Relationships with Hepatic Neoplasms and Other
Hepatic Lesions in English Sole," Carcinogenesis, 6, pp.M63-M69, 1985.
7. Neff, J. M., Polycyclic Aromatic Hydrocarbons in the Aquatic Environment;
Sources, Rues and Biological Effects, Applied Science Publishers, Essex,
England, 1979.
8. Nestler, F.H.M., Characterimtion of Wood Preserving Creosote by HO«"
cat and Chemical Methods of Analysis, USDA Forest Services Research
Paper FPL 195, US Department of Agriculture, Forest Products Laboratory,
Madison, WI, 1974.
9. Tetra Tech, Inc., Preliminary Investigation, Eagle Harbor, Bainbridgefttairf,
Hbshington, Prepared by Tetra Tech, Inc. and Black and \featch for the
Washington Department of Ecology, Nov., 1986.
10. Ward, a d.,RecomaissanceSur^of IruertidWProperty O»^by tyctoi
Company in Eagle Harbor, Memorandum from Battelle Northwest Marine
Research laboratories to Darrel Palmer, Wyckoff Company, 198&
11. \ake, etal., Chemical Contaminants in Clams and Crabs from Eagle Harbor,
Wishington State, with Emphasis on Polynuclear Aromatic Hydrocarbon!:
Washington Slate Department of Ecology, Water Quality Investigation Section,
Oct., 1984.
136 FATE
-------
Considerations for Discharge of Contaminated Groundwater
To a Municipal Sewer System and POTW
William B. Lindsey, RE.
Raul E. Filardi, Ph.D.
B & V Waste Science and Technology Corp.
Overland Park, Kansas
Stephen D. Chatman
Laura C. Perkins
McLaren
Springfield, Missouri
Ralph E. Moon, Ph.D.
Geraghty & Miller, Inc.
Tampa, Florida
INTRODUCTION
The remediation of contaminated groundwater at hazardous waste
sites often involves alternatives for its extraction, treatment and dis-
posal. One alternative that may be considered is discharge of contami-
nated groundwater to a municipal wastewater collection system for
conveyance to a publicly owned treatment works (POTW). This paper
presents observations made during extended use of a city's municipal
wastewater collection system and POTW for discharge of trichloro-
ethylene (TCE)-contaminated groundwater during aquifer and pilot
testing for an NPL site in Missouri.
As part of the Remedial Investigation (RI) conducted at the site, aquifer
testing was planned for a deep bedrock aquifer to evaluate its hydrogeo-
logic properties'. The deep bedrock aquifer was one of three area
aquifers, and only it produced an adequate quantity of water for a
drinking supply. It presently serves as the city's municipal water supply.
The city's wastewater collection system and POTW were proposed for
disposal of groundwater from the aquifer tests.
Information for this paper was gathered from three studies; a study
which examined the feasibility of using the existing sewer system and
POTW for discharge, the RI and the FS2. Data from these studies
were organized in two phases for this paper: (1) an evaluation of the
capacities and background contaminant levels of the existing sewer
system and POTW; and (2) an extended pilot testing program which
monitored TCE removal by air strippers, levels of TCE discharged to
the sewer from the air strippers and from an off-site well, and TCE
levels of the POTW influent and effluent.
To effectively and economically implement a remedial action for the
site, numerous site hydrologic characteristics were investigated during
the RI, including the extent of influence of recovery well pumping. To
characterize this information, aquifer tests were conducted at selected
on-site monitoring wells and abandoned Municipal Well No. 1 located
off-site. The aquifer tests consisted of pumping the wells at various rates
and measuring the water levels in surrounding wells to define the cone
of depression associated with potential operation of the recovery
well.3.
The proposed method for disposal of TCE contaminated groundwater
generated during aquifer and pilot testing was discharge to the city's
municipal sewer system and treatment at the POTW. This method was
agreed to by the Missouri Department of Natural Resources (MDNR)
and the city, subject to pretreatment limits to be discussed later. The
following goals and objectives were developed to answer questions
associated with the use of the municipal sewer system and POTW for
possible extended disposal during remedial actions.
• Evaluate the capability of the POTW to remove TCE contamination
• Establish the POTW's National Pollutant Discharge Elimination
System (NPDES) permit limitations or requirements
• Determine the flow capacity of the sewer system
• Determine existing TCE concentrations in the collection system flow
path from near the site to the POTW
• Determine if TCE volatilization or dilution were occurring in the
collection system
• Document the effectiveness of air strippers in removing TCE during
aquifer testing and conformance to pretreatment requirements
• Evaluate any health risks associated with TCE vapors in the collec-
tion system or emissions from the POTW
SITE INFORMATION
The city is located in Missouri and has a population of approximately
6,000 people. Figure 1 shows the location of the site in the city. A former
industrial and manufacturing site, it had been leased and operated by
a number of business concerns through the years. The site consists of
a 21,000 ft2 lot enclosed within a 6-foot high chain-link fence. A
former plant building, constructed before 1902, stood on the site. In
1979, the northern portion of the building was destroyed by fire. The
fire-damaged portion of the structure was demolished and the debris
was pushed into the basement under that portion of the building.
0 500 1500 MOO ttt(
o CW- _ Munjci(ul Wdl
».Ol«c Well
* RatKMl Sprin,
Figure 1
Location Plan
FATE 137
-------
In June, 1982, MDNR and the U.S. EPA selected the city's municipal
water supply wells for random sampling and analysis for volatile or-
ganic compounds as part of U.S. EPA's National Synthetic Organic
Chemicals Survey. The locations of the municipal wells arc shown in
Figure 1. TCE was detected at 15 mg/L in samples collected from
Municipal Well No. 1. This value was below Missouri's health-based
criterion which, at that time, was 27 mg/L. Additional samples from
all three city wells were collected in March, 1983. A TCE level of 10
mg/L was detected in Municipal Well No. 1 while none (at a detection
limit of 10 mg/L) was detected in Municipal Wells No. 2 and 3.
During the subsequent RI, volatile organic compounds, primarily
TCE, were detected in on-site subsurface soils and groundwater of the
three separate aquifer systems: the unconsolidated materials/fractured
shallow bedrock system, unfractured shallow bedrock system and deep
bedrock system. The unfractured shallow bedrock system is a minor
aquifer. The deep bedrock system is a major aquifer capable of yields
as high as 2,000 gpm. All of the city's municipal supply wells are located
in the deep bedrock aquifer. TCE contamination detected in the deep
bedrock wells varied, depending on the location of the well in relation
to the contaminant plume. TCE concentrations ranged from non-
detectable for most off-site wells, up to 200 mg/L for Municipal Well 1.
and up to 18,000 mg/L for on-site wells. Aquifer tests and pilot tests
were proposed for the deep bedrock wells to determine their hydro-
geologic properties for development of remedial pumping strategies.
BACKGROUND CONDITIONS
Based on historic data, it was anticipated that the aquifer testing and
pilot testing would generate up to 200 gpm of potentially TCE-
contaminated groundwater. Therefore, it was necessary to ensure that
the groundwater generated during these activities be properly controlled.
treated and discharged in a manner which met all RCRA, state and
local requirements. Because of the need for a readily available meant
of discharge, the city's wastewater collection system and POTW were
proposed for discharge of the groundwater from aquifer testing. Before
initiating discharge to the sewer, the feasibility of conveying test flows
through the sewer system as well as the potential effects on the POTW
and its anticipated performance were evaluated. The existence of any
background levels of contaminants in both the sewer system and POTW
also was established.
Sewer System
An evaluation of the sewer system was performed to determine if TCE-
contaminated groundwater could be discharged to the system. The
objectives of the evaluation were to determine the following:
• The presence of TCE in the sewer system
• The ability of the sewer system to convey the additional flows
Once the presence of TCE contamination in the sewer system was
established, other objectives were added.
• Determine if volatilization or dilution of TCE from the wastewater
was causing a reduction of TCE levels in the system
• Determine the possible effects of discharging untreated or treated
TCE-coniaminated groundwater to the sewer system, such as TCE
vapor build up, creation of health hazards to sewer maintenance
workers, or development of explosive atmospheres
Sewer System Description
A layout of the sewer system, sewer flow direction and year of sewer
construction are shown on Figure 2. Flows discharged from the site
would be conveyed to the POTW through both gravity sewers and force
0 COO 1200
^•M
FEET
LEGEND
TRIBUTARY BOUNDARY
GRAVITY SEHER LINE
MAKHOLE
Figure 2
Existing Sewer System
138 FATE
-------
mains. From the site, wastewater flows in 8- and 10-in. vitrified clay
pipe (VCP) sewers, constructed in the mid-1950s, and continues in a
15-in. reinforced concrete pipe (RCP) to the lift station, both constructed
in the 1980s. From the lift station the flow is conveyed in a force main
to 18- and 21-in. diameter gravity sewers leading to the POTW.
The condition of the sewer system had been evaluated during an earlier
infiltration/inflow (I/I) study4. The I/I study found the existing sewers
flowing at less than half capacity during normal conditions, but
surcharging during rainfall events. Infiltration/inflow was found to be
excessive in the older sewer sections. The sewers near the site were
found to surcharge during rainfall events. The I/I study recommended
that a sewer system evaluation survey (SSES) be conducted to further
evaluate the condition of the sewer system. The SSES included a physical
survey, television inspection and smoke testing of selected portions of
the system. The television inspection revealed that the condition of lines
near the site varied. As would be anticipated for VCP sewer lines con-
structed in the mid-1950s, many leaking joints and root intrusions were
observed. A sewer system rehabilitation program was conducted in 1984
to reduce I/I sources.
Sewer System Background Sampling
In December 1986, and January, 1987, the MDNR collected water
samples in Southwestern Bell Telephone manholes near the site and
found trichloroethylene (TCE) and other volatile organics. The detec-
tion of these chemicals suggested the introduction of contaminated shal-
low groundwater into the sewer system. Wastewater was sampled from
selected manholes and analyzed for priority pollutant volatile organics
to establish background conditions. In addition, wastewater flow was
measured when the samples were collected using a calibrated V-notch
weir. The sampling was conducted on three occasions: June, 1987; July,
1987; and May, 1988. Sampling point locations and TCE concentra-
tions are shown in Figure 3.
The wastewater in manholes on the gravity sewer leading away from
the site contained TCE in various concentrations. In general, the TCE
concentrations decreased with distance from the site, as would be
expected, because of dilution by incoming wastewater from downstream
branches.
The decrease in TCE concentrations away from the site raised two
questions. Was the decrease caused by dilution or volatilization in the
turbulent wastewater flow? If significant volatilization was occurring,
could vapor build up to levels which might present a health hazard to
sewer maintenance workers? To determine whether dilution or volatili-
zation was causing the decrease in TCE levels, instantaneous flow
measurements taken during the June, 1987 sampling period were
examined. TCE concentrations, flow rates and TCE mass loads for three
manholes where flow was measured are listed in Table 1. The manhole
numbers correspond to those on Figure 3.
Table 1
TCE Concentrations and Flow Rates in Sewers
(June 1987 Sampling)
Manhole No.
5
9
10
TCE
(ppb)
230
17
16
Flow
(gpm)
12
98
101
(pounds/day)
0.033
0.020
0.019
The decrease in TCE mass loading in downstream manholes indi-
cated that volatilization may occur within the sewer system. However,
the concentration of volatile organic vapors was measured before entering
the manholes but organic vapors were not detected above background
levels.
SRAVITY SEHER LINE
MANHOLE
LAMPHOLE
FORCE MAIN
LIFT STATION
Figure 3
Sewer System Sample Points and TCE Concentrations
FATE 139
-------
Sewer System Evaluation
The flow capacities of the newer sections of the sewer system were
calculated from construction drawings. Flow capacities of the older
sewer sections, for which no construction drawings were available, were
established by a field survey to obtain pipe diameters and invert eleva-
tions. The sewers had adequate capacity to simultaneously carry peak
domestic wastewater flows and aquifer test flows. However, the sewers
are known to surcharge during heavy rainfalls, which would prevent
discharge of aquifer test flows during these periods.
POTW
The POTW was evaluated to determine if TCE-contaminated ground-
water could be discharged to it. The objectives of the evaluation were
to determine the following:
• The presence of TCE in the POTW influent, effluent or sludge
• The ability of the POTW to remove TCE from groundwater
• The capacity of the POTW to handle the additional flows
• Any adverse impacts the TCE might have on plant operations or
performance
POTW Description
As shown on Figure 4, the POTW is an activated sludge plant using
brush rotor aerators and mixed media filters. The design average flow
was 926,880 gal/day or 644 gpm, and the hydraulic capacity was 7.34
mgd (5,070 gpm). The plant is governed by a MDNR NPDES permit
which stipulates monthly average limits of 10 mg/L for biochemical
oxygen demand (BOD) and 15 mg/L for suspended solids (SS). The
permit stipulates a monthly average TCE discharge limit of 2 mg/L,
to be measured once every 6 mo.
Raw wastewater enters the POTW through a manually cleaned bar
screen and is pumped by two enclosed 54-in. screw pumps. The flow
is measured by a Parshall flume before it enters a multiple channel
aeration basin. Aeration and mixing in the basin are accomplished by
brush surface aerators. Next the flow enters two 55-ft diameter clari-
fiers. The clarifier effluent flows to filters equipped with an automatic
traveling bridge backwashing mechanism. Filtered effluent passes
through a chlorine contact basin and is discharged to a reaeration
structure.
POTW Background Sampling
A background sampling program was conducted at the POTW to determine
ICE concentrations in the plant influent, effluent, and sludge. The sample num-
bers and locations are shown in Figure 4. Samples were collected on June 11
and 12, and on July 14, 1987; the results are summarized in Table 2. During
the June sampling period, influent and effluent samples were collected at different
times of the day to determine diurnal variations in TCE concentrations.
Table!
POTW Influent, Effluent and Sludge Samples
Sample AMALYTES
Number
Location
ill ill
(ppb) (ppb)
ill ill ill
(ppb) (ppb) (ppb)
The background TCE concentrations in the plant influent ranged from
non-detectable to 10 mg/L. TCE concentrations in the plant effluent
were below detection limits. There were no detectable TCE concentra-
tions in the plant sludge samples; however, toluene was detected at a
concentration up to 27,000 ppb.
TCE was detected in the POTW influent but it was below detectable
levels in the POTW effluent. The POTW apparently could reduce
influent TCE at the concentrations received during sampling to less
than the 2.0-mg/L discharge limit. From this evidence the ability of
the POTW to treat higher levels of TCE could not be determined. It
also appears that at the influent TCE levels found during sampling, (here
is no TCE carryover to the sludge.
Evaluation of POTW Treatment Potential
The potential of the POTW to remove the TCE that exceeded back-
ground levels and to meet the 2 mg/L discharge limit was evaluated.
A literature search was conducted to review the biological treatment
of TCE from contaminated groundwater. POTWs have reported signifi-
cant removals of volatile organic carbon (VOC) by various treatment
processes''". These reported removals include secondary treatment
processes such as activated sludge plants using surface aeration, diffused
air and pure oxygen; trickling filters; aerated lagoons; rotating biologi-
cal contactors; air strippers; and advanced wastewater treatment systems
which incorporate tertiary treatment, such as mixed media filtration.
The principal mechanisms involved in TCE removal at air activated
sludge plants are air stripping, adsorption on the microbial growth and
biodegradation. Volatilization by air stripping was reported to be the
primary mechanism involved in removal of TCE in activated sludge
plants. TCE can be volatilized into the atmosphere in the plant collec-
tion system, wet wells, grit chambers, aeration basins and post-aeration
devices (weirs). TCE removal efficiency values in activated sludge
processes reported in a U.S. EPA study0 ranged between 68 and 90%
and were up to 97% when followed by tertiary treatment such as effluent
filtration. The wide range of removal efficiencies underscores the fact
that removal estimation requires plant-by-plant evaluation. The removal
capacities of individual plants are found to be strongly influenced by
physical configuration.
The POTW average design flow was 926,800 gpd (0.93 mgd),
however, it was determined that the POTW could handle a flow of
1.46 mgd at the design per capita organic and solids loadings. The
POTW has a hydraulic capacity of 7.34 mgd and the capacity of the
clarifiers, at normal design loading rates, is 2.9 mgd. Flows from aquifer
testing, which have no organic or solids loadings, were not expected
to affect plant performance or its ability to meet NPDES requirements,
EXTENDED AQUIFER TESTS - RESULTS
Based on the background samples gathered from the sewer system,
the POTW and other information, MDNR, the pretieatrnent authority
for the city's POTW, recommended the following discharge limits for
the sewer system and the POTW which were adopted by the city:
TPl
TP3
TP2
TP2A
TP4
TP5
TPl
TP3
TP3A
TP5
Notei Concentration! below the detection limit of 2 ppb ere indicated by KD.
(1) Trichloroethjrlene
(2} Toluene
(3) - Chloroform
(4) trani-j.,2-Dlchloroethene
(3) - 1,1,1-Trlchloroetlune
Influent
Effluent
Influent
Influent
Effluent
Sludge
Influent
Effluent
Effluent
Sludge
8/11/87
6/11/87
6/12/«7
6/12/97
6/12/87
1/12/17
7/14/17
7/14/17
7/14/67
7/14/17
3.1
ND
«.«
10.0
KD
ND
ND
ND
ND
ND
3.»
ND
NDO)
2.2
ND
27.000
13.0
KD
HD
4.100
7.0
ND
31.0
J8.0
ND
ND
ND
KD
11.0
6. a
ND
ND
4.6
9.0
ND
HD
KD
KD
ND
ND
ND
ND
ND
ND
ND
ND
HD
ND
4.3
KD
Monitoring Point
Sewer Die-charge
POTW Influent
POTW Effluent
Table3
Pretreatment Discharge Limits
Allowable
TCK Concentration U«it
(ppb)
200
Established by the POTW NPDES permit.
The city agreed to allow discharge of fluids into the sewer system
and the POTW. The agreement allowed up to 200 gpm of groundwater
meeting the established limits to be discharged.
Groundwater was discharged to the sewers from two sources, untreated
groundwater from Municipal Well No. 1 and treated groundwater from
onsite wells. Groundwater from Municipal Well No. 1 was generated
140 FATE
-------
AUTOMATIC BACKWASH
FILTERS-
CHLORINE CONTACT
CHtMICft
ACCESS ROtO
OLD INLET
STRUCTURE
OLD CONTROL
•UILDING
OUTLET
STRUCTURE-
-TP-3
TP-3A
TP-«
PARTIAL PLAN
NTS
Figure 4
POTW and Sample Locations
during long-term aquifer testing. TCE concentrations in groundwater
from Municipal Well No. 1, located approximately 500 ft south of the
site, were below the 200 mg/L limit; therefore, no treatment was
required. Discharge of groundwater from Municipal Well No. 1 to the
sewer commenced in August, 1987, and was periodic through January,
1988 and then essentially continuous from January, 1988 through
January, 1989. The concentration of TCE during this period is shown
in Table 4. Until March, 1988, samples from Municipal Well No. 1
were taken at least once per day. After that date, the frequency was
changed to once every 2 wk with the city's approval since the TCE lev-
els in Municipal Well No. 1 were shown to be consistent.
Groundwater which was generated from on-site wells during the
remedial investigation contained TCE concentrations significantly higher
than the 200 mg/L sewer discharge limit. Two air strippers, with a flow
capacity of 150 gpm and operating in series, were constructed to reduce
groundwater TCE concentrations to acceptable levels. The air strippers
were operated intermittently for two periods: in September and October,
1987; and January through March, 1988. TCE concentrations in the air
stripper Tower 1 influent, Tower 1 effluent/Tower 2 influent, and Tower
2 effluent/sewer discharge, including the total TCE removals, are shown
in Table 5. The frequency of collecting samples from the Air Stripper
No. 2 effluent/sewer discharge was one or more times per day.
Concurrently with the discharge of groundwater to the sewers, the
POTW influent and effluent TCE concentrations were monitored.
POTW sampling began in June, 1987 and continued until January, 1989.
The POTW influent and effluent TCE levels are listed in Table 6. The
frequency of collecting POTW influent and effluent samples was once
per day until March, 1988, when the frequency was changed to once
every 2 wk. The City approved the request to decrease the frequency
for this sampling also because the TCE concentrations were shown to
be consistent.
Figure 5 shows the relationship between TCE concentrations in the
discharges from Municipal Well No. 1 and Air Stripper No. 2 to the
sewer and the corresponding POTW influent and effluent TCE con-
centrations. Based on expected sewer flow quantities, the travel time
from the discharge point of Municipal Well No. 1 or Air Stripper No.
2 to the POTW would be approximately 2 hr. Based on measured flow
rates, the average hydraulic retention time at the POTW during the
pumping was approximately 40 hr.
Air samples were taken from the sewers and at the POTW during
air stripper operation on Feb 3 and June 16, 1988 to determine the
presence and/or concentration of VOCs in the air. The chemicals present
were qualitatively identified using a portable gas chromatograph. Results
of the samples are shown on Table 7. The sewer air samples were taken
at Manhole MH-6, as identified on Figure 3, and the POTW air samples
were taken downwind of the aeration basin channels.
EXTENDED AQUIFER TESTS-DISCUSSION
During extended pumping of Municipal Well No. 1, the groundwater
TCE concentrations became stabilized in the range of 40 to 80 mg/L,
FATE 141
-------
Table 4
TCE Concentrations in Groundwater From Municipal Well No. 1
Table 6
POTW Influent and Effluent Concentrations
SAMPLE
DATE
08/10/87
08/10/87
09/02/87
09/02/87
09/03/87
09/03/87
09/03/87
09/03/87
09/03/87
09/03/87
09/03/87
09/03/87
09/12/87
10/24/87
10/25/87
10/26/87
1 0/27/87
10/28/87
10/29/87
10/29/87
11/04/87
11/04/87
11/05/87
11/06/87
11/07/87
11/08/87
11/08/87
11/09/87
12/15/87
12/16/87
12/17/87
12/18/87
TCE
CONCENTRATION
(PPB)
32
44
180
110
200
106
117
87
140
150
89
99
150
93
100
88
110
99
78
67
60
76
64
76
127
83
86
80
40
57
55
71
SAMPLE
DATE
01/01/68
02/04/88
02/27/88
03/01/88
04/15/88
05/15/88
06/02/88
06/15/88
07/15/88
08/01/88
08/15/88
09/01/88
09/15/88
10/14/88
11/01/88
11/15/88
12/01/89
12/15/88
01/01/89
TCE
CONCENTRATION
(PPB)
76
20
27
43
57
52
48
40
43
48
46
27
51
52
53
58
69
57
29
TableS
Air Stripper Influent and Effluent TCE Concentrations
DATE
09/21/87
09/21/87
09/21/87
09/24/87
09/24/87
09/24/87
10/16/87
11/03/87
12/10/87
01/18/88
01/19/88
01/20/88
01/21/88
01/22/88
01/23/88
01/24/88
01/25/88
01/26/88
01/27/88
01/30/88
02/03/88
06/16/88
1 2/23/88
12/23/88
TOWER 1
INFLUENT
TCE (PPB)
6800
1800
260
830
780
6700
360
-
99
3800
3200
4400
4800
4000
3700
3300
4900
3200
3600
2100
2500
4200
3800
3900
TOWER 1 EFFLUENT
TOWER 2 INFLUENT
TCE (PPB)
280
76
11
160
140
570
ND
110
33
430
440
370
250
320
230
300
240
470
240
46
-
110
150
160
TOWER 2 EFFLUENT
SEWER DISCHARGE
TCE (PPB)
33
21
80
11
-
58
28
26
57
4 7
57
60
60
59
53
46
4 6
63
3.8
ND
-
4 2
62
6.4
TOTAL
TCE
REMOVAL (%)
99 95
98.83
9692
9867
-
9991
9922
-
94 24
9988
9982
9986
9988
9985
9986
9986
9991
9980
9989
9990
_
9990
9984
9984
SAMPLE
DATE
7 / 17/87
7 / 17/87
7 / 18 /67
7 / 19 /87
7/29/87
7/30/87
7 / 31 /87
/ 1 /87
/ 2 /87
/ 5 /87
/ 6 /87
/ 10 1ST
I 10 /87
/ 12 /87
/ X /87
i 20 /87
1 21 /87
/ 22 /87
1 23 /87
/ 24 /87
/ 4« /87
8/47/87
8/49/87
8/31/87
9 / 1/87
9 / 2/87
9 / 3 /S7
91 12/87
9 / 17/87
9/17/87
9 / 18/87
9/18/87
9/24/87
9/25/87
9/28/87
9/29/87
9/30/87
10 / 1/87
10 / 2/87
10 / 16 /87
10 / 24 /87
10/25/87
10 / 26/87
10 / 27 /87
10/28/87
10/29/87
11 1 3/87
11 / 4/87
INt-LUtNI
(PPB)
NO
NO
2.7
4.4
NO
NO
2.3
28
14
48
NO
NO
IS
NO
7
10
34
i7
33
2
NO
83
76
42
2
NO
8.1
18
NO
6-6
4 4
44
8.4
2.8
5.5
27
3
ND
NO
NO
24
29
28
19
18
15
2.1
12
tf-KLUtN"!
(PPB)
NO
NO
ND
NO
NO
ND
NO
NO
NO
NO
NO
ND
NO
ND
ND
NO
NO
NO
NO
ND
NO
NO
NO
NO
NO
NO
NO
ND
NO
NO i
NO
NO
NO
NO
NO !
NO
NO
NO
NO
NO
NO i
NO
NO
NO
ND
NO
NO
NO
SAMPLE
DATE
11 / 6/87
11/8/87
11 / 7/87
11 / 8/87
11 / 8/87
11 / 9/87
11 / 10/87
12 / 5/87
12 / 16/87
12 / 17/87
12 / 18 «7
1 / 1/88
1/18/88
1/19/88
1/20/88
1/21/88
1/22/88
1/23/88
1/24/88
1/25/88
1/28/88
1/27/88
1 / X «8
2/22/88
2/27/88
3 / 1/88
3/15/88
3/21/88
4 / 1/88
4/15/88
5 / 1/88
5/15/88
6 / 2/88
6/15/88
7 / 1/88
7/15/88
6/1/88
8/15/88
9 / 1/88
9/15/88
10 / 1/88
10 / 15/88
11 / 1/88
11/15/88
12 / 1/88
12 / 15/88
1/1/89
INFLUENT
(PPB)
15
27
27
28
27
13
14
13
20
15
11
15
81
24
28
28
15
21
18
8.3
16
19
9.5
7.8
14
22
15
18
8.6
8.9
5.3
S.4
13
6.8
9.2
NO
4.7
6
2
38
73
7.4
NO
7.5
11
9.5
97
~ B+Lumri
(PPB)
NO
NO
NO
NO
4
ND
NO
NO
NO
NO
NO
NO
ND
NO
NO
ND
6.1
NO
ND
NO
2.2
38
NO
NO
NO
ND
ND
NO
NO
ND
ND
NO
ND
NO
NO
ND
ND
NO
NO
NO
NO
NO
ND
NO
NO
NO
NO
Table?
Constituents Detected in Air Samples
VolitUt Ormiict
(ppb)
02/03/88
Hflt""
06/16/88 02>Q3/«8 24/14/51
Benrtn* ND (23)
M>thjrl*n« Chlorldl 5000
Tecrachloroethcn* HD (50)
1.1.l-Trlchloro«th«n« NA
Trlchloro«th«n« NO (23)
ND (1.0)
NA
ND (5.0)
ND (100)
1.0
290
11000
720
NA
770
ND (1.0)
MA
MB (S.OI
HD (100)
1.0
142 FATE
-------
o
180 —1
160 —
„ 140
m
Q.
Q.
120 —|
100
80
o
u
LU 60
u
i-
40
20
MUNICIPAL WELL NO. 1 DISCHARGE
AIR STRIPPER NO. 2
EFFLUENT/SEWER DISCHARGE
POTW EFFLUENT
JUL ' ftUG ' SEP ' OCT ' NOV ' DEC ' JflN ' FEB MAR ' APR ' MAY ' JUN ' JUL AUG ' SEP ' OCT ' NOV ' DEC ' JflN
1987
1988
1989
SAMPLE DATES
LEGEND
TCE CONCENTRATIONS DURING CONTINOUS DISCHARGE
TCE CONCENTRATION FROM SINGLE SAMPLING EVENT-
CONNECTED FOR CLARITY
Figure 5
Ice Concentrations for Discharges from Municipal Well No. 1,
Air Stripper No. 2, and POTW Influent and Effluent
below the 200 mg/L discharge limit. Flow quantities from this well
ranged from 50 to 75 gpm which did not exceed the flow capacity the
sewer system at any time.
The two air strippers operating in series effectively reduced TCE con-
centrations in on-site wells below the 200 mg/L discharge limit. During
September, 1987, a period of intermittent air stripper operation, the total
TCE removal was greater than 94%. During January, 1988, a period
of continuous air stripper operation, the average total TCE removal was
99.8% or more, and TCE concentrations in the water discharged to the
sewer system were consistently below 10 mg/L. Flow quantities from
the air strippers ranged from 50 to 100 gpm which did not exceed the
sewer system flow capacity at any time.
Flows from Municipal Well No. 1 were not discharged to the sewer
during the January 1988 operation of the air strippers. However, it is
anticipated that even with a combined discharge from Municipal Well
No. 1 and the air strippers, the sewer capacity would not be exceeded
at any time except possibly during heavy rainfall events. During heavy
rains, groundwater pumping could be temporarily halted to prevent
surcharging the sewers.
Background TCE concentrations in the POTW influent ranged from
nondetectable to 10 mg/L. During the period from August, 1987 to
January, 1989, when groundwater from Municipal Well No. 1 and treated
groundwater from the air strippers were discharged to the sewer system,
TCE concentrations in the POTW influent ranged from non-detectable
to 29 mg/L. Ninety-five samples were collected during this period and
the number and the percentage of samples which exceed selected TCE
concentration ranges are shown in Table 8.
TableS
Influent POTW Flows
TCE
Concentration
Range
(ppb)
21 30
11 - 20
ND 10
Total
Number of
Samples
12
22
61
95
Percent
(Z)
13
23
64
100
Cumulative
Percent
(Z)
100
87
64
Sixty-four percent of the samples did not exceed the highest TCE
background level of 10 mg/L and 87% were below 20 mg/L. Only 36%
of the samples exceeded the background levels. Peak TCE concentra-
tions in the POTW influent appeared to increase slightly during
discharge from either Municipal Well No. 1 or the air strippers.
Background TCE concentrations in the POTW effluent were consis-
tently below the detection level of 2.0 mg/L, which is the NPDES dis-
charge limit for TCE measured once every 6 mo. During the period
when groundwater from Municipal Well No. 1 and treated groundwater
from the air strippers were being discharged to the sewers, 95 POTW
effluent samples were taken. In general, the POTW effectively removed
TCE to below detection limits. Four effluent samples did exceed the
detection limit of 2.0 mg/L. The TCE concentrations in these four
samples were 2.2, 3.8, 4.0 and 6.1 mg/L. Possible reasons why these
FATE 143
-------
samples exceeded detection limits were examined.
Three of the POTW effluent samples above 2 mg/L were obtained
between Jan 18 and Jan 30, while the air strippers were being operated.
Groundwater from Municipal Well No. 1 was not being discharged to
the sewers at the time. During this period, POTW influent TCE con-
centrations ranged from 8.1 to 28 mg/L. However, air stripper effluent
TCE concentrations ranged from non-detectable to 6.3 mg/L. In
Figure 5, it can be seen that air stripper effluent TCE levels at the time
were below POTW influent TCE levels, which suggests that other
sources of TCE may exist in the sewer system.
Air samples taken downwind of the POTW aeration basin contained
TCE concentrations at or below detection limits which indicated that
this basin was not a significant source of volatile emissions. TCE and
other volatiles were detected in the one sewer manhole sampled. The
concentrations of these volatiles were below the time-weighted average
for normal workday exposure of SO ppm established by the American
Conference of Governmental Industrial Hygienists (ACGIH).
CONCLUSIONS
The sewer system and the POTW were evaluated for potential con-
veyance and treatment of TCE-contaminated groundwater generated
during remedial activities at the site. Wastewater in the sewer system
was found to contain TCE at concentrations which decreased with
distance from the site. This decrease could not be attributed directly
to either dilution or volatilization. Air measurements did not indicate
any volatile chemicals in sewer manholes above background levels.
The sewers had extra capacity to convey remedial flows along with
normal wastewater flows, except for periods of heavy rainfall. The
POTW influent was found to contain TCE in concentrations ranging
from non-detectable to 10 mg/L, and the POTW effluent TCE concen-
trations were below the detection limit of 2 mg/L. The POTW had
adequate hydraulic capacity available to treat the increased flows.
The City agreed to allow the discharge of groundwater to the sewer
system and the POTW provided the fluids met the pretreatment limits
established by MDNR. TCE concentrations in groundwater from
Municipal Well No. 1 became stabilized between 40 to 80 mg/L during
extended pumping. These concentrations were below the 200 mg/L
discharge level and thus did not require treatment. During continuous
operation, the air strippers reduced groundwater TCE concentrations
from onsite wells by an average of 99.8%. Air stripper effluent TCE
concentrations normally were below 10 mg/L.
Peak TCE concentrations in the POTW influent appeared to increase
slightly during discharge from either Municipal Well No. 1 or the air
strippers. However, the TCE concentrations in 64% of the POTW
influent samples were below the background TCE level of K) mg/L.
The highest TCE concentration detected in the POTW influent was
29 mg/L. The POTW effluent TCE concentrations usually were below
the detection limit of 2 mg/L. Only in four of the 95 samples did effluent
TCE levels exceed the 2 mg/L detection limit concentration.
TCE levels in air were measured downgradiem of the POTW aeration
basins. During discharge from Municipal Well No. I or the air stripper
operation, TCE concentrations were at or below detection limits and
the aeration basins did not appear to be a significant source of volatile
emissions. TCE levels also were measured at one sewer manhole during
discharge from the air strippers. TCE was detected in the manhole air;
however at a concentration less than the level established by ACGIH
for normal work day exposure.
REFERENCES
I Geraghty & Miller. Kfmedial Investigation Report, (Final Draft), June 1989
2. Black & Vcaich, Feasibility Study. July. 1989
3. Geraghty & Miller, Pilot Program Statement of Work Remedial Investi-
gation/feasibility Study, (Draft) Jan 20. 1989.
4. Hood-Rich. Architects and Consulting Engineers, Hbstevater facilities Plan
for Green County, Missouri, Prepared for the Green County Sewer District,
Feb. 1984
5. Hanna. S., US. EPA, Cincinnati, OH, personal communication K>J. Sandino,
Black & Vsatch, July 13. 1987.
6. Grady, C P.L., Jr., Biodegradation of Hazardous wastes by CoDveffiJonil
Biological Treatment, Haz. Wastes and Haz. Mat., 3, pp. 333-365, 1986.
7. Lue-Hing, C.,elai. Effects of Priority Pollutants on the Disposal of Sludges
from Publicly Owned Treatment HbHu. Report No. 5-K), Dept. of Research
&. Development. Metropolitan Sanitary District of Greater Chicago, tt, 1981
8. Richards, D.J. and Shien. W.K.. Biological Fate of Organic Priority Mili-
tants in the Aquatic Environment, Htaer Res., 20, pp. 1077-1090. 1986.
9 Roberts. P.V., Munz. C and Dandliker, P.. "Modeling Vblatile Organic Sob*
Removal by Surface and Bubble Aerator," JWPCF. 56, p. 157. 1984.
10. Russell, L.L., Cain. C.B, and Jenkins. D.I., Impact of Priority Pollutants
of Publicly Owned Treatment Works Processes: A Literature Review,
Proceedings of the 37th Industrial Waste Conference, Purdue University,
Lafayette, IN, pp. 871-833, 1983.
11. Unger. M.T. and Claff. RE. Evaluation of Percent Removal variability
for Priority Pollutants in POTW's. Proceedings of the 40th Industrial Mae
Conference. May 1985. Purdue MiiwrjiA. Lafayette, IN, pp. 915-924, B85.
12. U.S. EPA. fate of Priority Pollutants in Publicly Owned Treatment Worts,
EPA 440/1-82/303, Sep. »82.
144 FATE
-------
Predicting the Fate and Transport of
Organic Compounds in Groundwater
Roger L. Olsen, Ph.D.
Camp Dresser & McKee, Inc.
Denver, Colorado
Andy Davis, Ph.D.
PTI Environmental Services
Boulder, Colorado
ABSTRACT
The rate of migration and the concentration of hazardous chemicals
in ground water is a major factor in determining potential extent of
migration, in performing risk assessment and in designing remedial
actions. To assess the rate of migration and concentration of chemicals
in ground water requires a thorough understanding of the geochemical
behavior of the hazardous chemicals in soil water systems. Organic
chemicals can undergo a variety of reactions in the subsurface including
hydrolysis, oxidation/reduction, volatilization, adsorption, and biodegra-
dation. The importance of each of these processes in effecting the fate
and transport of chemicals depends upon the site conditions and the
specific chemical compounds of concern. Generally, adsorption and
biodegradation are the major reactions effecting chemical transport in
ground water.
Adsorption can be evaluated and predicted using eight methods. These
include:
Use of empirical field data
Methods based on K^
Methods based on water solubility
Methods based on molecular structure
Methods based on surface area
Laboratory methods
Field column devices and injection tests
Methods based on plume location
Several of these methods require only minimal site data that can be
easily obtained. As many of the methods as possible should be used
depending on data availability and on the purposes of the prediction.
For example, laboratory studies may be necessary when a quantitative
prediction of desorption is needed to design a treatment plant in terms
of concentration and design life. In all cases, the prediction should be
compared to actual site data.
Of the processes which control mineralization of organic compounds
in the subsurface, biodegradation is the most important mechanism in
transforming short chain halogenated compounds in an anoxic environ-
ment, and in breaking one8 and two'ring compounds under aerobic con-
ditions. The reaction rates of these processes have been defined for both
laboratory and field conditions and are usually modeled using the power
rate law or the hyperbolic rate law.
Modeling contaminant transport in the subsurface relies on a large
body of site specific data including that required to represent adsorp-
tion, biodegradation and dispersion of the compound of interest.
Examples discussed include the lateral migration of trichloroethene and
benzene in ground water, percolation of tetrachloroethene through the
unsaturated zone, and volatilization of trichloroethene from the ground
water surface followed by adsorption in the overlying soil profile. Where
appropriate, adsorption and biodegradation are included in each
simulation.
FATE 145
-------
Geostatistical Decision-Making Process
For Plume Modeling In Cadillac, Michigan
Kevin A. Kincare
Michigan Department of Natural Resources
East Lansing, Michigan
Steven M. Aulenbach
GEDA Systems, Inc.
Boulder, Colorado
ABSTRACT
A geostatistical block model of plume geometry provides a sci of
powerful decision-making tools for enforcement and remedial design
at sites of environmental contamination. These three-dimensional block
models are most effective when geologic, hydrogeologic and historic
information are incorporated into the modeling of these complex sites.
The geostatistical modeling was part of an overall program to define
zones of contamination and aid in a design of remedial measures for
the industrial park. Data collected from various hydrogeological
investigations during the last 30 yr were examined to create boundary
conditions for the geostatistical evaluation. These data included
MDNR~ and PRP sponsored investigations for the contamination
problems that began to surface in 1978.
The MDNR investigation produced 795 vertical data points over a
period of three mo. These data were subjected to quality control
measures and then explored for patterns that might relate to the under-
lying contaminant hydrogeology. A representative and geologically
realistic block model was built with geostatistical techniques. The Final
block model was verified by further field sampling. Thorough
exploration of the sample data and an understanding of the geologic
setting yielded conservative, defensible kriged estimates of contami-
nation that were used as an enforcement tool. The results of the geo-
statistics provided insight into source location and further data needs.
INTRODUCTION
The city of Cadillac is located in northwestern lower Michigan in
Wexford county (Fig. 1). The Cadillac Industrial Park is in the north-
west corner of the city. Various facilities within the industrial park have
been under investigation since the first private wells were found to be
contaminated in 1978. During the subsequent 10 yr, eight sources of
groundwater contamination have been discovered within an area of just
over 0.5 mia a. The contamination and the proximity of the municipal
water supply have won the industrial park two spots on the
CERCLA/SARA NPL as well as funding from the Michigan Environ-
mental Response Act (Act 307, 1982).
In 1986, out of concern for the water supply, the Michigan Depart-
ment of Natural Resources began a program to define known and sus-
pected plumes of contamination. The major pollutant in the park is
trichloroethene. A chromium plume has been defined at a separate NPL
site. Over 180 wells have been drilled in an area of approximately 0.5
mi2 by contractors employed by the MDNR, the PRPs.
Vertical sampling in 88 of the well borings enabled the MDNR to
collect data with 795 discrete three-dimensional chemical analyses with
which to analyze the contaminant distribution. Using the screened auger
method1, vertical samples were taken between 30 and 180 ft below the
surface. Six contaminant plumes have been defined, one of which, the
Figure I
Wexford Coumv MI
East Plume, directly threatened the city well field.
These data were used to construct a three-dimensional, geostatistical
model of the plume geometry for enforcement and remedial design.
This study focused on the East Plume area. The East Plume data
included 280 vertical samples from 27 wells.
The samples were analyzed in the field with a Photovac 10AK) porta-
ble gas chromatograph. Sample results were reported in total volatile
organic carbon (VOC). Contract laboratory results showed trich-
loroethene to be 97 to 1004 of the VOC field-reported concentration.
Taylor and Serafini' showed the field results to be well correlated (r
= 0.83, n = 48 df) with the laboratory results. Their work demon-
strates that the use of field screening data is applicable for the pur-
poses of this project.
SITE GEOLOGY
Cadillac is situated on a basin of glacial origin (Fig. 2). It is located
at the southern end of the Cadillac outwash plain. This outwash plain
146 MODELING
-------
is hemmed in to the east and south by the Valporaiso Moraine and to
the north and west by the Lake Border moraine. This condition caused
intermittent ponding in the south and east of the outwash plain during
the Lake Border stand. The stratigraphy in Cadillac consists of out-
wash sands alternating with lacustrine clays. The sediments consist of
alternating clays and well sorted outwash sands. Four outwash layers
were described in the 290 ft maximum depth of exploration. The un-
saturated zone is 30 ft thick. The two uppermost clays pinch out in
the southeast half of the industrial park, resulting in the three upper
outwash layers becoming one in the northeast half. The deep clay ap-
pears to be a regional till.
01234 5mlM
tg>-j Lake Border Moraine I | Out Wash
V/ffll Valparaiso Moraine EgSSjj Water
Lacustrine Sand tc Gravel
Port Huron Moraine
Figure 2
Glacial Geology of Wexford County
The lake clays exist as "bowls within bowls," having been formed
by the infilling of the Cadillac basin. Sands filled the basin during peak
glacial melting. The clays were then deposited into successively smaller
lakes during times of low melt or inefficient drainage. Lakes Mitchell
and Cadillac (Fig. 2) are remnants of this process. Five discrete lake
clays have been observed in well logs from the Cadillac area, two of
which extend into the industrial parik. These two lake clays pinch out
in the park on a N30W strike (Fig. 3). The shallow lake clay pinches
out at an elevation of 1265 ft (MSL) and dips to the SSW at a gradient
of 0.02 (Fig. 4). The deep lake clay pinches out at an elevation of 1140
ft and dips to the SSW at a gradient of 0.01. The ground surface eleva-
tion averages approximately 1295 ft. The bottom of the basin is a regional
till clay at an average elevation of 1070 ft. The till lies above an older
outwash sand layer that extends to at least an elevation of 945 ft.
Bedrock elevation is approximately 545 ft.
The presence of sloping clay layers that pinch out in the middle of
the study area makes a complex situation. The bulk of the East Plume
data lay beyond the shallow clay but above the area of the lower lake clay.
HYDROGEOLOGY
The presence of two confining clay layers that both pinch out in the
industrial park make a complex hydrogeologic system, as well. Where
both clays exist, there are three aquifers above the regional till (Fig. 4).
Figure 3
Cadillac Industrial Park
Northern Limit of Glacial Lake Clays
1300 .,
1260.
land surface
1220.
1180.
1140-
1100
1060.
1040 -I . 1 1 . 1 , 3 1
500 1000 1500 2000 2500 3000 3500 40DO
A Distance ft. A'
Figure 4
Cadillac Industrial Park
Glacial Stratigraphy
North of where the clays pinch out, there is only aquifer. The city well
field is screened in an older outwash aquifer below the regional till.
Hence, depending on location, there are from two to four aquifers to
be considered. The groundwater in the uppermost aquifer flows north
to northeast above the upper clay. The regional flow in all other aqui-
fers is toward the northwest. There is a downward vertical gradient in
all aquifers. Therefore, three directions of groundwater flow have to
be considered.
MODELING 147
-------
The shallow aquifer is a water table aquifer above the shallow lake
clay. There are 30 ft of unsaturated sand above this aquifer. The satu-
rated thickness ranges between 5 and 10 ft where the clay pinches out,
to 60 ft at the south end of the industrial park. Groundwater flow in
this aquifer is toward the nearest edge of the clay. In the industrial park.
this causes the flow to vary from north to northeast (Fig. 5). There
is a downward vertical gradient in this and all other aquifers studied
for this project. The decrease in saturated thickness downgradient causes
an increased flow gradient to the north. The gradient increases from
0.0019 to 0.0029. With no change in hydraulic conductivity, this 65%
increase in gradient does not compensate for the 600% loss of saturated
thickness. It is obvious that there must be leakage through the shallow
clay. Lakes Mitchell and Cadillac are the recharge areas for the shallow
aquifer.
Figure 5
Groundwater Flow Directions in the Upper, Intermediate, and Lower Aquifers
The intermediate and lower aquifers lie below the shallow and lower
lake clays, respectively (Fig. 4). Both aquifers flow toward the north-
west. Their recharge area is the high country of the Valporaiso Moraine
to the east (Fig. 2). The saturated thickness of the intermediate aquifer
is approximately 130 ft. The gradient remains constant at 0.001. The
lower aquifer is approximately 40 ft in saturated thickness. Its horizontal
gradient is 0.001. Vertical gradients in these aquifers are also down-
ward. The groundwater discharge area is the Manistee River, 18 miles
to the northeast at an elevation of 810 ft (Fig. 2).
With the absence of the lake clays in the northeast portion of the in-
dustrial park (Fig. 4), all three of the above mentioned aquifers merge
into one. Consider that there are 12 in. of recharge per year to the water
table from precipitation. There is no change in aquifer thickness,
hydraulic gradient or hydraulic conductivity in the northeast part of
the study area. These observations alone lead to the conclusion that
there is leakage through the till clay into the deep aquifer.
The city's wells are screened in the deep outwash aquifer below the
regional till (Fig. 4). This aquifer is at least 110 ft thick; the bottom
confining layer has not been reached by any well. The well field
produces an average of 2.2 mgd with a capacity of 10.5 mgd from seven
wells. Groundwater flow is toward the well field as all monitoring wells
screened there were within the pumping zone of influence. Hydraulic
conductivity for all of the aquifers is in excess of 5 x K)-'-3 ft/sec.
A pump test performed on the city well field concluded that leakage
through the till contributed a significant portion of the well field pum-
page. As much as 5% of the pumpage is coming through the till where
it is overlain by contaminated portions of the upper aquifers (Fig. 6).
When this situation came to light, it was deemed an emergency. Im-
mediate steps were taken to further define the vertical and horizontal
extent of contamination for remedial design.
Figure 6
Location of VOC Plumes
The geostatistical study, therefore, had to consider the geological and
hydrogeological parameters outlined above. That is, three directions
of groundwater flow, movement through and around the confining layers
and leakage within the influence of an active well field.
EXPLORATORY DATA ANALYSIS
The East Plume data were thoroughly examined prior to the geo-
statistical estimation. Exploratory data analysis techniques were used
to identify patterns in the sample data. These patterns were compared
to the current hydrogeologic visualization of site conditions. Exploring
the data provides both a quality check on the data and a reality check
on the modeling process itself. Insight gained during this step can be
rapidly incorporated into the geostatistical block model to produce
superior estimates.
The data were first explored as a single collection of measurements.
A quantile plot was used to provide a picture of the distribution of the
data. A plot of the ordered sample values against their reported VOC
concentrations, the quantile plot highlights several patterns and
groupings (Fig. 7). Roughly a third of the observed samples were below
detection level or traces. These coded values are plotted in the lower
left hand corner of Figure 7. Such values serve to bound the kriged
estimates in the final three-dimensional block values.
The other three groupings (1 to 4 ji/L, 5 to 10000 n/L. and > 10000
may represent factors involved with the introduction and trans-
148 MODELING
-------
6.00 q
-4.00
0.00
0.20
0.40
0.60
'o'.so'
DATA
FRACTION OF
Figure 7
Quantile Plot of log 10 VOC Values
port of the contaminants. These factors include intermittent dumping,
rainfall passing through contaminated soils, the descent of dense non-
aqueous phase liquids (DNAPL) through the aquifer, hydrogeology and
glacial stratigraphy. Each of these "post-source" sources can produce
a different distribution signature in the aquifer.
The two extreme values found at the upper left of the quantile plot
represent two outliers, (observations that seem to lie too far from the
majority)2'3. An order of magnitude greater than the other samples,
these concentrations were measured in the same boring. Their impact
on the spatial continuity between samples was investigated during the
varicography phase of the process.
Vertical aspects of the East Plume data also were examined during
this stage of the process. Samples values were collected and displayed
using their elevation. Values above 1 mg/L were grouped according
to their vertical location within the three-dimensional block model. The
model consists of multiple levels. Each level is 10 ft thick. Levels are
numbered from the top of the model down. Level 1 starts at the ground
surface.
The box plot display of the Iog10 VOC samples by levels showed an
informative pattern; a cyclic pattern in median values (Fig. 8). The
median is the horizontal line within each box. Two local highs are clear
in this display with local maximums occurring in levels 8 and 14. This
pattern may indicate a vertical clumping of high VOCs within the East
Plume study area. VOC behavior can be explored further by examining
how the length of the box changes by level (same figure number). The
spread of the bulk of the data, the central 50 %, is shown by the length
of the box. Note how spread varies with depth. The middle levels exhibit
much less spread in their raw data values than the upper and lower levels.
u
§
Block Model Level
Figure 8
Box and Whisker Plot of East Plume Block Model by Level
The short box length of levels 10 and 11 also presents evidence that
two of the sample values, the two extreme values mentioned above, may
be out of the ordinary for these levels. These two, represented as squares
in the display, fall outside of the bulk of the data for their levels. The
box plot was also used in the interpretation of the varicography.
VARICOGRAPHY
The next step of the geostatistical decision-making process is to quan-
tify the spatial relationships that exist between sample pairs. This quan-
tification process is done with the variogram, a basic tool of geostatistics.
The variogram provides key information for the actual estimation
process, kriging.
By successively using each sample as a datum, the sample variance
for all predetermined intersample distance categories is calculated. Then
the distance (x-axis) vs. variance (y-axis) plot, the variogram, is drawn.
When the variance is calculated from data that fell within certain angular
windows from the datum, quantitative changes in the trend with direc-
tion can be determined. Figure 9 shows the ideal form of a spherical
model variogram4. The plot begins at the origin and rises until it
reaches a maximum variance, the sill (C), where the variance remains
constant for greater distances. The sill of the average variogram will
be equivalent to the total sample variance. The distance at which the
sill is attained is the range. At distances greater than the range, the rela-
tionship between samples no longer is influenced by distance.
distance
Figure 9
Idealized Spherical Model Variogram
If the plot has a Y-intercept greater than zero, that value is called
the nugget (C0). If the modeled variogram does not pass through the
origin, it is indicative of a high degree of variability over short dis-
tances. This anomaly can be the result of laboratory error sampling
error or the intrinsic microvariability of the environment itself.
Being directional in nature, the variogram is also an excellent tool
for investigating possible anisotropic conditions at the site. Several types
of variograms were calculated to verify if different measuring scales
showed consistent patterns. These were the general relative and indi-
cator variograms. By looking at the East Plume data from several
different viewpoints, consistency was built into the final block model
estimates.
The protocol for modeling the variograms was to first calculate and
model general relative variograms for different directions. Indicator
variograms using a median cut were used to temper the general rela-
tive ranges. Indicator variograms use the "cut" value as the datum by
which they compare all other values. Such indicator variograms are
more resistant to extreme values and thus provide a second, conserva-
tive estimate of the range.
An average general relative variogram which used all available data
pairs was used to estimate the nugget (C^ and the structured variance
(C,) (Fig. 10). Variograms were then calculated and modeled on the
four cardinal directions; azimuths 0, 45, 90 and 135 degrees. Vario-
grams were for azimuths 22.5, 67.5 and 157.5 to better investigate
anisotropic conditions.
MODELING 149
-------
40.00
«> 32.00
c
D
> 24.00
16.00
8.00 c
Co
Direction: 0.0 General Relative Model
Window: 180.0
-------
in bringing out trends that might not otherwise be quantifiable. It also REFERENCES
suppresses felse trends that might appear as a result of using techniques , Taylor T w and Serafmi, M. C., "Screened auger sampling: the tech-
that are not supported by the geologic model. This process, therefore, nique and two case studies," Groundwater Monitoring Rev, S(3), pp. 145-152,
yields a more reliable variogram with which to construct the kriging 1988.
estimation ellipse. The modeling methods used in this study gave con- 2. Chambers, J. M., et al., Graphical Methods for Data Analysis, Duxbury
servative, defensible, kriging estimates that were used as a tool for Press, Boston, MA, 1983.
enforcement and remedial design 3. Gilbert, R. O., Statistical Methods for Environmental Pollution, 2nd ed.,
John Wiley and Sons, New York, NY, 1986.
4. Clark, I., Practical Geostatistics, Elsevier, New York, NY, 1979.
MODELING 151
-------
Using a Three-Dimensional Solute Transport
Model to Evaluate Remedial Actions
for Groundwater Contamination at the
Picatinny Arsenal, New Jersey
Donald Koch
Engineering Technologies Associates
Ellicot City, Maryland
Ira May
U.S. Army Toxic and Hazardous Materials Agency
Aberdeen, Maryland
Thomas A. Prickett
Thomas A. Prickett and Associates
Urbanna, Illinois
Joseph Murphy
U.S. Army Toxic and Hazardous Materials Agency
Aberdeen, Maryland
Peter Mattejat
Engineering Technologies Associates, Inc.
Ellicott City, Maryland
INTRODUCTION
The Picatinny Arsenal is located in Morris County, New Jersey
approximately 4 mi northeast of Dover (Fig 1). The installation, officially
known as the U.S. Army Armament Research Development and
Engineering Center, performs research on munitions and weapons.
Recent investigations have found trichloroethylene (TCE) and other vola-
tile organic solvents in groundwater. The metal plating shop in
Building 24 has been identified as a possible source of contamination.
TCE and other solvents were used in decreasing operations at this metal
shop.
The contaminant plume was found in the water table (top) layer of
a three-layer aquifer system with some evidence of minor amounts of
contamination in lower layers. The objective of the study was to site
wells for a proposed remedial action plan that included the pumping
of contaminated groundwater and treatment in an air stripping tower.
One important question was whether the wells in the water table layer
of the aquifer system would effectively control gradients in the lower
aquifers and stop contamination that could be in the lower aquifers from
migrating off-site. To answer this question, a three-dimensional solute
transport model was used.
Groundwater modeling is a powerful tool that may be used to predict
contaminant transport at hazardous waste sites One and two-
dimensional groundwater flow and solute transport models are used
to predict contaminant transport. There arc situations, however, where
a three-dimensional simulation capability is necessary There are a
number of well-known groundwater flow •models with three-dimensional
capability, such as the USGS Modular Three-Dimensional, Finite-
Difference Ground-Water f-low, Model (MODFLOW)1 and the
Prickett-Lonnquist Aquifer Simulation Model (PLASM3D), but there
are relatively few three-dimensional solute transport models.
A new three-dimensional solute transport model. RAND3D was
developed as part of this project. The RAND3D model is a solute trans-
port model utilizing the random-walk algorithm. A preprocessor code
(PREMOD3D) was written to use the output of the MODI-LOW model
as input and create files of velocity vectors for the RAND3D model.
The RAND3D model runs interactively on an IBM PC while displaying
Figure 1
Location of Picatinm Arsenal
152 MODKI.ING
-------
the progress of the plume graphically. Figure 2 shows the conceptual
relationship between the data and computer models used on this project.
hydrogeologlc data
(llthologlc data)
(well water levels)
(punp test data)
1
MDDFLDW
USGS nodutar three dlnenslonal
finite difference flow nodel
heads
i
PREMDD3D
Fortran program written to
conpute velocities from MDDFLDW
output
solute data
(decay rate)
(retardation)
(sources)
velocities
RAND3D
three dbtenslonat, randon walk
solute transport nodel
ICS Of
concentrations
aquifer data
(porosity)
(dlsperslvlty tensor)
(screens)
grid
Figure 2
Conceptual Relationship Between Models and Data
RAND3D MODEL
The RAND3D program is a three-dimensional version of the random
walk algorithm developed by Thomas Prickett, et al., at the Illinois
Water Survey as an efficient algorithm for solving groundwater solute
transport problems2. The model originally was developed for two-
dimensional solute transport. Thomas A. Prickett and Associates
developed a three-dimensional version of the model, and further modifi-
cations and improvements were made to the model as part of this project.
The random-walk technique is based on the concept that dispersion
in porous media is a random process. A particle, representing the mass
of a specific chemical constituent contained in a defined volume of water,
moves through an aquifer with two types of motion. One motion is with
the mean flow (along streamlines determined by finite differences), and
the other is random motion (governed by scaled probability curves
related to flow length and the longitudinal and transverse dispersion
coefficients). A sufficient number of particles are included in simula-
tions so that their locations and density, as they move through a flow
model, are adequate to describe the distribution of the dissolved con-
stituent of interest. Each particle represents a fixed mass of solute. As
more particles, with correspondingly smaller masses, are used in a given
simulation, accuracy improves.
One major feature of the RAND3D model is its interactive opera-
tion on an IBM PC or compatible microcomputer. After velocity files
are prepared using PREMOD3D or some other suitable procedure, the
user may use this program to simulate solute transport and watch the
results on the monitor. The program operates from a menu. The user
is prompted for all data inputs. A major feature of the model is the
ability to display geographic features on the computer screen and su-
perimpose the plume simulation. The user may zoom in on any area
of the model to see a more detailed simulation. The geographic fea-
tures are input by the user in any convenient right-handed (x-y) coor-
dinate system in ft (such as a State Plane coordinate system). These
features may then be displayed on the screen as background reference
for the plume simulation.
The RAND3D model includes the following features:
• Calculation of horizontal advective transport based on a four point
interpolation of the input velocity vectors
• Calculation of vertical advective transport based on linear interpo-
lation between the input vertical velocity vectors at the top and bottom
of each layer
• Calculation of dispersion using constant dispersivities: longitudinal,
transverse and vertical
• Calculation of first-order decay
• Calculation of linear, reversible adsorption (retardation)
• The ability to originate solute (particles) in the model as sequences
of prisms, cylinders, or lines
• Calculation of solute concentrations exiting the model at sinks (wells
or gaining streams)
• Mapping of solute concentration in user selected areas of the model,
either plan view or cross-section concentration maps may be prepared
• Output of gridded solute concentrations by layer for plotting
• Interactive operation
• On-screen display of plume (particle) movement in user selected area
• On-screen display of user input geographic features at user selected
scale as background for the plume display
• Saving and viewing of screen slides
• Saving and restart of model parameters at any time
• Transient flow simulations may be simulated by inputting a series
of velocity files
The RAND3D model was designed for an IBM PC or compatible
microcomputer with 640K, a numeric co-processor, a hard drive and
a color monitor with a color graphics adapter. The program is written
in Microsoft Quick Basic Version 3.0. Current limits in the program are:
• Maximum input grid of 45 columns, 45 rows and three layers
• Maximum number of particles is 10000
• Maximum number of sinks (wells or gaining streams) is 99
• Maximum number of special feature files is 20
SITE GEOLOGY AND HYDROGEOLOGY
The study area is located in the drainage basin of Green Pond Brook,
a tributary to the Rockaway River. The Rockaway River flows into the
Boonton Reservoir, a water-supply reservoir for Jersey City. Green Pond
Brook runs through the middle of the Picatinny Arsenal.
The Picatinny Arsenal is located in the Green Pond syncline, a struc-
tural region within the New Jersey Highlands physiographic province.
The New Jersey Highlands is comprised of a northeast-southwest system
of folded and faulted Proterozoic to Devonian rocks that form a sequence
of valleys and ridges. The Green Pond syncline is a narrow, northeast-
trending, faulted syncline containing a thin section of Paleozoic sedi-
ments. Bedrock at the site consists of gneiss, quartzite, dolomite and
conglomerate. The bedrock is overlain by approximately 200 ft of glacial
deposits. The glacial deposits are stratified, consisting of sublacustrine
sands and gravels, lake-bottom silts and deltaic sands and gravels3.
Groundwater flow at the site generally follows the topography; ground-
water flows towards Green Pond Brook and down valley. Vertical
gradients are downward except around Green Pond Brook where there
is some upward movement of groundwater.
FLOW MODEL CALIBRATION
The USGS MODFLOW model was used to simulate the groundwater
flow at the site. The groundwater flow system at the site was repre-
sented as a three-layer model: the first layer was the water table aquifer
in the permeable glacial sediments near the land surface; and the second
layer was the confined glacial aquifer. The third layer was the frac-
tured limestone and dolomite underlying the glacial sediments. A 35
column by 43 row grid was defined as shown in Figure 3. The model
was calibrated to the existing observation well data assuming steady-
state conditions. Figure 4 shows the water table in the upper layer of
the model generated from the calibrated model. Observation well water
levels were compared to water levels predicted by the model. The average
error across 41 wells was 0.12 ft and the root mean-square error was
1.76 ft.
MODELING 153
-------
Figure 3
Model Grid
ALTERNATIVE ANALYSIS
Six different pumping schemes for remedial action were simulated using the
three-dimensional solute transport model, RAND3D. The first alternative
analyzed was the no action alternative The existing position of the plume was
input to the model and the movement of the plume towards Green Pond Brook
was simulated for 60 yr. After 5 yr, 60% of the contamination had entered Green
Pond Brook.
The first active remedial action scheme simulated was a group of three col-
lection wells, each pumping at 36 gpm. located to create a hydraulic barrier
between Building 24 and Green Pond Brook. These wells were input to the
MODFLOW model, the new water table was simulated and u set of velocity
flies that reflect the transient conditions in the aquifers wus created for input
to the RAND3D model. After 6 yr of pumping, 91% of the TCE has been
removed; 88% by the wells and 6% by the stream.
The third remedial action scenario simulated was the group of collection wells
plus injection wells. Four injection wells were assumed to be placed on the
upgradient site of Building 24 (the assumed source of the plume). Each collec-
tor well was assumed to be pumped at a rate of 72 gpm, twice that used in the
collector well scenario. Of the 216 gpm to be treated. 200 gpm would be in-
jected back into the water table aquifer. Letting some treated water discharge
to surface water insures that the system as a whole (total of pumping and injec-
tion) causes a slight depression in the water table, so if the assumptions are
incorrect, contamination will still remain in the area, rather than being pushed
away faster than it would without injection. Each injector well would recharge
50 gpm. After 6 yr of pumping, 98% of the TCE has been removed; 94% by
the collection wells and 4% by entering Green Pond Brook.
The fourth alternative simulated was using three collection wells and discharging
the treated water to Bear Swamp Brook, which is upgradient of Building 24
and the contamination plume. The assumption was that by increasing the flow
and depth of flow in Bear Swamp Brook, the recharge to the water table aquifer
would increase. The results of this simulation indicate that this alternative is
not significantly different from the three collection wells with discharge of treated
water to Green Pond Brook. Infiltration to Bear Swamp Brook was small.
The fifth alternative simulated was using four collector wells. The first three
wells were at the same positions as in the other pumping alternative simula-
tion* (equally spaced row between Building 24 and Green Pond Brook). A fourth
well was placed adjacent to Green Pond Brook, where substantial concentra-
tions of TCE had been measured in the water table aquifer. This fourth well
was assumed to be pumped at a rate of 44 gpm for I yr and then turned off.
After 6 yr of pumping, 94% of the TCE has been removed; with 92% by the
wells and 2% by entering the stream.
The sixth and final alternative simulated was variable pumping at the three
collector wells By pumping more from wells in the middle of the TCE plume
and less from wells at the edges, it was hoped that the overall efficiency of the
collection and treatment system would increase. Pumping more water from the
wells at early times would also capture more of the contamination that is between
the collection wells and the stream. The collection well in the middle of the
plume assumed to be pumped at a rate of 80 gpm for the first year, 60 gpm
for the second year and 54 gpm for the third year. One well at the edge of the
plume was assumed to be pumped at a rate of 18 gpm and the collection well
on the other side of the plume was assumed to be pumped at a rate of 36 gpm.
After six years of pumping, 95% of the TCE has been removed; 92% by the
154 MODELING
-------
1000'
I I I I
x Figure 4
Calibrated Steady State Water Table
Top Views
Cross-Section Views
After 60 days
After Two years
After 12 years
After 12 years
Figure 5
RAND3D Screen Graphics
wells and 3% by entering the stream.
Figure 5 shows several of the screen graphic displays generated by the RAND3D
program for the variable pumping scenario. The first column of results shows
the a top view of the plume after 60 days of pumping, after 2 yr of pumping
and after 12 yr of pumping. The second column of results shows the corresponding
cross-sectional views of the aquifer system.
Collection Wells with Variable Pumping
A sensitivity analysis was performed on the model predictions. The
parameters with the greatest amount of uncertainty that also had a sig-
nificant impact on the simulation results were the retardation coeffi-
cient (adsorption), the amount of TCE still leaching from the unsaturated
zone to the water table aquifer over time and the amount of TCE ad-
sorbed in the confining clay beds between aquifers. The sensitivity
results indicate that the cleanup (pumping and treatment) could extend
for more yr than predicted. With the maximum reasonable retardation
coefficient, 90% cleanup would take approximately 15 yr. With a
reasonable worst case scenario for TCE leaching into the water table
aquifer from recharge, 90% cleanup would take more than 20 yr as
TCE is continuously entering the aquifer. The collection of TCE that
may be trapped in the confining layer takes even longer. Assuming that
TCE is trapped in the confining layer near Building 24, after 50 yr
of pumping and treatment, 50% of the TCE is still present in the aquifers
and confining layers.
CONCLUSIONS
A practical model for simulating three-dimensional solute transport
in groundwater on an IBM PC has been developed. This model uses
the groundwater flow results of the MODFLOW model and simulates
solute transport using the random-walk algorithm. The model operates
interactively and generates graphic displays of plume movement as the
MODELING 155
-------
simulation takes place.
The results of the modeling of the Building 24 TCE plume at the
Picatinny Arsenal indicate that there is no clearly superior pumping
design for cleaning up the contaminated groundwater and preventing
TCE from reaching Green Pond Brook. All of the simulated scenarios
that do not include recharge wells upgradient of Building 24 achieve
similar long-term removal rates. Recharge wells would speed the removal
of TCE from the aquifer, but effective recharge wells may not be feasi-
ble because of a shallow water table and the likelihood of injection well
clogging. The pumping plans that remove groundwater from the aquifer
rapidly collect more TCE from the water table aquifer faster. All the
collection well scenarios simulated effectively formed a barrier to the
movement of TCE towards Green Pond Brook. Placing collector wells
closer to Green Pond Brook would effectively collect more of the TCE,
reduce the amount entering Green Pond Brook, but result in larger
amounts of pumpage containing lower concentrations of TCE.
REFERENCES
1. McDonald. M.G and Harbaugh, AW., A Modular Three-Dimensional Fmiie-
Difference Ground-Water How Model, Techniques of Water-Resource Inves-
tigations, of the United Stales Geological Survey, Washington. DC, 1988.
2. Prickctt, T.A., Naymik, TO., and Lonnquist, C.G., A Kandom-Hblk Solute
Transport Model for Selected Croundwaler Quality Evaluations, Illinois Water
Survey Bulletin 65, Champaign, IL.1981.
3. Sargent. B.P, Fusillo, T.V., Storck, D.A., and Smith, J.A., GroundwaurCon-
tamination in the Area of Building 24, Picatinny Arsenal, New Jersey, review
draft. United State-. Geological Survey, Trenton. NJ, 1988.
156 MODELING
-------
Statistical Modeling of Ambient Air Toxics Impacts
During Remedial Investigations at a Landfill Site
Louis M. Militana
Steven C. Mauch
Roy F. Weston, Inc.
West Chester, Pennsylvania
ABSTRACT
At landfills or other waste disposal sites, the off-site impacts due to
air toxics generated by intrusive activities are a principal concern. To
assess these impacts, the multivariate statistical technique of canonical
correlation has been applied to ambient air toxics sampling data col-
lected during a remedial investigation of a landfill in the metropolitan
area of Los Angeles, California. The goal of the analysis is to deter-
mine whether a site activity produces significant ambient air toxics
impacts in the area immediately downwind of the site.
Canonical correlation analysis of the data collected at the downwind
site reveals that the primary physical process occurring is dilution of
contaminants by wind, with secondary slight increases in contaminant
levels primarily due to boring activities. Although the canonical models
are not strong enough for quantitative predictions for this data set, they
do provide a realistic qualitative analysis of the physical situation.
INTRODUCTION
This paper presents the results obtained from application of canonical
correlation analysis to ambient air toxics sampling data collected down-
wind of a landfill site during RI activities. Canonical correlation is a
multivariate statistical technique that can be used to evaluate the rela-
tionship between groups of variables, in this case, meteorological
conditions, site activities and ambient air toxics levels. The technique
is an extension of traditional multiple regression analysis, which seeks
to relate a single variable to a group of other variables.
Canonical correlation was chosen as an analytical tool because of
its ability to provide information beyond the scope of traditional statistical
comparison techniques, such as simple tests for equality of means or
multiple correlation. The use of multivariate methods allows better
resolution of the complex interactions between the atmosphere and the
variety of air toxics compounds that may be present due to intrusive
activities on a landfill site.
SITE DESCRIPTION
The site chosen for this study was an urban landfill located in the
Los Angeles, California area. Historically, the site was used for dis-
posal of general construction-type debris, but petroleum wastes and
solvents also were potentially disposed of there. The site investigation
was prompted by plans for new construction over the landfill site.
The site is located at the intersection of two major thoroughfares,
with the upwind sampling location near the intersection. The down-
wind air sampling site was located beyond the northeast corner of the
landfill area. The heavy automobile traffic around the site had a definite
influence on the sampling results, particularly at the upwind site.
DATA COLLECTION
Ambient ah" samples were collected at the upwind and downwind
locations during the 3-wk site investigation. Wind speed, direction and
air temperature data were collected concurrently with the sampling.
Due to the consistent land-sea breeze circulation pattern at the site, day-
time winds were most frequently from the southwest. The wind rose
for the site activity period is shown in Figure 1. The upwind and down-
wind sampling locations were therefore the same for all samples and
were chosen based on this wind pattern.
•a . 79 2JOS
SCALE (KNOTS)
MIND SPEED (KNOTS) PER
0-3 3-6 HO 10-
N I.B2 D.IG 0.00 0.
WE 2.06 0-00 O.DD 0
NE 3.67 0,00 0 00 0
ESE 2. B6 0.16 0 00 0.
CENT 0
S
CCURRENC
1B-JI
.00
00
.00
00
E
.00
.00
00
S9E 1.76 Q.4B 0.00 0.00 0.00 0.00
KINO SPEED (KNOTS) PERCENT D
It-J M 6-10 10-16
5 3 83 0 64 000 000
SSH 9 91 1.28 0,00 0 00
SM 4.13 t! 02 1, SB 0 00
WNW 4 47 0 80 0 00 000
CCURRENC
UMI
00
00
.M
E
>21
.00
.00
.00
.00
NNH 0.96 0.00 0 00 0.00 0.00 0.00
Figure 1
On-site Wind Rose Activity Period
During the activity period, 31 high-volume air samples and 33 vola-
tile organics samples were collected. The compounds detected included
eight toxic volatile organic compounds (VOCs), copper, lead, zinc and
asbestos. The following eight VOCs were detected in at least 75 % of
the samples;
acetone
benzene
ethylbenzene
styrene
toluene
xylenes
tetrachloroethene
1,1,1-trichloroethane
MODELING 157
-------
The VOCs were collected using passivated stainless steel canisters
(U.S. EPA Method TO-14) and the rnetals were analyzed from high-
volume air samples of paniculate matter. Asbestos was determined using
low-volume personal pumps and filter cartridges.
Site activity was parameterized as the durations of the two principal
intrusive activities: boring (soil core samples) and drilling (ground-
water monitoring wells). Activity durations were obtained from the site
log books.
In addition to the activity period data, background samples were col-
lected on the 3 days immediately prior to the start of intrusive site ac-
tivities. The mean contaminant levels in these three samples arc used
to establish a benchmark level to assist in interpreting the activity period
results.
SAMPLING RESULTS
For the activity period, a complete range of descriptive statistics was
calculated for the upwind and downwind locations. The statistics include
the average, standard deviation, standard error, maximum, minimum,
median and the 25th and 75ih percentiles (lower and upper quartiles).
These statistics were used to generate the box-and-whiskers plots
presented in the next section, used for upwind vs. downwind
comparisons.
Mean concentrations of air toxics measured during the activity period
are compared to mean background concentrations for all contaminants
in Figure 2 for the solid contaminants and Figure 3 for the volatile
organics. Concentrations of the solid contaminants were higher during
the activity period than background means in all cases except for the
downwind lead and upwind zinc levels. For the VOCs, all concentrations
levels were higher than corresponding background levels. The concen-
trations of acetone, styrene, toluene and xylenes during the activity
period were markedly higher than their background levels at both
locations.
350
50
Upwind Location
Activity Period
Downwind Location
Activity Period
Upwind Location
Background Period
Downwind Location
Background Period
40
30
20
10
-
•
\
I
1
Up Upwui
iH Activt
Hm Upwin
Hii Backgi
SSS uowni
^ Activi
>^W uowni
^ffi Backgi
d Location
y Period
id Location
round Period
.vind Location
ty Period
rind Location
round Period
L I
^
r
n
'• n
in
Figure 2
Comparison of Activity and Background Levels
For Solid Contaminants
Benzene j Slyftiw j Toluene | Xytonu
Acetone EBiytxnztne PCE TCA
Figure 3
Comparison of Activity and Background Levels
For Volatile Organic Contaminants
UPWIND/DOWNWIND COMPARISONS
To assess the amount of contamination introduced into the ambient
air by site activities, a comparison of upwind and downwind means
can be used. Normally, a I-test for equality of means would be used
in the comparison. However, the t-lest assumes that both samples are
normally distributed. This is not a reasonable assumption for the air
toxics data being considered, as they tend more toward a log-normal
distribution. Rather than performing the log transform to "force" the
data to become normal, a nonparametric comparison of medians was
performed using the Wilcoxon two-sample test for independent sam-
ples. None of the upwind/downwind pairs of medians were significantly
different at the 10% level.
The upwind/downwind sample sets also were compared using side-
by-side box-and-whiskers plots. A box-and-whiskers plot (or simply a
box plot) is designed to display the distribution of a sample and allow
visual comparison of samples. The plot consists of a box bounded by
the 25th and 75th percentiles (lower and upper quartiles) of the data
set, with a bar between them indicating the 50th percentile (median).
The "whiskers" extend from the quartile edges of the box to die maxi-
mum and minimum values in the sample. In addition to these standard
features, a circle has been added at the sample mean, along with an
error bar stretching one standard error unit above and below the mean.
This addition allows visualization of the skew of the distribution and
allows easy comparison of means in side-by-side plots.
Interpretation of the plots is straightforward. A normally distributed
sample would have a perfectly symmetric plot, with the mean and
median collocated at the center of the box. Skewed samples have means
above or below the median and disproportionate whiskers.
Side-by-side box plots were used to compare upwind and downwind
sample sets. Figure 4 shows the data for all of the solid contaminants
and Figures 5 and 6 show the VOCs in two groups. The plots are
generated from the descriptive statistics for the activity period only.
158 MODELING
-------
90
75
60
45
30
15
0
: i
• .
i 1
•
(
j t
\
P
UP DOWN
Location
Asbestos
700
600
500
400
300
200
100
! ! !
.....]
i
,
I
H
\
\
UP DOWN
1
0 15°
1
nanograms per
8
111;
i
... j
i
I
1
I
UP DOWN
350
300
250
3
nanograms per
S
50
! ! !
1
<
r
i
T
III!
UP DOWN
Location Location Location
nanograms per cubic meter
fo *» o> co
08888
! !
p
. . . . . .
1
(
?!
i
» :
UP DOWN
Location
TSP Copper
Figure 4
Box Plots for Activity Period: Solid Contaminants
Lead
Zinc
These figures effectively portray the difficulty in determining whether
the downwind contaminant levels are in any way distinctly greater than
the upwind levels, using normal "yardsticks" such as means or medians.
Using these methods of comparison would lead to the conclusion that
site activities had no distinguishable impact on contaminant levels, with
the apparent exception of acetone. However, the broad overlap of the
box plots for acetone indicates that the two samples are not meaningfully
different (a conclusion supported by the Wilcoxon test).
DOWNWIND CANONICAL CORRELATION
The primary goal of this analysis was to determine the potential
impacts of site activity on nearby downwind (i.e., off-site) locations.
Therefore, the remaining discussion in this paper is limited to the down-
wind data collected during the activity period.
The canonical correlation procedure was performed based on the
correlation matrix for all contaminant variables. The northerly and
easterly wind speed components were mean values covering the period
from 0700-1700 L each day. The values of these variables for each day
are tabulated on Table 1. Note that the vector wind components are
directed to the north and to the east. Thus, a northeast wind would
have both components negative, while a southwest wind would have
both components positive.
Canonical correlation extends the sample correlation concept from
two single variables to two sets of variables. The two sets are analo-
gous to the dependent and independent variables in traditional regres-
sion analysis. The canonical correlation procedure finds the most highly
correlated pairs of linear combinations of the variables in each set. These
linear combinations are known as canonical variable scores and the
sample correlation between a pair of scores is the canonical correla-
Table 1
Daily Mean Wind Components
Day
1
2
3
4
5
6
7
8
9
10
11
12
13
North
fmph)
2.74
3.28
3.30
2.93
2.33
1.39
2.28
1.02
2.11
1.32
0.66
0.64
1.48
East
fniDh)
0.68
3.05
0.33
3.27
3.75
2.41
2.31
5.62
3.13
4.93
4.73
3.85
4.33
tion coefficient. The scores may be interpreted by examining the
component variables6 sample correlations with the resultant score.
Solid Contaminants
A summary of the results of the downwind canonical correlation
analysis for the solid contaminants is shown on Table 2. The first two
pairs of canonical variates are significant at the 10% level. The corre-
lations of the two pairs of scores with their component variables are
MODELING 159
-------
ft
i
1
(
r
,
-
i
1
0
1
1
I
s
!
3
r
UP DOWN
Loot on
Acetone
UP DOWN
Location
Benzene
1 1 1 1
'
(
j
y
L
*
r
UP DOWN
12
I I
I
I
UP DOWN
Locator Locate*
Tetrachloroethene (PCE) Trichloroethane (TCA)
Figure 5
Box Plots for ACIIUIN Period: Acetone, Bctucnc, PCE and TCA
shown in Table 3. In Table 3 and in all subsequent tables including
sample correlations, coefficients significant at the 10% level are flagged
with a (<) symbol.
Table 2
Summary of Canonical Correlation Results
Downwind Solid Contaminants
Number
1
2
3
4
Canonical
Correlation
0.9715
0.9027
0.7421
0.4159
Significance
Level
0.0069
0.0951
0.3274
0.5145
Based on the significant correlations, the first pair of vanatcs reflect
low copper levels occurring with southerly winds. Considering all the
correlations, the general relationship expressed by the first pair of scores
is lower contaminant levels and more southerly winds, which is con-
sistent with site geography. A stronger easterly wind component is neces-
sary to carry contaminants towards the downwind site. Therefore, the
southerly component would contribute to transport away from the site
(dilution.) Therefore, the first set of canonical variates appears to
represent the general reduction of contaminant levels at the downwind
site by dilution.
The second pair of variates reflects higher lead levels and longer
boring periods, based on the significant correlations. In general, the
relationship is between higher contaminant levels and longer boring
Table 3
Downwind Correlations of Canonical Variates
With Component Variables
Contaminant Scores
Asbestos
TSP
Copper
Lead
Zinc
First
Pair
-0.427
-0.170
-0.51K
-0.199
0.122
Second
Pair
-0.048
0.228
-0.190
0.669<
0.467
Wind/Activity Scores
North
East
Boring
Drilling
First
Pair
0.704<
-0.061
-0.209
0.383
Second
Pair
0.248
-0.427
0.939<
-0.150
times. There is also a relatively high correlation in the activity/wind
score with westerly winds. As previously discussed, westerly winds
(high easterly components) are primarily responsible for contaminant
transport to the downwind site Therefore, the second pair of variates
160 MODELING
-------
'
i
,
1
h
r
i
\
32
28
24
20
12
•
4
_
j
i
-i
i
.
*.....
\
i
-
r
1
i
i>i
6
2
i
<
i
i
1-1
n
1
1
i
i
1
<
•"I
,
1
I r
<
i-
r
i
UP DOWN
Location
Ethylbenzene
UP DOWN
Location
Styrene
UP DOWN
Location
Toluene
UP DOWN
Location
Xylenes
Figure 6
Box Plots for Activity Period: Ethylbenzene, Styrene,
Toluene and Xylenes
represents the general elevation of contaminant level at the downwind
site during site activity with more westerly 7sea breeze8 flow regimes.
The logical extension of this analysis would be to attempt to predict
the quantitative effects of varying levels of site activity on contaminant
levels. Constructing such a model would require establishing a solid
relationship between the variables and the scores. Unfortunately, the
correlations are too weak to be of predictive value. However, the
canonical correlation analysis does indicate that elevated contaminant
levels are qualitatively associated with increased boring activity.
Volatile Organic Contaminants
The canonical correlation analysis is summarized in Table 4, with
one significant pair of variates indicated. The correlations in Table 5
show no contaminants significantly correlated to the contaminant score.
this type of ambiguity occurs in canonical correlation analyses
whenever there are strong correlations between many variables in either
group. Such a high degree of correlation does exist amongst many pairs
Table 4
Summary of Canonical Correlation Results Downwind Volatile
Organic Contaminants
TableS
Downwind Correlations of Canonical Variates
With Component Variables
Contaminant Scores
Acetone
Benzene
Ethylbenzene
Styrene
PCE
Toluene
TCA
Xylenes
First
Pair
0.196
-0.218
0.013
-0.023
0.018
-0.082
-0.287
0.086
Number
Canonical
Correlation
Significance
Level
1
2
3
4
1.0000
0.9934
0.7811
0.7272
0.0000
0.1675
0.8137
0.6406
Wind/Activity Scores
First
Pair
North 0.716<
East -0.640<
Boring -0.329
Drilling 0.159
MODELING 161
-------
of VOCs, principally due to the influence of nearby traffic emissions.
The use of highly correlated predictor variables in linear regression
produces an analogous effect.
More definitive results might be possible if some of the highly
correlated contaminant variables were eliminated. Such an elimination
of variables would be arbitrary, based on available data and so was not
attempted.
CONCLUSIONS
In the interpretation of ambient air sampling data collected during
field investigations at a landfill, "traditional" statistical comparisons
(e.g., comparison of means) may fail to reveal meaningful relationships
between site activity and resulting air contaminant levels. This short-
coming is due to the inability of single-variable statistics to account
for the more subtle interactions often present in air toxics sampling.
The use of a multivariate technique such as canonical correlation allows
a more detailed examination of the interrelationships among sampling
variables.
When applied to a set of actual ambient air toxics data collected during
activities at a landfill, comparisons of the upwind and downwind samples
using box plots and Wilcoxon two-sample tests for equal medians did
not reveal any significant increase in contaminant levels. This was due
mainly to the wide variability inherent in the data.
Canonical correlation analysis of the solid contaminant levels and
the activity/wind variables at the downwind site shows that:
• There is primarily dispersion of contaminants across the normal sea-
breeze wind direction (southwest)
• Boring duration and elevated levels of metals are positively related
These canonical relationships are not strong enough for quanti-
tative use.
Canonical correlation analysis of the VOC data at the downwind site
arc rendered indeterminant due to a high degree of inter-correlation
among the volatile contaminants. These interrelationships are due mainly
to traffic on the thoroughfares bordering the site, which likely obscures
any relationships between VOC levels and site activity.
The overall conclusions of the canonical correlation analysis of
ambient air toxics sampling during remedial investigations at this site
may be summarized as follows:
• On-site activities resulted in slightly elevated concentrations of copper,
lead and zinc in the ambient air downwind (northeast) of the she
• The increases in levels of these contaminants are not statistically
significant
• No significant increases in toxic VOC levels were linked to site activity
The use of canonical correlation to analyze results from other air
sampling efforts in support of RI/FS operations proved useful in this
case and may prove to be of even greater use in the future.
162 MODELING
-------
RANDOM-WALK Modeling of Organic Contaminant
Migration from the Theresienfeld Landfill Located
in the Vienna Basin Aquifer of Austria
Richard Rudy, P.G.
Ecology and Environment, Inc.
Tallahassee, Florida
Gerald Strobel, RE.
Ecology and Environment, Inc.
Buffalo, New York
Wolfgang Widmann
ILF Consulting Engineers
Innsbruck, Austria
ABSTRACT
The Theresienfeld landfill, located 30 km south of Vienna, Austria,
contains several thousand drums of waste solvent material mixed with
municipal garbage. The landfill operated from 1972 to 1985. Since the
early operating periods at the facility, waste solvents have leaked and
leached into a highly productive aquifer of the region. As part of a feasi-
bility study/design process for remediation of this landfill, a limited
groundwater modeling effort was undertaken to provide a general under-
standing of continual contaminant migration and a relative prediction
of future solute transport under two scenarios: with remediation of the
landfill materials and without remediation of these materials.
This waste facility is situated in the southern portion of the Vienna
Basin, which is a large, elongated, trough-shaped depression created
by classic horst and graben tectonics. Basin sediments consist of Tertiary-
age clay deposits overlain by 100 to 150 m of Quaternary gravel inter-
mixed with thin silt and clay lenses. The unconfined aquifer within the
gravel deposit is a principal future groundwater resource for Vienna.
General hydraulic conductivity values for this aquifer are in the range
of 10~2 to 10~3 m/sec. Groundwater flow velocities within the aquifer
range from 6 to 20 m/day along the longitudinal axis of the graben
structure.
In the 1960s, gravel mining in this area was prominent. At this par-
ticular location, mining operations resulted in a large pit with dimen-
sions of 100 m wide by 750 m long by 20 m deep. Early in the 1970s,
mining was no longer profitable, and this site was sold to a local entre-
preneur who began using the pit as a landfill, but without the use of
contaminant prevention techniques such as liners and leachate control.
Leachate from the landfill has since seeped into the relatively fast-
flowing Vienna Basin aquifer and resulted in organic contamination of
several nearby downgradient monitoring, residential and industrial wells.
Specific contaminants of the plume consist primarily of chlorinated
hydrocarbons including trichloroethylene, tetrachloroethylene and
1,1-dichloroethane at total concentrations of approximately 500 to 1000
ja/L immediately downgradient of the landfill.
In addition to this landfill, there are several other contaminant sources
that likely are impacting aquifer ground water quality. Highly indus-
trialized cities such as Wiener Neustadt and Ternitz are situated upgra-
dient of the landfill. In particular, industrial facilities such as old steel
mills in these two cities likely have contributed to groundwater con-
tamination in the main recharge area of the aquifer.
Groundwater modeling consisted of evaluating average flow and solute
transport conditions in the general basin area surrounding the landfill
using the Analytical RANDOM-WALK Model. This model was used
to assess two-dimensional flow conditions under a finite difference
formulation, while integrating solute transport from particle-in-a-cell
for convective effects and random-walk techniques for dispersion in a
porous medium as a random process. This particular code was selected
for this study because of limited time constraints, limited data base and
the model's ability to simulate two-dimensional mass transport problems
in homogeneous/isotropic aquifers under steady8state water table con-
ditions. Thus, although the results of this study are at best qualitative,
they do provide a general and relative indication of long-term impacts
on the aquifer.
In this modeling effort, flow parameters and contaminant loads were
determined based on chronological assumptions and best available data
in the general basin area with dimensions of 23 km by 46 km. Three
general sources of contamination were incorporated in the model: the
landfill as a point source, and two areas upgradient of the landfill as
line sources to simulate existing and continual inflowing contaminated
groundwater. These conditions were then calibrated to the most current
data set to best simulate the actual contaminant plume extent in two
dimensions as it currently exists.
The calibrated modeling simulation showed clearly that remediation
of the landfill source immensely improved groundwater quality of the
aquifer. However, low level contamination slightly above Austrian
drinking water standards would persist in much of the aquifer without
remedial action on assumed upgradient sources in the cities of Wiener
Neustadt and Ternitz. Without remediation of the landfill, the plume,
as originally estimated, would become more concentrated with the
various contaminants and would increase in extent to impact a much
larger area of the groundwater resource.
INTRODUCTION
The Theresienfeld landfill is a large, uncontrolled hazardous waste
facility located in a rural area about 30 km south of Vienna, Austria
(Fig. 1). Environmental effects from this landfill resulted in national
publicity and a government investigation of this and several other waste
disposal facilities in the area.
The Theresienfeld site is situated in the Vienna Basin, one of the
major present and potential future groundwater sources for Vienna. A
few years ago, several communities installed drinking water produc-
tion wells at a location downgradient of the landfill. To the dismay of
these communities, the wells contained concentrations of chlorinated
solvents ranging from 20 to 30 /*g/L throughout the central axis of the
Vienna Basin.
As a result of detectable contamination in the water production wells,
the Austrian government began monitoring the basin aquifer and detected
a widespread problem. The Theresienfeld landfill, as well as other con-
taminant sources from Ternitz and Wiener Neustadt, industrial areas
hydraulically upgradient of Theresienfeld, seemed to be contributing
to the problem. Ternitz and Wiener Neustadt have been heavily indus-
trialized since before World War II.
MODELING 163
-------
I 3 -\ /. \. •
•
THERESIENFELD
LANDFILL
SOURCE: Gaologitche Bundaunitilt, Witn und Umgcbung, 1:200,000
SCALE
10
15 MILES
10
20 KILOMETERS
. . ^r- Vicniu
Innibrudi c O
AUSTRIA
l-i^ure I
Locution nl ihc Thcrcsicnfcld LanclHIl in (he Vienna Basin
164 MODELING
-------
§
D
DTheresienfeld
»_ Kr chenacker
SOURCE: Osterreichische Karte; Wiener Neustadt (1:25,000)
SCALE
1 MILE
.5
1 KILOMETER
O
Figure 2
Local Theresienfeld Landfill Area
-------
THERESIENFELD
LANDFILL
SOURCE: Geologischt Bundeunttah, Witn und Umgebung. 1:200,000
SCALE
0 5 10 15 MILES
10
IS
KEY:
••B160MM Water Table Elevation (meteri above jea level)
20 KILOMETERS
Innsbruck
1-igurc .1
Generalized Vienna Basin Aquifer Wuicr Table Isolinc
(source' Bergcr 198.1)
166 MODELING
-------
The purpose of this modeling effort was to illustrate a simplified rela-
tive impact on the aquifer of a total cleanup operation implemented
at the landfill and to determine if the aquifer can recover to a useful
condition if the contamination source is controlled. Generalized two-
dimensional groundwater modeling results are presented under two
scenarios that yield insight on potential future horizontal contaminant
plume migration: (1) the landfill is left in its current condition, and
(2) the existing contaminant source (landfill materials) is removed.
In order to simulate the contaminant plume, sources of contamina-
tion upgradient of Theresienfeld were simplified and taken into account.
These sources generate the base load contamination of the upgradient
groundwater, which is superimposed onto the load from the Theresien-
feld landfill.
For this study, only the horizontal plume migration was evaluated.
Furthermore, this evaluation was performed utilizing the best availa-
ble data, which are limited in amount. However, the results demon-
strate that the groundwater system will recover slowly if the source is
controlled.
This two-dimensional model provided reliable results because the
aquifer in this area is fairly homogeneous with increasing depth to the
impermeable zone and vertical head differences likely are not signifi-
cant. Because of limited data and funds, a vertical contaminant predic-
tion was not performed.
BACKGROUND
Mined-out pits evidently end up being used for deposition of waste
materials all over the world. This was the case at Theresienfeld.
Beginning in 1966, an extensive gravel mining facility existed at this
location. Sometime in the late 1960s, the gravel production business
experienced dwindling prices, and by 1970 many gravel producers in
the area shut down operations. The excavation at Theresienfeld was
one such operation and resulted in an elongated pit about 100 m wide
by 750 m long by 20 m deep.
In 1972, the owner leased the Theresienfeld gravel pit to a nearby
paint and solvent manufacturing and recycling plant. The paint and
solvent plant obtained a permit from local authorities to dispose of
drummed waste materials in the pit. According to the permit, all
drummed waste was to be deposited in layers, with 20 cm of fill material
placed between each layer. However, the landfill operation was initiated
without any contaminant migration prevention technology, such as im-
permeable liners, leachate collection or gas control systems, or even
security. Later, the new operator also obtained permission to dispose
of wastes from other industries in the pit. Within 1 yr after initiation
of the landfill operation, inspection authorities observed pooled
chemicals in various areas of the facility. The pools apparently were
the result of haphazard drum disposal or the dumping of free liquid
into the excavation. Improper disposal practices at the landfill continued,
and numerous problems were documented. Over the remainder of the
decade there were several chemical fires at the landfill. The operator
began to accept waste materials such as paint and solvent residue sludges
and shredded rubber, metal and manufactured items. Used oils were
spread over the shredded materials, apparently under the assumption
that they would serve as a sorbent.
By 1980, the landfill had begun to accept household wastes. By this
time, nearly half of the pit was full. In addition, approximately 200
drums were illegally buried in trenches at the bottom of the unfilled
half of the pit. Figure 2 is a map of the Theresienfeld area showing
the filled and unfilled portions of the pit.
HYDROGEOLOGY
The Theresienfeld area is situated in the central portion of the trough-
shaped depression known as the Vienna Basin (Fig. 3). In Austria, the
basin is approximately 60 km long and in the Theresienfeld area approxi-
mately 10 km wide. The basin extends in a north-northeast direction
from Neundirchen to the Danube River near Vienna. The basin was
formed by classic horst and graben tectonics. The associated structural
movement has resulted in down-thrown Tertiary- and Quarternary-aged
rock materials which make up the graben and upthrown Jurassic- and
Tertiary-aged units along the basin's flanks. Faulting has occurred deep
in the Tertiary sediments within the basin, perpendicular to the graben
structure. In the area of Theresienfeld, the result of this faulting
phenomenon is a sub-basin known as the Mittendorfer Senke, which
is one of four major sub-basins within the larger Vienna Basin.
In general, the stratigraphy of the Vienna Basin consists of Quar-
ternary gravel deposits which overlie Tertiary clays, clay marls and con-
glomerates. This gravel deposit was transported and deposited by the
numerous rivers that drain into the basin from the surrounding upland
terrain. The gravel formation is fairly homogeneous, with localized clay
and sand lenses sporadically located throughout. In the Vienna Basin,
the gravel varies in thickness from 3 m to 150 m. In the Mittendorfer
Senke, the gravel is approximately 100 m thick.
A thick Tertiary clay deposit exists below the gravel and extends
throughout the entire Vienna Basin. This formation consists of a blue-
gray clay intermixed locally with limestone fragments and layers of sand.
Most of this clay is impermeable. The average clay thickness in the
basin is believed to be approximately 300 m. The interface between
the gravel and clay is very distinct, and the depth of this boundary is
highly variable within the Vienna Basin. In the Theresienfeld area, the
clay surface is estimated to range in depth from 25 m in the area south
of Theresienfeld, to 100 m in the immediate vicinity of Theresienfeld,
to 25 m in the area north of Theresienfeld.
The unconfined, very permeable aquifer which resides in the gravel
formation of the Vienna Basin is believed to be one of the best fresh
water resources in Europe. The aquifer is recharged primarily by the
Schmarzau and Piesting Rivers which flow out of the mountainous
region located southeast and east of the Vienna Basin. These rivers,
along with several others to the south, are also the main surface water
features in the basin. All of the surface water eventually drains into
the Danube River which transects the basin at the northwest end.
The depth to the water table in the Theresienfeld area is approxi-
mately 20 m below the surface. The regional groundwater flow direc-
tion of this aquifer is generally north-northeast, parallel to the
longitudinal axis of the basin. In the Theresienfeld area, the local
groundwater flow direction is also northeasterly. Dye tracer tests per-
formed by local authorities have shown that flow rates in this aquifer
in the Theresienfeld area are very rapid, ranging from 6 to 10 m/day
near Theresienfeld, and increasing toward the recharge area at the south-
west end of the basin to as much as 20 m/day.
This surficial aquifer supplies drinking water to most of the popula-
tion of the Vienna Basin area, except for Vienna. The city of Vienna
currently receives its water via aqueducts from the uplands to the south-
west. However, the demand for water at Vienna is exceeding the ca-
pacity of current supply sources, and it is anticipated that the city will
have to begin drawing on Vienna Basin groundwater sources in the near
future.
In 1982, based on periodic sampling of several wells throughout the
Vienna Basin area, local authorities directed the Theresienfeld landfill
operator to sample and analyze groundwater from nearby existing wells
for chlorinated hydrocarbons and metals. As a result of this sampling
and regional ongoing investigations, the authorities concluded that the
landfill was contributing substantial contamination to the groundwater
with trichloroethylene, perchloroethane, toluene and 1,2-dichloroethane
being the major chemicals of concern.
During the period from 1982 to 1985, the Austrian government
installed a number of monitoring wells around the landfill perimeter
and sampled these wells and nearby private wells on several occasions.
The drilling installation methods and construction details of these wells
are unknown.
Based on the available groundwater sampling data, a lateral contami-
nation plume was defined (Fig. 4). In addition, in 1984, soil gas sampling
was performed in the landfill area by a German firm. The soil gas results
were quite similar to the groundwater sampling results. Thus, the data
clearly indicate that significant contamination is leaching from the land-
fill to the groundwater aquifer and migrating downgradient.
MODEL DESCRIPTION
A semi-quantitative mass transport model known as Analytical
RANDOM-WALK1 was employed for this study. The flow portion of
MODELING 167
-------
o
D
m
z
o
SOURCE: Water Authority of Lower Austria
KEY:
SCALE
10 25 u
-------
this model is based on groundwater velocities actually measured in this
basin aquifer.7 Expanded regional solutions for groundwater flow were
obtained analytically in this model based on an average of these
velocities.
Aquifer parameters were discretized by superimposing a finite dif-
ference grid over a map of the basin aquifer area. This grid covers a
large area downgradient of the site with equidistant x and y axis lengths
of 23 km by 46 km, respectively. This area was further divided into
100 distinct cells with dimensions of 2300 m by 4600 m (Fig. 4). Input
parameters for this particular model include both flow and solute trans-
port conditions. The final parameters used for this project are listed
in Table 1. The following discussion explains how these final conditions
were developed.
Table 1
Groundwater Model Input Parameters
English
Metric
Aquifer thickness (saturated) 264 ft 80 m
Hydraulic conductivity 2,846 gpd/ft2 1.34 x 10'3m/sec
Porosity (dimensionless) 0.3 0.3
Regional flow velocity 19.7 ft/day 6 m/day
Transmissivity 747,000 gpd/ft 1.06 x Kr'nrVsec
Storage coefficient (dimensionless) 0.2 0.2
Retardation (dimensionless) 10.0 10.0
Longitudinal dispersivity 60.0 ft 18.3 m
Transverse dispersivity 10.0 ft 3.05 m
Annual Mass Loading Rate 280,000 Ib 27,170 kg
Flow Parameters
The parameters for the groundwater flow portion of the model con-
sisted of transmissivity, hydraulic conductivity, storativity and flow
velocity obtained from the available data base and personal interviews.
The flow field parameters were based on local research sources7 and
showed that groundwater in this area commonly flows at 6 m/day in
the general area of the basin. In this simplified modeling effort, only
a single flow trend is used for mass transport conditions. Hydraulic
conductivity (K) was obtained from average field measurements made
during the November, 1985 geotechnical investigation of the empty west
pit area. Transmissivity (T) was obtained by simply calculating the
product of K and the estimated average saturated thickness of the aquifer
(80 m). As expected, this groundwater system has a relatively large
transmissivity. The storage coefficient was assumed at a value of 0.2,
which is typical for an unconfined aquifer.4
Solute Transport Parameters
Parameters for the solute transport segment of the model were
determined based on limited background information with regard to
total chlorinated hydrocarbons as the primary leachate at the landfill.
In the development of the solute transport parameters, a typical land-
fill leachate formation process has been assumed to occur. Under this
assumption, once the leachate begins to form by precipitation recharge,
it migrates slowly downward through the landfill where physical, chemi-
cal and biological forces act upon it. Eventually the leachate reaches
saturated strata, where it moves as defined by the hydraulic flow velo-
cities. From this point on, the leachate concentration will decrease due
to several phenomena, including dilution, filtration, sorption, microbial
degradation and dispersion. General input solute transport parameters
to simulate the process described above in this modeling effort consist
of contaminant mass flow rate, retardation and dispersion.
The leachate leakage rate was determined in a two-step process and
based on the equation:
QCo = Leachate mass/year
Where: Q= Source area X recharge rate
Co= Initial concentration, (^ig/1) of Leachate
In this equation, the source area and recharge rate were obtained from
the available data as 35,000 m2 and 716 mm/yr, respectively; thus, Q
equals 68.7 mVday. Co has not been measured except for very high
concentrations detected in samples collected from excavated drums. The
drum concentrations, however, cannot be directly used because once
the material leaks from the drums, mixing and dilution occur from
recharge wastes. Therefore, the leachate concentration was first esti-
mated based on the solubility of trichloroethlyene and then adjusted
to simulate concentrations that had been measured in the nearby wells.
Hence, an annual input of 78,972 kg 7pounds8 was used as a mass con-
taminant loading rate into the model. This acontaminant loading, in
turn, corresponds to an initial leachate concentration of approximately
3,146 mg/L in the landfill.
Retardation is used to represent the change in the solute concentra-
tions of the groundwater caused by chemical reactions within the aquifer
matrix. These reactions include absorption, organic fixation, chelation,
etc. Chemical reactions between the dissolved components and the
aquifer matrix have a tendency to retard contaminant movement rela-
tive to groundwater movement. The retardation value for this model
was determined by using the following equation (Gabarini and Lion
1986):
V
R =
= 1 +
(2)
Where:
Vc ne
R= Retardation (dimensionless)
Vo= Groundwater Velocity
Vc= Contaminant Velocity
Pb= Subsurface Bulk Density
nc= Effective Porosity
Kd= % Organic Carbon (KJ
Koc= Soil/Substrate-Water Partition Coefficient for the
Chemicals of Concern (chlorinated hydrocarbons)
and Normalized to the Substrate's Organic
Carbon Content
Dispersion of the leachate in the groundwater system essentially causes
the contaminant concentrations to decrease with increasing distance of
flow. It is caused by a combination of molecular diffusion and
hydrodynamic mixing. Dispersion can be both longitudinal and trans-
verse, the net result being a contaminant plume with a general conic
configuration downgradient from the continuous pollution source. The
contaminant concentrations are less at the margins of the cone and
increase in the middle toward the source. For this modeling, disper-
sion is input as dispersivity, the difference being that the dispersion
includes velocity.
The dispersivity parameters for this modeling evaluation were first
obtained from literature values given for a groundwater system of a
similar gravel aquifer in Loins, France1 and then adjusted by model
calibration using observed groundwater quality data from the landfill
vicinity so that the modeled contaminant levels generally paralleled
actual field measurements. As a result, the model output is most relia-
ble near the source and progressively decreases in reliability with
increasing distance away from the source.
Chronological Assumptions
The final step in setting up the model consisted of developing a
reasonable chronological history of the site based on previous waste
disposal practices at this location. Because of the large area covered
in this modeling evaluation, contaminant loading was performed for
both the continuous and discontinuous source simulations at a point
where the landfill materials currently exist. In both cases, the modeling
simulation began in 1972, which is the approximate date when disposal
MODELING 169
-------
18400
20700
23000
SOURCE: Geo. Bun.. Wlen und Umgcbung (] 2000 000)
KEY
I20BBBB1 Total Chlorinated
Hydrocarbon Itoline 1/4g/l)
Figure S
Modeled Contamination Plume — Year VfTi
(Calibration Phase)
-------
o
o
m
o
METER
2300
4600
6900
9200 Ł.-=:
11500
13800
16100
18400
20700
23000
SOURCE: Geo. Bun.; Wlen und Umgebung (1:2000.000)
SCALE
5
10 MILES
KEY:
I 20 ^H Total Chlorinated
Hydrocarbon Isoline 1/49/1)
10 KILOMETERS
Figure 6
Modeled Contamination Plume — Year 1982 (Calibration Phase)
-------
2
O
D
rn
O METER
2300
4600
69OO
9200
11500
138OO
16100
18400
20700
23000
SOURCE Geo. Bun.. Wlen urn! Umgebung (1:2000.000)
SCALE
5
tO MILES
KEY:
10 KILOMETERS
120MB Total Chlorinated
Hydrocarbon liolme 1/49/1)
Figure 7
Contaminant Plume Prediction Year
-------
operations were initiated.
Sources of contamination were assumed as follows:
• Theresienfeld landfill as a "point source"
• Upstream sources as two "line sources''
— One line source parallel to the groundwater flow to bring into the
model in 1972 the already existing contamination.
— One line source on the upstream boundary of the model, trans-
verse to the groundwater flow, simulating the inflowing contami-
nated groundwater from upgradient sources.
It was assumed that the contamination from upgradient sources re-
main constant at an average of about 30 g/L which is the present situa-
tion. As previously discussed, in 1982 detailed groundwater quality
measurements showed a contaminant plume extending beyond a near-
by well (W-83) located hydraulically downgradient of the site. The
horizontal longitudinal plume configuration is essentially parallel to
the northwesterly groundwater flow direction, which is toward two
recently installed Vienna production wells and the Danube River.
For the continuous source or no-action simulation, it was assumed
that contaminant release would occur indefinitely and predicted to the
year 2040, or 50 yr past 1990, which is the likely realistic point at which
the source would be mitigated. In the second simulation, contaminant
discharge was assumed to end in 1990 following implementation of
source control measures and removal of waste materials.
SOLUTE TRANSPORT SIMULATION
The transport simulation was performed in two phases. First the simu-
lation was calibrated to best parallel the limited data available. Next,
following calibration of the model, two predictions were made to show
what the horizontal plume extent effects are if: (1) the landfill situation
is left in its present state and release of contaminants continues, or (2)
the landfill materials are removed.
Calibration
Model calibration consisted of using the chronology and parameters
previously discussed to best simulate the actual contaminant plume
extent as it currently exists based on the known well data. As a result,
several runs were performed starting with the initial 1972 period in order
to achieve reasonable calibration. In each run of the model, input
parameters consisting principally of contaminant particle mass, retarda-
tion and dispersion were altered in various combinations to a point at
which the modeled contaminant plumes roughly resembled the current
plume configuration. Figures 5 and 6 show the estimated total chlori-
nated hydrocarbon plume for 1972 and 1982, respectively. Figure 7 shows
the predicted modeled plume for December, 1990, which has general
dimensions of approximately 9.2 km long by 4.6 km wide, or a total
approximate area of 42.3 km2. These figures illustrate how the system
was polluted over time and the associated contaminant spreading.
The variety of complex facts that control the movement of leachate
and the overall behavior of the contaminant plume are difficult to assess
accurately within the given data limitations in that the final effect
represents several factors acting simultaneously. Therefore, the illus-
trations shown here, and the predictions of concentration and plume
geometry that follow, are, at best, only to be used as relative estimates
that provide an idea of potential aquifer restoration if the landfill source
is removed.
Contaminant Movement Prediction: Landfill Source Continuous—
(No Action)
The first prediction was based on the premise that the landfill would
be left in its present state, i.e., no action and contaminant leaching con-
tinues. Figures 8 through 10 illustrate the modeled areal extent of the
plume for the years 2000, 2020 and 2040, respectively. Essentially, these
figures represent 10, 30 and 50 yr beyond the present time. Concentra-
tions are shown in /xg/L of total chlorinated organic contaminants.
As~depicted in the figures, the model predicts that the plume evi-
dently contaminates a wide and long strip of the aquifer. The predicted
50 yr contaminant plume extends approximately 19.5 km hydraulically
downgradient of the landfill and attains a maximum width of approxi-
mately 10 km. Hence, a total approximate area of 195 km2 of the
aquifer will be contaminated at a concentration greater than 40 ftg/L
according to this simulation. Thus, from these results clearly show that
as long as nothing is done to control the contaminant source, the resulting
plume will grow larger and impact downgradient resources to a much
greater extent. At the movement rate depicted by this simulation, the
contaminant plume front theoretically could reach the Danube River
area (about 40 km downgradient from the site) in 150 yr.
Contaminant Movement Prediction: Landfill Source Removed
A second prediction was made to show how the contamination will
spread with time if the landfill materials are removed or contained so
further leaching cannot occur. Figures 11 through 13 show the general
modeled areal extent of the plume for the years 2000, 2020 and 2040,
respectively. These figures represent time sequences 10, 30 and 50 yr
following source control of the landfill. Concentrations are shown in
mg/L of total chlorinated organic contaminants.
As shown in the illustrations, the model predicts that the plume will
migrate as a large slug parallel to the predominant groundwater flow
direction as expected. In addition, the 30- and 50-yr predictions show
a slight transverse spreading due to advection and dispersion mechanisms
as the plume migrates further. According to this prediction, the area
immediately downgradient of the landfill will begin to be restored to
natural conditions in 20 to 30 yr, provided that other upgradient sources,
and any heavier-than-water contaminants that have possibly accumu-
lated at the bottom of the aquifer below the landfill, do not exist.
It must be understood that this prediction does not account for what
may occur vertically; the vertical plume spatial extent was not evaluated.
Realistically, the effect of vertical relationships may be important because
certain contaminants, such as trichloroethylene, may exist at high con-
centrations at low areas along the impermeable layer interface of the
aquifer and thus act as additional sources. A verification of segregated
concentration levels with depth is required to consider this condition
further. The given prediction, however, provides a reasonable estimate
for planning purposes. It also should be noted that the given predic-
tions do not take into account additional increased pumpage effects
associated with water demand from the City of Theresienfeld, the gravel
pit operation or any other downgradient production wells as well as
any impact from the nearby canal.
CONCLUSIONS
The Theresienfeld landfill represents Austria's first experience with
hazardous waste site problems and the resulting effect on critical ground-
water resources. In response to the growing concern of how this par-
ticular landfill has and will impact the Vienna Basin Aquifer of eastern
Austria, a limited groundwater modeling study was implemented using
the very flexible and dynamic Analytical RANDOM-WALK code6.
Although the area modeled was very large and the existing data base
limited, through the use of this model, a qualitative prediction was per-
formed to demonstrate on a general scale the overall impact to the aquifer
under two separate scenarios: (1) with remediation of the contaminant
source area and (2) without remediation of the source.
In essence, the two separately calculated scenarios show that by
remediating the Theresienfeld landfill, i.e., total source removal, the
groundwater quality of the Vienna Basin Aquifer will significantly
improve. However, a contamination leachate slug that would still exist
once the source is removed, together widi other regional continuous
contaminant sources upgradient of the landfill (located in Wiener
Nuestadt and Ternitz), would still adversely impact this aquifer, and
contaminant levels will continue to exist in excess of Austrian drinking
water standards unless these sources also are remediated. Hence, this
limited RANDOM-WALK modeling study met an overall objective to
provide the public with a general understanding of existing Theresien-
feld landfill effects on the Vienna Basin Aquifer.
REFERENCES
1. Anderson, M. P., "Using Models to Simulate the Movement of Contaminants
through Groundwater Flow Systems." CRC Crit Rev. Environ Cont Vol 9
1979. ' ' '
MODELING 173
-------
o
o
o
23000 L-
METER
2300
4600
690O
9200 *?•=:
11500
13800
16100
18400
20700
SOURCE: Gao. Bun.. W1«n und Umgebung (I 2000 000)
SCALE
5
10 MILES
KEY:
I20BBB Total Chlorinated
Hydrocarbon Itoline (f4g/\)
10 KILOMETERS
Figure 8
Contaminant Plume Prediction landfill Source C'onnnueouv
(No Atlmn) -Year 2OOO
-------
I
o
tn
O
METER
2300
4600
6900
9200 i
11500
13800
16100
18400
20700
23000
SOURCE: Geo. Bun.; Wlen und Umgebung (1:2000.000)
SCALE
5
10MILES
KEY:
10 KILOMETERS
1 20
Total Chlorinated
Hydrocarbon Isoline
Figure 9
Contaminant Plume Prediction: Landfill Source Continuous-
No Action)—Year 2020
-------
2
O
O
m
H
O
13800
16100 -
18400
20700
23000
SOURCE: Gao. Bun . Wien und Umgebung (1:2000.000)
KEY:
I20MMM Total Chlorinated
Hydrocarbon I to I me (/
-------
1
0
13800
16100
18400
20700
23000
SOURCE: Geo. Bun.; Wlen und Umgebung (1:2000.000)
SCALE
5
10 MILES
KEY:
I 20 •• Total Chlorinated
Hydrocarbon Isoline
10 KILOMETERS
Figure 11
Contaminant Plume Prediction: Landfill Source Removed Year 2000
-------
o
D
m
r-
O METER
2300
4600
69OO
9200 ^^
115OO
13800
16100
18400
20700
23000
SOURCE: Geo. Bun.. Wlen ond Umgcbung (1:2000.000)
SCALE
s
\0 MILES
KEY:
1O KILOMETERS
<&
I 20 •• Total Chlorinated
Hydrocarbon Itoline
l-igurc \2
nniint Plume Prctliclion i.^niintl Source RcnuncO ">c;ir 2O-O
-------
I
w
o
23000
METER
2300
4600
6900
9200 :*Ł.—
11500
13800
16100
18400
20700
SOURCE: Geo. Bun.; WIen und Umgebung (1:2000.000)
SCALE
5
10 MILES
KEY:
10 KILOMETERS
\20mm Total Chlorinated
Hydrocarbon Isoline (/X9/0
Figure 13
Contaminant Plume Prediction: Landfill Source Removed Year 2040
-------
2. Berger, R., Unpublished Report of Groundwater Flow in the Vienna Basin. (12), 1986. 6. Pricken, T.A.. Naymik. T.G and LanquiM. CO., A
Vienna, Austria, 1983. "RANDOM-WALK" Solute Transport Model for Selected Groundwater
3. Brix, F. and Plochinger, B , "Geologic Map of the Republic of Austria," Quality Evaluations, Bulletin 654. Illinois State Water Survey, Champaign,
1:50,000; Plate 76 Wiener Neustadt, Geol. B. "A", Vienna, 1982, II, 1981.
4, Freeze, R.A. and Cherry, J.A., Groundwater. Prentice-Hall, Inc., Engle- 7. Reitinger, Personal Communication from Professor Reilinger, University of
wood Cliffs, NJ, 1979. Vienna, 10 Richard Rudy, Ecology and Environment. Inc. 1987.
5. Gabarini, D.R. and Lion, L.W., "Influence of the Nature of Soil Organics 8. Tollman, A., Geology of Austria, Volume* 1-3, Vienna, Austria. 1977, 1985,
on the Sorption of Toluene and Trichloroethylene." Environ. Sci Tech. 20, 1986.
180 MODELING
-------
Economic Analysis of Public and Private Management
of Remedial Actions
Marc Curtis, RE.
ERM-Southwest, Inc.
Houston, Texas
ABSTRACT
From 1983 to 1989, the two Superfund Projects analyzed in this paper
proceeded from preliminary investigations to construction. One project
was managed by a state agency (public), the other by PRPs (private).
The U.S. EPA provided varying degrees of oversight for both projects.
This paper discusses the magnitude and distribution of costs for the
two projects and compares the project management performance.
Overall, private management controlled construction costs much better
than public management, with bid prices for the Private Project being
$125,000/ac of clay cap, compared to $512,000/ac of clay cap for the
Public Project. However, the Private Project had significantly higher
administrative and engineering costs due to the central role the U.S.
EPA and its oversight contractors played in the remedial process. The
redundant management style imposed on the private project added
$1,425,000 to the design phase and will add up to $1,800,000 to the con-
struction phase.
The U.S. EPA should reduce its oversight of privately funded remedia-
tion projects to a compliance review role. A reduced role would improve
the design phase, the cost-effectiveness of projects and the rate at which
remedial construction is completed.
INTRODUCTION
Private management, with the qualities necessary to succeed in the
competitive marketplace, has the potential to outperform public manage-
ment of remedial actions. Public agencies are organized and staffed
to regulate, not construct, while private parties (industrial corporations)
routinely design, bid and con-struct complex facilities. The most
important quality necessary to manage projects is flexibility, the ability
to develop an approach which continually adjusts to the specific needs
of a project. When Public Agencies are placed in the unfamiliar role
of project management, they are required to apply inflexible procedural
requirements to all projects instead of developing an individual approach
for each project.
This paper compares performance of public and private management
by comparing two Superfund remediation project costs. Underlying
reasons for differences in performance are examined for the two projects.
The comparison is divided into three phases: (1) design, (2) bidding
and (3) construction. The general characteristics of the projects are
discussed first as background for the comparisons that follow.
GENERAL CHARACTERISTICS OF THE PROJECTS
Differences in the size, nature and location must be described before
comparing the engineering and construction activities of different
projects. For these projects there are a number of similar components
in the remedial design which can be directly compared; however, the
private site is remote and weather conditions frequently delay construc-
tion work.
Both projects proceeded in a similar path from preliminary investi-
gations to construction over the time period 1983 to 1989. The Public
Project has a clay/membrane cap which covers approximately 11 ac,
while the Private Project's clay/membrane cap covers 41 ac. Both projects
have slurry walls constructed using the excavated soil mixed with slurry
for backfill. The maximum depths of both walls are 40 ft and the total
square footage of the walls are similar: 100,000 ft2 for the Public
Project and 150,000 ft2 for the Private Project (not all of the Private
Project was contained by slurry walls).
The Public Project is in a major metropolitan area and is readily
accessible to the labor, materials and utilities necessary for construc-
tion. The Private Project is remote, the daily commute to the site is
approximately 40 mi and the site is not served by public telephone,
water or sewer lines.
Annual rainfall at the Public Project site is 42 in compared to 62
in at the Private Project site. The Public Project is occasionally delayed
by rainfall but site drainage is good. Construction schedules are strongly
affected by weather at the Private Project site with flooding and rain-
fall frequently causing extended delays in the work. Contractors, in
determining the cost of the work, considered weather an insignificant
factor for the Public Project and a significant factor for the Private
Project.
COMPARISON OF DESIGN PHASES
The design phase includes all of the work from preliminary site
investigations to final agency approval of the construction documents.
The RI/FS, consent decree, construction plans, specifications, worker
health and safety plan, construction quality assurance plan, operation
and maintenance plan, quality assurance project plan and all other
reports, studies and contract documents necessary to proceed with bid-
ding and contracting for the work are included in this phase. The mag-
nitude and distribution of design phase costs for the Projects are
compared in Table 1. It is evident that the Private Project has many
more types of expenditures than the Public Project. These additional
expenditures include negotiations with the agencies, monitoring U.S.
EPA site activities, U.S. EPA oversight, legal and administrative costs.
Negotiations with the agencies include the engineering and legal work
necessary to develop the comprehensive and detailed agreements
between the U.S. EPA and the private parties for remediation. Negoti-
ation of this agreement, the Consent Decree, added $340,000 to the
cost of the Private Project. There is not a directly comparable cost for
the Public Project as potential responsible parties did not enter into
a Consent Decree with the U.S. EPA.
Within the Private Project there is a broad area of duplicate effort
which includes monitoring, oversight and preparation and review of
duplicate reports. For example, both the U.S. EPA and the private parties
produced RIs and FSs because U.S. EPA's early policies regarding Con-
COST & ECONOMICS 181
-------
sent Decree terms were not acceptable to the PRPs. Duplicate effon
is an apparent significant burden on the Private Project, representing
an additional cost of $1,425,000. However, in the case of this site, il
definitely reduced overall project costs because it prevented U.S. EPA
from unilaterally selecting an overly conservative and more costly
remedy. This duplicate effort is in part a result of the organizational
set-up illustrated in Figure 1 in which the U.S. EPA and its Consul-
tants play a poorly defined parallel management role from beginning
to end of the project. The problems associated with this organizational
setup are aggravated by the high turnover rate in the U.S. EPA and its
oversight consultants. The U.S. EPA project manager and oversight con-
tractor changed three limes during the Design Phase of the Private
Project.
Table 1
Design Phase Costs
PRELIMINARY
Remedial Investigation
Monitoring EPA
Feasibility Study
Past Agency Response Costs
EPA RI/FS Critique
Consent Decree Negotiations
DESIGN
Geotechnical Investigations
Remedial Design
Oversight Review Response
EPA's Oversight
EPA's Oversight contractors
State's Oversight
ADMINISTRATION
Legal
Accounting
TOTAL
P\)t)?ig site
$670,000
0
2)0,000
0
0
0
n Site
$900,000
0
250,000
0
0
0
0
$250,000
$1,150,000
$270,000
67,000
104,000
600,000
43,000
340,000
$1,624,000
400,000
380,000
200,000
70.000
240,000
5,000
$1,295,000
400,000
30,000
"450,000
$3,349,000
STATE )7~| EPA
PRP
V
•
s
1=
f
-------
COMPARISON OF BIDDING PHASES
The Bidding Phase consists of preparing bid documents, bidding the
project and awarding the construction contract(s). Table 2 compares
the bid prices for the Public and Private Projects. The Public Project
has very high mobilization costs compared to the Private Project. The
high mobilization costs for the Public Project are the result of the Con-
tractor's perception that initiating a job with a government agency is
expensive and the specific mobilization requirements specified are more
extensive than those for the Private Project. Table 2 indicates that the
public management team selected off-site disposal as part of the
remediation which represents a large portion of the construction bid.
Excluding off-site disposal, the Public Project construction bid prices
were $512,000/ac of clay cap compared to $125,000/ac of clay cap for
the Private Project. Table 3 compares the unit costs for construction
common to both projects. With the exception of sand, the unit costs
are much higher for the Public Project than for the Private Project.
The higher sand costs for the Private Project are associated with the
remoteness of the site.
Table 2
Bidding Phase Results
Bid Amounts
Construction Costs
Public Site Private site
$4,590,000 $5,140,000
Non-Construction Costs
Bonds
Mobilization/Demobilization
Sampling/Analysis
Administrative Delay
Transportation/Disposal
TOTAL
117,000
730,000
210,000
27,000/day
0
145,000
0
0
10.500.000
$16,130,000
Estimated Construction Phase Engineering Costs
Oversight
Sampling/Analysis
EPA Oversight
TOTAL
GRAND TOTAL
$500,000+
$16,630,000+
$5,285,000
800,000
200,000
1.800.000
$2,800,000
$8,085,000
Table3
Unit Cost Comparison
Public Site Private Site
Clay ($/cu.yd.) 8.00 5.00
60 mil HOPE ($/sq.ft.) 0.50 0.35
Sand ($/cu.yd.) 15.00 16.85
Topsoil ($/cu.yd.) 18.00 2.50
Slurry wall ($/sq.ft.) 3.70 1.50
Seeding ($/sq.yd.) 0.65 0.25
Clearing ($/acre) 4,500.00 1,300.00
Bid prices depend on the Contractor's perception of the project and
the prices suppliers quote for the materials of construction. The Con-
tractor's perception includes how he perceives his competition, clarity
of the contract documents, the unit pricing structure, contract imposed
project overhead costs, contingencies and the working relationship with
the Owner. Material prices depend on how well the specifications foster
competition between suppliers.
Successful bidding requires the management team to do much more
than prepare plans and specifications and put them out to bid. They
must develop a bidding strategy and actively address concerns and
options raised by contractors and suppliers during the bidding process.
The bidding strategy must develop a unit pricing structure which iso-
lates contingencies but comes as close as possible to a single lump sum,
hard money contract. Where unit prices are used, there must be an
accurate method of selecting the quantity to bid and measuring quanti-
ties constructed. This may require a predesign site investigation more
detailed and directed toward the selected remedy than that accomplished
during the RI. An investment in such a predesign data gathering step
usually will return its cost several times over in lower construction costs.
The Bidders should believe that a good working relationship will be
established with the Owner. The management team must develop this
belief through the manner in which the pre-bid conference and other
communications with the bidders are conducted.
The cost differences summarized in Tables 2 and 3 are the result of
specific differences in the bid documents and different bidding strategies.
The Public Project required performance and payment bonds in the
amount of the contract price. These bonds added $117,000 to the cost
of the Public Project. The purpose of the bonds is to provide security
that, in the event the contractor fails, the work will be completed by
the bonding company at the prices bid and subcontractors will be paid
so that liens are not placed on the property.
Public agencies typically require bonds on construction projects, but
the costs often exceed the benefits. If there are many competitive bids
for a project, it is easy to find a replacement for a failed contractor
and the procedure for replacing a contractor is simpler if a bonding
company is not involved. With proper construction management, the
contractor always has completed more work than he has been paid for
at any given time in the contract, and it is not difficult to verify that
subcontractors are being paid. Bonds are expensive insurance for
problems that can be handled effectively through the contractor selec-
tion process and construction management.
The bid strategy for the Public Project resulted in 40, mainly unit
quantity, bid items. The bid strategy for the Private Project resulted
in 12 bid items, half of which were unit quantifies. The success of the
Public Project bid strategy depended on accurately estimating bid quan-
tities and completely defining the work with a large number of specific
bid items. The difficulty in accurately estimating quantities, especially
for hazardous waste work and without detailed site investigations directed
specifically toward the selected remedy unit quantities, leaves the Public
Project open to change orders and scope of work disputes.
The Private Project combined 85% of the work into two lump sums.
Unit pricing was used to isolate contingencies and allow adjustment
of certain field controlled activities where there was a potential for con-
siderable savings if properly managed and small downside risks. The
structure of the Public Project's unit pricing allowed payment of $700,000
for quality control, health and safety and mobilization before the start
of construction. Comparable costs for the Private Project were $120,000,
and this amount was paid after the start of construction.
The Public Project left contingencies in some bid items. For example,
there was only one unit price used to bid the slurry wall, but the specifi-
cations defined several actions the contractor would have to take if
affected materials were encountered. The contractor had no choice but
to assume the worst and include the contingency costs in the one bid
item. This contingency is one explanation for the higher slurry wall
unit cost for the Public Project as shown in Table 3.
The Public Project placed the primary responsibility for QA/QC on
the Contractor. This requirement distorts the engineering costs for the
Public Project by making the QA/QC costs part of the construction costs
and, more importantly, this does not provide independent construction
quality assurance as required by U.S. EPA guidance documents. The
construction quality assurance personnel and organization were indepen-
dent of the Contractor in the Private Project, providing clear definition
of engineering and construction costs and the separation necessary be-
tween the Contractor and QA/QC work.
The Public Project bid documents have many standard requirements
absent in the Private Project. These additional requirements include
pollution liability insurance, a procedure to obtain indemnification which
COST & ECONOMICS 183
-------
includes lengthy documentation, MBE/WBE requiremenis, Davis-Bacon
labor and wage requirements, federal audit, procurement and record-
keeping procedures and the requirement for preparation and submis-
sion by the contractor of several technical plans for quality control,
health and safety and spill control. Private management can be flexi-
ble in developing contract documents including only the requirements
necessary for particular projects and insurance and indemnification
necessary for potential risks. Private management also has the ability
to pre-qualify contractors and sub-contractors which can be used to
improve the bidding and contractor selection process.
Public management's rigid procedures and standard requirements dis-
courage all but a limited group of large national contractors who
specialize in bidding U.S. EPA-funded remedial work. This limits com-
petition and will drive bid prices up rapidly if the U.S. EPA increases
the rate at which projects are put out to bid. Private management has
the flexibility to limit bidding requirements and the ability to solicit
bids from local contractors. This flexibility creates a competitive bidding
environment, results in low unit prices for the work and the selection
of a prequalified contractor familiar with local working conditions.
tOOX
4 a a 10
MONTHS
PUBLIC PROJECT
12
COMPARISON OF CONSTRUCTION PHASES
The construction phase starts with the contractor mobilizing onto the
site and ends with completion of the work. It includes all of the con-
struction work necessary to complete the remedial action and the
engineering oversight required to assure compliance with the plans and
specifications. One means of measuring performance during the con-
struction phase is the comparison of estimated quantities, bid prices
and projected oversight costs to the actual quantities, prices and costs
The number and extent of change orders are also an indicator of per-
formance. Numerous or extensive change orders which increase the
contract time or price indicate problems in the bid documents and/or
project management.
As of June, 1989, the Public Project had been under construction
for 14 mo and the Private Project for 7 mo. While the projects are not
complete, there is sufficient information to measure the Public and
Private management performance through June. 1989 and discuss the
trends established.
Mobilization and site clearing were the only work items completed
during the first 14 mo of construction for the Public Project. Even with
die small amount of construction work completed, the effect of delays,
inaccurate quantity estimates and change orders on the project costs
have been established. Oversight, stormwater disposal and administra-
tive costs have continued to increase during the delays. Engineering
oversight and stormwater disposal costs have increased $1,000,000 from
the bid amounts. Administrative delays claimed by the contractor during
the first 14 mo will add $1.500jOOO to the Public Project. The site clearing
quantity, the only construction item completed during the first 14 mo,
was three times the bid quantity, increasing the cost for this item from
$50,000 to $150,000. Twelve change orders were issued for the Public
Project in the first 14 mo of construction, increasing the Contract Price
by approximately $200,000.
All of the site clearing work, drainage facilities and slurry walls con-
struction, and 60% of the geofabric, porewater drain system and cap
fill were completed during the first 7 mo of construction at the Private
Project. There have been no increases in the contract price due to delays,
quantity estimates or change orders. The project should be completed
substantially ahead of schedule and engineering oversight costs should
be at least $300,000 less than the originally estimated amount.
Construction phase results are shown in Figure 2 for the Public
Project. The shaded area illustrates the effect of having several con-
tract cost items adjusted by delays; the construction costs continue to
increase while the percentage of construction completed remains
unchanged. Figure 2 presents the same information for the Private
Project, illustrating that the project is on schedule and should be com-
pleted for the bid price. Tables 4 and 5 compare the costs as they are
projected in the construction phase to the bid amounts for two projects.
100X,
coNsmcnoN
-ACTUAL COST
10
12
« 6 6
MONTHS
PRIVATE PROJECT
Figure 2
Construction Phase Results
CONCLUSION
The results of the Contracting and Construction phases demonstrate
the ability of Private Management to outperform Public Management
in Superfund remediation projects. The Private Project is on schedule
and there have been no increases in the contract price during construc-
tion. The Public Project has experienced extensive delays, the contract
price has, in the initial phase of construction, already increased by 15%
and the per-acrc bid prices were four limes greater than those for the
Private Project.
The design phase demonstrates a problem with the U.S. EIWs
approach to oversight of private projects. The Private Project's con-
struction costs only account for 50% of the total project cost. Con-
struction costs, especially for projects of this size, should account for
at least 75 % of the total project cost. The redundant management style
imposed on the Private Project by the U.S. EPA's oversight role and
standard procedures are the primary reason for the imbalance between
184 COST & ECONOMICS
-------
construction and non-construction costs. The U.S. EPA does not take
an active central role in publicly funded remedial actions and should
not in privately funded actions. They are not staffed to do it and it does
not improve the site remediation process.
The U.S. EPA should regulate privately funded remediation in the
same manner they regulate industrial wastewater discharges. They should
issue clear standards and actively enforce those standards. If this were
done, the abilities of private management would be free to achieve the
U.S. EPA's goal of remediating Superfund Sites in a fast, cost-effective
manner.
Table 4
Public Project Construction Phase Results
TableS
Private Project Construction Phase Results
construction Costs
Non-Construction Costs
Bonds
Mobilization/Demobilization
Quality Assurance
Administrative Delay
Transportation/Disposal
TOTAL
Bid
$4,590,000
117,000
730,000
210,000
Projected Total
After 14 Mo. of
Construction
S 5,590,000
117,000
730,000
210,000
27,000/day 1,500,000
10.500.000
$16,147,000
10.500.000
$18,647,000
Bid
Construction Costs
Non-Construction Costs
Bonds
Mobilization/Demobilization
Quality Assurance
Administrative Delay
Transportation/Disposal
TOTAL
$5,140,000
0
145,000
0
0
$5,285,000
After 7 Mo. of
Construction
$ 5,140,000
0
145,000
0
0
$5,285,000
Estimated Construction Phase Engineering Costs
Oversight
Sampling and Analysis
EPA Oversight
TOTAL
GRAND TOTAL
500,000
0
Unknown
$500,000
$16,647,000
800,000
0
g
$ 800,000
$19,447,000
Estimated Construction Phase Engineering Costs
Oversight
Sampling and Analysis
EPA Oversight
TOTAL
GRAND TOTAL
BOO,000
200,000
1,BOO.OOP
$2,800,000
$8,085,000
600,000
100,000
500.000
$ 1,200,000
$ b,485,000
COST & ECONOMICS 185
-------
Use of the New OWPE CERCLA Cleanup Cost Data Base System
For Calculating Settlement Premium Payments,
Evaluating Cleanup Costs and
Reviewing Remedial Technologies
Tom Gillis
U.S. EPA
Washington, DC.
Joe Knox
Mark Johnson, MBA, ARM
PRC Environmental Management, Inc.
McLean, VA
ABSTRACT
The U.S. EPA Office of Waste Programs Enforcement (OWPE) has
developed a data base to provide a basis for estimating response action
costs at CERCLA sites for settlement purposes. OWPE may use these
estimates in the CERCLA settlement process to add a "premium" to
response action cost estimates. The premium payment is an additional
amount included in the settlement to account for the possibility of cost
overruns or additional, unforeseen response actions. This data base is
important to OWPE, the U.S. EPA Region, and the states because it
represents the first empirically based source of data and methodology
to calculate premium payments.
The estimates the data base provides are unique because they are
derived from information on both the estimated cost that usually is
developed at the ROD stage and the actual remedial cost that has accrued
over time. The difference between the ROD estimate and the actual
costs for a given technology can serve as a basis for the premium
payment.
The data base currently contains more than 350 records. Each record
represents an NPL site for which a ROD has been signed. Records
contain general background information on the site, the type of remedy
being implemented, the type of contaminants at the site and the con-
taminated media, the cost estimate for the remedy as set forth in the
ROD and the actual costs that have accrued over time. The actual costs
were generated from CERCLIS reports that show the outlays and obli-
gations for each NPL site.
The data base has additional uses besides calculating premium
payments and cost estimates. It can be used to quickly gather figures
on the number of sites using a particular remedy or the number of sites
with a particular type of contamination.
THE CERCLA CLEANUP COST DATA BASE SYSTEM (CCCDS)
The CCCDS combines cost, location and technical information from
the U.S. EPA RODs and the Comprehensive Environmental Response,
Compensation and Liability Information System (CERCLIS) to assist
the U.S. EPA in the Superfund settlement process and other Super-
fund activities. To date, the U.S. EPA has entered data from more than
350 RODs and corresponding CERCLIS data fields into CCCDS.
The data from the RODs include both general identification and
location information (such as the ROD identification number, site name,
ROD publication date, state and region) and more detailed technology,
contaminant and cost information. The data from CERCLIS provides
location and identification information (such as operable unit number.
U.S. EPA identification number, address, remedial project manager
name and telephone number) as well as cost information on remedial
design and remedial action obligations and outlays for each operable
unit.
CCCDS combines the ROD and CERCLIS information to produce
35 different data elements (listed and described in Table I) for each
record in the data base. Each record represents an individual Super-
fund site operable unit thai has a signed ROD. CCCDS has been
designed to generate reports or provide various analyses.
After the ROD information was included in the CCCDS, the
remaining data fields were completed with CERCLIS information. The
CERCLIS information was generated by two ad-hoc CERCLIS reports
that showed the site location information and the obligation and outlay
data for each site that had a signed ROD.
As the data base currently exists, the CERCLIS cost information is
a one-time, "snapshot" view of the actual costs. The cost information
is updated quarterly in CERCLIS as it accrues for each site, but was
entered only once into CCCDS. Updates of CCCDS will link the
quarterly CERCLIS updates directly into CCCDS.
USING CCCDS TO CALCULATE SUPERFUND
SETTLEMENT PREMIUM PAYMENTS
A primary purpose of CCCDS is to provide an empirical basis for
calculating settlement premium payments using historical Superfund
site data. The U.S. EPA's Superfund settlement policy ' allows the
Agency under certain circumstances to offer responsible parties at Super-
fund sites a limited release from liability in exchange for reimburse-
ment of response costs that may include a "premium" payment to cover
the risk of cost overruns or the need for additional response actions.
Additionally, as the Superfund program evolves, the use of historical
data becomes more and more viable as an analytical aid.
Background
Section 107 of CERCLA, as amended by SARA, holds responsible
parties liable for cleaning up a hazardous waste site — whether as current
or past owners, or as operators, transporters or generators of hazardous
substances. Through CERCLA, the Congress demanded that those
responsible for the presence of hazardous substances at Superfund sites
either carry out the site cleanup themselves or pay for the response
actions the U.S. EPA conducts.
The liability standard for cleanup under CERCLA is "strict, joint
and several," a so that the U.S. EPA may recover the entire cost of
cleanup from any contributor without obligation to identify or seek out
all liable parties. In practice, however, the U.S. EPA has attempted to
negotiate with responsible panics, though there may be hundreds at
a site, in an effort to persuade them to allocate costs among themselves.
The agency increasingly has encouraged out-of-court settlements, that
do not compromise protection of public health and the environment,
to procure PRP cleanup of the site or recover cleanup funding. The
U.S. EPA prefers to have the PRPs conduct the remedial actions rather
186 COST & ECONONflCS
-------
Table 1
CCCDS Data Elements for Each Site Record
CCCDS DATA
ELEMENT DESCRIPTION
S1TENAME Preferred name of the site
RODID Record of Decision identification number (EPA assigned number to a
particular ROD)
EPAID A unique identifier (either in Dun and Bradstreet or GSA format used to
indicate a hazardous waste site or an unanticipated removal (incident)
occurring at a location not previously identified as a site in the CERCLIS
inventory (e.g. oil spill)
OPUNITNUM A designation for the operable unit at which events are occurring.
Legitimate entries are '00' to '99'
REGION EPA Region in which the site is located
ADDRESS Street address, route number, or other specific identifier of the physical
location of the site or incident
CITY Name of the city, town, village or other municipality in which the site is
located or incident occurs. If the site is not located or if the incident did
not occur within such a jurisdiction, the nearest geographical place name
STATE Code that identifies the state or territory in which the site is located or
incident occurs
ZIP Code that identifies the U.S. Postal Service delivery area in which the site
is located or incident occurs
RCNAME Regional contact name
RCPHONE Regional contact phone number
RODDATE Date when the Record of Decision was signed
REMTECHTYPE Code(s) of the type(s) of remedial technology selected by EPA and
described in the ROD
REMTECHSP Code(s) of remedial technology specifications (design/engineering
specifications
ESTDESCOST Estimated design cost (cost to complete the remedial design)
ESTCONCOST Estimated construction cost (cost to construct or implement the remedial
technology after a final design has been completed)
DESCONCOSTS Total costs of design and construction (only if total is provided in the
ROD)
ESTOMCOST Estimated operations and maintenance O&M cost (cost to operate and
maintain the medial technology after construction)
NUMOMYEAR Number of years of operations and maintenance
ESTPRWORTH Estimated total present worth of the remedial technology (sum of estimated
design, construction and O&M costs only as listed in the ROD)
RDCOMPDT The actual completion date of the remedial design
RDPRIOROBL The dollar amount that was obligated (set aside) for the remedial design
for the prior fiscal year
RDCURROBL The dollar amount that was obligated (set aside) for the remedial design
for the current fiscal year
RDOUTLAYS The dollar amount outlayed (paid) for the remedial design to date
RACOMPDT The actual completion date of the remedial action
RAPRIOROBL The dollar amount that was obligated (set aside) for the remedial action for
the prior fiscal year
RACURROBL The dollar amount that was obligated (set aside) for the remedial action for
the prior fiscal year
HAOUTLAYS The dollar amount outlayed (paid) for the remedial action to date
KEYCONTAMN Code(s) of the key contaminant(s) at the site
DRUMS Drums as a contaminated medium or source of contaminated medium at
the site
BKLIQUID Bulk liquid as a contaminated medium or source of contaminated medium
at the site
SOIL Soil as a contaminated medium at the site
CROUNDWATR Ground water as a contaminated medium at the site
SURWATER Surface water as a contaminated medium at the site
AIR Air as a contaminated medium at the site
than simply provide cleanup funds.
Congress in SARA Section 122(f) authorized the U.S. EPA to enter
into covenants not to sue, empowering the agency to provide limited
releases from liability to PRPs in settlements. The covenants not to sue
usually include "reopeners" that allow the U.S. EPA to revisit the settle-
ment to recover additional costs incurred due to unknown conditions
or new information that arises after remedial actions begin — but that
may be waived when, for instance, the U.S. EPA has determined that
"extraordinary circumstances'' exist. The extraordinary circumstances
waiver may be applied based on the effectiveness and reliability, or per-
manence, of the remedy, the nature of remaining risks at the facility,
the demonstrated effectiveness of the technology, the involvement of
PRPs, litigative risks or "whether the Fund or other sources of funding
would be available for any additional remedial actions that might even-
tually be necessary at the facility."3 Under certain circumstances, the
U.S. EPA may thus waive the usual reopeners when PRPs have sub-
mitted a premium payment above baseline remedial costs.
METHODOLOGY FOR DEVELOPING CCCDS
The CCCDS data base was designed for easy use by the U.S. EPA
regional and headquarters staff involved in CERCLA settlements and
other activities. Because the CCCDS is menu-driven, little or no training
is necessary to begin using the data base. The CCCDS software has
been compiled for speedy operation using the royalty-free Foxbase run-
time PC software. Foxbase, however, is not required to run the soft-
ware. The U.S. EPA also is developing a users manual for routine
procedures such as data input, report generation, data querying, coding
form generation, data backup, data displaying and printer selection.
The U.S. EPA conducted a detailed review of each of the 350 RODs
in CCCDS and recorded data on coding sheets. The data on the coding
sheets were then entered into the system. Professional judgment was
sometimes required to match site-specific technologies with a master
list of control, removal and treatment technologies. Additionally,
extended review and professional judgment were required to identify
technical specifications associated with each technology.
The Premium Payment Concept
The premium payment concept is documented in the U.S. EPA policy
and guidance.4 The premium payment levies a surcharge above
cleanup costs. Similar to an insurance premium, the payment offsets
the risk the U.S. EPA assumes in providing PRPs with a limited release
from liability with a payment that exceeds the cost contemplated to com-
plete remediation. That premium should be enough to compensate for
both potential cost overruns and unexpected additional costs,5 yet may
provide an incentive to settlement by supplying a release from future
liability. When using a premium payment, releases from liability are
of two general types: (1) a release from responsibility for cost overruns
in implementing the remedy contemplated in the settlement agreement,
or (2) a release from additional site remediation if the selected remedy
is not protective of human health and the environment. Either type com-
monly carries reopeners that allow the U.S. EPA to recover additional
costs from PRPs if conditions arise that were unknown when U.S. EPA
determined that remediation was complete.
The U.S. EPA guidance states that the premium should be set at a
level that shields the government from having to bear potential cost
overruns and that provides funds "to protect public health and the en-
vironment in the event that additional response work will be needed
at the site." The premium should be adequate to protect against future
liability that may arise due to remedy failure or mistaken assumptions
about the effectiveness of the remedy. In addition, new information dis-
covered about a site, perhaps during the U.S. EPA's 5-yr review required
under CERCLA section 121(c), may demand further remedial work.
In such cases, the guidance says, both the likelihood and the cost of
future remediation should be considered, and the premium should be
allocated in terms of each PRP's percentage of the total estimated
remediation cost.
Still, the method of calculating the premium remains unresolved. The
current U.S. EPA OECM guidance provides only general guidelines.
A recent publication has suggested a framework of procedures that can
be used to derive a premium payment on a case-by-case basis.6 The
methodology relies on premium ratio multipliers that were derived from
statistical distribution functions that represent the consequence of the
risk the U.S. EPA retains balanced against the probability that the
premium will be sufficient to cover any additional costs. CCCDS
provides an empirical foundation for the premium payment previously
COST & ECONOMICS 187
-------
discussed. The system provides, for the first time, an historical picture
of cost overruns and unexpected additional expenses in the Superfund
remedial program.
Calculating Premium Payments Using CCCDS
Historical Data From Other Superfund Sites
Ideally, two types of cost information are needed to calculate a
premium payment: (1) the estimated cost of cleanup at the time of settle-
ment, and (2) the actual cost of cleanup at the completion of cleanup.
The difference between these two costs represents the premium pay-
ment that should be assessed during a cash-out settlement.
At the time of a cash-out settlement, however, both panics (the U.S
EPA and the PRPs) have only the estimated cost for each site specific
remedial technology that will be used in the cleanup. Because of the
nature of premium payment (included as part of an up-front, cash-out
settlement), actual costs are not available. Although actual costs are
not available for the specific site being settled, they may be available
for other sites that have implemented the same or similar technologies.
If, therefore, historical data are reviewed for other sites that have im-
plemented the same or similar technologies, then a premium payment
may be based on the statistical mean (or other measure of central
tendency) of the difference between the estimated costs and the actual
costs for each site. An alternative approach would be to calculate the
statistical mean of the difference between the site-specific estimated
cost and the actual costs at each of the other sites.
Sources of Estimated Costs
Estimated costs of cleanup are refined throughout the RJ/FS process.
Generally the closest estimate at the lime of settlement are the figures
that appear in the Superfund ROD. These estimates generally are based
on the results of the feasibility study and are calculated using a U.S.
EPA costing model. This model is a software-based system that incor-
porates the recommended remedial technology specifications and site
considerations into a site-specific cost estimate.
The ROD usually includes cost estimates, technology specifications
and other site characteristic data. Through a review of sites with similar
technology specifications and site characteristics, the estimated costs
from RODs at other sites can be used to help calculate a premium pay-
ment for settlement purposes.
Sources of Actual Costs
Actual costs are available only when the cleanup has been completely
designed and constructed and all operations and maintenance completed.
While this figure rarely is available, since most Superfund sites have
not reached this phase, different types of actual costs may be obtained.
One type of actual cost is the cost of implementing the remedial tech-
nology as indicated by the final remedial design. For premium pay-
mem purposes, the actual cost would be the sum of: (1) the cost of the
remedial design, and (2) the projected cost of implementing the final
design.
Another type of actual cost is the cost at the completion of the remedial
action (construction), but before long-term operation and maintenance
begins. For premium payment purposes, this would be: (1) the cost of
the remedial design, and construction, and (2) the projected cost of
the operating and maintaining the technology until the site is properly
cleaned.
These actual costs are shown in CERCLIS as either obligations or
outlays. Obligations are dollar set-asides that the Agency has committed
to spend. For example, an obligation for remedial action is based on
its projected cost as documented in the final remedial design. CERCLIS
shows both current year and prior year obligations. Outlays are dollar
amounts that the Agency has already spent. For example, an outlay for
remedial design or remedial action represents an invoiced amount that
the Agency has paid.
CCCDS Capacity To Calculate
CERCLA Premium Payments
CCCDS is designed to help the U.S. EPA calculate premium pay-
ments using many different combinations of data. As discussed above,
Table!
List of Technologlo Coded and Entered Into CCCDS
ENGINEERING CONTROL
TECHNOLOGIES
Technologies
Air I.mitsioni Control
pipe vents
ireacli venu
gas barrier*
gas collection
overpecking
Surface Wiur Control
surface if tit (opt)
surface water diversion
•ad collection systems
dikes end berms
ditchee. diveniom. waterways
chutes and downplpes
levees
seepage beams and ditches
terraces and benches
grading*
re vegetation
surface water pumping
Qroundwaler Control
• impermeable barrier
slurry walls
grout curuim
sheet pilings
• permeable treatment beds
• groundwaier pumping w«as, fallow
water table edjastmeni
plurot containment*
• kachaie control
subsurface drain
- drainage ditdMs
Asphalt Dryer
On-iilt RCRA Landfill
Temporary On-lite Storage
On-site Disposal
Aeration
OFF-STTI TRANSPORT. STORAGE,
TREATMENT OR DISPOSAL
Off-lilt Transportation to RCRA Landfill
Off-lite Transportation to RCRA Incinerator
Off-site Transportation lo other Treatment
REMOVAL TECHNOLOGIES
Escavatioa
Hydraulic Dredging
Mechanical Dredging
Provision of Alternative Water Supplies
individual treatment unio
water distribution systems
new wells In s new location or
deeper wells
cisterns
upgraded treatment Tor existing
distribution systems
Drum Removal
TREATMENT TECHNOLOGIES
Mrerl Waste Treatment
Biological Methods
• modified convention*! waslewaiet
t/eatnsent lachniqoea
• anaerobic, aerated, and
facuJutiva lagoons
• supported growdi biological
Chemical Methods
chlorinatipn
precipitation, ftocculaiioo,
•edimenttliofl
neutralization
equal infjon
chemical oiidatkn
decalorinaiion
Physical Methods
air stripping
carbon anaorptsoe)
permeable bed Inalaaeu
wet air oxidation
incineration
vapor phase abaortaio*
activated sludge
Trenutnl ef SoUs end Seettmeab
Incineration
Wet Air OiidaiKM
Solidification
• solidification w/naparation
In she T:
• solution mining, (soil washing or
flushing)
• nmruiDaiion/oetoairication
• asicrobiological degradation
Relocation of Residents
Building Removal
Tank Removal
Belt Uoaid Removal
Debrii Removal
Soil!
SAMPLING/MONITORING
TECHNOLOGIES
Install Monitoring Wells
Sampling awl Analysis (of welU, soil,
leechate. surface water, air etc)
SITE PREPARATION TECHNOLOGIC
Access Road
Grading
Reveteiation
Oeartng and Grubbing
Fencing
Decontamination
• Building Decontamination
these premium payment calculations are based on the actual costs
associated with sites that already have implemented similar technologies.
CCCDS is best suited to calculate the cost overrun component of
the premium payment. The cost overruns can be calculated using either
obligations or outlays. Because outlays more realistically represent actual
costs, the calculations using outlays are also more realistic.
To calculate overruns using outlays, CCCDS would first be queried
to list all sites where: (1) the total of remedial design and remedial action
outlays is greater than the ROD's estimated costs (total estimated cost
188 COST & ECONOMICS
-------
of design and construction), and (2) the ROD's estimated costs are
greater than zero. For each of these sites, CCCDS can calculate the
absolute value of the overrun (difference between total outlays and the
estimated cost) as well as the percentage of overrun. CCCDS can then
calculate a mean, median, variance, standard deviation or standard error,
among other measures, for the resulting sample of cost overruns.
CCCDS can perform a similar calculation using only the obligation
fields for remedial design/remedial action (RD/RA). Use of the obli-
gation fields would provide a greater sample size because many more
sites have funds obligated for RD/RA but have not yet had actual out-
lays. The obligations, however, would not provide premium payment
calculations that are as accurate as the outlays, for two reasons. First,
funds may be obligated for RD/RA and then deobligated. Second, the
obligation may be greater or less than the subsequent outlay. The cal-
culation of the premium payment would in any case proceed the same
as in the calculation for the outlays. CCCDS would first be queried
to list all sites where: (1) the total remedial design and remedial action
obligations are greater than the ROD estimated costs, and (2) the ROD
estimated costs are greater than zero. CCCDS would then calculate
statistical summaries of the cost overrun.
CCCDS also can calculate the cost overruns using a combination
of outlays and obligations. For example, it can list cost overruns for
all sites where the RD outlays plus the RA obligations were greater
than the ROD estimates. CCCDS also can display the results in dif-
ferent subgroups. Determining the cost overruns by different tech-
nologies probably would be the most useful for premium payment
calculations, but the system can further show cost overruns by the U.S.
EPA region, type of contaminant or contaminated medium, ROD date
and other aggregations.
CCCDS also is capable of providing historical data on the need for
additional, unplanned response actions, although this calculation is more
involved and less empirically based. Determining the need for addi-
tional, unplanned response actions would first involve a search of
CCCDS to determine which sites contain more than one operable unit.
A further review would be needed to determine if the additional operable
unit was described in the original site plan. If it was not, then the total
outlays for that operable unit would represent an additional component
of a premium payment. These calculations could be grouped easily by
technology, region or other components.
OTHER USES OF CCCDS
CCCDS can be used for purposes other than calculating settlement
premium payments. Because the system contains detailed information
on remedial technology types, remedial technology specifications and
contaminants and contaminated media, the system can be used to help
plan site-specific response actions at other sites. For example, this might
include reviewing other sites that already have implemented a particu-
lar technology, addressed certain contaminant types or any combina-
tion thereof. Table 2 shows a list of the technologies that have been
coded into the system.
CONCLUSION
CCCDS can help provide an empirical basis for calculating settle-
ment premiums at Superfund sites. The system can provide estimated
costs from Superfund RODs and actual costs from CERCLIS for each
site that has a signed ROD. CCCDS can perform various calculations
with the ROD and CERCLIS data to develop premium payments based
on: (1) cost overruns and (2) the need for additional response. CCCDS
also can be used to review technology types, specifications and site-
specific data for remedial planning.
REFERENCES
1. 50 Fed. Reg. 5034, Feb. 5, 1985.
2. The standard of "strict" liability in CERCLA Section 101 (32) is incorporated
from the Clean Water Act. Courts ruling in CERCLA cases have further held
that the standard is joint and several.
3. Public Law 96-510, Superfund Amendments and Reauthorization Act, Sec-
tion 122(f)(4)(F).
4. U.S. EPA Guidance on Premium Payments in CERCLA Settlements, the U.S.
EPA Office of Enforcement and Compliance Monitoring and Office of Solid
Waste and Emergency Response, signed Sept. 15, 1988.
5. Guidance on Premium Payments in CERCLA Settlements, p.2.
6. Johnson, Gillis and Fries, "Using the Premium Payment Concept to Pro-
mote Superfund Settlements," presented at the HMCRI Superfund Conference,
Nov., 1988.
COST & ECONOMICS 189
-------
De Minimis Settlement — A Success Story
Jay Nikmanesh
Frank Bissett
Sandra McDonald
George Duba, Ph.D.
TechLaw, Inc.
Lakewood, Colorado
ABSTRACT
Under Section I22(g)(l) of SARA, the U.S. EPA is provided the
explicit authority to enter into so-called "de minimis" settlement, in-
volving only a minor portion of the response costs, with certain classes
of responsible parties. De minimis parties would include, for example.
a landowner who did not contribute to a hazardous waste release and
did not conduct, have knowledge of or permit hazardous waste activities
at the site. During the past few months, a significant de minimis settle-
ment was reached involving a Superfund site in the midwest. The settle-
ment was reached based on a unique quantitative approach utilized
during the negotiation phase of site remediation. All site records were
organized, screened and analyzed for information regarding waste ship-
ments to the site. Quantitative information was entered into a
computerized data base. The data base was then manipulated to identify
hundreds of de minimis parties. These parties were approached by the
U.S. EPA for their portion of the cleanup. During negotiations, the data
base was modified using suggestions from the U.S. EPA staff to "fine
tune" the levels of responsibility for each de minimis party. The end-
result was one of the largest de minimis settlement to date - over 11
million dollars for site remediation.
INTRODUCTION
Section 122(g)(l) of SARA provides the U.S. EPA with explicit
authority to enter into so-called de minimis settlements with certain
classes of PRPs whose involvement at the site mandates responsibility
for only a minor portion of response costs. Although individually these
de minimis settlements usually include small sums of money, a de
minimis settlement with several hundred PRPs can yield quite a sub-
stantial amount. The case study described below is just such an
example— A "de minimis" settlement for over 11 million dollars.
SITE DESCRIPTION AND HISTORY
Located on less than 100 ac in the Midwest, the landfill site ("the site")
accepted industrial wastes for nearly 10 yr prior to the 1980 implemen-
tation of hazardous waste disposal regulations mandated by the passage
of RCRA. Close to 20,000,000 gal of wastes were indiscriminately
disposed of at the site, including hundreds of toxic chemical compounds.
Hazardous substances and wastes were dumped into unlined ponds and
barrels were deposited in an unlined pit and subsequently buried. Wastes
from the ponds and drums polluted hundreds of thousands of cubic yards
of soils, producing toxic sludges and contaminating local groundwater.
The U.S. EPA closed the site in the late 1970s after years of chronic
violations of its state permit and of industrial disposal laws. A high
hazardous ranking system (MRS) package score resulted in the site's
placement on the NPL in 1983. Cleanup of the site will involve resto-
ration of groundwater quality and require some form of removal and
destruction of contaminated soils and sludges. Site investigation and
remediation costs will probably stretch well into the nine figure range.
To date, several hundred PRPs have been identified. These PRPs
include generators and transporters of hazardous substances and wastes
disposed of at the site. Approximately one-third of these parties formed
a steering committee to participate in investigation and cleanup activities
at the site.
In the mid-1980s, the United Stales filed suit under CERCLA against
more than 30 responsible parties at the site, many of whom were steering
committee members, for implementation of the U.S. EPA's selected
remedy and for payment of response costs. Later, the steering commit-
tee filed a third-party action against more than one hundred other PRPs
not named in the U.S. EPA's earlier suit. The action sought to show
that these third-party defendants were also liable in connection with
activities at the site and asked the court to order these parties to pay
their share of cleanup costs.
Many of the defendants identified in the third-party actions were "de
minimis" parties, as determined by the U.S. EPA pursuant to Section
122 of SARA. This section of SARA provides for settlement with par-
ties whose waste contributions have been minimal in comparison with
the total volume of hazardous substances at a site. The U.S. EPA
negotiated with a group of the "de minimis" parties and the parties
agreed upon a consent decree for "de minimis" settlement and release
from liability in connection with the site.
The proposed Consent Decree: provides for final settlement of alleged
liabilities for site cleanup and response costs; raises revenues to be
applied to cleanup activities; and will greatly reduce the expense and
complexity of pending litigation with defendants and non-settlors.
Effectively, "de minimis" parties will be released from future liability
with regard to the site as long as no new information on their waste
contribution to the site is uncovered. By entering into the settlement
and resolving the liability issue, a "de minimis" party also will be
protected from the third-party action filed by the defendants.
Eligibility for participation in the settlement was based on a party's
waste contribution to the site. Volumetric waste allocations were
determined using a transactional data base developed by the authors
under a litigation support contract to the U.S. EPA. An alphabetic listing
of participating parties and their respective volumes and cash payments
was developed from the data base and appears as an attachment to the
proposed Consent Decree. Cash payments were calculated by
multiplying a party's percentage of total waste volume at the site by
the U.S. EPA's estimate of total past and future response costs plus
premiums. To date, the settlement includes over 170 "de minimis" parties
and is valued well in excess of 11 million dollars.
190 COST & ECONOMICS
-------
LITIGATION SUPPORT PROCESS
The authors were first hired to provide litigation support and evidence
audit services on the case in the mid-1980s. These services included
sample chain-of-custody audits, case file inventories, the development
of a transactional data base, and the ongoing compilation of a record
documenting the findings made by the U.S. EPA in entering into the
"de minimis" settlement. Through this 5-yr process, the authors became
thoroughly familiar with the elements of tide case and the case documen-
tation. This familiarity enhanced the authors' ability to develop a trans-
actional data base made up of more accurate volumetric waste
transactions.
Volumetric waste allocations forming the basis for cash payments for
parties entering the proposed "de minimis" settlement were then derived
from the transactional data base. It summarizes site waste transactions
for more than 350 parties over a 10-yr period.
Documentation for developing the data base was acquired by the U.S.
EPA at first by using its authority under RCRA Section 3007 and
CERCLA Section 104(e)to request information. Additional evidence
was gathered during the discovery phase of litigation and during settle-
ment negotiations with parties interested in participating in a "de mini-
mis" settlement.
Logs obtained from the owner recorded transactions at the site on
a monthly basis throughout its business life. Daily records, which
covered the last few years of operation, also were acquired. These logs
provided specific information on transactions by date, transporter and/or
generator and listed volumes and waste types.
Reports were obtained from state agencies detailing monthly trans-
actions by transporters. These reports specified the source of wastes,
the volume and waste type hauled and the destination for disposal. While
transporters from out-of-state who used the site were not required to
file monthly reports with state agencies, most kept in-house records
providing similar information.
A state-mandated manifesting system was in effect during the period
that the site accepted wastes. These manifests provided potential evidence
supporting alleged site transactions. Manifests typically were initiated
by the generator at his facility, executed by the transporter when a load
was accepted and signed by the site operator upon receipt of the load.
A completed and executed manifest provided agreement between these
parties on the date of a transaction, the source of wastes and the volume
and waste type disposed of at the site.
Shipping tickets and receipts provided additional information on site
transactions. Manifests and logs often cited receipts and tickets, provid-
ing additional evidence that a transaction had actually occurred. A trans-
action poorly supported by other documentation was often confirmed
by a receipt or ticket bearing the site owner's signature.
Many of the documents mentioned were provided with parties'
responses to the U.S. EPA's requests for information. Some parties
provided additional evidence on transactions in the form of narratives
or tables constructed from in-house records. These responses also
provided historical information on corporate affiliations and relation-
ships. Other information was obtained from depositions conducted pur-
suant to preparation for litigation.
Documents from the above sources were sorted, numbered and then
organized in transaction packets. Each packet contained documenta-
tion supporting a specific disposal event at the site. Packets were then
filed in chronological order in a folder assigned to individual PRPs.
All important information regarding a specific transaction was
extracted from the individual packets. Ideally, extractable data included
the generator, transaction date, waste type, volume, unit type, trans-
porter, document type and document number.
A written record was created for each PRP that contained its name,
an assigned code, addresses, the name and address of its contact and
the names of any other companies with which the PRP was associated.
Lists also were created for all the unique waste types and unit types
that were found in the transaction documents. Each unique waste or
unit was assigned a code number. This information was thus prepared
for entry into computerized data bases.
The computer system used to store extracted information was a
Table 1
Transactional Data Extraction Sheet
GENERATOR:
Extraction Date
Initials
DATE (WASTE TYPE|VOLUME|UNITS|TRANSPORTER]DOC. TYPE|DOC NO.
PRIME 2755 minicomputer operating under the PRIMOS operating
system. A resident software package known as the HENCO INFO data
base management system was used for this case. This hardware and
software combination was chosen because of its unique suitability for
handling the amount of information that was to be stored and retrieved.
Four main data file structures were developed to accommodate the
four major sources of information that the U.S. EPA provided. The first
three structures respectively held data from transporter reports filed
with state agencies, the site's monthly log and transporter in-house
records. The fourth data file contained data from the site's daily log,
as well as supporting evidence from waste manifests, CERCLA Section
104(e) responses, shipping tickets and receipts.
A series of data files called "match files" also was created to store
recurring data from the written PRP records and from the respective
waste and unit type lists. Screen drivers were developed to facilitate
on-line input and edit all data types. Once data entry was completed,
reports of entered data were generated using quality control programs
written specifically for that purpose. These reports were compared with
original documents to ensure the accuracy of data entry. Changes neces-
sary to correct extraction and entry errors were made to the data, and
a second quality control report was generated. This report was used
to verify that the paper changes were reflected in the data base.
Programs were developed for the production of summaries listing
particular information in the various data files. For instance, one
program converted all unit types into gallons, from which summaries
were generated that could rank generators or transporters by the amount
of waste contributed to the site and identify waste types associated with
each volume. These summaries, an example of which is shown in
Table 2, assisted the U.S. EPA attorneys in identifying the relative status
of PRPs.
Table 2
Generator Ranking
RANK
72 COMPANY A
73 COMPANY B
74 COMPANY C
WASTE TYPE GALLONS
ASBESTOS 3150.00
ASBESTOS INSULATION 3600.00
6750.00
CYANIDES
ETCHING-SOLUTION
NITRIC ACID
OIL
PAINT SLUDGE & WATER
ALCOHOL
CHLOROTHENE
ISOCYANATES
OIL
PAINT SLUDGE
PHENOLIC SAND RESIN
% OF TOTAL
0.061
0.070
0.131
0.019
0.063
0.028
0.018
0.002
0.130
0.004
0.016
0.009
0.029
0.029
0.032
0.119
COST & ECONOMICS 191
-------
Ultimately the four main data files were combined into a compre-
hensive data file which contained all the information extracted from
the site documentation. Software was developed to produce a transac-
tional summary printout of this data file. The transactional summary
report was organized alphabetically by facility/party. Transactional
information specific to a facility/party was listed chronologically by
month (Table 3). Monthly volumes were presented from up to four of
the data files containing the four categories of transactional data men-
tioned earlier. Of the volumes present for a specific month, the largest
quantity was reported in the last column. Totals for each data category
and the "largest quantities'' column appeared at the bottom of each
party/facility report.
Table 3
Transactional Summary Report
Note: All amounts given in gallons
GENERATOR: COMPANY D
ADDRESS: 121 KELP STREET
STATE
BEfiJBI
DEC
SEP
HOV
NOV
HOV
1975
1976
1977
1978
1979
400,
,00
160.00
100.
300.
.00
.00
HOHTHLY
LM
300
300
200
300
.00
.00
.00
.00
TRANS.
LM
DAILY
LOG
300.00
300.00
300.00
300.00
LARGEST
QUANTITY
400.00
160.00
100.00
100.00
300.00
WASTE
CQJBES.
11
11,12
11, S7
77
13,77
1360.00 1100.00
260.00 1200.00 1660.00
Using this transactional summary report, the U.S. EPA and was able
to identify parties it considered "de minimis." This process facilitated
the formation of a coalition of "de minimis" parties interested in
settlement negotiations with the U.S. EPA.
PROBLEMS
Early drafts of the transactional summary report were somewhat
inaccurate for a variety of reasons. "De minimis" parties were quick
to point out perceived discrepancies in their individual volumes. The
U.S. EPA attorneys were soon cognizant of the need for adjustments
to some reported volumes.
Volume discrepancies were attributable to a variety of factors.
Opinions on the proper conversion factors to be used in the data base
to convert some unit types to gallons were diverse. For example, the
U.S. EPA assumed that the amount of waste in a drum was 55 gal unless
irrefutable evidence to the contrary could be presented. PRPs often
claimed that the drums that they disposed of at the site were smaller
or contained a lesser amount of waste.
A recurring problem arose in reconciling different types of documen-
tation pertaining to a specific transaction. Discrepancies in reported
information between documents often led to the entry of a transaction
into the data base twice (i.e., double counting a single transaction).
For example, a daily site log might record receipt of a shipment a month
later than the pick-up date recorded by the transporter in his report
to a state agency. Similarly, two pieces of documentation on one trans-
action might report two seemingly different waste types, e.g., spent
hydrochloric acid in the generators 104(e) response and tank bottoms
on the site log, or incompatible units, e.g. cubic yards on an invoice,
pounds on a receipt. Once again, the potential for double counting
existed.
Another type of problem arose from the general task of party/facility
identification. The PRP files were originally compiled at the direction
of U.S. EPA attorneys and contained aliases, name changes and affilia-
tion information for parties identifying subsidiaries, parents, etc. During
the course of litigation and negotiations, some of these relationships
changed either because of acquisitions, mergers or other reasons. The
result was that some parties had transactions listed under multiple names
or had the same transaction attributed to two different entities not known
to be the same or related. In both situations, there was potential for
reporting the correct amount of waste for the PRP.
Finally, some transactions were simply not well documented. Only
by developing transactional packets and comparing supporting documen-
tation could they be substantiated. Certain generic assumptions had to
be made.
SOLUTIONS
Discrepancies were resolved by a variety of means. As negotiations
began to produce potentially realizable settlement terms, the authors
re-audited transactional packets and eliminated the cited sources of errors
in quantities. A hierarchy was established for ranking the quality of
data presented by the many document types, and standards for evaluating
the quality of documentation as evidence of a transaction at the site
were re-examined. Additional information obtained from depositions
and new or supplementary responses to information request letters clari-
fied many of the quantity, conversion factor and party/facility identity
problems. Negotiations and communications with some "de minimis"
parties on their volumes also resulted in the submission of additional
or clarifying information. Volumes were literally negotiated in a few
cases where irrefutable evidence of a transaction existed but the volume
or some other factor was unclear.
Changes were made to the data files, reflecting the resolution of dis-
crepancies, and reports similar to Table 4 were generated for use in
the "de minimis" settlement process. One of these reports ultimately
became pan of the volumetric allocation attached to the proposed
Consent Decree.
Table 4
Transaction*! Summary Report
Note: All amounts given in gallons
GENERATOR:
COKPAMY E
ADDRESS: 123 YOUR LANE
SEP 1979
NOV 1979
DEC 1979
FEB 1980
APR 1980
JUN 1980
STATE
KCEOJBX
1200.00
4200.00
4200.00
4200.00
DAILY
me
1600.00
4200.00
4000.00
1200.OO
4200.00
3600.00
REVISED HASTE
QUANTITY CBBE5
1600.00 81
3600.00 81
3600.00 84
3600.00 81
4200.00 65
3600.00 81
TOTALS 3200.00 8400.00 4200.00 22800.00 22200.00
TOTAL: 22200.00
RESULTS-PROPOSED CONSENT DECREE
FOR "DE MINIMIS" SETTLEMENT
The re-audit of the transactional files and the data base was com-
pleted last fall. The process was documented in a series of audit reports
submitted to the U.S. EPA, reporting on discrepancies identified and
their resolution at the direction of the U.S. EPA attorneys. Subsequent
to the generation of a list of "de minimis" parties, the negotiated Con-
sent Decree was distributed to parties for consideration. By early spring
of this year, over 150 parties, representing nearly half of those eligible,
had submitted executed Consent Decrees.
The U.S. EPA's decision making process in entering into the settle-
ment has been carefully documented in a "De Minimis" Settlement
Record. At this writing, the proposed Consent Decree still has not been
formally entered by the court. Discrepancies with all participating parties
have been resolved, and comments received during a public comment
period were addressed with the assistance of the authors.
192 COST & ECONOMICS
-------
CONCLUSIONS AND RECOMMENDATIONS
The de minimis settlement process described was a definite success
story, but the authors learned several lessons during the project. These
lessons are summarized in the recommendations listed below:
• Start the process early - identify as many de minimis settlers as
possible so the response costs can be spread and thus reduced for
any given party.
• Identify and collect all relevant waste transaction documentation-
new waste information will change totals.
• Organize—a complete document organization and control system will
facilitate location and retrieval of important waste contribution data.
• Automate—modifying waste contributions is much easier using the
computer.
• Communicate—all parties on both sides of negotiations must be aware
of all waste transaction assumptions used in building the data base.
• Check for accuracy—constant quality assurance will increase the level
of comfort for all involved in the process.
DISCLAIMER
This paper was prepared with the knowledge of the U.S. EPA's
National Enforcement Investigations Center (NEIC) and has been
reviewed by representatives of this agency office. Statements and
opinions expressed are those of the authors. No official support or
endorsement by the U.S. EPA or any other agency of the federal govern-
ment is intended nor should be inferred.
COST & ECONOMICS 193
-------
CERCLA Natural Resource Damage Release Request
Delaware's Approach
John T. Barndt
Diane E. Wehner
State of Delaware
Department of Natural Resources and Environmental Control
New Castle, Delaware
ABSTRACT
jus important that both Federal and State natural resource trustees
become' involved early in the remedial ~
"
doipg so.
resource
theTjtOD. When a
concerns
rdin
heard
is financing the remediation at a site, trustee
me identification of
incorporated into the RI/FS and
involvement is essential as the PRP may request a release from future
liability for natural resource damages after the ROD is signed. The Dela-
ware Department of Natural Resources and Environmental Control has
found the end result of this process to be successful compensation for
-------
Pre-Rl/FS
Trustee Identification
and Comment
Remedial Investigation
and Feasibility Study
Trustee Comment
Record of Decision
Trustee Concurrence
Remedial Action
Negotiations (Consent
Decree)
Trustee Comment
Remedial Design
and Remedial Action
NPL
Deletion
Fig. 1
Natural Resource Trustee Involvement in the Superfund Process.
resource trustee (the Secretary of the DNREC) oversees the manage-
ment of the State's environmental concerns in the Department's five
divisions. These divisions of the agency include the Division of Air
and Waste Management, the Division of Water Resources, the Divi-
sion of Fish and Wildlife, the Division of Parks and Recreation and
the Division of Soil and Water Conservation (Fig. 2). Those divisions
with affected resources were formally contacted to comment on the
RI/FS and ROD and to "sign off on the release for the specific
resources in question. The following pathway led to their eventual con-
currence on the damage release requested:
• Various division representatives were involved throughout the
remedial process for technical support.
• The investigation of any damages to natural resources was incorpo-
rated into the RI.
• The development of restoration/replacement activities was performed
during the FS.
• The selection of an appropriate remedial action to compensate for
any natural resource damages identified was made along with other
remedial decisions during the development of the ROD.
• The design of the appropriate remedial action was incorporated into
the remedial design document.
The agreement by the PRP to perform a remedial action was prompted
because of the authority the State and Federal trustees held with the
covenant not to sue. The end result was adequate compensation for any
natural resource damages.
DELAWARE DEPARTMENT Of
NATURAL RESOURCES AND
ENVIRONMENTAL CONTROL
Division of Air and
Waste Management
- Air
Division of Fish and
Wildlife
Fish and Wildlife
Division of Water
Resources
- Surface water
Grounduater
- Wetlands
Division of Soil and
Water Conservation
- Soil
Division of Parks and
Recreation
Rare/Endangered
Plants
Fig. 2
Structure of the Delaware Department of Natural Resources
and Environmental Control (DNREC).
MAKING THE PROCESS WORK: A CASE STUDY
The Wildcat Landfill is a 45-ac site located along the St. Jones River
in Kent County, Delaware, approximately 2.5 mi southeast of Dover
(Fig. 3). The landfill was privately operated, accepting both municipal
and industrial wastes from 1962 until it was ordered closed by DNREC
in 1973 due to numerous permit violations. The landfill was later in-
vestigated by the U.S. EPA and DNREC and placed on the NPL in
1982. From late 1985 to 1988, an RI/FS4 was conducted by DNREC
under a cooperative agreement with the U.S. EPA.
The major findings of the RI were:
• Groundwater beneath the landfill and to the southeast of the landfill
was contaminated with low levels of trace metals and organic con-
stituents.
• Landfill contents, including drummed wastes, were exposed within
and at the boundary of the landfill.
• Leachate seeps with inorganic and organic constituents were found
along the periphery of the landfill in the area of an adjacent pond.
• Surface water and sediments in the adjacent pond were contaminated
by inorganics.
• Aquatic fauna in the adjacent pond exhibited elevated levels of in-
organics.
Two RODs were developed to address the activities required to
remediate the site and the adjacent pond. The problems on and within
COST & ECONOMICS 195
-------
474000
478000
482000
474000
478000
482000
SCALE r • joxr
Fig. 3
Location of the Wildcat Landfill Site in Kent County. Delaware
the landfill were addressed in the first ROD in June, 1988'. The pond
adjacent to the site was addressed in a second ROD in November.
1988*.
The first ROD required a partial landfill cover, replacement of certain
private wells in close proximity to (he landfill, removal and disposal
of drummed wastes, placement of institutional controls on-site and to
the southeast of the site, and groundwater monitoring. The second ROD
required that the pond adjacent to the site be filled to eliminate possi-
ble future impacts to indigenous and migratory fauna. Additionally, a
replacement pond was to be created elsewhere on the owner's property
in an area unaffected by the landfill.
Both the U.S. EPA and DNREC agreed that the first ROD should
be finalized in the interest of keeping the remedial process moving.
Because the environmental assessment of the pond was ongoing at that
time, it was decided a second ROD would be developed pending the
outcome of the assessment.
Negotiations with the Wildcat PRPs began almost immediately fol-
lowing finalization of the first ROD and prior to the second ROD (negoti-
ations with the PRPs prior to initiation of the RI/FS were unsuccessful).
During these negotiations (for implementation of the selected remedy),
the PRPs requested that the second ROD be completed prior to entry
into a consent decree and that a release from future liability for natural
resource damages, from both Federal and State trustees, be granted.
The Federal trustees for this site included the Department of Interior
(represented by the U.S. Fish and Wildlife Service) and the Depart-
ment of Commerce (represented by the National Oceanic and
Atmospheric Administration). The State of Delaware's trustee was
DNREC.
The second ROD was finalized and required filling the adjacent pond
and creating a replacement pond on an unaffected portion of the
property. The alternative selected for the pond represented a worst case
scenario and was selected partly because of the request for release from
liability for future natural resource damages from the PRPs. Both the
U.S. EPA and DNREC regarded the biological evidence for the pond
as somewhat inconclusive and placed language in the second ROD such
that the selected remedy would be re-evaluated if it was not implemented
by the PRP group.
The language granting the natural resource damage release was in-
corporated into a consent decree negotiated between the PRPs, the U.S.
EPA and DNREC. with considerable comment by the Federal and Stale
trustees. The wording of the release is as follows: "The United Stales
and the {State of Delaware] hereby waive and release with respect to
the Settlers any claim that they may have for damages to natural
resources at or arising from the Wildcat Site resulting from releases
or threats of releases at or from the site for which the Settlers are alleged
to be liable pursuant to Section 107 (a) of CERCLA or from the imple-
mentation of the Remedial Action pursuant to the Decree." Note that
the natural resource damages, as defined in the release language, can
be either from releases from the landfill or from implementation of
the selected remedy.
The damages associated with releases from the landfill include con-
tamination of groundwater. contamination of surface water and sedi-
ments, and bioaccumulation of inorganics in mummichogs (Fundulus
heieroclitus) and painted turtles (Chrysemys picta) in the pond. The
damages associated with implementation of the remedy include the loss
of wetlands around the periphery of the pond and potential effects on
rare plants on the site. It should be noted that 29 acs of wetlands were
originally lost at the time the landfill was operated due to the direct
placement of landfill wastes upon prior existing tidal wetlands. This
loss of wetlands occurred prior to the existence of Federal or State sta-
tutes protecting wetlands. Further, this loss is not the result of release
of any hazardous substances from the site nor from the remedial action.
Consequently, the Federal and State agencies did not pursue recovery
of these resources.
The Secretary of DNREC, as the State's designated trustee, required
the concurrence of numerous agencies within the Department, including
the Division of Water Resources (for groundwater, surface water and
wetlands), the Division of Parks and Recreation (for plants) and the
Division of Fish and Wildlife (for fauna). As discussed earlier, an effi-
cient internal mechanism was required to ensure the timely input by
these divisions prior to agreement by DNREC to grant the release. The
Division of Air and Waste Management, responsible for the Superfund
program, coordinated involvement of the other divisions in the remedial
process (RI/FS and ROD stages) and also in the subsequent develop-
ment of the remedial action work plan attached to the consent decree.
This assured that the State's natural resource concerns were adequate-
ly addressed during all stages of the process. Concurrently, the U.S.
EPA coordinated involvement of the Federal trustees, assuring that their
natural resource concerns were adequately addressed.
Both the Federal and State trustees were signatures to the consent
decree. These agencies will also be involved in review of the remedial
design prior to initiation of remedial action at the site.
Though the Federal and State natural resource trustees took a risk
in granting the release request, they were comfortable in doing so
because of their participation throughout the process. The consent decree
contains a reopencr clause should the remedial actions not be com-
pleted by the PRPs or not meet the requirements defined by the two
RODs. Nonetheless, the PRPs felt that the remedies would be success-
ful and agreed to enter into the consent decree in spite of the inclusion
of the reopener language.
CONCLUSION
The State of Delaware^ experience on the Wildcat Site has led us
to move forward on the development of a policy to address natural
196 COST & ECONOMICS
-------
resource damages resulting from any releases of hazardous substances
or oil. It has taught us the importance of early notification and coordi-
nation of natural resource trustee agencies to effect a favorable resolu-
tion to environmental issues surrounding Superfund settlements.
The proposed policy calls for DNREC to identify damages to natural
resources resulting from any release of hazardous substances or oil.
Upon discovery of the release, a central coordinator within DNREC
is notified. This individual then works with the appropriate DNREC
division responsible for managing the investigation and cleanup of the
release. Together, along with the Federal agencies in cases of joint
trusteeship, the assessment of any damages to natural resources is com-
pleted and, if necessary, the appropriate compensation is pursued. In
this way, the necessary natural resource damage assessments of inci-
dents ranging from slow releases at NPL sites to major spill events can
be addressed in a consistent and organized fashion.
The implementation of the remedies selected for the Wildcat Land-
fill represents a landmark for interagency, intra-agency and PRP co-
operation at a Superfund site. The development of an efficient
mechanism for the identification of natural resource damages and the
subsequent damage claim release procedures by the State allowed for
successful and timely completion of negotiations by Federal and State
government and private parties in rectifying environmental problems
at Superfund sites within the State of Delaware.
FOOTNOTES
1. Two sets of Natural Resource Damage Assesment rules were published by
DOI. The Type A rule, finalized on Mar. 20, 1987 in 52 FR 9042, addresses
assessments for spills of hazardous substances or oil in coastal and marine
environements. The rule uses a computer model to perform simplified
assessments. The Type B rule, finalized on Aug. 1, 1986 in 51 FR 21674,
addresses more complex assessments of damages in other environments.
2. A federal ruling made on July 14, 1989 requires DOI to revise both the Type
A and Type B natural resource damage assessment procedures.
3. See CERCLA/SARA Section 107 (f) (2).
4. CH2M Hill Southeast, Inc. Wildcat Landfill Remedial Investigation Report,
Volume 1, May, 1988.
5. Record of Decision. ROD Decision Summary: Wildcat Landfill Site, Kent
County, Delaware., June, 1988.
6. Record of Decision. ROD Decision Summary: Wildcat Landfill Pond, Kent
County, Delaware., Nov., 1988.
COST & ECONOMICS 197
-------
Characterization and Washing Studies on
Radionuclide Contaminated Soils
William S. Richardson, Ph.D.
Auburn University at Montgomery
Montgomery, Alabama
Tonya B. Hudson
S. Cohen and Associates, Inc.
Montgomery, Alabama
Joseph G. Wood
Charles R. Phillips
U.S. EPA
Montgomery, Alabama
ABSTRACT
Soils from sites in Monte I air and Glen Ridge, New Jersey, are con-
taminated with radium-226 and thorium-230. Barium-radium sulfaie.
partially extracted ores and other radiommerals, allegedly an artifact
from a radium extraction mill, are found mixed to varying degrees with
the native soil and constitute a radiological hazard characterized by
elevated levels of radon and gamma radiation.
Soil samples from the site were characterized with respect to
radionuclide distribution and particle size by wet screening and radio-
chemical analysis. In both soil samples, a significant amount of
radium-226 and thorium-230 activity is found in the smaller-sized soil
fractions Based on the results, a washing process thai includes vigorous
mixing of the soil with water and physical separation of particles by
size was developed.
INTRODUCTION
Soils from residential and business communities in Montclair stopped
and Glen Ridge, New Jersey, are contaminated with radium-226 and
thorium-230. The contamination allegedly was produced by a radium
extraction mill that operated nearby in the early pan of the century.
As a result of the subsequent use of this radium residue as landfill during
construction, approximately 300,000 yd' of soil on more than 95 ac
are contaminated; almost 1,700 people in more than 500 homes are
affected to some degree by elevated levels of gamma radiation and
radon-222 gas. The radon, produced by the radioactive decay of
radium-226, is of particular concern since it may enter homes con-
structed on areas containing contaminated fill material. The most
significant contaminants producing the gamma radiation are radon and
radium-226, ranging from approximately 40 to 1,000 pCi/g of soil, and
thorium-230, ranging from approximately 20 to almost 900 pCi/g':
The contamination is the result of the presence of process residue
containing barium-radium sulfate precipitates, partially extracted ores
and other radiominerals that are mixed to varying degrees with the native
soils1. Earlier studies on uranium mill tailings indicated that volume
reduction by physical separation and chemical extraction might be a
feasible means of remediation of the Montclair and Glen Ridge sites4.
PROCEDURES
Determination of Particle Size and Size Distribution
Wet sieving was performed on two soil samples from the Montclair
site and on one soil smaple from Glen Ridge. One Montclair sample
was labeled "Montclair," while the other was labeled "Representative"
since, as reported in COM Reports1', the specific activity of
radium-226 contamination in the latter represents an average value for
the overall sites. The soils were dried at 60°C and wet-sieved using
a Brinkman, model VS, vibrating siever.
Radiochemical Analysis
All soil samples and soil fractions were dried at 60°C and prepared
for analysis of radium-226 by gamma-ray spectroscopy using high purity
germanium detectors'. Radium-226 was identified and measured using
the 186 KeV photopeak. Since only very small quantities of uraruum-235
were found in the samples (less than 01%), interference by the
uranium-235 185 KeV photopeak was not a significant consideration.
Selected samples with low specific activity were counted after being
sealed in air-tight containers for 30 days to allow for equilibrium of
radon-222 and its daughters.
Soils and selected wet sieved fractions and samples were also analyzed
for thonum-2303. Aliquots were completely solubilized in acid
mixtures, and the thorium was separated by ion-exchange chroma-
tography and counted by alpha spectroscopy using thorium-234 as a
tracer to determine the chemical yield of the procedure.
Wash Studies
Four soil fractions were identified for wash studies: (+4 designates
material retained by a number 4 sieve), -4/+16 (-4/+16 designates
material that passes through a number 4 sieve but is retained by a number
16 sieve), -16/+30 and -30/+50. Based on the literature survey4, water
and several salt or salt/acid solutions were selected as initial wash
reagents (Table 6).
Samples of the selected soil fraction, prepared by dry screening on
a Gilson TM-4 screener. were analyzed for radium-22& The samples
were then mixed with water or the selected wash solution (S mL/g)
in a 1-gal container and shaken between 100 and 350 rpm on a Lab-
Line Orbit Shaker, model 3590, at room temperature for 1 hr. At the
end of that time, they were rinsed with water, and the solid residue
(Rl) was collected over the appropriate sieve (i.e., number 4 for a +4
soil fraction or number 16 for a -4/+16 fraction). The residue was dried
at 60°C, weighed and analyzed. The filtrate was subsequently filtered
through a Whatman No. 1 filter paper and then a 0.4m polycarbonate
filter. These residues were dried and weighed. The volume of the filtrate
was measured and, along with the residues, analyzed for radium-226.
Two
-------
of the study. In the second step, a new soil fraction was washed with
the filtrate from the first step. The filtrate from the second step was,
in turn, used to wash a third new soil fraction.
Combined Washing and Wet Sieving of Total Soils
Soil samples were weighed and analyzed for radium-226. After
analysis, the samples were mixed with tap water (5 mL/g) in a 1-gal
container and shaken at 350 rpm at room temperature for 30 min. The
soil mixtures then were sieved under vacuum on a Gilson Wet-Vac Sieve
Tester using the selected sieve sizes.
RESULTS AND DISCUSSION
Particle Size and Radiochemical Distribution
Table 1 gives the average specific activity of radium-226 and
thorium-230 in the Montclair, Glen Ridge and Representative soil
samples based on their dry weights. The specific activity of radium-226
is approximately six times higher in the Glen Ridge soil than in the
Montclair soil, while the Representative soil contains less than one half
that of the Montclair soil. Thorium-230 specific activity in the Montr
clair and Glen Ridge soils is equivalent to that of radium-226. The
thorium isotope specific activity in the Representative soil is
considerably less than that of the Montclair soil.
Table 1
Total Soil Radiochemical Analysis
Table!
Montclair Soil Wet Sieving
Soil
Montclair
Glen Ridge
Representative
Ra-226
(pci/gr)
132 ± 18
828 + 51
54 ± 10
Th-230
(pCi/gr)
123 ± 12
826 ± 28
18+2
Error for specific activity represents + 1 sigma.
The distributions of radium-226 and thorium-230 by particle size are
indicated in Tables 2, 3 and 4 for the Montclair, Glen Ridge and
Representative soil fractions, respectively. Note in Table 2 that the
radium-226 specific activity is moderate, less than 100 pCi/gr, in the
Montclair fractions larger than 600 micron (30 mesh size), but it
generally increases as the particle size decreases. There is a noticeable
increase between the -10/+16 and the -16/+30 fractions and between
the -200/+400 and the -400 fractions and an unexpectedly high value
for the -16/+30 fraction—more than twice the value of the preceding
fraction. The thorium-230 values are, with the exception of one fraction
(-16/+30), less than that of radium-226.
Table 3 shows that radium-226 is distributed in a similar manner in
the Glen Ridge soil, but the increase in specific activity is not as uniform
with decreasing particle size. There is, again, a noticeable increase from
the number 16 to 30 mesh size and from 400 to -400 mesh; a doubling
in activity, with a very high activity in the -400 fraction. Thorium-230
activity specific activity is also inversely related to the particle size with
the activity, doubling between the 16 and 30 mesh size and between
the 400 and -400 fractions. In each fraction, however, the thorium-230
is less than that of radium-226.
Table 4 indicates that the radium-226 is more evenly distributed in
the Representative soil, but an increase in specific activity is observed
with relatively significant increases from the -30/+50 fraction to the
-50/+100 fraction and from the -100/+140 to the -140/+200 fraction.
Each fraction contains less thorium-230 than radium-226.
The elevated specific activity in the fine soil material is clearly demon-
strated by these data. Thus, partial remediation of the soils by wet sieving
techniques appears to be feasible.
Ra-226
Size Weight Percent* (pci/gr)
+4
-4/+10
-10/+16
-16/+30
-30/+50
-50/+100
-100/+140
-140/+200
-200/+400
-400
*Percentage of
1.46% is trash
18.25
7.94
3.23
4.54
7.46
14.16
6.74
5.55
10.85
21.28
100.00
sieved
Percentage error for
44
26
39
84
117
113
138
170
194
382
material; 3.34%
specific activit
+ 20%
± 24%
± 31%
+ 15%
± 12%
+ 12%
+ 11%
± 8%
± 11%
± 8%
of soil
Th-230
(pci/gr)
7 + 6%
12 ± 9%
15 ± 8%
175 ± 4%
71 + 5%
62 + 5%
68 + 5%
115 ± 4%
132 + 4%
283 ± 5%
is large rocks and
y represents ± 2 sigma
Table 3
Glen Ridge Soil Wet Sieving
Size
+4
-4/H-10
-10/+16
-16/+30
-30/+50
-50/+100
-100/+140
-140/+200
-200/+400
-400
Ra-226
Weight Percent* (pci/gr)
31
9
3
4
5
11
5
4
7
15
99
.78
.74
.61
.93
.85
.09
.64
.02
.62
.70
.98
346 +
307 +
268 ±
535 +
492 +
472 ±
498 ±
677 ±
1,006 ±
2,855 +
9%
7%
10%
8%
5%
5%
5%
5%
4%
3%
Th-230
(pCi/gr)
76
154
108
211
289
302
365
500
987
2801
± 6%
± 4%
± 5%
± 4%
± 4%
± 4%
± 3%
± 3%
± 4%
+ 5%
*Percentage of material sieved; 0.65% of soil is large rocks and
0.30% is trash.
Percentage error for specific activity represents + 2 sigraa
error.
Tables 2 through 4 also summarize the particle size distributions of
the material sieved. In the Montclair soil, Table 2, approximately 30%
of the soil is retained by the number 16 sieve; 34% is retained up to
the number 30 sieve (600 micron). Table 3 indicates a similar trend
for the Glen Ridge soil. At least 45% of the' sample is retained up to
the number 16 sieve during wet sieving; 50% is retained up to the num-
ber 30 sieve.
Table 4 indicates that the Representative soil is similar to the Mont-
clair in distribution of particles by weight. However, it contains
approximately 10% more fine material (-400 mesh); unlike the Mont-
clair soil, no large rocks (> 2 in.) are present in the soil.
Soil Wash Studies
Examination of the distributions of radium-226 concentrations in the
Montclair and Glen Ridge soils along with data from the geological
characterization3 indicated that preliminary wash studies should be
performed on +30 soil fractions. These fractions had been separated,
CONTAMINATED SOIL TREATMENT 199
-------
Table 4
Representative Soil Wet Sieving
Size
+4
-4/+10
-10/+16
-16/+30
-30/+50
-50/-HOO
-100/+140
-140/+200
-200/+400
-400
Weight
15.
6.
2.
4.
7.
12.
5.
4.
10.
Ra-226
Percent* (pCi/gr)
79
70
65
74
73
29
55
56
48
29.51
100.
00
1< ±
22 ±
27 ±
25 ±
25 +
33 i
33 ±
52 ±
58 ±
105*
9*
9*
10*
9*
7*
5*
25*
16*
10*
Th-230
(PCi/gr)
5 ±
8 +
B ±
9 ±
16 ±
23 ±
23 i
39 ±
55 ±
"
9*
6*
6*
5*
5*
i\
5*
5*
4*
•Calculated fron total activity of the sample sieved and
percentage of the fraction.
••Not measured.
Percentage error for specific activity represents ± 2 signa
error.
by the methods described above, from Montclair and Glen Ridge soils
obtained from the New Jersey site in October, 1987.
Table 5 is a summary of the initial results of single-step wash studies
with water and gentle shaking. With one wash, water removes approxi-
mately 50% of the radium-226 activity from the +4 fraction and
approximately 85% of that in the -4/+16 fraction. The data indicate
that these results were primarily accomplished by removing fine soil
particles and suspending them in the wash water — note the weight
percentage of sample recovered during washing, especially the per-
centage of the -4/+16 and -16/+30 fractions recovered. It is not sur-
prising that less sample is recovered (more is lost) from the smaller-sized
fractions during washing, since more surface area is available for
adherence of fine material on these fractions, and this fine material
should be removed during the wash process. In addition, these smaller
fractions would be expected to contain a larger percentage of loose fine
material from dry screening than the larger-sized fractions. In each case,
the filtrate contains little to no activity (data not shown in Table S).
The final average specific activity of the Montclair samples ranges from
K) to 71 pCi/g. Although the Glen Ridge samples follow the same trend,
the final activity is well above 71 pCi/g (121 to 330 pCi/g), since the
activity of the samples initially is high.
TableS
Summary of Results from One-Step Ufesh Study with Water
Initial rlnal
•p. Act. «p. Act.
Ra-334 Ra-114
goil alia fpgi/^rl fpgl/prf
H
e
(.)
•4(a) it t
-4/*ll(b) 104 1
-la/*10(o) 141 1
»4<«) 1»1 1
-4/*ll(b) 190 ±
-U/.IOIC) 1,0.1 J
fUpraaanta tha avaraga and
7.0 10 1 1 . 7
14 11 i 1.1
19 71 i 1.7
17 121 t 21
54 111 i 91
17 110 t 44
atandard davlatlon
•aroant Might
of Total Paroant
Activity of faavla
92
14
• 1
40
• 2
17
at
t 7.0 14 i 9.1
1 1.4 49 t 1-0
1 J.« 14 i 1.9
i. 14 12 t 4.1
t 1.9 44 i 1.9
i 1.1 44 1 1.3
aavan runa.
Xnt./Plnal
•p. Act.
n-no
lagi/arl
17/1
104/19
191/99
1,097/11
•11/101
7*4/147
The final specific activity of the Moniclair +4 and -4/+16 fractions
indicates a promising trend for remediation by washing and screening,
since their average values after washing are 10 pCi/g and 33 pCi/g,
respectively. Thorium-230 values are lower than those of radium-226,
indicating that ingrowth of radium-226 would not be a long-term
problem.
In most instances, the salt solutions produce similar, and in several
cases slightly better, results (Table 6). The data generally indicate,
however, that, relative to water, salt solutions increase the activity of
radium-226 in the filtrate, apparently by solubilizing more of the
radionuclides.
Table 6
Summary at Results from One-Step Wash Study with Sato
Initial
•p. Act.
fta-114
Final
•p. Art.
Of TOtAl
nativity
*
•
-«/•
•*/•
-4/*
-4/*
.
*
I
-4/»
-4/»
-«/•
HaCl 19 ll
Kl 20 10
GBCl,/*^! 2t 14
•Dl* 24 11
• «acl >i 41
4 RC1
a CaCl»/llCl 1
• ton i
UC1
m
CaClj/HCl
4 *aci
• m
• OKC1./IIC1
K
14
12
104
14]
113
If 920
11 1*4
M 119
-4/*ll nT*" 1) 244
104
71
911
10*
1«7
1,1*0
•
191
769
4. Ml
•
0
1.4tl
9.UO
•
40
70
94
a4
74
• 1
M
to
91
99
44
40
«}
M
fl
11
II
><
14
19
4*
44
a
M
H
u
n
70
19
42
a«
(b) R«pr«>anti tha avaraga and atandard davlatlon of algtat runa.
(0) Rapraaanta tha avaraga and atandard davlation of four runa.
Shaking valoclty waa 100 rpa.
arvaklna velocity ma IOC rpai.
An important consideration in a large-scale remediation process using
water is the amount of water required. If the wash water can be recycled,
an appreciable amount of water will be conserved during volume reduc-
tion. Further, recycling will avoid the necessity of disposal or treat-
ment of large volumes of radioactive liquids. In a study designed to
examine the feasibility of water recycling, a -4/+16 soil fraction was
washed first with deionized water; the filtrate was collected after filtering
through a micropore filter and used to wash a new -4/+16 fraction.
The filtrate from the second wash was used, in turn, to wash another
new fraction. In each step of the wash process, the same percentage
of activity is removed leaving samples with comparable specific
activities. The activity of the filtrate in each case is less than S pCi/L.
Thus, the study indicates that wash water filtered through a micropore
filter to remove suspended particles may be recycled at least twice with
no significant decrease in removal efficiency.
The effect of two- and three-step washing also was examined. With
each fraction, the study indicates that the two-step process, compared
to the one-step process, removes a greater percentage of radium-226
activity. Like the single-step procedure, each step of the process removes
some mass from the sample. The first step removes the majority of
the associated fines, but visual examination of the sample after two wash
steps indicates that the material has less fine particles associated with
it than does a comparable sample washed only once. The loss of material
during the second wash step is approximately 5% of the initial sample
weight. In every experiment, the specific activity of the filtrate is less
than 5 pCi/L. The results of the three-step wash study with water indi-
cate that only a very small amount of additional sample is removed
by the third wash step. Examinations of the residues from the two- and
three-step studies support this observation since there is no visual
physical difference in comparable residues. There is no significant
increase in the loss of total activity of the samples after the third wash.
and the specific activity is essentially the same.
A preliminary study of washing rocks with water was initiated. Similar
to the +4 soil fractions, the geometry of the rock sample presents more
of a problem for radium-226 analysis by gamma-ray spectroscopy than
those of smaller fractions. The Montclair rocks, however, indicated a
specific activity of less than 15 pCi/gr and were not washed. On the
other hand, the Glen Ridge rocks with more coal-like and coaly-slag
200 CONTAMINATED*SOIL TREATMENT
-------
character have a specific activity of 260 ±217 pCi/g, but the wash study
is not conclusive.
Combined Washing and Wet Sieving Studies of Total Soils
The results of the wet sieving and water-wash studies indicated that
the examination of a combination of the two processes applied to a total
soil sample would be appropriate. The results presented in Table 7
demonstrate that by combining vigorous shaking (350 rpm) with vacuum
sieving, up to 35 % of the Montclair soil can be separated with an average
radium-226 specific activity of 15 pCi/g, a specific activity very simi-
lar to that obtained in the preliminary studies. With the inclusion of
the -50/+100 fraction, however, almost 43% of the Representative soil
can be recovered with a radium-226 specific activity of 15 pCi/g. It
is important to note that 56% of this soil sample can be recovered with
a specific activity of 16 pCi/g and 67% can be recovered at 19 pCi/g.
Table 7
Final Studies of Vigorous Shaking and Subsequent Sieving of
Soils on the Wet-Vac Siever
CONCLUSIONS
Considering the need to develop a simple, safe, economical, on-site
treatment process that would produce a significant volume of remediated
soil to remain on-site, the results of these studies indicate that water
washing is a prime candidate for a process that meets these criteria.
Using water exclusively would eliminate the necessity for removal of
salt and/or acids by processes that would require one or more steps,
possibly including, among others, ion-exchange, neutralization, or
precipitation. Since the data indicate that little radium-226 is present
in the filtrate after washing the soil fraction up to three times with water,
it is likely that the water could be disposed directly or, more impor-
tantly, be recycled several times during the washing process. Thus, a
wash process that would include wet screening of the Montclair soils
to separate the +50 or +100 fraction would be followed by filtration
of the -50 or -100 fraction to remove wash water that in turn would
be recycled in the process. The -50 or -100 fraction could then be col-
lected for disposal or additional treatment.
-100/+200
-200/+400
Height
Percent
Ra-226
(pci/gr)
Weight
Percent
Ra-226
(pCi/gr)
13.46 22
67.49 19**
32.5^ 180
13.61
13.12
28.02
100.00
Height
Percent
Ra-226
(pci/gr)
+4
-4/+16
-30/+50
-50/+100
11.06
5.59
4.10
7.99
leUia
12
21
14
14
ii
21.94
5.69
2.67
_4^4S
34.79*
10.46
15
15
16
^1B
15**
42
18.68
11.73
2.91
5.52
38.84*
11.63
102
151
175
Ifii
134**
174
11.41
8.23
246
484
Th-230 epecific activity for each fraction wee lese then the specific ectivity of Ra-226.
•Cumulative weight percent.
**Heighted average of specific activities of above fractions.
Although vigorous shaking and wet sieving with vacuum do not
produce a sufficiently remediated Glen Ridge soil, the process does
separate approximately 55 % of the soil (+30) with less than half the
specific activity of a sample that has been shaken gently (125 rpm),
120 pCi/g compared to 290 pCi/g.
REFERENCES
1. Remedial Investigation Study for the Montclair/West Orange and Glen Ridge,
New Jersey Radium Sites, Vol. I, Camp Dresser and McKee, Inc., Roy F.
Western, Inc., Clement Associates, Inc., ICF, Inc., U.S. EPA Contract No.
68-01-6939, New York, Sept. 13, 1985.
2. Appendices for Remedial Investigation Study for the Montclair/West Orange
and Glen Ridge, New Jersey Radium Sites, Vol. n, Camp Dresser and McKee,
Inc., Roy F. Weston, Inc., Clement Associates, Inc., ICF, Inc., U.S. EPA
Contract No. 68-01-6939, New York, Sept. 13, 1985.
3. Nieheisel, J., Characterization of Contaminated Soil from the Montclair/Glen
Ridge, New Jersey Superfund Sites, Inhouse Report, U.S. EPA, Office of
Radiation Programs, Washington, DC, 1988.
4. Richardson, III., R.S., Snodgrass, G.B. and Neiheisel, J., Review of Chemi-
cal Extraction and Volume Reduction Methods for Removing Radionuclides
from Contaminated Tailings and Soils for Remedial Action, U.S. EPA, Office
of Radiation Programs, Analysis and Support Division, Washington, DC and
Eastern Environmental Radiation Facility, Montgomery, AL, U.S. EPA, July
24, 1987.
5. Lieberman, R., ed., Eastern Environmental Radiation Facility Radiochemistry
Procedures Manual, U.S. EPA, Report No. 520/5-84-006, Montgomery, AL,
June, 1984.
CONTAMINATED SOIL TREATMENT 201
-------
Evaluation of U.S. EPA Soil Washing Technology for
Remediation at UST Sites
Richard P. Traver, P.E.
Anthony N. lafuri, P.E.
U.S. EPA
Release Control Branch
Edison, New Jersey
Myron S. Rosenberg, Ph.D., P.E.
William K. Glynn
Mary E. Tabak, P.E.
Michael Whitehead
Camp Dresser & McKee Inc.
Boston, Massachusetts
M. Pat Esposito
Bruck, Hartman & Esposito
Cincinnati, Ohio
ABSTRACT
It has been estimated that approximately three to five million under-
ground storage tanks in the United States are used to store liquid
petroleum and chemical substances. Further estimates indicate that
100.000 to 400,000 of these tanks and their associated piping systems
may be—or have been—leaking. The resulting soil and groundwater
contamination, especially to primary drinking water aquifers, have left
the United States with a massive cleanup problem, and in some cases,
with the necessity of abandoning water supply wells for indefinite periods
of time.
The U.S. EPA through its Risk Reduction Engineering Laboratory's
Release Control Branch has undertaken research and development efforts
to address the problem of remediating contaminated soils resulting from
leaking underground storage tanks. Under this initiative, (he Releases
Technology Staff is currently evaluating soil washing technology as an
economically viable and technically feasible cleanup remedial alterna-
tive to the current practice of hazardous landfill disposal.
Soil washing is a high energy, dynamic physical volumetric reduc-
tion and feedstock preparation process in which soluble contaminants
are extracted from the solid fraction into liquid medium, usually water.
In addition, the separation of the highly contaminated fine soil pani-
cles (silts, clays and colloids) from the bulk of the soil matrix is
accomplished through the mechanism of volume reduction. As a result,
significant fractions of the contaminated soil can be "cleaned" and
returned into the original excavation or used as cleaned "secondary"
fill (i.e., road beds, bridge foundations) or aggregate material for con-
crete and asphalt production. Since the contaminants of interest are typi-
cally concentrated in the fine soil fractions, their separation and
segregation from the bulk soil increases the overall effectiveness of the
process. Treatment of the spent wash solution prior to recycling is
required. The "enriched" contaminated "fines" fraction has now been
readied for an appropriate ultimate treatment technology such as solidifi-
cation/stabilization, biological treatment, solvent extraction, low tem-
perature desorption, incineration, etc.
The soil washing program is evaluating the effectiveness of soil
washing technology in removing petroleum products (unleaded gaso-
line, diesel/home heating fuel and waste crankcase oil) from soils and
treating the generated residuals. The program consists of testing of soil
washing technology at the bench scale, pilot scale, and through filed
demonstrations in order to develop the applicability and design criteria
for full scale implementation as a long-term corrective action at leaking
underground storage tank sites.
INTRODUCTION
Based on the Hazardous and Solid Waste Amendments of 1984 and
its Land Ban Regulations, (he U.S. EPA has discouraged the excava-
tion and landfill disposal practices of the past for contaminated soils
resulting from leaking underground storage tanks (USTs). The U.S. EB\
has encouraged the use of on-site treatment technologies, however,
problems have plagued the development of on*site treatment technolo-
gies for the treatment of petroleum contaminated soils. Technical sup-
port is needed to develop effective long-term corrective actions at leaking
underground storage tank sites, design cleanup program guidance, and
help implement state programs.
The remedial options available for the treatment of contaminated soils
from UST sites are broadly segregated into two main categories, namely
those which remove the contaminants without excavation (in-situ tech-
niques) and those which require excavation of the soil and subsequent
cleaning on-site. The former group of remedial options have not yet
been demonstrated for high efficiency removal of contaminants from
the subsurface. These techniques are plagued by the uncertainty of soil
contamination levels in the subsurface after treatment. Soil excavation
followed by extensive cleaning of (he soil will ensure a more complete
and expedient removal of contaminants over in-situ techniques which
require long periods of time.
On-site soil washing of excavated soils is a viable alternative to in-
situ techniques and has been shown to be effective for the cleanup of
applicable Superfund and leaking underground tank sites. The goal of
this effort is to demonstrate the feasibility of soil washing for cleaning
up petroleum contaminated soils.
The U.S. ER\ developed soil washing technology is a physical process
in which excavated soils are contacted with an aqueous based wash
solution with selective additives predicated upon the soil characteris-
tics and specific contaminants of interest. The highly water soluble con-
taminants in the soil matrix are extracted from the solid fraction into
the liquid medium. The two principle cleaning mechanisms include
the dissolution of the contaminants into the extractive agent and/or the
dispersion of the contaminants into the extraction phase in the form
of panicles (suspended or colloidal).
In addition, the separation of the highly contaminated fine (>74
micron) soil particles (silts, clay and colloidal) from the bulk of the
soil matrix is accomplished through the mechanism of volume reduc-
tion. As a result, a significant fraction of the contaminated soil is cleaned
and can be put back into the original excavation following testing and
202 CONTAMINATED SOIL TREATMENT
-------
approval by the appropriate lead agency. Since the contaminants are
more concentrated in the fine soil fractions due to their typically higher
cation exchange capacity and "relatively" hugh surface areas, their
removal from the bulk soil increases the overall effectiveness. Subse-
quent treatment is typically required for the spent wash waters and the
fine soil fractions. The information developed from this project is
assisting the U.S. EPA in defining the criteria for developing soil washing
as a long-term corrective action at leaking underground storage tanks.
PROGRAM STRATEGIES
Under Phase I of the U.S. EPA's research program, a surrogate soil
matrix containing a range of petroleum products at varying concentra-
tion levels was prepared and subjected to bench-scale performance evalu-
ations of soil washing technology. This paper covers the formulation
and characterization of the U.S. EPA's surrogate soil matrix described
as the Synthetic Soil Matrix (SSM).
Prior to spiking the full scale quantities of SSM, several bench scale
experiments were performed to develop a dose/response relationship
between the quantity of petroleum product added to the soil matrix and
the analysis quantification. The petroleum products evaluated during
this study include unleaded gasoline, diesel oil and waste crankcase
oil. The full scale SSM was then blended with a specific quantity of
petroleum product to obtain a predetermined concentration level. TPH
analysis was performed to verify the concentration levels for diesel and
waste oil and BTEX analysis was performed to verify the concentra-
tion levels for gasoline.
The bench-scale washing experiments were designed to simulate the
U.S. EPA-developed pilot-scale Mobile Soils Washing System (MSWS)
or also known as the "mini-washer." Bench-scale experiments simu-
late the pilot- and full-scale drum-screen washer which separates the
>2-mm soil fraction (coarse material) from the <2-mm soil fraction
(fines) by use of a rotary drum screen. In the pilot and full scale system,
high pressure water knives operate at the head of the system to break
up soil lumps and strip the water soluble contaminants of the soil par-
ticles and separate the highly contaminated fines from the cleanable
coarse fractions.
SYNTHETIC SOIL MATRIX CHARACTERIZATION
The basic formula for the SSM was determined by the U.S. EPA under
the Best Demonstrated Available Technology Program from an exten-
sive review of contaminant groups and soils types found at Superfund
sites throughout the United States. The SSM was blended from a
predetermined mixture of clay, silt, sand, top soil and gravel in two
15,000 pound batches.
A review of the existing soil characteristics were made and additional
tests were conducted to further delineate the physical and chemical
properties of the SSM. The tests included particle size distribution,
moisture retention curve, Atterberg limits, cation exchange capacity,
base saturation, organic matter, chemical constituents and mineralogy.
Quantification and assessment of these specific properties will assist
the technical community to understand the differences that may be
observed between the performance of soil washing technology on the
SSM and on actual site specific UST site soils.
The SSM is composed of 60 percent sand, 19 percent silt and 21 per-
cent clay as determined by particle size distribution analysis (Table 1
and Figure 1). Based on this composition the SSM would be classified
(USDA) as having a sandy clay loam texture. Particle size distribution
data may be used to estimate hydraulic properties (Mishra et al., 1989),
residual saturation (Hoag and Marley, 1986), capillary movement, bulk
density, and surface area of the soil prior to more extensive analyses.
The moisture content of the SSM ranged from 33.1 percent at satu-
ration (0 bar) to 8.7 percent at the permanent wilting point (15 bars).
The moisture content at field capacity (0.1 bar) was 21.0 percent. The
moisture-retention curve (Fig. 2) developed from the moisture content
data was indicative of a finer textured soil. The moisture content data
can be used to evaluate moisture and chemical characteristics of the
SSM. For example, the amount of soil water that can be extracted from
the SSM under typical environmental conditions (0 to 15 bars) will be
Table 1
U.S. EPA Synthetic Soil Matrix Particle Size Distribution (USDA)
USDA(%)
USCS (%)
GRAVEL
SAND
TOTAL
V. COARSE
COARSE
MEDIUM
FINE
SILT
CLAY
60.0
16.0
8.8
11.7
23.5
19.0
21.0
58
15.2
26.8
i i < i i ! ; | '
a - -i
a
a
>B
IB
38
3B
7PI
IB
0
2e
I
1 —
l\.
a 100
I \
i \ \
1:
i'l
! !
10. 6
I -. i
4-
1 \
)V
f
— ; .
1 .0
13 ? *
| _ |.
! 1
i i
: j
N
; i
i i
: i
: j
— -
! |
1
0. 1
~
0.31
. V^
—
— -
G.eo
GRAIN SIZE -
Figure 1
U.S. EPA Synthetic Soil Matrix Particle Size Distribution Curve
24.4 percent. The remaining soil water is considered as "unavailable"
which can be removed by artificially induced vacuums or pressures.
Some similarities exist between the moisture content and residual satu-
ration of petroleum hydrocarbons in the soil. Generally, stronger com-
petitive adsorption of water for soil occurs and displaces non-ionic
organic chemicals that are present in petroleum hydrocarbons (Chiou
Malrlc Palenllal(.bars)
Figure 2
U.S. EPA Synthetic Soil Matrix Moisture-Retention Curves
CONTAMINATED SOIL TREATMENT 203
-------
ei al, 1989). Residual saturation will be dependent on moisture con-
tent and decrease with increasing moisture.
Analysis of the SSM for concentrations of various exchangeable ions
indicated that phosphorous was moderate, potassium was low, and both
magnesium and calcium were very high (Table 2).
Table 2
Chemical Characteristics of SSM
Table 3
UST-Soil Wishing Dote/Response Bench-Scale TwU
PARAMETER
ORGANIC MATTER
PH
CATON EXCHANGE
CAPACITY (CEC)
TYPICAL
UNITS W>NGŁ
% 13 0 4 100
8 4
meqnOOg 21 7 4 34
BASE SATURATION
ca
Mg
K
H
AVAILABLE PHOSPHORUS
WEAK BRAY
NaHCOS
POTASSIUM
MAGNESIUM
CALCIUM
ppm
ppm
ppm
ppm
99%
86.2
124
1.3
0
20
31
1 12
324
3740
The pH of the SSM was 8.0. This pH value is generally the result
of the presence of bases such as Ca2* and Mg1* ions. Such bases may
be readily removed with addition of water and/or other ions which may
displace the Ca1* and Mg1* ions and thus lower the pH. The pH of
the SSM will in part effect the CEC of the pH dependent fraction of
the soil (primarily organic matter), and the adsorption of metals.
Mobility of most metals such as lead (Pb) will be minimal as long as
the SSM pH exceeds a value of 6.5.
Cations exchange capacity of the SSM was 21.7 meq/lOOq. This CEC
value is somewhat typical of soil with a texture finer than a sandy loam
or with elevated organic carbon content. Determination of CEC is es-
sential in the evaluation of the fate and transport of charged ionic species,
but will have little influence on the non-ionic organic compounds present
in petroleum hydrocarbons.
The base saturation of the SSM was 99.9 percent, and was dominated
by the Ca1* ion (86.2 percent). Addition of water or any leaching
solution should considerably reduce the base saturation of the SSM
as the Ca2* is replaced by H* and Al'* ions.
The results of the dose/response tests are shown in Table 3. The lab
tests indicate that the soils reach a level of liquid saturation at about
23% liquid (both water and gas or diesel). The tests were conducted
such that the soils were all prepared to a 20% water level. However.
this limited the amount of gas or diesel which could be mixed into the
soil mixture.
At 20% water, the highest achievable BTEX concentration was about
3000mg/kg. For diesel, at water content of 20%, the highest TPH con-
centration was 60,000 mg/kg.
The dose/response curves are plotted in Figures 3 and 4. A linear
regression of the data yielded the following relationships for
dose/responses:
Gasoline:
G = gasoline concentration, mg/kg
B = BTEX concentration (sum of benzene, toluene, ehtylbenzene and
Eq. (2) G 13.33(8) - 375
The correlation coefficient r, for this equation is 0.998.
Diesel:
II »
II 1
II I
I tt*
It *K
D = diesel fuel concentration, mg/kg
T = TPH concentration, mg/kg
Eq. (3) D = 0.675(T) + 6268
The correlation coefficient r, for this equation is 0.995.
10000 40000 00000 10000 100000
Qtoolln*
Figure 3
Dose/Response values for Gasoline
« TPHCoramntta
10000 tOOOO 30000 40000 80000
Dlcotl (mg/kg)
Figure 4
Dose/Response values for Diesel Fuel
204 CONTAMINATED SOIL TREATMENT
-------
An additional experiment was conducted to determine if a higher level
of BTEX could be achieved if a lower moisture content was used. In
this experiment, a sample of soil was moistened to a 10% water con-
tent and then saturated with gasoline. The results, shown in Table 4
as "Jar 9," indicated that the BTEX content was increased to 4670 mg/kg.
The percent overdose required to achieve desired BTEX concentra-
tions for gasoline are shown in Table 4. Theoretical BTEX added to
the soil was calculated by assuming that the gasoline used for these
tests consisted of 16% BTEX by weight, as referenced in Table 4.
Theoretical amount of BTEX was compared to the lab data to obtain
the percent overdose. The amount of overdose obtained based on the
above assumption ranged from 80% to 194%.
Table 4
UST—Soil Washing Gasoline/BTEX Overdose Results
Jar No. Theoretical
Added BTEX
Concentration"
1 976
2 4,829
3 9,335
4 2,343
9 13,753
Lab Determined
BTEX
Concentration
(mg/kg)
406
2,200
4,420
1,280
4,670
% Overdose
140
120
11,1
83
194
This number represents 16% of the added gasoline concentration as referenced by
Hoag, G.E.; Bruell, C.J.; Marley. M.C., 1984. A study of the mechanisms controlling
gasoline hydrocarbon partitioning and transport in groundwater systems.
Storrs, CT: Institute of Water Resources, University of Connecticut. Prepared for U.S.
Department of the Interior, Geologic Survey Reston, VA. Project No. USGSG832-06,
NTIS NO. PB85-242907.
Equations 1 and 2 can be used to determine the amount of diesel
or gasoline to add the SSM to reach the desired concentrations. Based
on these calculations, estimates were made to determine the amount
of gasoline and diesel fuel to add to the SSM to obtain the desired con-
centrations of BTEX and TPH for the bench scale experiments. The
SSM blends were prepared in the U.S. EPA SSM Blending Facility in
Edison, NJ in 50 Ib batches for use in the bench scale soil washing
experiments.
SUMMARY AND CONCLUSIONS
With the SSM being fully characterized and the dose/response tests
completed, the next step was to conduct bench scale soil washing tests.
These experiments involved washing the SSM spiked with gasoline,
diesel fuel, or waste crankcase oil under several operating conditions
to obtain sensitivity analysis curves on various parameters affecting soil
washing efficiency.
The experiments were conducted by contacting approximately 1400
g of soil with varying amounts of washwater. The contact time varied
according to experiment as did the rinsewater volume. The washing
of the soils was conducted by shaking the soil and washwater in a
2-gallon jar in a shaker table operating with a stroke and frequency
of 1.6 inches and 4 Hz respectively. The rinsing of the soils was per-
formed in a Gilson Wet-Vac Model WV-1 which both rinsed the soils
as well as separated the particles into three fractions using No. 10,
No. 60 and No. 140 sieve trays. The process of the washing and rinsing
yielded five distinct fractions - the soils on the three sieve trays, a wash-
water, and a rinsewater. All fractions were measured for mass (or
volume) as well as contaminant concentration. A measure of total BTEX
(benzene, toluene, ethylbenzene and o-, m-, and p-xylenes) was used
on gasoline spiked soils, and total petroleum hydrocarbons (TPH) was
used on diesel spiked soils.
Preliminary screening tests were conducted on soils spiked with diesel
and gasoline to determine the optimum conditions for contact time,
washwater volume, rinsewater volume and washwater temperature.
Figures 5 through 8 present some of the data obtained from the screening
experiments. The results of the screening indicate that the optimal wash-
water parameters for SSM spiked with diesel and gasoline are: 20 to
30 minute contact tune, 1:1 soil to washwater mass ratio, 3:1 rinsewater
to washwater volume ratio, and ambient temperature for the washwater.
These conditions resulted in a 90+ % removal of TPH and BTEX in
the No. 10 and No. 60 sieve fractions.
No. 60 Siev
No. 140 Sle
Type 01 SSM: High Dle5el
Constant Paremeters:
Additive - None
Soil : Washwater . 1:1
Rinsewater : Washwater * 3:1
Wash Temperature - 77-84 F
10 15 20 25 30 35
Contact Time (mln.)
Figure 5
Contact Time Effect on Percent TPH Removal
° No. 10 Sieve
D No. 60 Slave
Type ol SSM: High Diesel
Constant Parameters:
Rinsewater : Washwater - 3:1
Additive - None
Wash Temperature * 77-84 f
Contact Time - 30 mln.
Washwater to SSM Mass Ratio
Figure 6
Washwater to Soil Ratio Effect on Percent TPH Removal
No. 10 Sieve
No. 60 Sieve
No. 140 Sieve
Type ol SSM: High Gas
Constant Parameters:
Additive • None
Soil : Washwaler Ratio - 2:1
Wash Temperature - 77-84 F
Contact Time - 30 mln.
Rinsewater to Washwater Volume Ratio
Figure 7
Rinsewater to Washwater Ratio Effect on Percent BTEX Removal
CONTAMINATED SOIL TREATMENT 205
-------
o No 10 SHVI
0 No W Suv.
• No 140 Stov*
TyB. 01 SSM Hlgn 0»
Coniltnl Piftmtwt
Adaillv* • None
ConlftCI Tlmt 10 mm.
20 40 to >o 100 1:0 140 i(o 110 100
Tampvratur* (F)
Figure 8
Temperature Effect on Percent BTEX Removal
It should be noted that the conditions stated above represent the most
cost-effective operating conditions for bench scale treatment of SSM
using soil washing technology. Operating conditions for each site soil
may vary and should be determined on a case by case basis.
The SSM can be characterized as somewhat alkaline, sandy clay loam
with a moderated CEC and low organic matter content. The alkaline
nature of the SSM is due to the presence of dolomitic limestone. The
clay formulation of the SSM is only partially represented by kaolinite
and montmorillinite. The swelling of the SSM is minimal.
Soil washing of the SSM should decrease the pH and base saturation
while removing a considerable amount of the Ca14 and Mg1*. As the
limestone is removed, the texture of the SSM should become coarser.
Organic carbon determinations should be made only after removal of
the inorganic carbon.
The dose/response tests provided the necessary information to
determine how much gasoline and diesel fuel should be added to the
SSM to obtain the desired concentrations of BTEX and TPH respec-
tively.
Preliminary soil washing bench scale screening tests indicate that
removals of greater than 90% of BTEX and TPH can be obtained for
soils in the No. 10 and No. 60 sieve fractions using the following oper-
ating conditions: 20-30 minute contact time, 1:1 soil to washwater mass
ratio, 3:1 rinsewater to washwater volume ratio, and using washwater
at ambient temperature.
Further work is being conducted to determine what, if any, effect
additives to washwaters have on the removal of TPH on soils spiked
with waste oils. The additives being investigated are CitriKJeen (an
organic-based solvent), and a surfactant. Bench scale experiments will
also be conducted using actual site soils where leaking underground
storage tanks have resulted in soils contaminated with gasoline and
diesel. The results and conclusions of these experiments will be
published in future papers.
These initial results show great promise in providing a feasible and
cost-effective technology to the user community in the remediation of
soil contaminated with petroleum hydrocarbons from leaking under-
ground storage tanks, large buck storage tank farms, refineries and
associated transportation and handling accidents. Stay tuned for fur-
ther developments!
REFERENCES
I. Chiou, CT. "Theoretical considerations of the partition uptake of oonkmk
organic compounds by toil organic matter." In B.L. Sawhney and Brown (ok.).
"Reactions and movement of organic chemical in soil." Soil Science Society
of America, Inc. American Society of Agronomy, Inc. Madison, WI, 1989.
2. Hoag. G.G. and Marley. M.C. "Gasoline residual saturation in iinsanirated
uniform aquifer materials" / of Em'I Eng.. 112 (3): pp. 586-604. 1986.
3. Lindsay. W L. "Chemical Equlibha in Soils" John Wiley A Sons. New tori.
NY. 1979.
4. Mishra, S.. Parker iC. and Singhal. N.. "Estimation of soil hydraulic proper-
ties and their uncertainly from particle size distribution data "7. of Hydrology,
108 pp. 1-18. 1989
5. Nelson, D.W. and Sommery L.E . "Total Carbon, and organic matter." In
A L Page (ed.), "Methods of Soil Analysis Part 2. Agronomy monognphy
No. 9 (2nd ed.)." Soil Science Society of America. Inc. Madison, WI, 1982.
6. PEI Associates. Inc. "CERCLA BOAT SARM Preparation and Results of
Physical Soil Washing Experiment." 1988.
7. Traver. R.B, "Development and Use of the EPA's Synthetic Soil Matrix
(SSM/SARM)" US. EPA Releases Control Branch. Risk Reduction
Engineering Laboratory. Edison. NJ. 1989
206 CONTAMINATED SOIL TREATMENT
-------
Bench- and Pilot-Scale Case Studies for Metals and
Organics Removals from CERCLA Site Soils
Marilyn E. Kunze
John R. Gee, RE.
EEC, Inc.
Philadelphia, Pennsylvania
ABSTRACT
Efforts are being made to devise technologies and treatment systems
to remediate contaminated soil-on site without generating significant
wastes for off-site disposal. Two technologies under current study are
washing excavated soils in above ground treatment units (soil washing)
and flushing soils in place, without excavation (soil flushing). A recent
bench-scale soil washing study was performed on soil samples with
high organics and metals contamination obtained from a CERCLA site.
The soil washing study demonstrated greater than 90% removal of a
large number of the contaminants using various surfactant, organic
solvent and acid washing solutions. For a second CERCLA site, a pilot
soil flushing study is being conducted by passing water, or water with
modifying additions, through columns of undisturbed soils. The three-
phased soil flushing study has, to date, demonstrated significant removals
of metal and organic contaminants. The pilot study has also yielded
preliminary identification of relationships between various flushing pro-
cess parameters, which will be confirmed in the third phase of the study.
INTRODUCTION
In concept, soil washing consists of applying a solvent solution to
excavated soil placed in an above ground treatment system and pro-
cessing the soil until adequate amounts of contaminants are removed.
The process of soil washing, depicted in Figure 1, consists of segregating
excavated soils into appropriate size fractions, feeding the soils into a
tank containing a solvent and allowing the solvent to dissolve soil con-
taminants into the liquid solvent phase'. The excavated soils may
require dewatering to remove excess liquids prior to washing the soil
solids. These steps are followed by separating the resulting solid phase
for further treatment and/or disposal. The treated solids may require
dewatering prior to disposal. Frequently, the used solvent (or soil
washing solution) is collected and treated to allow recycle back to the
treatment system to reduce costs.
Transfer of contaminants from the soil solids to the liquid phase can
occur by dissolution, chelation or shearing of the contaminants bound
to a soil matrix due to the action of the solution. The exact nature of
the soil washing solution required depends on the chemical nature of
the contaminants to be removed and the mineralogy of the soil.
Selection of the optimum washing process would be based on data
derived from treatability studies, beginning with the type described in
this paper, as well as pilot study for testing the most promising process
options. Process options for consideration would include: a continuous-
mix batch reactor, a high-pressure washer, a soaking system, a counter-
current or concurrent flow system or any combination of these processes.
Soil flushing, as described in this paper, is the application of a solvent
solution (usually water) to the ground surface (or at depth) of an un-
excavated, undisturbed soil, allowing the solution to percolate downward
Figure 1
Concept for Soil Washing
and "flush" the entire soil contamination zone. A number of methods
could be used to apply the solution, including infiltration by surface
ponding, subsurface drainage fields, spray irrigation or pumping to the
subsurface using wells or well points. At the base of the contamination
zone, the flushing effluent is recovered at the groundwater table using
subsurface drainage pipes, trenches, wells or well points. Frequently,
the effluent from flushing is treated to allow recycle back to the flushing
system. In the majority of cases, water is used as the flushing solution,
although dilute acids or bases, chelating agents, selected minerals,
aqueous surfactant solutions and organic solvents have been suggested.
In this paper, a soil washing and a soil flushing bench-scale study
are discussed and the results from each are presented. Each study was
conducted on soils from separate hazardous waste (i.e., CERCLA) sites.
SOIL WASHING STUDY
Site Conditions
The Soil Washing Study was conducted on soil from a non-operating
commercial tract of land (approximately 6 ac) located within the
Piedmont geologic province of New Jersey. As shown by the cross-
section in Figure 2, materials encountered at successively lower depths
in the subsurface are: (1) fill soil having variable compositions and
particle sizes, (2) peat, (3) silt and (4) clay2 Over the site area, the
fill soil ranges from about 3 to 11 ft.
Figure 3 presents a representation of the fill soil components which
include natural soil particles (including clay and peat) as well as waste
fragment materials from site construction. The total volume of
contaminated fill and peat is approximately 115,000 yd3 (c.y.). The site
is covered with construction debris. A sludge disposal or spill area
contains dark, greasy sludge underneath a dry soil crust. A 10,000-gal
waste tank, which contains less than 20 c.y. of highly contaminated
sludge, is present on-site.
Information from past site investigations indicates that there are three
groundwater aquifers on-site: (1) a water table aquifer, (2) a till aquifer
and (3) a bedrock aquifer. The water table is typically only 2 ft below
CONTAMINATED SOIL TREATMENT 207
-------
UPROXIMATI
IHVAT1ON
(FMI FROM GROUND
+ 8
+ 6 '
+ 4
+ 2
0
-2
-4
-6
CONITRUCTIOH
7 MI ..4
_
B-1 SOIL OR WELL
BORINQ
Figure 2
Cross Section of Site for Soil/Wasle Washing Study
(Reference: ERM, Draft Feasibility Study)
WASTE PARTICLE
OLA88 SHARD
-PEAT
FRAGMENT
SAND PARTICLES
Figure 3
Representation of Fill Soil Components
(Enlarged)
the ground surface. To date, possible groundwater contamination in these
aquifers has not been fully characterized.
Study Objectives
The overall objective of the Soil Washing Study was to make a
preliminary evaluation of onsitc soil washing technologies for
remediating the contaminated Till soil, as well as tank and pit wastes.
at the CERCLA site. Tank and pit wastes were evaluated since their
relatively small volume may make them candidates for coprocessing
with the soil. The specific objectives of the Soil Washing Study were
to provide a preliminary indication of the following:
• The feasibility of extracting metals and organic chemicals from the
site soil and waste
• The type(s) of solvents that may remove significant percentages of
the soil/waste contaminants
• The types of contaminants that may be difficult to remove
• The solution contact time required for removal of contaminants and
• The levels of contaminants that may be transferred to the used
solutions
Site Sampling
Sampling locations were selected to obtain soil and waste sample com-
posites representative of the range of contaminant types and concen-
trations at the site. Composite samples were obtained to represent the
highest detected concentrations of individual contaminants and overall
site contamination. Consequently, soil washing tests were conducted
on the following six soil and waste samples:
• A soil composite containing the highest concentrations of all soil con-
taminants ("High Contamination Soil")
• A soil composite for high lead contamination ("High Lead Soil")
• A soil composite for high PCB contamination ("High PCB Soil")
• A soil composite containing all soil contaminants, but at concentra-
tions less that for soil areas (i.e.. "hot spots") with contamination
similar in magnitude to that of the highest detected concentrations
of contaminants ("Overall Soil Composite")
• A waste composite for high base neutral organics contamination
("High Base Neutrals Waste") and
• A composite of all wastes from the Waste Tank and the Waste Pit
("Overall Waste Composite")
These composite samples consisted of combined grab samples of soil
and tank/pit wastes. Actual sampling locations and depths were selected
based on a previous site investigation. Soil sampling depths varied
between 0 and 5 ft below the ground surface. Grab samples were
collected using back-hoes followed by shovels or hand trowels. One
or more grab sample layers were placed in a Teflon-lined, plastic col-
lection bag, depending on the type of composite sample required. Con-
siderable amounts of rubble were present at some of the sampling loca-
tions. Rubble material was not included in the soil/waste samples.
Soil/Wfeste Characterization
All sample analyses in this study were conducted by a U.S. EBV
certified Contract Laboratory Program (CLP) laboratory. Analyses con-
ducted are listed in Table I. All analyses were reported as individual
organic compounds (or elemental metals), except that petroleum
hydrocarbons were reported as a total compound class. CLP protocol
followed included the Statement of Work (SOW) 1086 for organics (with
revisions through SOW 7/87 for inorganics). Either the CLP Low Con-
centration Method or the CLP Medium Concentration Method for solid
samples was used in analyzing the study samples for organics. The High
Concentration Method included in the CLP protocol was not used for
this study, since it was not approved by U.S. EPA. At the time of the
study, U.S. EPA had not approved any laboratories to perform the High
Concentration Method.
All sample types had high organic chemical and heavy metal con-
centrations, even the sample type (Overall Soil Composite) which
represented the average soil concentrations. Certain sample types con-
tained fragments of concentrated organic solids or sludges. The High
Contamination Soil Composite contained notably higher concentrations
than the Overall Soil Composite of the following contaminants: copper,
lead, cadmium, xylene, ethylbenzene. toluene and trichloroethene.
The High Base Neutrals Waste Composite contained no detectable
levels of base neutral organic compounds. The lack of base neutrals
may be due to either; (1) uncertainties in the data base used to select
the sampling locations or, (2) a high variability in the contamination
levels in the Waste Pit Area. The lank/Pit Waste Composite contained
high concentrations of total PCBs, total xylenes, lead, copper and
chromium.
Selection of Solutions
In designing the Soil Washing Study, various chemicals and chemical
solutions were evaluated for possible use. This evaluation was based
on the results of a literature survey and direct communications with
chemical vendors, various research and development branches of the
U.S. EPA and investigators conducting similar studies. The solutions
used in this study had been used at other sites to reduce organic and/or
inorganic contaminant levels in soils. In the Soil Washing Study, higher
solution strengths were used than those reported in the literature, since
contaminant levels at the site were higher.
208 CONTAMINATED SOIL TREATMENT
-------
Table 1
Summary of Raw Sample Analyses for Selected Parameters
CONCENTRATION (In aw/ko) IN STUDY SAMPLES
HIGH
CONTAMINATION
MtmumuHTS SOIL
Kte
Aroclor 1242 190
Aroclor 1254 22. OJ
Aroclor 1260 HO
upljtile Oraanlcs
Hethylene chloride 5.6 J
Acetone 64.0
trans-l,2-D1chloro- 4.7 J
ethene
Chloroforii 140
1,2-Dtchloroethane 93
1,1,1-Trlchloroethane 47
Trlchloroethene 270
4-Hethyl-2-pentanone 28.00
Trichloroethene NO
Benzene ND
Tetrachloroethene 210
Toluene 530
Chlorobenzene 7.3 J
Ethyl benzene 68
Xylenes (total) 390
Base Neutral
Oroanlcs
Phenol ND
1,2-Dlchlorobenzene 17.0 J
Naphthalene 31.0 J
1.2,4-Trlchloro- ND
benzene
2-Nethyl naphthalene 15.0 J
bls(2-Ethylhexyl) 150
phthalate
Otethyl phthalate 15.0 J
Butylbenzylphthalate ND
Phenanthrene 19.0 J
DI-n-Butylphthalate 29.0 J
Metals
Antimony 3.9
Arsenic 7.7
Berylllm 0.6
Cadmiui 18.6
Chromliun 81.3
Copper 1,790
Lead 979
Kercury 6 . 7
Nickel 20.8
Stlenlu. 1.5
Zinc 612
OVERALL PCBs LEAD BASE NEUTRALS
SOIL SOIL SOIL UASTE
NO ND
ND ND
ND 5.2 J —
1.3
2.4
0.7
3.8
ND
KD
NO
0.16
1.8
0.45
7.6
3.6
0.45
1.2
14.0
ND ND
ND ND
ND ND
ND --- --- NO
NO --- --- ND
27 ND
ND NO
ND ND
NO --- — ND
ND ND
3.8
14.8
0.4
0.7 -
89.8
399
596 •-- 1,540
8.5
26.4
1.3
874
OVERALL
HASTE
130,000
16,000
ND
970
2,100
ND
4,800
1,700
600
1,500
2,400
1,500
970
70,000
29,000
ND
3,000
12,000
4,000
16,000
290
920
140
2,700
2,800
300
ND
270
421
33.8
0.0
361
6,060
4,020
59,700
103
35.6
11.1
2,510
NOTES: (Ij Source: ERH, Draft Feasibility Study, 1989
(2) Study Simple Identification; are is follows:
KB Soil
Lead Soil
High Contamination Soil
Overall Soil
Base Neutrals Waste
Overall Uaste
Nigh PCBs Soil Composite
High Lead Soil Conposlte
High Contamination Soil Composite
Overall Soil Compos Us
High Base Neutrals Uaste Composite
Overall Uaste Composite
(3) An unexpected result was the detection of no base neutral organlcs In the
High Base Neutrals Uaste sample, which was collected from previously identified
sampling points. The cause of this result Is unknown.
--- Not analyzed
NO Not detected
This study included soil and waste sample washing with four solu-
tions: (1) a 5% (by weight) solution of aqueous surfactant (i.e., Triton-
X-100), (2) a 10% (by volume) solution of hydrochloric acid and (3)
both a 5% and (4) a 10% (by weight) solution of citrate solvent (i.e.,
Citrikleen). Water was used as a base for preparation of all wash solu-
tions. Each wash solutions was analyzed for the same contaminants as
the soil samples. The Triton-X-100 surfactant did not contain any of
the organic compounds detected in the soil/waste samples and the
Citrikleen contained only one (i.e., acetone) of these compounds.
Total petroleum hydrocarbons were detected at levels of 5.2 and 4.1
mg/L in the 5% Triton-X-100 and the 5% Citrikleen solutions, respec-
tively. The surfactant solution contained less than 0.5 mg/L each of
copper, lead and zinc. The citrate solvent solution contained less than
0.5 mg/L each of copper, lead, zinc and chromium. These hydrocarbon
and inorganic contaminants could have been introduced into the solu-
tion by the laboratory or by the manufacturers of the Triton-X-100 or
Citrikleen. The hydrochloric acid solution contained trace amounts of
inorganics but no organics.
Figure 4
Soil Washing with One Fluid
liable 2
Soil Washing Study
Soil/Waste Process Trials
TABLE Z
SOIL HASH I KG STUDY
SOIL /WASTE PROCESS TRIALS
TYPE OF
PROCESS TfllAL
Washing with
One Fluid
Wishing with
Multiple
Fluids
(Sequential)
SOLUTION TYPES/
STRENGTH
Hydrochloric
add/ 10%
TrHon-X-100/5%
PT-24
CHrlkleen/SX
C1tr1kleen/I0%
PROCESS
TRIAL
HUHBER
PT-1 through
PT-12
PT-13 through
PT-24
PT-25 through
PT-36
PT-37 through
PT-48
Triton-*- 100/5X PT-49 through
Hydrochloric acld/lOX PT-51
Cltrikleen/10*
MOTE: The fluid application r
for all process trials.
atlo was maintained
SAMPLE TYPE
IHVESTIgATED
High Lead Soil
High Contamination Soil
OveraU Soil Compos Ue
Overall Waste Composite
High PCB Soil
High Contamination Soil
Overall Soil Composite
Overall HasU Conposlte
High PCB Soil
Overall Soil Ctmpostte
High Base Heutrals Uaste
Overall Uaste Composite
High PCB Sol,
Overall Soil Composite
High Base Neutrals Uaste
Overall Uaste Composite
High Contani nation Soil
Overall Soil Composite
Overall Uaste Composite
at a constant (10 9 fluid:
SYSTEM
CONTACT
TIMES
45/90/180
45/90/180
US/90/ 1 BO
4 5/90/1 BO
45/90/180
45/90/180
45/90/180
45/90/180
45/90/160
45/90/180
45/90/180
4S/90/180
45/90/180
45/90/180
45/90/160
45/90/180
270
270
270
1 g sample)
STAGE NUMBER
CONTACT OF
TIMES STAGES
(HIN.) (HIN.)
15/30/60 3
1 5/30/60 :
15/30/60 i
15/30/60 :
15/30/60 :
15/30/60 I
15/30/60 :
is/30/ eo :
15/30/60 :
15/30/60 :
15/30/60 ;
15/30/60 :
15/30/60 :
15/30/60 ;
15/30/60 :
15/30/60 :
90 :
90 :
90 :
1
1
s
)
1
1
1
J
J
J
1
1
)
]
)
1
3
3
J
All solutions at their application strengths had relatively low viscosity.
Both the surfactant and citrate solutions were opaque and tended to foam
when agitated. The hydrochloric acid solution tended to emit acid fumes
when agitated. The surfactant had a musty odor and the citrate solvent
had an odor of oranges.
Soil/Waste Washing Trials Using One Solution
Figure 4 provides a simplified representation of the washing trials
using one wash solution (i.e., process trials PT-1 through PT-48). Each
trial consisted of three individual washes (or process stages) perform-
ed on the same, preweighed volume of a single sample type.
Most of the process trials run in this Soil Washing Study involved
washing soil or waste samples with only one of the selected solutions.
The limited time available for completing the trials (2 wk) and sample
analyses (3 wk) required the study to focus on only two process
variables: the type of washing solution and the total time in which a
sample type was in contact with a solution (i.e., the "system contact
time"). Forty-eight different trials were conducted using the six sam-
ple types, four different washing solutions and various system contact
times. Table 2 contains a list of the different combinations used.
As shown, contact time varied among trials with the acid solution,
the Triton-X-100 and the Citrikleen. For each sample and solution type,
three process trials were performed, each with a different system con-
tact time, to determine the effect of contact time on contaminant removal.
The weight of sample per weight of solution (i.e., the "application ratio")
was maintained at a constant (25 g of sample per 250 mL of solution)
for all 48 trials.
CONTAMINATED SOIL TREATMENT 209
-------
Sequential Washing Trials Using Multiple Solutions
In addition to the process trials for testing u single solution type, three
"sequential trials" were performed which consisted of successively
washing a single sample type with more than one solution. Figure 5
provides a simplified representation of the three sequential trials: PT-49,
PT-50 and PT-51. In each trial, the following three solutions were ap-
plied to the same sample volume, each solution being applied in a
separate washing stage: 5% aqueous surfactant, 10% hydrochloric acid
and 10% Citrikleen. Each solution was applied for 90 min in a single
stage (unlike the earlier trials), so that three washing stages resulted.
The same application ratio was used as for the previous 48 trials.
Table 3
Summary of Percent Rcmovali' for Selected Contaminants
Rirtmx IOLIDI
—I row n*i
ExnUCTIon
StCUCATl |
2 1IU .
rVACTXOlU
AHALTU
•UO>LX
•OLID!/
OSED rLOIOl
Figure 5
Sequential Washing with Multiple Fluids
Results of the Soil Washing Study
For the trials with a single solution, the process samples collected
and analyzed included: (1) the treated sample solids prior to a final water
wash, (2) the final, treated sample solids and. (3) a composite of the
solution volumes used in each washing stage. For the three sequential
trials, the process samples collected and analyzed included: (I) the final.
treated sample solids. (2) the used aqueous surfactant from the first
stage, (3) the used hydrochloric acid from the second stage, (4) the us-
ed Citrildeen solution from the third stage. (5) the water wash at the
end of the second stage (6) the sodium carbonate wash following the
second stage water wash and (7) the final water wash. The carbonate
wash was used to neutralize the sample prior to washing with the citrate-
based solution (the last stage). Table 3 summarizes the results of these
process sample analyses.
The percent contaminant removals achieved in the washing trials are
summarized in Table 3. As shown, washing with hydrochloric acid
generally removed a high percentage of lead from all sample types in-
vestigated. The acid washing also removed a high percentage of copper,
but a lesser percent of chromium. Washing with 5% surfactant remov-
ed a high percent of PCBs (between 66 and 78%) and total xylenes
(between 87 and 99%). The 5% Citrikleen solution removed between
91 and 98% of a number of volatile and base neutral organics, but the
degrees removal of PCBs with this solution could not be determined
due to analytical difficulties. Table 5 summarizes some observations
about how certain process variables affected contaminant removals.
This preliminary study did not attempt to remove the maximum possi-
ble contamination, but generally showed that high removals arc possi-
ble. Used wash solutions contained high levels of meuil and organic
contaminants. Notable levels of selected contaminants, as detected in
the used solutions, are summarized in Table 4.
Residual concentrations (i.e., following washing) of lead, copper and
chromium of up to 733, 363 and 68 ppm, respectively, remained in
the washed High Contamination Soil sample (Table 3). Between 80 and
450 ppm of lead residual was detected in the washed High I-cad Soil.
Residuals of PCBs, xylenes and toluene of up to 7.4, 9.6 and 4.8 ppm,
respectively, remained in the High Contamination Soil sample. Residuals
of xylenes and tetrachloroethene of up to 1.8 and 0.96 ppm, respec-
tively, remained in the washed Overall Soil Composite. However, the
degree of removal of PCBs, xylenes and toluene in this sample was
inconclusive due to analytical difficulties.
UMHU. rot truer s*mci(2>
CMTANIIUKH
Volilllt Drain 10
lllnchloroilliini
IjllMl (lotll)
Notfljrlint Chlorldt
EthjrlbmliAi
f... H.ulnl Oraimci
HIGH
CONTAMINATION
SOIL
81 9i
87 98
M)
94 99
M 97
use
OVUAU. fCI IXAO NOTTIM.
MIL an mi jam.
62-87
u -n
n
M
OKHU
JBHL
N-M
ii -n
K-n
80 t>
4- Trlclllorobtnffnt
pMhllltl
Kfc
Aroclir 11W
Aroclor I2S4
ArKlor I HI
Total PftrolctlB
MO
99
IIC
V>
7i
M)
-!t
n
N)
Ml
NO
K>
•C
IK
HC
J!
»
w
R-91
ll-W
U 81
K
II-M.9
n
Copofr
HOIIS
HO
RC
•
II)
111
O)
Rot
Hot
Hoi
0 M
(0 91
fS-M
Sourct CUM. Drift f«ll
Stud; >«^l« lo>ntlflct
p tvloutljr I0>ntlfl*4 IJdT
,„ IJ,M
111 XtK)
con< lut Ivff
ss t) M-;I
18 U n-K
M 91 '0 W • 7S-B1
si in, stirfr. HOT
oni tri it follom
Higk Kit Soil Coapoilti
Hi«k nm Soil Cowoiltl
Hlfk CMtWlMtlM Soil Cc»polltl
Oririll Soil Coniltl
HIO» |4U iNtrlll Ktstl COIpOlltl
Ovirill tflltl CaapolIU
'.'•t 0>tlCtlO« of no bill MUtril amulet
wi;tl l^v't. oklcfe MIS colltctld fm
B'ift9 polatt Tht dull of tflli ritult
Table 4
Soil Wishing Study Cootaminanl Levels in Used Solutions
.vUMiM mvTfn
.') ).«M ; I M4 l ).«!
;»-i?e
•rM.m
Suitability of CLP Methods
During sample analyses, significant analytical difficulties were
encountered from the use of U.S. EPA CLP methods. These difficulties,
listed below, often required the laboratory to report residual contami-
nant concentrations as less than between K) and 2.500 ppm, rather than
the lower, actual value expected.
1. The methods did not include a procedure for analyzing samples con-
taining surfactants. Consequently, a foam was generated during the
preparation and analysis of samples previously washed with surfac-
tant solution. The foaming problem interfered with the introduc-
tion of the preparation extract into the GC/MS and could only be
mitigated by repeated sample dilutions.
2. Contaminant concentration values were obtained for treated samples
and used solutions in cases where contaminant levels were lowered
to within the calibration range of the analytical instrument (i.e., the
GC/MS). However, the methods frequently did not allow determina-
tion of precise contaminant concentrations in solid and liquid pro-
cess samples. This problem is relatively common during con-
taminated soil analysis and was due to the high levels of multiple
210 CONTAMINATBD SOIL TREATMKNT
-------
IbMe 5
Observed Effects of Process Variables
Figure 6
History of Waste Disposal
PROCESS
VARIABLES
FLUID TYPE
SYSTEM CONTACT
TIME (SCT) (1)
[—Elongated Trenches
(1962)
GENERAL EFFECTS OBSERVED
Higher PCB removals were achieved with citrate-based
solvent that with surfactant (Triton-X-100)
Similar removals of toluene, xylenes, ethylbenzene,
and tetrachloroethene were achieved with the
citrate-based solvent and with Triton-X-100.
PCB removal via extraction with Triton-X-100
increased as the SCT was increased over 45 minutes.
Removals of toluene, xylenes, ethylbenzene, and
tetrachloroethene did not increase with increasing
contact time, fo extractions with Triton-X-100.
Lead removal via extraction with hydrochloric acid
increased as the SCT was increased over 30 minutes.
This trend was not observed for copper and chromium
removal .
USE OF A WATER WASH
Overall metal removals
of a final water wash.
Overall organics removals were
applying a final water wash.
increased after application
not increased by
NOTE: (1) The system contact time is the total time (over all
washing stages) that the soil/waste sample was in
contact with the extraction fluid.
contaminants, making high sample dilutions necessary to keep the
chromatographic peaks for the contaminants within the instrument
range.
SOIL FLUSHING STUDY
Objectives
The Bench-scale Soil Flushing Study was initiated at a hazardous
waste (CERCLA) site to investigate and determine the feasibility of soil
flushing as a remedial technology for subsurface soils. A second and
related purpose was to determine the contaminant retardation
characteristics of soils within the existing groundwater contamination
plume of the site. This information is required for the design of facilities
to recover and treat groundwater.
The flushing study was planned in three phases:
• Phase I Preliminary Bench-scale Investigation
• Phase II Flushing Water Bench-scale Investigation
• Phase III Final Bench-scale Testing
In Phase I, a preliminary estimate of flushing feasibility were made.
In Phase H, the effects of different additions to the water used for flushing
were evaluated. The purpose of Phase in is to confirm the feasibility
of flushing by bench-scale testing using the methods and procedures
optimized in the first two phases.
To date, Phase I and n have been completed and are the subject of
this report.
Site Conditions
The hazardous waste site investigated is located within the Atlantic
coastal plain of the eastern United States. The site was operated as an
uncontrolled hazardous waste dump from the early 1950s to 1962. Wastes
reportedly were dumped at the site in drums or as bulk liquids and
either burned or buried in open pits or trenches. The surface of the
site is characterized by the absence of vegetation, the presence of a black
tar-like or asphalt-like material and areas containing corroded drums,
broken glass and other debris. A composite view of waste disposal
activity, taken from past aerial photography of the site, is shown in
Figure 6.
The site has remained virtually unchanged since 1962. An initial in-
vestigation by the U.S. EPA in 1983 lead to the site being listed on the
NPL. In 1986, a remedial investigation of the site was begun by the
U.S. EPA. The U.S. EPA has issued a draft feasibility study for the
General Refuse
(1956, 1962)
Standing Liquid
(1962)
Vegetation
• Depression With
Black Standing Liquid
(1956)
\\
\\
\\
site, and a ROD for remedial action will be issued by the end of 1989.
The site topography is relatively flat with occasional small hills. The
32-ac denuded area of the site is surrounded by forest. Soils at the site
and in surrounding areas consist of well drained, bleached, weathered
sand underlain by a sandy loam subsoil containing minor percents of
clay minerals. These soils have a low organic content, a low pH and
exhibit a low cation exchange capacity (CEC). Unsaturated soils beneath
the site range in depth from 16 to 18 ft.
The sandy aquifer beneath the unsaturated soils extends to a depth
of over 180 ft until reaching the first confining layer. Groundwater flows
with an average horizontal hydraulic gradient of approximately 0.003
ft/ft. The transmissivity of the sandy aquifer beneath the site ranges
between 75,000 and 150,000 gal/day/ft. Groundwater quality is naturally
acidic and contains high concentrations of iron, manganese, carbon
dioxide, dissolved and suspended solids. The natural pH, iron and
manganese groundwater concentrations exceed federal drinking water
standards.
Disposal of waste materials in the 1950s and 1960s has contaminated
the surface soil, subsurface unsaturated soils and groundwater. Plans
for remediation of surface soils include off-site encapsulation and off-
site treatment of liquid and semiliquid wastes unsuitable for encap-
sulation. Due to high transmissivity of the aquifer beneath and downgra-
dient of the site, a groundwater contamination plume extends approx-
imately 1 mi from the site. To accurately estimate the duration of treat-
ment plant operation, site-specific information must be obtained regar-
ding the retardation of contamination in the sandy aquifer. Contami-
nant retardation is expressed as the rate of groundwater flow versus
the normally slower rate of contaminant movement in an aquifer.
One potential method of remediating subsurface unsaturated soils
involves flushing these areas with treated groundwater to remove con-
tamination from the soil. If feasible, in situ soil flushing could reduce
subsurface remedial costs by at least 90% over other methods involv-
ing excavation. Bioremediation of this site had been previously evaluated
by EEC4 and was determined to be only partially effective.
CONTAMINATED SOIL TREATMENT 211
-------
Some of the hazardous volatile and semi-volatile compounds found
at significant concentrations (1 to 100 ppm) in subsurface unsaturated
soils at the site are:
Chlorobenzene
Styrene
1,1,2,2-lbtrachloroethane*
bis (2-ethylhexyl) phlhalate*
Toluene*
bis (2-chloroethyl) ether*
Chloroform*
4-methylphenol*
1,2-Dichloroethane*
2-methylphenol*
Ethylbenzene*
Naphthalene
Trichloroethpn*-"
Phenanthrene
Tetrachloroethene*
Phenol
Benzene*
1.2-Dichlorobenzene
2-Butanone
1.4-Dichlorobenzene
Acetone
DDT*
Xylene
ODD*
Compounds with an asterisk are those identified by the U.S. EPA
as chemicals which pose a potential health risk and are. thus, of primary
concern at the site. The distribution of contaminated subsurface soil
is shown in Figure 7 as unsaturated contaminated soil thickness isopleths.
Figure 7 also shows the edge of existing surface waste. AS shown, con-
taminated soils extend to the water table in the west central portion of
the site where standing liquids were observed in 1956 and 1962
(Figure 6).
Existing conditions at both sites indicate that in situ soil flushing may
be feasible. Soils at the site are well drained and have a relatively low
cation exchange capacity (CEC) and organic content, which limits their
ability to retard contaminant flow. The results of site characterization
work indicate areas of subsurface unsaturated soils which are con-
HOCKO
: : ftOADWAY
- CONTAMINATED SOIL
THICKNESS I30H.CTM
COOt 0' 3URF»C( WASTE
laminated with soluble organic compounds. Since the site has been
"naturally flushed" by percolating rainfall for over 30 yr, the majority
of soluble site contaminants should be located at depth, just above or
in groundwaler.
This assumption is partially confirmed by past site investigation which
shows that groundwaler beneath and downgradient of the site contains
compounds at concentrations one order-of-magnitude less than those
in contaminated subsurface soils. In addition, subsurface soil volatile
organic contamination was found to be relatively widespread with the
highest concentrations usually occurring at the groundwaler table.
To assist in determining both the effectiveness of flushing subsur-
face soils and the time required for groundwater pumping and treat-
ment plant operation, a bench-scale soil flushing study was conducted.
Phase I
Phase I of the Soil Flushing Study involved the collection of three
contaminated-undisturbed soil columns for bench-scale testing. One soil
column was obtained near the surface of the site and the remaining two
from just above and just below the groundwater table at the location
shown in Figure 7.
Soil flushing bench-scale testing of all three soil columns was con-
ducted using the apparatus shown in Figure 8. As shown, the apparatus
utilizes a stainless steel Shelby lube to obtain an undisturbed column
of soil for testing. The only materials in contact with the soil or flushing
water are Teflon, glass or stainless steel. A steel plate and rod assembly
was used to seal the lop and bottom Teflon end caps securely against
the stainless steel Shelby tube. To aid in the prevention of water loss.
Teflon tape was placed over the threaded ends of the Teflon connectors
and Teflon 0-rings were placed between connections. Prior lo use, all
parts of the apparatus were decontaminated.
Figure 8
Bench-Scale Rushing Apparatus
- TEFLON INFLUENT §»0
GLASS TUBE
QLASS BELL TOP
STEEL PLATE
THREADED STEEL MOD
STAMLESS STEEL
•HELSY TUBE
STEEL PLATE
GLASS TUBE
TEFLON
TEFLON CAP
GLASS BEADS
SON.
GLASS WOOL
GLASS BEADS
TEFLON CAP * VALVE
TEFLON EFFLUENT BAO
Figure 7
Contaminated Subsurface Soil Distribution
The apparatus is air tight and uses Teflon gas bags for influent and
effluent collection. Flow is under gravity conditions and is regulated
by a valve in the bottom Teflon end cap. For Phase I, distilled water
was used as influent. Effluent was sampled every 12 to 24 hr for analysis
212 CONTAMINATED SOIL TREATMENT
-------
of 12 subsurface site contaminants on U.S. EPA s Target Compound
List (TCL). Effluent samples, as well as soil samples, were obtained
from the soil columns for analysis of all TCL parameters at the beginning
and end of the bench-scale test. Soil samples taken in the field adja-
cent to those used for bench-scale testing were analyzed for gradation
(% sand, % silt and % clay) organic content, cation exchange capaci-
ty, pH, density and porosity.
Results from Phase I of the bench-scale flushing test show that effluent
concentrations from flushing the two columns, taken just above and
just below the groundwater table, significantly decreased during the
test. This effluent concentration decrease, however, was directly cor-
related to a significant decrease in flow through the soil (i.e., perme-
ability) also experienced during the test in the two columns taken just
above and just below the groundwater table. Similar reductions in
permeability during bench-scale testing did not occur with the soil col-
umn taken near the surface. Effluent concentration changes during
bench-scale testing of the surface soil column were mixed, with some
contaminant concentrations decreasing and others having wide varia-
tions during the test.
As previously indicated, subsurface soils at the site contain minor
percents (2 to 4%) of clay minerals. Review of the Phase I soil flushing
results indicated that flushing with distilled water removed double-
valence, positively charged ions of calcium and iron as well as single-
valence, positive ions of sodium from the soil matrix. Removal of these
ions, specifically from the clay in the subsoil, may have induced disper-
sion and swelling of the clay which, in turn, reduced soil permeability.
As a result, three additional soil columns taken adjacent to those sub-
jected to soil flushing, were tested for permeability. Permeability testing
of these soil columns was conducted using tap water which contained
approximately 43 parts per million (mg/L) of calcium, 35 mg/L of
magnesium and 26 mg/L of sodium. This tap water permeant had a
very limited capacity to remove either double or single-valent, positive
ions from the soil matrix. Thus, permeability reductions from clay
dispersion and swelling should have been prevented. Significant reduc-
tions in permeability, however, still occurred in the soil column taken
below the groundwater table.
Based on Phase I testing, it became apparent that the phenomenon
of permeability reduction during flushing was more complex than anti-
cipated. Other factors which may have contributed to reduced
permeability include dissolution and/or precipitation of inorganics in
response to changes in pH or precipitation of iron in the presence of
dissolved oxygen.
The pH of the water used to conduct soil flushing and permeability
tests was approximately 7 and 8.3, respectively. Changes in pH may
have induced the formation of metal complexes. Metal ions present
within the soils can complex when a sufficient amount of hydroxide
becomes available in solution. These complexes form within the in-
terstices between soil grains and restrict flow. Soils at the site contain
a significant amount of iron, which can precipitate and clog soil pores.
A pH increase also produces additional negative charges on the clay
particles present in the soils at the site. Increases in the negative charge
of clay particles cause the clay to disperse and swell, whether the in-
crease is from a rise in pH or removal of positive ions.
The water used for permeability and soil flushing tests had a dissolved
oxygen content of approximately 6 and 7 mg/L, respectively. Ground-
water at the site contains no measurable dissolved oxygen, but high
amounts of carbon dioxide. Therefore, the addition of dissolved oxygen
to the soil could easily result in the precipitation of iron from the soils.
Reductions in soil permeability (clogging) may be a major impediment
to the soil flushing process. Phase II of the Soil Flushing Study was
designed to investigate additions to flushing water influent which may
prevent reductions in soil permeability during flushing.
Phase II
Flushing influent variables which have been identified during Phase I
are pH, dissolved oxygen content (DO) and calcium ion concentration
(Ca). To determine the optimum combination of these variables
economically, bench-scale testing in Phase H was planned using the
23 statistical experimental design of these variables, shown in Table 6.
Table 6
Phase II Soil Flushing Study Experimental Design
ACTUAL INFLUENT CONCENTRATION
Soil Column/Solution
Number Ca (ppm)
39.32
381.37
39.31
383.26
38.52
386.94
42.13
387.26
0.00
fiH
4.2
4.2
5.7
5.8
4.2
5.5
5.6
5.6
6.3
DO (ppm)
1.19
1.17
1.16
1.16
8.35
8.70
9.14
8.96
9.30
As shown in Table 6, nine soil columns were tested during Phase II.
Distilled water was applied to column 9 to match Phase I conditions.
The solutions used for Columns 1 through 8 were made by adding the
appropriate concentration of calcium chloride. Lower pH solutions were
obtained with the addition of hydrochloric acid. Decreased DO was
achieved by bubbling nitrogen gas through the solution.
Each soil column tested during Phase n was obtained from just below
the groundwater table at the sampling locations used in Phase I. In Phase
I, soil taken from just below the groundwater table clogged using both
distilled and calcium enriched tap water. To determine the effects of
different Ca, pH and DO influent concentrations on soil permeability
as well as on contaminant removal, all soil columns for Phase n were
taken from soil zones with as near identical conditions as possible. The
horizontal and vertical location of samples were surveyed to match the
Phase I location. In addition, each column contained 18 in of soil.
The same apparatus used in Phase I was used for Phase n. As in
Phase I, effluent was sampled every 24 hr for analysis of 12 TCL sub-
surface contaminants. To determine if the solutions used affected the
soil matrix, effluent was also analyzed every 24 hr for total suspended
solids, calcium, chloride, iron, silicon, sodium and zinc. As in the first
phase of testing, soil samples were obtained from the soil column for
analysis of TCL parameters at the beginning and end of bench-scale
testing. Unlike Phase I, however, the actual soil in each column was
analyzed for gradation organic content, cation exchange capacity, pH,
density and porosity.
Throughout Phase n of the Soil Flushing Study, no reduction in soil
permeability was evident. Flows through all nine soil columns remained
consistent. Effluent concentrations of the 12 TCL compounds measured
from Columns 1 through 7 decreased quickly as compared to Columns
8 and 9. After approximately 15 pore volumes, effluent concentrations
of the volatile compounds monitored were reduced to nondetectable
levels (<1 ppb). To determine if the influent solutions used inhibited
contaminant removal and/or prevented a reduction in soil permeability,
the influent of all nine soil columns was changed to distilled water after
approximately 15 pore volumes. After changing the influent to distill-
ed water, no additional contamination was removed and reductions in
soil permeability did not occur. After Phase II bench-scale testing was
completed, soil column concentrations of all site-related TCL volatile
compounds were nondetectable.
To evaluate the effects of each solution on soil permeability, the
inorganic effluent results obtained during solution testing were reviewed.
Although Column 9 showed no apparent signs of reduced permeability,
it was the only column in which total suspended solids were removed
during flushing. This indicates a breakdown of the soil matrix. Removal
of sodium from the columns also indicates an increased potential for
CONTAMINATED SOIL TREATMENT 213
-------
dispersion and swelling of any clay present. Sodium was removed from
all columns. A preliminary statistical analysis of the data, however, in-
dicates that removal of sodium from the soil matrix is reduced using
the higher calcium, low pH, low DO solutions.
To determine the effect of each solution on contaminant removal, soil
contamination concentration data obtained from bench-scale testing were
used to fit the following first-order decay equation.
C (V) exp-
(1)
Co
Where: V = Number of pore volumes
C(V) = Soil contaminant concentration at pore volume V
Co = Soil contaminant concentration at zero
pore volumes.
exp = napierian base
K = Decay constant for a particular compound
and soil
In Equation 1, the larger the value of K, the lower the number of
pore volumes required to remove the contaminant. A statistical analysis
program known as MINITAB' was used to fit Phase D experimental
data to Equation 1 using multiple regression analysis. For each com-
pound and each soil column, a regression coefficient, K, was
determined.
The ranges of K for each compound found in the soil columns are
shown in Table 7. In the gas chromatograph used for effluent analysis,
1,2-dichJoroethane and bis(2-chloroethyl)ether coelute. Thus, the K
values shown in Table 7 are for 1,2-dichloroethane and bis
(2-cnloroethyl)ether combined. As shown, K values for each compound
have significant variations. Since effluent concentrations decreased
quickly from Columns 1 through 7, K values were generally half an
order-of-magnitude greater than for Columns 8 and 9.
Table?
Phase II Soil Flushing Study Regression Coefficients
Compound
l,2-D1chloroetlune/
b1s(Z-chloroethyl)ether
Benzene
Trlchloroethene
Toluene
p-and B-Xylene
o-Xylene
Ethyl benzene
1,1,2,2-Tetrachloroethane
Regression K
Value
0.196 to .991
0.090 to 1.73
O.US to 1.17
0.1S8 to 0.898
0.117 to 0.865
0.130 to 1.25
0.089 to O.S8S
0.3M to 1.06
Mater
Solubility1
»20°C(pp»)
8690/10200
1780
1100*
SIS
198"
175
152
150*
1 from Verschueren, 1983, Handbook of Environmental Data on Organic
Compounds
* 1,2-dichloroethane and bfs(2-chloroethyl) ether coelute during
effluent analysis. K values shown are for the combination of both
compounds.
* At 25°C
** For p-Xylene at 25°C
To determine if K values were related to the influent solutions used.
the regression K values were statistically compared to influent calcium,
pH and dissolved oxygen content, as well as soil density, porosity.
percent sand, percent silt, percent clay and percent organic matter. In
addition to influent and soil physical characteristics, regression K values
also were compared to a total concentration of the 12 TCL compounds
measured in each soil column (total Co) and total
bis(2-ethylhexyl)phthalate (BEHP) initially found in each soil column.
BEHP is a common contaminant in each column and the only major
semi-volatile present.
Preliminary results of this analysis indicate that:
• For relatively soluble compounds of 1,2-dichloroethane,
bis(2-chloroethyl)ethcr and benzene, the regression K values are
related to the Total Co in the soil column, the concentration of BEHP
initially present and the dissolved oxygen content in the influent
solution used for flushing. The regression K constants are larger for
soils with low Total Co and BEHP concentrations and low influent
dissolved oxygen concentrations.
• For moderately soluble compounds of trichloroethcne. toluene,
chlorobenzene, xylene, ethylbenzene and 1.1.2.2-tetrachloroethane,
regression K values are directly related to the BEHP concentration
initially present in the soil and the pH and/or dissolved oxygen con-
tent of the flushing influent. As with the higher soluble compounds,
the regression K values are larger for soils with low initial BEHP
concentration and low influent pH and dissolved oxygen
concentrations.
These results are consistent with accepted theories of partitioning
between the solid and solution phase V>(6) (i.e., soils with greater
organic content tend to retain organics).
CONCLUSIONS
From results of the Bench-scale Soil Wishing Study the following
conclusions can be made:
• Removal of volatile and semi-volatile organics, petroleum hydro-
carbons. PCBs and a variety of heavy metals is possible using soil
washing methods.
• Both the aqueous surfactant and the aqueous citrate-based solutions
are effective for high removals of all classes of organic compounds
tested. The citrate-based solution appears to be slightly more effec-
tive than the surfactant for PCB removal.
• Chromium, nickel, mercury and arsenic are contaminants that may
be more difficult to remove to acceptable levels using a hydrochloric
acid solution.
• Single-stage soil washing is capable of removing high percents of
contaminants. However, due to the level of residual concentrations
in this study, a greater number of stages would be required to meet
cleanup goals.
• Used solutions would require significant pretreatment for reuse due
to their high contaminants levels. For the site investigated in the soil
washing study, complete replacement of spent washing solutions with
fresh solution volumes may be necessary during initial stages of soil
washing. Complete replacement and disposal of used solutions may
involve significant cost which, in turn, each impact the feasibility
of soil flushing.
• The CLP protocol used for process sample analysis was not suitable
due to the high constituent concentrations.
From results of the Bench-scale Soil Rushing Study, the following
conclusions can be made.
• In situ removal, to nondetectable levels, of many volatile compounds
from soils by soil flushing is possible.
• Soil flushing can cause reductions in soil permeability which, in turn.
can prevent the removal of soil contaminants.
• Selection of soil flushing influent should be based on a review of
site soil chemistry to prevent permeability reduction during flushing.
• The first-order decay equation (Eq. 1) accurately expresses the reduc-
tion in volatile soil contamination during soil flushing. The decay
constant is related to the initial total concentration of volatile and
semi-volutile contaminants present in the soil.
Both studies require additional bench) and/or pilot) scale work to
fully demonstrate feasibility of the soil remediation technologies.
214 CONTAMINATED'SOIL TREATMENT
-------
HFFFRFNCES ' ee' °ln "' "Feas'b'''ty of Bioremediation at Hazardous Waste Sites" Proc
o/ f/ie Eleventh Annual Madison Waste Conference, University of Wiscon-
1. U.S. EPA, Handbook - Remedial Action at Waste Disposal Sites (revised). sin, Madison, WI, 1988
U.S. EPA-625/6-85/006, US. EPA, Cincinnati, Ohio, 1985. 5. Ryan, B.F., Joiner, B.L., Ryan, T.A., MIN11AB Handbook, Second Edition.
2. Dames and Moore, Remedial Investigation Report, Sept. 1988. (Client main- PWS-Kent Publishing Company, Boston, MA, 1988.
tained confidential for purposes of this paper). 6. Lyman, W.J., Reehl, W.F., Rosenblatt, D.H., Handbook of Chemical Pro-
3. U.S. EPA, 1988 and Penetone, Personal Communication with M.E. Kunze, perty Estimation Methods, McGraw-Hill Book Company, New \fork, NY,
EEC, Inc. 1982.
CONTAMINATED SOIL TREATMENT 215
-------
- Soil Stabilization/Solidification
at the Tacoma Tar Pits
Gretchen Rupp, P.E.
Environmental Research Center
University of Nevada Las Vegas
Las Vegas, Nevada
James Pankanin
PRC Environmental Management, Inc.
Seattle, Washington
ABSTRACT
The Tacoma Tar Pits Site is a Superfund site on the Puyallup River
in Tacoma, Washington. Coal tar and hydrocarbon liquids are present
in shallow ponds and as discontinuous lenses within site soils. Soils
and shallow ground water are contaminated with lead, PCBs and coal
gasification products including phenols, polycyclic aromatic hydro-
carbons, benzene, toluene and xylene. Portions of the site are covered
with up to 3 ft of automobile shredder fluff and associated debris.
The principal remedial technique named in the ROD on for the site
is on-site fixation (stabilization/solidification) of shallow soils and auto
fluff. In a bench-scale soil fixation treatability study, site soil, coal tar
and auto fluff were mixed in various proportions and fixed via a
proprietary product. The resulting monoliths underwent testing for con-
taminant concentration, physical-engineering properties and leaching
behavior. These properties were determined:
• contaminant concentrations as compared to site cleanup goals and
other ARARs
• volume increase due to fixation
• wet/dry durability
• compressive strength as a function of time and monolith composition
• permeability
• the effect of wet-dry stressing on monolith strength, permeability
and leaching
• effect of fixation on the TCLP leaching of five classes of contaminants
• ANS 16.1 leaching behavior, with and without wet-dry stressing.
Using calculated diffusivity values for the contaminants of concern,
long-term leaching from a full-size monolith at the site was projected.
INTRODUCTION
The Tacoma Tar Pits is a Superfund site in the industrial district of
Tacoma adjacent to the Puyallup River. A coal gasification plant operated
on the 30-ac site from 1924 until 1956. Since 1967, a scrap-metal
recycling facility has occupied most of the site. Current site features
are shown in Figure 1.
RI/FS of the site has been conducted by the PRPs.1 •' The Rl showed
that wastes from the coal gasification and metal recycling operations
are distributed over the entire site. Several acres arc covered to a depth
of 1 to 3 ft with decomposing auto fluff—the foam, rubber and non-
ferrous metal products of an automobile shredder. There are approxi-
mately 5000 yd3 of coal tar within the shallow soil (chiefly in a "tar
boil area") and in a small pit. Organic non-aqueous phase liquid (NAPL)
forms lenses within the soil over much of the site.
These wastes are the source of several types of contaminants. Phenols,
benzenes and polycyclic aromatic hydrocarbons (PAHs) are the chief
components of the coal tar and NAPL. In the auto fluff and underlying
Fig I
The Tacoma Tar Pits Site.
fill, concentrations of inorganic contaminants (principally lead, mercury
and arsenic) are elevated. PCBs are found in the auto fluff and the
underlying fill. Varying levels of metals, PAHs, phenols and benzenes
have been measured in the shallow aquifer on the site.
ROD for the Tar Pits Site established cleanup goals for site soil,
surface water and ground water.' Goals were established in each
medium for lead, benzene, total PCBs and fi\Hs (the sum of benzo(a)pv-
rene, bcnzo(a)anthracene, benzo(b)fluoranthene, benzo(k)fluoranthene,
dibcnz(a,h)anthracene and indeno(l,2,3-cd)pyrene). The ROD named
stabilization/solidification of shallow soils and auto fluff as the prin-
cipal component of site remediation.
Using materials from the Tar Pits site, a bench-scale treatability study
was conducted. Its purpose was to test fixation as a treatment technique
for multiple, physically heterogeneous waste matrices containing several
classes of contaminants.
MATERIALS AND METHODS
Study Design
The study was designed not to optimize treatment parameters but,
more fundamentally, to determine whether stabilization/solidification
would even work on the waste matrices from the Tar Pits. Figure 2
shows the study design. The first phase of study planning established
the properties of the raw and fixed wastes that would be tested in
216 CONTAMINATED SOIL TREATMENT
-------
assessing the success of the treatment.
Based on these properties, specific chemical, physical and leaching
tests were chosen. The parameters chosen to track were the four
contaminant classes named in the ROD, plus phenols. To assess the
physical integrity of the monoliths, tests of permeability, durability and
compressive strength were specified. The leaching tests selected were
one regulatory test (the TCLP) and one test that would allow projec-
tions of long-term leaching (American Nuclear Society Test 16.1).
Composition Analysis
T«r
/ •*
),
'
Mix
1:0 SolltFluff
1:1 SolliFluff
3:1 Sol I: Fluff
blank sand
0:1 SolUTar
1:1 SolhTar
TCLP
i
Analysis
of Extracts
billty
st
Permeability
Test
Fig. 2
Material Flow Through the Testing Procedure
The second phase of study planning involved (1) selecting the dif-
ferent waste materials and combinations of materials to be treated and
(2) the degree of replication at each step. The number of samples that
could be generated was constrained by the high analytical costs asso-
ciated with the five contaminant classes. Six matrices were selected
for testing:(l) site soil; (2) a l:l(w/w) mix of site soil and auto fluff;
(3) a 3:1 soil-fluff mix; (4) coal tar; (5) a 1:1 soil-tar mix; and (6) blank
sand (a clean commercial sand). Although tar fixation was not proposed
in the ROD, tar was chosen for testing to define the limits of fixation
technology for site materials. This testing would also help determine
what would happen if, during remediation, subsurface tar were encoun-
tered and inadvertantly incorporated into the feed stream of the soil
treatment unit.
Once the study design was finalized, additional planning steps
included determining the appropriate sizes of the fixed coupons, cal-
culating the needed volumes of site materials and devising the site
sampling plan. The project workplan5 sets forth all of the methods
used and the rationale for their selection.
Field and Laboratory Methods
The sampling and fixation experiments took place in January 1988.
The sampling design was biased towards the collection of highly-
contaminated material to provide maximum challenge to the fixatives.
Site materials were collected over the full depth interval that would be
excavated for site remediation. Seven soil samples were collected from
pits 1 ft deep. They were composited on-site and screened through 3/8-in
hardware cloth. Approximately 150 Ib of field-moist soil were collected.
Auto fluff was collected from four locations where the RI indicated
NAPL contamination. The greatest thickness of fluff sampled was
2 ft. Particles larger than 3/8 in were screened out, and the fluff was
homogenized on-site. Approximately 100 Ib of fluff were sampled. A
total of 3 gals of liquid tar were taken from the tar pit.
Before processing, site samples were stored at ambient temperature
(40 to 50 °F) for 5 days. Mixing and fixation took place at the U.S.
EPA Region 10 laboratory in Manchester, Washington. All mixing was
done in 2-kg batches using a Hobart rotary mixer. After site materials
were mixed in the predetermined ratios, sub-samples of the unfixed
materials were created for chemical analysis and TCLP extraction. Then
the mixtures were treated by measured addition and mixing of tap water
and one or mor^propaejary reagents>Fixed mixtures were poured into
small cylindricatmolds of three different sizes, as required by the dif-
ferent tests. These were capped, labelled and stored at 100% relative
humidity in closed plastic trays.
Materials to be tested for compressive strength were dispatched to
the physical-testing laboratory the day after fixation. All other coupons
were cured at ambient temperature for 28 days before further handling.
Phenol analysis was by colorimetry. All other chemical analytes were
quantified according to Contract Laboratory Program methods by CLP
laboratories. Physical/engineering tests were performed by U.S. EPA's
Center Hill Laboratory according to its standard protocols for stabi-
lized/solidified materials (6). Each type of monolith was tested for
ultimate (28-day) unconfmed compressive strength by a modification
of ASTM D1633-84. The development of strength was traced by testing
cylinders of the 3:1 soil-fluff mixture at 3, 7, 14, 22 and 29 days. Perme-
ability was tested in a pressurized upflow apparatus by a method similar
to EPA Method 9100-2.8(7). Durability was examined by a wet-dry stress
test similar to ASTM D-559. The test entails 10 cycles of 6-hr immer-
sion in deionized water followed by 18 hr of drying at 140°F(60°C).
Seasonal wet-dry exposure of the full-scale monolith via shallow ground-
water is possible at the Tar Pits site, and monolith strength and perme-
ability might be affected. Therefore, two sets of fixed samples were
subjected to wet-dry stressing, then one was strength tested and the
second was tested for permeability.
Raw and fixed materials were crushed and leached by the TCLP. Both
standard and zero-headspace extractors were used. Fixed materials were
also extracted via a modification of the ANS 16.1 test. In this test, a
cylinder was suspended in deionized water under quiescent conditions.
After 1, 3, 5, 7, 14, 28 and 90 days, the leachant was replaced by fresh
water and subjected to chemical analysis. A full set of fixed coupons
underwent this testing. In addition, a set of 3:1 soil-fluff cylinders
underwent wet-dry stressing (6, 14 or 24 days) followed by 90 days of
standard ANS leaching. In these tests, the initial leachates were collected
during the wet-dry stressing period.
Resource constraints prohibited a fully-replicated experimental and
analytical design. Instead, precision was assessed at every step by testing
triplicate samples of the 3:1 soil-fluff mixture.
Volatile compounds were analyzed in several types of samples, with
the understanding that the reported concentrations may not be repre-
Table 1
Contaminants in Site Materials
Concentration (mg/kg)
Contaminant Material
Reported
from RI
Treatability Study
Raw Fixed
Lead
Phenols
Benzene
Soil
Fluff
1:1 Soil-Fluff
Tar
Soil
Fluff
1:1 Soil-Fluff
Tar
Soil
Fluff
1:1 Soil-Fluff
Tar
73-12,900
2910-4700
37
<,
150
<0.001
430
2490
3670°
3080
377
790C
584
200
0.002
<0.007
877
4663°
2770
52
<1450
<1970
<1050
Total PAHsa Soil
Fluff
1:1 Soil-Fluff
Tar
Total PCBsb Soil
Fluff
1:1 Soil-Fluff
Tar
<0.7-135
11-204
5.6
3240
6.2
58=
32
<198
13.2
<348
2950
6.6
26C
16.2
<28
a7Sum of benzo(a)anthracene, benzo(b)fluoranthene,
benzo(k)fluoranthene, benzo(a)pyrene, indenod,2,3-cd)pyrene and
dibenz(a,h)anthracene
b. Sum of Aroclors 1016, 1221, 1232, 1242, 1248, 1254 and 1260
c. Calculated from the concentrations in the soil and 1:1 soil-
fluff mixes
CONTAMINATED SOIL TREATMENT 217
-------
sentative. At no stage of sample handling was any attempt made to
capture or quantity escaping volatiles.
RESULTS AND DISCUSSION
Chemical Concentrations
Contaminant concentrations in selected raw and fixed materials are
summarized in Table 1 and compared with data from the RI Concen-
trations in the raw materials generally fell within the previously-reported
ranges. Lead concentrations in soil and fluff were 2000 to 5000 mg/kg.
The soil collected for the treatability study was contaminated with
NAPL. as evinced by the phenol and PAH \alues. but volatile con
tarninants were nearl\ undetectable. PAH concentrations in fluff were
less than half those in soil (<1 vs 13 mg kg), reflecting the fact that
the auto shredder began operation long after the coal-gasification wastes
were released on the property. The shredder fluff is the chief source
of PCBs at the site, and total PCB concentrations in the sampled flufl
exceeded 50 mg/kg (calculated from the soil fluff mixes).
The dilution factor caused h\ adding fixation reagents to the raw
materials was approximately 70 rr . The pro- and post-fixation concen-
tration results did not reflect this factor closely (Table 1). The source
of the variation is not clear; data from triplicate samples indicate that
between-sample variation was 10 to 15 "f
Physical Properties
'Volume increases of site materials resulting from fixation ranged from
45 to 95%. These relatively large volume increases reflect the high bulk
density of the in-place soil (about 130 Ib/ft1) and the high reagent
dosages used in the treatability study.
In the wet-dry stress test of fixed monoliths. durability was measured
by monolith weight loss. There are no standard criteria for acceptable
weight loss, but a weight loss of 5% or less generally is judged accept-
able." In the treatability study, tar and tar-soil cylinders had weight
losses exceeding 5% (11.8 and 92%. respectively). Among the soil and
soil-fluff monoliths, one experienced a 5.3% weight loss; the average
weight loss was less than 3%.
Figure 3 traces the development of monolith conipressive strength
over time In concrete work, the 28-day strength is the measure of
"ultimate strength." As shown in Figure 3. the 3:1 soil-fluff cylinders
were still measurably gaining strength at 29 days. The inclusion of fluff
and organic compounds probably slowed the development of strength;
this phenomenon has been widely demonstrated.'1 In such materials.
a longer test (50 or 60 days) may give a more accurate measure of ulti-
mate strength.
Figure 4 depicts the 29-day compressive strengths ol the \arious
monoliths. Water contents of all the mixes were similar. The mixes with
higher proportions of soil developed greater strength; both auto fluff
and tar detracted from monolith strength. Among unstressed cylinders,
fixed soil had the greatest strength (895 psi) while fixed tar had the
least (zero). For every material type, cylinders subjected to wet-dry
stressing were stronger than unstressed cylinders During the dry por-
tions of the stressing cycles, cylinders were held at 140' I- (60T) which
accelerated their curing. Because ol this, effects due solely to cyclic
wetting and drying cannot be distinguished
There currently are no general standards lor the strength ol monoliths
created during remediation of NPL sites. The Nuclear Regulatory Com-
mission requires a strength of 150 psi in rigid waste materials The
proposed minimum strength for land-disposed RCRA solid wastes is
50 psi. At the Tar Pits, the overburden pressure on a full-si/e monolith
would be 2 to 5 psi. Most of the tested mixes could readily support
this load.
Figure 5 shows the permeabilities of the fixed materials. Among the
unstressed materials, the most permeable was the 1:1 tar-soil cylinder
(5x10 " cm/sec); the least permeable was fixed tar (1.2x10 ' cm/sec)
Permeabilities of the soil and soil-fluff monoliths were all within a (actor
of three, suggesting that the inclusion ol fluff did not increase permea-
bility. For every type of fixed material, wet-dry stressing increased
permeability by more than an order of magnitude. This increases in
permeability may have been an effect of accelerated curing Altcrna-
w
Q.
c
I)
v>
u>
Q.
Ł
o
700
600-
500-
400 -
300-
200-
100-
i
20
25
5 10 15
Curing Time (days)
— Average • Experimental value
Fig. 3
Development of Compressive Strength in a
Fixed 3.1 Soil Auto Fluff Mn
30
1400
unstressed
Sand
Blank
I wet/dry stressed
Fig. 4
I'IK on fined Comprx-ssivc Strengths of Fixed Materials.
Results lor .VI Soil:Fluff Mixture are
Means ol Triplicates; All Others are
Single Determinations.
218 CONTAMINA11.D SOIL TRKATMKNI
-------
lively, it may have been caused by micropores near the cylinder sur-
faces, created through dissolution or other mechanisms. Leaching data
(see below) indicate that such dissolution did occur. These results suggest
that the physical integrity of an in-place monolith would be greatly
enhanced by keeping it dry.
Table 2
Summary of TCLP Results
u
13.6
>12.2
>9.2
>12.5
12.2
>13.2
11.7
>12.2
>13.4
10.9
>13.2
>15.2
10.3
>12.7
>14.7
a. Average of seven leaching intervals totaling 90 days.
b. Initial concentrations in the fixed cylinders not
quantified; assumed to be half those in the raw materials.
A leachability of 5 would indicate a fairly mobile compound, while
a value of 15 would indicate a compound essentially immobile in this
system10). Most of the contaminants of interest in the Tar Pits materials
exhibited leachability indices greater than 10, indicating they were rela-
tively immobile in the fixed coupons. Furthermore, there was no time
trend in the leaching rates of most organic and inorganic contaminants
(calculated leachability indices were the same for all seven leaching
intervals). The sole exception was phenols: in all types of cylinders,
CONTAMINATED SOIL TREATMENT 219
-------
I
a
20 40 60 80
Elapsed Time (days)
800
600
E 400
3
o
01
200
20 40 60 80
Elapsed Time (days)
•Soil 1:1 Soil:Fluff 3:1 SoiliFlufl
Fig. 6
ANS Leachaic pH and Conductivity vs Time.
phenol leaching decreased over lime. Since phenols are weak acids.
their concentration might have been expected to increase as leachatc
pH fell. The fact that it decreased instead suggests that diffusion through
the solid cylinders was the factor limiting phenol leaching.
The leachability indices were used to make rough projections of con-
taminant release from a full-scale, on-site monolith. The projections
assumed an unfractured monolith with the underside contacted by
slowly-moving groundwater. Under these conditions, it was projected
that less than 10% of the lead in the monolith would be released in
1000 yrs (Fig. 7). Leaching of organic compounds from a soil or soil-
fluff monolith also would be less than 10% over 1000 yrs. On the other
hand, phenols, naphthalene and possibly other PAHs were projected
to leach from a soil-tar monolith at higher rates, such that 50% or more
would be released over 1000 yrs. Hydrologic modeling would be required
to ascertain whether these compounds, very concentrated in the soil-
tar monolith, would ever be present at measurable levels in the shallow
ground water.
10"
•o
,0
.-»
I
I
10"
..''..•'
Total PCBi
Phenol*
1 10 100 1000
Yean After Fixation
Fig. 7
Projected Leaching of a Full-Scale 1:1 soil Huff
Monolith at the Tar Pits Site.
Since a full-scale monolith at the Tar Pits is likely to undergo alter-
nating periods of wetting and drying, one set of 3:1 soil-fluff cylinders
was subjected to wet-dry stressing followed by ANS 16.1 leaching.
Surface weathering of the cylinders (as expressed by leachate pH and
conductivity) was accelerated by the wet-dry cycles (which were con-
ducted at room temperature).
However, leachate concentrations of all individual contaminants were
low, and calculated leachability indices were similar to those generated
from simple ANS tests. There was no evidence for enhancement of
contaminant leaching by wet-dry stressing.
CONCLUSIONS
In this study, various physical and leaching tests were used to ascer-
tain whether stabilization/solidification is a viable remedial technology
for the materials at the Tacoma Tar Pits site. Test results indicate that.
for site materials not containing high levels of tar, stabilization is a
promising technology. The physical tests suggest that a monolith of soil
or a soil-auto fluff mixture would be sufficiently strong, durable and
impermeable to meet site-specific remedial goals. A fixed 1:1 mixture
of tar and soil probably would have sufficient physical integrity, but
pure fixed tar probably would not.
TCLP results indicate that in most fixed materials the contaminants
of concern could be immobilized effectively enough that the leachates
would nearly meet cleanup goals without dilution. However, leachates
from fixed tar could not readily meet the goals. Results of the ANS
tests suggest that long-term contaminant leaching from a full-scale
monolith would be very slow. Organic compounds leached from tar-
rich monoliths might be measurable in slowly-moving groundwater.
ACKNOWLEDGEMENTS
This project was funded by U.S. EPA's Environmental Monitoring
Systems Laboratory - Las Vegas, project officer Ken Brown. John Barich
of U.S. EPA's Region 10 directed the study. Time and materials were
donated by the vendor of the fixation product. Silicate Technology
Corporation. Joseph Simon and Sons, the property owner, gave per-
mission for sampling and the use of site materials in the treatability
testing.
220 CONTAMINATED SQIL TREATMENT
-------
DISCLAIMER
Although the research described in this article has been supported
by the U.S. EPA through assistance agreement CR 814701 to the Univer-
sity of Nevada, it has not been subjected to agency review and there-
fore does not necessarily reflect the views of the agency, and no official
endorsement should be inferred.
REFERENCES
1. Applied Geotechnology, Inc., "Remedial Investigation, Tacoma Tar Pits,
Tacoma, Washington," AGI, Bellevue, WA, 1987.
2. Envirosphere Company, "Final Feasibility Study of the Tacoma Historical
Coal Gasification Site." Envirosphere Co., Bellevue, WA, 1987.
3. U.S. EPA, "Decision Summary and Record of Decision, Remedial Alterna-
tive Selection, Final Remedial Action, Tacoma Tar Pits, Tacoma, Washing-
ton." U.S. EPA, Region 10, Seattle, WA, 1987.
4. Rupp, G.L., "Bench Fixation of Soils from the Tacoma Tar Pits Superfund
Site." Project Report. EPA/600/8-89/069.
U.S. EPA Environmental Monitoring Systems Laboratory, Las Vegas, NV, 1989.
5. MSI Detoxification Inc., "Final Protocol for Bench Fixation Trials for Soils
from the Tacoma Historical Coal Gasification Site." MDI, Bozeman, MT,
1988.
6. Center Hill Research Facility, "Summary Test Report, Tacoma Tar Pits."
Technical Assistance for Evaluation of Solidification/Stabilization Treatment
Technologies. University of Cincinnati, Cincinnati, OH, 1989.
7. U.S. EPA, Test Methods for Evaluating Solid Waste. EPA SW-846, 3rd
Edition. U.S. EPA Office of Solid Waste and Emergency Response, Washing-
ton, DC, 1986.
iullinane, M.J., Jones, L.W. and Malone, P.G, Handbook for Stabiliza-
tion/Solidification of Hazardous Waste. EPA/540/2-86/001. U.S. EPA
hazardous Waste Engineering Laboratory, Cincinnati, OH, 1986.
9j/Jones, L.W., "Interference Mechanisms in Waste Stabilization/Solidifica-
tion Processes." U.S. EPA Hazardous Waste Engineering Laboratory, Cin-
cinnati, OH, 1989.
10. PEI Associates, Inc. and Earth Technology Corporation, "Stabiliza-
tion/Solidification of CERCLA and RCRA Wastes." US. EPA Risk Reduction
Engineering Laboratory, Cincinnati, OH, 1986.
CONTAMINATED SOIL TREATMENT 221
-------
Evaluation of Chemically Stabilized/Solidified Soils
Using the California Waste Extraction Test
M. John Cullinane, Jr., Ph.D., P.E.
E. Fleming
Teresa Holmes
U.S. Army Engineer Waterways Experiment Station
Vicksburg, Mississippi
ABSTRACT
Contamination of soils with heavy metals is a significant problem
at many uncontrolled hazardous waste sites. Chemical stabiliza-
tion/solidification (CSS) has been proposed as a technology that reduces
the mobility of heavy metals and other contaminants.
This paper presents the results of a bench-scale investigation of the
capability to reduce the mobility of arsenic, cadmium, copper, lead and
zinc from contaminated soils through application of CSS technology.
Individual samples of contaminated soils were collected from three sites.
Three generic CSS processes were evaluated: Portland cement, lime/fly
ash and cement kiln dust. Physical strength of the specimens was
evaluated using unconfmed compressive strength (DCS). Contaminant
mobility characteristics were evaluated using the Stale of California
Waste Extraction Test (WET).
The results of DCS testing indicated substantial strength development
for all binders tested. The results of the WET procedure indicated
substantial reductions in leaching of all contaminants as a result of CSS.
However, leachate concentrations for arsenic and lead exceeded the State
of California's promulgated criteria.
INTRODUCTION
Chemical stabilization/solidification (CSS) is a process thai involves
the mixing of a contaminated soil with a binder material to enhance
the physical and chemical properties of the soil and to chemically bind
any free liquid7 The CSS process involves the addition of water and
binder material to the soil followed by mixing and a curing period. A
schematic flowchart of CSS processing is shown as Figure 1. Typically,
the binder is a cement or pozzolan. Proprietary additives also may be
added. In most cases, the CSS process is changed to accommodate
specific contaminants and soil matrices. Since it is not possible to dis-
cuss completely all possible modifications to a CSS process, discus-
sions of most CSS processes arc related directly to generic process types.
The performance observed for a specific CSS system may vary widely
from its generic type, but the general characteristics of a process and
its products usually are similar. General discussions of CSS processes
are given in Malone and Jones'; Malone, Jones and Larson*; and U.S.
EPA1.
IT IT
CSS systems that have potential application to contaminated soils
include both Portland cement processes and pozzolan processes. Port-
land cement processes use Portland cement to produce a type of soU/con-
crete composite. Contaminant migration is reduced by
microencapsulation of the contaminants in the concrete matrix and con-
version of the metals to a less soluble form. Pozzolanic processes use
the finely divided, noncrystalline silica in fly ash and the calcium in
lime to produce low-strength cementation. Contaminant containment
is produced by the same mechanisms in for the cement processes.
Proprietary and non-proprietary admixtures, designed to enhance one
or more properties of the mix. may be added to all CSS processes.
The specific objectives of this study were to determine if CSS tech-
niques can be applied to soils contaminated with heavy metals to reduce
contaminant leaching and to characterize the effect of CSS on the con-
taminated soils. Three solidification processes were used to stabilize/
solidify the contaminated soil and are differentiated by the type of binder
material used in the process. The three processes included: Portland
cement, kiln dust and lime/fly ash.
MATERIALS AND METHODS
Materials of Interest
The materials of interest were contaminated soils obtained from three
sites. Contaminants of interest included arsenic, cadmium, copper, lead,
nickel, selenium and zinc. Analytical results from performing the
California Waste Extraction Test (WET)J on the untreated soils are
presented in Tables 1 and 2. The soils were generally classified as clay.
The moisture content of the untreated soils was approximately 55, 25
and 27% for Soils 1, 2 and 3, respectively. A 5-gal composite sample
of each soil was collected. The sample was collected from the top 12
in. of soil. Upon receipt at the laboratory, samples were placed in cold
storage until implementation of the CSS evaluation protocol.
Table 1
Total Threshold Leaching Concentration (TTLC) Meals for Untreated Soil
Figure I
Schematic of the Chemical Stflbili/alion/Solidificution Process.
Paraautar
Arienlc
Cadalim
Copper
Load
Nickel
S.Unlu*)
Zlnr
Criteria
500
100
2.500
1,000
100
5.000
Detection LLcilt
(Hi/kit
0.01
0.005
0,025
0.2
0.04
0.005
0.02
Soil 1
1500 0
» }
170.0
10.0
8* 0
3.0
220.0
Soil 2
93 0
250.0
1,550
37.100
BDL
12.0
233.000
Soil >
(••/ki)
s; o
.2.0
1,000
19.600
70,0
74
97.300
222 CONTAMINATED SOIL TREATMENT
-------
Table 2
Soluble Threshold Leaching Concentration (STLQ
Metals for Untreated Soil
Parameter
Arsenic
Cadmium
Copper
Lead
Nickel
Selenium
Zinc
Criteria
(me/1)
5.0
1.0
25
5.0
--
1.0
250
Detection Limit
(me/I)
0.01
0.005
0.025
0.2
0.04
0.005
0.02
Soil 1
(me /I)
16.0
0.1
1.3
BDL
BDL
BDL
12.0
Soil 2
(me/1)
1.8
7.6
61.0
370.0
BDL
0.21
12,500
Soil 3
(me/1)
1.1
2.7
68.0
500.0
BDL
0.36
10,700
Initial screening test
The objective of the initial screening test was two-fold: the first
objective was to determine the appropriate water to soil ratio (W/S or
WSR), by wet weight, for each CSS process; the second objective was
to narrow the range of binder to soil ratios (B/S or BSR) used for detailed
evaluation. The soil was moist; however, it was necessary to add water
to the contaminated soil to provide sufficient water for effective hydra-
tion. WSRs and BSRs selected for initial evaluation were based on the
previous experience of testing personnel.
Determination of the appropriate WSRs and BSRs for detailed
evaluation was based primarily on the results of the Cone Index Test
(CI) performed on the Initial screening test samples after they had cured
for 48 hr. The CI measures the resistance of a material to the penetra-
tion of a 30-deg, right circular cone using the method specified in TM
5-5304. The CI value is reported as force per unit surface area, in psi,
required to push the cone through a test material at a rate of 72 in./min.
Preparation of Specimens for Detailed Evaluation
Specimens were prepared by mixing water and binder with the
contaminated soil in a Hobart K455S mixer. The resulting slurry was
poured into 2-by-2 in. brass molds. Immediately after the slurry was
placed in the molds, the molds were vibrated to remove voids. At the
higher BSRs, the mixture was very viscous and vibration was an
ineffective method for removing voids. These specimens were tamped
according to ASTM C 109/86'. The molded specimens were cured in
the molds at 23 °C and 98 % relative humidity for a minimum of 24
hr. Specimens were removed from the molds when they developed suffi-
cient strength to be free standing, and curing was continued under the
same temperature and relative humidity conditions until further testing.
Unconfmed Compressive Strength
Unconfined compressive strength (UCS) was used to define and
characterize the effects of the CSS process on the physical characteris-
tics of the soil. The UCS of the treated soil test specimens was deter-
mined using ASTM method C 109/86'. UCS testing was performed
on cubes after they had cured for 21 and/or 28 days. UCS was reported
as the force per square inch, in psi, required to fracture the cube.
BSRs that exhibited UCS values greater than 50 psi were selected
for evaluation of the contaminant-release characteristics of the treated
soil. A UCS criterion of 50 psi was based on Office of Solid Waste
and Emergency Response (OSWER) Policy Directive 9487.00-2A7.
Contaminant Mobility Testing
Since the sites under investigation are located in California, the
California Waste Extraction Test (WET) procedure was used to evaluate
contaminant mobility2. The California procedure requires evaluation
of both Total Threshold Leaching Concentration (TTLC) and Soluble
Threshold Leaching Concentration (STLC). The TTLC/STLC analyses
were performed by IT Corporation, Cerritos, California. Specimens
selected for the extraction tests were forwarded to IT Corporation under
chain-of-custody. TTLC and STLC extracts were analyzed for metals
according to the methods and within the time constraints summarized
in the Federal Register3 and specified in SW-84610
DISCUSSION OF RESULTS
Initial Screening Test Results for Soil 1
In the initial screening test for the cement binder, WSRs of 0.1 and
0.5 were evaluated. The 0.1 WSR did not sufficiently hydrate the mix-
ture, resulting in a dry, powdery specimen. At the 0.1 WSR and 0.4
BSR, the CI value was 647 psi after 48 hr of curing. Although the CI
values were relatively high, the 0.1 WSR did not provide sufficient water
for efficient hydration. In contrast, the 0.5 WSR resulted in very wet
mixtures and yielded specimens with relatively low CI values, 93 and
260 psi for the 0.4 and 0.6 BSRs respectively. Based on past experience,
observation of test specimens and the initial screening data, a WSR
of 0.2 and BSRs of 0.2, 0.4 and 0.6 were selected for detailed evaluation.
The results of the initial screening test using kiln dust as the binder
were similar to the cement binder results. Although relatively high CI
values were obtained at the 0.4 and 0.6 BSR (693 and >750 psi, respec-
tively) at the 0.1 WSR, the samples were not adequately hydrated and
dry, powdery mixtures were produced. Using a WSR of 0.5 and BSRs
of 0.4 and 0.6, the samples were very moist and developed relatively
low CI values (185 and 167 psi, respectively). Based on past experience,
observation of test specimens and the initial screening data, a WSR
of 0.2 and BSRs of 0.2, 0.4, 0.6 and 0.8 were selected for detailed
evaluation.
The results of the lime/fly ash initial screening test were similar to
cement and kiln dust binder results. At 0.1 WSR, the samples evaluated
were not sufficiently hydrated and at 0.5 WSR the samples were too
wet for efficient hydration. Based on past experience, observation of
test specimens and the initial screening data, a WSR of 0.2 and BSRs
of 0.2L/0.2F, 0.2L/0.4F, 0.4L/0.2F and 0.4L/0.4F were selected for
detailed evaluation.
UCS Results for Soil 1
The results of the UCS tests for Soil 1 are shown in Figure 2 and
discussed below.
KILN DUST .2 LIME/FLY FISH
BINDER
.4 LIME/FLY HSH
Figure 2
Twenty-eight Day UCS for Soil 1 Using Cement, Kiln Dust and
Lime/Fly ash as Binders
For the cement binder, the 28-day UCS increased as the BSR in-
creased, with the UCS doubling for each 0.2 increase in BSR. For
example, at 0.2, 0.4 and 0.6 BSRs, the average UCSs were 134, 278
and 640 psi, respectively. Similar results were obtained when the UCS
tests were run on 21-day cubes, indicating that the sample reached near
maximum UCS at a 21-day cure time.
UCS results for the kiln dust binder were similar to those obtained
with the cement binder. The 28-day UCS increased as the BSR was
increased and a BSR of 0.8 developed three times the 28-day UCS of
the 0.2 kiln dust BSR. The contaminated soil treated with a 0.2 kiln
CONTAMINATED SOIL TREATMENT 223
-------
dust BSR developed a 28-day UCS of 119 psi, and the 0.8 BSR developed
a 28-day UCS of 370 psi. As the BSR increased, the UCS also increased
but at a decreasing rate with each increase in the BSR. The 0.6 and
0.8 ratios had approximately equal 28-day UCS results, indicating little
or no strength gain for BSRs greater than 0.6.
The interpretation of the lime/fly ash UCS data is more difficult than
the cement and kiln dust UCS data because both the lime BSR and
the fly ash BSR were varied. The 0.2L/0.2F BSR had UCS results.
155 psi, similar to the 0.2 cement and 0.2 kiln dust BSRs (134 and 119
psi, respectively). The average 28-day UCSs for 0.2L/02F and
0.2L/0.4F were 154 psi, and 182 psi, respectively. A 02 increase in
the lime BSR resulted in doubling the UCS. wuh 04L/0.2F and
0.4L/0.4F UCSs of 330 psi. and 389 p.si. respectively compared i<> 155
and 189 psi, for the 0.2L/0.2F and 0.2L/0.4F.
Extraction Test Results for Soil 1
As shown in Figure 2. all the binders, at the BSRs investigated,
developed 28-day UCS well above the 50 psi. selection criterion, hence,
the specimens with the minimum BSR were selected for WFT analy-
sis. The BSRs selected for extractions included: 0.2 cement. 0.2 kiln
dust, 0.2 lime/0.2 fly ash. The results of the WKT for treated Soil 1
are given in Table 3. The TTLC resulis reflect the dilution resulting
from the addition of the binder material. The STLC results reflect sub-
stantial reduction in the apparent leachability of the coniaminants.
However, leachate arsenic concentrations for all binders exceeded the
STLC arsenic criterion (5 mg/L) by a factor of three to seven. Other
contaminants were less than their respective criteria. Of the binders
evaluated, Portland cement appeared to perform the best, with kiln dusi
and lime/fly ash demonstrating roughly equal performance.
Table 3
Results of Solidification/Stabilization Studies on Soil I
ITU: TTL. true ITU:
ITU TTLC ITU.
M/ll
3«* e 10 IMO o it i no j
106 c 10 • i a 10 i) or
ZMt e 21 » iro o I > l»o a
looe o to 10 o «>» 0 1*0 or
1M 0 I • 10 OD«« « » 0 «KO 0}
MM 0 I* • 120 0 II 0 Ul 1) 0 ir 1M 0
to or lit) » at •*
o zl it • o i>
» i) it* or oe
1 tr Ml 11 I 01
Initial Screening Test Results for Soil 2
In the initial screening lest for the cement binder. WSRs of 0.2 and
0.5 were selected for evaluation. A 0 2 WSR did not thoroughly hydrate
the sample, but a 0.5 WSR provided a mixture of good consistency and
adequate CI values The 48-hr CI value at a 0.5 WSR and BSR.S ot 0.4
and 0.6 were 133 psi, and 467 psi, respectively. Based on past
experience, observation of test specimens and the initial screening data.
a WSR of 0.5 and BSRs of 0.4. 0.6, 0.8 and 1.0 were selected for detailed
evaluation.
Based on the moisture contents and the results of the initial screening
test for the cement binder. WSRs of 02 and 0.5 were also used lor the
kiln dust binder initial screening lest For the same reasons listed for
the cement binder initial screening lest, a 0.5 WSR was selected lor
detailed evaluation of kiln dust. Because the initial screening test results
for a 0.5 WSR and 0.4 BSR were low, a 0.6 BSR was selected as the
lowest BSR for detailed evaluation. Other BSRs selected lor detailed
evaluation were 0.8, 1.0 and 1.2
For the lime/fly ash binder, WSRs of 0.2 and 0.5 were initially evalu-
ated. Because both lime and fly ash arc used, the mixtures using a 0 2
WSR were much dryer than the cement and kiln dust mixes prepared
at a 0.2 WSR. A 0.5 WSR resulted in mixtures of good consistency
and CI values, ranging from 125 to 517 psi, for the BSRs evaluated and
was selected as the WSR for detailed evaluation. The 0.2I./0.2F BSR
was selected as the lowest ratio for detailed evaluation because sub-
stantial strength was gained after only 2 days of cure. Based on past
experience, observation of test specimens and the initial screening data,
the BSRs selected for detailed evaluation were 0.2L/0.2F, 0.2L/0.4F,
0 2L/0.6F, 0.4L/0.21, 0.4L/0.4F. 0.4L/0.6F, 0.6L/0.2F. 0.6L/0.4F and
0.6L/0.6F
l< S Results for Soil 2
The results of the UCS tests for Soil 2 are shown on Figure 3 and dis-
cussed below.
B:-OER
Figure 3
Twcmyeight Da> LCS (or Soil 2 Using Cement. Kiln Dust and
Lime'Fly ash 15 Binders
Each BSR for the cement binder exceeded the 50-psi, criterion. At
the lowest BSR, 04. the average UCS was 198 psi.. The average UCS
increased approximately 100 psi, with each 0.2 increase in the BSR.
Each BSR for the kiln dust binder exceeded the 50-psi. criterion.
As expected, the 0.6 kiln dust BSR (114 psi.) did not develop as much
strength as the 0.6 BSR for cement. As the BSRs were increased by
increments of 0.2, the UCS also increased until the BSR reached 1.2.
At this point, the kiln dust dehydrated the sample very quickly producing
an extremely dry mixture, with a resulting decrease in UCS
For the lime/fly ash binder, the average UCS did not increase sig-
nificantly as the BSR was increased. At 0.2L/0.2F, the average UCS
w-as 85 psi, and at the 0.6L0.6F BSR the average UCS was 260 psi.
The most significant increase in UCS w-as at the 0.4L 0.6F BSR, which
exhibited an increase of 100 psi. over the 0.4U04F BSR. Similarly
to Soil 1 results, the increase in lime had more effect than an increase
in fh ash
Extraction Test Results for Soil 2
As shown in Figure 3. all the binders, at the BSRs investigated.
developed a 28-day UCS well above the 50 psi. selection criterion. The
BSRs selected for WET extraction included: 0.6 for cement, 0.8 for
kiln dust and 0.2/0.4F for lime/fly ash. The results of the WET for
treated Soil 2 are given in Table 4. The TTLC results reflect the dilu-
tion resulting from the addition of the binder material. The STLC results
reflect substantial reduction in the apparent leachability of the con-
taminants. However. Icachalc lead concentrations for all binders
exceeded the STLC lead criterion (5 mg/L) by a factor of seven to 22.
Zinc exceeded the STLC criterion for the lime/fly ash binders by a factor
of 10. Other contaminants were less than their respective criteria. Of
the binders evaluated, Portland cement appeared to perform the best,
with kiln dust and lime/fly ash demonstrating roughly equal
performance.
224 CONTAMINATED SOIL I HI ATMENT
-------
Table 4
Results of Solidification/Stabilization Studies on Soil 2
Standard Untr.tat.vd Soil Camant Kiln Dual Lima/Flyaih
TTLC STLC TTLC STLC TTLC STLC TTLC STLC TTLC STLC
250.0 233000.0 12500.0 147333.0 213.0 119333. t
49.0 135333.0 2500.0
Initial Screening Test Results for Soil 3
The two WSRs evaluated for the cement initial screening test were
0.2 and 0.5. The 0.2 WSR resulted in samples that were not adequately
hydrated resulting in a dry, powdery mixture. In contrast, the 0.5 WSR
was very moist. The 0.5 WSR and 0.6 BSR had a CI value over 750
psi. At 0.5 WSR and 0.4 BSR, the CI value was 150 psi. These results
indicated that by lowering the WSR, the BSR also could be lowered,
while obtaining similar strength results. Based on past experience, ob-
servation of test specimens and the initial screening data, a WSR of
0.35 and BSRs of 0.2, 0.4 and 0.6 were selected for detailed evaluation.
WSRs of 0.2 and 0.5 were used for the kiln dust initial screening
tests. The 0.5 WSR resulted in a very wet mixture, with low CI values
of 33 psi, and 73 psi, at 0.4 and 0.6 BSRs, respectively. At a WSR of
0.2 and BSRs of 0.4 and 0.6, the CI values were much higher, 550 and
567 psi, respectively. Although the CI values for the 0.2 WSR were
adequate, a WSR of 0.35 was chosen for detailed evaluation because
laboratory notation indicated that the samples were not effectively
hydrated at the 0.2 WSR. BSRs of 0.2, 0.4 and 0.6 were selected for
detailed evaluation.
WSRs of 0.2 and 0.5 were used for the lime/fly ash initial screening
tests. For the 0.2 WSR, CI values ranged between 490 and 533 for the
BSRs evaluated. For the 0.5 WSR, CI values ranged between 63 and
343 psi,. Like the cement and kiln dust, a 0.2 WSR as too low and
a 0.5 WSR was too high. The 0.5 WSR did not achieve significant
strength except at the highest BSR, 0.4L/0.4F, evaluated. A 0.35 WSR
and BSRs of 0.2L/0.2F, 0.2L/0.4F, 0.4L/0.2F and 0.4L/0.4F were select-
ed for detailed evaluation.
UCS Results for Soil 3
The results of the UCS tests for Soil 3 are shown on Figure 4 and
discussed below.
300
225
150
75
BINDER RHTIOS
D .8
KILN DUST .2 LIME/FLY SSH .a LIME/FLY PSH
BINDER
Figure 4
Twenty-eight Day UCS for Soil 3 Using Cement, Kiln Dust and
Lime/Fly ash as Binders
For the cement binder, the UCS of the 0.2, 0.4 and 0.6 BSRs were
69 psi,, 173 psi, and 254 psi, respectively. As the BSRs increased, the
rate of increase in UCS decreased. From 0.2 BSR to 0.4 BSR, the UCS
increased by 104 psi, but from 0.4 BSR to 0.6 BSR, the increase was
only 81 psi,.
The kiln dust binder results contrasted with those of the cement binder
evaluation. The UCS of the 0.2, 0.4 and 0.6 BSRs were 19 psi, 64 psi,
and 128 psi, respectively. The 0.2 BSR did not obtain enough strength,
only 19 psi, to pass the 50-psi, criterion. As a result, 0.2 BSR was not
evaluated further. The UCS increased with increases in the BSR, tripling
from 0.2 BSR to 0.4 BSR and doubling from 0.4 BSR to 0.6 BSR. As
expected, kiln dust was not as effective as cement in this segment of
the solidification/stabilization of Soil 3.
The BSRs tested for lime/fly ash were 0.2L/0.2F, 0.2L/0.4F, 0.4L/0.2F
and 0.4L/0.4F with a WSR of 0.35. The respective average UCSs were
111 psi, 143 psi, 206 psi, and 200 psi. Increasing the fly ash BSR by
0.2 resulted in a 30-psi, increase in UCS; however, increasing the lime
BSR by the same amount doubled the UCS. Increasing both lime and
fly ash caused the UCS to triple at 0.4L/0.4F.
Extraction Test Results for Soil 3
As shown in Figure 4, all the binders, except the 0.2 BSR kiln dust,
developed a 28-day UCS above the 50-psi, selection criterion. The BSRs
selected for extraction included: 0.4 for cement, 0.4 for kiln dust and
0.2L/0.4F for lime/fly ash. The results of the WET for treated Soil 3
are given in Table 5. The TTLC results reflect the dilution resulting
from the addition of the binder material. The STLC results reflect sub-
stantial reduction in the apparent leachability of the contaminants.
However, leachate lead concentrations for all binders exceeded the STLC
lead criterion (5 mg/L) by a factor of nine to 46. Copper concentra-
tions slightly exceeded the criterion. Other contaminants were less than
their respective criteria. Of the binders evaluated, Portland cement
appeared to perform the best, with kiln dust and lime/fly ash demon-
strating roughly equal performance.
Table 5
Results of Solidification/Stabilization Studies on Soil 3
Standard
TTLC STLC
atad Soil
STLC
Comont
TTLC STLC
Kiln Duat
TTLC STLC
Llma/FXyash
TTLC STLC
Afsenlc
Cadmium
Copper
Laad
Hlckol
Salenium
Zinc
0 06
2500 0
1000.0
1000.0
10600.0
70.0
68.0 090.0
500.0 13067.0
HD<4 64.0
653.0
14600.0
69.7
0.13 53 0
0.21 35.7
30.0 675.7 33.0
66.0 13666.7 220.0
HIK0.4 72.7 KIKO.lt
CONCLUSIONS
A laboratory study was conducted to investigate the effects of three
CSS processes on a contaminated soil. Both UCS and WET tests were
performed on the stabilized/solidified specimens and based on the results
of these tests, the following conclusions can be drawn:
• Small quantities of binding agents produce materials with UCS well
above the 50-psi, criterion.
• Water must be added to the contaminated soil in order for the binders
to develop strength.
• The binders can be easily mixed with the contaminated soil.
• The stabilized/solidified soil sets within 24 hr and no free liquid was
observed after this 24-hr period.
• The CSS processing of the soil effectively reduced the mobility of
the contaminants in the soil.
• Because of the high concentrations of the contaminants and/or the
aggressiveness of the WET procedure, none of the CSS processes
produced a product that meets the California Department of Health
Services TTLC/STLC criteria.
CONTAMINATED SOIL TREATMENT 225
-------
REFERENCES
1. American Society for Testing and Materials, Construction; Cement; Lime;
Gypsum. Vo\ 0401, Annual Bank ofASTM Standards. Philadelphia, PA, 1986.
2. California Administrative Code, "California Hazardous Waste Regulations."
Department of Health Services, Section 66700, July 29, 1985.
3. California Department of Health Services. "California Site Mitigation
Decision Tree," Sacramento, CA. 1986.
4. Headquarters. Department of the Army, "Materials Testing," Technical Manu-
al No. 5-530, Section XV, Washington. DC, 1971.
5. Malone. P. G. and Jones, L. W., "Survey of Solidification/ Stabilization
and Technology for Hazardous Industrial Wastes," EPA-600/2,79056, U.S.
EPA, Cincinnati, OH, 1979.
6. Malone, P. G., Jones, L. W. and Larson, R. J., Guide to the Disposal of
Chemically Stabilized and Solidified Waste. SW-872, Office of Water and
Waste Management. U.S. EPA, Washington. DC. 1980.
7. US. EPA. Office of Solid Waste and Emergency Response (OSWER) Poll
cy Directive 9487.00-2A. Office of Solid Waste and Emergency Response.
Washington. DC. 1986.
8. U.S EPA. Handbook for Stabilization/Solidification of Hazardous Wastes,
Hazardous Waste Engineering Research Laboratory. Cincinnati, OH, 1986.
9. U.S EPA. Federal Register. 51. No. 142, Office of Solid Waste, Washing-
ton, DC, Nov. 7. 1986.
10. U.S EPA. Tea Methods for Evaluation Solid Waste: Physical/Chemical
Methods. SW-846, 3rd ed.. Office of Solid Waste and Emergency Response.
Washington. DC, 1986.
ACKNOWLEDGEMENT
The tests described and the resulting information presented herein,
unless otherwise noted, were obtained from research conducted by the
U.S.A.E. Waterways Experiment Station, Vicksburg, Mississippi. The
study was sponsored by the U.S. Department of the Navy, Facilities
Engineering Command, Western Division, San Bruno, California.
Mr. Carl Schwab was the Department of the Navy project manager.
Permission to publish this information was granted by the Chief of
Engineers.
226 CONTAMINATED SOJL TREATMENT
-------
In Situ Remediation of Groundwater and Soils:
Seminar Outline
Scott B. Wilson
Groundwater Technologies, Inc.
Concord, California
BRIEF HISTORY
• 1946 - CE Zobell publication: Hydrocarbon Microbiology
• 1950s - Single cell protein development showed relationship of
bacteria/oil/water
• 1967 Focus on ocean oil spills and bacterial degradation: first
environmental focus
• 1970s - Microbial enhanced oil recovery technology showed rela-
tionship of bacteria/oil/water/mineral
• 1972 - First application of in situ aquifer bioremediation, Sun
Oil Company under the guidance of Paul Yaniga (Groundwater
Technology, Inc.)
• 1980s Bioremediation technologies shown to be a proven and
cost-effective remediation strategy for many organic waste con-
taminated sites.
BIOMECHANISM OF ORGANIC CHEMICAL DEGRADATION
• Aerdbic Oxidation
• Use of oxygen as terminal election acceptor
• Most rapid biomechanism for degradation of most organic
wastes
• Products are oxidized intermediates, carbon dioxide and water
• Co-oxidation
• Use of co-metabolite to degradation of an otherwise recal-
citrant compound.
• Harder to control than simple aerobic oxidation.
• Rates vary greatly depending upon pollutant, co-metabolite,
etc.
• Anaerobic Dehalogenation
• Simple biologically mediated redox reaction
• Requires very anaerobic condition
• Simply dehalogenates substrate, not mineralized
BIODEGRADABILITY OF ORGANICS
• Unsubstituted Hydrocarbons
• Rapidly degrade
• Multiple ring structures are more resistant
• Resistance imparted by substitution
• Halogens
• Ethers
• Methoxy groups
• Etc.
• Molecular Complexity
• Greater the complexity - greater resistivity
BACTERIAL ATTACHMENT TO ORGANIC POLLUTANTS
• Cell surface hydrophobicity
• Hydrophile - lipophile balance of cell surface
• Oil/water partitioning
Biosurfactant production
Activity of cell surface
Surface activity of biomolecules
Action on pollutants
forms of pollutants available to bacteria
Droplets
Dissolved molecules
Microemulsions
BACTERIA CAPABLE OF IN SITU BIODEGRADATION
• Dependent upon pollutant
• Dependent upon environmental factors
• New organisms being discovered regularly
BACTERIA FOR USE IN AN IN SITU
DECONTAMINATION SYSTEM
• Genetically engineered strains
• Not likely due to regulatory constraints
• Acclimated strains
• Generally unnecessary for most applications
• Almost impossible to use effectively in subsurface ground-
water systems.
• Stimulation of indigenous strains
• Generally best way to approach bioremediation
• Appropriate bacteria generally present
NUTRIENTS NECESSARY FOR IN SITU AEROBIC
BIOREMEDIATION SYSTEMS
• Oxygen
• Air spargers
• Hydrogen peroxide
• Soil aeration
Nitrogen
Phosphorous
Trace elements
pH
Temperature
NUTRIENTS NECESSARY FOR
CO-OXIDATION BIOREMEDIATION
• Same as for Aerobic System with the Systems Addition of a
co-metabolite.
NUTRIENTS NECESSARY FOR ANAEROBIC
DEHALOGENATION SYSTEMS
• Same as aerobic oxidation systems except:
CONTAMINATED SOIL TREATMENT 227
-------
• No oxygen should be added
• Need supplementary carbon sources
IN SITU BIOREMEDIATION SYSTEM DESIGN
• Saturated Zone
• Hydraulic control
• Infiltration
• Closed loop recycle
• Nutrient system
• Monitoring
• Unsaturated zone
• Soil aeration system
• Forced air
• Negative pressure
• Nutrient percolation
• Soil pore moisture
• Monitoring
ON-SITE BIOREMEDIATION SYSTEM DESIGN
• Soils/Sludge treatment
• Cell construction
• Aeration
• Nutrient addition
• Off-gas treatmeni
• Monitoring
CASE STUDIES
CASE n
In Situ Bioremediaiion: A Case Study through Closure.
* Contaminated site setting
• Bioremediation system design
• Performance of system/site data
• Post-closure monitoring
• Associated costs of project versus other treatment technologies
CASE rtt
In Situ Bioremediation: Unsaturated zone and Saturated Zone.
• Contaminated site setting
• Bioremediation feasibility study
• Comprehensive system design
• Performance of system/site data
• Associated costs of project
CASE/«
On-Site Bioremediation of Heavy Oil Contaminated Soils.
• Site setting
• Feasibility study
• Site excavation/construction
• Nutrient amendment
• Performance data
• Associated costs of project versus other treatment technologies
Appendix I
In Situ Bioreclamation: A Cost-Effective Technology
to Remediate Subsurface Organic Contamination
WtttonmJKd^nlA. Avwi
Abstract
le sau bOTeclaeaatloa • • prove* letbisusugy that ooft of
microor|ifiiuH to «jt bydrocartkom «m dcptadtw upon
environmental factors, u well as iht chemical makeup of
the organic substrate Tht appbcalion of hydrocarbon-
degrading bacteria u not new and ulerwure u rcaddy
available on rtUud top.cs, nich as iht production of
surfactant* in hydrocarbon fermentation (Suzuki cf al
1969) and the bacterial extraction of bcturaca from tar
sandt (Genoa and Zjj* 1977 and Vryndham and Caster-
Ion IOTI) Atlas(l9tl)haspreacnu>daaci and Kt uhunatr
(ate dciermlnt whelhtr aa la situ trratmrni promt can
be uaed or whether coauiament or phyaieal renoval 9
more appropruu ThcbaftibdjtyofaraoMdiauonptoccu
M dcurmined by raatchtrtf. vvailabk tachaolofy wuh aa
imormandirtf, of contanuaanl trampon and reaction
contamiaaau pott ntuujy CAOI u the tubtui
'ace at a vapor pnaM and three condemed phaata
(I) raobde free product (phaae arparaud), (1) natduaU*
coatanuaatad foil (aorbad ahaae). and (!) coniarrunawl
fround water (diaaorvad phlic). Tbc dailnbuuoa of coa-
taraioanu inlo theae duTeraat peuBea, whur a rouh of
dynamic tranapon. a otonalery a runcuon of their phyaol
and chemical properuea aad the hydroaaolo|fcal and
(eochenucal characuriatio of lie formation Ox mun
eiamine lot phaa. datrwuUaa ta lemi of the volumt of
the tuhnirfaoe impacted by a phaae aad the amount of
the conlaminani within a abate,
Tabk I repr»BUlhUmaunaf
t taioline tpU in a medium land aqtafej (Of oundw.iti
Technolofy Inc. 1913), There are tcveral ftneraluatiora
that can b< made froea ihJa. Ftnt. pound waur flow a
the primary loiti^erra medunlem for the tpread of wn-
uuninalloii once the frae product layer hat achievad fkw
equilibrium. Became of the hgjr, mobttey of ground
water, the volume of ground wauw eontainiiiMed a greater
than the other phaata Thb ahatc, however, conuira only
Winter Itn CWMH IT)
TABU I
Phne, DhnaVtUon of a Trpaol MJM CiaVal C«
SpaT
Ca *«
%af
Adsorbed (toj)
Doaorved (water)
100
ZWjOQO
96MOO
\»
SO
no
a
a
1-5
•Mnbn mid a>»der. araU v •
II lea
• im.U.rrKUoooi'tteioi^pr.^unfiloJ.^raM) MM
of the comamuuni (63 pcnx« of the a
proem M (rar-phuc aatentL
*T}x9KoadtibfrM>im\obt»»At*\!kM\i*m
o(o..aer.aJ in ihe (rouMl wcm • imifl eo.BpdUf^ w ta«
rtl.un.rf.naK.^a..tfu(l)pmri»i Ub-vHoo «•••>•<>
h*« iho«n (Browa ci ii M7) ihai tmokr* rnKtaal
uiuranon capwity M4i7 -«ai vncrf ffoa* Utt% •"•»•
b«/kf i«J (JIJOO ppo) (« eo..w uad tta-, 1 MM
IUM| •Hate/Matt*
diamrurl Saadt at (Seld caaanty t
lower residual saturation faftantm than dry t
Al Arid capacity mom m coastas. reiahal satoauoa
otpaesua rantjnt bom » Slafasttae ais the soj pore moisture and atfsster waten
tj nimaiaiiiuiln tppmraialf nariitns in tar rttarar
fan. nwarasaauoa a anatted aad ocp»M by the
previoissry described ostchaasstas at ptaoe « a ow in sstu
ireatmeai proccas. Tbst o list oaly ia BAB treatateat
nstthod praveti la date that addreoes both the tasorxd
coatajBunaMs, as we! as aa aouror, the ressoval sod
PrindnJe. of Apptkatloti
Twt proper appbcaiMB cl ia situ bioncUtaaliM
microbioloty. aad eafuteenng For biortclasBatioa to
be svoottaiul there ansi be aa appropriate bactenal
comrounlly. the ayirototy asusl alkjw for litMt)
uaiupon through the rubsurf ace and the soil and wam
chemist ry BtttM be coeapalibk with the iaucjtfaclMa of
necessary tmlrieau.
Lahentory KM S>wa>
Prior lo the field apphcauoa of any ia ot. bioredt
malioo system, aa anesunenl thould be made of the
ground WBWT aad tog ouatty acraa the trie with roped
to chemical and physical parameters, as weB at bactenal
enumeration and contaminant ennctmralnnv A labors
228 CONTAMINATED SOIL TREATMENT
-------
n
o
>
o
2
r
REGIONAL
CROUNDWATER
GRADIENT
PASSIVE IN SITU BIORECLAMATION SYSTEM
. Ptaivc b Au bionduBrtfan tjtltm.
lory pilot study should be conducted to discern the
biodegradability of the contaminants and the subsurface
parameters necessary to stimulate maximum degradation.
This is accomplished by a microcosm evaluation, which
consists of asoil/water slurry of site material designed to
emulate the site environment. Inorganic nutrient (e.g.
nitrogen and phosphorous sources) and oxygen (where
applicable) concentrations are adjusted to give the greatest
rate of degradation. In complex mixtures, the rates and
sequences of specific compound degradation should be
evaluated. Based on the microcosm studies, an estimate
of remediation time can be derived under these optimum
conditions. Once this is accomplished, an engineering
evaluation should be conducted to determine the feasibility
of obtaining these conditions in the subsurface and to
define potential problems that could arise during the
implementation of the in situ program.
Conceptual System Design
Proper design of an in situ bioreclamation system
provides for mass transfer into and out of the contami-
nated area. This can be achieved through a passive system
by simply infiltrating nutrients into the contaminated
subsurface (Figure 1). Often, however, at sites where
ground water is being recovered, rcinfdtration of aug-
mented site water is performed in a dynamic system
design (Figure 2). The advantage of a dynamic system
design is that induced hydraulic gradients can be created
to control flow and thus movement of nutrients and
contaminants.
It is generally unnecessary to add bacteria to the
subsurface. Bacteria capable of degrading a wide range of
organic contaminants have been shown to exist in sub-
surface environments (McKee, et al. 1972, Litchfield and
Clark 1973) and can be stimulated to degrade the contam-
inants of concern as indicated by positive laboratory pilot
study results.
Oxjj«n Supply
For those systems requiring aerobic microbial pro-
cesses, oxygen is generally the limiting factor to the bio-
reclamation process. Support for this is shown in reports
by many workers including a study by Wilson el al.
(1985), in which it was concluded that oxygen supply was
limiting creosol biodegradation in contaminated aquifer
material. One method of supply is the feeding of atmos-
OXYGEN AND NUTRIENT
ADDITION.
REGIONAL
GflOUNDWATER
GWUMENT
DYNAMIC IN SITU BIORECLAMATION SYSTEM
Fl(we 2. Drrumfc k> ritu MortebmtfM tjtttm.
Winter 1989 GWMR
pheric oxygen through air spargers. These spargers,
generally of silicon carbide, have the capacity to diffuse
atmospheric air into ground water at a rale of up to 10
cubic feet of air per minute when placed in wells and
driven by an adequate air compressor. Problems asso-
ciated with this approach to aeration, however, can be
critical. Biofouling (sliming) of the sparging surface occurs
in the presence of the induced aeration, often inhibiting
flow of air outward into the well bore. Biofouling of the
well^ filter pack can also occur, decreasing flow and
diffusion into and out of the native soils. Remedying
these problems generally entails the laborious pulling of
the spargers for cleaning and the treatment of biofouled
wells with an appropriate chemical.
Oxygen can also be supplied to ground water chemi-
cally by the use of hydrogen peroxide solutions. Hydrogen
peroxide decomposes naturally in the presence of heavy
metal catalysts and certain microbial enzymes (e.g., cata-
lase) to produce water and oxygen:
2H2O2 - 2HjO + Oj.
This decomposition is completed over time and has
been shown to present no hazards lo ground water (Texas
Research Institute 1982). Although used as an antiseptic
at high concentrations (i.e., 3 percent solution), hydrogen
peroxide can be supplied to hydrocarbon-utilizing bacteria
at up to 2500 ppm without showing cytotoxic effects
(Texas Research Institute 1983). The proper application
of hydrogen peroxide to the subsurface allows for adequate
oxygenation of the contaminated ground water while
controlling the biofouling of the infiltration points.
Nutrient Supply
The nutrients necessary to stimulate bacterial degra-
dation in the subsurface should be studied and defined at
the laboratory pilot study stage. Generally, however, nut-
rient requirements consist of phosphorous and nitrogen.
Phosphorous can be supplied in one of several forms but
common sources for ground water augmentation are
orthophosphate and polyphosphate salts. Ii should be
recognized that phosphates are readily adsorbed onto
soils and if improperly applied can precipitate from solu-
tion, affecting the hydraulic conductivity of contacted
strata.
A wide range of nitrogen sources have potential for
use in situ. Common supplies arc ammonium salts, as
these are inexpensive, easy to handle, and the ammonium
ion is readily assimilated in bacterial metabolism. Any
time a nitrogen source is added to ground water (particu-
larly when inducing aerobic conditions) it is important to
monitor levels of the inorganic nitrogen species, including
ammonia, nitrite, and nitrate. This allows for an under-
standing of the total quantity of inorganic nitrogen avail-
able and of the subsurface oxidation-reduction potential
Trace dements, by and large, are available in sufficient
quantities in ground water environments. These include
calcium, magnesium, manganese, iron, sulfur, etc. How-
ever, in situ biodegradation systems have been applied
where native ground water quality was augmented with
trace elements to ensure maximum biodegradation as
defined in initial laboratory pilot studies.
176 Winter 1989 GWMR
Monitoring and Amu>*
The proper operation of any in situ bioreclamation
system includes the monitoring of ground water quality
with regard to dissolved oxygen, pH, inorganic water
chemistry, organic water chemistry, and bacterial enu-
meration. Particularly critical is the initial startup period
when infiltration of nutrients (and oxygen) is begun
Such shakedown monitoring establishes subsurface
transport patterns, nutrient transport times, oxygen con-
sumption rates, and contaminant release and degradation
rates. Throughout operation, monitoring should be con-
ducted to ensure optimum conditions are being maintained
in the subsurface to affect maximum degradation rates.
Monitoring of the system may also be necessary to ensure
that any hydraulic control requirements are being met 10
contain contaminant and nutrient migration.
Case Study
The case study lo be discussed is the remediation of a
gasoline contamination site. The leak was estimated at a
volume of 900 gallons of leaded gasoline, the dissolved
fraction of which was impacting a municipal water supply.
The soils impacted were calcareous silly sandy clays and
calcareous argillaceous silty and/or gravelly sands. This
was underlain by a highly fractured and solution-
channeled impure limestone and dolomite aquifer of high
productivity. The area! extent of the dissolved contami-
nation within the aquifer was within a 100-foot radius of
the spill site.
After adequate assessment of the site conditions, a
comprehensive ground water remediation program was
undertaken. The program involved hydraulically con-
trolling the initially observed dissolved hydrocarbon con-
tamination using a water-table depression pump. This
central pumping point served as a collection for phase-
separated hydrocarbons using a dual-pump system
(Yamga 1982). Recovered water was treated via a
counter-current flow air stripping system. The total
amount of phase-separated product recovered was
approximately 100 gallons, with excavation of contami-
nated soils removing an estimated 50 gallons of hydro-
carbons. The balance of contamination was bound in the
soil as the adsorbed phase and dissolved in the ground
water.
A laboratory pilot study showed a hydrocarbon
degrading consortium of bacteria present within the con-
taminated aquifer (approximately 1.0 x 10* cfu/mL).
This consortium was shown to provide maximum bio-
degradation of the contaminants under aerobic conditions
with ammonium chloride and sodium phosphate nutrient
augmentation. The geochemistry of site soil and water
appeared to be compatible with this mixture of inorganic
salts, showing no effects upon soil permeability or infil-
tration even when subjected 10 hydrogen peroxide feed.
An in situ bioreclamiiion system was implemented at
the site. Recovered water treated through the air stripper
was augmented with ammonium chloride and phosphate
salts, as well as hydrogen peroxide and reintroduced to
the contaminated subsurface through an infiltration
gallery in a dynamic-type in situ bioreclamation system.
m
N)
N)
'•O
-------
0
o
Z
>
m
a
\ddu>onjj laadnf of aatneati and peroude w« ooe-
ducicd by bucfa (oaJux| of tuuepc vck
AAo 11 maatte of opentt« tht in «B bMHviMnMiaa
tysicaa. aparojuauic}} 99 peraari of the imdaaJ wtl-
bound and dosoMrf hydrocarbon* had bem degraded,
•mm qualflr to the recovery v«g atdaraKd I 0 pan pn
bdboa toul peirafcrum hydrocarbon* (F^wm > * M
Ttm upracau DOR thaa (JO fiflom of peiraanNn
hydrocarbon budcpvded, cocapond lo oaf> » (aBoa*
of nducina attributed to av anpcMf of the rectwred
The coct-cflectmaeB of • Ma
ocber tiuunrM tecfeaoiofio • UK ran* of r^> (anon
Fmi, bnradvuuoa «tD um boU UK toi «d tac
ia. tft*B dmmtmc tfcc KWTX afc
*cfl u the iTtapt
a aton rtfNd thae exiraaioa and lk« ibe U8K of
a a feaervffy * very c
ud Saaih t9l5)
The cam far ippeyuil b
reflect (I) the phenol vnp ic j.
i : i uiuou.il o( equqHOenL. (3) the
».V (t\ the •cmoes nqvarvd, bock aaaryucal and
B (ar «npeNO( Umr). aa aaarym caa be
made of ibt COM beaefji of copto-paf ia an bwreelaa»a-
IMM owcr lAeceni aamntind ma pump-wnj « required
S*mce reqnrcd
PbyiiuJ Klnp
Phn«:»l in up
N.P.O demand per lb comamiruni
Projtti lime
TotaJ notrwni dcnuod
Sumtxr M wilt. Bquipmeni
Proycdtiinc
Nvaobcr of ««lls, eqinpocni
Couof *clh
Ptojeci nine
The pellet uuiixJ izotu of implementing the in MB projected towrai rale) or proyeoed COB mnbuted lo
an additional opcrHiA| tune of iwo yean by pump-and-
irtaiiccbaototyaioca. Howw. wih«pom H^r «»n Summary
OCWfatma) (he purr.p *nd irrji technotofy «ouU h**e la «• hwccdaauixM 3 i pro^a techAotof> thjt
*rhurxr irmrdi*i'. and m*erob»o*oft of ihr t^r
rcTkct ihe praciH ««nh of opendnm (capit«t and orer- Fo* bioree&aa>atM>a u> be ncccHfuJ then man be KB
Mioa and muri'trum com o*rt operBUOA umc •' a appropruac bacvnal c
WVatarmiCWMI
-------
illow lor timely transport of nutrients, and Che soil and
water chemistry must be compatible wilh the nutrients
introduced.
Proper system design includes a laboratory pilot study
to discern the biodcgradability of contamination under
site conditions and an engineering evaluation to study
effects of implementation upon site soils and the aquifer
itself. Nutrients (and oxygen where applicable) can be
added to the subsurface both passively and in dynamic
systems requiring recovery and infiltration of water.
In situ bioreclamation rapidly attacks the residual
contaminant: napped within site soils, as well asdissolved
within the aqueous phase. Because of this simultaneous
reduction of both the symptom, as well as the source of
contamination, in situ bioreclamation has proven to be a
very cost-effective remediation alternative.
References
Atlas, R.M. 198!. Microbial degradation of petroleum
hydrocarbons: An environmental perspective.
Microbiat Rev., v. 45, no. I, pp. 180-209.
Brown, R.A., G.E. Hoag, and R.D. Morris. 1987. The
Remediation Game: Pump, Dig, or Treat. Water Pol-
lution Control Federation Conference. Oct., Phila-
delphia, Pennsylvania.
Dhalback, B., M. Hermansson. S. Kjelleberg, and B.
Norkrans. 1981. The hydrophobicity of bacteria—an
important factor in their initial adhesion at the air-water
interface. Archives of Microbiology,?. 128, pp. 267-270.
Davis, J.B. 1972. The migration of petroleum products in
soil ground water: Principles of countermeasures.
American Petroleum Institute Publication No. 4149,
Washington. B.C.
Engineering-Science Inc. 1986. Cost model for selected
lechnologiesfor removal of gasoline components from
groundwater. American Petroleum Institute Publica-
tion No. 4422, Washington, D.C.
Gerson, D.F. 1985. The biophysics of microbial growth
on hydrocarbons: Insoluble substrates. Int. Kore-
sources J.. v. 1, pp. 39-53.
Gerson, D.F. and J.E. Zajic. 1979. Bitumen extraction
from tar sands with microbial surfactants. Canada
-Venezuela Oil Sands Symposium, pp. 195-199.
Groundwatcr Technology Inc. 1983. Estimation of lost
product distribution in the subsurface. Confidential
Client. GT1, Chadds Ford, Pennsylvania.
Utchfield, J.H. and L.C. Clark. 1973. Bacterial activities
in ground waters containing petroleum products.
American Petroleum Institute Publication No. 4211,
Washington, D.C.
McKee, J.E., F.B. Uvcrty. and R.N. Henel. 1972. Gaso-
line in groundwater. / Water PolL Corn. Fed., v. 44,
pp. 293-302.
Raymond. R.C, V.W. Jamison, J.O. Hudson, R.E. Mit-
chell, and V.E. Farmer. 1978. Field application of
subsurface biodegradation of gasoline in a sand for-
mation. Final report submitted to American Petroleum
Institute, p. 137.
Rosenberg, M., E. Rosenberg, and D. Gutnick. 1980.
Bacterial adherence to hydrocarbons. R.C.W. Berkeley,
J.M. Lynch, J. Mefflng, P.R. Rutter, and B. Vincent
(ed.) Microbial Adhesion to Surfaces. Ellis Norwood,
Chinchester, England, pp. 541-542.
Suzuki, T, K. Tanaka, I. Matsubara, and S. Kinoshila.
1969. Trehalose lipid and alpha-branched beu-hydroxy
fatly acids formed by bacteria grown on n-alkanes.
Agric. Kol Own., .v. 33, no. II, pp. 1619-1626.
Texas Research Institute Inc. 1982. Feasibility Studieson
the Vse of Hydrogen Peroxide to Enhance Microbial
Degradation of Gasoline. American Petroleum Insti-
tute, Washington, D.C.
Texas Research Institute Inc. 1983. Progress Report:
Bioslimulalion Study. February.
Velankar, S.K., S.M. Barren. C.W. Houston, and A.R.
Thompson. 1975. Microbial growth on hydrocarbons
- Some experimental results. Biotech. Geoeng., v. 17,
pp. 241-257.
Wilson, S.B. 1985. In situ biosurfactanl production: An
aid to the biodegradation of organic ground water
contaminants. In Proceedings of the Petroleum
Hydrocarbons and Organic Chemicals in Ground
Water— Prevention, Detection and Restoration.
National Water Well Association/ American Petroleum
Institute. Houston, Texas, pp. 436444.
Wyndham, R.C. and J.W. Costerlon. 1981. In vitro
microbial degradation of bituminous hydrocarbons
and in situ colonization of bitumen surface within the
Athabasca oil sands deposit Appl Diviron. Microbial,
v. 41, pp. 791-800.
Yaniga, P.M. 1982. Alternatives in the decontamination
of hydrocarbon contaminated aquifers. In Proceedings
of the Second National Symposium on Aquifer Resto-
ration and Ground Water Monitoring. National Water
Well Association, Dublin, Ohio, pp. 47-57.
Yaniga, P.M. and W. Smith. 1985. Aquifer restoration: In
situ treatment and removal of organic and inorganic
compounds. Groundwaler Contamination and Rec-
lamation. American Water Resources Association, pp.
149-165.
Zobell, C.E. 1946. Action of microorganisms on hydro-
carbons. Bacterial. Rev., v. 10, pp. 1-49.
Biographical Sketches
ScottB. WilsonisdireaorofBiorecuunarionServuxs
for Groundwater Technohfy Inc. (40X1 FOut Ln., Sidle
B, Concord, CA H520). He ct responsible for overseeing
the design and operation of biorecumatum systems Jar
the decontamination of water and sots aid directs the
operations at the Groundwater Technology Inc. Biorec-
tamation Laboratories baited in Chadds Ford, Pennsyl-
degree Jrom the University ofSanDtego and his master's
degree from the University of Tens at El Paso where he
studied app&ed microbiology endgeohgy,
Kidard A. Bivwn Is tecrmlcalinanater for Ground-
water Technology Inc.', BioredamaOoa Services (220
NcfWOodParkS.,Norwood,MA 121X21 at the northeast
and mid-Atlantic United Stales. During the last 10 yean
he has been Involved In the development of methods
tamghydrogenperoxidetftsourceojoxygenmground
water and sous to stimulate indigenous bacterial detjra-
dation of petroleum hydrocarbons. He holds aB^t.in
chemistry from Harvard University and a PhJ). in Inor-
ganic chemistry front Cornet University.
Winter 1989 CWMR
179
microorganisms to proliferate; these materials Bust already be
present or be supplied in the proper form and ratios to the
requisite microorganisms. Extremes or temperature, pH, salinity,
and contaminant concentrations can also markedly influence the
rates of microbial growth and substrate utilization. The nature
of the limiting environmental factor(s) will often help dictate the
strategy in applying biotechnology to hazardous waste treatment.
In most cases the organic pollutants themselves are able to
supply the carbon and energy required to support heterotrophic
microbial growth. However, the introduction of carbonaceous
materials to soils and groundwater aquifers can cause an imbalance
in the natural biodegradation processes, limiting the microbial
transformation of the organic pollutant. For example, when labile
carbon is introduced to an aerobic aquifer, the microorganisms
consume oxygen along with the carbon substrate. An anaerobic
aquifer can be expected whenever the rate of aerobic respiration
exceeds the rate of oxygen input to the site. To sustain aerobic
microbial growth, oxygen, therefore, must be supplied to the
subsurface microorganisms.
The importance of oxygen supply to in-situ biodegradation was
well documented recently in a study of a wood treating site in
Conroe, Texas (S). A downgradient portion of the contaminant plume
was characterized by low levels of organic pollutants and dissolved
oxygen, while inorganic contaminants (i.e., chloride), which were
associated with the organic wastes, remained at elevated
concentrations. The authors suggested that oxygen was consumed
during the aerobic metabolism of the organic contaminants by the
indigenous micro-organisms. Hydrocarbons persisted in areas of the
plume where oxygen levels were insufficient to support aerobic
biological activity.
Artificially increasing the oxygenation of subsurface
environments will dramatically increase the growth of heterotrophic
bacteria. In a study of petroleum hydrocarbon degradation, sand
columns were used to determine the effect of oxygen supply on
bacterial growth and degradation of gasoline.
Appendix II
OXYGEN SOURCES FOR BIOTECHNOLOOICAL APPLICATIONS
Richard A Brown. Groundwater Technology. Inc.
It is well recognized that microorganisms play prominent roles
in the transformation and degradation of organic chemicals in
virtually every major habitat except the atmosphere. Microbial
communities in nature exhibit a truly impressive biochemical
versatility in the number and kinds of synthetic organic compounds
that they are able to metabolize (1,2).
Virtually the only natural transformation of polluting
chemicals that can result in complete mineralization occurs via
microbial metabolism. However, there are limits to the metabolic
versatility of microorganisms. Many xenobiotic substrates are
transformed so slowly that they cause some degradation of
environmental quality. This resistance to biodegradation, though,
is not a feature that is strictly associated with exotic chemical
compounds. Decomposition is a function of the a) structure of the
particular contaminant, b) the existing environmental conditions,
and c) the physiology of the requisite microorganisms (3,4). Of
these, the environmental limitations are the easiest to rectify.
In order to grow, microorganisms need a suitable physical and
chemical environment. Microorganisms, like all other forms of
life, are primarily composed of C, H, O, N, P, S, although a
variety of other elements are also found in trace amounts. These
substances are required to varying degrees in order for
Several columns were prepared under identical conditions using
SO mL of wet sand sieved to 40-60 mesh. Fifty Billiliters of
gasoline were added to each column and allowed to drain through.
An average of 4.3 mL of gas was retained. The columns were then
washed with 2 liters of nutrients made up in groundwater.
Different levels of oxygen were supplied to the columns by using
air, oxygen or hydrogen peroxide dissolved in groundwater. The
columns were treated for two weeks. At the completion of the
experiments the columns were drained, and analyzed for gasoline
content, total organic carbon (TOC), total bacteria and gasoline
utilizing bacteria.
Bacterial counts in the interior of the column showed a very
strong dependence on the oxygen level:
DEPENDENCE OP BACTERIAL GROWTH ON AVAILABLE OXYGEN
Bacteria, Colony Forming Units (CFU)
Available Oxygen,
DDm 1 Ave. )
8
40
112
200
Correlation w D.O.
Ratio of counts *
Heterotrophic Bacteria
(x lo'i
.05
5.S
75
207
.979
4 X 105
/ Gram Dry Soil
Gasoline Utilizing
Bacteria fx 10*1
.0001
.7
27
31
.933
3 X 10!
CONTAMINATED SOIL TREATMENT 231
-------
As can be seen from the data, the bacterial count* Increased
dramatically with increasing available oxygen. Gaaollne utilising
bacterial ara even »or« sensitive to oxygen level* than are general
heterotrophic Bacteria.
Available
Oxygen
pom fAve.l
a
40
112
200
Correlation w D.O. .994 -.tJ
Ratio g«« degraded » J.27 .19
,u pp» D.o. : • pp» D.O.
Based on Average of J.u g originally present.
Casollns Bio-
degraded
oraaa i1
.111 12.9
.901 16.9
.77) 35.4
1.272 42.4
Caaoline
Out
grama
.71
.77
.99
.4*
riushed
t'
JJ.4
19. •
19.4
16.1
iflllf bftTlfc-
Total Gasoline
Removed
or«»« I'
1.091 36. »
1.271 42.4
1.111 49. •
1.7*2 91.7
day an air epargar provide* per well for different hydraulic
conductivities end gradient*. The table assumes a 10 foot
saturated thickness and that the lateral influence of the well 1*
1 ft.
u
Hydraulic Conductivity
aala / Day / Ł1 fair!
10* (gravel) 4
10* (medium sand) .0*
10 ' (eilt) 6X10 '
•ruaiM.
(high)
(oxygen)
30
SIlMLt WILL
Hydrsulie Gradient
ft/ft
(•edium)
(elr) (oxygen)
.* ).
4x10'" IxlO''
(lOV)
-Ml
(air) (oxygen)
.04 .1
sxio"* mo'1
A* can be seen air sparging la a limited source of oxygen.
Sparging pure oxygen instead of air will increase the pound* per
day by a factor of five so that the maximum on the matrix would be
10 lb*. oxygen per day Inatead of 5.
Several things should be noted from this data. Flrat, the
•or* oxygen that was supplied, the more gasoline that vas
biodegraded. Second, the rate of biodegradatlon under highly
oxygenated conditions was greater than the rate of physical
reiaoval/dissolution.
These cand column studies demonstrate that bacterial growth
and metabolism are very dependent on oxygenetion. As a result, an
important part of the biological treatment of hazardous waste 1*
oxygen supply.
There are basically two method* of oxygen supply - physical
•nd chemical. Physical supply Involves forcing air and/or pure
oxygen into the contaminated matrix. Chemical oxygen supply
involve* the addition of *ub*tance* which can be converted to
oxygen, *uch a* hydrogen peroxide (6); or eubatance* which can act
•* terminal election acceptor* directly such as nitrate (7,1). All
of these methods have been used In treating contaminated coil* and
aquifers.
The choice of an oxygenation method depend* on aeveral
factors. Basically, on* want* to achieve maximum efficiency in
oxygenation. Too little oxygen supply relative to the amount of
contamination result* in much longer remediation time*. Too much
oxygen relative to the amount of contamination being treated can
result in elevated remedial comte. The principal is to balance
oxygen *upply with oxygen demand. The factor* that must be
considered In achieving this demand are:
oxygen mass transfer, pound* per unit time, supplied
by each method
• contaminant load and location
ease of transport/utilization
First, considering oxygen man transfer, it 1* easy to
calculate the amount of oxygen supplied by the different methods.
The more oxygen supplied per unit tine, the greater the potential
level of bloreclamation.
Air sparging, one of the simpler techniques, provides oxygen
by diffusing air/oxygen Into a well bore. This is accomplished by
using a porous stone, sclntured metal or fitted glaas diffuser.
The water In the well bore ii saturated with oxygen and diffuaea
out into the formation. The amount of oxygen aupplled 1* a
function, therefore, of the rat* of water flow by the well bore.
This, In turn, Is a function of the hydraulic conductivity, the
gradient and the surface area of the formation effected by the well
bore. The following matrix calculate* the pound* of oxygen per
A second (ystee 1* to pump air/oxygen saturated water into a
contaminated equifer. The pound* per day of oxygen (applied is •
function of injection rate:
Inleetlofi Bate.
1
10
100
•imau nu.
aereted water
IB m» P.O.
.12
1.2
12.0
oxygenated water
SO nee D.O.
Air venting *y*tem* ar* an efficient means of eupplylng oxygen
through unsaturated contaminated colls. Thi* technique i* used in
treating vadoae zone contamination or in treating excavated coll
pile*. Air can be added by either injection or by withdrawal. In
vadose tone treatment, the common method la vacuum withdrawal.
This method ha a the added advantage of physically removing volatile
contaminant* in addition to *upplylng oxygen. The amount of oxygen
•upplied 1* a almpl* function of the air flow rate*. The following
table uses a 201 oxygen content for air to calculate elr cupply:
tQPMM »»*. DAT orrotn »O»»LI»P
^y TBKT BTBTtM t OMBATOmATEP aOlm
air flow rate
1
5
10
20
50
100
oxygen *upply
Ibs / Pay
21.
116.
211.
466.
1.166.
2.112.
Finally, there are two chemical carrier aystems Hydrogen
peroxide and nitrate. While both of thea* material* are highly
•oluble, their common use rate i* about 1000 pom (.It). The number
of oxygen equivalents eupplied ia dependent on the chemistry
involved. Hydrogen peroxide 1* converted through decompoaitlon to
oxygen:
»,o, •» H,O + i/a o,
Each part of hydrogen peroxide aupplie* .47 part* of oxygen.
Nitrate la, on the other hand, directly utilised a* a terminal
election acceptor. It* oxygen equivalents can be calculated by
comparing the amount of nitrate required to oxidise a aubstrat*
versus the amount of oxygen. Take, for example, the oxidation of
Hethanol:
Oxygeni CM,OH « 1/2 O, * CO, «. JH,O
Nitrate: NO,' 4 1.0«CH,OH « H* —-» .06»C>NrNO, » .47N, *
.7600, * 2.44H,O
232 CONTAMINATED SOIL TREATMENT
-------
Based on these above equations, one part of nitrate is equivalent
to .84 parts of oxygen.
The oxygen equivalents supplied by these two chemical carriers
is a simple function of injection rate.
tOWDS mt DAY OXYOEK EQUIVALENTS BPPFLrED
BV CHEMICAL CARTERS. BISOLE WELL
t 1000 DDB
Injection rate
gpm
1.0
5.0
10.0
20.0
50.0
H,02
eouiv i
5.6
26.0
56.0
112.0
260.0
1.84 eouiv o:/part NO3)
10.0
50.0
100.0
200.0
500.0
The second factor in considering an oxygen source is the
contaninant load and location. Contaminant location is important
in that vent systems require unsaturated environments and will,
therefore, be excluded in treating contaminants below the water
table. Contaminant load, on the other hand. Impacts all means of
oxygen supply, in that it determines oxygen demand. What drives
contaminant load is the phase distribution.
Petroleum hydrocarbons exist in the subsurface as three
condensed phases: mobile free product (phase separated),
residually saturated soil (adsorbed phase), and contaminated ground
water (dissolved phase). The distribution of hydrocarbons into
these different phases, while a result of dynamic transport, is
ultimately a function of their physical and chemical properties,
and the hydrogeological and geochemical characteristics of the
formation. One must examine the phase distribution by two means:
the areal extent of contamination or the volume of the subsurface
impacted by a phase and the severity of contamination or the amount
of the contaminant within a phase, measured as either total weight
or concentration. The following table gives the phase distribution
for a gasoline spill in sand and gravel:
PHABE DISTRIBUTION OT OA80LINI ID BAM) AMP ORATEt
Ext*nt of
Contamination
i)
iter)
Volume,
cu. yd.
780
2,670
11,120
% of
Total
5.3
18.3
76.3
Ifej.
126,800'
11,500
190
Matt
Distribution
Cone.
DDIB
2,000
15
\ of
Total
90.9
8.2
0.3
Free phase1
Adsorbed (soil)
Dissolved (water)
1 Actual value recovered from site
There are several generalizations that can be made from the
above data concerning the distribution of petroleum hydrocarbons
between the different phases. First, groundwater flow is the
primary long term mechanism for spread of the contamination once
the free product layer has achieved flow equilibrium. Thus, the
areal extent of groundwater contamination is typically greater than
that for other phases. However, the amount of material in the
groundwater is small compared to that retained in the soil matrix,
less than St. The residually saturated soil, if untreated, is a
continuing source of groundwater contamination.
In looking at the contaminant load, the presence of and the
distribution between the different phases is an important factor.
The following table gives the pounds per cubic yard of aquifer for
dissolved and adsorbed phase contamination. The calculation
assumes a porosity of 30» and a dry soil bulk density of
27001b/yd5. The soil levels are generally two orders of magnitude
higher than dissolved levels.
Dissolved
Phase «
1 ppm
10 ppm
100 ppm
DISSOLVED AND
i / yd1 or AO
I/yd5
5X10''
5x10''
5x10'*
ADSORBED PRASE
DI»R HATERIAL
Adsorbed Phas
Phase t
100 ppm
1,000 ppm
10,000 ppm
I/yd5
.27
2.7
27.0
From this data it is obvious that contaminated soil drives the
contaminant load. The more the volume of contaminated soil and the
higher the level of contamination, the greater the contaminant
load. One cubic yard of soil contaminated at only 100 ppm contains
as much contaminant as 5.4 yd3 of contaminated aquifer material
(dissolved phase).
The third factor in considering an oxygen source is the ease
of transport and utilization. This involves the means of
application, the maintenance of system, and the rate/degree of
utilization.
An air sparger system uses a small compressor able to deliver
-1CFM per well. The sparger itself is either a porous stone, a
•cintured metal diffuser, or a fritted glass diffuser. Power
consumption is minimal. The transport of the aerated water is
limited by the rate of groundwater flow. The most significant
operating cost is an air sparger system is maintenance of the
compressor and of the diffuser and well screen. Biofouling or
inorganic fouling of the diffuser and well screen can be
significant and well therefore require a high degree of
maintenance. Bacterial utilization of the dissolved oxygen is very
high.
Injection of aerated/oxygenated water is a relatively simple
system. The simplest approach is to use an air stripper to aerate
the water. Often in treating a contaminated aquifer, groundwater
is recovered and air-stripped to achieve hydraulic control of the
contaminant plume. Reinjectlon of the stripped groundwater, can
therefore, be accomplished for relatively low cost. The main cost
of operation is controlling fouling of injection system. Transport
of the oxygenated water is dependent on the geology (hydraulic
conductivity). Bacterial utilization of the injected dissolved
oxygen is very good.
Venting systems, while limited to unsaturated soils, are very
efficient means of oxygen supply. The primary capital cost is the
vacuum pump(s) needed to drive the system. Maintenance of the
pumps is fairly simple and power consumption is minimal. The
efficiency of the vent system is enhanced by volatile removal. The
largest potential cost with o vent system is treatment of the vapor
discharge. This can be accomplished by using disposable carbon,
regenerable carbon or catalytic oxidation. Regenerable carbon and
catalytic oxidation are capital systems.
A hydrogen peroxide system is generally a low capital, easy
to maintain system. It does entail a fairly high OtM cost due to
the chemical cost of the hydrogen peroxide. The cost of hydrogen
peroxide is dependent on the volume used. Small quantities cost
more per pound than do large quantities. On a per pound of oxygen
basis, the cost will range from $1.50 to $2.50. The biggest cost
factor involved with hydrogen peroxide is how quickly it
decomposes. There are two mechanisms of decomposition - Biological
and metal catalysis. Ideally, one would like minimal metal
catalyzed decomposition. In some soils, however, that contain high
levels of iron or manganese, metal catalyzed decomposition can be
severe. In such cases the solubility of oxygen is rapidly exceeded
and the water phase degassed, loosing available oxygen and
drastically reducing the efficiency of the system.
Finally, nitrate systems are a potential electron acceptor
alternative. Operationally, these systems have not been proven.
Capital costs for a nitrate system would be fairly low consisting,
as with peroxide, of a supply tank and metering pump. Chemical
costs for nitrate are $.60 - .70 / Ib oxygen equivalent. The issue
with nitrate, however, is not the cost or ease of addition, but
instead the biochemistry of utilization and the regulatory Issues.
In a recent test of nitrate utilization, it was found that even
with an extremely labile substrate such as sucrose, there was a
significant lag phase in the utilization of the nitrate when oxygen
CONTAMINATED SOIL TREATMENT 233
-------
was also available at low level*. It would appear that nitrate
utilitation requires low oxygen condition!. If the biochemistry
of nitrate 1* complicated, the regulatory lasuei becone
significant. Nitrate level* in ground water are regulated at 10
ppm. If nitrate ia not rapidly utillied, injection would have to
be tightly controlled and nay be precluded.
To put the above costs and analyses into perspective, one can
compare the operation of the different syitems for a ssmpls
gasoline problen. The site characteristics are ai follows:
Area of contamination
soil - 100 x 50 c 5ft (Jft above; 2ft below water)
groundwater 250 x 70 x Jft
Contaminant concentrations
soil - 1200 ppm
groundwater - II ppm
Aquifer characteristic*
hydraulic conductivity S x lo' bal/day/ft'
hydraulic gradient .01 ft/ft
pumping rate/well 35 gpm
saturated thickness 25 ft.
DTK 15 ft.
Contaminant distribution
•oil
cu
Total
2500 lb.(1500 Ib above/1000 Ib below
10.0 Ib.
25)0 Ib.
The configuration of the systems would vary.
out* vould be as follows:
The basic lay-
Air Sparging: Sparging would be through 15 wells spaced
through-out the plume. There would be no recovery well*.
Mater In»peetlon: The systen would consist of 2 recovery
wells pumped at IS gpm each. Each well would have a 25 ft.
cross-gradient capture radius. The 70 gpm water would be
• ir stripped through a 2 ft. X II ft. air stripper and
rainjected through two upgradient galleries.
Vent Syst«»: Venting would be through 4 welle placed on 40
ft. center*. Each well would be pumped at 40-45 CFX. The
syste* would require • single 2 H.P. high vacuusi blower.
Vapor discharge would be treeted through a catalytic
oxidizer.
Peroxide System: Hydraulic control is maintained by 2
down gradient recovery wells each operated at )5 gpm.
Of the 70 gp* recovered, 15 gpn would be reinjected
through a gallery and 2 upgradient injection welle.
Peroxide would be added et 1000 pps. Excess water
would be >ir stripped and sewered.
nitrate System: Because of the concern with off site
•igratlon of nitrate, there would be 4 downgredlent
recovery wells. Two of the welle, on the leading edge of
the plume, would be operated at 15 gpm each. This water
would be amended with nitrate (1000 ppm) and reinjected
through 4 upgradient well* and an Injection gallery. The
two additional recovery well* would be> pieced 50 ft.
downgradlent of the plum* to create • barrier to migration
of nitrate. Th«y would be opereted et 30-40- gym.
Using this dsta, the capital and operating costs for tich
system csn be calculated. The following table give* a comparison
of the different systems.
As can be seen, there is s wide verlance in both cost
effectiveness and In treatment effectiveness. In terms of cost
performance, the order ie:
Venting > > peroxide > nitrate > air eparger > water Injection
In order of treatment effectiveness the order is:
Peroxide - nltrste > water Injection > venting > air sparging
While venting Is • very cost effective method it is limited to
treating the vadose tone. Consequently, it'* treatment
effectiveness Is limited.
This sbove analysis it given for s situation with extensive
contamination. If the degree of contaminants is changed so that
the eoil contamination is minimal, the enalyses would change.
Assuming that there is no soil contamination above tbe water table
and that the eoil levels are <100 ppm, the performance of the
different systems would be as follows, sll other fsctors remaining
constent:
cosT/rmosjouict, unr osaui or COVTVMIHITIOI
System
lbS/d*y
Tlm« of Treatment
S Ib
Air sparging
Meter Injection
Venting
Peroxide
Nitrate
Rot Applicable
110
211
110
DO
1*0
240
4).5J
l». 55
When the degree of contamination 1* less, simpler systems such as
air sparging become more cost effective. When the contamination
is only the dissolved phase, an air sparger system is ths best
choice. The following table summeritee the best choices for
different contemination eituetions.
The choice of an oxygen aupply la dependent on the contaminant
load, the maas transfer and the ease of transport/utilisation.
Depending on what the degree of contamination is, different system*
will be most effective.
ureiuiictt
1) Alexander, H., I'll. Biodearadation of Chemicals ef
tnvironmentsl Concsrn. SCIENCE, 211:1)2-1)1.
2) Xobeyaaki, H. and B.E. Rittmann, 1*12. Mieroblal Removal
of Haiardous Organic Compounds. ENVIRON SCI TECHNOL, 1*:
170a-ll)a.
Co»t/Performu»ivce Co«peri»on for Various Oxytjon Symtesia
Hiqh Dourer of Contamination
4. Coeta «
Syete* Capital Operation Maintenance
U»/D«y
Oxygen
Performance ---—___ __.
% Site Utilization Tlsw of S/U, orraen
Treated Efficiency % Treatment Ua*>d
Air Sparging
Mater Injection
Venting Syatesi
Peroxide Syate*
Nitrate Syatesi
$39,000
$77,000
$••,500
$«00/s>onth
$1200/Bonth
flSOO/aonth
••0,000 $10.0OO/oonth
$120,000 $«900/m«nttt
SUOO/mxjnth
SlOOO/month
11000/aonth
$lSOO/m>onth
»1000/SK>nth
6
•
4000
190
211
41
7S
«0
10O
100
70
SO
9
IS
12.S
•SI daya
1580 daya
132 daya
330 day*
)3S daya
S2».«2
$ 3.«2
$22.0*
234 CONTAMINATCD SOIL TKI.ATMHNT
-------
Prioritization of Systems for Different Contaminant Situations
(1-Best)
Types and location of Contamination
System
Dissolved Only
Soil Above and Below
Dissolved and Water Table and
Soil Below Water Dissolved
Soil Above
Water Table
Air Sparger 1 2
Water Injection 3 3
Vent System
Peroxide 2 1
Nitrate (Not Recommended at present)
3) Alexander, M., 1965. Blodeoradation! Problems of Molecular
Recalcitrance and Hicrobial Fallibility. ADV APPL MICROBIOL,
7:35-80.
4) Alexander, M., 1973. Nonbiodeoradable and Other Recalcitrant
Molecules. BIOTECH BIOENGINEER, 15:611-647.
5) Wilson, J.T., J.F. McNabb, J.w. Cochran, T.H. Wang, M.B.
Tonson and P.B. Bedient, 1985. Influence of Hicrobial
Adaption on the Fate of Organic Pollutants in Ground Water.
ENV TOXICOL CHEM, 4:743-750.
6) Brown, R.A., R.D. Norris and R.L. Raymond, 1984. Oxygen
Transport in Contaminated Aquifers. Proceedings of the
NWWA/API Conference on Petroleum Hydrocarbons and Organic
Cheaicals in Ground Hater-Prevention, Detection and
Restoration. Nov. 5-7, 1984, Houston, TX.
7) 'Use of Biotechnics in Water Treataent: Feasibility and
Performance of Biological Treatment of Nitrates," A.
Leprince, V. Richard, Aqua Sci. Tech. Rev., 1982 (5), pp.
455-62.
«) "Nitrogen Removal in a Subsurface Disposal Systea," A.
Andreoli, R. Reynolds, M. Bartllucci, R. Forgione, Water
Science Tech. 1981, 13 (2), pp. 967-76.
3
2
3
1
2
System Layout
Air Sparging
Grountfweter Plum*
Sparger Wells
Water Infection
Air Stripper
/
/
\
^ \
Injection
Gellertn
N
\J
Contaminated Soil
\
-O
Wells
)
S
System Layout
Venting
Blower Catalytic OxMtzor
Groundwater Plum*
/^ u-u-*
Is
K o-1 w ^
OA
<"v
V
Contaminated Soil
Peroxide
HzOj NutitonlB
Groundwater Plume
Injection/ Gallery
\ Contaminated Soil
h-O
System Layout
Nitrate
Groundwiter Rume
Treatment
CONTAMINATED SOIL TREATMENT 235
-------
Appendix III
CROUNOWATtR CONTAMINATION AND MCLAMAT1ON
AMtUCAM VATl* UiOUtCU A1MCUT1ON
AQUIFER RESTORATION IN SITU TREATMENT AND REMOVAL
OF ORGANIC AND INORGANIC COMPOUNDS
II IV (WlMtoMIt
M* tMiAb km earn* "* |
,
riMa ia»diir4 t«ofp"K yirumitn li dlttuwtf
M . ro*. —• -wfc- a* hi-*** TW -~H- TV kiy M * imiMM »4 •bn*m«i of rJili M
^ju-^v* *. *.« **. r- y^fcn.« — F- ^^^ ^ ^ |wbe|(M. rf M w-lu ho,,^,,,,
*Mtk arflHm «UMM| MlM ky4foc«'tv>«
*U mUMam vt «Mrttxti tnrf acvfr* M (V |
., _„. •*» *• » "f *"»•" "1*« rf l«)**|«« ^«fo«id« a
MHMM.QAM- CMwWnrckialali
x D •(•k^tlM *
rcrr njua •«•*•( im 11 mi IL > mrf^i»
ctMivtriiuiJoa
IWTWODUCTfOH
law ^murf aon(« link Tht tr«a of At kw b »Mdf rl
Ip^rtUMIltr » tO T f«*1 -f nd^OM. V**y A »»t*.
.
kM W- bwnM Ui MMW. Tfc.
iriM H • fa^iji «f 30 w 2) fra Wkm |n4i.
•fenM
•nvt
TV
Mrf VNWMIM in"- V* • *
to HIM il«4*n4*tto* »rO«|Mb
DISSOLVED HYDROCARBON CONTAMINATION -^
RECOVERY WELL
OBSERVATION WELL
DOMESTIC WELL
htOIUM A5SUUEKT
TV vnl «IUM wrf «•*!!»•) W ifc* ***** «
DISSOLVED HYDROCARBON CONTAMINATION
0 RECOVERY WELL
• OBSERVATION WELL
A DOMESTIC WELL
t «• t4 rwr
236 CONTAMINATED SOIL TREATMENT
-------
Acjulfci Ktnoniloi VI. Acukmid la Sin, Itabp.lntu orOitulc Cor.umhu.il
Aqul/cr RMIOnllon Vb Acceknlcd In Sin BlodeinditioB ofOfink Coounlnull
DISSOLVED HYDROCARBON CONTAMINATION
• RECOVERY WELL
• OBSERVATION WHLL
A DOMESTIC WELL
2 FEB 84
A. IMCOII Map MI/IIWIC Connnlnllon (ppm). November 1911.
RECOVERY WELL
OBSERVATION WELL
DOMESTIC WELL
AIR SPARGING WELL
AJR COMPRESSOR
INFILTRATION GALLERY
AIR STRIPPING TOWER
AREA OF CONTAMINATION
AIR. LINE
WATER DISCHARGE UNE
m
Ik
Figure 4. Schema lie uf BiorccUmatlon Syifern.
RECOVERY WELL
OBSERVATION WELL
DOMESTIC WELL
Ftjure 3B. Itocoo Mip (ton Concentration (ppm). Novcmbei 1911.
Further analysti, beyond organic scam, showed i need lo ad-
drea Inorganic, u well as. organic compounds. The analytic*]
iciulli dictated the development of i physko/cheroloJ treat-
ment process thai relied on neutraflutfon. carbon adiorpcfcw.
ton exchange, ind biologic control (Fl|ur« I). This comblni-
tlon of treatment uepi produced tejiheMcaDy uiablc wawr.
which wti Tree of any residual dissolved orpntc compound*.
INITIATION OF A BATE MENT/ AQUIFER RESTORATION
In Implementing the designed program, i practical approach
was kept In mind (hat look into account (he nature of the
problem, the »urce. the configuration of (he plume, the na-
ture of the groundwaier system, and the character of the
community. The program development wu carried out In •
logical sequence which Included:
* Development. Installation and shakedown lest ing of the
physico-chemical water treatment systems on domestic web.
• Excavarlon and disposal of highly contaminated loD tn
the lank pit area.
• Conversion of the lank pii. via backfill with crushed
stone, to an infiltration gallery.
• Construction of a pumping well located In the center of
the plume to control the water (able and movement of con-
taminated groundwater.
• Pump test of I he central weD to aOow calculations of
the expected radiui of influence, to assess the well'i capability
to control the migration of the plume.
• Construction and erection of an air stripper Tor viola-
Efle organic removal.
• Development of nutrient mix ratios for addition to the
groundwatet synem to accelerate hydrocarbon-utilizing
bacteria Tor reduction of the fugitive organic*.
• Development of mechanical means of air supply and air
sparging to deliver oxygen into the groundwater lyslem. Pnt-
rtoudy existing observation weDi were used u air iparging and
nutrient addition pointl,
• Development and construction of i nutrient mix link la
the area of the Infiltration gallery for batch feed of nutrlend
(o (he contaminated tank pit area.
• Shakedown letting of the lyrTcm to ensure operational
efficiency In the control of the organk plume.
The system, u designed, was to initiate pumping at the
central wed. Inducing water In the plume lo flow rajUlly In-
ward from the periphery. The recovered contaminated water
wai then passed through an air stripping tower where volatile
organic! were removed and oxygen was added. Nutrients wen
(hen added to (he hydrocarbon -free/oxygen -rich water, which
w» (hen passed through the con lamina ted softs and ground-
water system thus accelerating (he In-dtu reduction of organic
compounds via the increased numben of hydro carbon-utilizing
bacteria. The control" of the ipread of nutrients/oxygen and
biologic community was maintained by the central pumping
well, which had redirected groundwater movement to that
point. The treatment fw the organici was enhanced via the
addition of oxygen and nutrients on (he periphery of the
plum*. Thb water WH then pulled bRk through the con-
taminated zone to the central pumping wtfl.
RESULTS OF THE ABATEMENT PROGRAM
The result] of the aquifer restoration program were quite
good. The phytico/ekanictl treatment ytttm for the do-
mestic wetb furtcttooed wtfl, producing i reuabty unble water
tupply (Flfuret St tad ft). Ai the tott) aquifer clean-up pro
pun moved forward, decreased frequency of Ueitmenl medii
exchange wu required, diui ittesting lo overall conUmfntnt
reduction. Tht centrd pumpbtf uvff con (lined and eoniroDed
the pi tune configuration Dee in ta-dtu treatment vesxl (Fig-
ure 6). The tlf itripphf rower, subsequent to ihakedowD
letting, performed u dedpwd with greater than 98 lo99 per-
cent efficiency for removal of volatile organic* (Figure 7). Tht
designed infiltration gaOoy. located In the former tank pii.
proved functional In accepting the 30000 to 35.COO pOoni
per day of treated oxygen and nutrient rich water. Heavy
spring rain and recharge caused some concern regarding over
topping of (he giDcry, which, however, did not occur.
The air iporfint Vtiem, consisting of mechanical air com-
pressors, air Unei, and down weD diffusera, proved lo be ef-
fective to partially effective In delivering needed oxygen lo
peripheral areas of the plume ouuide the infiltration gallery
Major limitation focused on the maximum quanlty of oxygen
that could be Induced Into the groundwiur system (10 ppm}
at the sparging point and the fouling/plugging of the sparging
points by the development of thick biologic growths. Theu
two items precluded optimum oxygen transfer to the fractured
bedrock synera and required frequent mechanical cleaning.
Despite this non-optimum condition are/off efficiencies of
dean-up over the first eleven months showed a general SO to
85 percent reduction in organic contaminanti. Several well]
proved 10 be absent of any organic contaminants a( this point
(Figure 8 and 9).
While pleased with (be overall results, the ipedTics of
oxygen transfer rates were EmJUng biologic community growth
and lengthening the project restoration tlmeframe: therefore.
a program to accelerate dus problem was developed. The pro.
gram involved a comprehensive approach lhat Included:
• laboratory research,
• fklduudret,
• further hydrogeologk and engineering assessment, and
• information/educational meetings and contact with re
presemarjvet of the community and regulatory agencies.
The results of (he appued efforti wu (he development of i
comprehensive approach to deliver Increased quantities of
oxygen to the grouncrwaier »ynem via the trickle feed and
dbasiodatlon of dilute quantities of hydrogen peroiide
Laboratory studies conducted jointly by Mr. Richard L Ray-
mond and FMC (Richard A. Brown) showed lhat lnducins
dilute concentrailoni of hydrogen peroxide did not kill de-
sired hydrocarbon utilizine. bacteria. On the contrary, thu
increased iheir numbers and (he rale of hydrocarbon re-
duction. Field studies by R. L. Raymond also showed tknilar
CONTAMINATED SOIL TREATMENT 237
-------
TOP VIEW
UnliiiUd walar to
ouuioa lap
Incoming
Notes:
1. Neutralizing ftIter;
preliminary sediment. iron,
and manganese removal.
). Portable ion exchange units
(in parallel); removal of
hardness and further iron and
minganese re-ioval.
I A 2*. taste and odor filter;
activated charcoal with high
surface area for removal of
taste and odor (removal of
hydrocarbons).
Fit in S. Schematic: Walcr Treatment Sytlmilo Improve Anlbelk Water Quality la Rnidencei ArTecttd by S«tfnc1 Il)r4roca>toa Cc«m«li>iltn«.
6. Ultra violet light;
bacteriological treatment
unit for call form treatment.
I
I
WELL V
KEY
l^lFLu
E • EFFLUENT
WELL V
WELL U
*
a
I
4
2
E
0
WELL a
WELL U
!*•! NOVtMKK
WELL B
WELL L
>I9«I NCVCMBER
WELL
rifud ]A- HtntiBf H CorKtr
!)•! NOCMBCfl
• lion fit u>d foil fhrl'clJ-Ow'r.kil T'lJImUL
WELL L
IHI NOVCUOt:*
rt|vif II lra
lltl NOVCMBCR
238 CONTAMINATED SOIL TRRATMI-.N I
-------
DISSOLVED HYDROCARBON CONTAMINATION
WELL 3
I 'U 4IU
1992
WELL 2
rfln
'Ł• 4 HAN
1382
WELL 9
n n_r-i
14 MOV 4 JAM zrit 4MJM
RECOVERY WELL
OBSERVATION WELL
DOMESTIC WELL
Retire 6. Wiler TaM* Gradient (data ufcen from Gfoundwaler Technolocr manitoriti| weHi, April 1,1901).
SAMPLES WERE RETRIEVED BUT NOT
ANALYZED DURUM IEPT TO DEC u».
UONTHLY 3AUPLINQ AND ANALYSIS PMO-
QRAU WILL CONTIWJC IN JAN IMS.
0 INFLUENT
•— EFFLUENT
(DATA FROU INFRA-RED PROCEDURE)
WELL 6
nrffl
1EFT I4MOV
1991 1992
WELL 0
n
•22:
1301 1902
DATE OF
1991
SAMPLING
aVa «S5«
Figuie I. TotaJ Hydrocarbon Conceniratioru for Core Homo.
Aquifer Reitoutlon VU Accelerated In Sllu Blode^idnlon of Orgwik Conumlninu
DISSOLVED HYDROCARBON CONTAMINATION
DATE OF SAMPLING
Ffeiir* 7. Total Hydrocarbon Conctnin.fciu for Air Stripping Towtr.
results. In Ihe dc»etopmeni of che handling, delivery, and ip-
pllaiwn technique i tot hydrogen p«foxlte ob«fvarion points. TT.e icsalu showed «i
acceptabk Increase, from 0.5 ppm dioolved oxygen to 8.0
ppm dissolved oxygen in a 24-hour period. The Increase In
diuoJred oxygen also nlmutaied in inertJK in microbiolo|k
•clivity and a decrease in hydrocarbon concca[rations.
The concentration of hydrogen peroxide u«d In I he on-
going program for enhanced bioreclaination ii 100 ppm. yield-
ing SO ppm of dissolved oxygen Tor uptake and utilization
by Ihe microbiologic community. Hydrogen peroxide is cur-
rently being added to the ground water system u [he she. both
at the inHltration gallery and former air tptijjni wells. An
added benefit of the hydrogen peroxide UK in the weDi h
that when introduced to the well bore at 100 ppm, it keep
(he well free of heavy biogrowth, thut allowing more equal
and Quicker iranimiuion of needed oxygen to the impacted
areas of the (roundwater lyitem. The moil receni results from
the Kite show overall hydrocarbon concentration level] to have
declined in the core area, with only five homeowner welli
stm ihowing degradation (Figure 10).
41 RECOVERY WELL
• OBSERVATION WELL
A DOMESTIC WELL
Figure 9. liocon M»p of Hydrocarbon Contamination (ppm) of Domenie WeDi, December 1. 1982.
CONTAMINATED SOIL TREATMENT 239
-------
SUMMARY
TV .pp[M4 Uchfliqutl of phrtfaJ/clwmial lifiimM of
domttk wtD wtiti combined «*lh frovitdwiin m»«ip*ib-
rto* ** i pwnplni wefl. ib arlppM* fw *oblOf orpftk ra>
«MmJ, u»4 tnh»nc*4 bkniinMililkm/btomUiniiion of iiowna*
•*(« lyiicira cqnumiitai^ by wpnk compoundi pt|M*.
•M ut §MI«U AitMV •)
DISSOLVED HYDROCARBON CONTAMINATION
Itll
r, Itll
«4 A«W«i
t «rf liufanl mt fUflMtf
IN. DTD Ft**!* VM«
Ii af dM IVU C»H»'"
_ itatMMM.
v*«to. r M« D o*«t^ iHi MM
WMM
*t n
•MM
wtLL
wtLL
A DOMESTIC WELL
240 CONTAMINATED S0IL TREATMENT
-------
Groundwater Extraction System Design for a
U.S. EPA Superfund Site
Donald M. Dwight
Metcalf & Eddy, Inc.
Wakefield, Massachusetts
ABSTRACT
Baird & McGuire, Inc. operated a chemical mixing and batching
facility in Holbrook, Massachusetts, producing household and indus-
trial products including floor waxes, wood preservatives and pesticides.
Widespread contamination of the bedrock and overburden aquifers by
over 200 chemicals including pesticides, volatiles, semi-volatiles and
metals has occurred due to improper handling and disposal of products
and waste. The town well field, located 1,500 ft from the site, was aban-
doned in 1982 because of volatile chemicals detected in the water. In
1982, the site was scored on the hazardous ranking system and cur-
rently is ranked high on the U.S. EPA NPL.
In 1988, Metcalf & Eddy, Inc. was contracted to design remedial
measures for the site including the design of a groundwater extraction
system, a recharge system and a groundwater treatment plant. Moni-
toring wells were installed and two aquifer tests were performed to obtain
design information. By utilizing a state-of-the-art electronic data
acquisition system developed by the Illinois State Water Survey, a high
density of accurate data characterizing aquifer behavior was collected.
The data ?/om the pumping tests were used to construct and calibrate
a three-dimensional groundwater flow model of the site. The model
was utilized to predict the effects of various extraction well schemes
on the flow regime and to locate wells strategically to result in quick
and efficient groundwater cleanup. The USGS Modular Groundwater
Flow Model (MOD3) was used in this study. Calibration of the model
involved replicating aquifer behavior exhibited during the pumping tests
and matching assumed steady-state head distributions. Model simula-
tions provided a better understanding of aquifer interconnections,
recharge areas and interactions with a river flowing through the site.
Ultimately, the model was used to determine the optimal well con-
figuration based on an evaluation of capture radii. The design of the
groundwater extraction system was based on modeling results indicating
pumping well location, pumping rates and well screen depth and
intervals.
INTRODUCTION
This paper describes the approach taken to the design of a ground-
water extraction system for a U.S. EPA Superfund Site in Holbrook,
Massachusetts. The system is part of a groundwater remediation scheme
involving groundwater extraction, treatment and reinfiltration. The
groundwater at this site is contaminated with over 200 chemicals
including pesticides, volatiles, semi-volatiles and metals. The objec-
tive was to design a groundwater extraction system to capture contami-
nated groundwater in three somewhat distinct aquifers and to promote
flushing of the aquifers.
The approach that was used to develop the data necessary for design
included three basic steps:
• Exploratory drilling and the installation of monitoring wells to gather
information about the potentiometric surface, the bedrock, the over-
burden geology and groundwater quality
• Pumping tests, utilizing a state-of-the-art electronic data acquisition
system, to characterize the flow characteristics of the aquifers, iden-
tify hydrologic or impermeable boundaries and determine the degree
of hydrologic connection between the aquifers
• Three-dimensional groundwater flow modeling to determine optimal
extraction well locations, screened intervals and pumping rates
BACKGROUND
For over 50 yr, Baird & McGuire, Inc. operated a chemical mixing
and batching facility producing household and industrial products
including floor waxes, wood preservatives and pesticides. In 1982, the
town well field, located only a few hundred feet away from the site,
was closed due to die detection of volatile chemicals in the water supply.
After site investigations, it became obvious that improper handling and
disposal of waste products had resulted in contamination of bedrock
and overburden aquifers.
The bedrock is a gabbro-diorite formation, the top 20 ft of which
are fractured and weathered. The overburden consists of two hydro-
geologically distinct units. The lower layer is glacial till consisting of
sand, silt, cobbles and boulders. The upper layer is stratified drift con-
sisting of fine to coarse sand with trace amounts of silt. Within the strati-
fied drift, two layers with distinct textural differences were identified.
The lower part of the stratified drift is predominantly coarse-grained
sand. The upper part consists mostly of silty fine sand.
An RI/FS performed by other engineering consultants indicated that
groundwater contamination exists predominantly in the bedrock, till
and coarse-grained sand layers. Information from approximately 70 wells
installed under the RI/FS was used to determine piezometric surfaces
and the direction of groundwater flow. Groundwater quality data from
the RI/FS and an additional sampling episode conducted by M&E were
used to identify the plume boundaries. From these studies it was
determined that contaminated groundwater flows toward the Cochato
River that flows through the site. A downward vertical gradient was
identified in the western portion of the site, while an upward vertical
gradient was identified in the vicinity of the river.
The horizontal extent of the plume with respect to the river and site
boundaries is illustrated in Figure 1. The studies indicated that con-
taminated groundwater migrates horizontally and vertically downward
to deeper layers in the source areas located in the western extent of
the plume and eventually discharges to the Cochato River. The portion
of the plume existing on the other side of the river exists predominantly
in the till and bedrock, indicating that some contamination migrates
past the river.
CONTAMINATED GROUNDWATER CONTROL 241
-------
Figure 1
Silc Map With Plume Delineation
WELL INSTALLATION
Additional field work beyond thai conducted under the RI/FS was
needed to fill in data gaps. In particular, information regarding areal
and vertical extent of contamination near the river and the source area
was lacking. Therefore, the first phase of the field work involved in-
stalling additional monitoring wells in stratified drift, till and bedrock
in those areas to further define the extent of contamination. These wells
also were used to refine the interpretation of the piezometric surface.
M&E sampled 50 new and existing wells to determine the present-day
plume configuration. The data indicated that the greatest contamina-
tion levels exist in the coarse-grained layer of the stratified drift. Till
and bedrock were found to be moderately contaminated.
In addition to the wells installed for sampling purposes, 16 observa-
tion wells and two extraction wells were installed for use during the
pumping test. Information obtained from an exploratory boring in the
vicinity of the pumping test location was used to determine screen size
and screen location of the pumping wells.
AQUIFER TESTING
The second phase of the field work involved performing two ground-
water pumping tests to determine aquifer flow characteristics. The tests
were performed in the till and stratified drift aquifers. Originally, a
test was to be conducted in the bedrock aquifer as well. However, during
well development, it was observed that groundwater yields within the
consolidated bedrock was too low to permit aquifer testing.
The pumping tests were designed by performing pretest calculations
utilizing data from previous geological site studies. Utilizing transmis-
sivity values obtained from slug tests and grain size analyses, an esti-
mate of the zone of influence of the extraction wells was made to locate
observation wells an adequate distance away from the extraction wells.
Observation wells were installed in clusters consisting of stratified drift,
till and bedrock wells. The intent was to monitor the drawdown in all
layers during the pumping of each overburden layer to characterize the
degree of hydraulic connection between layers. The well clusters were
installed at right angles to the extraction wells to determine the degree
of anisotrophy. A sketch of the aquifer test set-up is provided in
Figure 2 Based on the estimated low yield of the bedrock pumping
well, and the drillers' opinions that the weathered bedrock holds a con-
siderable amount of water, a decision was made to install two addi-
tional observation wells in the upper fractured and weathered bedrock
zone. These wells were monitored during each lest to determine the
influence of pumping the till and stratified drift on the weathered
bedrock.
The pumping rates and the length of each test were governed K> some
extent by the eventual fate of the pumped water. The pumped water
was expected to be highly contaminated, precluding the possibility of
convenient discharge to a surface water body or sewer system. After
investigating various discharge alternatives, it was decided to store the
water on-sitc in a holding tank. A 500,000-gal storage tank was con-
structed approximately 200 ft away from the pumping wells. The struc-
ture was a rectangular tank consisting of galvanized steel sides and
support frames. The tank was fitted with an HOPE liner and floating
cover to prevent the escape of water and volatile chemicals to the en-
vironment. The capacity of the tank governed the pumping rate and
the length of each test. Based on this, two consecutive pumping tests
of the till and stratified drift aquifers consisting of 3-day pumping and
3-day recovery periods were conducted. The pumping rates were 20
gpm and 75 gpm for the till and stratified drift aquifers, respectively.
242 CONTAMINATED GROUNDWATHR CONTROL
-------
o
B-3
O
S-3
6" Bedrock Well
O
S-4
O
B-4
O
T-4
Shed
O
T-3
NOT TO SCALE
Os-i
Strat Drift Pump Well
• Till Pump Well
O-
-T-1
WB-1
O
B-5
O
T-5
O
S-5
o o o o
B-2 S-2 T-2 WB-2
Storage
Building
LEGEND
O 2 Inch Monitoring Wells
® 6 inch Monitoring Wells
Figure 2
Aquifer Test Area Sketch Map
Thirty-five wells were monitored during the aquifer tests. Fourteen
wells located within a 100-ft radius of the extraction wells were moni-
tored by an electronic data acquisition system built by the Illinois State
Water Survey. The system utlilizes submersible pressure transducers
to monitor water pressure. Anaolog-to-digital and current-to-voltage con-
version circuitry allow direct transmittal of water pressure measure-
ments from submerged pressure transducers to a laptop computer. The
data are then stored in a floppy disk. The laptop computer allows the
data to be viewed as they are being recorded. This added convenience
allowed the test operator to evaluate the behavior of each aquifer dur-
ing each test and was helpful in making decisions regarding operation
of the test.
The data acquisition system proved to be an extremely useful tool
in this application. By logarithmically logging a high density of accurate
data, the response at each aquifer was clearly defined, thus facilitating
analysis. The data from the pumping test were plotted and analyzed
utilizing various software analysis packages. Time-drawdown, distance
drawdown and recovery analysis resulted in somewhat consistent results.
Values of transmissivity, storage coefficient and vertical permeability
of each layer were computed.
An evaluation of the shapes of the curves (clearly defined by the high
density drawdown measurements) indicated that the upper stratified drift
behaved as a unconfined aquifer with delayed yield, while the deeper
till and bedrock units behaved as semi-unconfined aquifers (unconfined
with leakage from higher permeability confining layers). The conclu-
sion was that the stratified drift and till aquifers are highly connected
with differences in transmissivity of only a factor of three. The analy-
sis of the drawdown data from weathered bedrock wells indicated that
the unit responds similarly to the till unit and has a similar hydraulic
conductivity. The values of transmissivity, storage coefficients and ver-
tical permeabilities as well as the shapes of the time drawdown cures
and the results of the evaluations of the relative response of the aquifers
were used to construct and calibrate a representative flow model of the
site.
GROUNDWATER MODELING
The final step in the extraction system design involved construction
and calibration of a representative groundwater flow model and then
utilization of the model to predict the effect of various extraction well
schemes on the groundwater flow regime. The model used in this study
was the three-dimensional modular groundwater flow model (MODS)
written by Michael G. McDonald and Allen W. Harbough of the
U.S.G.S.. A three-dimensional model was needed in this study to simu-
late the hydraulic relationship of the aquifers with each other and with
the river flowing through the site. It was apparent at the outset that
it would be pertinent to incorporate the effects of these relationships
into the design of the extraction system. The model consisted of three
layers.
• Silty-fine sand layer—the uppermost layer of the stratified drift.
• Coarse-grained layer—the lower-most layer of the stratified drift.
• Till and weathered bedrock—the lower most overburden unit.
A graphic interpretation of a vertical cross-section through the model
is shown in Figure 3. As the figure indicates, neither the silty-fine sand
aquifer nor the coarse-grained sand aquifer is continuous throughout
the site. Both layers pinch out to the west. Due to the similar response
exhibited during the pumping tests, the glacial till and weathered bedrock
were modeled as one unit. This layer (layer 3) extends down to compe-
CONTAMINATED GROUNDWATER CONTROL 243
-------
tent bedrock where the hydraulic conductivity is assumed to be
negligible.
Figure 3
Model Cross-Scclion
A steady-state calibration of the model was performed by adjusting
transmissivity, vertical permeability, river bed conductance, recharge
and boundary conditions until the model-computed piezometric head
distribution for each layer matched the observed distributions. The dis-
tributions were observed to be very similar for each layer except in
a few areas where vertical gradients were detected. Due to the lack
of seasonal data, the steady-state head distribution was assumed to be
the most recent interpretation of head distribution. A mass balance calcu-
lation was perfomed to determine the amount of groundwater flowing
into the river and out of constant head boundaries. This value was used
in evaluating the effectiveness of each extraction system scheme. The
calculation results indicated that most of the water existing in the system
eventually flows into the river while a small amount of water flowing
in deeper layers migrates past the river. This explained the observed
contamination in deeper layers on the oposite side of the river.
Utilizing data from the aquifer tests, a transient calibration was per-
formed to verify the predictive capabilities of the model. In this effort.
each aquifer test was simulated by inserting extraction wells and running
the model for a 3-day pumping period. Model-computed drawdowns
in each layer were compared to those exhibited during each test. Input
conditions were adjusted further until a good match was achieved. The
parameter that had to be varied the most was the vertical permeability.
This was expected due to the relative uncertainty in the calculation of
the value. Thirty-two comparisons of computed head and observed head
over two tests were made utilizing 17 wells within the model area. The
average difference between model computed head and observed head
was 0.51 ft. At this point it was assumed that the predictive capabilities
of the model were adequate.
The next step in the modeling effort was to utilize the model to predict
the effect of various extraction well schemes. The objective was to de-
termine the optima] extraction scheme that would effectively remove
the contaminated groundwater by forming a cone of depression that
encompassed the contaminant plume. Due to schedule constraints,
preliminary design of a groundwater treatment system has already com-
menced, based on an estimated maximum flow rate of 200 gpm. All
model simulations were performed at a rate of 200 gpm or less. The
steady-state head distribution utilized as initial conditions for layers 1,
2, and 3, respectively, are illustrated in Figures 4, 5, and 6. The figures
illustrate the horizontal extent of each layer. Areas in which the con-
tours are discontinous represent locations where the deeper layers crop
out at the surface.
Various extraction well configurations were modeled with different
well locations, screened intervals and pumping rates. System flow rates
were varied from 100 gpm to 200 gpm. It was found that the full
200 gpm would be needed to capture the entire contaminant plume.
RESULTS
The final design was reached through an iterative process of varying
individual well pumping rate, well depths and well locations and
determining the resultant impact on the aquifers. The design includes
two wells located in the till and weathered bedrock and four wells located
Figure 4
Steady-Slate Head Distribution—Layer 1
Figure 5
Steady-Slate Head Distribution—Layer 2
Figure 6
Steady-Slate Head Distribution—Layer 3
244 CONTAMINATED GROUNDWATER CONTROL
-------
in the coarse-grained sand layer which is the most transmissive. Figures
6, 7, 8 and 9 illustrate the cone of depression created in each layer.
The groundwater mound formed in the right side of each figure is created
by the basin recharge system. The water will be conveyed to this sys-
tem after it has been treated in the on-site treatment plant. The figures
indicate that the cone of depression will effectively contain and remove
the contaminant plume.
Figure 7
Head Distribution After Long Term Operation of
Extraction And Recharge Systems—Layer 1.
Figure 8
Head Distribution After Long Term Operation of
Extraction and Recharge Systems—Layer 2
Figure 9
Head Distribution After Long Term Operation of
Extraction and Recharge Systems—Layer 3
Precise predictions of the time it takes to remove all contaminants
from groundwater could not be made utilizing a flow model alone.
Furthermore, the large array of chemicals existing in the groundwater
makes it difficult to estimate retardation effects. However, rough esti-
mates of pore-volume removals indicate that considerable cleanup may
occur in less than a decade. Calculations show that more than 10 pore
volumes may be removed from the contaminated area in this time,
indicating that a significant amount of flushing and removal will occur.
CONCLUSIONS
The design of the extraction system discussed in this paper was based
on the results of three-dimensional groundwater flow modeling. Keeping
in mind that the representativeness of model predictions is governed
by the quality of the field information used to construct the model, the
field work performed under this project were tailored to meet the data
needs of the model.
The use of an electronic data logger proved to be a cost-effective way
of acquiring a high density of accurate drawdown measurements during
aquifer tests. In effect, it simplified the evaluation of aquifer behavior
and enhanced the representativeness of data used to construct the ground-
water model. MODS proved to be an effective tool for optimizing
extraction well schemes. Although groundwater modeling results often
are viewed with some skepticism, it cannot be argued that a well-
constructed and representative model can provide valuable insight into
the behavior of a complex hydrogeological system. Used correctly, a
groundwater model can be a useful design tool.
CONTAMINATED GROUNDWATER CONTROL 245
-------
Evaluation of the Effectiveness of
Groundwater Extraction Systems
Jennifer L. Haley
Caroline Roe
U.S. EPA
Washington, D.C.
John Glass, Ph.D.
CH2M Hill
Washington, D.C.
ABSTRACT
The most common method for addressing contaminated groundwater
is extraction and treatment. Tb evaluate the effectiveness of this process
in achieving concentration goals in the groundwater, data from 19
ongoing and completed groundwater extraction systems were analyzed.
This analysis indicated several trends including: containment of ground-
water plumes was usually achieved; contaminant concentrations initially
decreased significantly followed by a leveling out; after the period of
rapid decline, the continued decreases in containment concentration
were usually slower than anticipated; and data collected during the
remedial investigation were often insufficient to optimize system design.
design.
Factors limiting the achievement of concentration goals fell into four
basic categories: hydrogeological factors, such as subsurface hetero-
geneity, low permeability units and presence of fractures; contaminant-
related factors, such as high sorption to soil and presence of non-aqeous
phases (dissolution from a separate non-aqueous phase or partitioning
of contaminants from the residual non-aqueous phase); continued migra-
tion from source and size of the plume itself; and system design factors,
such as pumping rate, screened interval and extraction well location.
The findings of this study indicate that groundwater extraction is an
effective method for preventing further migration of contaminant plumes
and achieving risk reduction by removing a substantial mass of con-
taminants from the groundwater; however, the findings indicate that
in certain situations, it may not always be practicable to achieve health-
based cleanup concentrations throughout the groundwater to fulfill the
primary goal of returning groundwater to its beneficial uses. Where
cleanup to health-based concentrations throughout the groundwater is
not practicable, extraction and treatment can be operated to to optimize
contaminant mass removal and contain the groundwater plume. Con-
clusions that can be drawn from the study arc: plume containment should
be considered early during site management planning; certain data not
currently collected on a routine basis should be gathered to better esti-
mate restoration time frames and system response; and groundwater
remedies should be flexible to allow for modification to the system.
INTRODUCTION
Laboratory researchers and hydrogeologists involved in groundwater
contamination cleanup have been encountering several conditions that
can limit the rate at which contaminants can be removed from the
subsurface. The project described in this paper was initiated to assess
the validity and prevalence of these findings in actual experiences with
groundwater extraction to date. The purpose of the project was to assess
the effectiveness of groundwater extraction systems in achieving speci-
fied goals at sites where groundwater extraction systems had been
operating for a period of time long enough to generate performance
information.
Several sources of data were reviewed in an effort to identify ground-
water extraction systems currently in operation and actions that had
been completed and pumping terminated. Information on 112 sites
including Superfund, RCRA and industrial sites where groundwater
response actions were being implemented by the U.S. EPA, other Federal
Agencies, Slates or responsible parties, was collected and organized
in a data base for review. The majority of these sites, however, had
not reached a full implementation phase and consequently were not
useful for this study. Nineteen cases were identified as good candidates
for more in depth evaluation based on the data available on actual
performance.
This paper presents the findings of the study and provides examples
from the 19 case studies examined in detail that illustrate the various
factors that can affect the performance of groundwater extraction
systems. Finally, recommendations based on this study are summarized.
BACKGROUND ON CASES
The 19 case studies form a representative sample of the variety of
conditions frequently encountered when performing groundwater
extraction. Pertinent aspects of the 19 sites is provided in Table 1. Several
general characteristics are worth noting.
In all cases, one of the goals of the extraction systems was to prevent
further migration of contaminants. Twelve of the cases also specified
quantitative concentration or contaminant mass reduction goals as well
as containment. The seven remaining cases generally indicated a desire
for contaminant mass reduction but did not clearly specify this as a
goal. Three of the sites involved treatment at existing well-heads;
however, this action was incorporated into the groundwater extraction
system, generally not with the goal of reducing contaminant concen-
trations but with the intent of preventing further migration beyond the
wells. The existing wells acted as a barrier system which prevented con-
taminant migration to other drinking water wells.
The period of operation of the 19 extraction systems at the time avail-
able data were reviewed ranged from 5 mo to 6 yrs. In most cases,
the systems had been operating longer than the projected lime required
for cleanup; however, concentration-based goals had not yet been
attained and extraction was continuing.
The variety of contaminants encountered in these sites was limited.
The primary contaminants in all but two cases were volatile organic
compounds (VOCs). This finding is not surprising since VOCs are the
most prevalent groundwater contaminants found at Superfund sites and
tend to be more mobile than other classes of compounds. Semi-volatiles
were also present in two cases. Chromium, pesticides and creosote were
detected at one site each.
The 19 case studies represent a broad spectrum of geologies from
246 CONTAMINATED GROUNDWATF.R CONTROL
-------
Table 1
Summary of Case Study Site Characteristics
Site Name
Amphenol Corporation
Black & Decker, Inc.
Des Moines TCE
Du Pont Mobile Plant
Emerson Electric Company
Fail-child Semiconductor
General Mills, Inc.
GenRad Corporation
Harris Corporation
IBM Dayton
IBM San Jose
Nichols Engineering
01 in Corporation
Ponders Corner
Savannah River Plant
Site A
Utah Power Light
Verona Well Field
Ville Mercier
Date of Initial
Extraction
January 1987
May 1988
December 1987
December 1985
December 1984
1982
Late 1985
Late 1987
April 1984
March 1978
May 1982
January 1988
1984
September 1984
September 1985
August 1988
October 1985
Hay 1984
1983
Remedial
Objective
Restoration
Restoration
Restoration
Containment
Restoration
Containment
Restoration
Restoration
Well-head treatment
& Restoration
Was restoration,
now containment
Restoration
Restoration
Containment
Well-head
Treatment
Restoration
Restoration
Restoration &
Containment
Restoration &
Containment
Containment
Chemicals
Present
Organi cs
Organi cs
Organi cs
Organi cs
Organi cs
Organi cs
Organi cs
Low Sorption
Organi cs
Organi cs
Metals
Organi cs
Organi cs
Organi cs
Organi cs
Low Sorption
Organi cs
Low Sorption
Organi cs
Organi cs
Organi cs
Organ ics
High and Low
Sorption
Organics
Geologic Environment
Unconsolidated glacio-f luvial
sediments
Glacial till & fractured
Unconsolidated glacio-f luvial
sediments
Alluvial sand & clay
Sand
Alluvial sand & gravel
with silt & clay layers
Peat, glacial deposits,
& fractured rock
Glacial sand, gravel
Sand & shell with a
clay layer
Sand with clay layers
Alluvial sand & gravel
with silt & clay layers
Weathered & fractured shale
Unconsolidated glacio-f luvial
sediments
Unconsolidated glacio-f luvial
Coastal plain sand,
silt & clay layers
Limestone & sand
Alluvium & fractured
basalt
Glacial sand, gravel,
& clay
Unconsolidated glacial
sediments & fractured rock
Innovative
Technologies
Fracture
enhancement
Slurry wall
Intermittent
pumping
Well points
Well points
Reinjection
Vapor extraction
Intermittent
pumping
Vapor extraction
CONTAMINATED GROUNDWATER CONTROL 247
-------
various geographic locations. Two of the sites are located in the north-
western United States, seven are located in the southeast, six in the
northeast, two in the southwest and two in the midwest. This geographi-
cal distribution was considered an important factor in assessing the role
that varying hydrogeologies may play in impeding or promoting
extraction of contaminants.
GENERAL OBSERVATIONS
Several trends were observed in looking at the overall performance
of the systems. As discussed above, a common goal of all the actions
was containment of the contaminant plume. In the majority of the cases,
this goal appeared to be successfully achieved. Groundwater gradient
data indicated an inward gradient toward the center of the plume and
little or no movement of contaminants beyond plume boundaries that
existed at the initiation of the action.
Contaminant mass removal was usually significant. Removal of
thousands of pounds of contaminants (up to 130,000 Ibs in one case)
was not uncommon. However, the rate of mass removal often declined
quickly to low levels. This initial drop in removal rate is thought to
be the result of a combination of removing groundwater faster than the
contaminants can desorb from the soil, lowering water tables below
the most contaminated portions of the subsurface and diluting concen-
trations by drawing in less contaminated groundwater from surrounding
areas. Although concentrations in the groundwater appeared to be
reduced significantly, the levels remaining were generally above health-
based standards for drinking water, which was the most common con-
centration goal of the actions.
FACTORS AFFECTING PERFORMANCE—CASE EXAMPLES
The factors affecting die performance of the extraction systems
examined in this study fell into the following four primary categories:
• Aquifer properties such as subsurface heterogeneity and presence
of low permeability units or fractures
• Contaminant properties such as level of sorption to soil, immisci-
bility (dissolution from non-aqueous phases or partioning of other
contaminants from residual non-aqueous phase) and density
• Adequacy of source removal and size of the plume itself
• System design such as pumping rate, location of extraction wells and
depth/length of screened interval
The following sections illustrate the impact these factors may have
on the performance of groundwater extraction systems using examples
from the case studies reviewed.
Aquifer Properties
All of the cases reviewed in this study reflected complications resulting
from the heterogeneous nature of the subsurface. Well-sorted
homogenous hydrogeologica] systems below a contaminated site tend
to be the exception rather than the norm. At a chemical plant site in
Alabama, it appeared that the implications of the heterogeneous nature
of the subsurface material may not have been accounted for in the design
of the extraction system. The water level data from monitoring wells
located around the site indicated that plume capture had been achieved.
However, a mass balance on the system revealed that about half the
contaminant mass was escaping the recovery wells. A possible expla-
nation for this apparent conflict is that contaminants were moving be-
low the screened interval of the extraction wells. This explanation is
supported by the fact that the hydraulic conductivity of the subsurface
material increased with depth and all the on-site wells were screened
in the upper, less permeable portion of the aquifer. In addition, a near-
by production well screened at the lower depths continued to operate
during this period and may have increased the vertical migration of con-
taminants.
The impact of low permeability units in the subsurface is illustrated
by the Pbnder's Corner site in Lakewood, Washington At this site, the
variation of contaminant concentrations with depth was assessed and
correlated to the subsurface stratigraphy. This analysis indicated that
almost 90% of the primary contaminant, tetrachloroethylene (PCE).
present was located in a very low-permeability silt and clay unit. Con-
taminant removal rates are limited not only by the slow rate at which
groundwater can be pulled through this unit, but also by the fact that
the soil in this zone has a higher organic carbon content and conse-
quently enhances sorption of the PCE to the soil.
Several of the case studies involved sites where fractures played a
role in contaminant movement At the Black and Decker Site in Brock-
port, New York, the identification of discrete fractures lead to the con-
clusion that recovery of TCE-contaminated groundwater would be very
difficult. In order to create interconnections between the discrete
fractures, explosives were set off in the contaminated zone.
Contaminant Properties
Another factor that plays a role in virtually all the case studies
reviewed is sorption. The amount of contaminants sorbed to the soil
often is not accounted for in estimating restoration time-frames or in
confirming that final cleanup goals have been attained. At the Savannah
River Plant in Aiken, South Carolina, the contaminant mass in the
groundwater was estimated based on groundwater concentrations. After
3 yrs of extraction, a comparison was made between the mass removed
at the extraction wells and the difference in the estimated mass remaining
insitu based on the groundwater concentrations before and after extrac-
tion. The mass actually removed by the system was I48j000 Ibs; however,
the groundwater concentration comparison indicated that only 23jOOO
Ibs had been removed. The discrepancy can be attributed to contaminants
sorbed to the soil that were dissolving into the groundwater as it was
drawn to the extraction wells.
The presence of non-aqueous liquids that either float or sink in the
aquifer can substantially increase the restoration time by acting as a
continuing source of contaminants to the groundwater. At the IBM Day-
ton Facility in South Brunswick, New Jersey, the extraction system was
operated for 6 yrs, and concentrations appeared to be stabilizing at a
level determined to be acceptable to the State. Extraction was then ter-
minated. Continued groundwaler monitoring revealed that containment
concentrations were increasing. It was concluded that this was the result
of contaminants present in a non-aqueous phase more dense than water
that had sunk within the aquifer. Because it would be very difficult
to locate and completely remove the pockets of contamination, the goal
of the extraction system was changed to containment. Extraction was
re-initiated at a lower pumping rate and was projected to continue
indefinitely.
Problems can result from non-aqueous liquids that are less dense than
water, as well. At the Verona Well Field site in Battle Creek, Michigan,
a non-aqueous phase liquid layer approximately 1 ft thick was detected
floating on die water table. Traditional product recovery techniques in-
volving creation of a drawdown cone into which product would flow
and could be recovered were used to reduce this layer to approximately
1 in. At this point, product recovery techniques were no longer effec-
tive, but the remaining floating layer was sufficient to provide a source
of contaminants to the groundwaler at levels above the cleanup goals
established for the site. A vapor extraction system was then installed
to remove the remaining product.
Adequacy of Source Removal
One of the more obvious factors that can affect the ability of ground-
water extraction systems to achieve concentration reductions in the
groundwater is the adequacy of measures taken to prevent continued
contaminant migration from source areas. Soil cleanup levels often are
based on an evaluation of direct contact threats and may not account
for the continued migration of contaminants to the groundwaler. At an
industrial site in Minnesota, concentrated wastes were removed from
a disposal pit during the source action. Contaminated soil below the
waste was not removed, despite sampling results which indicated that
significant levels of contaminants were present in the soil. This factor
probably contributed to the difficulty experienced in efforts to reduce
groundwater concentrations during extraction at this site.
System Design
Another factor affecting extraction performance is the design of the
extraction system. In the case of the Alabama site previously discussed.
248 CONTAMINATED GROUNDWATER CONTROL
-------
the screened interval of the extraction wells may have been too shallow
to contain the plume of contaminated groundwater. At an industrial site
in Florida, portions of the contaminant plume were not captured by
the extraction system since the extraction system did not address con-
taminants in the upper aquifer. Fortunately, the extraction system
included the operation of barrier wells (existing water supply wells with
well-head treatment systems) downgradient from the restoration system
which were expected to capture the portion of the plume that escaped.
CONCLUSIONS
The results of this project highlight factors and approaches that are
prudent to consider in developing and implementing groundwater
response actions. These findings do not alter the primary goal of
returning groundwater to its beneficial uses in a time-frame that is
reasonable given the particular site circumstances. Rather, they argue
for the collection of data to allow for the design of an efficient cleanup
approach that more accurately estimates the time-frames required for
remediation and the ultimate cleanup levels achievable. The conclusions
cover three basic aspects of site remediation: consideration of early ac-
tion, site characterization and remedy specification. In addition, con-
sideration on a more routine basis of various methods to enhance the
effectiveness of groundwater extraction appears warranted.
Conclusion 1: Plume Containment Should
Be Considered Early
One of the program management principles identified in the revised
NCP is the intent of the U.S. EPA to balance the goal of definitively
characterizing site risks with the bias for initiating response actions
as early as possible. Where groundwater contamination is involved,
this bias for action should be reflected by considering, early in the site
management planning process, measures that can be implemented to
prevent further migration of contaminants if these measures will prevent
the situation from getting worse and provide useful information to design
the final remedy. Because the data needed to design a containment system
often are more limited than that needed to implement full remediation,
it will in many cases be valuable to prevent the contaminant plume from
spreading while the site characterization to select the remediation system
progresses.
The determination of whether or not to implement such a system
would be based on existing information, best professional judgment
and data defining the approximate plume boundaries, contaminants
present and approximate concentrations. The justification for taking
the action would be based on a comparison of the benefits of taking
an action and the possible benefits of waiting to act until the investiga-
tion has been completed.
The advantages of early action include prevention of further con-
taminant spreading and the generation of useful data on the response
of the hydrogeologic system. If it is determined that a containment action
should be implemented, the advantages of initiating an action should
be maximized by carefully monitoring system response. In particular,
groundwater flow should be monitored frequently, immediately before,
during and immediately after initiation of the action to obtain informa-
tion on system response.
Conclusion 2: Data That Will Assist In
Assessing Contaminant Movement and Likely Response to
Extraction Should Be Collected
In addition to the traditional plume characterization data normally
collected, assessments of contaminant movement and extraction effec-
tiveness can be greatly enhanced by collecting more detailed informa-
tion during construction of monitoring wells. Frequent soil or rock
coring and the use of field techniques to assess relative contaminant
concentrations in the cores are ways that might be used to gain this
information. Analysis of contaminant sorption to soil in the saturated
zone can also provide the basis for estimating the time-frame for
reducing contaminant concentrations to established levels and identifying
the presence of non-aqueous phase liquids. Cores taken from depths
where relatively high concentrations of contaminants were identified
might be analyzed to assess contaminant partitioning between the solid
and aqueous phases.
Conclusion 3: Flexibility Should Be Provided
In The Selected Remedy That Allows For System
Modifications Based on Information
Gained During Operation
In many cases it may not be possible to determine the ultimate con-
centration reductions achievable in the groundwater until the ground-
water extraction system has been operated and monitored for some
period of time. Remedies should provide flexibility that allows for
modifications, or should indicate that the initial action is an interim
measure and that the ultimate remedy will be evaluated at some speci-
fied future date. This iterative process of system operation, evaluation
and modification can effectively result in the optimimum system design.
Three options for describing remedies that account for the uncertainty
in system response to extraction are outlined here. The appropriate-
ness of a given option relates to the level of confidence associated with
the expected performance of the extraction system with respect to
achieving specified concentration goals. The options are listed below
in order of decreasing confidence that specified concentration goals
are practicable to attain:
1. Select a remedy designed to achieve specified concentrations in the
groundwater that reflect achievement of the basic goal of returning
groundwater to its beneficial uses. If the achievement of these goals
is determined to be impracticable based on data gathered during im-
plementation, the remedial action would be continued or modified
to achieve the secondary goal of optimizing contaminant mass
removal. The methods used to evaluate when optimum mass removal
is achieved and any associated ARAR waivers would be fully
described in advance as a contingency remedy.
2. Select an interim remedy that will be monitored carefully for some
specified period of time; e.g., 5 yrs, to determine the practicability
of returning the groundwater to its beneficial uses. At the end of
this defined observation period, the effectiveness of the remedy would
be evaluated and the final action determined.
3. Select a remedy designed to optimize mass removal, reducing risks
to the extent praticable, over those portions of the aquifer where
contaminant concentrations cannot be reduced sufficiently to return
the groundwater to its beneficial uses. Any ARARs, such as MCLs
or State standards, that would not be acheived in the area of attain-
ment would be waived. Institutional controls would be implemented
in perpetuity to prevent access to portions of the groundwater where
contaminants remain above health-based levels and containment
measures would be continued to prevent migration of contaminants
at concentrations exceeding health-based levels to clean groundwater.
The decision to use this option must be based on data that clearly
indicate the impracticability of returning groundwater to its benefi-
cial uses. A contingency should be included to the effect that if
operation of the system indicates that health-based goals can be
attained, the remedy should be operated to achieve this goal.
Under all of the options, groundwater monitoring should continue
for at least 2 to 3 yrs after active remediation measures have been com-
pleted to ensure that contaminant levels do not begin to increase. For
cases where contaminants remain above health-based levels, a review
after 5 yrs would be required.
If it is determined that the primary goal cannot be met over some
portion of the area of attainment, an evaluation of when optimum mass
removal has been achieved must be made. This evaluation might be
based on reaching a point of diminishing returns; that is, concentra-
tion reductions are no longer significant although contaminant mass
continues to be removed (concentrations approach an asymptotic level).
Alternatively, the evaluation may be based on the concentrations that
would be expected to migrate from the site should the extraction system
be shut off. Experience to date on this phase of groundwater remedia-
tion is limited and more definitive guidance can only be developed with
collection of data during actual system operation. When the point of
diminishing returns has been reached, however, this should be viewed
as a signal that some re-evaluation of the remedy is warranted.
CONTAMINATED GROUNDWATER CONTROL 249
-------
Conclusion 4: Methods to Enhance Extraction
Effectiveness and Efficiency Should be Considered
It is clear from many of the case studies that variations on system
design and operation can improve the effectiveness and efficiency of
extraction. Some of these methods such as infiltration/re-injection and
construction of slurry walls are fairly traditional. Others, like vapor
extraction in conjunction with groundwater extraction and fracture
enhancement, are relatively new and appear promising for certain types
of situations. It may be appropriate to use some innovative technolo-
gies, such as in-situ biorestoration, in a treatment train where extrac-
tion is used to achieve initial concentration reductions followed by the
use of the innovative technology to reduce concentrations an additional
increment. Finally, some alterations of traditional pumping systems may
be worth consideration in the majority of cases. This includes inter-
mittent pumping to allow for containment and water level
re-equilibration. Another consideration is how operation of the system,
e.g., location of operating extraction wells, can be progressively
modified based on observation of aquifer and plume response.
SUMMARY
Groundwater extraction will continue to be a primary method for
addressing contaminated groundwater to reduce plume spread and
remove contaminants from the groundwater. An evaluation of several
representative cases of groundwater extraction indicates that there are
svcral factors and circumstances that can limit the overall performance
of extraction. These factors should be recognized during site investi-
gation through more detailed data collection. Also, remedies should
be modified during system operation in response to data collected. In
addition, it is valuable to consider the benefits of implementing a con-
tainment system prior to full site characterization to prevent contaminant
migration as the investigation is completed.
250 CONTAMINATED GROUNDWATER CONTROL
-------
Field Investigation to Characterize Relationship Between
Ground Water and Subsurface Gas Contamination
At a Municipal Landfill
Ken Smith
Camp Dresser and McKee
New York, New York
Jeanne Martin
Camp Dresser and McKee
Boston, Massachusetts
Edwards Als
U.S. EPA
New York, New York
ABSTRACT
Under the direction of the U.S. EPA, a remedial investigation was
performed at an inactive municipal landfill. Leachate from this
municipal landfill had contaminated an upgradient municipal well in
an adjacent residential area with organics. Methane and other gases
were historically detected in homes in the same area, apparently entering
the homes from the subsurface.
Determination of the source/pathway from the landfill to the municipal
well was complicated by the fact that the well was upgradient and that
rapidly moving inorganic contaminants associated with leachate (i.e.,
cations, anions) were never detected in the well discharge. Thus, it was
suspected that there was a relationship between contaminants in the
subsurface gas phase and those detected in groundwater at the municipal
well. In order to determine the nature and extent of contamination and
to design appropriate remediation, both ground water and subsurface
gas contamination needed to be assessed.
A field investigation and sampling protocol were designed to study
the relationship between contaminants in various media including gas,
groundwater and soil. Groundwater monitoring wells were installed
upgradient and downgradient of the landfill. Soil samples were collected
from the vadose zone during the installation of subsurface gas monitoring
wells. Gas monitoring wells with multilevel probes were installed in
the adjacent neighborhood. Groundwater samples were collected to co-
incide with seasonal gas sampling events. Samples of landfill leachate
and gas emanating from the surface of the landfill also were collected,
and an evaluation of the landfill gas migration control (extraction) system
was performed.
INTRODUCTION
The landfill site described in this paper is located in New York State,
on Long Island, where residents rely almost exclusively on ground-
water for drinking water. The regional geology consists of bedrock,
overlain by Cretaceous deposits including the Lloyd aquifer, the Raritan
clay (a significant aquitard) and the Magothy formation. These Creta-
ceous deposits lie under Pleistocene glacial deposits over most of the
Island. The Lloyd aquifer, Magothy formation and Upper Glacial for-
mation all are used for water supply in communities neighboring the site.
The landfill is located in a former sand and gravel pit which is adja-
cent to a residential area. The pit was excavated towards the residential
area, creating a sand cliff. Refuse was placed in this pit (Fig. 1). Because
of poor construction records, it is not clear if a continuous clay barrier
was installed between the refuse and the residential area.
Construction and filling of the landfill began in 1974. As refuse was
collected, a 20-mil PVC liner was installed until an area of 29 ac was
covered. There were several documented leachate spills prior to com-
NOTE:
• - Concentrations Representative Ol ToM Volatile Organic Priority Pollutants.
Figure 1
Conceptual Block Diagram of Contaminant Transport
pletion of the liner. Approximately 260,000 tons of municipal and con-
struction debris were disposed of each year until the landfill ceased
operation in 1981. There also were undocumented reports of illegal
dumping and drum disposal in the landfill. Refuse is still received in
an active fill area adjacent to the inactive landfill.
During the winters of 1979-1981, small explosions (furnace "puff-
backs") occurred in several homes in the residential area. Air monitoring
by the local fire department and the department of health revealed that
methane levels exceeded the lower explosive limit (LEL) in several
homes as well as the subsurface. Other gases, including vinyl chloride,
benzene, toluene and tetrachloroethene, also were detected in homes.
In 1981, a passive gas venting system was installed around the periphery
of the landfill. Soon after, additional vents were installed and blowers
were attached to create an active gas extraction system for the landfill.
Also in 1981, volatile organic contaminants were detected in the upgra-
dient municipal supply well located in the adjacent neighborhood
(Fig. 1). The municipal well was closed because of this contamination.
Current operations at the inactive landfill include a gas extraction
system (active system) consisting of stainless steel and PVC active vents
CONTAMINATED GROUNDWATER CONTROL 251
-------
(located around the periphery of the landfill) attached to a manifold
blower system. Gases are pumped to a horizontal combustion unit for
destruction. There is also a passive gas venting system consisting of
PVC vents around the periphery of the landfill and large concrete cisterns
(passive vents) throughout the fill. A leachate collection system was
installed in 1976, and leachate is collected and aerated at the base of
the inactive landfill before being pumped to a POTW. This landfill is
not capped, and rainfall entering the fill, generates significant quanti-
ties of leachate.
The understanding of contamination prior to initiation of field
activities indicated that if the landfill were to be the source of upgra-
dient groundwater contamination at the municipal well, then there would
have to have been a transport mechanism that could account for the
organic contaminants alone (i.e., without inorganic contaminants).
Because of the gas contamination which had occurred in the subsur-
face surrounding the fill, it was suspected that contamination existing
in the gas phase was contributing to contamination detected in the
groundwater.
FIELD PROGRAM DESIGN AND IMPLEMENTATION
The sampling program was designed to evaluate the nature and extent
of contamination and to evaluate the source/receptor pathway for the
contamination detected off-site. Historically, methane and volatile
organic vapors (such as vinyl chloride and tetrachloroethene) had been
detected off-site. The installation of the gas venting system at the land-
fill had reduced significantly off-site migration of methane gas, but trace
concentrations of volatile organics continued to be detected in subsur-
face wells designed to monitor the effectiveness of the gas venting
system. The quality of groundwater hydraulically upgradient of the land-
fill had not been characterized except for the historic detection of vola-
tile organics at the municipal well and at monitoring wells directly
adjacent to the landfill.
Prior to the start of the field investigation, various contamination
sources and pathways were considered for evaluation. These pathways
included discharge of leachate to the groundwater (although the lack
of inorganic contaminants at the upgradient municipal well limited this
hypothesis), gas to soil to groundwater partitioning of organic con-
taminants in the subsurface, condensation of gas phase contaminants
when warm moist air from the landfill encountered cooler ambient soil,
and direct gas to groundwater partitioning when rainfall percolated
through the vapors in the subsurface surrounding the landfill.
The field investigation set out to first, fill in the data gaps for both
the groundwater and the subsurface gas, and second, to assess whether
the landfill was the source of upgradient groundwater contamination.
The effort to supplement the existing groundwater data included the
installation of an extensive groundwater monitoring well network, the
sampling of new and existing groundwater wells, and the performance
of hydraulic permeability and pump tests. In order to address the sub-
surface gas data gaps, additional, multi-level gas monitoring wells were
installed off-site, and samples of the subsurface gas were collected from
the new and existing gas wells. Finally, an assessment of the effective-
ness of the landfill gas extraction system was performed, and samples
were collected to determine the emission rate of volatile organic com-
pounds (VOC) through the surface of the landfill. The results from the
groundwater, subsurface gas and landfill assessment investigations were
compared to address whether the landfill was the source of upgradient.
groundwater contamination.
Groundwater Investigation
Groundwater monitoring wells were installed upgradient of the landfill
between the fill and the contaminated municipal well and at background
locations. Additional wells were installed hydraulically downgradient
of the landfill to determine if additional contamination existed at down-
gradient locations and to compare contamination patterns. The
monitoring wells were installed in the Upper Glacial formation with
several clusters which included wells screened in the Magothy formation.
Two rounds of groundwater samples were collected to coincide with
two of the seasonal subsurface gas sampling events. The basic contami-
nation pattern indicated a significant plume of volatile organic con-
taminants (total VOC concentrations > 500 mg/L) associated with the
monitoring wells just upgradient of the landfill. Lower concentrations
of these contaminants were detected in wells further upgradient from
the landfill, closer to the municipal well. Contamination was not de-
tected when the municipal well was sampled. However, the well had
been off line for over 6 yr.
Groundwater contamination was also detected downgradient of the
landfill. In comparison, this contamination was similar to that which
might result from leachate contamination. Downgradient groundwater
contained not only some VOCs, but also various semi-volatile and in-
organic compounds. Contamination was detected at much lower con-
centrations (generally < 10 mg/L for a few individual volatile and
semi-volatile organic contaminants). The inorganic compounds were
detected at proportions very similar to the leachate collected from the
landfill. However, many more organic contaminants were detected in
the leachate than in the downgradient groundwater.
The results of earlier pump tests which had been performed on the
closed municipal supply well indicated that the zone of capture of this
well extended well under the inactive cell of the landfill during both
summer (maximum) and winter (minimum) pumping conditions. The
pattern of contamination observed during groundwater sampling (highest
groundwater contamination closest to the landfill, with decreasing con-
centrations with distance from the landfill) led to the conclusion that
the landfill was the likely source of upgradient contamination. Com-
parison to observed downgradient concentrations indicated that upgra-
dient contamination probably was not the result of leachate migration.
Also, the known historical pattern of contamination indicated that con-
centrations and types of contaminants did not result from leachate.
Subsurface gas investigation
The focus of the Rl into the subsurface, vapor phase contamination
was to define the nature, degree and extent of this contamination within
the unsaturated zone off-site and to determine whether this was inter-
related with the groundwater contamination problem.
In an effort to define the nature, degree and extent of the vapor con-
tamination problem, a complex gas monitoring well network and
sampling plan was devised. The goals of this approach were: 0) to
determine whether subsurface vapor contamination was related to landfill
generated gases in terms of composition and concentration, and (2) to
evaluate the potential for landfill generated gases to migrate from the
landfill to the vadose zone off-site, by performing a mass balance
approach on the landfill and by analyzing the effectiveness of the existing
gas migration control system.
At the initiation, and during the course of the study, several concep-
tual pathways were identified to explain the observed patterns of con-
tamination in off-site gas and groundwater. Several»of these pathways
included vapor phase contamination from the landfill affecting con-
centrations of organic contaminants in groundwater. It was recognized
that the result of the study would probably indicate that a combination
of pathways was acting. The conceptual pathways are outlined as follows:
I. Gas-Water Partitioning:
Relationships for vapor-liquid equilibrium for dilute aqueous solu-
tions arc very well defined and are governed by Henry's law, which
states:
p = Hx
where p = partial pressure of a substance at a given temperature
(T) and pressure (P)
H = the Henry's law constant at a given T and P,
x = the liquid phase concentration at a given T and P
Therefore. Henry's law defines the distribution of a substance between
the vapor and liquid phases for a system in equilibrium. If we assume
that gases within the subsurface are in equilibrium with infiltrating rain-
water, then Henry's law defines a pathway for the transport of vapor
phase contaminants to the groundwater.
2. Gas-Soil-Water Partitioning
Several researchers1'2 have studied the sorptive characteristics of
soils for volatile organic compounds (VOCs) under conditions
252 CONTAMINATED GROUNDWATER CONTROL
-------
representative of the saturated and unsaturated (or vadose) zones.
These studies have shown the soil-vapor partition coefficients to sig-
nificantly differ such that the soil-vapor partition coefficient (soil
concentration to vapor concentration) is orders of magnitudes greater
in unsaturated conditions rather than saturated. Thus, gas partitioning
to the solid/soil phase, with subsequent partitioning to the liquid
phase from infiltrating rainwater could be a significant mechanism.
3. Direct or Indirect Landfill Condensate Discharge
Gases generated within a landfill carry large quantities of water, along
with any primary gases or trace contaminants. In fact, gases within
a landfill are saturated 'relative humidity of 100%) with landfill tem-
peratures ranging from 80 to 140 °F. Due to this high moisture con-
tent, condensate traps are standard requirements for all landfill gas
extraction systems. The condensate formed when these warm,
moisture-saturated gases are extracted to cooler surroundings usually
contains high concentrations of VOCs, and, in fact, can form a free
organic phase.3 Due to a poorly designed manifold system between
active wells at this site, it was observed that condensate from ex-
traction vents within the landfill could collect in the headers and
then drain into active vents external to the landfill and into native
soil. Additionally, it was theorized that indirect discharge could result
from condensate forming as moisture laden landfill gas migrated
from the warm landfill (80 to 140 °F) into the vadose zone off-site
where subsurface temperatures dropped to approximately 50 °F.
The subsurface gas field investigation was designed to test all the
pathways contributing to the contamination, except for the discharge
of leachate, and included several different field activities. The activities
are described in the following paragraphs.
Monitoring Well Network Installation
To define the nature, degree and extent of off-site, subsurface gas
contamination by VOCs, a monitoring well network using multilevel,
nested, subsurface gas monitoring wells (denoted landfill gas wells or
LFG wells) was installed. The purpose of the wells was to determine
the lateral and vertical (with depth) distribution of vapor phase con-
taminants. Typical well construction for the LFG wells is shown in
Figure 2. The wells contained four probes (labeled A through D) set
at approximately equidistant intervals. The probes were 0.5-in. O.D.
virgin Teflon-TFE tubing with the bottom 5 ft perforated with 0.25-in.
holes. Each perforated section (designated as the probe section) was
screened in a 15-ft zone of #1 moire gravel and isolated from the other
probes by a 2- to 5-ft layer of bentonite/cement slurry. Each well was
installed using hollow-stem auger drilling techniques.
The LFG wells were placed both to supplement existing LFG wells
and to complement new and existing groundwater wells (Figures 3a
and 3b). During installation of the LFG wells, soil samples were col-
lected in specially designed split-spoon sampler liners to evaluate the
vapor-soil-water contaminant pathway. The samples were collected in
two foot intervals to correspond to the intervals in which the gas
monitoring probes would be set. The intervals were generally 20-22,
50-52, 100-102, and 120-122 below grade. The soil samples were col-
lected in decontaminated stainless steel sleeves inserted into 21 split
spoon samplers. The ends of the samples were covered with air tight
plastic caps and Teflon inserts. This method minimized volatilization
of soil contaminants.
Sampling of the LFG Wells
A large amount of landfill gas monitoring data had been collected
from the site before the initiation of the RI. This included an investiga-
tion by U.S. EPA's FIT team and monthly sampling of existing landfill
gas wells by the local municipality. However, up to this point most of
the gas sampling data lacked the quality to withstand litigation. For
this and other technical reasons (including a need to cover a wide range
of concentrations, the affects of humidity on sorbent traps and analyti-
cal reproducibility) SUMMA canisters were chosen as the method of
sampling.
SUMMA canisters are stainless steel canisters with a specially pas-
sivated internal surface (passivated by the SUMMA process) that makes
m
- 0.5"00.x 0.375"ID. TFE Teflon Tubing
- Probe Section: Last ,5 ft. of tubing perforated with 0.25" holes
Q
D
B
. Backfilled Soil
, 0.25"x 0.5" Pea Gravel
Horie #1 or #0 Sand
Bentonite/Cement Slurry Seal
Figure 2
Typical EPA LFG Well Construction
them inert to the adsorption, desorption and degradation of VOCs.
Although, the SUMMA technology is relatively new, U.S. EPA-RTP
has tested the stability of VOCs within the canisters4'5 and has had suc-
cess using them in ambient air studies. It appears that U.S. EPA may
eventually support SUMMA canisters as the recommended method for
air and vapor sampling for VOCs, replacing EPA draft methods TO-1
and TO-2.
At the time of the field investigation, several methods for the analy-
sis of gas/air samples from SUMMA canisters had been developed,6-7
but the U.S. EPA's draft method TO-14 (for the analysis of VOCs from
SUMMA canisters) was still under preparation. A modified version
of water method 624 (GC/MS), where the gas from the canisters was
passed through a sorbent trap and then desorbed to the GC/MS sys-
tem, was applied and proved to be very successful.
Four rounds of LFG well sampling were performed, one in each
season, in an attempt to measure seasonal effects on the concentrations
of VOCs within the subsurface off-site. Additionally the LFG well
sampling events were performed concurrently with the four rounds of
landfill surface emission rate sampling and the two synoptic rounds
of groundwater sampling.
Landfill Assessment
In order to assess the potential for landfill-generated gases to migrate
from within the landfill to the unsaturated zone off-site, and to provide
a basis for selection of remedial alternatives for the FS, an assessment
of the effectiveness of the existing landfill gas migration control system
(or gas extraction system) was performed. This assessment of the gas
extraction system involved the following activities:
• The installation of multi-level, nested, pressure-probe wells at radial
distances from a representative extraction vent. These probes, when
monitored under the varying conditions of the vent testing, would
provide an estimate of the sphere of influence for the extraction wells
CONTAMINATED GROUNDWATER CONTROL 253
-------
m • flipnmm rnt smn THM on-sat Ltnana Git Montortng Wrti (1-7)
• - Ktmtmu na four EPA on-sat Linaau OH won/wring WHII
N
Figure 3a
Location of Landfill Gas Monitoring Wells
LEGEND:
N-H20 • County Will
THH-3 • Exlitlng Town win
CPA-105 • US emlmnminul Pmitciion Agincy win \S
0 400 MO 1MO "*'
Kilt (
l (Ml
Figure 3b
Location of Groundwatcr Monitoring Wells
and thereby determine the proper well spacing within the refuse;"
• The collection of extraction vent gas samples for primary gas (O2,
CO1, N1, CH') analysis and for VOC analysis and the measurement
of well head vacuums, flow rates and methane contents under con-
ditions of varying applied vacuum;
• An examination of the physical condition of the extraction system
including the measurement of water and sediment levels in extraction
vents;
• Sampling of condensate collected within the extraction system
manifold for VOCs.
Also, under the heading of the landfill assessment, but concurrent
with the LFG well sampling, samples of the gas venting through the
landfill surface were collected using specially designed flux isolation
chambers (flux boxes). The flux boxes were adapted versions of those
used by Radian and the U.S. EPA" The purpose of these samples was
to quantify emission rates from the landfill to the ambient for use in
the public health evaluation (PHE) for the RI
Other studies have suggested that surface flux emission rates along
with data from pressure probes and vent tests could be used to perform
a mass balance analysis on a landfill" This approach was examined
during the RI to see if it could be used to determine estimated quanti-
ties of gas migrating off-site.
The general approach for performing a landfill gas mass balance
analysis on a landfill involves estimating or measuring each stream
leaving the landfill and estimating an overall emission rate. Consider
a control volume around the landfill, which identifies each of the fol-
lowing four streams exiling this particular landfill:
• Migration out of the landfill into the subsurface off-site
• Gas removed via the active gas extraction vents
• Gas emitted from the passive venting plastic vents and concrete cistern
vents
• Gas emitted through the surface (soil cover) of the landfill
By estimating the gas production rate, P, and measuring streams 2,
3 and 4, one can determine the quantity of gas migrating into the sub-
surface off-site. However, it was determined in the study that because
an overall gas production rate and an overall surface emission rate are
both gross figures, from measured point samples, and due to the large
heterogeneity in trace gas compositions (of VOCs) at any point within
or at the surface of the landfill, the mass balance approach lacks the
accuracy and precision required for the RI.
RESULTS OF THE LANDFILL GAS INVESTIGATION
Landfill Gas Well Investigation and Sampling
Table 1 contains data on the concentrations of VOCs detected in the
LFG wells during three rounds of sampling and compares them with
the results from neighboring groundwater wells. Although it is not shown
on this table, only three (probes B, C, and D) of the four probes (A,
B, C. and D) from the U.S. EPA wells were sampled. The shallowest
probe (probe A) showed no observable variation from the barometric
pressure, and was assumed to be in good hydraulic connection with
the atmosphere. Generally, the medium (C) and deep (D) probes for
all the LFG wells consistently exhibited the greatest number of con-
taminants and the higher concentrations. However, some of the detec-
tions from the B probe exhibited the highest concentration for a
particular contaminant. Nine VOCs were identified as the major con-
taminants in the off-site, subsurface gas based upon their frequency of
detection and also their presence in the groundwater:
1,2-Dichloroethane
Trichloroethene
1,1,1-Trichloroethane
I.l-Dichloroethene
Tetrachloroethene
Chloroform
1.2-Dichloroethene (total cis/trans)
Vinyl Chloride
1,2-Dichloroethane
Except for chloroform, each of these contaminants also was present
254 CONTAMINATED GROUNDWATER CONTROL
-------
in the gas sampled from the gas extraction vents and the flux boxes
(Tables 2a/2b and 3). The presence of these contaminants in both the
subsurface gas off-site and in the gas collected directly from the land-
fill establishes that the off-site, subsurface gas contamination most-likely
originated from the landfill. Ideally, to further verify the correlation
between the off-site gas and the landfill, the relative concentrations
between the two should compare favorably. However, due to the hetero-
geneous nature of a landfill, the comparison is poor. In feet, the vents
within the landfill do not compare well with each other (Table 3).
Finally, there are no general trends in the lateral distribution of sub-
surface gas contaminant concentrations off-site. This probably is due
to the unsteady-state nature of this landfill. Each round of LFG well
sampling only represents a "snapshot" of the subsurface gas concen-
trations at the time of sampling. It is believed that the subsurface gas
concentrations are relatively dynamic due to the influence of varying
weather conditions (e.g., barometric pressure, temperature and rain-
fall) and due to the discontinuous operation of the gas extraction system.
Therefore, the measured LFG well concentrations may not be represen-
tative of any type of average condition, if one exists.
The soil samples collected during the installation of the LFG wells
indicated the presence in the groundwater of very few VOCs (TCE,
toluene and 2-butanone) each estimated at 10 mg/L or less (Table 4).
Based upon these results, there does not appear to be a significant degree
of partitioning to the soil phase.
Landfill Assessment
Results from the landfill assessment consisted of data and samples
collected during the flux box emission rate sampling, vent testing, and
condensate sampling events. The analytical results for the flux box and
condensate samples are shown in Tables 2a/2b and 5 respectively. The
system performance data recorded during the vent testing (for use in
the FS Alternatives analysis) are not necessarily pertinent to this paper
and therefore are not presented.
The primary purpose of the landfill assessment was to evaluate the
effectiveness of the landfill gas extraction system and determine whether
gases were migrating from the landfill to the subsurface off-site. At
the time of the vent testing program, the landfill gas extraction system
had fallen into disrepair. Numerous sections of the extraction manifold
piping were blocked by collected condensate (which caused surging)
the extraction blowers were operating below design capacity and the
entire system was operated 8-12 hrs out of every 24 hr day. Additionally,
methane was found in 1 of the LFG wells at concentrations up to 3 %
of the lower explosive limit (LEL) and in another well at up to 100%
of the LEL. Due to the fact that the vent testing could not be performed
Table 1
Correlation Between Detected Landfill Gas and
Groundwater Contamination
T OBSERVED CONCENTRATIONS
CONTAMINANT
1,1'Diehloroethane
Trich Ioroethene
1,1,1-TricMoroethBnc
1.1-Oichtoroethene
wthene
Chi
1,2-Diehloroether
Vinyl Chloride
1,2-Dlchloroethar
1,1-Oichloroethar
1rieh Ioroethene
1,1,1-Trichloroer
1,1-Dfchloroether
1,2-Dfchloi
Vinyl Chi01
1,2-Dfehloi
, -Dichloroethane
r ehloroethene
, ,1-TrichloroelhBne
, -Dichloroethene
e rachIoroethene
,2-Dichloroethene (total
inyl Chloride
,2-Oichloroethane
,1-Diehloroethane
nchl roethene
,1,1- r ehloroethane
,1-Of h oroethene
etrac I roethene
,2-Dl h oroethen.
inyl h cride
,2-Oi h oroethani
1,1-Dichloroetharx
Trichloroethene
1,1,1-Trichloroeth
1 1-Oichloroetherw
1,2-Oichloroethene (t(
Vinyl Chloride
1,2-Dichloroethane
, -D chloroethnne
r ch roethene
, ,1 richloroethane
, -D hloroethene
e ra loroethene
h or orn
,2-0 hloroethene (tot
inyl hloride
,2-Di hloroethane
:s: J - Estimated Value
••- Reported value fc
tions measured for both n
to its fullest extent (surging caused inaccurate flow and wellhead vacuum
measurements and the applied vacuum could not be varied) and the
detection of elevated levels of methane off-site verified the ineffective-
ness of the extraction system, in its current condition, the vent testing
program was terminated early. Interpretation of the results from this
abbreviated testing indicated that landfill gas was migrating from the
landfill to the subsurface off-site.
The flux box sampling results were only used for the Public Health
Table 2a
Flux Box Analytical Results (Round 2)
a
VhylCIWde
Cteroefim
tonne Fqi
t.l-OcNMMIhene
CWorotorm
1.2 ttertomeOufle
l.l.l TrtWwocOww
V«|4tacUM
bttnxKHoramctune
Wnynoehtoronwhmc
Beivonc
us U Ocrtofopropcne
Oionwto™
<-Urftri 2'Perumn*
1.1 .7 J Tc*iJJ»Jv*l>unt
CMonftouenc
TotXitone
U
U
FV
u
u
fv
u
! U !
"u
U
»- --
2 UJ 3
1 U 1
U .
U
u
n
it
FV
S- -
"
u
U !
"U
"
«
Ul !
U 1
jj
U
U I
U
. F*
FV
U
u
u
UJ FIJI
u
u"
u
«
jj
u
"
u
u
_ ."r
'V
u
u
u
F«
u
u
u
u
u
u
U F
"
u
U 10
FV
u
U X
u
F4
UJ
u
u
u
~u
UJ
jj
r
J
u
U 11 U MIU 1CQU IU 711
E 11
N
TI
u
3
u
u
Fq
UJ
u
u
.11
~u
UJ
' — f
u
»w
P"
1IGO
E m
I F
E F
ui e
U U E H
J MID E «
liOtt) Ł 01
U
ax
EH
UJ
u
1 U
JU
E F
Ul ?
UJ 7
UJ *
Ul
n »u u
n U * U u
p FV FV fV
„ FV f»
BU 1U
1 U
e u
A
9 U
» f
» J
M UJ
I U
n u
r u
0
u u
» u
1 UJ
1 Ul
ui n u
uj ns
u
u
FW
u
I
u
u
u
Ul
Ul
u
u
u
u
J
u
u
"
u u
fv fv
V U
u u
u u
U u
u u
u u
UJ U
Ul 11
U UJ
U UJ
u u
U U F
u u
u u
U 7U 7U IU
u u u u
u u
FV FV
u
u
u
Jj
u
u
UJ
UJ
Ul
u
u
u
u
:
u
u
u
FV
u
UJ
u
u
u u
u
u
UJ
UI
u
F
«
u
0
u
u
u u
Ul U
UJ u
UJ U
u u
i U
IU U
s u
CONTAMINATED GROUNDWATER CONTROL 255
-------
Table 2b
Flux Box Emission Rates (Round 2)
Evaluation (PHE). Due to the high variability in on-site gas concen-
trations and emission rates, and the generally poor precision of overall
surface emission rates and gas production rates, the mass balance
analysis was not useful and was therefore not performed.
Interrelationship Between Groundwater and Gas Concentrations
Comparing the observed subsurface vapor concentrations with the
observed groundwater concentrations shown in Table 1, we first con-
cluded, does not appear to be a direct relationship between the vapor
phase and groundwater concentrations of VOCs. However, if the data
are compared using the Henry's Law relationship," one sees that
much of the data are comparable, i.e., the calculated equilibrium water
concentrations are close to the observed groundwater concentrations.
Table 6 presents an example of this calculation.
Two notable exceptions to the close correlation are ground water wells
TNH-6 and TNH 10/9 (wells 10 and 9 are a nested pair of wells at
the same general location). These wells exhibited much higher con-
centrations than predicted according to the vapor phase concentrations.
The explanation of the anomalies at these two wells may be several-
fold. For instance, the ground water and vapor phase concentrations
of VOCs near TNH 9 and K) had been greater prior to the installation
of the gas migration control system and when the public supply well
was in operation (and drawing water from beneath the landfill). Since
that time, the concentration of VOCs in the water from these two wells
has been decreasing with lime. The current groundwaier concentra-
tions may then be a result of residual contamination.
However, this trend noted above does not explain the results from
TNH-6, which has remained relatively constant over time. In this in-
stance, two explanations have been proposed. First, TNH LFG-4 usually
does not measure representative subsurface gas concentrations due to
the well's close proximity to two gas extraction vents (as a result, when
the gas system is on, ambient air is drawn into the LFG monitoring
well, thereby diluting the subsurface vapor concentrations). Second,
it has been postulated that the consistently high concentrations at this
well may be a result of the indirect discharge of concentrated conden-
sate due to the change in temperature between the landfill and the
surrounding soils.
CONCLUSIONS AND RECOMMENDATIONS
The results of the field investigation into the various sources of con-
tamination from a municipal landfill not only provided a basis for com-
pleting the RI and FS. but also verified the applicability of and the
Table 3
Volatile Organic Analysis from On-Site Gas Vents
Actlv. v.ntt r.»l»<
Compound
Vinyl chlorid.
Chlorofora
n.thyl.n. chlorid.
1,1-dichloro.th.n.
1 , 1-dichloro.th.n.
1,2-dlchloro.th.n.
Tr.n. 1 , 2-dichloro.th.n.
BroaodichloroM.th.n.
Trlrhlaro.th.Hb
1 , 1 , 2 , 2-k.t r.chloro.th.n.
1,1, l-trlchloro.th»n.
T.tf >thlaio*th*H»
Carbon t.t r nchlorld.
lil , 2-trlchloro.th«n.
B.nfe.n.
Chlorob.nt.n.
V.nt
Ho . 101
IB0747I
il.S
l.SIU
1 . 46U
9.11
8.50
15. t
2J.5
1 55U
It 9
10.1
1.11U
67. J
J.OIU
4 51
14)
151
v.nt
Ho. IU
IB0742I
19. »
1 5IU
1 . 4tU
0.92U
• t»
1 .94
2.55
l.JSU
2. JO
11 . 10
1 )1U
B 69
1 .OIU
1.52U
It. )
12)
V.nt
Ho . lit 1 dup 1
IB0719)
J7.
1 .
1 .
0.
1.
1 .
2.
1.
1 .
I
1
5
1
1
77
11)
5
iou
17
7tU
7t
45
55
27U
It
. )2U
.OIU
.91
.71U
. 25U
.)
vont v.nt
Ho. 122 Ho. t
1107121 IB0744I
J.
1.
1.
0.
0.
0.
0.
1 .
1
29
1
1
2
1
1
3
14 142
5IU 1.1IU
4tU 1.4tU
t2U 20.7
IU 121
9IU t.U
I9U 1.2*
5JU 1.5SU
.27 11.2
.4 l.tu
.11U ll.f
.71 5).t
.010 2. OIU
. 52U 1.5JU
tt 127
.OtU It)
Concr.t.
Cl.t.rn
00*041
141
1 .
1.
7.
1).
1.
1*.
I
i
1
1
1*
2
1
SI
1
4IU
2*
(4
1
17
0
*su
.41
.IU
. llu
.2
.OIU
. S2U
.1
. OtU
Concr.t*
Cl.t.tn
Idiipl
(807941
141
1.
2
7
12
1.
If
1
t
1
1
11
2
1
49
(0
5IU
.95
.71
.t
.1*
.4
. ssu
.00
. tu
.11U
.4
.OIU
. 51U
.t
.1
• lomr
1*1. t
II07I4I
71.0
l.SIU
11.1
12. »
11.0
l.t)
24. S
1.55U
21. S
il.t
1.29
Sl.l
2. OIU
4.09
151
ito
HOTCS:
U > Und.t.ct.d >t th. d.t.ction linit thown.
billion on • volua. per voluu bam.
256 CONTAMINATED GROUNDWATER CONTROL
-------
Table 4
Analytical Results from Soil Samples Collected
During Installation of EPA LFG 201 to EPA LFG 204
Uett No.
EPA LFG 201
EPA LFG 202
EPA LFG 203
EPA LFG 204
Sampling Interval
B«(ow Grade (ft)
20 to 24
50 to 54
100 to 104
126 to 130
29 to 34
60 to 65
90 to 95
120 to 125
10.5 to 15.5
29.5 to 34.5
92 to 97
15 to 20
50 to 55
Ho volati le organ
Results
c compounds detected
No volatile organic compounds detected
Ho volatile organic conpounds detected
Ho volatf le organ
c conpounds detected
Toluene detected at a concentration of 2J
Toluene detected
it a concentration of 5J
Ho volatile organic conpounds detected
Tetrachloroethene detected at a concentration of 2J
Toluene detected
No volati Le organ
No volati le organ
No volati le organ
Ho volati le organ
Ho volati le organ
it a concentration of 2J
c conpounds det ted
c conpounds det ted
c conpounds det ted
c conpounds det ted
c conpounds det ted
^^
f. Moisture
Content
15
a
U
12
13
13
15
10
9
6
10
12
15
2
2
1
AU concentrations are reported in ug/kg.
J = estimated concentration.
success in using several new and innovative monitoring and sampling
techniques.
As for the RI, the following conclusions could be made based upon
the results of the investigation and upon historical data:
• Landfill generated gases containing trace constituents of VOCs have
historically migrated and continue to migrate from the landfill into
subsurface soils off-site
• The large degree of heterogeneity in landfill gas concentrations both
on- and off-site make it difficult to positively prove, by a "finger-
print" comparison of relative VOC constituent concentrations, that
off-site vapor contaminants originated from the landfill
• The presence of the same nine VOCs in samples of landfill gas, off-
site subsurface gas and groundwater, plus the absence of leachate
characteristics indicates that somehow VOC contaminants (possibly
vapor borne) are being transferred to the groundwater upgradient of
the landfill. Probable mechanisms for this interphase transport are
from the combined action of indirect condensate discharge (through
vapor condensation via a temperature change) and by vapor-to-liquid
partitioning of VOCs to infiltrating rainwater
• Except for several anomalies and despite the large heterogeneity in
both on- and off-site gas concentrations, comparison of groundwater
and gas data generally support the vapor/infiltrating rainwater parti-
tioning mechanism.
lableS
Results of Condensate Sampling
Sa»pl* t
Saapl* I.D.
Coapound
ChloroB*than*
Broa>oa*than*
Vinyl chlond*
Chlo rot than*
M*thyl*n* Chlorid*
Ac* ton*
Carbon Diaulfid*
1 , l-Dichloro*th*n*
1 , l~0ichloro*than*
l,2-Dichloro*th*n* (total)
1 , 2-oichloro*than«
2-Butanon*
1,1, l-Trichloro*than*
Carbon T*t rachlorid*
Vinyl Ac*tat*
BroModichloroia*than*
1 , 2-Dichloropropan*
cia-1 , 3-Dichloroprop*n*
Trichloro*th*n*
Dlbro*ochloro**than*
1,1 . 2-Trichloro*than*
B*nc*n*
trana-l , 3-Oichloroprop*n*
Broamf or*
4-H*thyl-2-p*ntanon*
2-H*xanon*
T*ttachloro*th*na
1,1.2, 2-T*trachloro*than«
Tolu*n*
Chlorob*nz«n*
Ethy lb*nz*n*
Styr*n*
Xylan* (total)
1
BR 821
ND
NO
ND
NO
ND
15,000
no
ND
ND
no
ND
ND
4,600 J
ND
ND
TQA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
2
BT 257
ND
ND
ND
ND
ND
2,900 J
ND
ND
ND
ND
ND
ND
1,600 J
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
)
BT 258
ND
ND
ND
ND
ND
ND
ND
ND
ND
24
ND
ND
NJA
ND
ND
ND
ND
ND
ND
ND
ND
ND
5
ND
ND
35
11 J
ND
ND
12
5
1!
ND
16
n«ld Blank
BT 259
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
fQA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
4
BT 2(0
ND
ND
ND
ND
ND
560
ND
ND
ND
21 J
ND
ND
87 J
ND
ND
fQA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
17 J
ND
ND
ND
ND
Trip Blank
BT 261
ND
ND
ND
ND
2 J
ND
ND
ND
ND
ND
4 J
ND
TQA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
5
BT 262
ND
HD
ND
ND
ND
630
ND
ND
ND
47
ND
ND
200 J
ND
ND
r
ND
ND
ND
ND
RD
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
BT 266
NO
ND
HD
ND
ND
12,000
ND
ND
ND
ND
ND
ND
3,200 J
ND
ND
FQA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND Not D*t.ct*d.
TQA rall.d Quality Aaauranc*.
J Estl«at*d Valu*.
BT 264 ll a duplicate »a»pl* of BT 261
BT 266 ia a duplicate aaapl* of BT 265
All concentrations ar* in (uqyl)
CONTAMINATED GROUNDWATER CONTROL 257
-------
•Able 6
Calculation of Equilibrium Water Concentrations
1
\
1
1
COHMIBU1 |
MWM1I.VW
•TMIM**
> KJH *'•*(*«•
Mf**l»m
'•"•"•-
.I-»UMBr*«(h»
CMOTI*
•*»,-. IM 1
C«,-M (I)) |
•l/l 1
1
••I/I 1
«,'•» 1
a in
e m
•-HI
om
G '*!
» til
I»H
•
»*0
U'«
in*
i
j
•
w
,
1
M
H
9*
in
in
MO
toe
C*Uvl*<*4 |
i*.mt |
W*l*r CWV
J*
' U
1 lt«
M
1 r*>
1 ^
I
! i
t tOUli
1
|in '
i ' 'i
i
t >i »
t i ••
i i t.
MO*
Mil*
l«
AW
.,
tt
*t
W
'!
H
» am
f»* IN
• 1 «
' -1
t it
• *
' ••
!H
Regarding sampling and monitoring techniques the following were
successful applications of innovative technologies:
• Split-spoon liners were successfully employed to collect undisturbed
soil samples for chemical analysis
• SUMMA canisters were demonstrated to be a viable technique for
the collection of subsurface gas and landfill gas samples
• Nested probe design landfill gas monitoring wells were successfully
used to monitor the migration of subsurface gas from a landfill
Further areas of research in the field of subsurface and landfill gas
migration should include numerical modeling of migration in order to
better predict the impact of subsurface gas sources, such as landfills.
on surrounding areas.
REFERENCES
1. Chion. C.T. Sharp. T.D "Soil Sorption of Organic Vapor; and Effects of
Humidity on Sorptive Mechanism and Capacity." Environ. Sci Tech., 19
(12). pp. U96-1200. 1985.
2. Peicwon, M.S.. Lion, L.W.. and Shoemaker. C.A. "Influence of Vfopor-
Pha&e Sorption and Diffusion on the Fate of Trichloroethylene in an
Unsaturaled Aquifer System," Environ. Sci. Tech., 22. (5), pp. 571-577, 1988.
3. Shusicr, K.A., "Environmental Issues: Condensate," Proc. of the GRCDA
7th International Landfill Gas Symposium, pp. 139-142, GRCDA. Silver
Spring, MD.
4. Oliver. K.D . Pleil, J.D., and McClenny. W.A. "Sample Integrity of Trace
Level Volatile Organic Compounds in Ambient Air Stored in SUMMA
Polished Canisters," Atmos. Envion 20:p. 1403. 1986.
5. Holdrcn, M.W. and Smith. D L. "Stability of Volatile Organic compounds
while stored in SUMMA Polished canisters." Final Report. US. EPA Con-
tract No. 6^-02-4127, Research Triangle Park. NŁ.. Batielle Columbus
Laboratories. Columbus, OH. Jan.. 1986.
6 McClenny. W.A., Pleil. J.D.. Holdren, J.W.. and Smith. R.N. "Automated
Cryogenic Preconcenlralion and Gas Chnomalographic Determination of
Volatile Organic Compounds," Analytical Chemistry, 56 (14), 1984.
7. Day ion. D.P.. and Rice. J., "Development and Evaluation of a Prototype
Analytical System for Measuring Air Toxics," Final Krport. Radian Corp.
for the U.S. EPA. EMSL. Research Triangle Park. NŁ., Nov. 1987.
8. Lofy. R ). "Study of Zones of Vacuum Influence Surrounding Landfill Gas
Extraction Wfclls." Argame National Laboratory Rpt. No. ANUCNSVTM-U3,
Argonne National Lab.. Argonne, III,, March 1982.
9. Kunz. CO and Lu. AH "Methane Production Rate Studies and Gas Flow
Modeling for the Fresh Kills Landfill." Interim Report. N.Y. Slate Energy
Research and Development Authority Rpt. No. 80-21. N.Y ERDA. Albany,
N.Y. Nov. 1980.
K> Klenbusch, M.R. "Measuremenl of Gaseous Emission Rales from Land
Surfaces Using an Emission Isolation Flux Chamber." U.S. EPA Rpt No.
EPA/6OO/8-8&008, U.S. EPA. EMSL. Las Vegas. Nevada. Feb. 1986.
U. Lyrwynshyn. G.R., etal.. "Landfill Methane Recovery Pan II: Gas Charac-
terization," Final Report, Argonne National Laboratory Rpt. No. ANLJCNSV-
TM-IIS. Argonne National Lab., Argonne, 111.. Dec. 1982.
12. Gossett. J M. Measurement of Henry's Law Constants for Cl and C2 Chlo-
rinated Hydrocarbons. Environ. Sci. and Tech.. 21. No. 2. 1987.
258 CONAMINATED GROUNDWATER CONTROL
-------
Use of Hydraulic Controls in an Aquifer Impacted
By Tidal Forces
David Tetta
U.S. EPA
Seattle, Washington
Brian Y. Kanehiro
Berkeley Hydrotechnique, Inc.
Berkeley, California
Paul D. Fahrenthold, Ph.D.
Fahrenthold and Associates, Inc.
Berkeley, California
ABSTRACT
The Wyckoff Company facility on Bainbridge Island occupies
approximately 40 ac at the mouth of Eagle Harbor adjacent to Puget
Sound in the state of Washington. The facility, which currently is used
for log peeling, storage and shipping, had been used as a wood
preserving and treating plant since the early 1900s. Contamination of
soils and groundwater exist in the operations area comprising approx-
imately 15 ac of the facility. Groundwater contamination exists in
immiscible and miscible phases.
In 1987-1988, the U.S. EPA and Tetra Tech evaluated alternatives for
dealing with identified contamination as part of an expedited response
action. As part of that evaluation, a hydraulic barrier well system along
the perimeter of the site was identified as the preferred means of
addressing seepage of product and contaminated groundwater into Puget
Sound.
The site is particularly complex because of the multi-phase nature
of contaminants and very large tidal influences. Four pump tests,
including monitoring of tidal responses, were performed to develop a
better understanding of the site. Additional analysis based upon the
results of the pump tests suggested that the hydraulic barrier well system
would not be a viable option for the site. The proposed option is
hydraulic control in the form of pumping from the central portion of
the site where the majority of the contaminants occur and more modest
pumping in shoreline areas where product is known to leave the site.
INTRODUCTION
The Wyckoff Company facility on Bainbridge Island occupies ap-
proximately 40 ac at the mouth of Eagle Harbor, adjacent to Puget Sound
(Fig 1). The average ground surface elevation is approximately 10 ft
above mean sea level (MSL). The ground surface is composed primarily
of permeable fill, with some paved surfaces over operational areas of
the site. The facility borders approximately 0.8 mi of shoreline along
its eastern and northern edges. A bluff at the southern boundary of
the facility ascends toward the island interior to an elevation exceeding
200 ft.12.
The area of concern at the facility has been identified as the opera-
tions area which occupies approximately 15 ac in the northern portion
of the facility. This 15 ac area of concern is referred to as the Wyckoff
site.
Operations at the Wyckoff site included aromatic oil and creosote
unloading and storage, chemical storage, wastewater treatment, untreated
pole and pile storage, log rafting, log peeling, wood preserving, treated
wood storage and shipping. The site currently is used for log peeling,
storage and shipping.
The site had been used for wood-treating since about 19105. The
original wood-treating operation was constructed on a small peninsula
tto iiiiiiiiiiiiim"""" ("
aim "" PEBlMeTCROITCH \
Figure 1
Site Map of the Wyckoff Bainbridge Island Facility'-1
formed by longshore currents in Puget Sound that pass across the mouth
of Eagle Harbor. The area of the harbor between the peninsula and
the shoreline of Bainbridge Island formed a cove which, before it was
filled in the 1920s, was used as a untreated and treated log storage and
shipping area.
Prior to 1929, the eastern and northern shoreline of the peninsula
(facing Puget Sound) were protected from tides and wave erosion by
a bulkhead located inshore of the present bulkhead. The site has
undergone at least two major changes, once in the 1920s and again in
the 1940s.6. A significant amount of fill has been added to the site,
extending the shoreline into Eagle Harbor and Puget Sound.
Contamination at the site includes soil contamination, buried sludges
and groundwater contamination. Chemicals of concern include
Polyaromatic Hydrocarbons (PAH) and chlorinated phenols. Ground-
water contamination exists in three phases at the site. A floating pro-
duct phase exists with total PAH concentrations ranging from 6.9 to
36%. Groundwater at the site also shows soluble PAH concentrations
ranging from less than 80 to 166,000 ug/L. Finally, sinking product
identified in several wells at the site has shown total PAH concentra-
tions ranging from 0.5 to 50%u. Petroleum products seepage into
Puget Sound from this site has been observed for at least 25 yr12.
Expedited Response Action
In 1987-1988, the U.S. EPA and Tetra Tech conducted an evaluation
of alternatives for dealing with identified contamination at the site as
part of an expedited response action. As part of that evaluation, a
CONTAMINATED GROUNDWATER CONTROL 259
-------
hydraulic barrier consisting of six wells spaced along the perimeter of
the site was identified as the preferred means of addressing the seepage
of product and contaminated groundwater into Puget Sound. The U.S.
EPA and Wyckoff Company signed an Administrative Order on Con-
sent in 1988 to perform this response action.
Prior to implementation, the practicality and effectiveness of a
hydraulic barrier well system was re-evaluated. Among the major con-
cerns were the estimated material properties for the site and the large
tidal effects. The tidal water level fluctuation in Puget Sound can exceed
15 ft.
These concerns were initially addressed by numerical modeling and
sensitivity analysis by Wyckoffs consultant, Hydrotechnique. The
preliminary analyses indicated that the existing data were not sufficient
to provide reasonable assurance of the effectiveness of a hydraulic barrier
well system. Accordingly, four pump tests and additional analyses of
a hydraulic barrier well system were performed. The better estimates
of material properties derived from the pumping test and the results
of the additional analyses indicated that the hydraulic barrier was not
a viable option for the Eagle Harbor site.
This paper includes a discussion of the analysis of the four pump
tests that were run at the site, analysis of the hydraulic barrier wells
based upon the results of the pump tests and a discussion of proposed
strategy for the site.
PUMP TESTS
The major factors complicating the performance and analysis of the
pump tests were the large tidal effects and proximity of the ocean. The
fluctuation in water levels due to tidal responses in the near-shore areas
can actually be larger than the drawdowns that can be achieved by
reasonable pumping rates in testing wells. The distances from the areas
of interest to the shoreline are also comparatively small and the tests
were therefore susceptible to irregular boundary conditions.
Methods
Four pumping tests were performed to estimate properties. Each test
consisted of pumping out of one 8 in diameter pumping well and
observing drawdowns in the pumping well and two to four 2-in diameter
observation wells. The wells were 39 ft deep with screening from
approximately 5 to 35 ft. Test durations ranged from 5 to 24 hr.
Tidal corrections were applied to both pre-lest pumping and the long-
term pumping tests at each of the four test locations. V&ter levels of
the ocean and in available wells were monitored for a period of
approximately 24 hr prior to pumping. Tidal correction factors based
upon attenuation of the amplitude of water level fluctuation and phase
Predicted Tidal Response
000 0*0 080
0 PW-2 (*neo»u'etj)
I 20 I 60
lime (mlnulei)
lags were determined from these data. The corrected data for each well
were obtained by subtracting the estimated tidal influence from the
uncorrected data for each well. Comparison of the estimated tidal in-
fluence at the well with the actual recorded water level in lest well PW-2,
shown in Figure 2, indicates good agreement.
Analysis of Pumping lest Data
Estimation of hydrologic properties for each test was performed on
the basis of an unconfined model. Selection of the unconfmed model
was based upon comparison of data to unconfined, confined and leaky
aquifer solutions. An analytic solution for analysis of pumping test data
from an unconfined aquifer developed by Neuman" was used for the
analysis. The major assumptions include:
• The fluid is isothermal and single phase wiih constant viscosity and
density
• The aquifer is homogeneous and its principal directions of hydraulic
conductivity are oriented horizontally and vertically
• The aquifer overlies an impermeable horizontal layer and is of infinite
lateral extent
• The well fully penetrates the aquifer and is of infinitesimal radius
(i.e. no wellborc storage effects)
• No skin effects
• The initial drawdown in the aquifer is zero and a constant flow of
rate Q is imposed at the wellbore at time t = 0
An inverse fitting technique was used to analyze the data from the
four Eagle Harbor pumping tests. The governing equation was evaluated
numerically with standard numerical integration schemes".
Analysis of the long-term pumping test data suggests that flow in the
areas of the four tests occurs under unconfined conditions. This con-
clusion is consistent with observed geohydrologk conditions at the site.
An example of drawdown versus time for an observation well associated
with pumping well PW-1 located in the central portion of the site is
shown in Figure 3. The horizontal hydraulic conductivities for the four
tests ranged from 4 x 1O5 m/sec to 2 x IO"1 m/sec, respectively, ver-
tical hydraulic conductivities ranged from 5 x K>' m/sec to 2 x 10*
m/sec. The storativities estimated ranged from 2 x 10' to 8 x XT'. The
specific yields for the four tests range from approximately 0.1 to 0.3.
Some uncertainty is associated with the specific yields because they
were derived from the later portions of the various tests where the ocean
boundary condition may be significant.
100 -3
O
Q
<
ct:
Q
10 i
1 -
0.1 -
0.01
0.01
0.1 1 , 10 v
TIME (min)
100
1000
Figure 2
Observed and Predicted Water Level versus Time Data for
Eagle Harbor and Well PW-2"
Figure 3
Drawdown versus Time for Observation Well OB1-1. Pumping Test 1
(Tidal Corrections Applied)1.
ANALYSIS OF HYDRAULIC BARRIER WELL SYSTEM
The preliminary sensitivity analyses of the hydraulic barrier well
system performed before the pumping tests, involved a numerical model
260 CONTAMINATED GROUNDWATER CONTROL
-------
with pumping wells and time-varying boundary conditions. A simpler
approximate model was used in the analyses described here. The
numerical model was not used because the preliminary analyses sug-
gested that the hydraulic barrier would not be viable for the types of
material properties derived from the pumping tester the numerical
modeling suggested that the approximate model provided reasonable
accuracy and the results of the approximate model would not be used
for design of a hydraulic barrier well system.
The analytical model used for the analysis is based on combining
two sub-models, namely a pumping well model and an ocean tide model.
The pumping well model is used to analyze distributions of ground-
water levels due to pumping. The ocean tide model is used to analyze
distributions of groundwater levels resulting from response to the ocean
tides. Because the governing equation and boundary conditions for
groundwater flow in an unconfined aquifer can be approximated as
linear9, the distributions of groundwater level at the site can be ap-
proximated by combining of groundwater levels predicted by the
pumping well and ocean tide models.
To illustrate the performance of a hydraulic barrier well system, a
pumping well located in the areas of Milwaukee Dock, shown in
Figure 4, was considered where shoreline seeps are visible during the
outgoing tide. The distances from the pumping well to the eastern,
western and northern shorelines are approximately 30.5 m, 244 m and
213 m respectively. Because the eastern shore is much closer to the
pumping well than the northern and western shores and the tidal
response drops off exponentially with distance, the influences of the
northern and western boundaries were ignored.
Y
PUGET SOUND
Figure 4
Location of Cross-section AA for Hydraulic Barrier Well System Analysis.
Note that subsequent figures illustrating the models and
results are presented with A on the left and A' on the right
(i.e., AA' as opposed to AA)3.
Pumping Well Model
In order to predict drawdowns in such a system, the method of
images' was adopted. An imaginary infinite system, shown in
Figure 5, was used. The system includes one real pumping well and
one imaginary injection well. The distributions of water levels resulting
from pumping and injection can be predicted by the unconfined aquifer
solution presented by Neuman'. Tidal effects were considered by a
tidal response model. The material properties used in the analysis were
those derived from pumping tests.
Ocean Tide Model
In general, in an unconfined aquifer, fluctuation in groundwater levels
in response to the ocean tide decreases with distance inland from the
shoreline. Analysis of tidal influence on groundwater flow in the areas
near the shoreline was approximated by solving a one-dimensional flow
in a semi-infinite domain with a sinusoidal approximation of the ocean
tide. The governing equation for the one-dimensional problem can be
expressed as:
Injection Well
C-Q)
OCEAN
Piezometer
.O
Pumping Well
(Q)
LAND
X
Figure 5
Schematic Representation of the Method of Images for the
Pumping Well Model2
ajH = _S_ 3H a)
3x2 T 3t
where
H = water level with reference to the mean sea level
x = distance inland from shoreline
T = transmissivity; (K=rb)
S = effective storage coefficient
t = time
To solve the above equation, it is assumed that the initial water level
of the aquifer is zero (mean sea level) and that boundary conditions
include H = H0 sin 2 7rt/t0 at x = 0 (shoreline) and H = 0 at x =
ala, where t0 is the tidal period. The steady periodic condition at the
site was expressed as:4-13
H = H0 exp [-x (irS/tJ*] sin [2irt/t0 - x^S/tJ)"4] ^
In Equation 2, H0 represents the amplitude of the ocean tide for the
major tidal period, t0 and T/S represents the average hydraulic dif-
fusivity of the aquifer. To application of Equation 3 to the Wyckoff site,
the value of T derived from the long-term pumping tests was used. S
is derived by solving Equation 2 for S with field observations of tidal
response to a 12 hr tide. Through this calibration process, it was found
that the effective storativity S is on the order of 0.02.
The solution for the ocean tide is one-dimensional and for a confin-
ed aquifer situation. It was selected for simplicity. The approximation
with the large storativity is stated by Toddu to be reasonably good for
an unconfined situation and the overall solution derived from combining
the pumping and tidal influences to corresponded reasonably well with
the results of the numerical model used in the preliminary sensitivity
studies for the hydraulic barrier well system. The large storativity which
was derived from calibration with field data can be viewed in a sense
as a lumping of storativity and specific yield.
CONTAMINATED GROUNDWATER CONTROL 261
-------
Figure 6 shows the range of groundwater fluctuation versus distance
from shoreline predicted by Equation 2 with T = 1.83 x 10J mVsec,
S = 0.02, t0 = 12 hr and H0 = 1.68 m. Groundwater fluctuations
shown in Figure 6 are reasonably consistent with those observed at the
site, indicating that the model and the parameters used are appropriate
for the site.
6-
Ł 5-
4-
g 3-
< 2-
I— 1 -
UJ-3-
o
2-4-
(T -5-
-6-
cldcl period - 13 houn
uplltudl of oca«n tld« . 1 M • () 1 (II
triniBUilvltjr - 1,1) II 10'* •'/••<
•ff«cctva •taratlvlty - 0 03
b so 100
DISTANCE
so"
Tso
200 250 300 350 400 450 500
FROM SHORELINE (ft)
Figure 6
Range of Groundwater Level Response to Ocean Tides'
pumping race - 40 gpm
Kr - 2 X 10'4 n/sec
Kv - 2 X 10'6 m/sec
storacivity - 2 X 10"3
specific yield - 0.1
b - 9.15 m (30 ft)
pumping period - 6 hours
tidal period - 12 hours
amplitude of ocean tide - 1.68 ID
transraissivity - 1.83 X 10"3 ra2/sec
effective storativicy - 0.02
Figure 7
Contour of Groundwater Levels at Low Tide Predicted by
Combination of Pumping Well and Ocean Tide Models
(Pumping of 40 gpm from a well, Case I)3
Superposition of Models
Evaluation of a hydraulic barrier well system was made by combining
of the pumping well and ocean tide models presented in the previous
sections. For a scheduled pumping of 40 and 100 gpm from a well in
the areas of Milwaukee Dock, distributions of groundwater level for
low tide in these areas are shown in Figures 7 and 8, respectively.
Figure 7 shows that the pumping rate of 40 gpm would not be suffi-
cient to create a hydraulic barrier to movement of groundwaier flow
from the site to Puget Sound during the low tide period. At 100 gpm
(Fig 8), the gradient between the well and the ocean is nearly revers-
ed. The drawdowns illustrated in Figure 7 were in reasonable agree-
ment with drawdowns observed during the pumping tests.
\
\
I. / . / . /
pumping race - 100 gpra
K - 2 X
m/sec
Kv - 2 X 10'6 m/sec
storativity - 2 X 10~3
specific yield - 0.1
b - 9.15 m (30 fc)
pumping period - 6 hours
tidal period - 12 hours
amplitude of ocean tide - 1.68 m
transmissivity - 1.83 X 10"3 m^/sec
effective storacivity - 0.02
Figure 8
Contour of Groundwater Levels at Low Tide Predicted by
Combination of Pumping Well and Ocean Tide Models
(Pumping of lOOgpm from a weU, Case 2)'
Evaluation of these results indicated that the well spacing required
to form a continuous hydraulic barrier is approximately IS m and 30 m.
For the present purposes, a well spacing of 23 m is adopted. Because
the perimeter 30 m inland from the shoreline of the site is on the order
of 550 m, the number of wells required for the barrier system is
approximately 24 wells, resulting in a very large total pumping rate
of approximately 1,200 gpm (average over the full tidal cycle).
After reviewing this analysis and considering other potential problems
discussed in the following section, the hydraulic barrier approach was
abandoned in favor of the hydraulic control approach discussed below.
HYDRAULIC CONTROL STRATEGY
The major complexities at the Eagle Harbor Site are the very large
tidal influences, multi-phase nature of the contaminants, existence of
262 CONAMINATED GROljNDWATER CONTROL
-------
multi-phase contaminants off-shore and nature of the expedited response
action. As in the case of the hydraulic barrier well system, these fac-
tors suggest that immediate cessation of seepage off-shore or contain-
ment probably can not be practically achieved by implementation of
reasonable physical barriers.
Because of their complex nature and lack of data, direct analyses
of movement of contaminants at the site were not performed. Accurate
estimates of movement of the floating, sinking and soluble phases in
the heterogeneous, variably saturated media at the site would require
a multi-phase numerical model. The lack of data such as relative con-
ductivity and capillary pressure curves and 'retardation1 data for various
contaminants suggests that such an undertaking would not be justified.
It is known, however, that product, especially floating product, is
significantly more abundant in the central portion of the site and that
contaminated groundwater occurs in most portions of the site, but with
higher concentrations in the central area and in the area of the old sump.
There are three considerations: stopping off-shore seepage, removal of
contaminants from the central portion of the site and prevention of move-
ment of contaminants from the central area to the near-shore areas of
the site.
It is likely that the central area is ultimately the major contributor
to off-shore seepage. The central area likely remains significantly more
contaminated than the near-shore areas not only because of the initial
locations of spills, but also because net movement seaward associated
with the regional gradient is far smaller that the cyclic landward-seaward
movements of groundwater associated with the tides. Near-shore
activities directed at immediate containment and stoppage of off-shore
seeps will not address the overall problem and in the case of a very
high pumping rate hydraulic barrier well system could potentially worsen
the problem by accelerating the movement of contaminants from the
central to near-shore portions of the site.
Given the logistic as well as technical problems associated with the
hydraulic barrier, attention was directed at more modest near-shore pum-
ping and direct control of the contaminated central portions of the site.
The strategy that is described below is being developed in cooperation
with Wyckoff s consultants (Hydrotechnique and Fahrenthold and
Associates), the U.S. EPA , the Washington State Department of Ecology
and the U.S. EPA 's consultant (CH2M Hill).
There are two basic objectives. In the central portion of the site, the
objective is control, in the sense of removal of contaminants and
minimization of migration to near-shore areas. In the near-shore areas,
the objective is to intercept product and soluble contaminants that are
leaving the site because of the regional gradient and tidal influences.
It would appear that the central portion of the site is more important
to the long term solution, but that timeliness dictates that a large, im-
mediate effort be directed at the near-shore areas.
There is general concern about movement of immiscible phases from
more contaminated to less contaminated areas both horizontally and
vertically. Movement of oil from a contaminated area through a clean
area can result in residual saturation of oil in the originally clean area
that will be very difficult to remove. The large drawdowns associated
with very high pumping rates have the potential to pull a floating layer
of product down through or laterally across a comparatively clean
medium.
This finding suggests that contaminates should be attacked at their
sources, but that initial pumping rates should be modest and caution
should be exercised in near-shore areas. Given the uncertainties
associated with movement, particularly of immiscible phases, a phas-
ed startup of pumping in the central portion and near-shore portions
of the site is proposed. The first phase will include relatively low
pumping rates from each of three wells located in the central portion
of the site. Well startup will be staggered so that product thicknesses
and concentrations of soluble contaminants in pumping and observa-
tion wells can be monitored. This monitoring will provide a better
understanding of the movement of contaminants.
Because of the importance attached to stopping off-shore seepage,
it is proposed that pumping in the near-shore areas be started as soon
as the information derived from pumping the central portion suggests
that such pumping may be safely initiated. The proximity of the ocean
will result in smaller drawdowns in the near-shore areas and in effi-
ciency because most of the water ultimately will originate from the
ocean. This has a negative impact on the area from which contaminants
may be drawn but, on the positive side, lessens the possibility of drawing
contaminants from the central portion of the site.
As in the case of the wells in the central portion of the site, the start-
up of wells in the near-shore area will be staggered and start at a com-
paratively low pumping rate. Given the importance attached to off-shore
seepage and the expected small drawdowns, however, it is expected that
pumping in the near-shore areas may be increased more rapidly than
in the central portion of the site. Subject to findings during startup,
the total pumping rates in the near-shore areas may exceed those in
the central portion of the site. The near-shore extraction system will
include four wells whose locations are based primarily upon areas where
off-shore seepage is known to occur. The first well is northeast of the
transfer pit. The second well is east of the retorts, and the third well
is in the area of the Milwaukee Dock. An additional well is being add-
ed next to a recently installed monitoring well near the Milwaukee Dock
where the most visible seeps are observed. The area is also farthest
from the central area contamination and will be started first.
It is expected that the early phases of operation of the near-shore
wells will provide a better understanding of movement of contaminants
in this area of high tidal influence. This information will allow for final
adjustment of pumping rates in these areas.
CONCLUSIONS
The large tidal effects and multi-phase nature of the contaminants
at the Eagle Harbor site make it difficult to develop a definitive remedia-
tion strategy. Preliminary analyses suggest that direct containment by
a hydraulic barrier well system is not a viable option for the site. The
proposed source control and more modest pumping in the near-shore
areas is believed to be a practical alternative. There is little question
that such source control will, as a minimum, have a positive impact
on conditions at the site. Monitoring early phases of pumping will help
to provide the additional information that is necessary to more fully
evaluate the system. Final analyses on the design of the system are
presently underway.
REFERENCES
1. Bear, J., Dynamics of Fluids in Porous Media, Elsevier Publishing Co., New
York, NY, 1972.
2. Berkeley Hydrotechnique, Inc., "Final Pump Test Report for Eagle Harbor
Site—Analyses," Jan. 24, 1989.
3. Berkeley Hydrotechnique, Inc., "Groundwater Extraction System for the
Wyckoff Eagle Harbor Site," Draft, June, 1989.
4. Carslaw, H.S. and Jaeger J.C., Conduction of Heat in Solids, 2nd Edition,
Oxford University Press, 1959.
5. Entrix, Inc., "Data Report for the RCRA 3013 Investigation," Prepared for
the Wyckoff Company, Eagle Harbor, 1986.
6. Joy, J., Memorandum: Eagle Harbor Facilities Tour and Historical Review.
Washington Department of Ecology, Sept. 4, 1984.
7. Neuman, S.P., "Theory of Flow in Unconfined Aquifers Considering Delayed
Response of the Water Table," Water Resources Res., 8, pp. 1031-1045, 1972.
8. Neuman, S.P., Supplementary Comments on "Theory of Flow in Uncon-
fined Aquifers Considering Delayed Response of the Water Table," Water
Resources Res., 9(4), 1102, 1973.
9. Neuman, S.P., "Analysis of Pumping Test Data from Anisotropic Uncon-
fined Aquifers Considering Delayed Gravity Response," Water Resources
Res., 11, pp. 329-342, 1975.
10. Press, W.H., Flannery, B.P., Teukolsky, S.A. and Vetterling, W.T., Numerical
Recipes, the Art of Scientific Computing. Cambridge University Press Cam-
bridge, 1986.
11. Struck, R.G. and Adams, M.A., "Final Pump Test Report Wyckoff Com-
pany Eagle Harbor Site - Data," Jan 24, 1989.
12. Tetra Tech, 'Assessment of Expedited Response Actions, Wyckoff Company-
Bainbridge Island," Feb 2, 1988.
13. Todd, O.K., Groundwater Hydrology, John Wiley & Sons New York
NY.1959.
14. U.S. EPA, Consent Order No. 1088-02-17-106, in the matter of Wyckoff
Company, Eagle Harbor Facility, July 29, 1988.
CONTAMINATED GROUNDWATER CONTROL 263
-------
A Review Of ULTROX®
Ultraviolet Oxidation Technology
As Applied To Industrial Groundwater,
Wastewater And SUPERFUND Sites
Jerome T. Barich
Jack D. Zeff
Ultrox International
Santa Ana, California
INTRODUCTION
Ultraviolet oxidation technology has had limited exposure to the
various engineering disciplines despite having been commercialized
almost 9 yr ago with the installation of an ULTROX® system at IBM
in Boulder. Colorado. The application of this technology is steadily
expanding, however, as it offers a means of solving many of the problems
created by the toxic, water soluble organic chemicals that are found
in groundwater, wastewaters, leachate and drinking water supplies.
More conventional or better known unit processes and operations
such as liquid/solids separation, reverse osmosis, air stripping, biotrcat-
ment or granular activated carbon can remove many toxic organics from
water. However, these methods may solve the one problem only to create
a problem in another medium. Air stripping removes VOCs from water
only to discharge them into the ambient air; reverse osmosis generates
a reject stream of concentrated contaminants that must be dealt with;
granular activated carbon requires either regeneration or burial; and
liquid/solids separation obviously creates sludges requiring disposal.
Therefore, it is significant that UV/oxidation, when used in tandem
with some of the above mentioned process. , or as a stand alone treat-
ment process, can effectively destroy or render non-toxic many of the
organic chemicals found on the priority pollutant list.
Chemical oxidation without ultraviolet enhancement has been used
in water treatment for a number of years. Potassium permanganate,
chlorine and chlorine dioxide also have been used to treat solutions
containing organics such as phenol, and hydrogen peroxide with a
catalyst such as Fenian's Reagent has been used to oxidize phenol.
There is a need for more powerful oxidizing methods which do not
produce hazardous by-products. This paper describes the experience
of Ultrox International in applying ultraviolet/oxidation for the destruc-
tion of organic chemicals in wastewaters, drinking waters, leachates
and groundwaters. The oxidants used in these applications are ozone
(Oj) and hydrogen peroxide (H,O,). Ultrox International was issued a
process patent in 1988 covering the application of UV light, ozone and
hydrogen peroxide to a broad range of organic compounds. Excerpts
from the company's patent application arc shown in Table 1.
DESCRIPTION OF THE UV-OXIDATION PROCESS
Ultraviolet light, when combined with O, and/or H3O produces a
highly oxidative environment significantly more destructive than that
created with O3 or H,O; by themselves or in combination. UV light
significantly enhances ozone or H;O2 reactivity by:
• Transformation of O, or HO to highly reactive (OH) radicals
• Excitation of the target organic solute to a higher energy level
• Initial attack of the target organic by UV light
Table. 1
Oxidation Results for Several Systems Treaty
Methylene Chloride and Methanol
15
15
CQIfTBQl.
100
100
100
IH
100
5»
42
100
4t
17
100
11
11
100
it
i*
loo
It
T.i
0
10
CQHTHQL
75
73
73
75
75
75
nv/o,
75
II
73
I.I
COKCZNTRATIOHS -
The importance of the conversion of the ozone or H,O2 to (OH) can
be more easily understood after studying the relative oxidation power
of different chemicals.
Species
Fluorine
Hydroxyl Radical
Atomic Oxygen
Ozone
Chlorine Dioxide
Hydrogen Peroxide
Perhydroxyl Radicals
Hypochlorous Acid
Chlorine
• Based on chlorine as reference (=1.00)
Oxidation
Potential
tblts
3.06
2.80
2.42
2.07
1.%
1.77
1.70
1.49
1.36
Relative
Oxidation
Power*
2.25
2.05
1.78
1.52
1.44
1.30
1.25
I.K)
1.00
The design of the equipment in the ULTROX* system is based on the use
of components which arc highly reliable and require very little maintenance.
These systems operate either in a continuous flow or batch mode. They utilize
high efficiency LTV lamps with a long life and a micro-processor to control and
automate the process.
The ULTROX* UV-oxidation system consists of a UV-oxidation reactor and
tin oxidation source—an ozone generator with an air preparation system and/or
a hydrogen peroxide feed system. Reactor volumes range from 75 to 5,000 gal
treating up to 250 gpm.
264 CONTAMINATED GROUNDWATER CONTROL
-------
The reactor is fabricated from stainless steel. The UV lamps are
enclosed within quartz tubes for easy replacement and are mounted
vertically within the reactor. Depending upon the size of the reactor
and the type of water to be treated, the reactor can have four to eight
stages. Lamps are installed either in all stages or in designated stages,
depending upon the type of treatment specified.
When ozone is used as the oxidant, it is introduced at the base of
the stage. The ozone is dispersed through porous stainless steel diffusers.
The number of diffusers needed will depend upon the type of organics
being oxidized and the degree of removal required. If hydrogen peroxide
is substituted for ozone, it is directly metered into the influent line to
the reactor.
Within the reactor, the water flows from stage to stage in a sinusoi-
dal path using gravity flow. When the reactor uses ozone, the residual
ozone in the off-gas is decomposed back to oxygen by the use of a fixed-
bed catalytic unit operating at 150 °F (66 °C). The off-gas is then vent-
ed to the atmosphere with less than 0.1 ppmw O3 (OSHA Standards).
APPLICATION OF UV-OXIDATION TO VARIOUS WATERS
The UV-oxidation equipment developed in the past few years can be
used for a wide variety of waters. Table 2 shows compounds found in
groundwaters and wastewaters that have been successfully treated with
UV/oxidation.
Table 2
Common Industrial Effluents and Groundwater Contaminants
Amines
Analine
Benzene
BIS (2-Chloroethylether)
Chlorinated Solvents
Chlorobenzene
Complex Cyanides
Creosote
Dichloroethylene
Dioxins
Dioxanes
Freon 113
Hydrazine Compounds
Isopropanol
Methyl Ethyl Ketone
Methyl Isobutyl Ketone
Methylene Chloride
PCB's
Pentachlorophenols
Perchloroethylene
Pesticides
Phenol
Polynitrophenols
Polynuolear Aromatics (PAHs)
RDX
1,1,TCA Tetrahydrofuran
Trichloroethylene
TNT
Toluene
Triglycol diethyl ether
Polynitrophenols
Vinyl Chloride
Table 3
Direct Operating & Maintenance Costs
for UV/Oxidation at Industrial Installations
TYPE OF WATER
Wood Treating
Uaatavater
wood Treating
Groundvatar
Fuma
Scrubbar
Hatar
contaninatad
Goundvater
ContaBinatad
Groundvatar
Contaminated
Groundvatar
Haatavater
pentachloro-
phanol and
phanol
Pantachloro-
phanol and
phanol
Hydrazine,
Honoaathyl-
hydrazine
Dnayanatrical-
dlxathyl-
hydrazina
TCE, trail*
DCE, HaClj
TCZ, TCA,
DCA, PCE,
XaCl, V1C1
Drinking watar PCE
150 pp»
5 pp»
5,000 ppa
5 pp.
600 ppb
ipp.
90 PP.
15 ppb
POTH
1000/gal
Biotreat-
mant Plant
On-Sita
VOLUME
TREATED
PER DAY
30,000
86,400
DIRECT DIM
COST RAHGE
S1.25-S1.35/
$0.90-$1.00/
1000 gal
$O.OB6/gal
Surface 300,000 $0.47/1000 gal
HatWT
000 gal
216,000 $0.39/1000 g«l
Distrib-
ution syatea
to gal
I gal
Vaata uatar
cyanidaa
75,000 $2.29/1000
Table 3 reports direct O&M costs encountered at commercial projects
treating industrial wastewater, groundwater and drinking water. Con-
taminants in these waters include phenols, chlorinated solvents, hydra-
zine, dimethylnitrosamine, tetrohydrofuran and formaldehyde.
Commercial systems have been designed, built and installed to treat
flows varying from 1,200 to 1,360,000 g pd.
Standard equipment designs are used in all of these installations.
Reactor size varies from 300 to 4,800 gal. Ozone generators range from
21 to 150 Ib/day. In several cases, hydrogen peroxide is used in place
of or with ozone.
Before designing a treatment plant treatability studies are carried out
in the laboratory using glassware equipment to determine the feasibility
of treating the water with UV/O3 or UV/H2O2.
If the results are encouraging, the next step in the study involves the
installation of a skid-mounted, pilot plant on-site. Sufficient design and
economic data normally are collected within 2 to 4 wks. Specifications
for the full-scale system are then prepared. Standard reactors, ozone
generators and hydrogen peroxide feed systems are utilized.
Full-scale systems, in most cases, are automated using microprocessor
control. The system usually requires periodic monitoring (once per shift
or once per day). The systems are designed to operate in a batch or
continuous mode depending upon treatment requirements.
In a number of cases, UV-oxidation is used as part of a treatment
train. For example, at wood treating sites prior to the UV-oxidation
treatment, the wastewater or groundwater requires breaking of oil/water
emulsions and removal of suspended matter as well as adjustment of pH.
CASE STUDY-LORENTZ BARREL AND DRUM SITE
The U.S. EPA has established a formal program to accelerate the
development, demonstration and use of new or innovative technolo-
gies to be used in site cleanups. This program, called the Superfund
Innovative Technology Evaluation (SITE) Program, has four goals:
• To identify and, where possible, remove impediments to the develop-
ment and commercial use of alternative technologies
• To conduct a demonstration program of the more promising inno-
vative technologies for the purpose of establishing reliable per-
formance and cost information for site characterization and cleanup
decision-making
• To develop procedures and policies that encourage selection of avail-
able alternative treatment remedies at Superfund sites
• To structure a development program that nurtures emerging tech-
nologies
Each year, the U.S. EPA solicits proposals to demonstrate innova-
tive technologies. To identify the best available technologies, an exten-
sive solicitation is necessary. A screening and selection process follows,
based on four factors:
• The technology's capability to treat Superfund wastes
• The technology's performance and cost expectations
• The technology's readiness and applicability to full-scale demon-
strations
• The developer's capability and approach to testing
Ultrox was selected in the third year to the SITE program to demon-
strate its UV/oxidation technology. The Lorentz Barrel and Drum
Superfund site in San Jose, California, was selected for the demon-
stration project.
The Lorentz site was used for drum recycling for nearly 40 yr. Over
this period of time, the site received drums from over 800 private com-
panies, military bases, research laboratories and county agencies in
California and Nevada. Drums arrived at the site containing residual
aqueous wastes, organic solvents, acids, metal oxides and oils.
Since 1968, there have been several regulatory actions at the Lorentz
site. In 1987, the Lorentz facility ceased operation and the U.S. EPA
assumed lead agency responsibility for site remediation. Investigations
revealed that the groundwater beneath the site was contaminated with
a number of chlorinated solvents, chlordane, toxaphene and PCBs.
An ULTROX® P-150 pilot plant was moved to the site on Feb. 21,
1989. Thirteen tests were conducted between Feb. 24 and Mar. 9,1989*,
on extracted groundwater from the site. During the treatability bench
CONTAMINATED GROUNDWATER CONTROL 265
-------
studies, TCE, TCA and DCA were chosen to monitor the progress
of the pilot.
The final report has not yet been issued by the U.S. EPA. However,
based on the preliminary results, the UV/oxidation process was
successful in the reduction of all of the VOCs present in the ground-
water at the Lorentz site to below drinking water standards.
The bicarbonate level of the groundwater was extremely high (1200
mg/L). Because of this, treatment costs are higher than what would
be experienced in more normal groundwater applications. Based on
the conditions tested at the site, treatment costs were estimated to be:
Flow Rate:
Influent Concentration:
Effluent Concentration:
Treatment Costs:
Ozone (@ $0.06/kwh)
H2O2 (@ $0.75/lb)
UV (incl. power and annual lamp
replacement
O&M Cost
Capital Amortization
(16%/year)
Total Treatment Cost:
100 gpm
250-1000 M/
pesticides.
< 10 Mg/L
S/1000 gal
$ 0.370
0.156
VOCs.
PCBs
0.836
1.36
0.75
$2.11/1000 gal
risk to the environment, e.g., air stripping without emission control
equipment?
Ultraviolet/oxidation certainly destroys toxics on-site and does not
create residual problems. However, one must be aware of the limita-
tions on UV/oxidation systems. Table 4 is a comparison of features
of two different UV/oxidation methods. Table 5 contains an econom-
ic comparison of UV/oxidation systems.
Table 4
Performance Comparison Between UV/O/HLp, and High Pres-
sured UV with H,O2
O*M
UV/O3/H,0, Hltkw
UV/H.O,
Lo.
Hifhw
»*,
Vino
UV/OJ/H,0, Oood
UV/HjO, Poof
LOT
Hi*
lorn
CASE STUDY-AUTOMOTIVE PARTS
MANUFACTURER, MICHIGAN
Water tested beneath a Michigan automotive parts manufacturer
revealed significant VOC contamination. TCE levels of 5,000 to 10,000
Mg/L were recorded as well as trace levels of other chlorinated sol-
vents. The Michigan Department of Natural Resources required that
the manufacturer pump and treat the groundwater.
The manufacturer investigated air stripping with GAC off-gas treat-
ment, aqueous phase GAC and UV/oxidation as possible treatment
alternatives. Bench-scale studies were conducted at a GAC supplier
and at Ultrox*s laboratory. While all treatment techniques could pro-
vide the required removal levels, UV/oxidation was the most econom-
ical. A pilot-scale treatment system was delivered to the site. Testing
over a 2 wk period confirmed the data obtained in the laboratory. A
full-scale treatment system was ordered and installed in April, 1989.
The system is currently operating and achieving the following results,
which exceed Michigan requirements:
Flow Rate:
Influent Concentration:
Effluent Concentration:
Treatment Costs:
Ozone (@ $0.06/kwh)
H202 (@ $0.75/lb)
UV (incl. power and annual
lamp replacement
O&M Cost
Capital Amortization
(16%/year)
210 gpm
5500 Mg/L TCE
1 /ig/L TCE
S/1000 gal
$0.119
0.188
0.133
0.44
0.29
Total Treatment Cost: $ 0.73/1000 gal
OPERATING AND MAINTENANCE CONSIDERATION
When selecting a treatment process, a number of factors obviously
must be considered. Does the process destroy toxics on-site? Does it
solve an immediate problem but create a new problem or a long-term
TaMe5
Economic Comparison of UV-Oxidation of TCE in Groundwater
Flow fUte
tnfl. TCE COK.
EflLTCECooc
I ppb
ttag/L
UVPOMT*
CsPowcr*
Vuioieaancc " * '
U !«*/«—•
Operuut Ubor
YEARLY COST
10.11
an
aos
in
OJt
JB.
KL0
sa2.sos.oo
1144
Ul
0.13
Oil
Jffl.
OL11
1221,800.00
• Ekartal Euan - HL07/KWH
•• H.O, (100%) Cni - t0.75/B>.
"• Induda lamp npUeUMau, one. • yui tut UV/O,/H|O, «nd 3 » 4 tytu far M|> pwmn i
••" 10 ywr itapreduka/IOK luww
CONCLUSIONS
Over the last IS yr, UV/oxidation has progressed from research and
development to commercial operation. During these years, Ultrox has
advanced its design through applied bench testing, pilot studies and
full-scale systems that remove contaminants from a wide variety of
wastewaters and groundwaters.
UV/oxidation technology is not suitable for every organic contami-
nation problem. It can, however, effectively address a wide range of
contamination problems. This form of on-site chemical oxidation can
offer real advantages over conventional treatment techniques and should
be considered when evaluating water treatment alternatives.
266 CONTAMINATED GROUNDWATER CONTROL
-------
Hydrological and Geochemical Controls Limiting
Contaminant Transport in Groundwater
at Weldon Spring, Missouri
Kathryn Johnson, Ph.D.
Anne Connell, M.S., RE.
Paul Patchin
Felicity Myers
Morrison-Knudsen Environmental Services
Boise, Idaho
ABSTRACT
The Weldon Spring site, considered for inclusion on the NPL, presents
a combination of unique geologic features which precludes the appli-
cation of conventional contaminant migrating transport modeling to
determine the fate of contaminants from four on-site raffinate pits. A
contaminant transport model provides a numerical solution to a con-
tinuum of contaminant dispersion along a groundwater flow path with
the option of a linear retardation term. This approach, however, is not
suitable for the modeling of the discontinuous geochemical and hydro-
logical events along a flow path which occur at Weldon Spring.
Contamination at this site, located near St. Louis, Missouri, stems
from TNT production during the 1940s and uranium refining opera-
tions in the 1950s and 1960s. Although significant sources of uranium
and related radionuclides, nitrate and sulfate, heavy metals and nitro-
aromatic compounds have been identified on-site, off-site receptors show
only low levels of uranium, nitrate and nitroaromatics. This paper dis-
cusses the transport of uranium and nitrate from the raffinate pits and
does not address nitroaromatics.
The off-site contamination by nitrate and uranium has resulted from
a series of discontinuous events. Seepage from the raffinate pits, chemi-
cal retardation at the raffinate pit/overburden interface, groundwater
flow through porous media and subsurface conduits, recharge from
losing streams and spring discharge all contribute to transport.
To better understand the sequential effects of these factors, a sound
conceptual model identifying the discrete segments of the flow paths
and the dominant processes controlling contaminant transport, followed
by an analytical solution when appropriate, proved to be a practical
and useful approach.
INTRODUCTION
The Weldon Spring site is located approximately 30 mi west of
St. Louis in western St. Charles County, Missouri. The site includes
the four raffinate pits constructed to contain wastes from uranium
refining and the Weldon Spring Chemical Plant. Together these are
encompass 217 ac. The U.S. Department of the Army produced TNT
from 1941 to 1944 at Weldon Spring, and the U.S. Atomic Energy Com-
mission operated a uranium feed material plant between 1957 and 1966.
Weldon Spring is located on the drainage divide between the Missis-
sippi and Missouri River basins (Fig. 1). Nearby streams include Schote
Creek, a tributary of Dardenne Creek north of the site and the southeast
drainage, which is an unnamed tributary of the Missouri River south
of the site. Three lakes, known as Lakes 34, 35 and 36, have been con-
structed on the drainages in the August Busch Wildlife Area to the north
of the site.
Several springs and seeps, some of which flow only after a rain, are also present
in the vicinity. Burgermeister Spring, a major perennial spring located imme-
diately upstream of Lake 34, is hydraulically connected to both groundwater
Fig. 1
Weldon Spring Site and Vicinity
(from MK-F and JEG, 1989)
and surface water discharge from the site. This spring has been affected by
uranium and nitrate transport.
TNT production at Weldon Spring introduced nitric and sulfuric acids, metals
and nitroaromatic compounds to the site soils and water. This contamination
generally was confined to areas near the TNT processing plants, wastewater
discharge lines, lagoons, the wastewater sludge incineration areas. The wastewater,
commonly called red water, was stored in lagoons constructed within surface
drainages, and historical records suggest that the lagoons frequently overflowed
into ditches and streams. The major component of the wastewater was sodium
sulfite, used in the purification of TNT. A surface water impoundment, called
the Frog Pond, was constructed as a settling basin for TNT wastewater discharge
(Fig. 1).
CONTAMINATED GROUNDWATER CONTROL 267
-------
Uranium metal production from processed uranium ore, or yellowcake, at
Wfeldon Spring involved the use of nilric, sulfuric and hydrofluoric acids; magne-
sium; sodium carbonate and hydroxide; and olher chemicals. The raffinate pil-s
were constructed of local soils to contain neutralized acidic wastes (Fig. 1). Decant
water from the pits was discharged off-site into the southeast drainage. Ash Pond.
north of the raffinate pits, was constructed to contain ash from an on-site coal-
fired steam generating plant and from incineration of materials contaminated
with uranium (Fig. 1).
The groundwater and surface water transport pathways from on-site sources
to off-site receptors are controlled by geochemical processes wilhm a complex
hydrogeologic regime. Two primary mechanisms contribute to the transport of
contaminants to the groundwater: leaching and seepage from the surface and
subsurface sources through the unsaturatcd zone into the groundwater, and
infiltration of contaminated surface water from streams off-site.
Analysis of the site's geohydrology, contaminant distribution in surface and
groundwater and the geochemical processes immobilizing contaminants has shown
that the Wcldon Spring site does not meet the requirements for conventional
contaminant modeling. The examples given in this paper locus on the flow path
from the raffinate pits and Ash Pond, the major sources of contamination, north-
westward to the streams and ultimately to Burgcrmcisier Spring.
Many of the data referenced in this paper were developed during the site charac-
terization and Remedial Investigation conducted by Morrison-Knudsen and Jacobs
Engineering Group for the U.S. Department of Energy.1 During these
studies, site geology, hydrology and geochemistry were interpreted based
on borehole logs, hydrologic measurements and water chemistry analysis
from geotechnical or monitoring wells drilled to various depths in the
overburden and bedrock.
Data presented here were selected from an extensive project data base
to illustrate the transport of nitrate and uranium from the raffinate pits.
Dye tracing studies to identify the conduit flowing from the site were
conducted by the Missouri Department of Natural Resources.2 Local
hydrologic information was derived primarily from U.S. Geological
Survey investigations.1
TYPICAL BOREHOLE
THICKNESS RANGE (Ft.)
o
(E
3
a>
cc.
in
>
o
TOPSOIL/FILL
LOESS
- 30
10
:RRELVIEW FORMATION 0-22
o
5
2
BO
• COMPETENT
LIMESTONE
Fig 2
Representative Stratigraphy of the Wcldon Spring Site
(from MK-F and JEG. 1989)
GEOLOGY
The geology beneath the site is characterized by 15 to 60 ft of clayey
overburden overlying an argillaceous cherty limestone bedrock of the
Burlington/Kcokuk Formation (Fig. 2). The overburden has been
divided into six recognizable units based on physical characteristics.
These are. in ascending order: residuum, basal till, clay till, Ferrel-
view Formation, loess and topsoil/fill. The overburden generally U
thickest over bedrock lows. Much of the original upper overburden
stratigraphy and the original surface drainage system across the site
were obliterated by cut and fill operations during construction of the
uranium feed material plant.
The Mississippian Burlington/Kcokuk bedrock has been divided into
two units distinguished by the degree of fracturing and weathering
exhibited in the rock. The upper weathered unit ranges in thickness
from 9 (o greater than SO ft. The competent unit extends to about
130 ft to another unit of limestone.
The bedrock surface exhibits a high on the eastern portion of the
site and a low on the north/northwest portion of the site (Fig. 3). The
upper unit is highly weathered at the top, exhibiting solution features
ranging from pinpoint vugs to small cavities which generally are filled
with clay. No large-scale closed depressions characteristic of sink hole
development have been identified on the surface of the bedrock. Linear
depressions developed on the bedrock surface are interpreted to be
preglaciul drainages. The formation of these features appears to have
been controlled by northeasterly and northwesterly trending joint sets.
Fig. 3
Contour Map of Top of Limestone Bedrock
at the Wcldon Spring Site
(from MK-F and JEG. 1989)
HYDROLOGY
The Weldon Spring site is divided into three general drainage systems.
Ash Pond and the raffinate pits drain to the northwest, Frog Pond and
related streams drain in the northeast portion of the site and the southeast
drainage flows from the site to the Missouri River. Surface run-off from
Ash Pond and the outside embankment of the raffinate pits flows off-
site via an NPDES-permitted stormwater discharge point into a tribu-
268 CONTAMINATED OROUNDWATER CONTROL
-------
tary of Schote Creek and into Lake 36. Surface water leaving the south-
eastern portion of the site flows through the southeast drainage toward
the Missouri River.
Most local streams are intermittent and are characterized by losing
and gaining stream reaches. These streams have highly variable flows
and derive most of their water from direct run-off. They lose water
by seepage through the stream bed and gain by inflow from springs,
creating a dynamic connection between surface water and groundwater.
Lost discharge resurges at springs downgradient in the same drainage
or in adjacent drainages. Both wet-weather springs and perennial springs
are present.
With respect to contaminant transport, the aquifer of importance below
the site occurs within the upper zone of the Burlington/Keokuk
Formation. Depth to water ranges from approximately 35 to 65 ft. The
potentiometric surface shows groundwater flowing in a northerly direc-
tion from the site, forming a trough toward Burgermeister Spring
(Fig. 4). An east-northeasterly trending divide exists across the site,
which roughly corresponds to the regional surface water divide (Fig. 5).
The groundwater divide passes beneath Raffinate Pits 1 and 2, within
approximately 330 to 660 ft of the site's southern border. Groundwater
to the north of the divide flows toward the Mississippi River; south
of the divide, groundwater flows to the Missouri River.
I KILOMETER
Fig. 4
Potentiometric Surface of Upper Bedrock Aquifer in the
Vicinity of the Weldon Spring Site
(from Kleeschulte and Emmett, 1987)
Results from packer tests, insitu (slug) tests and pump tests show that
the hydraulic conductivity of the limestone below the site is highly vari-
able Hydraulic conductivities, as determined by the tests, range from
approximately 10'3 to 10'8 cm/sec, with a general tendency for values
to decrease with depth. Both spatial and vertical variability can be
500 250 0 500 IOOO
Fig. 5
Potentiometric Surface of Shallow Bedrock Aquifer
Weldon Spring Site from MK-F and JEG, 1989
attributed to the degree of weathering and fracturing, type of fractures
and interconnection of the fractures in the limestone.
Groundwater movement in the limestone aquifer beneath the site is
believed to occur predominantly by diffuse flow along horizontal bedding
planes and, to a lesser extent, through vertical fractures. Due to its higher
degree of weathering and fracturing, and generally higher hydraulic
conductivities, the upper 10 to 20 ft of saturated bedrock may provide
a preferred zone for groundwater transport. As the intensity of
weathering and fracturing decreases with depth, the aquifer becomes
less homogeneous, flow paths are more widely spaced and the influence
of vertical fractures is more limited. Based on hydraulic measurements
and water quality data, flow to lower zones and deeper aquifers is
believed to be insignificant.
Groundwater flow off-site occurs by diffuse flow as well as through
free-flow conduits. Although specific conduits, such as that from Ash
Pond to Burgermeister Spring, have been identified, no evidence has
been found for conduit flow immediately beneath the site. Dye tracing
tests show that surface flow from the site is lost to the subsurface
immediately west of the site and reemerges at Burgermeister Spring
approximately 48 to 72 hr later, depending on precipitation condi-
tions.2 The straight-line subsurface distance is approximately 6,500 ft.
Comparison of daily flow hydrographs of Burgermeister Spring and
corresponding rainfall indicates that the discharge from the spring
responds quickly to rainfall (Fig. 6).
CONTAMINANT DISTRIBUTION
Water within the four raffinate pits is a source of several elements.
Elements not present at high concentrations in the raffinate water appear
to be elevated above background concentrations in on-site groundwater.
This contamination is assumed to be due to the proximity of contami-
nated soils. Nitrate and uranium, however, are the principal elements
CONTAMINATED GROUNDWATER CONTROL 269
-------
a »•
5 «•
5 o
o.a
0.6
8
Ł0.4
cc.
Ill
O- O.E
« 0
0> 4.0
U
UJ
V>
IE
8
3JC
1C
RAINFALL
.II . ^ Ji. I. 1,
BURGERMEISTER SPRING
WET-WEATHER SPRING
APR MAY JUNE JUUf AIM SEPT OCT NOV DEC JAN FES MAM APR
Fig. 6
Hydrographs of Bur-germeister Spring and Nearby Wet-Weather Spring and
Rainfall. March 1985 to April 1986 from Kleeschulte and Emmett. 1987
which appear in off-site receptors at concentrations above background
(Table 1).
Table 1
Mean Values of Elements in RafTinate Pits, Groundwater,
Ash Pond, and Burgermetster Spring
*. ma ma ma ma ma ma ma ma
ma ma ma ma
nt i1
mi'
It M
t.4 m
ct.i tn n
a 10 «n
M
MM»>
«MOO>
•OCI11
•om>
MM>>
n <« cn
•21 ota »a «oa
HO M Utt U
IH HO in
n UT mo n
41 01 4ttl
(II 41 4M
47 (0. U 00
M tO.l - (40 H
' 04U trm Mi t
' Otu ttr MHlt4Hn« «4lll rfJMOiil 14 riffliuu rill fr«» m-r*r4«lwi kn« jM«tl kftn*4riiq OrMO,
The groundwater distribution patterns of uranium and nitrate, shown
in Figures 7 and 8, illustrate the processes of geochemical retardation
and dispersion. Nitrate is typical of those elements whose concentrations
are controlled primarily by dispersion. Uranium concentrations, on the
other hand, are more affected by geochemical retardation processes.
270 CONTAMINATED GlfOUNDWATER CONTROL
OKAINMM
CURRCNT
PMC IM4
i \
/ \y
' y^WELOON SPRING
SITE BOUNDARY
\
SCAl| IN F|ET
v» no o
900
Fig 7
Mean Nitrate Concentrations (mg/L) in Surface Water
and Groundwater in Selected Monitoring Locations
at Weldon Spring Site (from MK-F and JEG. 1987)
DUtlVlCCS
CUKUfNT
Fig. 8
Mean Uranium Concentrations (ug/L) in Surface Water
and Groundwater in Selected Monitoring Locations
at Weldon Spring Site (from MK-F and JEG, 1989)
-------
Groundwater monitoring wells MW-3007, MW-3008, MW-3009 and
MW-3013, listed in Table 1, were installed adjacent to the raffinate pits.
Comparison of uranium concentrations in the raffinate pit water to con-
centrations in the groundwater adjacent to the pits shows a significant
retardation between the base of the raffinate pits and the groundwater.
This retardation is attributed primarily to the geochemical process of
precipitation as uraninite. The concentration of nitrate at thousands of
mg/L in the groundwater downgradient of the pits, however, suggests
minimal geochemical retardation. The concentrations of nitrate and
uranium measured in three downgradient monitoring wells situated along
the flow path, MW-2003, MW-2002 and MW-2001, are attributed to
hydrodynamic dispersion.
Within the raffinate pits, uranium concentrations average approxi-
mately 829 ug/L, decreasing to an average of 6 ug/L in MW-3007 and
to background levels (less than 7 ug/L) in MW-2003, MW-2002 and
MW-2001 (Fig. 7). Geochemical thermodynamic calculations performed
using PHREEQE4 suggest that this decrease in uranium concentrations
between the raffinate pits and groundwater is caused by precipitation
of uraninite. Precipitation is encouraged by chemically reducing con-
ditions within the overburden and limestone formations, caused by the
presence of organic carbon and sulfide minerals in the soils and rock.
Highly soluble iron concentrations, up to several hundred ug/L, within
the groundwater confirm that this medium is also characterized by the
presence of chemically reducing conditions. These reducing conditions
allow the precipitation of arsenic, molybdenum and vanadium in a
fashion similar to the natural processes occurring during the forma-
tion of uranium roll front deposits.5 The retardation of radium is
attributed to adsorption and coprecipitation.
The pattern of nitrate concentrations in the groundwater illustrates
the dispersion of elements along the groundwater flow path, since nitrate
is not significantly affected by geochemical retardation processes.
Although the reaction from nitrate to nitrite is thermodynamically
favored at the redox potentials of the groundwater, slow kinetics of this
reaction limit the significance of this process.
Within monitoring wells MW-3007, MW-2003, MW-2002 and
MW-2001, mean nitrate concentrations are 4,259 mg/L, 2,811 mg/L,
2,532 mg/L, and 28 mg/L, respectively. The spacing between these
four wells is roughly equidistant. The decrease in nitrate levels from
greater than 4,000 mg/L to approximately 2,500 mg/L over a distance
of approximately 600 ft between MW-3007 and MW-2002 is typical
of hydrodynamic dispersion. The decrease from approximately 2,500
mg/L to 28 mg/L in MW-2001 over approximately 300 ft suggests that
a significant inflow of water with low nitrate concentrations is causing
a dilution of nearly 1 to 100. This contribution of water appears to be
from the subsurface conduit which extends from Ash Pond. This con-
duit carries a large flow of groundwater recharged from the ground-
water divide as well as surface water lost to the subsurface outside of
Ash Pond. The surface water from Ash Pond contains a mean nitrate
value of 48 mg/L. Nitrate concentrations in the discharge from Burger-
meister Spring range from 11 mg/L to 203 mg/L, with a mean value
of 68 mg/L.
The contribution of surface water from Ash Pond to the groundwater
is illustrated by the uranium concentrations in the groundwater, surface
water from Ash Pond and Burgermeister Spring. The uranium concen-
tration in the groundwater near Ash Pond is within background con-
centrations. The surface water from Ash Pond contains greater than
2,000 ug/L of uranium. However, uranium concentrations in Burger-
meister Spring range from 31 ug/L to 240 ug/L, with a mean value
of 125 ug/L. This decrease in uranium concentrations between Ash Pond
runoff and Burgermeister Spring is due to a combination of dispersion
into the groundwater system and removal by precipitation due to the
changes in chemistry between the surface water and the groundwater.
The comparison between the mass flux of nitrate and uranium dis-
charging from Burgermeister Spring illustrates the differences between
the transport patterns of nitrate and uranium (Fig. 9). The mass flux
of nitrate is inversely proportional to the discharge, whereas the mass
flux of uranium is directly proportional to the discharge. The inverse
relationship between mass flux of nitrate and discharge suggests a
dilution of nitrate concentration by the increase in discharge, which
is primarily from surface run-off lost to the subsurface. The direct rela-
tionship between mass flux of uranium and discharge supports the idea
that the major source of uranium is the surface run-off lost to the sub-
surface. The concentration of uranium remains relatively constant, and
the increase in mass flux is attributable to an increase in discharge.
800
700-
600 -
500-
4OO-
300-
200-
100-
0.2 0.4
Discharge (cts)
2400-
2200-
3000-
1800-
I6OO-
1400-
1200-
IOOO-
800-
600-
400-
200-
02 04
Ol.ehorg. (cfi)
Fig. 9
Contaminant Flux vs. Discharge of Burgermeister
Spring, 1987, 1988 (from KM-G and JEG, 1989)
CONCLUSIONS
Waters discharging from Burgermeister Spring are contaminated with
nitrate and uranium. Nitrate concentrations are reduced by a factor of
50 between the groundwater at the site boundary and Burgermeister
Spring. Concentrations of uranium in the surface water leaving Ash
Pond are reduced by a factor of 15 before resurging at Burgermeister
Spring.
The mechanisms of contaminant transport between the raffinate pits
and Burgermeister Spring comprise a series of discontinuous geochemi-
cal and hydrologic events. These events are:
• Seepage from the raffinate pits
• Chemical retardation of uranium and most of the other contaminants,
with the exception of nitrate and other anions, to levels near back-
ground within a few hundred feet of the raffinate pits
• Dispersion of nitrate and low levels of other contaminants in the area
between the raffinate pits and Ash Pond
• Mixing of nitrate-contaminated groundwater with infiltration of
surface run-off, from Ash Pond which is lost to the subsurface
• Chemical precipitation of uranium from Ash Pond waters as the
surface water mixes with the groundwater system
• Conduit flow of groundwater to Burgermeister Spring
CONTAMINATED GROUNDWATER CONTROL 271
-------
This sequence of discontinuous events acting on contaminant trans-
port between the on-site sources of nitrate and uranium and the major
off-site receptor illustrates that conventional contaminant transport
modeling would not appropriately describe the existing contaminant
distribution.
REFERENCES
1. MK-Ferguson Company and Jacobs Engineering Group. Draft Remedial
Investigation Report far the Weldon Spring Site Remedial Action Project,
Weldon Spring. Missouri. U.S. Department of Energy, Oak Ridge Opera-
tions Office, 1989.
Dean, T J, Croundwater Tracing Projea-Wcldon Spring Area. Interim Report,
Pan I, Missouri Department of Natural Resources, 1984.
Kleeschulle, M.J. and Emmctt, L.F., Hydrology and Htuer Quality at the
Weldon Spring Radioactive Wave-Disposal Sites, St. Charles County, Mii-
souri, Waler Resources Investigation Report 87-4169. U.S. Geological Survey
1987.
Parkhursl, D.L., D.C. Thorstcnson and Plummcr, L.N., PHREEQE—A
Computer Pmgramfor Geochemical Calculations, Water Resources Investi-
gation Report 80-96, US Geological Survey. 1980.
Harshman, E.R., Distribution of Elements in Some Roll-Type Uranium
Deposits. Proc of a Symposium on formation of Uranium Ore Deposits, pp.
169-183, Athens. GA, 1974.
272 CONTAMINATED GROUNDWATER CONTROL
-------
Bioremediation Cleans Up the Groundwater Supply
Of a Small Mid-Atlantic Community
Paul M. Yaniga, PG
Frank Aceto, PG
Louis Fournier, PhD
Charlton Matson
Groundwater Technology, Inc.
hadds Ford, Pennsylvania
ABSTRACT
A community maintenance facility experienced a loss of 1000 gal
of gasoline from an underground storage tank. The loss caused an eco-
nomic impact far beyond the cost of the lost fuel. Like many older com-
munities, the location of added-on service facilities grew from an
as-needed perspective. As a result, the fuel storage facility was located
near the community well field. The loss of fuel resulted in adsorbed,
dissolved, vapor and phase-separated organics in the same carbonate
aquifer from which the community drew its 1,000,000 gal of water a day.
An immediate response program, coupled with a comprehensive
short-range aquifer restoration program, eliminated the dependence on
imported water and restored the wells within the impacted aquifer to
service. The elements of on-site/in situ treatment included:
• Volatilization of adsorbed phase organics above the vadose zone with
a soil vent system
• Phase-separated organic recovery from the aquifer proper using a
Scavenger™ two pump system
• Air stripping for removal of dissolved phase organics in the
groundwater
• Enhanced Natural Degradation END™ via native microorganism
stimulation to remove residual adsorbed phase organics in the vadose
and water saturated zone.
INTRODUCTION
A small community in Northeastern Pennsylvania was the first to
use Comprehensive Site Remediation (CSR™) to save its drinking
water supply wells from hydrocarbon contamination. Using CSR™,
soil and groundwater pollutants decreased to U.S. EPA-acceptable risk
levels following implementation of this technology, which eliminated
99%+ of the total contamination.
The contamination occurred in late January, 1985, in Catasauqua,
Pennsylvania, a borough 20 mi north of Allentown. Approximately,
1,000 gal of regular leaded gasoline leaked from a 20-yr-old under-
ground storage tank at the borough's Public Water Works, 50 ft from
one of the supply wells. Analytical data indicated contamination of
municipal well #1.
After discovering the leak, borough officials closed all three wells
and all water pumps and purchased water from a neighboring commu-
nity. Quick resolution of the contamination problem was essential; many
residents were without water and the borough could not afford to buy
water indefinitely. The normal combined yield of the wells was approxi-
mately 1,000,000 gal/day.
SITE DESCRIPTION
The site is approximately 300 ft east of and uphill from, the southern
flowing Catasauqua Creek. The site area is underlain by impure lime-
stone and dolomite of the Epler Formation. This formation is locally
a highly fractured and solution-channeled limestone that produces large
quantities of water from the joint sets and solution channels. Two of
the observation wells (OW 1 and OW 5) encounter competent bedrock
at approximately 25 to 30 ft below surface grade. Calcareous silty sandy
clays and calcareous argillaceous silty and gravelly sands overlay the
bedrock.
The shallow groundwater system, as monitored in the installed
observation wells, is under water table conditions. Depth to ground-
water in the immediate spill area varies from 15 ft to 28 ft below grade.
The water table gradient at this site is relatively flat. This flat lying
gradient limited the rate of groundwater contaminant migration from
the spill site under non-pumping conditions. The natural gradient
measured from the installed wells at the site is approximately 0.1 ft/100 ft
towards Catasauqua Creek (west of the loss area).
INVESTIGATION
To define the direction of groundwater flow and confirm the presence
of gasoline in the groundwater system, the project team constructed
five observation wells around the loss area. Interceptor wells also moni-
tored contamination at the water supply wells. The project team con-
structed the wells to penetrate the upper 10 to 15 ft of the aquifer. This
is the zone typically impacted by free/phase-separated, dissolved product
and adsorbed phase organics.
The annular space of all wells was gravel-packed with 1/8-in by 1/4-in
gravel as a filter medium. After gravel packing, they developed the wells
by air lifting to remove solids.
The project team then surveyed well elevations and tied them into
a USGS topographic map. They collected water level data from all five
observation wells and the Municipal Well #1. The direction of the
groundwater gradient was west, toward Catasauqua Creek with a slope
of approximately 0.01 ft/100 ft.
The retention capacity of soils for gasoline is dependent on soil type.
The finer grained the soil, the greater the capacity to retain hydrocar-
bons. In order to reach the water table as a dissolved or free floating
phase at Catasauqua, the gasoline needed to overcome the formation's
retention absorption capacity.
After installing the observation wells, which showed an initial (stabi-
lizing) absence of free floating product on the water table, GTI con-
ducted a detailed soil analysis. The analysis defined elevated soil
concentrations in the tank pit area (> 30,000 mg/kg). The team con-
ducted limited excavation of this area, as the elevated concentrations
of materials remaining in the soil could leach as dissolved components
to the groundwater system and the water supply wells. Depth to ground-
water (less than 20 ft) and the physical configuration of the site limited
the effectiveness of this excavation to less than 10% of the total loss.
CONTAMINATED GROUNDWATER CONTROL 273
-------
After tank removal, the team found a concrete pad (approximately
12 in thick) in the bottom of the tank pit. The pad had prevented direct
infiltration of gasoline downward from the entire tank pit base. After
removal of the concrete slab, staining and odors in the formation showed
gasoline infiltration was primarily in the northwest area of the tank pit.
Based on these data, excavation of contaminated soil continued in this
area of the tank pit.
During excavation, samples were collected and analyzed for total gaso-
line. Gasoline concentrations in the soil ranged from 0.26 gal/ft1 to
less than 0.09 gal/ft' with an average of 0.14 gal/ft' As a field guide
to excavate accessible soils, project geologists selected the most con-
taminated material by field scanning with an HNu P101 photoionizer.
Depth and anthropogenic features limited access to this soil.
Laboratory analysis of excavated samples determined it was not pos-
sible to recover more gasoline by soil excavation. The results indicated
that excavation of more than 15 yd' of soil recovered only about 0.15 %
of the estimated 7,000 Ib loss. This calculated recovery value was based
on the average concentration of 3.7 gal/yard'. As a result of these
findings and calculations, no further excavation was necessary.
After GTI geologists defined the plume, they also installed a 6-in
recovery well and began pumping the contaminated water. In initiating
the CSR™ program, the technicians equipped the recovery well with
a two-pump oU/water Scavenger™. The Scavenger had a two-fold
purpose: (!) containment of the plume, and (2) recovery of phase-
separated product. In addition, a small-scale aeration apparatus was
utilized. The system stripped groundwater of dissolved constituents and
allowed it to be discharged into a local surface stream. Recognizing
that pump and treat alone could address less than 15% of the contami-
nation at this site, scientists gathered samples for alternative cleanup
feasibility strategies.
The project team examined the traditional cleanup processes including
massive excavation, pumping, treatment, removal, storage and replace-
ment of contaminated soil and determined none of them to be practical
them not feasible. The contaminated plume extended under the water
works and outlying buildings. Successful excavation would require
demolition of these buildings, which was not feasible. Also, the high
costs of such an approach, with the long-term liability of storing con-
taminated soils, was a major concern for borough leaders.
A bioremediation feasibility study helped to determine the safely and
cost-effectiveness of using in situ biodegradation as an alternative to
long-term conventional pump-and-treat techniques. The CSR™
program design and implementation focused on reducing soil-adsorbed
and dissolved phase hydrocarbons. Scientists accomplished the soil by
using both standard pump-and-treat technology with the END™
process for in situ treatment. The focus on the adsorbed-phase organics
became paramount because residual adsorbed phase organics accounted
for more than 80% of the residual aquifer impacts at concentrations
in the tens of thousands of mg/kg.
Microbiologists, after designing and piloting the nutrient mix pro-
gram, began adding hydrogen peroxide and nutrients to the contami-
nated water to enhance the natural degradation process. Hydrogen
peroxide acted as an oxygen source to overcome certain oxygen trans-
fer limitations through the silly sand soil residue.
RESIDUAL CONTAMINATION CLEANUP
Losses of hydrocarbons to the subsurface can lead to a four-fold
problem, dependent on the type and quantity of loss and the nature
of the underlying geologic/hydrogeologic system.
Small losses can lead to the development of an adsorbed phase only
with possible generation of vapors. Larger losses create an adsorbed
problem followed by (he development of a dissolved phase within the
aquifer Each type of loss can generate a vapor phase problem. In cer-
tain environments, losing significant volumes decreases the ability of
the soil to adsorb hydrocarbons and the groundwater system to dissolve
hydrocarbon, creating a free-product phase. In this case, all phases con-
tribute to the evolution of vapors. With this background knowledge,
one must be aware of the various phases and knowledgeable about their
potential environmental impact.
In most losses, the largest volume of contaminants is in the adsorbed
phase. Smaller amounts normally are present in the vapor phase.
The free-product phase can represent varying amounts from 0 to 40%
or slightly more in certain coarse-grained aquifers. The dissolved phase
usually amounts to less than K)% of the overall problem. In this case,
12% of the hydrocarbons were recovered as phase-separated organics,
less than I % as vapor and approximately 2 to 3 % as dissolved organics.
The CSR™ process incorporates the natural elements of hydro-
geology and soil microbiology to construct an in situ bioreactor. The
program involves a comprehensive scientific approach which:
• Uses native groundwater samples containing the indigenous com-
plement of bacteria for evaluation of beneficial hydrocarbon-
consuming species
• Samples impacted water for analysis of the type and concentration
of organic contaminants present
• Samples impacted soil for analyses of contamination to define vertical
and area! impact of fugitive organics
• Defines the specific hvdrogeology or rates pre-existing data for the
application of such information into the construction of an in situ
bioreaction cell, of circulating hydration
• Rates the response of the native microorganisms to nutrient stimula-
tion and enhanced degradation of organic compounds
• Optimizes the nutrient additives to create a formulation of enrich-
Site location with groundwater
plume before CSR™
Site location with groundwater
plume lifter 18 months of CSR™
Figure I
Contaminated Plume
C Site location with groundwater plume
at initiation of project phase out
274 CONTAMINATED 'GROUNDWATER CONTROL
-------
AIR STRIPPING TOWER
DECONTAMINATED SOIL AND WATER
CONTAMINATED WATER AND
SOIL WITH DISSOLVED AND
ADSORBED HYDROCARBONS
Figure 2
Cross Section of Typical CSR™
•"i/' \ /:
,«>;••' \ ,^J
•y _^r\ v---
TE
FCfl 1915 FEET
Figure 3
Project Location Borough Water Works Catasauqua, PA
ment added to the circulating loop of hydration to initiate and sustain
a bioreaction
• Designs an applicable program for site-specific installation of equip-
ment which will initiate an accelerated in situ biorestoration of
impacted soil and groundwater
BOIREMEDIATION PROCESS
For several years, Groundwater Technology, Inc. has used hydrocarbon
bioremediation as part of an overall CSR™ program. Bioremediation
is a program utilizing a naturally occurring process for the oxidation
of organic contaminants by indigenous bacteria. Naturally existing
bacteria flourish within the groundwater and soil of the subsurface
environment. The bioremediation enhancement process provides tech-
M FCfi 84 FEET
OBSERVATION WELL
MUNICIPAL WELL
^ 7^\
COVCKCO IESCXVOM-4 j
Figure 4
Dissolved Hydrocarbons (ppb) Borough Water Works Catasauqua, PA
nical or food grade constituents necessary for increased growth of
bacterial proliferation to those native microbes. This program rapidly
accelerates this natural process (by increasing cell numbers and
metabolism rates) whereby the aerobic species of bacteria biochemi-
cally oxidize organic compounds to CO2 and water. Bacterial con-
sumption of these compoi.mds as energy and carbon sources reduces
their concentrations in the environment. Simultaneously, the dissolved
phase concentration of then organics decreases as a result of bacterial
consumption as does the amount of contamination in the adsorbed phase.
The primary target of bioremediation is the adsorbed phase of con-
tamination. Reduction of contaminants from the adsorbed phase
mitigates not only the dissolved phase but also reduces the amount of
volatile organic vapors.
In a bioremediation system, a carefully balanced pump/injection cell
supplies oxygen- and nutrient-enriched water over and through the area
of concern. This external supply system reaches the adsorbed phase
target. By supplying a properly balanced amount of inorganic nutrients
and oxygen-rich water to the bacteria in the presence of the organic
contaminants, a bloom of beneficial hydrocarbon-consuming bacteria
occurs.
The natural consortium of hydrocarbon-consuming bacteria incor-
porates some of the carbon compounds into cell mass via reproduc-
tion; other carbon is oxidized to CO2. The bacteria, once formed,
further use the organic constituents as an energy source, removing the
contaminant concentrations from the subsurface environment. The by-
products of aerobic degradation are more bacterial cells CO and
H2O. 2
Construction of a system capable of bringing all these factors into
balance and accomplishing a steady-state bioreaction is very compli-
cated. The sciences of hydrogeology, engineering and microbiology work
together to formulate a comprehensive plan of action to accomplish this
technology. Site-specific conditions which need careful analysis and
examination include: geology, hydrogeology, chemistry—both organic
and inorganic, biochemistry and microbiology.
CONTAMINATED GROUNDWATER CONTROL 275
-------
FIELD APPLICATIONS and RESULTS
Feasibility studies revealed a rich and plentiful microbial population
in the ground water at Catasauqua. Of the normal soil saprophytes
present, there was a small percentage present capable of degrading
petroleum hydrocarbons. Although the microbial consortium decreased,
its fate was favorable for biostimulation based on laboratory studies
of nutrient enrichment. Microbiologists expected that an effective
nutrient formulation could bring about a rapid bloom of hydrocarbon-
utilizing bacteria.
The recommended nutrient mixture was the following:
Amount (parts)
400 pts
2 pts.
1.5 pts.
0.5 pts.
Constituent
Ground water
NH4C1
Na,HPO4
NaH.PO,
The inorganic chemistry and geochemistry of this site were not totally
favorable to biostimulation but the overall feasibility for I-ND™ with
design, changes was good. Microbiologists recommended the process
highly, based on the laboratory data.
Due to the nature of the geology and the resistance to infiltration
exhibited via the native soils, GTI decided to consider chemical oxygen
supply by hydrogen peroxide additions. This enhanced oxygen source
could offset certain permeability limitations. GTI determined, in bench-
scale studies, that a period of acclamation by the bacteria would precede
effective use of HKy.
The bacterial tolerance of hydrogen peroxide was very low in non-
acclimated populations. In unadapted samples, the lethal concentra-
tion of hydrogen peroxide was only 30 to 50 mg/L. Samples of ground-
water, enriched by nutrients and allowed to acclimate bloom in the
presence of oxygen for 10 days before hydrogen peroxide additions, were
more tolerant to the peroxide additive. Sub-lethal effects only began
when hydrogen peroxide concentrations reached 500 mg/L and lethal
effects occurred above 1000 mg/L (0.01%). Total sterility occurred at
concentrations above 35,000 mg/L (3.5%).
These studies indicated that direct injection of concentrated hydro-
gen peroxide is lethal to bacteria, but that dilution would use of hydro-
gen peroxide as a soluble oxygen source.
The saturation for oxygen at ambient groundwater temperatures is
approximately 100 mg/L; exceeding that level is wasteful. The system
should maintain a balance between tolerance levels and usable concen-
tration. If anything interrupts the constant supply of oxygen, the micro-
bial cells are stressed. The bacteria] consumption of hydrocarbons is
most effective if the oxygen source is constant and consistently regu-
lated to supply safe, usable concentrations. In designing the system,
it was extremely important to deliver the proper concentration of hydro-
gen peroxide directly to the groundwater. The design of the delivery
system included certain restrictions for safety and liability of handling
concentrates. The peroxide was diluted to avoid shock or lethal injury
to the microbes within the zone of contamination.
The preliminary study empirically determined that hydrogen peroxide
additions are necessary to the groundwater below the zone of contami-
nation. If the zone of adsorbed contamination is as much as 20 ft below
the water table, joint injection of (he hydrogen peroxide must be K) ft
below that or 30 ft below the water level.
At the infiltration gallery influent, the scientists calculated that the
chemical concentrations were in a usable range as the hydrogen peroxide-
enriched water reaches the contaminated zone. Microbiologists took
care to not overdose the system by using a diluted hydrogen peroxide
solution in the beginning followed by steady, gradual increases in hydro-
gen peroxide concentration. Microbiologists expected consumption of
all the hydrogen peroxide. It was not the intention of the operators to
exceed the requirement to prove delivery of the hydrogen peroxide.
Tracer studies by other chemical methods proved that hydrogen perox-
ide was derived. Operators took care in the startup phase of the nutrient
additions not to sub-lethally or lethally shock the system.
After the first 3 mo of cleanup, Catasauqua began limited use of its
drinking water supply. Groundwater Technology, Inc. stopped adding
nutrients temporarily in March, 1987. At that time, the water table stabi-
lized at a historic high level, limiting GTI's effectiveness to treat the
zone of impact. Soon thereafter, before a low water table, GTI micro-
biologists began to add nutrients again. The additional nutrients caused
a bloom of the naturally occurring bacteria to occur as the water table
was sinking. This finding verified remediation at lower zone levels in
the aquifer where low concentrations of contaminants remain adsorbed
to soil.
An examination of dissolved hydrocarbons showed nearly complete
remediation of soil and groundwater. Project phase'out of biodegrada-
tion began in the spring because of the excellent rate of degradation.
The project achieved more than 99%+ reduction to 1 mg/L total
hydrocarbons in the recovery well. The town's water supply returned
to normal.
CONCLUSION
Through a review of the collective data, the applied multi-
disciplined/multifaceted. comprehensive site remediation CSR™
employing Enhanced Natural Degradation (END™) was most effec-
tive in reducing or eliminating contaminant load and restoring aquifer
use.
Key elements to the success of the program included:
• Recognition of the adsorbed phase of organics at 2 to 3 orders of
magnitude greater concentration than dissolved as the loon-term
source of impacts
• Development of a closed loop in situ reactor CSR™ system that
addressed impacts at the source in a contained manner
• Use of hydrogen peroxide as an oxygen enhancer to accelerate biologic
degradation of the organics bound in the silty sand vadose zone
276 CONTAMINATED GROUNDWATER CONTROL
-------
Integration of Borehold Geophysics and
Aquifer Testing to Define a
Fractured Bedrock Hydrogeologic System
William K. Richardson, Jr.
G. L. Kirkpatrick, P.G.
Stephen P. Cline, P.G.
Environmental Resources Management, Inc.
Exton, Pennsylvania
ABSTRACT
Volatile Organic Compounds were detected in a manufacturing facility
production well constructed in a fractured bedrock aquifer of Triassic
age. In order to design a remedial alternative for the contaminated
aquifer, interval packer tests and long-term pumping tests were planned.
A borehole geophysical logging program was designed and implemented
to characterize the structural and lithostratigraphic properties of the
aquifer. The information provided by the geophysical logs and nine
interval packer test results were used in the design of a long-term pump
test. The geophysical data and pump test information were integrated
to define the hydrogeologic system and to design a remedial alternative
for the recovery of groundwater beneath the site.
The borehole geophysical logging program was designed to obtain
required hydrogeologic information within a specific budget. Informa-
tion needs included fracture location and formation geometry as well
as other lithostratigraphic parameters necessary to characterize ground-
water flow at the site. Natural gamma ray, single point resistance, spon-
taneous potential, compressional wave acoustic and caliper logs were
recorded in eight groundwater monitoring wells. Digital geophysical
logs in excess of 600 ft were obtained.
Geophysical logs were interpreted and cross-sections prepared illus-
trating spatial variability of lithostratigraphic units at the site. Fracture
zones were identified based on caliper and acoustic logs for each well.
The geophysical information developed was used to identify potential
water yielding zones within wells. Using the geophysical data, interval
packer tests were designed to provide information on both individual
zone yield and cumulative yield for the wells. The packer tests also
provided a comparison of the water quality of each testing zone.
A long-term pumping test was designed for an interval packer tested
well incorporating eight additional site observation wells. The test design
was based on the interpreted hydrogeologic information obtained from
the geophysical logs and interval packer tests. The long-term pumping
test results were incorporated with geophysical logging and interval
packer test data to further define the site groundwater flow system.
The interpretation of the geophysical logs for the monitoring wells
provided a better understanding of the subsurface hydrogeologic con-
ditions than previously possible using only lithologic drilling logs. Using
geophysical logs, it was possible to better define discrete lithostrati-
graphic units and structural features and to design an effective packer
test and pump testing program. Integration of three major components
of the investigation, specifically, geophysical logging, interval packer
testing and long-term pumping tests, made characterization of the
hydrogeologic system possible. Investigation results are being used to
guide selection and implementation of a site remediation program.
INTRODUCTION
Volatile Organic Compounds (VOCs) were detected at a manufacturing
site production well in Southeastern Pennsylvania during a routine water
sampling program conducted by the state regulatory agency. Investi-
gations were conducted to confirm the existence and determine the extent
of VOCs in soils and groundwater at the facility. Investigation results
indicated the locations of potential source areas and the lateral extent
of on-site groundwater contaminants. Unfortunately, the existing facility
data base did not provide adequate information to design an appropriate,
cost-effective site remedial action. A borehole geophysical logging
program was completed to expand the existing data base by charac-
terizing subsurface geology. The information provided by the borehole
geophysical logs was used to design interval packer tests and long-term
pumping tests to define the site groundwater flow regime.
The site is located within the borough of a small suburban town which
uses surface water resources for its water supply. Well records indicate
that no water supply wells are used within a 0.5-mi radius of the site.
Few industrial water supply wells are located between 0.5 mi and
1 mi of the site. Because of site conditions and limited groundwater
use, public health risk from site contamination was considered limited.
The site is underlain by a Triassic age (240 to 205 million years old)
sedimentary formation subdivided into three members: Upper, Middle
and Lower. The Middle Member, which directly underlies the site, is
approximately 2,300 ft thick. It is composed of fine-grained arkosic
sandstone interbedded with red siltstone and shale.1
These rocks were deposited as coalescing alluvial fans deriving their
sediment from nearby crystalline highlands.1 This depositional en-
vironment typically forms rocks that exhibit abrupt vertical and lateral
changes in both lithology type and texture. Mineral constituents of the
fine-grained sandstone include from 50 to 70% quartz, 30 to 50%
feldspar and 1 to 3% iron minerals.1 Much of the feldspar, originally
present in the matrix, has undergone retrograde metamorphism altering
to sericite and other clay weathering products.1
Continued deposition of overlying deposits caused downwarping of
the formation units resulting in a simple homoclinal formation dip of
12° North. During downwarping and subsequent loading/unloading
events, the formation was subject to fracturing and faulting. Figure 1
illustrates fracture sets developed in formation outcrops perpendicular
as well as parallel to bedding. The frequency and interconnection of
fractures were considered to be potential factors in groundwater move-
ment beneath the site.
The primary objective of the investigation was to define the hydro-
geologic system to the extent necessary to design an effective ground-
water recovery system for the identified contaminant plume. Critical
to this task was an adequate definition of subsurface geology and its
influence on groundwater occurrence and movement. At the initiation
of this investigation, site-specific information concerning the subsurface
lithologies, condition of the rock matrix, or fracture frequency was not
available. The effect of a pumping recovery well on the groundwater
CONTAMINATED GROUNDWATER CONTROL 277
-------
Fig I
Well Developed Vertical Kraclurc Sets and
Weathered Begging Planes in Outcrop
The Pen in Photograph Center is 5 in Long
system was unknown. Additionally, the vertical extent of aquifer
contamination due to VOCs was not known. This paper summarizes
the integration of several technologies including borehole geophysics
and pump tests required to characterize the site, to the degree neces-
sary, for the selection of a remedial approach.
TECHNICAL APPROACH
The results of previous environmental consultant investigations
indicated elevated concentrations of VOCs to 7.5 ^g/L in groundwater
beneath the site. The verticaJ extent of this contamination had not been
defined, and the deep production well repeatedly contained elevated
VOC concentrations.
Prior to this investigation, a simplified site groundwater flow model
was used. This model was developed using limned information gathered
during investigations designed for source area and groundwater plume
definition. Three types of geologic information pertaining to the site
existed: (I) general geologic agency information.1 (2) geologic infor-
mation obtained during the drilling of site monitoring wells and
(3) structural and fracture trace information gathered from outcrop study
and fracture trace analysis. The above information, although useful,
did not provide an adequate interpretation of site hydrogeology
The presence of dominant fracture sets, perpendicular and parallel
to bedding planes, and the environment of deposition make the ground-
water flow beneath the site complex and not easily predictable Prior
to this investigation it was assumed that some component of ground-
water flow occurred in fractures and bedding planes as well as through
primary matrix pores The degree to which fractures, bedding planes
and lithologic changes controlled groundwater flow beneath (he site
was unknown. Specifically, the effect of aquifer amsolropy on ground-
water flow to a pumping well was unclear
The following tasks were designed to meet the informational needs
of the investigation:
• Examination of cuttings collected during groundw-jtcr monitoring wvll
installation
• Conducting a borehole geophysical investigation including natural
gamma ray, spontaneous potential, single point resistance, culiper
and compressional wave acoustic logs in nine site wells which range
in depth from 38 to 600 ft
• Design and performance of interval packer tests on a 2**) ft deep
production well
• Design and performance of a 3-day pumping lest of the production
well
• Performance of short-term pumping tests on shallow monitoring wells
to estimate the variability in transnnssivilies over the site
The above approach was designed to provide a maximum amount
278 CONTAMINATED GROUNDWATI-.R CONTROL
of information concerning the hydrogeologic system and vertical extent
of contamination, while performing (he work within a relatively short
time-frame and restricted budget. Borehole geophysical logging of site
wells provided a cost-effective characterization of site geology. Litho-
stratigraphic and structural details were made available for interpre-
tation (including a well that previously had no information) on the day
that logging was performed. This approach provided a broader data
base concerning the subsurface geology for the design of nine interval
packer tests and one long-term pumping test that would later provide
additional hydrogeologic data.
Scanning Electron Microscopy (SEM) was used to examine under-
lying bedrock petrography. Cuttings collected during drilling of ground-
water monitoring wells were examined with emphasis placed on
diagcnctic changes specific to the sandstone and inierbedded siltstones
and shales
The borehole geophysical program was designed by analysis of
regional geology, site-specific well construction data and pump teat
requirements Geophysical tool selection was guided by bedrock chanc-
icristics and included an examination of primary luhologies as well as
fracture occurrence For most wells at the site, there was little or DO
information available concerning the materials penetrated. Selection
of geophysical tools was also based on engineering parameters (casing
si/e. depth and integrity) of interest.
The most cost-effective logging suite for the investigation was then
determined To identify luhologies. logs selected were natural gamma
ray. spontaneous potential and single point electrical resistance. To locale
open bedding planes and fractures, a sensitive three-arm caliper tool
was selected Compressional wave acoustic logging was completed for
specific monitoring wells as an independent verification of porosity and
fracture occurrence
Field data were acquired only after the optimum borehole geophysi-
cal logging program was designed. Operations included tool calibra-
tion, parameter measurement and recording, digital data processing and
QA/QC logging runs.
Natural gamma ray values were recorded in standard American
Petroleum Institute (API) units. Three-arm caliper measurements were
made with a sensitive caliper device capable of detecting borehole wall
geometry changes to O.I in. The caliper logs w«re plotted at an exag-
gerated scale for easy fracture bedding plane identification.
Geophysical tool accuracy was assessed by repeating tool measure-
ments over a critical 100-ft borehole section. Overlay comparison of
the original geophysical data and repeat geophysical data for each
measurement provided assurance that the recorded data were correct.
Packer test intervals were determined on the basis of borehole geophysics
interpreted lithologic changes as well as locations of fracture and
weathered bedding plane locations. Packer test intervals were also
selected based on the caliper logs such that packer locations would pro-
vide an effective seal against the borehole. Figure 4 presents interpreted
luhologies as well as interval packer test results.
The interval pucker lest of the facility production well was designed
using the data provided by the geophysical logging program. The packer
was configured lo isolate nine potential subsurface water-yielding zones.
These /ones were identified for isolation and testing based on inter-
preted borehole geophysical data.
A long-term, constant rate pumping test was conducted on the same
facility production w-cll after the step drawdown tests were completed.
Information used in the design of the lest included: (I) geophysical logs,
(2) information obtained from the interval packer tests and (3) yield
information from a step drawdown test performed after the interval
packer test The long-term pumping test was conducted for a period
of approximately 3 days while continuously monitoring water levels
in the plant production well and eight monitoring wells.
DISCUSSION OF RESULTS
Sl-M data for shale samples indicate that primary depositional fabrics
have been destroyed by diagcncsis. Specific grain boundaries of clastic
sill and clay sized particles are obscured by silica overgrowth and al-
teration of feldspars lo scricitc and other clay mineral weathering
products. Little or no intergranular porosity was observed within the
-------
shale interbeds. In contrast, SEM data for sandstone samples indicate
that, in addition to both silica cementation and feldspar alteration, secon-
dary rhombohedral calcite precipitation has significantly reduced in-
tergranular porosity (Fig. 2). Calcite filled fractures were documented
during drilling at the site. It appears that little interconnected pore space
exists in the sandstone matrix.
Fig. 2
SEM Photograph of Sandstone Encountered While
Installing Photograph of Sandstone Encountered
While Installing Monitoring Well.
Porosity Obscured by Sericite and Silica Overgrowth
(Photo Courtesy West Chester University).
Even though sandstone and shale were subject to varying degrees
of diagenetic alteration, the effect on the hydraulic properties of the
units are similar. Destruction of interconnected pore systems indicated
that critical hydraulic characteristics of the rock beneath the site may
not be wholly attributed to primary matrix porosity. The original deposi-
tional properties of the shales and sandstones did not appear to be a
dominant factor influencing the groundwater flow system.
Interpretation of the geophysical logs was completed after the data
had been verified as accurate and precise. Natural gamma ray, spon-
taneous potential, acoustic and single point electrical resistance logs
were used to aid in lithology interpretation. The three-arm caliper log
was used to identify fracture/bedding plane positions.
Figure 3 illustrates both lateral and vertical site stratigraphy. In
general, the site is underlain by interbedded sandstone, siltstone and
shale of varying fracture density.
Three major sandstone units were encountered by the wells at the
site. Small interbedded silty shale units were identified in these sand
units. Fractures and weathered bedding planes appeared to be better
developed in the shale units rather than the overlying and underlying
sandstone units. Open fractures and weathered bedding planes were
well developed in shale sequences.
From the borehole geophysics completed at the site, outcrop exami-
nation and petrographic microscope work, it appeared that groundwater
movement may be strongly controlled by fracture interconnection.
Specific depth intervals were targeted for packer testing to determine
relative hydraulic characteristics (transmissivity, and storativity).
As shown on Figure 4, packer test intervals were designed to pro-
vide coverage of the entire borehole. Each zone was pumped in step-
drawdown fashion while recording the change in water level within each
zone and in the open borehole above. Water quality samples were
obtained for each zone that yielded a sufficient amount of water. The
interval packer test data collected were used to evaluate the vertical
distribution of VOCs entering the well, vertical head distribution of
the nine zones tested, approximate yield of each zone and to estimate
Fig. 3
Stratigraphic Section of Subsurface Site Stratigraphy.
CONTAMINATED GROUNDWATER CONTROL 279
-------
Zone 1 Pecker TMI
Pumped Dry: No Recovery
Potemtometric Heed
Elev.: 14041
Zone 2 Pecker Ten
14GPM TMI 1,20OPM Te«! 2
2.73 cX Drawdown
Poienftomeinc Heed
Elev.: 140.31
Zone 3 Packer Teit
13GPM Tetl 1.20GPM Teit 2
4.96 ot Drawdown
Potenbomelric Heed
Elev.-148.43
Zone 4 Pecker TMI
126PU Ten 1.16GPM Tetl 2
56.79 ol Drawdown
Po unborn* We Heed
Elev.:1SO.B4
Zone 5 Pecker Tetl
3GPMTe«1
121.IB of Drawdown
Potertttarnetrtc Heed
Elev.: 150.34
Zone 6 Pecker Tetl
7GPM Tesl 1.1GPM Tetl 2
112.44 o( Drawdown
PoMntkxnetrlc Heed
Elev.: 1S734
Zone? Pecker TMI
3GPM Ted 1 Pumped Dry
158.18 of Drawdown
Polenttometrtc Heed
Elev.: 150.02
Zones Pecker TMI
3QPMTetl1 Pumped Dry
185.420) Drawdown
PoMnHomelrtc Heed
Elev.: 158.43
Zoned Pecker TMI
10GPM Ten 1.12GPM Ten 2
103.58 ol Drawdown
PolenUometrtc Heed
Elev.: 150.91
HAUMl
Fig. 4
Production Well Data Showing Interpreted Lithology,
Geophysical Logs, Packer Tesl Results.
280 CONTAMINATED OROUNDWATER CONTROL
-------
the relative hydraulic conductivity of the zones tested. The results also
provided information concerning the amount of vertical interconnec-
tion between water producing zones.
Based on the vertical profile of VOCs entering the well, it was deter-
mined that only the upper 200 ft of the plant production well were con-
taminated. A packer was placed in the well at 200 ft so that only the
upper 200 ft of the well were tested during the long-term aquifer test.
An elongated cone of depression developed over the site during the
long-term aquifer test. The orientation of the major axis of the water
table surface at an angle to bedrock strike suggests that ground water-
flow is fracture controlled. This is consistent with the observed features
documented during geophysical logging and with the hydraulic properties
of the zones packer tested.
In general, the geophysical logs delineated the locations of lithologic
and structural features. Petrographic analysis indicated that sandstones
and shales had little or no primary matrix porosity. This observation
was further confirmed by the results of the packer testing, which showed
very little yield from tested sandstone units, but relatively high yield
from tested zones containing frequent open weathered bedding planes
and fractures. This information indicated that the hydrogeologic system
was heterogeneous and anisotropic. Further, the packer test results
demonstrated the potential for sandstone units to act as confining units
within the aquifer. This information aided in the selection of an ap-
propriate analytical solution for determination of hydraulic properties
from aquifer test data. The long-term aquifer test documented the
anisotropic nature of the flow system and helped define the dynamics
of groundwater flow to a pumping well under the site conditions. This
information and other data collected during the investigation were used
in the selection and design of a remedial alternative for the site.
CONCLUSIONS
The geophysical logs provided a continuous, quantitative and quali-
tative record of the site geology. Subsurface stratigraphic correlations
were more straightforward than available drilling logs and provided data
on wells for which none were previously vailable. Aquifer tests were
then designed to characterize the subsurface hydraulic characteristics
of the lithologic and structural features identified using borehole geo-
physics.
The integrations of geophysical logs and pump test data were useful
in defining the hydrogeologic system present beneath the site. Investi-
gation results are being used to guide selection and implementation of
a site remediation program. Based on pumping test results, the most
effective remediation system was determined to target identified frac-
ture systems rather than specific lithostratigraphic units.
REFERENCES
1. Rima, D. R., Meisler, H. and Longwill, S., "Geology and Hydrogeology
of the Stockton Formation in Southeastern Pennsylvania," Pennsylvania Geo-
logic Survey Topographic and Geologic Survey Bulletin W-14, 1962.
CONTAMINATED GROUNDWATER CONTROL 281
-------
Remedial Action—A Success Story
Clyde Hutchison, RE.
Virginia Bretzke, EIT
B&V Waste Science and Technology Corp.
Overland Park, Kansas
ABSTRACT
The cleanup of hazardous waste sites presents many challenges above
and beyond typical construction projects. At the Petro-Chemical Systems
site near Liberty, Texas, the remedial action was completed success-
fully. Elements of interest associated with the design and construction
at the site include: health and safety issues; challenges of an expedited
schedule and limited data; and projecting and monitoring costs. When
the project was completed, the objectives of the remedial action, to
improve site access and minimize potential for contaminant exposure,
were effectively met with a final construction cost less than 2% above
the original bid price.
INTRODUCTION
The Petro-Chemical Systems, Inc. site, originally the site of a waste
oil processing company, was identified in 1970 as a potential environ-
mental problem. While the company was in operation, waste oils were
stored in several pits on approximately 4 to 6 ac of the site. Records
indicate waste oils were also spread on the access road. Frontier Park
Road, as a dust suppressant. All waste oil disposal operations were dis-
continued in June, 1970, at the request of the Texas Water Quality Board.
The State granted a commercial disposal facility permit to Petro-
Chemical Systems in 1971, but the permit application was subsequently
withdrawn in 1974.
Preliminary sampling conducted in 1982 and 1984 by the Texas Water
Commission (TWC) and the U.S. EPA indicated elevated concentra-
tions of several polyaromatic hydrocarbons in the formerly used dis-
posal pits. The documented presence of hazardous constituents on-site
led to placement on the NPL in late 1984.
In March, 1985. the State initialed a Rl/FS for the Petro-Chemical
Systems, Inc. hazardous waste site. The initial phase of this effort con-
centrated upon Frontier Park Road, the primary access to the site.
Frontier Park Road was an unimproved road extending eastward from
State Road FM-563 approximately 2 mi to a crossing at Turtle Bayou
(Fig. 1). The area was heavily wooded along both sides of the road,
except for sections cleared at residences. The area is relatively flat with
little topographic relief. The crossing at Turtle Bayou, consisting of
several culverts encased in concrete, had washed out. [hie to the ab-
sence of an adequate drainage structure at Turtle Bayou and at local-
ized drainage ways along the road alignment, the road was impassable
much of the year.
The RI documented the extent of contamination along the road. The
highly and moderately contaminated areas of the roadway are shown
on Figure 1. Highly contaminated soils are defined as having greater
than KX) ppm polyaromatic hydrocarbons (RVHs) or total wlatiks (TVs),
while moderately contaminated soils have between K) to 100 ppm PAHs
or TVs. The highest levels of contamination in these areas were gener-
ally within the upper 2 ft of the roadway.
Based on the findings of the RI/FS, a ROD was issued in March,
1987. The purpose of the remedial action authorized by the ROD was
to provide access to the site so to conduct a thorough RI of the entire.
Highly contaminated Moderute/y contaminated \fa(,\i
Moderately
contaminated
Decontamination
station of vault
-Moderately contaminated
•••*
j »
Sn *
tt
! I |
||
}
\
! ri^
rtv?; !
fc»>w/ crossing
Decontamination station]
at excavation areo^_/
u— -
1
6ENERAL SYMBOLS
—*•— rmn*n u«
—*/fc EAUMCMT UNI
—•/»•- mw
-------
The selected remedy for the remediation and reconstruction of the road
highlighted the following features:
• Excavate contaminated soil to below 100 ppm PAHs and/or 100 ppm
1 VS.
• Temporarily dispose of contaminated soils in an on-site storage
facility, designed in accordance with RCRA guidelines. (Final
remediation of contaminated soil will be evaluated in the full site
RI/FS.)
• Construct a new asphalt road over excavated areas and existing
roadway.
• Temporarily relocate on-site residents during construction.
The work was authorized to be completed as an Expedited Response
Action (ERA), which has a statutory limit of $2 million and a 1-yr limit
for design and construction. This fast-track approach was selected so
that site access would be available for the second phase of the RI. Black
& Veatch, as an associate firm of CH2M HILL under the REM IV
Contract with U.S. EPA, prepared the design and served as Construc-
tion Manager (CM) for die project.
The scope of work for the construction subcontract was defined in
detail in the plans and specifications that were developed. The design
included the following elements:
• Excavate approximately 5000 yd3 of highly contaminated soil in
Frontier Park Road between FM-563 and Station 18+00.
• Construct a double-lined on-site storage facility (vault) for temporary
secure storage of the excavated material.
• Backfill the excavated area and other portions of the road which were
below finished grade with uncontaminated native soils.
• Construct a road from FM-563 to east of Turtle Bayou, a length of
approximately two mi. The road work consisted of excavation and
filling to the subgrade elevation and construction of a lime stabilized
subbase, a flexible crushed rock base course and an asphalt surface
course.
• Shape and grade drainage ditches and install corrugated metal pipe
culverts and a structural aluminum-plate crossing at Turtle Bayou.
The existing drainage structure at Turtle Bayou was demolished and
removed from the site.
• Construct two vehicle decontamination stations.
• Provide site security, including fences and a guardhouse.
Seven bids for the construction contract were received, with Tricil
Environmental Response, Inc. selected as the lowest responsive, respon-
sible bidder.
HEALTH AND SAFETY
Health and safety issues are of utmost concern when working at a
hazardous waste site, and they differentiate remedial actions from stan-
dard construction projects. The remedial action must be designed to
minimize potential exposure to contaminants and to provide protection
to the health of residents and construction workers in accordance with
OSHA regulations (40 CFR 1910).
Before initiating any on-site activities, the CM developed a site safety
plan for activities by their personnel and the surveying and geotech-
nical testing subcontractors, while the Contractor was responsible for
his own site safety plan. Based on the contaminants at the site and the
activities being performed, the site safety plan specified the type of
protective clothing to be worn and methods for health and safety
monitoring. Most of the work in contaminated areas was conducted
in Level D protection. Level C protection was used when there was
a potential for respiratory exposure to contaminated dust, such as when
the highly contaminated material was excavated from the roadway and
placed in the vault. All personnel working on-site were required to be
medically monitored and health and safety trained, including a 40-hr
training course.
Because of concerns about protecting the health of the residents during
construction, residents living along Frontier Park Road were temporarily
relocated. The road was blocked and site security was set up at the
entrance to control access to the site. A temporary bypass road was
constructed around the first 1,600 ft of roadway, which was highly con-
taminated, to prevent contact with the contaminants during construc-
tion. Use of the temporary bypass road also controlled the spread of
ontaminants by limiting traffic through the highly contaminated area.
The plans and specifications also addressed methods for preventing
the spread of contamination to other areas of the site. Vehicle decon-
tamination stations were included near the section of roadway with the
highly contaminated material and near the vault to avoid contaminating
clean areas of the site while transporting highly contaminated material
to the vault. A material tracking system was established to document
that all excavated contaminated material was placed in the vault and
that the transporting vehicles were decontaminated.
Water from vehicle and personnel decontamination areas, as well as
water that contacted contaminated material in the excavation area and
the vault, was collected and treated. The water had to meet discharge
requirements established by the State before being released to Turtle
Bayou. The Contractor provided large frac tanks for water storage. The
first tank was used for contaminated water while the second was used
for water treated by the on-site treatment system. The treated water was
retained until laboratory results confirmed State discharge requirements
were satisfied. Careful planning of construction activities to periods
of dry weather minimized the volume of contaminated water requiring
treatment.
The storage vault was constructed in conformance with OSHA guide-
lines to protect workers during construction and to prevent releases of
contaminants. Following clearing and grubbing, a 12-in lime-stabilized
subgrade was constructed. Since a portion of the area where the vault
was constructed was contaminated, workers wore Level C protection
for the subgrade construction and for construction of the first lifts of
the 3-ft clay base. Equipment remained within the exclusion area until
these activities were completed and was decontaminated prior to removal
to clean areas of the site.
Levels of protection were downgraded for the remainder of the base
construction since work was being performed on clean material and
air monitoring showed no respiratory exposure from surrounding areas.
This approach resulted in improved worker productivity while main-
taining health and safety protection. A similar approach was followed
for placement of the contaminated material in the vault. Compaction
equipment remained in the vault for the duration of the filling.
Level C dermal and respiratory protection was worn during place-
ment of the contaminated soil and installation of the first lifts of the
multilayer cap. This protection was downgraded after the waste was
covered and workers could work on clean imported material.
SCHEDULE AND DATA LIMITATIONS
One of the challenges of this project was the expedited schedule under
which the project had to be completed. Because the project was initially
pursued as an ERA, the schedule was established to satisfy the 1-yr
completion requirement. Three months were allowed for design, 2 mo
for advertisement, bid and award of the construction contract, and 7 mo
for construction activities. A key element to completing the work on
an expedited schedule was the cooperation among all entities involved,
including the U.S. EPA, the CM, the Contractor, representatives from
State and local agencies and the local residents.
The schedule constraints were a primary factor in developing the de-
sign approach. The key personnel on the design team visited the site
soon after project initiation to evaluate existing conditions and to con-
tact local utilities and agencies. Because some of the existing site data
were not adequate to complete the design, methods were needed to
collect the necessary data without impacting the schedule.
The geotechnical data for the site were limited to soil borings along
the roadway. These data were adequate to establish the structural foun-
dation requirements for the road and on-site vault. However, additional
geotechnical data were required to design the lime-stabilized subgrade
for the roadway and vault. The permeability of the native soils also
needed to be established to evaluate design options for the RCRA vault.
Five bulk soil samples were collected and analyzed for these physi-
cal properties. In addition, geotechnical testing was required during
construction to confirm that elements of work met the specified require-
SITE REMEDIATION 283
-------
ments. Considering the expedited schedule for the project, a single
procurement was used for both design and construction phase geotech-
nical testing support. Because the extent of geotechnical testing during
construction was not well defined al the beginning of the design phase,
the subcontract was bid on a unit price basis, allowing for adjustments
based on the final construction specifications and the field conditions
encountered during construction.
Surveying was required during the design phase to assess the topo-
graphy of the proposed vault area and to ensure that the vault was con-
structed within the appropriate property boundaries. Benchmarks were
established along the roadway, and the road design was prepared based
on a limited survey of the road conducted during the RI. The roadway
was surveyed in detail by the contractor at the beginning of construc-
tion. Adjustments to the designed roadway and drainage patterns were
made based on this detailed survey. For example, no drainage existed
south of the roadway as projected on the USGS map. Consequently,
the drainage design was adjusted to direct runoff in roadway ditches
east to Turtle Bayou. During clearing of the roadway alignment and
the subsequent survey, a number of additional driveway culverts were
identified which had not been discovered in the RI survey. The timely
discovery of the drainage and culvert features allowed early adjustments
and eliminated schedule impacts. Unit prices bid for culverts and ditch
construction allowed equitable cost adjustments, eliminating potential
schedule delays due to change order negotiation.
Because the RI concentrated on the roadway, little data were availa-
ble on the extent of contamination in the vault area. Grab samples of
soil were collected and tested for organic and inorganic contaminants.
The analyses results were used to evaluate whether clear and grub
material from the vault area had to be processed as contaminated
material. If untested, the clear and grub material would have had to
be considered contaminated and placed in the on-site vault. The ulti-
mate cost of final disposal of contaminated material justified the col-
lection of samples so that a significant portion of the material could
be handled as standard clear and grub material.
Other provisions were included in the contract documents to accom-
modate data gaps that could not be filled within the design period. In
some cases, U.S. EPA helped define the scope of work and set guide-
lines so the work could proceed on an expedited schedule. For example,
U.S. EPA directed that the limits of highly contaminated material
excavated from the roadway be as defined in the RI. Although samples
were collected at the bottom of this excavation to verify that the con-
tamination was removed, the area could be backfilled immediately,
without waiting for the analytical results. The U.S. EPA also made the
necessary arrangements to get access to the property during construc-
tion. This process included both temporary access during construction
and permanent extension of the roadway easement for extension of the
ditch construction. The U.S. EPA also made arrangements to relocate
local residents for the construction period and worked with the con-
tractor to accommodate resident access needs during construction.
Good communications and timely decisions were important elements
in keeping the project on schedule during construction. The on-site Con-
struction Manager represented the U.S. EPA and the design team and
worked with the Contractor on changes due to field conditions. Monthly
project meetings with representatives of the Contractor, the CM and
the U.S. EPA were held to expeditiously handle any problems that arose.
Turn-around time for the review of shop drawings and preparation of
change orders was minimized. Shop drawings had to be reviewed within
14 days of when they were received; often, however, they were reviewed
immediately upon receipt. Geotechnical field testing was also conducted
in a timely manner, with the testing firm available to perform field tests
within 48-hr notice. This cooperative effort allowed the construction
to be substantially complete in the required 7 mo.
COSTS
This project was initially authorized as an ERA and had a funding
limit of $2,OOOjOOO. All costs associated with the project had to be within
this limit, including construction and other subcontracts, engineering
fees, operation and maintenance costs and expenses incurred by the
U.S. EPA. Based on the preliminary cost estimate for the project of
$1.27 million presented in the FS, it appeared that the entire project
could be completed under the $2,000,000 limit. However, when all the
elements of the design were included, it became apparent that the
$2,000,000 limit would be exceeded. The actual low bid for construc-
tion was $1,690,889. In order to have enough funding to cover other
project expenses, as well as allow for contingencies in the construc-
tion cost, it was determined that this project was better suited to be
conducted as a Remedial Action (RA), which does not have budget
limitations, rather than as an ERA.
Although it was recognized in (he early stages of design that the cost
would be more than $2,000,000, it took time for the U.S. EPA to change
the funding mechanisms. The overall schedule for the project was
lengthened almost 4 mo, between the time the design and bid phases
were complete and the time the contract could be awarded. This delay
emphasizes the importance of making sure FS costs are realistic and
account for design details, since budgets are often based on this estimate.
Because of the limited data available at the time of design, the con-
struction contract was established with a mix of fixed prices and unit
prices. For cases where the scope of work was well defined, such as
the vault base and the decontamination facilities, fixed prices were used.
Otherwise, unit prices were used to allow for field modifications and
to account for actual quantities of materials used. Since the level of
personnel protection required has a significant impact on productivity,
a level of protection was listed on the bid form for each item to serve
as a basis for estimating the probable cost of the work. For some items,
the work was divided into two levels of protection. For example, the
road subgrade was treated with lime in both moderately contaminated
and non-contaminated areas. Therefore, unit prices were given for
preparing the lime-treated subgrade in both Level C and Level D
protection Unit prices were also bid for upgrading or downgrading
from the assumed level of protection that was listed on the bid form.
During construction, the on-site Construction Manager monitored
the work, reviewed the Contractor's payment requests, negotiated the
payment allowances based on personal observation and records of the
work and matte recommendations regarding payment. He also worked
with the Contractor in responding to changed field conditions, and
negotiated change order costs. Costs were controlled by this careful
monitoring, so that the final construction cost (including K) change
orders) was less than a 2% increase over the bid amount.
EFFECTIVENESS OF REMEDY
In addition to the objectives slated in the ROD, the following criteria
for the remedial action were developed in the RI/FS Report to meet
the cleanup objectives established by TWC and the U.S. EB\ for Frontier
Park Road:
• Improve access for equipment to the site to facilitate the planned RI
sampling and monitoring of the on-site waste disposal areas and to
facilitate future remedial actions
• Prevent contact with highly contaminated soils, defined as R\Hs
and/or TVs in excess of 100 ppm
• Minimize direct contact with moderately contaminated soils, defined
as PAHs and/or TVs between 10 and 100 ppm
The asphalt road was constructed to provide access to the site, in-
cluding access across Turtle Bayou. Additionally, the decontamination
station by the vault was left in place for use during future site investi-
gations. Therefore, the remedial action effectively met the first cleanup
objective for improved site access.
The highly contaminated materials were excavated to the limits indi-
cated in the RI/FS, as directed by the U.S. EPA, with revisions made
to the limits based on visual observations. Samples were collected from
the bottom of the excavated area to*document the level of cleanup
obtained. The cleanup goal established in the ROD targeted the 100
ppm level of PAHs and/or TVs. Sixteen of the 18 samples, or 89%,
met the combined cleanup criteria. The cleanup goals were based on
a residential setting with the potential for multiple routes of exposure
due to the unimproved condition of the road. The remedial action at
the site has reduced this potential for exposure. The majority of the
highly contaminated material has been removed from the roadway and
284 SITE REMEDIATION
-------
is secured in an on-site facility. The placement of clean fill and instal-
lation of an asphalt road over this area (and moderately contaminated
areas) has further reduced exposure potential and effectively fulfills
the second and third cleanup objectives.
Hence, the requirements of the ROD were met by the remedial action.
Most of the highly contaminated soils were excavated and disposed of
in the project-constructed on-site storage facility, thereby mitigating the
risk of human exposure to contaminants. A new asphalt road was con-
structed, providing access to the site and covering areas of moderate
contamination. The residents were temporarily relocated during con-
struction, and the Contractor cooperated with the residents to provide
them access to their property. Through the cooperative effort of all
entities involved with the remedial action, the work at Frontier Park
Road was successfully completed.
SITE REMEDIATION 285
-------
Case History: Fort A.P. Hill
Dioxin incineration Project
Thomas O. Mineo, P.E.
Metcalf and Eddy Technologies, Inc.
Somerville, New Jersey
Dominique K. Edwards
U.S. Army, USATHAMA
Aberdeen Proving Ground, Maryland
ABSTRACT
Metcalf & Eddy Technologies, Inc., under contract to Q H. Materials
Corporation, performed the remedial action in which dioxin-
contaminated soil and building debris at Fort A.P. Hill, Virginia, were
thermally treated and destroyed. The site, located 60 mi south of
Washington, D.C., contained 190 tons of dioxin-contaminated material.
Rotary kiln incineration was chosen as the most effective method to
achieve the U.S. Army's goal of ultimately disposing of the dioxin-
contaminated material stored at the Fort.
This paper describes the process Metcalf & Eddy Technologies, Inc.
used to develop the engineering and design report as well as the process
utilized in the field to perform the work. The U.S. Army Toxic and
Hazardous Materials Agency represented Fort A.P. Hill for this project
and required that all actions be performed in accordance with CERCLA
and ARARs.
The Engineering and Design Report was written to consider the
remedial design plan, incineration technologies, standard operation
procedures, sampling and analysis plan, site-specific health and safety
plan and the project management plan. As the report was developed,
Metcalf & Eddy Technologies' field management was consulted on all
details to assure a practical design.
The description of the process includes a discussion of materials
handling operations, materials shredding operations, rotary kiln incinera-
tor operations, ash handling operations and sampling operations. Brief
description of the quality assurance and site-specific health and safety
requirements also are presented.
INTRODUCTION
In February and March of 1989, Metcalf and Eddy Technologies,
Inc., under contract to O.H. Materials Corporation, performed a
remedial action in which dioxin-contaminated soil and building debris
were thermally treated and destroyed on-site by rotary kiln incinera-
tion. The remedial action was performed at Fort A.P. Hill on behalf
of the United States Army Toxic and Hazardous Materials Agency
(USATHAMA).
BACKGROUND
The Fort A.P. Hill Site, a U.S. Army installation located in Bowling
Green, Virginia, had a small storage building on the facility that was
contaminated by leaking containers of herbicides which were stored
inside. The herbicides included 2,4-D, 2,4,5-T and silvex. Dioxins,
which are known contaminates of the herbicides, were discovered in
the building's wooden floor, block foundations and soil adjacent to the
building. The bulk of the material contained dioxins in concentrations
of 0.001 to 0.002 ppm, with the highest concentration equal to 1.030 ppm.
The contaminated materials were excavated and isolated in 35-gal
fiberboard drums and subsequently over-packed into 55-gal steel drums.
A total of 1138 drums of material was stored in a secure building on
the site. USATHAMA initiated a FS in July, 1987 to analyze alterna-
tives for ultimate disposal of the dioxin-contaminated materials being
stored at Fort A.P. Hill. The study was performed in accordance with
the NCP which governs procedures for such studies. The draft final
FS was completed in March, 1988. The preferred disposal alternative
identified in the study was on-site incineration. Following review by
U.S. EPA Region III and the Virginia Department of Waste Manage-
ment, the FS was finalized in August, 1988.
A two-phased contract task was awarded to O.H. Materials. Corp.
to prepare operating and design plans (phase I) and conduct incinera-
tion of the dioxin-contaminated material (phase II). Metcalf &. Eddy
Technologies, Inc. was subcontracted to prepare the design and operating
plans and direct the field activities.
ENGINEERING AND DESIGN REPORT
Metcalf & Eddy, Inc. was assigned to prepare the Engineering and
Design Report for the project. The report describes the specific proto-
cols for each task that was to be carried out as a pan of the remedial
action. Specifically, the report included the following sections.
• Summary report
• Remedial design plan
• Review of incineration of acutely hazardous wastes
• Standard operating procedures
• Sampling and analysis plan
• Site specific health and safety plan
• Management plan
The schedule set by USATHAMA required destruction of all dioxin
by Apr. 30, 1989. In order to meet this schedule, it was necessary to
prepare the first draft of the over 200-page document in a period of
6 wk. This draft report was completed on time and submitted to
USATHAMA for review and approval.
REGULATORY REVIEW
Regulatory participation was an integral part of the successful dis-
posal of the dioxin-contaminated material stored at Fort A.P. Hill. All
plans and reports were reviewed by both federal and state regulators
as well as various Army agencies and commands. Adequate review
periods were incorporated into the schedule to allow draft plans and
reports to be thoroughly evaluated by the necessary parties. Timely
review was crucial to maintain the stringent schedule. The U.S. EPA,
Region III and the Virginia Department of Waste Management were
involved with meetings with the Army prior to the conduct of the FS.
Once the FS identified incineration as a potential remedial alternative,
the Virginia Air Pollution Control Board became involved with reviewing
286 SITE REMEDIATION
-------
plans and reports.
In accordance with the CERCLA and SARA, a public comment
period which included a public meeting was held at the completion
of the FS. When these events had been completed, the FS was finalized
and a ROD was prepared. The ROD contained a Responsiveness Sum-
mary which documents all questions and comments received during
the public comment period and responses to the questions which were
provided through the USATHAMA Public Affairs Office. Information
Repositories of convenient access (e.g., local libraries) were located
near Fort A.P. Hill to allow the public to review plans and reports.
The ROD documented the decision to implement on-site incineration
using a mobile rotary kiln.
Multiple copies of the draft operating and design plans were sub-
mitted to USATHAMA as required under contract. USATHAMA
retained a copy for review and forwarded the remaining copies to the
Fort A.P. Hill Environmental Office for distribution to the appropriate
reviewers. Copies of all plans and reports were sent to the U.S. EPA,
Region HI, the Virginia Department of Waste management, the Vir-
ginia Air Pollution Control Board, Fort A.P. Hill, the U.S. Army Toxic
and Hazardous Materials Agency, the U.S. Army Environmental
Environmental Hygiene Agency, the U.S. Army Forces Command Envi-
ronmental Office and the Department of Army Environmental Office.
Once the comments about the draft design and operating plans were
received, a meeting was held in January, 1989 to discuss issues. The
draft plan was then revised to incorporate the appropriate changes. This
revision was a very critical part of the program since it provided the
reviewing regulatory agencies with the opportunity to input their recom-
mendations and assured the Army that all federal, state and local regula-
tory requirements were met.
FIELD OPERATIONS
Field Operations starting with site preparation and equipment mobili-
zation began after the design and operating plans were finalized. The
USATHAMA project officer was present for the majority of the field
operations to monitor the contract, to verify that field activities were
in accordance with the operating plans and to provide timely decisions
when modifications to the operating procedures were required.
Fort A.P. Hill personnel routinely inspected the site to monitor progress.
U.S. EPA Region HI personnel and their contractor provided round-
the-clock oversight.
The field operation, with the exception of preliminary site prepara-
tion and ash drum disposal, took place on a 24-hr schedule for 7 wk.
During this period, a group of specialty subcontractors, brought together
by Metcalf & Eddy Technologies, integrated their unique capabilities
to successfully carry out this remedial action project.
Mobilization
Mobilization took place in a period of 2.5 wk. During this time, a
chain link fence was installed to delineate the drum storage area and
the exclusion work zone area. Electric power and telephone services
were brought in from the road 650 ft away. Five site support trailers,
two water supply trailers and four LPG tanks were set up. As the site
was being prepared, equipment arrived. The thermal destruction unit
(TDU) was the largest single piece of equipment. It included an
incinerator trailer, a pollution control trailer, a feed hopper and con-
veyor, an ash discharge hopper and conveyor, a lamella separator and
vacuum filter, and several tanks for caustic and water. All of the TDU
equipment was supported on steel plates placed directly on the crushed
stone and required no special foundations. The other major equipment
utilized on-site was a trailer mounted low speed double shredder. Wood
cribbing directly under the shredder portion of the trailer was all that
was required to adequately support the shredder.
A 1200-ft2 temporary building constructed of modular wood framing
was erected on the area between the shredder and TDU. The building
was placed directly on the ground and anchored. Three strategically-
located HEPA filter exhaust fans were placed inside the building to main-
tain negative pressure and minimize dust within the building.
All 1138 drums were brought over from the storage site to the work
area; a distance of 3 mi. The drums were placed in the drum storage
area. The warehouse where the drums had been stored was cleaned
with HEPA filtered vacuum cleaners.
Finally, field quality control personnel set up their field sample
collection facility. They also set up an on-site computerized drum
monitoring data base designed to track the materials during operations.
Health and Safety personnel reviewed the site conditions and made final
revisions to the site Health and Safety Plan. All personnel attended on-
site safety briefings, and the operations phase was ready to begin.
Operations
After a final walk through by the U.S. EPA, USATHAMA, Fort A.P.
Hill personnel and contractor representatives, the site was declared
operational. All work for the next 4 wk would be carried out in Level
C protection.
The operations activities included the following:
• Feeding material into the shredder
• Shredding the material
• Staging discharged shredded material within the building enclosure
• Feeding material into the TDU hopper
• Incinerating the material
• Feeding ash discharge into drums
• Storaging ash drums
The 55 gal steel drum overpacks were opened on the concrete decon-
tamination pad which was set at the gate joining the work area (Level C)
and the drum storage area (Level D). A forklift with a drum handling
device carried each drum to the shredder and emptied the 35 gal fiber
board drum into the hopper. The empty 55 gal steel overpack was
returned to the decontamination pad for decontamination.
The material entering the shredder was reduced in size and discharged
within the building enclosure adjacent to the shredder. The discharged
material went into one of 10 steel bins used to supply a 12-hr supply
of material for the TDU.
The bins within the building enclosure were emptied into the TDU
feed hopper which also was located inside the building. The material
was moved within the building by a small forklift and was discharged
into the TDU feed hopper by means of a self dumping mechanism on
each bin.
Material was fed continuously into the TDU 24 hr/day. Material
entering the TDU was carefully controlled and regulated by means of
a sophisticated control system which monitored and controlled primary
kiln temperatures, secondary combustion chamber temperatures, gas
flow and other parameters as required by the approved Engineering and
Design Report and government regulations. The unit operated using
rotary kiln combustion technology. Pollution control equipments
including a wet scrubbers was monitored and controlled integrally with
the combustion unit by means of a computerized control center. Auto-
matic shutdown sequences and alarms built into the control system
ensured an environmentally safe operation.
The ash from the rotary kiln was discharged into 55 gal shipping
drums. Forklifts fitted with drum handling devices carried these ash
drums to the decontamination pad where they were given an external
cleaning and transferred into the drum storage area. A separate forklift
in the Level D drum storage area carried each drum to its final staging
area prior to off-site disposal.
During all operations, Metcalf & Eddy quality control personnel
maintained records, gathered samples and checked the overall quality
of work. Ash samples taken each day were sent to an off-site labora-
tory for 48-hr turn-around dioxin analysis.
Health and Safety personnel held daily briefings for each shift and
continuously monitored on-site personnel for conformance to safety
requirements. Everyone was required to present proof of OSHA CFR
1910.120 training before being allowed into the work area. No serious
injury or exposure occurred during the remedial action.
Demobilization
All equipment was systematically decontaminated in order to minimize
the potential for dioxin contamination after TDU shutdown. All equip-
ment was demobilized in a period of 1.5 wk.
SITE REMEDIATION 287
-------
OFF-SITE DISPOSAL CONCLUSION
The drums filled with ash residue were approved for disposal at a The remedial action at Fort A.P. Hill met the U.S. Army's objec-
Class I hazardous waste landfill in Oklahoma. The ash drums were lives of ultimately disposing of the dioxin-contaminated material stored
removed from the site in June, 1989. Carbon filtered process water also at the Fort. In achieving this objective, the U.S. Army was extremely
was removed from the site in June, 1989 and sent to an approved understanding of local community concern. A tour given to the news
industrial water treatment facility in New Jersey. media of the mobilized site prior to the start of operations is one exam-
ple of Fort A.P. Hill openness in carrying out this remedial action.
All dioxin-contaminated material was treated and ash residues were
removed from the site before the end of June, 1989.
288 SITE REMEDIATION
-------
Decontamination of Explosive Contaminated
Sructures and Equipment
U.S. Army Toxic and Hazardous Materials Agency
Craig MacPhee
Aberdeen Proving Ground, Maryland
ABSTRACT
As a result of past operations, the U.S. Army has numerous buildings
and large quantities of process equipment which are contaminated with
explosives. The Toxic and Hazardous Materials Agency has been con-
ducting a three-phase study aimed at developing a safe, economical
and non-destructive method of rendering explosive contaminated
materials ready for reuse or disposal.
In Phase I of the study, 56 technologies were assessed for possible
use. Of the 56, the five most promising technologies underwent labora-
tory scale testing in Phase II. A thermal process which uses hot gas
to heat the contaminated materials to approximately 500 °F emerged
from the Phase II tests as the best technology. The hot gas process was
selected because of its relative safety and low labor costs.
Phase II of the development program which is now in progress is
full-scale pilot testing of the hot gas decontamination process. To test
how well the process works on structures, an extensive test program
on a building previously used for explosive munition demilitarization
was completed in August, 1987. In order to determine how effective
the hot gas process is on explosive contaminated processing equipment,
a flashing chamber in the Western Area Demilitarization Facility at
Hawthorne Army Ammunition Plant, Nevada, was modified for hot
gas testing. The series of tests was completed in September, 1989.
A wide vaiety of materials such as contaminated sewer lines, piping,
electrical motors and mixing kettles were successfully decontaminated.
Exposing items to 500 °F for 12 hr. removed all the explosive con-
taminants present (TNT, RDX, HMX, NC and NG).
INTRODUCTION
Probably the two most common methods in present use for removing
explosive material contamination are steam cleaning and decontami-
nation by fire (burn to the ground). Steam cleaning is, in most cases,
effective but provides only surface decontamination and is not effec-
tive on hard-to-access areas. It is difficult to completely decontaminate
concrete with steam. Steam cleaning of complex items such as motors
cannot assure that interior areas are cleaned.
Burning of structures contaminated with explosives has several draw-
backs. If other structures are nearby, burning the building may be risky.
Additionally, buildings with asbestos should not be burned. Finally,
open burning of a contaminated structure can be viewed as an uncon-
trolled release of toxic substances, local and state regulators may prohibit
intentional building fires.
In 1982, the U.S. Army Toxic and Hazardous Materials Agency
(USATHAMA) began a project aimed at developing new, improved
procedures for decontaminating structures and equipment contaminated
with explosives. The goal of this ongoing project is to develop a method
which will be safe, will produce little or no waste and will assure a
high degree of decontamination. Target compounds for removal are all
the major military explosives (TNT, RDX, HMX, NG, Tetryl, etc.).
The process to be developed would have to effectively remove con-
taminants from metal, wood, painted concrete and bare concrete. An
additional goal of the project is to develop a decontamination method
which is universally applicable and thus can be used on large struc-
tures as well as process equipment. The first phase of this project was
a review of existing techniques and the consideration of novel techniques.
Thermal Decomposition Concepts
Flashblast Contact Heating Hot Plasma
Microwave heating Flaming Hot Gases
Solvent Soak/Burn Infrared Heating
CO2 Laser
Electropolishing
Sandblasting
Ultrasound
Vacu-blast
Abrasive Concepts
Acid Etch Scarifer
Demolition Drill and Spall
Cryogenics Hydroblasting
Extractive Removal Concepts
Solvent Circulation Supercritical Fluids Rad Kleen
Surfactants Strippable Coatings Manual Steaming
External Steam Vapor Phase Solvent
Generator Extract
Radical Initiated
Decomp.
Molten Decomp.
Microbial
Ultraviolet and Cat.
Nucleophilic
Displacement
Solid State
Chemical Concepts
Base Initiated
Decomp.
Sulfur Base Reduct
Reduction Cleavage
Gamma Rad.
Ozone
Gels
Decomp. with DS2
Sodium Borohydride
Reactive Amines
Chromic Acid
Ascorbate
Foams
PHASE I OF DEVELOPMENT PROGRAM,
TECHNOLOGY SCREENING:
Under contract to USATHAMA, Battelle Columbus Laboratories per-
formed an analysis of existing explosives decontamination techniques
and also developed descriptions of novel concepts. Information was
gathered from government and private industry manufacturers of
explosives. Government facilities were visited to inspect contaminated
structures and equipment. In a July, 1983 report, Battelle documented
the detailed analysis of the following technologies:
SITE REMEDIATION 289
-------
Hydrogenation
Various combinations of methods also were considered. Each tech-
nology was evaluated based on destruction efficiency, mass transfer,
safety, damage to buildings, penetration depth, applicability to com-
plex surfaces, operating costs, capital costs and waste treatment costs.
Among the thermal decomposition concepts, hot gases received the
highest ranking overall and received high scores in all categories. The
hot gas process involves exposing contaminated materials to hot gases
in order to vaporize or decompose the contaminants. The hot gases
together with the vaporized explosives and breakdown products arc sent
to an afterburner for complete destruction.
The burn-to-ground method received high scores in most categories
but received the lowest possible scores for safety and building damage.
The only thermal concept recommended for further development was
hot gas treatment.
All of the abrasive concepts received poor scores for waste treatment
costs. The abrasive concepts also received low scores for penetration
depth. None of the abrasive concepts were considered for further
development.
External steam generator (pumping steam into the structure) scored
the highest of the extractive removal concepts. However, the low solu-
bility of some explosives in hot water prevents the steam method from
being universally applicable. Vapor circulation was the only extraction
technology selected for further development.
Three chemical decomposition techniques were selected for further
development. The concepts selected were radical initiated decomposi-
tion, base initiated decomposition and sulfur based reduction.
From the combination methods evaluated, only a combined chemi-
cal/hot gas concept was considered to be worthy of further development.
In all, 55 technologies or combinations of technologies were consi-
dered. Six concepts were selected for further investigation. The selected
technologies were hot gases, combination chemical/hot gas, vapor cir-
culation, radical initiated decomposition, base initiated decomposition
and sulfur based reduction.
PHASE II, LABORATORY TESTS
In Phase n, the technologies selected from Phase I were developed
in more detail. Probably the most important aspect of the development
work was the laboratory tests. Test coupons of steel, painted concrete
and unpainted concrete were spiked with known quantities of 2,4 DNT,
2ft DNT, TNT, TETRYL, RDX and HMX. The test coupons were
then subjected to the processes under investigation. After appropriate
treatment times, the coupons were inspected for residual explosives.
Hot gases and the combination of chemical/hot gases yielded the highest
degrees of explosives removal. In many cases the residual explosive
levels were below detection limits. Although each of the six processes
evaluated in the laboratory phase of testing offered some advantages
and disadvantages for particular operations, it was the hot gas process
which had a greater range of applications and provided the most
complete decontamination.
The laboratory tests did identify some potential problems with the
hot gas process. During testing, the formation of explosive crystals on
the outside surface (originally uncontaminated) of concrete test coupons
indicated that hot gases may cause explosives to migrate through con-
crete. This finding raises the concern that during decontamination of
a concrete structure, the explosives may be driven out of the structure
rather than destroyed. It was also noticed that the hot gas process dried
out and thus weakened the concrete.
Pretreatment of concrete with a caustic chemical led to quicker
destruction of explosives and allowed hot gas decontamination to proceed
at lower temperatures. Quicker destruction of explosives reduces the
possibility of migration. Operating at a reduced temperature lessens
the drying effects on concrete. Thus, it was concluded that the combi-
nation of chemical treatment and hot gases would be the best route to
complete decontamination without migration of explosives and with
minimal damage to concrete.
The hot gas process, complemented by chemical pretreatment,
emerged from the laboratory tests as clearly the most promising tech-
nology for widespread application. The next step was to see how well
the process would perform outside the laboratory on a contaminated
building.
PHASE II PILOT TESTS:
The Cornhusker Army Ammunition Plant (AAP) Tests:
Pilot tests of the chemical/hot gas decontamination method were
conducted at Cornhusker AAP in 1987. The tests were conducted for
USATHAMA by Arthur D. Little, Inc. The objectives of these first
pilot tests were to:
• Determine the effectiveness of hot gas with and without chemical
p ret real mem
• Evaluate the effects of test conditions on the integrity of an actual
structure
• Provide design criteria for full-scale systems
• Provide test data for regulatory permitting of the process
After numerous potential sites were considered, a projectile washout
building at Cornhusker AAP was selected as the test site. The building
has concrete walls, a concrete floor and a wooden ceiling. Dimensions
of the building were 25 ft. long, 25 ft. wide and 11 ft. high. Some modifi-
cations to the building were necessary such as construction of a false
ceiling to protect the wooden roof, replacement of the windows and
doors with sheet metal and insulation of the outside of the building.
Although inspection of the building revealed some TNT contamina-
tion, the level of contamination was too low to sufficiently challenge
the decontamination method. This problem was resolved by placing
TNT contaminated concrete blocks, which were removed from a sump
cesspool, inside the test building.
Hot gas was supplied to the building through ductwork by a 3jO million
BTU/hr. propane-fired burner. Gases exited the building into a propane-
fired afterburner. Gases entering the building, exiling the building and
exiting the afterburner were analyzed. In tests where chemical pretreat-
ment was used, a solution of sodium hydroxide and dimethylformamide
was employed. Theromocouples were used to monitor temperatures in-
side the building during treatment. Concrete samples were subjected
to mechanical properties tests before and after hot gas treatment.
Conclusions drawn from the Cornhusker pilot tests were:
• Hot gas decontamination of a building is safe and feasible.
• Although treatment of surfaces with caustic chemicals did increase
explosive removal on the surface of concrete, it has no effect on
interior contamination. Further, longer treatment with hot gas alone
should be capable of providing complete decontamination.
• The hot gas decontamination process caused the concrete block to
loose 5 % of its compressive strength and 20 to 30% of its bend (ten-
sile) strength. The effects of this loss in strength would have to be
judged on a case by case basis for each building treated. Of course,
if the building is not going to be reused, the condition of the con-
crete after treatment is of no concern.
• Initial design criteria and cost estimates for decontamination of small
and large buildings were developed.
• Process data, such as composition of effluent gases from the after-
burner, was collected and can be used for applying for regulatory
permits for future operations.
The Hawthorne AAP Pilot Tests:
Further pilot tests of the hot gas process (without chemical pretreat-
ment) were conducted in the summer of 1989 at Hawthorne AAP. These
tests were conducted for USATHAMA by Roy F. Weston, Inc. This
test series was directed towards the decontamination of process equip-
ment used in explosives operations. The objectives were to:
• Test the process on a variety of materials (vitrified clay, steel and
aluminum) with variety of contaminants (TNT, NC, NG and ammo-
nium picrate).
• Test the process on a variety of items including intricate equipment
which has areas inaccessible to other treatment processes (pumps,
pipes, ship mines, risers and transfer containers).
• Determine the temperatures and treatment times required to reduce
290 SITE REMEDIATION
-------
contaminant levels to below detectable limits. Define a process that
will render equipment items fit for unrestricted use or disposal.
• Render large quantities of contaminated equipment fit for unrestricted
use or disposal.
A flashing chamber at Hawthorne AAP was modified to accommo-
date the hot gas process. The same burner and afterburner that were
used at Cornhusker AAP were used at Hawthorne AAP. Hawthorne
AAP has a large store of equipment and munition items which require
treatment. Test items were selected from Hawthorne AAP's stores,
placed in the modified flashing chamber and treated with hot gas. Test
samples also included highly contaminated clay pipe removed from what
was once the West Virginia Ordnance Works.
Test items were sampled for explosives prior to testing. Some items
were spiked with explosives. After testing, the items were sampled for
residual explosives. The detection limit for explosive contamination was
approximately 10 mg/m2.
Conclusions for the Hawthorne pilot tests were:
• The hot gas process successfully decontaminated all items tested.
TNT, RDX, HMX, DNT, NC and NG were completely removed
for both exterior and interior surfaces.
• Heating contaminated items to 500 °F for 12 hr. rendered the items
completely decontaminated.
SITE REMEDIATION 291
-------
Stabilization of Petroleum Sludges
Jeffrey C. Evans, Ph.D., P.E.
Stephen Pancoski
Bucknell University
Lewisburg, Pennsylvania
ABSTRACT
Petroleum refineries historically have produced large quantities of
waste acidic petroleum sludges which typically were disposed of in open
pits. These practices have resulted into the need to develop a cost-
effective method to prevent the migration of these materials into the
environment. This paper describes the work performed in the first half
of a 3-yr research effort designed to investigate methods to effectively
stabilize and solidify acidic petroleum sludges.
Specific additives to achieve stabilization and solidification have been
investigated including several commercial products and processes.
generally proprietary in nature. Conventional stabilization agents, such
as cement and fly ash, along with more innovative agents such as
organically modified clays (organophilic clays) were utilized in these
laboratory investigations.
A large number of stabilization agents was evaluated with regard lo
their effectiveness in stabilizing petroleum sludge. The solidification
was evaluated quantitatively in an unconfined compression test. The
Toxicity Characteristic Leaching Procedure was used to evaluate the
leachability of the treated material. Results of these laboratory studies
are presented along with recommendations for further testing.
The laboratory tests were found to be limited in their ability (o
differentiate the stabilization effectiveness for the materials tested.
Despite the limitations of the test techniques, the effectiveness of a
variety of stabilization mixes was assessed in relative terms, based upon
comparisons of the stabilization mix test results.
In genera], the organically modified clay mixes have shown the most
promise in stabilizing the petroleum sludge. In the case of the organo-
philic clay mixes, the higher the cost of the mix, the better the perfor-
mance with regard to the measured test parameters. These clays, used
in conjunction with some type of binder material such as portland
cement, appear to provide the system necessary to adequately stabilize
and solidify organic-bearing hazardous wastes. In this system, (he
organic contaminants are contained by the clay and also arc trapped
in the physical matrix formed by the cemenl or other pozzolanic
material.
INTRODUCTION
These investigations consisted of an evaluation of various stabiliza-
tion agents for effectively stabilizing and solidifying an acidic petroleum
sludge, typical of those produced by oil refineries in the period from
approximately 1920 to 1970. This paper describes the results of the
studies conducted during the first half of a 3-yr project designed lo
develop a stabilization technique which effectively stabilizes and
solidifies the organic (hydrocarbon) constituents of a specific acidic
petroleum sludge.
At the start of the investigation, a bibliography of applicable litera-
ture was compiled and reviewed and a survey of commercial vendors
was undertaken to identify applicable candidate technologies available
in the marketplace. These literature and vendor surveys provided the
basis for selecting and evaluating additives and processes to be further
studied in the laboratory'
Tabte I.
Stabilization Mix Summary
Mil NO. HIK COMPOS HI OK
I Hudst/»ll«pulgll«/My Ath/QulckllM (1/0.42/0.31/0.08)
fludgt/lanJlon*/My Ath/0ulckl !•» (1/0.4/0.3/0.08)
lludfl./itnlonll./fly Aih/OulcklIM (I/O.4/0.3/0.M)
Sludgc/londlon«/gentanl(*/My Ath/OulckllM (1/0.4/0.2/0.3/0.08)
Sludgc/Cliyton* APA/fly Aih/Oulckl IM (1/0.4/0.3/0.M)
Sludge/AttipulflU/Hy Aih/OulckllM (1/0.4/0.1/0.04)
Sludg*/fu«p«ntan*/riy Adi/Oulckl IM (1/0.4/0.1/0.00)
Sluda*/Cliyton* 40/fly Aih/Oulckl IK (1/0.4/0.1/0.01)
lludg*/lond(on*/OI*IOMXtou* Cirth/OulckllM (1/0.4/0.3/0.08)
10 Sludg*ton ( City ton* APA (1.0/1.0)
28 Sludgc/CtMnt Kiln Dull (1.0/1.0)
29 tludg*/CM*nt Kiln Ouit (1.0/1.5)
30 lludge/ltfllonlte/MlcroHn* Cment (HC-500) (1/0.4/0.25)
31 Sludgc/lcnlonltc/Mlcroflnc Cement (KC-100) (1/0.4/0.25)
32 Sludge/Cement Kiln Oult/Solubl* SodI in Illicit* (1/1/0.1)
31 !lud«(/Cmen< Kiln Ouit/Sotubl* Sod I in SltlcKe (1/1/0.2)
34 Iludtc/Cmenl (l)/Solub(e Sod I in Slllc.tt (1/1.5/0.175)
35 Sludfle/Cl.yton. APA/C«nent (l)/Satubtt Slllcite (1/0.4/0.25/0.15)
36 «(ud««/P-40/P-27 (Illicit* technology) (1/0.3/0.3)
]7 lludg«/P-40/P-27 (Illicit* technology) (1/0.6/0.6)
38 Sludgt/Sorbond C«S II (1.0/1.0)
39 Slud«e/2Ht/C«Mnt (I) (1/0.4/0.5)
40 tludg«/1S-S5/C«
-------
A laboratory testing program was developed which included physi-
cal and chemical characterizations of the untreated sludge and of the
treated material after mixing and after a 2-wk curing period. Results
of these initial laboratory studies will be used to refine the stabiliza-
tion methods in the second half of the investigation which will include
full-scale field studies.
The stabilization agents employed in the investigation include those
used in mixes described in the literature. Products from commercial
vendors as well as other generic materials were incorporated into this
investigation. The mix ingredients are listed on Table 1 and described
more fully elsewhere2. In summary, the testing program included
mixes from previous studies3 and custom mixes employing sorbents
such as processed clays, organically modified clays, binding agents,
soluble silicates and proprietary agents from vendors. The ingredients
were mixed with the sludge in various combinations and proportions
aimed at stabilizing the acidic petroleum sludge. The mixing proce-
dures and methods of evaluating the treated material are described
elsewhere4. Studies also were conducted to evaluate the sorption
capacity of the various organophilic clays5.
The laboratory testing program consisted of physical and chemical
testing of both the untreated sludge and the stabilized product. Table 1
presents a summary of the first 50 stabilization mixes, including the
proportions by weight of each of the stabilization mix ingredients.
The unconfined compression test results for those mixes which con-
tained sludge ranged from 2 to 90 psi. These data emphasize the inhi-
bition of hydration reactions due to the presence of the organic sludge.
For the mixes which contain sludge, five of the six strongest mixes,
as evaluated by the unconfined compression test, contain an organophilic
clay. This result indicates the strength benefits which result from the
addition of an organophilic clay.
The chemical testing consisted of performing a modified TCLP test
on each specimen after it had been tested in unconfined compression.
The concentrations of the identified chemical constituents were com-
pared with those limits specified in the TCLP. In all cases, the reported
concentrations of the selected organic compounds were below the
maximum concentration levels specified by the TCLP.
ANALYSIS OF TEST RESULTS
The following sections view the data from the perspective of spe-
cific groups of mixes. Average values for the entire set of mixes are
compared to the average values for each subset of mixes (Table 2). The
subsets consist of mixes which contain specific stabilization agents.
Fly Ash and Lime Mixes
Fifteen of the 50 stabilization mixes contained fly ash. The fly ash
and lime mixes initially were stronger than the mixes without these
ingredients but, with curing, this strength advantage disappeared. High
phenol concentration and sum of organics values indicate that the organic
constituents present in the sludge were not contained by the mixes with
fly ash and lime as well as by some of the other mixes.
Cement Mixes
Of the 50 stabilization mixes, 23 contained cement. As seen in
Table 2, there are only slight differences between the averages for the
cement mixes and the averages for all the mixes, with regard to the
strength and chemical parameters. These similarities in average values
between the cement mixes and the entire set of mixes are due, in part,
to the large number of cement mixes contained in the data set.
Most of the stabilization mixes with Type I, II or V portland cement
resulted in mixes with favorable total organic carbon values but the rela-
tive hydrocarbon concentration values for these mixes varied. The mixes
containing microfine cement, MC-500 or MC-100, had favorable rela-
tive hydrocarbon concentration numbers but mixed total organic carbon
results. The strengths, as measured by the unconfined compression test,
were inconsistent for both the cement and the microfine cement mixes.
The stabilization mixes were intentionally designed to limit the amount
of cement (binder material) so that the effects of the adsorbent (primarily
the organically modified clays) could be observed. With the addition
of more binder material, it is believed that the strength of the treated
product would increase.
Organically Modified Clay Mixes
The mixes containing an organically modified clay include 24 of the
50 stabilization mixes. The as-compacted strengths for the organoclay
mixes are similar to those of all of the mixes. However, after the
organophilic clay mixes had been cured, their strengths were higher
than those of the other mixes. The ability of the organophilic clay to
adsorb the organics reduces the organic inhibition for the cement hydra-
tion process.
In the chemical analyses, the carbon analyses, phenol concentrations
and sum of organics are lower for the organoclay mixes than for the
entire data set but the relative hydrocarbon concentration averages are
approximately the same. The results for the mixes containing an
organically modified clay generally are favorable with respect to con-
taining the organic constituents of the sludge.
Table 2.
Test Average for Stabilization Mix Goupings
GROUPING
ALL MIXES
FLY ASH/LIME
CEMENT
ORGANOCLAYS
BENTOH I TE/ATTAPULG I TE
SOOIUH SILICATE
CEMENT KILN DUST
PROPRIETARY
NO. OF
MIXES
48
15
23
24
15
6
5
7
INITIAL
POCKET.
PENETR.
(psi)
10.9
15.3
10.0
13.4
13.7
11.8
5.3
5.8
VOLUME
CHANGE
«)
44.6
45.5
42.3
45.1
38.7
58.7
49.4
42.2
WET
DENSITY
(9/cm3)
1.34
1.29
1.36
1.27
1.33
DRY
DENSITY
(9/cm3)
1.03
1.00
1.02
0.97
0.99
1
1.38 I 1.06
1
1.50 | 1.20
1
1.35 I 1.03
CURED
POCKET
PENETR.
(psi)
38.6
35.9
38.9
43.6
33.3
51.0
42.4
30.6
use
(psi)
15.5
11.7
16.9
21.5
10.9
22.4
9.5
18.1
UNIT
STRAIN
(X)
7.3
8.1
8.1
6.0
10.1
3.5
TOTAL
ORGANIC
CARBON
(PPM)
118.15
130.14
112.61
76.21
172.53
111.23
4.7
9.1
156.84
105.87
TOTAL
CARBON
(PPM)
130.92
147.84
118.04
81.03
RHC
(PPM)
1.32
1.42
1.21
1.33
190.39 | 1.13
121.17
198.25
110.97
PHENOL
CONC.
(PPB)
289.18
418.36
292.17
248.91
SUM OF
ORGANICS
(PPB)
512.73
770.00
478.70
439.00
I
461.10 I 721.67
1.45 197.00
457.33
I
1.44 258.20 I 627.60
1.04
69.17
1 256.14
HAZARDOUS MATERIALS TREATMENT 293
-------
Bentonite and Aitapulgite Mixes
The mixes which contained an unmodified clay, either bentonite or
attapulgite, comprised 15 of the 50 stabilization mixes. The unmodi-
fied clay mixes were weaker and more plastic, on average, than all of
the mixes being studied.
With regard to the chemical analyses, the average values for the
carbon, phenol and organics summation analyses are higher for the
unmodified clay mixes than for the entire set of mixes. However, the
average relative hydrocarbon concentration value for these mixes is
slightly lower than for all of the mixes.
Soluble Sodium Silicate Mixes
Of the 50 mixes prepared in these investigations. 6 contained solu-
ble sodium silicate. The average cured strength of the silicate mixes
is higher than the average cured strength of all of the mixes. The solu-
ble sodium silicate brought about a noticeable increase in the strength
of the stabilization mixes. The chemical analyses revealed that the
average test parameter values for the soluble sodium silicate mixes did
not significantly vary from the average values for the entire set of mixes.
It is concluded, therefore, that, should strength increases be needed,
soluble sodium silicates may be added to the stabilization mix.
Cement Kiln Dust Mixes
Cement kiln dust mixes accounted for five of the 50 stabilization mixes
analyzed. The average computed bulk densities for these mixes were
greater than the overall mix averages. In the case of most soil-like
materials, denser materials correspond to greater strengths but these
values indicate that, for the cement kiln dust mixes , the densities are
higher, on the average and the u neon fined compressive strengths are
lower than the entire set of mixes.
The chemical parameters indicate that the average leachate total
organic carbon values for the cement kiln dust mixes are higher than
the overall average mix values. This indicates that, with respect to the
carbon analyses, the cement kiln dust mixes arc less effective in con-
taining the organic constituents of the sludge than the entire set of mixes.
Proprietary Mixes
The proprietary mixes accounted for seven of the 50 mixes tested.
These mixes included ingredients from American Colloid Company,
Waste Solutions International and Silicate Technology Corporation,
These mixes were grouped together to compare the results of commer-
cial mixes to those of generic products. Overall, the proprietary mixes
are not as strong as the other mixes.
The chemical analyses reveal that, for the proprietary mixes, the
average values for relative hydrocarbon concentration, phenol concen-
tration and summation of organic constituents are lower than the over-
all average mix values. A few of the proprietary mixes were exceptionally
soft and plastic and, as a result, did not provide an acceptable surface
area for leaching in the TCLP.
These leach test results are artificially low as a result of the plastic
nature of the material. In general, the data revealed that the proprietary
mixes contained the petroleum sludge to a greater degree than the entire
set of mixes.
TEST RELATIONSHIPS
The strength of the samples, as measured by a pocket penetromeier
and an unconfined compression test, was analyzed. Figure I. As-
Compacted Pocket Penetrometer vs. Unconfined Compressive Strength,
did not reveal a well-defined relationship. This result demonstrates that
the pocket penetrometer is of limited usefulness in predicting the
unconfined compressive strength of these stabilized sludges.
The relationship between the chemical test parameters and the
unconfined compressive strength of the samples was also examined.
The total organic carbon was plotted against the unconfined compres-
sive strength in Figure 2. No relationship was apparent between these
two test parameters. The graph of relative hydrocarbon concentration
versus unconfined compressive strength is shown on Figure 3. The data
are scattered, with relative hydrocarbon concentration values varying
for similar unconfined compressive strength values. The stronger sam-
^^
o.
t
Tu
Ł
c
1 1 1 ) t .
0 10 20 30 40 SO 60 70 80 90 100
Unconfined Compressive Strength (psi)
Figure 1,
As-Compacted Pocket Penetrometer vs. Unconfined Compreisrve Strength
pies had similar relative hydrocarbon concentration values which were
approximately in the middle of the range of recorded values.
^ 2*0-
"? 220-
CL
jx 200-
g ISO-
1 ISO
,5 i«o-
o 120-
e 100-
cT ao'
•5 60-
o 40-
*" 20-
o
•
^ *•* •
%*'
••%• • .
" ,
» V V
»» •» ^
»». ' V
• • »
.*
1 1 1 1 D 1 1 1 t
0 10 20 30 40 50 60 70 80 00 100
Unconfined Compressive Strength (psi)
Figure -
Tolal Organic Carbon vs. Unconfined Compressive Strength
4.0
35
3.0
2.5
2.0
1.5
1.0
0.5
0.0
t
* »
w *
0 10 20 30 40 SO 60 70 80 90 100
Unconfined Compressive Strength (psi)
Figure 3.
Relative Hydrocarbon Concentration vs.
Unconfined Compressive Strength
Phenol concentration data and unconfined compression test data were
compared in Figure 4 There was no distinct relationship between these
two test parameters.
The relationships between the chemical test parameters were also
investigated. Total organic carbon and total carbon concentrations were
plotted on Figure 5. A strong relationship existed for these data. For
lower values, the total carbon values were approximately equal to the
total organic carbon values. As the concentrations increased, the total
294 HAZARDOUS MATERIALS TREATMENT
-------
'ft
p>
o.
C
o
•3
Ł
M
?
u
u
C
o
u
"o
C
V
Ł
CL
700-
600-
500-
400-
300-
200
100
0
y
v V
V
9V ^f
v _
0
V V
' V
. 77
V 7
• *'*
•" "*.
^
T g |*T 1 1 1" ' 1 1 1 1 1
0 10 20 30 40 50 60 70 80 90 100
Unconfined Compressive Strength (psi)
Figure 4.
Phenol Concentration vs. Unconfined Compressive Strength
carbon values were slightly greater than the total organic carbon values.
In general, most of the carbon measured in the leach extract was or-
ganic carbon.
280
6 240-
A
o.
~C 20°"
o
.o
a 160-
u
I 120-
a
a 80-
o 40-
0--
40
80
120
160
200
240 280
Total Carbon (ppm)
Figure 5.
Total Organic Carbon vs. Total Carbon
Since total organic carbon was related to the sum of organics, Figure 6
and the sum of organics was weakly related to the relative hydrocarbon
concentration, Figure 1, it was anticipated that the total organic carbon
and the relative hydrocarbon concentration would be related in some
way. As shown in Figure 8, there is no distinct relationship between
the total organic carbon values and the relative hydrocarbon concen-
tration values. The data points are scattered, with low relative hydrocar-
bon concentration values existing with both high and low total organic
carbon values. This finding gives rise to the dilemma as to which
parameter is a better indicator of performance, total organic carbon
or relative hydrocarbon concentration.
In summary, there is not a well-defined relationship between strength,
as determined by an unconfmed compression test and organic concen-
tration in the mix extract, measured by total organic carbon, total carbon,
relative hydrocarbon concentration, phenol concentration or summa-
tion of organic concentration. Mixes with high strengths did not neces-
sarily have low total organic carbon or relative hydrocarbon
concentration values. As expected, when greater quantities of cement
were added, the strength of the stabilization mix increased. However,
it was not necessarily true that the stronger mixes prevented contaminants
from leaving the solidified matrix during a leach test. The possible
exception may be indicated by the observation that at strengths above
35 psi, leachability generally was reduced.
MIX REPRODUCIBILITY
The ability to obtain consistent test results for the same mix design
1500 +
I
.2 1000 +
o
'o
G
in
500 +
50 100 150 200 250
Total Organic Carbon (ppm)
Figure 6.
Sum of Organics vs. Total Organic Carbon
300
1500 +
.o
o,
p.
.5 1000
500 +
Relative Hydrocarbon Concentration
Figure 1.
Sum of Organics vs. Relative Hydrocarbon Concentration
HDU -
240 -
T 220-
J| 200-
C 180-
5 160-
5 1*0-
.2 120-
is 100-
| 80-
•a 60
o 40
*" 20
n
V
"7 v TV'
7 V TO '
„» * *
VS v
' 7
V » v
7 V 7 V
v v vv v7 v
V V ,V v * ' '
W v V
7
1 1 1 1 1 1 1 — — —
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Relative Hydrocarbon Concentration
Figure 8.
Total Organic Carbon vs. Relative Hydrocarbon Concentration
4.0
was evaluated by testing replicate samples. All of these mixes are shown
in Table 3, along with the corresponding test results. In general, the
physical test parameters reveal that good reproducibility exists among
the replicate samples. The chemical results also reveal that there is
generally good reproducibility. In summary, this limited replicate testing
indicates a high degree of confidence in the results of the chemical
analyses.
HAZARDOUS MATERIALS TREATMENT 295
-------
Table3
Mix Reproduclbillly
MIX NO.
6
21
17
30
46
47
48
49
50
REPLICATE NIXES AND BLANKS
NIX COMPOSITION
Sludge/Att«pulg (1/1.2/0.75)
Deioniied Water/Be
ntonite/Cemem (1) (1/1.2/0.71)
Sludge/Cement < 1 )/Bemonlt«/Fly Ath (1/0.4/0.1/0.1)
Sludge/C
Sludge/C
encnt (1 VBentonl te/Fly Aih (1/0.4/0.1/0.1)
ement (l)/8entonl tt/Fly Ath (1/0.4/0.1/0.1)
MIX NO.
6
21
TOTAL
ORGANIC
CARBON
(PP«)
133.1
199.2
RELATIVE J
HYDROCAR80
CONC.
(RHC)
1.708
2.592
17
30
150.8
173.3
46
47
2.7
2.5
48
49
50
139.4
161.2
155.6
0.596
0.431
UNTREATED SLUDGE
F*
7.0
6.0
6.0
6.0
•
•
S.O
6.0
6.0
BULK
DENSITY
<9/o«3)
0.99
1.01
1.02
1.01
•
•
.04
.03
.03
WATER
CONTENT
<*_>...
43.3
47.7
24.4
44.7
•
•
J7.6
36.4
36.8
LOSS ON
TREATED SLUDGE (2-WEEK CURE))
IGNITION pH
(X)
88.8
93.2
90.9
87.9
•
•
86.0
87.5
87.6
9.0
11.0
WET
DENSITY
(g/cn3)
1.26
1.29
dryi 1.33
dry 1.23
<""f\
dry
8
8
g
1.34
1.59
1.49
DRY
DENSITY
(9/ca3)
0.87
0.95
0.95
0.85
0.98
1.09
1.18
1.4J 1.10
1.45 1.07
WATER
CONTENT
(X)
30.8
26.0
28.2
31.2
26.8
31.6
20.8
22.9
25.9
1
1
LOSS ON
IGNITION
(X)
J5.7
16.8
40.2
44.7
24.8
7.3
36.1
14.1
32.6
UCJ
(Pfl)
23.13
30.52
6.22
4.29
505.61
266.76
1.59
1.96
1.75
M
PHENOL
(PPB)
348
638
650
NO
i
0.159
0.111
0.712
0.608
0.540
SELECTED CHENICAL ANALYSES OF HE TREATED SLUDGE (2-WEEK CURE)
OECANE
(PPB)
NO
4
15
NO
BENZYL
ALCOHOL
(PPI)
38
NO
3
NO
•
•
•
•
*
) i
NETHYL
PHENOL
(PPi)
257
BENZYL
ACETATE
(PPB)
NO
NAPHTHALENE
(PP8)
12
614 29 9
NO
NO
OIBUTYL
PHT HAUTE
(PPB)
NO
NO
CHROXIUM
(PPH)
0.02
0.07
LEAD
(PPM)
0.00
1
NICKEL CADMIUM
(PPM) ! (PPM)
0.00 0.00
0.00 0.00 0.00
i '. i: f
29
NO
7
NO
I
•
•
•
•
•
NO 0.64
2 I 0.50
•
•
0.00
0.27
•
| •
•
0
11
0.00
0.00
0.10
0.00
0.00
0.10
o.oo
0.00
0.00
0.00 0.00
0.20 i 0.00
COPPER
(PPM)
0.10
0.00
ZINC
(PPM)
0.02
0.22
0.00
1.75
0.03 3.00
! ;
^ 0.05 0.00
0.34
0.32
0.00
0.00
0.00
0.00
0.43 0.00
0.00
0.07
0.02
0.00
0.00
0.01
0.37
0.35
0.76
0.95
NON-DETECTABLE
NOT AVAILABLE
CONCLUSIONS AND RECOMMENDATIONS
Test Applicability
The stabilization mixes in this study were evaluated based upon a
chemical leach test (a modified form of the Toxicity Characteristic
Leaching Procedure) and a strength test (unconfined compression test).
The results from these tests did not correlate well with each other. There
was not a well-denned relationship between strength and organic con-
centration in the TCLP mix extract, measured by total organic carbon,
total carbon, relative hydrocarbon concentration or specific compound
concentrations.
Stabilization Mix Investigations
From the investigations described herein, the effectiveness of a variety
of stabilization mixes was assessed and some general conclusions were
reached:
• Mixes with greater quantities of cement generally were stronger
However, the stronger mixes did not necessarily prevent contaminants
from leaching from the solidified matrix during the leach test.
• The mixes containing fly ash did not have high strengths and did
not effectively immobilize some of the organic contaminants, as
reflected by the relative hydrocarbon concentration values and the
phenol concentration values.
• The unmodified clays, attapulgite and bentonite, also had little success
in containing the organics, as indicated by their total organic carbon
values.
• Soluble sodium silicates may be added to the stabilization mixes to
increase strength.
• The microfme cement mixes provided some success with regard to
containing the hydrocarbons, as quantified by the relative hydrocarbon
concentration values but had mixed success with regard to other
organic parameters.
• In general for the mixes which contain an organophilic clay, the more
expensive the mix, the better the mix performance with regard to
the measured test parameters.
The organically modified clays have shown the most promise and
will be investigated further in order to optimize their impact on organic
contaminant immobilization. These clays, used in conjunction with some
type of binder material such as ponland cement, may provide the sys-
tem necessary to adequately stabilize and solidify organic-bearing
hazardous wastes.
The principal problem with the organophilic clays is their high unit
costs. These clays were much more expensive than any of the other
stabilization agents and. although technically more successful than some
of the other mixes, were not economically feasible as a treatment
alternative. As a result, mixes with lower proportions of organophilic
clays will be investigated in an attempt to reduce the total mix cost while
effectively stabilizing the acidic petroleum sludge.
The conclusions reached in this paper are based upon the initial lest
data, which have limited statistical significance. Further testing is neces-
sary to strengthen and confirm these findings. Also, these conclusions
may only be directly applicable to the specific petroleum sludge used
m these studies. However, the findings may be useful as a starting point
for stabilization studies involving other types of hazardous organic
wastes.
ACKNOWLEDGEMENTS
These investigations were funded by the Ben Franklin Partnership
of Pennsylvania, the Sun Refining and Marketing Company and the
Earth Technology Corporation. The following individuals were involved
296 HAZARDOUS MATERIALS TREATMENT
-------
in the project: Dr. Arthur Raymond, Dr. Michael LaGrega, Dr. Elaine
Keithan, Kate Toner, Jim Spriggle, Holly Borcherdt, Lewis "Bud" Albee
and Jason Strayer.
REFERENCES
1. Pancoski, S.E., Evans, J.C., LaGrega, M.D. and Raymond, A., "Stabiliza-
tion of Petrochemical Sludges," Hazardous and Industrial Waste, Proceedings
of the Twentieth Mid-Atlantic Industrial Waste Conference, Ed. Ma. M. Varma
and J. H. Johnson, Jr., June, 1988, pp. 299-316.
2. Pancoski, S.E., "Stabilization of Petroleum Sludge," Masters Thesis, Civil
Engineering, Bucknell University, Lewisburg, PA, May, 1989, University
Microfilms International.
3. Van Keuren, E.L., "Advanced Laboratory and Pilot Field Study of Hydro-
carbon Refining Waste Stabilization with Pozzolans," Masters Thesis, Civil
Engineering, Drexel University, Philadelphia, PA, June 1987.
4. Evans, J.C., LaGrega, M.D., Pancoski, S.E. and Raymond, A., "Metho-
dology for the Laboratory Investigation of the Stabilization/Solidification of
Petroleum Sludges," Superfund '88, Proceedings of the Ninth National Con-
ference, Hazardous Materials Control Research Institute, Silver Spring, MD,
pp. 403-408, November, 1988
5. Evans, J.C.; and Pancoski, S.E., "Organically Modified Clays," Transpor-
tation Research Board Publication, Preprint Paper No. 880587, Jan, 1989.
HAZARDOUS MATERIALS TREATMENT 297
-------
Composting as a Method for Hazardous Waste Treatment
Michael K. Mays, RE.1
Westinghouse Environmental Services
Atlanta, Georgia
Dr. Lawrence J. Sikora, Ph.D.
U.S. Department of Agriculture
Beltsville, Maryland
James W. Hatton1
Shirley M. Lucia1
Westinghouse Environmental Services
Atlanta, Georgia
ABSTRACT
Westinghouse Environmental Services and Geoiechnical Services,
Inc (Westinghouse) conducted a bench-scale compost treatability study
of an ethylene glycol waste sludge from a CERCLA site associated with
a fiber manufacturing plant. The study was performed as pan of a feasi-
bility study of remedial alternatives for disposing of the sludge. The
site required remediation because leachate from the landfill contami-
nated an aquifer supplying drinking water to nearby residences. The
waste sludge's composition changed with location in the landfill, but
contained up to 3.200 mg/kg ethylene glycol, 128.000 mg/kg TOC and
6,400 mg/kg antimony. The objectives of the treatability study were:
• To evaluate the potential for biodegredation of the ethylene glycol
sludge using the compost process
• To determine the degree to which metals in the sludge would be
immobilized
• To determine sludge to bulking agent mix rates and other operating
parameters
Results indicated that more than 99% of the BOD, COD and TOC
were removed. Ethylene glycol was reduced by more than 94%. With
the exception of Barium, RCRA metals were not leachable. However.
Barium in the leachate was below the U.S. EPA EP toxicity standard
of 100 mg/L. A mix of 15% waste sludge to wood chips, by weight,
was identified as an optimal mix for the composting process. Pile
temperatures of up to UOaxaF were observed for that mix.
INTRODUCTION
Composting is a biological process used primarily for the stabiliza-
tion of organic materials that are relatively high in volatile solids
such as manures and sludges (Sikora and Sowers, 1985)'. The
degradation of the volatile solids results in the production of heat
and in a subsequent temperature increase characteristic of the
composting process. The use of composting as a means of degrading
organic industrial wastes has received considerable attention recently.
An in-vessel composting system may be beneficial for the degradation
of toxic constituents that may be subject to volatilization and/or
leaching. Rose and Mercer (1986)3 found that the insecticides
diazonin, parathion. and dieldrin degraded rapidly when composted
with cannery wastes. Deever and White (1978)' found significant
reductions in toluene-hexane extractable grease and oil after
composting petroleum refinery sludges. Sikora et al. (1982)' showed
in a preliminary laboratory study that composting however was no
more efficient in degrading pentachlorophenol and pentachloro-
nitrobenzene than degradation at a constant temperature of 25 "C.
Although composting as a treatment method for municipal wastes
and sludges has been done for some time, (Wilson and Dalmat,
1984)5, composting of industrial and hazardous wastes is rarely
(1) Westinghouse Environmental and Geotechnical Services, Inc.
(2) U.S. Department of Agriculture, Agricultural Research Service
Beltsville, Maryland
done. This paper details the bench-scale testing procedures and results
of a study conducted by Westinghouse Environmental and Geotech-
nical Services. Inc. (Westinghouse) on the composting of ethylene
glycol wastes at a Superfund site.
BACKGROUND
A remedial investigation (Rl) conducted by Westinghouse in June,
1986, found groundwater contamination linked to material landfilled
at the industrial plant in North Carolina. The primary source of the
contamination was wastes from the Glycol Recovery Unit (GRU)
buried in trenches during the early 1960s. The Rl, feasibility study
(FS) and remedial design (RD) for groundwater remediation at the
site have been completed and are the source of another paper. This
paper focuses on the treatment of the GRU material.
The FS for the source material identified composting as a method
applicable for ethylene glycol waste treatment. To verify that theory.
a treatability study was conducted on the GRU material.
The objectives of the treatability study were:
• to assess the ability of composting to degrade organic materials in
the GRU sludge
• to determine the proper mix ratios of GRU sludge to bulking agent,
inoculants, etc.
• to determine operational parameters such as run time, optimal pH
and moisture content, peak temperature, etc.
• to determine if the resulting residual from the composting operation
system would be classified as hazardous based on U.S. EPA's definition
of hazardous waste
• to determine operational efficiencies for the development of cost
estimates for full-scale operation
MATERIALS AND METHODS
The GRU material used in this study was white to grey in color and
had a consistency ranging from soft cottage cheese-like material to dry
friable material. The GRU material had an average content of ethylene
glycol, chemical oxygen demand (COD), and total organic carbon (TOC)
of 1,900 mg/kg. 120.000 mg/kg and 120,000 mg/kg, respectively.
Additives used for composting included wood chips, top soil and
sewage sludge, dry molasses, ammonium nitrate, 10-10-10 fertilizer and
lime. The purposes of the additives are as follows:
• The wood chips were added as a bulking agent and an additional
source of carbon.
• The top soil and sewage sludge were added as a source of diverse
microorganisms or "inoculants."
• Dry molasses was added to provide a readily available carbon source
to stimulate microbial growth during acclimation.
• Ammonium nitrate and 10-10-10 fertilizer were added as a supple-
mental source of nitrogen, potassium and phosphorus.
• Lime was added periodically to maintain the pH above 6.0 for
maximum microbial activity.
Reactors were constructed using 32-gal plastic trash cans. Air was
298 HAZARDOUS MATERIALS TREATMENT
-------
supplied using a blower and PVC pipe diffuser and distribution system.
Air was controlled using valves located at each reactor. Excess moisture
was removed using an underdrain. Ten reactors were used for this study.
A typical reactor is shown in Figure 1.
18" 01A—.
PVC
TOP view
ORAIN
3.0'
SIDE VIEW
Figure 1
Compost Reactor Design
The reactors were set up to simulate composting using different con-
centrations of GRU material and wood chips. Parallel reactors at 20%,
10%, 5%, 2.5% and 0% GRU material by volume were evaluated. The
reactors containing no GRU served as controls. Table 1 contains data
on the compost mixes.
Table 1
Compost Mix Ratios
Ratio of
GRU Sludge
to Uood
Chips
0:1001
2:5:97, SX
5:«X
10:90X
20; 60H
Gallons
of GRU
Reactor Sludge
1, 10 0
8, 9 1
6, 7 2
2, 3 4
4,5 8
Gallons
of Wood
Chips
40
39
36
36
32
Gallons Gallons
of Dry of Seuage Gallons
Molasses Sludge of Soil
5 2 3
223
2 2 3
2 2 3
223
Cups
of Cup
Ammonium 10:
Nitrate Fer
1 1/2
3/4
1
1 1/4
^ 3/4 i
S of
0:10
ilizer
4
2
4
4
2
The samples were mixed in a portable mixer. To thoroughly coat the
wood chip particles, the GRU sludge material was first mixed with tap
water until fluid. Fertilizer, lime and molasses were added to the
liquefied GRU sludge. This mixture was then blended in the mixer with
the wood chips. The sewage sludge and soil were then added. Water
was added into the mixer to a predetermined moisture content. A 20-gal
sample of the mix was placed in the reactor.
For process control, temperature, percent oxygen, moisture and pH
were monitored throughout the test. The temperature was monitored
daily as an indication of biological activity. The percent oxygen was
monitored daily, and the air flow was adjusted so that at least 10% oxygen
was maintained in the exhaust gases. Moisture and pH were measured
once a week to assure that the pH was maintained above 6 and that
the moisture was sufficient for microbial growth (40-60%, w/w) but
not in excess so as to result in anaerobic conditions.
Analyses were conducted using methods similar to that found in
Standard Methods (1975)6 to determine:
• removal efficiencies of biochemical oxygen demand (BOD), COD
and toxic organic compound (TOC)
• degradation rate of ethylene glycol
• nutrient levels - to assure sufficient nutrients to maintain growth
• initial and residual compounds and the mobility of those compounds
RESULTS
Summary of removal efficiencies is presented in Table 2. Figures 2,
3, 4 and 5 present changes in BOD, COD, TOC and ethylene glycol
concentration plotted against the test time. The data indicate:
• Removal efficiencies generally were in excess of 95% for the
parameters evaluated.
• The removal of the ethylene glycol was accomplished in less than
30 days with all reactors except the 20% mixtures. The 20% mix-
tures took considerably longer for acclimatization than the other
reactors.
A consistent and strong increase in temperature was observed in all
reactors except the 20% mixture indicating microbial activity. Reactor
temperatures exceeded 105 °F which is indicative of the biological
activity and the insulation of the mass in the reactor. Percent oxygen
in the reactors which proved useful for controlling process airflow
rate also indicated significant microbial activity throughout the test.
Tkble 2
Contaminant Removal Efficiencies
Reactor BOD
Removal
Percent
1 97.9
2 96.5
3 88.9
4 82.
5 75.
6 98.6
7 96.6
8 93.4
9 96.3
10 98.4
H.OO
„ 12.00'
GT i
x
o 10.00
_ B.OO
t—
<
-------
8/3/88
— 0 X GRU
•»• 10 % CRU
8/10/88
8/22/88
9/20/88
•*• 2.5 X GRU
-»- 20 X CRU
Figure 3
TOC vs, Time
GRU Compost
• 5 X GRU
12.00
9/20/88
•«- 2.5 X G"U
•*• 20 x GRU
Figure 4
TOC vs. Time
GRU Compost
• 5 x cnu
3"
<
H
i
<
ec.
\
8
z
_J
10.00
8.00
800
4.00
2.00
k — • — ' " ' " "~~— _^_* " n^
^~^v X^^
''•••. V- ^\
X . ^^
**•-.
" *,
*•--
~"~ -™~ *-,
"^
7/20/88 8/3/88 8/10/88 8/22/88 9/20/8B
— 0 X GRU -2S1GRU «SXCRU
*- 20 % GRU
Figure 5
Ethylene Glycol vs. Time
GRU Compost
Moisture and pH controls were adequate for the microbial growth
during the test. Results of pH monitoring indicate that a pH of 7
to 8.3 results in rapid BOD reduction, that the pH of the reactors
dropped approximately one pH unit during the course of the process,
and that 100 grams of lime a per pound of GRU sludge is adequate
to maintain the pH. Moisture content did not appear to have a sig-
nificant effect on the test as long as it was kept within a range of
40% 10 60%.
Composting with a bulking agent would necessitate a screening step
to reduce the volume of the final product for disposal. Sieve analysis
indicated that 30% to 40% of the final product would pass a No,
4 sieve. Literature suggests that an overall volume reduction of 20%
to 30% can be achieved during composting (Sikora et al., 1981);'
however, in the bench-scale tests a 10% increase in volume was
observed, probably because of handling of the mixtures to add lime
that took place during the runs.
EP Toxicity cxtractable metals analysis indicated extractable, RCRA
controlled, metal levels present in the compost equal to that in the
control. The concentrations allow the compost to be land disposed.
Extractable Target Compound List (TCL) compounds were detected
in the compost but only at concentrations from about 700 ug/kg and
3700 ug/kg. Subsequent analysis of screened compost materials in-
dicated the presence of benzoic acid and di-n-butyl phthalate near
their respective detection levels of 1000 ug/kg and 230 ug/kg. The
concentrations of these components would allow land disposal of (he
compost.
CONCLUSIONS
Data indicate that microbial activity was occurring in the compost
reactors, and that over 95% removal rates is achievable for the
parameters evaluated. All mixes composted rapidly except the 20%
mix by volume, which experienced a considerable lag period before
showing microbial activity.
Based on the data a number of conclusions can be drawn:
• Composting is possible for use in treating the GRU sludge in
preparation for land disposal
• The 10% (volumetric) GRU mix would be the best mix
• Process control data and monitoring procedures for composting were
developed
• Analytical data indicate the final product can be landfilled or land
applied
RECOMMENDATIONS
Although bench-scale test results show substantial degradation of the
organic components of the ethylcnc glycol wastes, pilot-scale testing
is recommended prior to full-scale use. To further test the applicability
of this technology and further define operational parameters, the
following tests are recommended:
• Conduct tests using concentrations between 10% and 20% GRU by
volume to determine a possible higher concentration than 10% which
can achieve give acceptable results
• Test the effect of an acclimated seed on the process
• Field test mixing and sieving equipment
• Consider the use of a mechanically-mixed system versus the static
pile method
• To confirm test results, send a sample of the wastes to a commercial
vendor of composting equipment for testing
REFERENCES
I. Sikora, L. J. and Sowers, M.A. "Effect of temperature control on the com-
posting process." J. Environ. Qual. W:434-438, 1985.
2 Rose, W. W. and Mercer. W. A. Fate of pesticides in composted agricultural
wastes, National Canners Association, Washington, D.C., 27p., 1968
3 Dcevcr. W. R. and White, R.C. Composting petroleum refinery sludges.
Texaco, Inc., Port Arthur. TX p. 24, 1978.
4 Sikora. L. J., Kaufman. D D., Ramirez, M.A. and Willson. G.R Degrada-
tion of pentachlorophenol and pentachloronitroabenzene on a laboratory
composting. In Pmc. of the Wight Ann. Research Symp., Cincinnati. OH.
EPA-600/9-82-002, pp. 372-381, 1982.
5. Willson, G. B. and Dalmat, D. Sludge composting facilities in the U.S.A.
Biocyrle J. of Waste Recycling, Sept/Oct issue, pp. 20-24, 1983.
6. Standard Methods For The Examination Of Hbler And MfaMnwuer. Mth Ed.
Arncr. Public Health Assoc. Washington, D.C . 1193p.
7. Sikora. L. J., Willson, G.B., Colacicco D. and Parr, J.F. Materials balance
in aerated static pile composting. Jour. Mfeier Ml. Control Rd. 53:1702-1707,
1981,
300 HAZARDOUS MATERIALS TREATMENT
-------
Recycling of Battery Casings At A Superfund Site
David A. Tetta
U.S. EPA
Seattle, Washington
ABSTRACT
The NL/Gould site is a former battery recycling facility located in
Portland, Oregon. As secondary lead smelting facility was in operated
on the site between 1949 and 1980. Facility operations consisted of:
lead-acid battery recycling; lead smelting and refining; and lead oxide
production. Approximately 80,000 tons of battery casing materials
remain on-site.
The Record of Decision for this site includes predesign studies to:
• Define recyclability criteria for the casings that will be used to
determine the volumes that can be recycled
• Determine process requirements to separate casings in a manner that
minimizes fugitive emissions
• Determine the modifications required to adapt existing separation
technology to conditions at the site
Predesign studies currently are being performed. The bench-scale
test program indicated approximately 20% of the waste on the site is
recyclable by separation alone. This recyclable material consists of lead
oxide/sulfate sludge (17%) and a lead concentrate (3%). Additional
cleanup can be achieved by treating the excavated material. At the present
time, approximately half of the waste would require stabilization in order
to be left on-site. This material includes 14% matte and 23% that is
treatment products.
INTRODUCTION
The Gould uncontrolled hazardous waste site is located in the Doane
Lake area of Portland, Oregon. A secondary lead smelting facility went
into operation in 1949. Activities included lead-acid battery recycling,
lead smelting and refining, zinc alloying and casting, cable sweating
(removal of lead sheathing from copper cable) and lead oxide produc-
tion. Operations continued under a variety of owners until 1980. In 1981
U.S. EPA and State of Oregon DEQ began investigating the site; it was
placed on the NPL in 1983. NL Industries, Inc. and Gould, Inc., under
an administrative order on consent with U.S. EPA, contracted with
Dames & Moore to perform an RI/FS.
During the smelter's period of operation, of 86,900 tons of battery
casings and 6,570,000 gal of battery acid were estimated to have been
disposed of at the site. In addition to acid and battery casings, a third
waste product called matte was produced by the smelting operation.
Matte disposal was estimated at 11,800 tons'
The battery casings consist of hard rubber, ebonite, plastic casings,
metallic lead, lead oxides and associated soil and debris. Lead con-
centrations (mostly lead oxide) ranged from 7,600 mg/kg (0.76%) to
190,000 mg/kg (19%). All of the battery casing samples had EP Toxicity
(EP Tox) results for lead above the regulatory limit of 5.0 mg/L. These
values ranged from 21 mg/L to 220 mg/L1.
Figure 1 shows the locations the casings and other wastes. Approxi-
mately 2% of the total volume of battery casings is located in surface
piles on the Gould property, the remaining 98% is located as fill on
the Gould and adjacent properties and in the sediments of East Doane
Lake. The subsurface casings are in direct contact with groundwater
underneath the site. The characteristics of the surface piles of casings
differ somewhat from the subsurface piles. During the RI, the surface
piles were found to contain a higher percentage of plastic and metallic
lead relative to subsurface casings on the Gould property or from the
Rhone-Poulenc property, which contain a higher percentage of rock
and slag. The metallic lead, plastic, ebonite and lead oxide components
of these casings have been considered potentially recyclable. The esti-
mated fractions of the various components in the surface and subsur-
face casings as determined in the RI are shown in Table 1.
The matte materials consist of metallic sulfide chunks containing
primarily iron and lead. Lead concentrations in the matte samples ranged
from 6.4% to 11%. All of the samples had EP Toxicity results for lead
above the regulatory limit of 5.0 mg/L. Low concentrations of arsenic
and cadmium also were detected in the EP Toxicity leachates. These
concentrations were within the regulatory limits (5.0 mg/L and 1.0 mg/L,
respectively).
In addition to battery casings and matte, large quantities of soil at
the site are contaminated with lead and can serve as secondary sources
for lead transport. The quantity of surface soil at the site considered
to be a secondary source is approximately 3,400 yd3. The volume of
subsurface soils estimated to be a secondary source is 12,800 yd3.
Sediment samples collected from East Doane Lake contained total
lead concentrations ranging from 160 mg/kg to 12,000 mg/kg. Based
on these results, the estimated quantity of contaminated sediment in
East Doane Lake is 5,500 yd3.
Lead contamination at the site has impacted groundwater in the
shallow fill aquifer as well as an alluvial aquifer deeper down. Lead
concentrations at points in these aquifers has exceeded the MCL for
lead of 0.05 mg/L. Total lead migration from the site into the ground-
water is estimated to be from 0.3 to 0.6 Ib/yr1
U.S. EPA'S DECISION ON REMEDIATION
In March, 1988, U.S. EPA issued an ROD for this site. The remedy
that U.S. EPA selected focused on attempting to recycle the battery
casings at the site. It included:
• Excavation of all of the battery casing fragments and matte from the
Gould property and adjacent properties where casings have been
identified
• A phased design program to determine the amount of material that
can be recycled and to minimize the amount of material that must
be RCRA landfilled
• Separation of the battery casing fragments
HAZARDOUS MATERIALS TREATMENT 301
-------
• Recycling of those components (or portions of components) that can
be recycled, off-site disposal for non-recyclable components that fail
the EP Toxicity test and on-site disposal of non-hazardous, non-
recyclable components
• Excavation, fixation/stabilization and on-sitc disposal of the remaining
contaminated soil, sediment and matte
SURVEY OF RECYCLING AT SUPERFUND SITES
Prior to beginning predesign studies at the Gould site, a survey of
battery recycling attempts at waste sites and general industry capabili-
ties was performed1. Several previous recycling and/or separation
attempts on battery scrap piles were identified by personnel from the
U.S. EPA and other environmental agencies. These prior recycling
attempts include efforts at the Sapp Battery site in Marianna, Florida.
in September, 1984, to separate approximately 4,000 yr' of battery
scrap. Separation equipment also was used to conduct an engineering
study on separating battery scrap components at the Granite City.
Illinois, Superfund site. Neither of these attempts to recycle ebonite
casings was successful. An attempt to use a commercial facility to recycle
casings from the Gould site was also performed during the RI/FS. The
casings did not pass EP Ibx.
A number of commercial vendors were contacted in the search for
a process that could recycle the Gould battery casings. Several facili-
ties feed the ebonite component of the battery casings directly to a
smelting furnace as a source of fuel and carbon. When this is done,
the lead content of the ebonite is not a factor; in fact, higher lead con-
tents of the feed to the furnace are desired to make the process more
profitable. Most of these companies expressed reluctance to accept the
Gould battery casings because the amount of recoverable lead in the
ebonite is low and it would slow down lead production capacity1
Current industry recycling practices are shown in the generalized
process flow diagrams in Figure 2. The battery casings pass through
several steps designed to protect the process equipment by removing
large rocks, chunks of slag, large pieces of scrap metal and other debris
such as automobile bumpers, discarded equipment and wood. These
steps usually include an electromagnet to remove ferrous metals, a screen
to remove large items and a manned inspection station to further remove
possibly damaging materials.
A hammer mill reduces the remaining material to 1/2-in. to 1-in par-
ticles and helps loosen and remove some of the lead oxide caught in
cracks covering the surface of the ebonite. The particles are passed over
a screen and washed with various agents such as water, surfactants or
acid, with water only being most common of the processes. The solu-
tion washes out the fine particles of lead oxide and soil, which are then
clarified and dewaiered.
A series of wet classification separators is used to separate the plas-
tic, lead and ebonite components. The separators are usually flotation
separators or heavy-medium countercurrent separators with a screw
auger or drag chain to remove settled solids. However, air separators
and separators using clean water washing rather than a heavy-medium
also are used. If a flotation separator is first, the plastic is removed
and the lead/ebonite stream is sent to a heavy-medium or other separa-
tor. If a heavy-medium separator is first, the lead is removed and the
plastic/ebonite stream is sent to a flotation separator. Once the ebonite
has been separated from the other battery casing components, it can
be washed again with water or surfactants1.
None of the companies contacted had successfully separated a waste
battery pile and produced an ebonite product that meets the EP Toxicity
standard for lead. Even a company that successfully processes whole
batteries or battery casings will have trouble cleaning battery wastes
from a Superfund site for the following reasons:'
• The presence of rock and slag; these materials will have to be removed
to avoid damaging the process equipment
• The presence of soil present will two problems: foaming and degra-
dation of the lead oxide product. The soil will usually remain with
the lead oxide because of similar panicle size. Foaming problems
can be solved by adding appropriate anti-foaming chemicals
• Lead oxide may be more firmly embedded in the ebonite as a result
of storage in the ground for a long time. These two materials may
thus be very difficult to separate.
The Bureau of Mines has successfully cleaned casings at the bench-
scale level. Using a I hr pre-wash with a carbonate solution, granula-
tion to less than -3/8 mesh and soaking 1 hr in a nitric acid solution
resulted in casings with a lead level of less than 100 ppm and an EP
Tox level of less than 0.2 mg/L1.
PREDESIGN STUDIES
In 1989. U.S. EPA and NL signed a consent decree which required
NL to perform a scries of predesign studies. The work is being per-
formed by Canonic Environmental and Hazen Research. Key features
of the studies include:
Site Reference Materials. Because of the high variability of lead con-
tent in casings and soils on-site. Samples of Site Reference Materials
(SRM) were prepared which are representative of the materials that
will require treatment. SRM samples include battery casings, subsur-
face soils, surface soils and matte. SRM materials were evaluated for
physical and chemical characteristics.
Recycling Pilot Studies. A series of bench-, pilot- and demonstration-
level studies is being developed. The purpose of these studies is to in-
vestigate the requirements and feasibility of separating the buried battery
casings into output streams of ebonite, plastic, lead oxide and metallic
lead. Major features of these pilot studies include:
• Optimizing throughput capacity
• Water washing, crushing and screening techniques for cleaning the
ebonite and plastic casings enough to that they can pass the EP
Toxicity test for lead
• Evaluation of fugitive emissions during the processing of the casings
and an investigation of the effectiveness of various mitigation
measures; the purpose of this task is to investigate the feasibility of
reducing airborne lead levels and suspended paniculate matter to meet
applicable standards
This work is being performed in three phases. Bench-scale and pilot-
scale studies are being performed at the Hazen Research facility in
Golden, Colorado. Demonstration-scale studies will be performed at
the Gould site.
BENCH-SCALE STUDIES
The purpose of the bench-scale studies was to develop a treatment
process which will meet the three criteria mentioned earlier. The RI
indicated that the contaminated material on site is extremely hetero-
geneous, consisting of a mixture of casings and furnace wastes which
were randomly landfilled over a period of some 30 yr. The first task
was to sample the wastes at the site. Casings at four areas of the site
are the focus of the bench-scale work. These samples include casings
from the Gould surface piles, Gould buried casings, Doane Lake casings
and Rhone- Poulenc buried casings. A summary of the waste composi-
tion is presented in Table 1.
Once the materials had been characterized, the approach for the
process development was determined. This approach, illustrated in
Figure 1, involved extracting the components of the waste which are
recyclable and treating the remaining components to produce recycla-
ble products or clean material which can be backfilled without stabili-
zation. The remaining materials could be stabilized for on-site disposal
in a monolith.
The bench-scale test program indicated approximately 20% of the
waste on the site is recyclable by separation alone. This material con-
sists of lead oxide/sulfate sludge (17%) and a lead concentrate (3%).
The remaining 80% of the waste required treatment to produce
additional products for recycling and clean products for on-site disposal.
The treatment process was developed by inspecting the materials to be
treated and selecting a variety of unit operations to accomplish the
cleaning task. These processes were tested using different combina-
tions of flow rates, liquid solids ratios, etc., until an optimal combina-
tion of parameters was reached which resulted in cleaning to meet the
required criteria of 5 mg/L lead in the EP Ibx extract.
Using this method, an additional 20% of recyclable material was
generated. The material consisted of 2% plastic and 18% clean ebonite.
302 HAZARDOUS MATERIALS TREATMENT
-------
JO
§
s
tn
5
>
r
00
H
1
H
AMERICAN STEEL MDUSTOES,
INC.
SECONDARY SOURCE AftEA
•ATTEHV CASMOS
MATT*. CASMOS AMD OTHER
ACC PU* OP •ATTBrT* CASMOS
APPROXIMATE SHREDDED AUTO SOOT FILL SIT*
APMWXWATE MAST PUNMACC MATTE PILL SIM
•uarccno •Amur CASMQ OI«M»AL MTI
> APPHOIIUATE ACmrUM MESIOUE F«J- SITE
MATTI TEST PIT LOCATION*
-dj. SATTEAV CASMQ TEST PIT
^^ LOCATIONS
LOCATIONS
A SATTEdr CASINO FRAOMENT
SAMPLE LOCATIONS
—- ESTMATEO SOUNOAMV OP WASTE OCBMIS
—— CONTHOLLEO BOOWOAHV OF WASTE DESMIS
too
M« ••r*»y ••• «>l«d May tS. (SSS.
kr O.E. U«(« » AM*eUI». at«>n«». Or*..
lor O«v«» A M*«r»
SCALE 1* 20O*
PEET
Figure 1
Location of Battery Casings & Matte
-------
Table 1
Estimated Battery Component Quantities
Rhone-foulenc &
Gould Subsurface
Ebonite
Plastic
Metallic Lead
Lead O.lde/Mud
Rock/Slag
Qther
Moisture
Total
Density
(Ibs/cu ft)
68.00
46.56
297,46
238.37
105.56
74.28
62.30
79.80
Volume
(cu. yds)
69.008
4,070
117
2.703
1.938
1.264
79.100
Height
Tons
63.349
2,558
469
8,700
2.762
1.268
-UJJ
85.218
Per Cent
(•eight)
74 3
3 0
0 6
10 2
3.2
I.S
7.2
Table 2
Physical Makeup of Battery Wtante Materials
mi«M Otrctnt)
Could CouU Uit OMM imw-PwitK
Surftn CailMi eurlea Haiti L4U milt
FUltlC
Httt«. tockl, u4 triik
U4 I4tt»rj fmtt
Htttlllc 1*44 >M Oltrli
II
U.I
S J
$.1
1.0
11. S
K 1
II.)
I.I
2S.O
3.1
0.0
n.i
1.1
Could Surface
llld Sul'ltt/Ollat illMI IM Olrt II.I
IJ.l
li.O
Ebonite
Plastic
Metallic Lead
Lead Oilde/Hud
Rock/Slag
Moisture
Total
65 81
45.06
287 88
230 69
102.16
62 30
70.07
899
595
6
52
148
1.700
799
362
24
161
204
59
1,609
50.0
22.5
1 5
10 0
12.7
37
1«bte3
Lead Distribution of Ebonite and Ptartk Wuer Wish Products
GouU
Surfict
Ft»a Ktur'ill H.t It TOJ
Gould Cut OO*M
'K Witt lilt Mitt Mitt
.I t> Toi ni.i CP TM n.t O ta
IB
CASES
] OEWATtBEO
OXCOES
Figure 2
M.A. Industries Simplified Process Flow Diagram
CONCLUSIONS
It is not certain at this time whether the cleaned ebonite can be recy-
cled. At a minimum, however, this material is sufficiently clean to allow
backfilling on-site without stabilization. The lead concentration and
EP Tox levels of the material before and after cleaning are presented
in Tables 2 and 3.
C&Oftit» tnd
"title
HtJtlC
Ebonite
046 (4 »
I 54 M4
117 IZJ.4
0.064 2 •>
0031
0 141 3.M
0 OM ) U d OH t.U
0042
The remaining material which cannot be treated consists of the mane
and waste products produced by the stabilization process. Trials with
several different processes did not successfully produce a recyclable
lead product or clean material from the matte which would pass the
EP Tox test. Significant progress was made in identifying the specific
portions of the waste stream which cause this material to rail the EP
Tox text. As indicated in Table 4. it was possible to significantly reduce
the toxicity of this material even at high lead contents.
During the treatment of the materials, the lead removed from the
casings was concentrated in a fine fraction which does not pass the
EP Tox lest for lead. Treatment of this material has not successfully
produced either a lead concentrate or a clean tailing. The bench-scale
test work indicated that 23% of the waste on-site consisted of this fine
fraction. It is anticipated that the amount of fines can be reduced sig-
nificantly by improving size reduction methods during the pilot plant
phase of the program. At the present time, approximately half of the
waste would require stabilization in order to be left on-site. This
materials includes the 14% matte and 23% treatment products.
The proposed technology as designed will enable approximately 40%
of the waste to be recycled, resulting in removal of 89% of the lead
on the site. A summary of the disposition of products is presented in
Table 5.
304 HAZARDOUS MATERIALS TREATMENT
-------
Table 4
Toxicity Reduction of Matte by Treatment Based on Lead
Material/Project
Gould matte as received
Quenched matte
Slowly cooled matte
S1lka stabilized matte
Percent pb
6.41
6.98
5.08
3.69
1,460
496
146
64
Tables
Preliminary Material Balance Based on the Results of the
Bench-Scale TES Work
Excavated Materials
Potential Recycle Products:
Lead Sllnes
Clean Plastic
Clean Ebonite
Metallic Lead
Tons
127,577
21,630
2,670
22,579
3,827
% by
Welaht
17
2
18
3
SITE MATERIALS
SEPARATE
RECYCLABLE
OR CLEAN
COMPONENTS
NON-RECYCLABLE
COMPONENTS
TREATMENT
RECYCLABLE OR
CLEAN PRODUCTS
Figure 3
Bench-Scale Testing Approach to Cleanup of the Gould Site
Potentially Stabilized Materials
Matte, Rocks, Lead, and Trash 3,601 3
Ebonite Fines 20,774 16
Note:
1. The separation between coarse and fines 1s made at 10 mesh.
2. Disposition of contaminated Is not Included Inthe materialbalance.
It is anticipated that future work through pilot and field demonstra-
tions will increase the amount of recyclable material and the amount
of lead removed from the site.
REFERENCES
1. Dames and Moore, Remedial Investigation and Feasibility for NL/Gould
Superfund Site, Final Report, February, 1988.
2. LPEI, Technical Assistance to U.S. EPA Region 10 for the Gould Superfund
Site, Portland Oregon - Survey of Commercial Battery Recyders, draft Report,
PEI Associates, Inc, Cincinnati, OH, contract It 86-03-3419, Work Assign-
ment i 19, 1988.
3. Bureau of Mines, Personal Communication between David Tetta, U.S. EPA
Region 10 and Ernie Cole, Bureau of Mines, 1989.
HAZARDOUS MATERIALS TREATMENT 305
-------
Start-up of an Innovative UV/Peroxidation
Groundwater Treatment System in the
Era of Superfund and RCRA
Corrective Action Programs
Nancy W. Gossett, RE.
James Bausano
CH2M Hill
Bellevue, Washington
John Oldham
Reichhold Chemicals
Tacoma, Washington
ABSTRACT
Site remediation under RCRA presents new challenges for design
professionals. In some cases of remediation, such as at Reichhold,
Tacoma. well established technologies are not always available to meet
the performance objectives that have been established.
The challenge at the RCI Tacoma site has been the treatment and
destruction of pentachlorophenol contaminated ground water. Due to
the land ban restrictions, all technologies which resulted in significant
volumes of solid generation were avoided due to the inability to dispose
of this material at this time. This criteria ruled out a significant number
of well established technologies.
Chemical oxidation, although not a commonly used technology, was
selected as an applicable technology at this facility. This involves des-
truction of organics through chemical oxidation using hydrogen peroxide
and ultraviolet light. The initial start-up phase of this unit has been
completed and many performance and operational questions still remain
unanswered. A rigorous performance testing program is currently in
progress to evaluate the entire treatment system. The results of this test
program will provide a better understanding of the parameters which
affect the performance of this technology.
INTRODUCTION
Reichhold Chemicals, Inc. (RCI). owns and operates a manufacturing
facility on about 52 ac. in the Tacoma, Washington Commencement
Bay industrial area. Since the facility began operations in 1956, a variety
of chemical products have been manufactured at the facility.
Pentachlorophenol (PCP) was a major product of the Tacoma facility
over the years. PCP is a chlorinated phenolic compound, used exten-
sively in the treatment of wood and lumber products. RCI discontinued
PCP production and dismantled the production area in 1985.
In January of 1988, Reichhold Tacoma applied fur a U.S. EPA RCRA
Part B permit to address past practice issues. This permit was granted
in November of 1988 under RCRA. The Pan B permit contains many
conditions which apply to the operating plant under RCRA. Most
important to this discussion is the Interim Corrective Action Plan (ICAP)
incorporated into the Permit which establishes the framework under
which the current interim control measures and site remedial activities
are being conducted.
The interim corrective actions underway at the Reichhold facility are
based on the conditions of the Permit and on site assessment work con-
ducted under a previous consent order. The current interim corrective
actions underway involve control and isolation of surface water and
groundwatcr. Additionally, final corrective action measures implemented
to date include soils treatability technology screening and scheduled
pilot demonstrations of several remedial technologies for on-site treat-
ment of soils.
The constituents present on-sile in the soils and groundwater include
pentachlorophenol and other phenolic compounds. Regulatory restric-
tions on wastes containing pentachloropnenol from manufacturing (F021)
under 40 CFR Part 261 do not permit off-site incineration, disposal
or treatment. Therefore, the interim corrective measures were proposed
for isolation and containment of soils and groundwater until the
appropriate technologies for permanent destruction or detoxification
of the hazardous constituents in the site soils are selected and im-
plemented.
INTERIM CORRECTIVE ACTION PLAN
As stated previously, the purpose of the interim corrective actions
currently being implemented at the site are to protect human health
and the environment until implementation of the on-site soils cleanup
and final closure. The objectives of the ongoing interim corrective
actions are to divert precipitation, surface water and groundwater away
from contaminated soils to prevent contact of clean water with hazardous
constituents found in the site soils and groundwater; and to prevent off-
site migration of contaminated groundwater. The specific actions which
have been implemented or are currently under construction at the facility
include:
• grading and placement of a site cover (concrete, asphalt and gravel
sections) over the contaminated soils to divert precipitation and surface
runoff away from these areas.
• installation of a french drain through the shallow aquifer at the facil-
ity perimeter. The drain intercepts contaminated water before it moves
off-site. Recovered water from the shallow aquifer is pumped to the
on-site water treatment system.
• installation of intermediate aquifer extraction wells in areas where
contaminants above action levels have been detected in the ground-
water. Eight extraction wells have been installed to date.
• installation of an on-site water treatment system to process recovered
groundwater prior to discharge to the local public sewer system.
WATER TREATMENT SYSTEM DESIGN AND INSTALLATION
Design Basis
Prior to selection of the most appropriate treatment technology, it
was necessary to establish the objectives of the treatment system, and
assess the site specific regulatory and physical restrictions under which
the treatment plant would be required to function.
The Water Treatment System (WTS) is regulated under Section 307(b)
of The Clean Water Act. and is not specifically regulated under RCRA
since it is considered a wastewater pretreatment system for discharge
to the sanitary sewer. However, the provisions of the RCRA permit re-
quire the groundwater in the system to be managed as a hazardous waste.
306 HAZARDOUS MATERIALS TREATMENT
-------
This affects the health and safety practices of plant operations. It also
requires that all solids removed from the process or any material which
comes in contact with the process feed water (recovered groundwater)
be managed as a listed hazardous waste. In this case, the land ban re-
quires that all waste or solids be stored indefinitely or treated on site.
Given the thirty year planned operating life of the system, water treat-
ment processes generating large amounts of solids were not considered
during the technology screening analysis.
Pending final negotiations with the local sewer district, discharge
criteria to the public sewer system were initially assumed to be at or
near OCPS pretreatment criteria. For POP, a discharge limit of 20 to
50 parts per billion (ppb) was used for the initial technology screening.
An ultimate plant influent flow rate of between 100 and 200 gallons
per minute, and 5000 parts per billion influent PCP concentration were
used as initial design assumptions. Due to the complexity of the site
geology and to meet the tight schedules for implementation of the in-
terim corrective actions, these initial assumptions were used for the
treatment system technology screening evaluations. At the same time,
parallel hydrogeological assessments were conducted concurrently to
refine the estimates of groundwater recovery rates and influent water
quality. These design parameters have been reevaluated later as more
data becomes available.
Initially, the scope of work and schedule were established with the
following tasks identified to implement the water treatment portion of
the groundwater recovery and treatment system:
• Technology assessment and screening
• Bench-scale treatability testing
• Technology/vendor selection
• Revise design basis
• Field pilot demonstration
• Evaluate performance and revise design basis
• Engineering/procurement/construction of full-scale system
• Start-up and handover
This initial work plan was established at the onset of the project with
the expectation that further revision to this plan may be necessary. The
complexity of the site geology, the uncertainty associated with using
an innovative treatment technology, and the regulatory aspects involved
with implementation of the ICAP dictated that the scope of work remain
flexible to accommodate new developments in the overall implementa-
tion of the ICAP.
Technology Screening and Assessment
The initial screening studies identified several technologies that were
technically and economically feasible. Biological degradation (activated
sludge), carbon adsorption, and various methods of chemical oxida-
tion were selected as potential treatment technologies. Solids disposal
issues quickly eliminated both biological treatment and activated carbon
adsorption as impractical choices due to the regulatory restrictions
placed on solids handling. Chemical oxidation, although not a commonly
used technology, was selected as a potentially applicable technology
for use at the Reichhold facility. This technology involves the use of
one or several oxidants to destroy organic constituents in a water stream
and therefore does not generate a waste stream or solids.
Treatability Testing
Three vendors of chemical oxidation water treatment systems were
selected to perform bench-scale treatability testing on groundwater sam-
ples collected from the site. Small quantities of groundwater were trans-
ported to vendors' treatability labs. All tests were witnessed by CH2M
Hill technical staff, and parallel chemical analyses of the test waters
were conducted at the CH2M Hill CLP (Contract Lab Program) ana-
lytical laboratories under rigid quality assurance protocol for confir-
mation of the vendors' results. Field visits to operating facilities were
conducted using the enhanced oxidation units to evaluate full-scale sys-
tems in operation and discuss system performance with operating per-
sonnel. Upon completion of the witnessed bench-scale treatability testing
and the site visits, each vendor submitted proposals for both
demonstration-phase and full-scale treatment systems. A comparative
technical and economic evaluation was performed. Peroxidation Sys-
tems, Inc. was chosen to provide a leased demonstration unit for the
demonstration-phase installation.
Revise Design Basis
Initially, the field pilot demonstration system was scoped as a
10 gal/min unit. However, it was determined that a larger system would
be necessary to treat substantial amounts of water generated during con-
struction activities, and testing of the well and sump systems. The final
design throughput for the demonstration-phase treatment system was
revised to 70 gpm with an influent PCP concentration of 5,000 ppb.
A plan for a phased installation of the treatment system was then
adopted. It was compatible with other ongoing field investigations, con-
struction, and the start-up of the many components of the interim cor-
rective action implementation. Per the revised plan, the
demonstration-phase system would be designed to treat groundwater
in batches, and, if the system met performance objectives, would later
be expanded and modified to operate continuously after start-up of the
entire groundwater recovery system. The objectives established for the
demonstration-phase treatment system were established as follows:
Perform a field demonstration of the selected equipment under actual
site conditions prior to final commitment to a full-scale system.
• Collect operating data to be incorporated into the design basis for
the full-scale system expansion.
• Provide a water treatment system to treat and discharge wastewater
generated during ongoing hydrogeological assessments and construc-
tion activities.
Demonstration-phase Installation
Design and installation of the demonstration-phase water treatment
system proceeded after selection of the Peroxidation Systems, Inc.
chemical oxidation system. In March of 1989 the initial startup of the
demonstration-phase UV/hydrogen peroxidation system began. The start-
up of the system was to occur in three steps:
• Mechanical shakedown of the completed system with potable water
to insure that the system was mechanically complete and functional
before introducing contaminated process streams into the system.
• Batch processing of water which was collected during on-site con-
struction activities.
• Batch processing of groundwater received from the intermediate
aquifer extraction wells. Two wells located in one of the areas with
higher detected contaminated levels were completed in May. Installa-
tion of the treatment system would allow continuous pumping of these
wells to obtain additional hydrogeological and chemical data on the
groundwater.
A start-up and testing plan was, prepared prior to completion of the
installation. This plan was established as a guideline. Flexibility was
written into the plan to accommodate changes which might be neces-
sary due to water quality conditions, schedule changes, and other con-
straints imposed by other ongoing site activities. Objectives were
established as a basis for the test program:
• Equipment performance guarantee.
Demonstrate that the vendor-supplied UV/hydrogen peroxide treat-
ment system can perform in accordance with the performance
guarantee.
• System Optimization and Operation.
Demonstrate that the entire water treatment system and each indi-
vidual component function according to the performance criteria
established.
• Baseline Data Collection.
Collect initial chemical and physical data on the recovered ground-
water from intermediate aquifer extraction wells.
• Effluent Discharge Compliance.
Establish a discharge monitoring and performance history for evalua-
tion by the local sanitary district. As negotiated with the City, dis-
charges would be on a batch by batch basis during the demonstration
HAZARDOUS MATERIALS TREATMENT 307
-------
phase program. Each batch would be approved for discharge by the
City during this demonstration program.
Final issue of the long-term discharge permit would follow a suc-
cessful record of discharges during the demonstration-phase program.
The initial start-up proceeded cautiously with a mechanical shakedown
of the various process components. Potable water was used in the sys-
tem to eliminate the possibility of any accidental discharges of ground-
water or contaminated construction water during this period.
The First available source of water for testing the system was con-
struction water. Well water was then processed after successful treat-
ment and discharge of the construction water. Routine operations
proceeded as batches of construction and well water generated from
other site activities were treated and discharged. It was soon noted that
the water coming to the treatment system from various sources exhibited
a large variability in water and chemistry. It was not well understood
at the time exactly which chemical parameters had the largest impact
on system performance, but large variations in system performance were
observed as the influent water characteristics varied. Given the varia-
bility of the feed water, it was only possible to establish generalizations
about system performance during this operating period. Although the
system was adequately providing treatment of construction water to allow
other site activities to progress, the changing contaminant levels and
chemistry of the groundwater resulted in ambiguous pcntachlorophenol
destruction rate data. It was not possible to conduct an adequate
assessment of system performance under the variable operating
conditions.
It was observed during this period that several variables in the feed
water seriously affect system performance. Although monitoring well
data had indicated the presence of iron in the intermediate aquifer
groundwater, iron concentrations in water pumped from the wells have
been up to four times higher than the initial estimates. Since performance
of a UV/hydrogen peroxide system is a function of ultraviolet light trans-
mission, treatment of turbid water with suspended iron paniculate has
proven especially difficult. Other variables such as the presence of other
organic compounds and the presence of suspended silt have also affected
the rate of peniachlorophcnol destruction. The effect of other water
quality variables such as alkalinity and pH are unknown at this time.
Prior to initiating the system expansion, a rigorous performance lest
program is being performed to evaluate the effect of various operating
parameters. The test plan is based on testing three discrete and
homogeneous water sources under a variety of operating conditions.
The results of the test will be used to establish a predictive operating
model to facilitate routine system operations. This level of understanding
is essential prior to proceeding into design of the expanded system.
CONCLUSIONS
As with any plant design and construction project, careful attention
should be given to establishing a work plan for a remedial corrective
action. However, uncertainty of scope, a shirting regulatory environ-
ment, uncertain and complex site conditions, and the use of new, untested
technologies can require substantial alteration to that plan at any stage
of the project. It i.s essential to acknowledge this and build contingen-
cies into work plans and schedules. Due to schedule constraints, it may
be impossible to resolve issues related to establishing a design basis,
and ihus may require that the engineer design considerable flexibility
into a system.
308 HAZARDOUS MATERIALS TREATMENT
-------
Pilot Scale In Situ Vitrification at
Arnold Engineering Development Center
Arnold AFB, Tennessee
J. Kent Lominac
Robert C. Edwards
AEDC/DEEA
Arnold AFB, Tennessee
Craig L. Timmerman
Pacific Northwest Laboratory
Richland, Washington
ABSTRACT
The Department of Defense has the Installation Restoration Program
(IRP) to identify and permanently remediate hazardous material dis-
posal sites at its military bases across the United States. Pursuant to
this guidance, Arnold Engineering Development Center (AEDC)
selected In Situ Vitrification (ISV) to remediate an old fire training area,
Fire Protection Training Area (FPTA) No. 2 at Arnold AFB in
Tennessee.
The ISV technology was developed by Pacific Northwest Laboratory
(PNL), Richland, WA for the U. S. Department of Energy (DOE) and
will result in for the destruction and encapsulation of the petroleum-
oil-lubricants (POL) and heavy metal-constituents found at the FPTA
and adjacent over-flow pond. ISV operates by passing a measured current
of electricity into the ground through a set of electrodes. The resulting
heat causes the soil to melt and form a solid vitreous (glass) mass similar
to naturally occurring obsidian or basalt. In the process, organic con-
stituents will be pyrolyzed (changed by heat) by the ensuing heat whereas
the non-organic material will be incorporated into the glass matrix.
Successful bench-scale tests were accomplished during the summer of
1988, and a successful pilot-scale test was accomplished in February
of 1989.
• SOUTH
•CAROLINA
SCALE
Figure 1
Arnold Engineering Development Center Location
INTRODUCTION
Situated in the rolling countryside of middle Tennessee lies the Air
Forces' best kept secret—Arnold Engineering Development Center
(AEDC) at Arnold AFB, Tennessee. AEDC is located in Coffee and
Franklin Counties, Tennessee, midway between Chattanooga and Nash-
ville (Fig. 1). The entire AEDC reservation encompasses 39,081 ac
devoted to testing, research and development facilities. AEDC was con-
structed in the early 1950s with initial testing starting in 1953. The Center
has, since its beginning, conducted a wide range of tests and simula-
tions in aerodynamics, propulsion and aerospace systems.
The U. S. Department of Defense (DOD) has developed a program
to identify and evaluate past hazardous material disposal sites on DOD
property, to control the migration of hazardous contaminates and to
control hazards to health or welfare that may result from these past
disposal operations. This program is known as the Installation Res-
toration Program (IRP). The IRP initially had four phases consisting
of: Phase I, Installation Assessment and Records Search; Phase n, Con-
firmation and Quantification; Phase m, Technology Base Development;
and Phase IV, Operations and Remedial Actions. The U.S. DOD now
follows the terminology of the U. S. EPA: PA/SI, Preliminary Assess-
ment and Site Inspection; RI/FS, Remedial Investigation and Feasibility
Study; RD/RA, Remedial Design and Remedial Action; and LTM,
Long-Term Monitoring.
During the investigation of Arnold AFB, 17 sites were initially were
identified as being potentially hazardous. Subsequently, two sites were
added as a result of discoveries made by base and/or contractor per-
sonnel. Five of the original sites were subsequently dropped from any
further investigative work after the initial investigation. With the
exception of the two newest sites, all others are well into the RI phase
of investigation. One site is in the RD phase and one site is getting
ready to begin RA. This paper concerns the activities at the site in the
RA-Site 10, composed of three identifiable activity areas: Fire Protec-
tion Training Area (FPTA) No. 2, Burn Area No. 1 and Landfill No. 1.
Site 10, which comprises approximately 14.5 ac is located northwest
of the Model Shop (Bldg 451) and northeast of Gate 5 (Fig. 2). The
FPTA was constructed in 1973 and was closed in April, 1988. The
training area consisted of an unlined gravel burning area connected by
drains to a small overflow pond.
During a typical fire training exercise, water was first applied to the
burn area surface. Combustible material, typically lighter than water,
was added then ignited. The ignited area was then used as a training
exercise for AEDC fire protection personnel. Contaminated petroleum
fuels, fuel filters, waste oils, thinners, solvents, and some propellants
were burned up to the late 1970's. Since then, the materials burned have
consisted primarily of JPD4 fuel and some sodium-potassium alloys.
A typical burn consumed 500 to 600 gal of fuel and occurred up
HAZARDOUS MATERIALS TREATMENT 309
-------
OrWRSiQN
smucuOK
OECOMUB
ISfRVOIR -I
v-
Figure 2
Site 10 Location
to 21 times per year. Landfill No. 1 served as a disposal site for many
installation wastes, including refuse, garbage, construction debris and
some shop wastes. The shop wastes generally were placed in a trench
located at or near Landfill No. 1 and burned (this area is know as Burn
Area No. 1). Some of the shop wastes included waste oils, contaminat-
ed fuels, solvents thinners and other combustible wastes.
IN SITU VITRIFICATION (ISV)-HOW IT WORKS
ISV is a thermal treatment process that coverts contaminated soil into
a chemically inert, stable glass and crystalline solid. Four electrodes
are inserted into the ground in the form of a square array to a predeter-
mined treatment depth (Fig. 3). Since soil is not electrically conduc-
Figure 3
In Situ Vitrification Process
live once the moisture has been driven off, a conductive mixture of
flaked graphite and glass fit is placed between the electrodes to act as
a starter path. A high silica content, fiber material acting as an insulating
thermal blanket is placed on top of the starter material and soil to aid
in initial heat retention (the blanket is consumed in (he melting process).
An electrical current is applied to the path. The resulting power heats
the starter path and surrounding soil up to 3600°F, well above (he soil's
initial melting temperature. The starter eventually will be consumed
by oxidation with the current being transferred to (he miw electrically
conductive molten soil.
During the growth of the vitreous zone, non-volatile elements (com-
pounds) are incorporated into the melt while organic components are
destroyed by pyrolysis. The pyrolyzed by-products migrate to the surface
of (he vitrified zone, where they oxidize in the presence of oxygen.
A hood placed over the processing area provides confinement for the
combustion gases which are drawn into the off-gas treatment system.
The hood area is larger than the area to be vitrified to assure that the
gases driven off during the process are captured. The hood has a skin
along the bottom edge to contain the gases and to aid in developing
a partial vacuum, both important criteria in the treatment system for
off-gases.
As the melt grows downward and outward, power is maintained at
sufficient levels to overcome the heat losses from the surface and to
the surrounding soil. In general, the melt grows outward to a total width
of approximately 50% of the spacing of the electrodes. Therefore, if
the electrode spacing is 18 ft center to center, a melt width of approxi-
mately 27 ft would be observed under normal conditions. The molten
zone is a roughly square with slightly rounded corners, reflecting higher
power densities around the electrodes. As the resistance decreases during
the melting process, the voltage is constantly monitored and adjusted
via electrical transformer voltage taps to maintain a constant operating
power.
The equipment necessary to produce a vitrified mass can be divided
into five major groups: (I) electrical power supply, (2) off-gas hood,
(3) off-gas treatment, (4) off-gas support and (5) process control. All
of these components, except the off-gas hood, are contained in three
standard size trailers (Fig. 4).
The off-gas hood and associated piping are dismantled and transported
on a flat-bed trailer between sites. The off-gas trailer is the most
expensive and complex of the three. This system cools, scrubs and fillers
the gaseous effluents exhausted from the hood. A glycol cooling unit
cools the scrub solution to extract built-up thermal energy. This cooling
process allows the scrub solution to be recycled back through the sys-
tem. Equipment necessary for the off-gas hood includes a small crane
to lift and position the hood for each melt and a small bulldozer (if
necessary) to grade and level the area before the hood is set in place.
Craft support, such as electricians, pipefitters, riggers, operators, etc.,
is required to set up the ISV equipment as well as move the equipment
from setting to setting.
sufvom IRAH.IR
EUCIRICM STSTtu;
WfP» W.1W
HOUSK
(UCIROM
OH-GAS HOOD
COVtR
Figure 4
Large-Scale Process Equipment for In Situ Vitrification
The long-term stability of the glass is, of course, a significant factor
in (his process. The glass has been subjected to a variety of leach tests,
including the U.S. EPA Extraction Procedure Tbxicity Test (EP Tbx)
and Toxic Characteristic Leach Procedure (TCLP). These tests show
a uniformly low leach rate for heavy metals of approximately 1 x 10-5
Ib/ff'/day or lower. Additional testing along with comparisons to
naturally occurring obsidians indicate that the mean life of the vitri-
fied material would be on the order of 1.000,000 yr',
BENCH-SCALE TESTING
Bench-scale (esting was performed on AEDC soils. Initial bench-
scale testing was conducted at AEDC during May. 1988. A follow-up
bench-scale lest was performed at PNL in July, 1988. Soils were tested
from both IRP Site 10 and IRP Site I. IRP Site 1 is a landfill/leach
pit area. The initial bench-scale test at AEDC was unsuccessful. The
310 HAZARDOUS MATERIALS TREATMENT
-------
problem was identified as the lack of electrically conductive elements
or fluxing agents (e.g., sodium carbonate) in the AEDC soils. Addi-
tional bench-scale testing using fluxing additives was conducted at PNL
using a modified single-phase electrical connection of the engineering-
scale power supply. The engineering-scale system was used instead of
the bench-scale system because the power supply better simulates the
power density, melt rate and control of the larger scale of operational
ISV systems.
The assessments of Site 1 and Site 10 soil compositions showed the
AEDC soil to have a high alumina/silica content. This material is suitable
acceptable for making a good glass product, but ISV also requires a
sufficient quantity of alkali elements (Li, Na and K) to lower the melt
temperature and provide electrical conducting. Typically, 5 % of the
alkali material is required for ISV to perform effectively. AEDC soils
assessed for bench-scale testing contained 11% of these materials; there-
fore, soil fluxing additives were required. Ten percent (by weight)
sodium carbonate was added to ensure successful ISV processing.
As a result of the bench-scale testing, the following was concluded:
(1) with the addition of a fluxing additive such as sodium carbonate,
ISV could process the contaminated soils from Sites 1 and 10 into a
more compact and environmentally stable (immobilized) form, (2)
organic contaminants were effectively destroyed, (3) leach testing results
from the Extraction Procedure (EP) Toxicity and Toxic Characteristic
Leach Procedure (TCLP) showed that metals of concern were below
the maximum permissible limit which indicates that inorganic con-
taminants are immobilized and (4) 5% to 10% sodium carbonate addi-
tions are necessary to process AEDC Sites 1 and 10 soils. Based on
the results of these tests, it was recommended to perform a pilot-scale
test at the AEDC Site 10 fire training pit to verify the efforts of the
bench-scale tests prior to actual remediation of the site with ISV2.
PILOT-SCALE ISV TEST
The pilot-scale test was conducted at AEDC during February 1989.
The pilot-scale test system used at AEDC utilized four electrodes with
3 ft, separation and consisted of a power control unit, off-gas contain-
ment hood over the test site and an off-gas treatment system housed
in a portable semi-trailer. The actual pilot-scale setup at AEDC IRP
Site 10 in shown in Figure 5.
Figure 5
Pilot-Scale Setup at AEDC IRP Site 10
The pilot-scale power system utilizes a Scott-Tee connection to trans-
form a three-phase input to a two-phase secondary load on diagonally
opposed electrodes in a square pattern. The 500-kW power supply may
be either voltage or current regulated. The alternating current primary
is rated at 480 V, 600 A, three-phase, and 60 Hz. This three-phase
input feeds the Scott-Tee connected transformer providing a 2 phase
secondary. The transformer has four separate voltage tap settings—1000
V, 650 V, 430 V and 250 V. Each voltage tap has a corresponding
amperage rating of 250 A, 385 A, 580 A and 1000 A per phase, respec-
tively. The amount of three-phase input delivered from the transformer
is controlled by silicon controlled rectifiers (SCRs). During the pilot-
scale test, this power system very effectively maintained a balanced load
to the electrodes.
The upper section of the off-gas containment and electrode support
hood (Fig 6) is 10 ft by 18 ft long, and is constructed from seven panels
of 20 gauge stainless steel bolted together. The lower structure is a sup-
port structure covered and sealed with a high temperature fiberglass-
based, silicon coated fabric. The fabric was bolted to the upper struc-
ture and covered with soil at the base to form a seal with the ground.
The overall hood height was 6 ft.
Figure 6
Off-Gas Containment And Electrode Support Hood
The off-gas system is shown schematically in Figure 7. The off-gas
passes through a venturi-ejector scrubber and separator, Hydro-Sonic
scrubber, separator, condenser, another separator, heater, one stage of
HEPA filtration, one stage of activated carbon filtration and a blower.
Liquid to the two wet scrubbers is supplied by two independent recir-
culation tanks, each equipped with a pump and heat exchanger. The
entire off-gas system has been installed in a 45 ft long semi-trailer which
makes the system portable. Equipment layout within the trailer is illus-
trated in Figure 8.
Figure 7
Off-Gas System Schematic for the Pilot
Figure 8
Pilot-Scale Process Equipment
HAZARDOUS MATERIALS TREATMENT 311
-------
The primary objective of the pilot-scale lest was to confirm the bench-
scale testing results on the actual Site 10 Fire Training Area contami-
nated soils. The testing evaluates process operations effectiveness and
off-gas behavior of volatile or entrained materials. The earlier tests
showed the need for additional fluxing additives to increase the elec-
trical conductivity to allow ISV processing.
The additives were placed differently in the pilot-scale tcsi than they
were in the initial tests. The initial tests were of a smaller scale and
allowed total blending of the fluxant (Na,CO,) with the entire soil
volume. AEDC did not want to disturb (he contaminated soil at Site
10. therefore a "cover layer" method was used to blend the fluxant.
For the pilot-scale test, the fluxant was mixed with clean soil and placed
over the contaminated soil. The depth of the "cover layer" was 3 ft
and had a fluxant concentration of 27%. The fluxing additives have
been shown to mix with the contaminated soil as the melt progresses.
The cover soil also serves to enhance the process destruction efficiency
of the organic contaminants and allows ISV to establish a molten /one
of soil prior to contacting the organic. This thermal inertia mass of
molten soil achieves pyrolysis of the organics as opposed to volatilizing
the organics if ISV were started at the contaminated soil surface.
Four 2-in diameter molybdenum cores with 6-in diameter graphite
collar electrodes were placed to the 10-ft depth from the cover soil sur-
face. The electrodes were positioned on a 3-ft square separation. Two
of the electrodes had fiber optic, depth monitoring transmitters attached
to them which enables the depth progress of t)>e ISV melt to be tracked.
The test was performed on the southern edge of the Fire Training
Area. This location was chosen to allow application of ISV to an actual
portion of the contaminated site and to allow thermal transport
monitoring to the clean surrounding soil. Pretest and post-test soil core
sampling was performed to obtain before and after soil profiles. A
surface view of the ISV block after processing is shown in Figure 9.
Figure 9
Surface View of ISV Block After Protevking
PILOT-SCALE TESTING
Analyses of the data from the pilot-scale tesl regarding the perfor-
mance of the ISV process to AEDC Site 10 soils provide the following
conclusions:
I. Fluxing additions are needed to process AEDC soils. Addition of
fluxing additives to the 3 ft of cover soil allowed ISV to treat a por-
tion of the Site 10 soil, but did not result in efficient ISV process
operations or effective fluxant mixing to the desired depth with AEDC
soils. Therefore, to ensure achieving the desired depth, these fluxing
additives should be added to the entire vitrification volume by soil
mixing or injection techniques instead of concentrating them in the
cover soil layer.
2. The pilot-scale ISV electrical and off-gas treatment system operated
effectively within design constraints throughout the 168-hr operating
period.
3 The operation produced a 15-ton vitrified block measuring 5 ft deep
and 8 ft wide on each side. A greater melt depth was desired, but
soil composition variations, the amount and type of fluxants added
and the method of adding the fluxants affected the depth achieved.
4 Inorganic paniculate releases from the melt to the off-gas system
were minor No detectable paniculate releases were measured out
the stack after off-gas treatment. The ISV process effectively retains
inorganic materials within the melt. Of the small quantities released
(0.25 Ib). the off-gas treatment system performs a very efficient scrub-
bing and filtering of the paniculate-.
5. Organic contaminants were effectively destroyed to the 89% level
for the fuel oil-contaminated Site 10 soil solely by the ISV melt ex-
clusive of any off-gas treatment. The overall ISV system destruction
and removal efficiency (DRE) was 99.85% which included the off-
gas treatment system.
6. Leach testing results passed both the Extraction Procedure (EP)
Toxicity and Toxic Characteristics Leach Procedure (TCLP) leach
tests and showed that all metals of concern are below leach release
limits. This result indicates thai inorganic contaminants are immobi-
lized to a level that should allow the site to be listed as non-hazardous
material according to regulatory criteria.
7. Pretest soil samples showed the highest organic concentration in the
surface soil samples at the original surface grade position. Post-test
analyses showed that the samples in the close proximity region to
the vitrified block (< 1 ft away) displayed a noticeable decrease in
the organic concentration between the pre and post-test samples from
the same relative positions. The available data indicate that ISV
processing will deplete a zone near the block of organic material
but does not thermally transport the organic species away from the
vitrification zone.
In summary, pilot-scale testing confirms the potential for ISV treat-
ment of organic contaminates soils from the Fire Training Area at IRP
Site 10. Based on the results of the bench-scale and pilot-scale tests,
ISV is a potential solution that could be used to remediate the soils
at the AEDC Site 10 Fire Training Area.
REFERENCES
I Lomiruc. J. K. and Julius. ). F. K . "In Situ Vitrification at the Arnold En-
gineering Development Center. Arnold AFR Tennessee." Pnx. 6lh National
Ctmfrrrncr on Hayirdous Htuies and Hazardous Materials, New Orleans.
LA. pp 377-379. HMCRI. Silver Spring. MD, 1989.
2 Timmcrman. C L.. "Feasibility Testing of In Situ Vitrification of Arnold
Engineering Development Center Contaminated Soils." Picific Northwest
Laboratory Project No 14384 for Oik Ridge National Laboratory. Oak Ridge
Tennessee, March 1989
312 HAZARDOUS MATERIALS TREATMENT
-------
Combining Innovative and Traditional Technologies for Effective
Remediation of PCBs and VOCs Contamination
Marc J. Dent
David S. Towers
David G. VanArnam, RE.
O'Brien & Gere Engineers, Inc.
Syracuse, New \brk
INTRODUCTION
The presence of PCBs and VOCs in the soil and groundwater present
a challenge in designing a complete and effective remediation program
due to the different chemical nature and mobility of these compounds.
This paper discusses the problem in a case study of a site in Pennsyl-
vania that has PCBs and VOCs contamination in a complex hydro-
geologic setting. Innovative and traditional technologies were evaluated
and applied to effectively remediate the site. All investigations and
remedial plans were developed with the concurrence of the State Regula-
tory Agency in this voluntary site cleanup.
Through the implementation of a phased hydrogeologic investigation,
the nature of the subsurface aquifer system and the extent of the con-
tamination were defined and characterized. It was determined that PCBs
and VOCs were present in the unsaturated zone soils, a perched water
PERCHED WATER
ZONE
PROPERTY LINE CTYP!)
FENCE CTYP)
FORCE
CTYI>:
AIR SIRWPER
COLUMN
STORM SEWER
Figure 1
Manufacturing Facility Site Plan
HAZARDOUS MATERIALS TREATMENT 313
-------
table, the regional sand and gravel aquifer and a lower bedrock ground-
water zone. Also, two separate free'phase petroleum product pools were
detected on the groundwater.
Remedial program challenges required simultaneous treatment of soils
in place, recovery and treatment of groundwater, recovery of the free
phase product and immobilization of the PCBs. These goals were suc-
cessfully accomplished by employing a creative combination of in situ
air stripping, soil bentonite slurry cutoff walls, granular activated carbon
treatment and counter-current packed column air stripping technologies.
BACKGROUND
This machining facility had, in the past, four types of disposal and
chemical usage areas, including the following:
• Underground storage tanks and pits
• Electrical transformers
• Drum storage area
• Miscellaneous construction debris and rubble disposal sites
The analysis of the soils beneath the removed tanks and pits prompted
a hydrogeologic investigation. Electrical transformers and waste oil tanks
had existed on-site and may have contributed to the PCBs detected.
Overflow of drums had stained surface soils which were later removed
in 1975 and 1980. Miscellaneous disposal areas also had existed and
consisted of a variety of materials including construction debris, and
rubbish.
A hydrogeologic investigation conducted at the site during 1987 iden-
tified a contaminated soil volume of approximately 18,000 yd'
containing up to 50 ppm VOCs (Fig. 1). Approximately 2.000 Ib of
VOCs are projected to be contained in the soils located within a half
paved-half grassed 125,000-ftJ area.
The hydrogeologic investigation also identified three contaminated
groundwater zones:
• Perched water table
• Overburden Aquifer
• Bedrock Aquifer
Because silt and clay are the upper natural soil unit at the site, rain-
water that percolates through the overlying Fill material is trapped in
the perched zone above this clayey zone. The perched zone is approxi-
mately 12 to 17 ft below ground and is configured approximately a half
circle with a radius of about 400 ft. A PCB concentration up 18 mg/L
and VOCs up to a concentration of 113 mg/L were detected in the perched
zone soils.
Underlying the perched zone is a regional groundwater table occurring
at depths of 16 to 32 ft. The water table in the overburden was deter-
mined to be unconfmed and exhibited a saturated thickness of 25 to
30 ft. The concentration of VOCs and PCBs from this aquifer ranged
from 21 mg/L VOCs and 0.083 mg/L PCBs. Flow in this aquifer flows
to the southwest and discharges to the river opposite the perched zone.
Two product pools were found on this water table in two locations.
Approximately ISjOOO to 25,000 gal of product contain from 5 to 7 mg/L
PCBs.
Pump test analysis conducted within (he overburden aquifer showed
a range of transmissivity values of 12,000 to 21,500 gal/day/ft for the
coarser material while values for (he finer grained soils ranged from
2,400 to 12,000 gal/day/ft. Average hydraulic conductivities ranged from
416 to 125 gal/day/ft2
Beneath the overburden aquifer, groundwater is in the fractures of
the underlying limestone and shale bedrock. Groundwater within the
bedrock flows toward the river and produces yields of 75 gal/min in
shallow levels to 300 gal/min in deeper levels. The groundwater in the
bedrock contains VOCs and no detectable concentrations of PCBs.
SOIL AND GROUNDWATER CLEANUP OBJECTIVES
The objectives of the soil remediation portion of the project were
to remove the source of VOCs in the soil to prevent continued addi-
tional contamination of the groundwater and to permanently immobi-
lize in place the PCBs in the soils to prevent them from further migrating
into the groundwater.
The objective of the groundwater portion of the project were to remove
and treat the contaminated groundwater and prevent the contamination
plume from migrating to off-site receptors.
SCREENING OF REMEDIAL ALTERNATIVES
In developing a remedial plan for soils and groundwater, a screening
of remediation techniques was conducted to determine which techniques
are applicable, which techniques require pilot or bench testing to
determine their applicability, and which techniques can be eliminated
as potential treatment components of the remedial plan.
There is a limited number of remedial alternatives for the treatment
of soils containing a mixed matrix of VOCs and PCBs. The following
alternatives were evaluated for consideration:
No action
Construction of physical barriers
In situ treatment (biological treatment)
In situ air stripping
Excavation and disposal off-site
Excavation and incineration
The No Action alternative was eliminated as a remedial alternative
since it would allow the soil contaminants to migrate into the ground-
water which was not considered to be acceptable.
The construction of physical barriers consisting of an impermeable
cap and slurry cut-off wall would be appropriate for preventing recharge
of the perched zone from either rainfall or seasonal overflows from
a nearby stream. The presence of a natural silty clay layer below the
perched zone would allow the slurry cut-off wall to be "keyed" in pro-
viding complete encapsulation. Since PCBs are only slightly soluble
in water and would migrate primarily in paniculate form, this tech-
nology would be effective. However, this technology might not be
effective for the VOCs since the transport of the VOCs through the soil
is not entirely governed by the moisture in the soil.
While in situ biological treatment was eliminated based on its failure
to degrade PCBs and chlorinated VOCs. in situ treatment employing
air stripping has been extensively proven effective in the field for VOCs
but not effective for PCBs.
For small quantities of VOC- and PCB-contaminated soil, disposal
off-site or incineration is a viable alternative. However, since the volume
of material is significant at this site to incinerate or dispose of the soil
off-site would result in an enormous cost.
In addition to the cost of incineration, it frequently is difficult to ob-
tain sufficient landfill or incinerator capacity at the tune of disposal
and there is additional liability associated with transporting the material
to an acceptable facility. For these reasons, incineration and disposal
off-site were not considered except for small quantities of material.
There are a number of alternatives for the treatment of groundwater
containing elevated levels of VOCs and PCBs. Considering site-specific
conditions and contaminants of concern, the following alternatives were
evaluated for consideration:
• No action
• Construction of physical barriers
• In situ treatment (biological treatment)
• Pump-and-treat with ozone
• Pump-and-treat with carbon adsorption
• Pump-and-treat with air stripping
The No Action alternative was eliminated since it would allow the
VOC- and PCB-conlaminated groundwater to migrate off-site to possi-
ble receptors. Physical barriers including low permeability caps and
groundwater cut-off walls could be constructed to inhibit migration of
contaminants off-site. The low permeability caps would prevent rain-
fall from leaching additional VOCs and PCBs into the soil and the
groundwater. The cut-off walls would prevent horizontal seasonal migra-
tion of groundwater from the nearby stream to the perched zone and
subsequent migration of VOCs and PCBs in a horizontal direction.
Although the capping and cut-off wall are appropriate for preventing
intercommunication between the seasonal stream bed and the perched
zone, they would not prevent the intercommunication of groundwater
314 HAZARDOUS MATERIALS TREATMENT
-------
between the overburden aquifer and the bedrock at the site.
Like biological treatment of soils, treatment of groundwater can be
effective only if the microorganisms can use the contaminants as a food
source. For the same reasons biological treatment was not applicable
to VOC- and PCB-contaminated soil, it also would not be effective for
groundwater cleanup.
Using ozone to oxidize the VOCs and PCBs could be effective.
However, since ozone is not selective in its oxidizing ability, excess
ozone would be required to achieve the stringent water discharge limits
for PCBs. The health and safety issues relating to ozone usage in this
industrial atmosphere also do not support its application.
Adsorption of VOCs onto granular activated carbon(GAC) is a poten-
tial method for meeting the required effluent criteria. The adsorption
tendency of organic compounds is a function of molecular weight and
water solubility. Most of the VOCs found in the groundwater at this
site are not efficiently adsorbed onto GAC and therefore do not pro-
vide a high loading capacity onto GAC. However, PCBs found on-site
are readily adsorbed onto GAC and will preferentially replace VOCs
under loading conditions. Since there are high levels of VOCs and low
levels of PCBs in the groundwater, GAC treatment to remove the PCBs
coupled with a technology to remove the VOCs may be the most cost-
effective solution.
Air stripping has widespread application for the removal of VOCs
from groundwater since the technology makes use of the moderate to
high volatility of VOCs. Typically, air stripping columns are able to
remove 90% or more of the VOCs but are not able to effectively remove
PCBs. This alternative in conjunction with GAC treatment provides
the capability to remove the VOCs and PCBs to the required discharge
limits.
Based on the preliminary screening of alternatives, the following treat-
ment components were proposed.
Soil treatment
• In situ air stripping for VOC removal (pilot testing required)
• Asphalt capping and cut-off wall for PCB containment (pending in
situ air stripping testing)
Groundwater Treatment
• Pump-and-treat with GAC for PCB removal
• Pump-and-treat with air stripping for VOC removal
PILOT TESTING
An in situ air stripping pilot program was conducted to determine
if this technology would be effective for in situ treatment of VOCs at
the site, therefore enabling design of a full-scale system to remediate
the balance of the contaminated soils. The goal was to remove 95%
of the VOCs from the soils by in situ air stripping.
The design of the in situ air stripping system incorporated air with-
drawal wells and air inlet wells. A portable trailer'mounted blower was
connected to one of the five withdrawal wells. Flow and concentration
data from the withdrawal well and pressure readings from the air inlet
wells were collected to determine the effectiveness of the system during
a 1-mo period.
The concentration of Trichloroethylene (TCE) in the air removed from
the withdrawal well verses time is given in Figure 2. The figure shows
that the concentration of TCE in the discharge air initially ranged from
60 to 100 ppm TCE, but was reduced over time to a level of 30 ppm.
The initial large changes in concentrations were a result of increases
in blower speed.
The mass flow rate of TCE verses time is given in Figure 3. The
Figure shows that the mass flow rate of TCE discharged, which initially
ranged from 0.5 to 0.8 Ib/hr of TCE, was reduced with time to approxi-
mately 0.2 Ib/hr. Initially, the large changes in mass flow rate of TCE
were attributed to the increases in blower speed.
The cumulative mass of TCE discharged from the withdrawal well
verses time which reflects the change in concentration and air flow rate
is given in Figure 4. The slope of the graph shows that the TCE removed
was initially at the highest rate and asymptotically reduced over time.
PPM TCE
150
140 -
130 -
120 -
110
100
90
80
70
60
50
40
30
20
10
0
50
100 150 200 250
TIME (HOURS)
Figure 2
Concentration TCE vs. Time
300
350
400
IBS. TCE/HOUR
1
50
150 200 250
TIME (HOURS)
Figure 3
Mass Flow Rate TCE vs. Time
300
350
400
50
300
350
400
150 200 250
TIME (HOURS)
Figure 4
Cumulative Mass
TCE Discharge vs. Time
Based on observed differential pressure readings in the air inlet wells,
a minimum radius of influence of 20 ft was realized.
The data collected during operation of the in situ air stripping system
show that the technology effectively removes VOCs. According to mass
balance calculations using pre-treatment soil boring data and in situ
treatment system data, soils subject to treatment during the pilot test
HAZARDOUS MATERIALS TREATMENT 315
-------
were remediated from approximately 20 ppm to I ppm. There!i>re, dill-
scale in situ treatment will be u component of the final remedial plan.
REMEDIAL DESIGN FOR SOILS
To remediate the soils, a full-scale soil air stripping system will be
installed in the perched zone to remove VOCs from the unsaturated
soils. The system will consist of a series of withdrawal (air is with-
drawn from the soils) wells and air inlet (air is permitted to flow into
the soils) wells. The system will use a total of 32 air withdrawal wells
to remediate the soils coupled with their respective air inlet wells.
An asphalt cap and slurry cut-off wall was constructed atop and around
the PCB contaminated portion of the perched /.one. This consiruction
effectively isolated the area of the perched zone containing PCBs from
percolating rainwater and inflow from the nearby stream.
Approximately 25% of the ground surface above the perched zone
was bare ground or covered with gravel. This area was covered with
new asphalt material. In addition to this area, asphalt capping was in-
stalled following the installation of: (1) underground pipelines for (he
groundwater treatment system: (2) the in situ air stripping system; and
(3) the slurry cut-off walls within the perched zone area.
The slurry cut-off wall was approximately 800 ft long (Figure I).
extended IS to 20 ft below ground and was keyed into the undisturbed
silty clay below the perched zone. The slurry cut-off wall consisted
of a soil bentonite mixture that achieved a permeability rate of 1 x 10
-7 cm/sec.
REMEDIAL DESIGN FOR GROUNDWATER CLEANUP
Seven recovery wells were installed at various predetermined loca-
tions at the facility to collect groundwater for treatment. The wells were
constructed of carbon steel casings with stainless steel screens. One
recovery well was located in the perched water zone, four in the over-
burden aquifer and two in the bedrock aquifer. Due to the presence
of an oily product in the overburden aquifer, three of the recovery wells
included product recovery pumps and groundwater pumps while the
remaining recovery wells only included groundwater pumps. Oily
product is pumped to above ground storage tanks for off-site disposal.
Groundwater from the seven recovery wells is pumped to an influent
vault where two submersible pumps (one operating and one in standby
status) transfer the groundwater to one of two multimedia filters. The
multimedia filters (one operating with one in standby status) remove
solids, oil and grease in the groundwater prior to its entering two GAG
contactors, operated in series. PCBs are preferentially adsorbed in the
GAC units prior to entering the packed column air stripper where VOCs
are removed. Treated effluent from the air stripper is discharged to a
river via the existing municipal storm sewer system in compliance with
strict (NPDES) discharge limits. As shown in Figure 5. the process
equipment is housed in a 40-ft by 40-ft structure.
PRODUCT
«CCOVERT
TAMl >*XT|- "ŁŁ)*
*U'Ł«
\ - IJ>
Figure 5
Treatment System Schematic
INFLUENT
VAULT
GRANULAR
ACTIVATED
CARBON #2
EFFLUENT
AIR
STRIPPER
EFFLUENT
1/12/89
2/7/89
3/10/89
4/20/89
5/10/89
6/8/89
7/11/89
8/2/89
9/6/89
1/12/89
2/7/89
3/10/89
4/20/89
5/10/89
6/8/89
7/11/89
8/2/89
9/6/89
1/12/89
2/7/89
3/10/89
4/20/89
5/10/89
6/8/89
7/11/89
8/2/89
9/6/89
1,1 DCE
230
160
0
0
0
140
0
0
0
0
0
0
0
0
14
150
0
0
0
0
0
0
0
0
0
0
0
1.1 DCA
620
490
350
270
220
150
160
190
100
0
0
0
0
1
260
130
210
0
0
0
0
0
0
0
0
0
0
Table 1
Groundwater Analytical Results
1,1.1
t-1,2 DCE TCA TCE
19000
13000
9100
6400
6800
5000
6100
6000
5000
0
0
0
0
0
2900
5300
6200
4300
0
0
0
0
0
6
25
14
7
14000
5200
3500
2800
2600
2100
2700
2800
3200
0
0
0
0
0
0
2000
560
1200
0
0
0
0
0
0
0
0
0
1700
1600
900
610
630
590
640
680
1300
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
(Concentrations in ppb)
VC CA MC
0
0
0
0
0
0
0
0
0
0
0
0
0
6
24
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
TOTAL
VOCs
35550
20450
13850
10080
10250
7980
9600
9670
9600
0
0
0
0
19
3198
7580
6970
5500
0
0
0
0
0
6
25
14
7
PCBs
15
1.3
0
0
0
0
0.4
0
0.9
0
0 *
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
NPDES LIMITS
LEGEND:
36
1,1 DCE
1,1 DCA -
t-1,2 DCE
1,1,1 TCA
200
300
1,1 Dichloroethene
1,1 Dichloroethane
t-1,2 Dichloroethene
1,1,1 Trichloroethane
* No Samples Taken
NA Not Applicable
450
525
16
NA
NA
TCE - Tcichloroethene
VC - Vinyl Chloride
CA - Chloroethane
MC Hethylene Chloride
PCBs - Pol/chlorinated Btphenyls
0.2
316 HAZARDOUS MATERIALS TREATMENT
-------
Backwash from the multimedia filters is discharged to an off-line back-
wash holding tank. Solids are allowed to settle in the tank prior to pump-
ing the decant through a bag filter and into the influent vault. The average
and maximum design flow of the treatment system components are 400
and 600 gpm, respectively. Empty bed contact time (EBCT) for each
GAC contactor at average and maximum flow is 9 and 13 min, respec-
tively.
Analytical data collected since January, 1989 indicate that the system
is meeting design and regulatory requirements. VOC and PCB con-
centrations in the influent vault ranged from 35 to 8 mg/L and 15 mg/L
to non-detectable (less than 0.0002 mg/L), respectively, while the VOC
and PCB concentrations in the second GAC contactor effluent ranged
from 7.5 mg/L to non-detectable for VOCs and non-detectable for PCBs.
Effluent from the air stripper indicated maximum VOC and PCB con-
centrations of 0.025 mg/L and non-detectable, respectively, as shown
in Table 1.
CONCLUSION
The presence of PCBs and VOCs in the soil and groundwater present
a challenge in designing an effective remediation program. Evaluating,
selecting and implementing the proper innovative and traditional tech-
nologies can provide an effective and reliable treatment system as
described above. The groundwater and recovery pumps and ground-
water treatment system has been operating for approximately 7 mo. The
slurry cut-off wall and asphalt cap have been installed. Full-scale oper-
ation of the in situ air stripping system is anticipated to begin in late
1989 and continue through 1990.
HAZARDOUS MATERIALS TREATMENT 317
-------
Membrane-Like-Material Extraction of Oily Wastes
From Soils and Solids.
James Keane
Kenterprise Research Inc.
York, Pennsylvania
ABSTRACT
Membrane-Likc-Matenal (MLM) is a new kind of Liquid Membrane.
but it is not like those that have been investigated in recent years. It
has. instead, very low permeability to gases and oily compounds and
has the property of picking up a layer of oily material that is adherent
to a solid surface and leaving in its place a layer of water. The material
that has been picked up is then used in the process of formation of the
membrane, which occurs at ambient temperature.
The MLM effect has been used to strip many different oil compounds
from solid surfaces and experiments have shown that as much as 40%
bentonite clay can be present, and the water/oil exchange process still
works. The very First successful separation that was achieved in the
laboratory was on an oil-soaked Bentonite clay, for which the oil con-
tent was reduced to below the values that were obtained on the same
sample, using Soxhlet Extraction.
The process is emerging from the laboratory and offers the promise
of a new way to reduce the volume of materials contaminated with such
things as oils. PCBs, dioxin and other materials found at Superfund
sites, or at the aftermath of oil spills such as the Valdez Disaster. For
each of these applications, the key point is that once an oil-wet surface
has been converted to a water-wet surface. Further applications of oil
or solvents will not re-contaminate the solid surface.
MEMBRANE-LIKE-MATERIAL
This paper is based on the discovery that a temporary membrane will
form at the interface between certain solvents, water and a compound
extracted from Athabasca Bitumen. The membrane, called MLM. for
Membrane-Like-Material, will partition solvent and oily solutions from
aqueous mixtures, capture and hold oily materials from solid surfaces
and depress the residual amounts of partly water soluble solvents in
water. Two applications of the technology arc represented in (he
proposal: (I) one to remove oily contaminants from solid surfaces such
as soils and (2) the other to separate the oily part from the non-oily
part of the water used in the MLM soils cleaning process, or which
was taken into the process with the waste material.
THE MLM LIQUID MEMBRANE
MLM (Membrane-Like-Malerial) is a new chemical class of Liquid
Membranes, quite unlike those described in the literature. It is an
inclusion type of Liquid Membrane, so called because it includes the
target species during its initial formation, from which the included
material is recovered, following a further separation step. The MLM
appears as sludge, with the target species trapped in it. While it remains
within the water phase, it is very stable, with the target specie firmly
held. However, it does not form a compound with the target and it
promptly dissociates at the air to water interface. The extraction of oily
compounds using MLM-formmg solvents, cannot be compared with
similar extractions using repealed solvent wishes, especially when the
solvents are selected to be the same for each method. In an MLM bawd
extraction, the solvent itself is counted as an oily compound and b
removed with the target species. In solvent extraction, the solvent residue
may be many times larger than the volume or mass of the target species;
it usually is not counted as an oily compound, and must then be extracted
in a further step. In the case of soils, this solvent residue can amount
to 30 to 60% of the soil by weight and the most frequently cited method
of removing it is the application of heal. To compare the two methods
on the same basis, the Tina) removal of solvents for both methods must
be considered. When this is done, the comparison becomes moot, since
the nature of the comparison has then been changed to comparing the
extraction of solvent using MLM. to the extraction of the same solvent
using Heat.
Figure I - 8 show an experiment that was performed as an MLM
demonstration for the Valdez operation. The MLM stripping was com-
pared with conventional solvent extraction, in the presence of water.
As can be seen, the conventional method has no effect, but the MLM
method has completely removed the oil from the sand. The original
proposal consisted of two stripping stages, one an in situ spray-on method
for the beach, and the other a water-oil separation for the residues.
Thia proc-e«« d*«ion.t rat ion cenpare* • Naptha Solvent
Extraction, with HIM Oil to Mater Surface Ch««l«lry exchange. UN
•and xmple waa beach land that <•«» dried end soaked with
Peruvian Crude Oil, to Ululate beach sand.
318 HA/.ARDOUS MATKRIAl.S TREATMENT
-------
Figure 2
Split Sample
Figure 5
Naptha Side - Stirrea
Figure 3
Naptha - Left. MLM- Right
Figure 6
MLM Side at 10 Seconds of Stirring
Figure 4
Water Added to Both
Figure 7
Water Wash is Completed at 3 Minutes
The spectrum of target materials and associated applications for the
MLM extraction process, is related to the solubility of the solvent
selected and thus is very wide when the solvent is properly chosen.
It is fortunate that the best solvents for the formation of MLM are like-
wise the best solvents for oily compounds that are the most trouble-
some. On the basis of the criteria described, the actual oily residues
must be divided by the solvation ratio used immediately prior to MLM
extraction. Since these values can be as high as 40 to 1, oily residues
can be very low indeed. The MLM Process might therefore be described
more as a method of extracting a solvent rather than as a solvent ex-
HAZARDOUS MATERIALS TREATMENT 319
-------
Figure 8
MLM Clewed. Sample Tkken. Napthi Side Has Been Drained
MLM SOLIDS PROCESSING
The MLM Process was discovered in 1981 during work on a process
that used kerosene as the agent in a surfactant based water exchange
process. The purpose of the work was to develop a process suitable
lor the extraction of tar sand. The process under development was found
to be only 60% effective and had to be supplemented with soils washing
and water cleaning techniques to make it usable. The use of surfactants
at or near the Critical Micelle Concentration (CMC) was the attractive
feature of this older technology. Working at or near the CMC leads
to surfactant values near 200 mg/L instead of the 2% or 20.000 mg/L
and up for other methods. It was necessary to develop a Sand Washing
technique, derived from our then current water treatment program. This
resulted in three unit processes needed to make this system work.
The challenge was thus established by these difficulties, to find other
ways to exchange oil for water at a solid interface, once we knew that
it was possible. It was found that certain solvents, in the presence of
a component found in Athabasca Bitumen, would form a film at an
interface between a solution of the solvent and water. This film was
remarkably stable in the water phase, but obviously had one side ac-
tive in water. The other side was adjacent to the oily phase and initial
experiments showed that the film was not only stable in the oil phase.
but also that the film would dissolve until the area of the film equalled
the area of the interface. Other experiments showed that the film thick-
ness was not of molecular dimensions, but was thick enough to be
manipulated and had palpable tensile strength, considering its origin.
A condensation effect was observed that released oily compounds cap-
tured by the film, into the oil phase. If the area of the surface was sud-
denly reduced, then wrinkles and folds would occur in the skin, that
slowly dissolved, leaving a new layer of MLM that occupied the reduced
surface area. It was found that this material could be literally cut out
of the surface and lifted away into the water phase, where it could be
manipulated, measured and handled "like" a membrane. A new layer
of MLM would form in the cut out area, and over perhaps 30 min,
the region would be completely healed.
It was postulated that since the water side seemed to be so stable,
that perhaps the MLM could be formed adjacent to an oil-coated
panicle, with the water side facing into the panicle. It turned out that
conditions could be established to ensure the formation of the MLM
in the proper manner to effect removal of the oily compound. The com-
pound, with the solvent, and the bitumen extract in contact with the
panicle, was then lifted off and used to form the MLM. with the
introduction of water. Thus it became evident that the heretofore
impractical separations of oily materials from surfaces would be
possible.
The spreading rates were determined, using very simple sand sub-
strate experiments. They were approximately two orders of magnitude
greater than in surfactant based systems. The method used in these
experiments was to oil-wet a known surface area of sand, typically 2000
ft' and then to effect the separation, while determining the time for
the separate process to occur. This time was the spreading rate of the
MLM throughout the sand mass. It was noted that the MLM formed
a front that was remarkably well defined. In a 4 in deep sand bed in
a separator? funnel, the interface front moved through the sand with
a fully oil-wet to fully water-wet exchange occurring in only 0.125 in;
moreover, this front was flat, with no fingering.
The zone in which the exchange of oil and water was taking place
had a turbulence that could be clearly seen. Since the relatively heavy
solvent was being displaced upward by the lighter water, it was postu-
lated that the driving force was the weight of the solid particle falling
through the MLM after it was coaled with water. Later work showed
that cylindrical vessels could be used and also that die range of opera-
tion extended from large excess liquid amounts to almost dry, provided
thai the starting condition for the water exchange step was properly
determined.
Experimental results showed that the conditions of the soil and oily
contaminant, at the start of the MLM formation step, were important.
with the heavier molecular weight materials being harder to remove.
Once it was learned that the solvent used must penetrate all the way
to the actual solid surface, so that the MLM can form there, these
difficulties were circumvented. If the solvent does not reach the solid
surface, the MLM. if it formed as a layer, could be oriented the wrong
way or would form as an amorphous mass and be ineffective. If the
MLM formed in this manner, then the last molecular layer of oily
material would not be available to effect the formation of the MLM.
If the MLM formed an amorphous mass, then the layering, fssrntial
to effect clean separation, would not be present. The starting condi-
tions are the bench-level determinations of the best cuuccuuation of
solvent and oily compound that will permit the •*****& step to go
to completion. These conditions vary principally with the concentra-
tion of oily material present and the molecular weight of the target
species.
The greater the molecular weight of the compound to be captured.
the more solvent must be present at the solid interface for the MLM
to form. A typical concentration of material for which this process was
originally developed, was ISOjOOO to 200000 mg/L. of oily material.
with molecular weights all the way up to the heaviest components of
heavy tars and bitumens, for which we expected to see residues of under
20 mg/L of solvent plus oil, with a final solvation ratio, at the end of
the counter flow extractor, of as much as 40 to I.
CONVENTIONAL SOILS WASHING
We found during the early development work that surfactants could
act as an interferant in the MLM water exchange step. The puzzle was
solved when our glassware was divided into two parts. The first was
cleaned in conventional detergent cleaners, rinsed and dried. The other
was then cleaned using solvent extraction, followed by acid cleaning.
The extra steps in the cleaning process eliminated the interference effects.
This finding lead to attempts to model the surfactant operation, and
to re-investigate the question as to why the original surfactant-based
technology that we started with was only 60% effective in establishing
water-wet surface area.
The model that was proposed was that the oil-soluble tails of the
surfactant molecules would enter the residue oil layer, leaving the water-
soluble heads to form a layer that would then be above the oil layer.
This process inhibits all further oil removal, since neither the solvents
nor the surfactants can subsequently approach the oil layer underneath,
due to the presence of this water layer. WE believe that this effect is
real and that it represents a limit to the effectiveness of all soil washing
processes that use surfactants. It might be noted that this same limit
also applies to enhanced oil recovery. The surfactant wash will cer-
tainly increase the oily material recovery somewhat, but it can never
approach the performance of a replacement system where the oil is
physically replaced with water.
320 HAZARDOUS MATERIALS TREATMENT
-------
THE MLM PROCESS FLOW DIAGRAM
A simplified flow diagram is shown in Figure 1. This process flow
diagram is based on the detailed Mass Flow Process Diagram that has
been prepared for API Separator Waste. The actual flow diagram was
prepared from testing at bench-scale level on waste samples from a
refinery site. This particular site had material characteristics that varied
from 5% to 95% spent catalyst, which is an inherently water-wet
material, and which makes solvent extraction very difficult.
The operation of the MLM Soils Washing System, is described below.
A conveyor, such as a screw conveyor, feeds a constant stream of
input material into the receiving hopper of the MLM Solids Treatment
Unit, prior to the first mixer stage.
The material is then passed to a second mixer stage, where conven-
tional solvent extraction of most of the oil is performed. The solvent
extraction step is divided into two stages to allow two different solvents
to be used as a process option. A non-halogenated solvent is used in
the first stage and a halogenated solvent, suitable to form MLM, is
used in the second. Only one solvent extraction step is shown in Figure
9.
After solvent extraction in the third mixer stage, the oil-coated, solid
particles are then coated with solvent, plus the MLM Extract. Thus,
small amounts of contaminant materials on the soil have been largely
substituted with larger amounts of a solvent. The solvent extraction
process ensures that sufficient time has elapsed to allow the two oil
phase components of the MLM to penetrated all the way to the solid
surface. If the target contaminant is present at only the mg/L level,
then the composition of the solvent and MLM extract alone are adjusted
to ensure that MLM forms in the correct orientation and strength, once
the water part is added. Upon the completion of the solvent extraction
to the degree required, the solid material is pumped to the fourth mixer,
where water is added and the mixture is stirred to ensure complete
coverage of all of the material with MLM.
The MLM-coated material is then passed to mixer number five, where
the MLM residue is removed using a water flush. The water films around
each soil particle will not permit either oily material or solvent to re-
attach to the soil particle. Flushing must be sufficient in this stage to
effectively remove all of the MLM, and any solvent that may have been
used. Solvent residues are believed to be primarily from the interstitial
areas, where solvent resides after washing. The cleaned soil is then
sent to a dewatering unit, after which it exits the system and is ready
for disposal.
Meanwhile, the collected water from the oily rinse stage and the
MLM-flushing stage is returned to the MLM water treatment unit, where
all of the oily fraction is collected, also using MLM; the oil contain-
ment leaves the unit as a solvent-bearing sludge. This sludge is then
taken away for further treatment in a solvent recovery system.
The solvent recovery unit can be a Wet Still, where the condensa-
tion occurs in a counterflow water stream. Solvent recovery must take
into account the vapor pressures of the site contaminants that are
expected, to ensure that any with lower boiling points than the solvent
selected will be recovered separately.
EARLY CLAY TESTS
The MLM process was developed to recover Tar and Bitumen at ef-
fectively 100% recovery, leaving a water-wet residue. A definitive test
was made to determine the final residue of the solvent on the water-
wet sand and from a starting condition of 130,000 mg/L, the solvent
plus oil residue was determined by GC methods to be less than 20 mg/L.
Several dozen samples of various waste materials and tar and oil sands
were tested for separation efficiency. In all cases, separation was es-
sentially complete. For a deposit of light oil in Bakersfield, Califor-
nia, which is in a clay formation, complete separation was achieved,
even though the sample consisted of 40% bentonite, with the oil bound
on the clay. No other potentially commercial process has been found
to process this material.
The results on clay-bound oil proved to be important in that the process
was shown to be effective in removing oily compounds from soil, and
few soils have 40% bentonite. In the U.S. EPA SBIR Project, dioxin
was shown to be reduced by an order of magnitude for each extraction.
However, it is likely that the extraction was more effective than this,
IN SOLVENT
IN MLM EXTRACT
O04 IBS,
O1987 KENTERPRISE
|N API SEP. WASTE
OJfWKE
t TOH/tfi
EXCESS
WATER
OUT
a'tas^ ISiibY'is
WATER SLUDGE SOLIDS
(COKE)
OIL
(< I ppm SOL\ENT>
WATER-\NET SOLIDS
(< 20 ppm SOLVENT)
Figure 9
HAZARDOUS MATERIALS TREATMENT 321
-------
since the MLM concentrate could not be analyzed by the U.S. EPA
method. The MLM apparently binds the dioxin within its temporary
structure to an extent that defeats the GC analytical process. All of the
other parts of the sample were successfully measured, so that the dioxin
removal could be inferred, but it could not be measured.
In comparing removal efficiencies in clay-type soils relative to oily
wastes and other media, it should be pointed out that the two cases
cited compare a dioxin sample at 10 ppm starting concentration with
an oily waste at 190jOOO mg/L. The samples were, in fact, sandy material
at 0.020 to 0.030 in in diameter for the dioxin case and bentonite clay
for the oily waste case, with the particle size extending down into the
low micron size range. Thus, the comparison takes the highest oily con-
tent, on the Finest material and compares it with the lowest oily con-
tent on the most coarse material. This comparison shows a broad
effective performance range for MLM, regarding the soil type and par-
ticle size.
An additional example of (he potential application of this process is
the treatment of API Separator waste which is refinery residue that can
consist of large amounts of spent catalyst, sand, rust and other wastes.
A preliminary study, undertaken as a prerequisite to the project, con-
sisted of 10 different waste types including Tank Bottoms and Drilling
Mud residues which were processed in a demonstration. All of the sam-
ples were cleaned, yielding a solid residue that showed no signs of oil
in a solvent extraction step done after heat drying of the water-wet solids.
THE MLM SOLIDS PROCESSING
A typical selection of equipment includes a five-pass Ribbon Mixer.
consisting of identical stages, with the material lifted from one stage
to the next. Our requirements are that all of the mixers have a water
blanket, with the solvent contained all the time. Available mixers arc
able to meet these needs and one or more stages can be modified to
effect the MLM formation.
THE MLM PROCESS STEPS
In these discussions, "water-wet" means that the water is in direct
adherent contact with the surface and is a Surface Chemistry condi-
tion. Likewise, "oil-wet" means the direct contact of the oil with the
surface. Each condition implies the absence of an underlying layer of
the opposite species. It is not possible to change from one condition
to the other without intervention. The MLM process provides the
intervention necessary to move the surface conditions from oil-wet to
water-wet, after which it is not possible to again oil-wet the solid sur-
faces without a further intervention, such as heat far above the boiling
point of water.
The process consists of the steps discussed below. Water Wash is used
first to separate the non-oily material from the mixture, if the water-
wet material is in the form of fines and if it is not contaminated. An
example of that kind of material is spent catalyst in API separator waste.
If the contamination level is in the low ppm range, so that the fines
are likely to carry the contaminant out with the wash, this step may
not be used. An example of the latter is a very small amount of a material
such as PCB in the presence of a large amount of fine clay, where it
would be preferable to wet the fines with solvent, rather than to try
to separate them in a water wash.
For spent Catalysts in API waste, the fines have been found to be
water-wet and constituting as much as 99% of the total material present.
In this case, the oily material is attached to large particles and Us density
is likely to be well over 1.0, so that the relatively small amount of heavy
oily material separates well from the large amount of non-oily material.
MLM will form around large lumps of water-wet fines and prevent
the solvent from penetrating into the mass Thus, for wastes from API
Separators, the spent catalyst must first be removed with a water wash
and the fines will not carry out contaminants because they are not con-
taminated.
In many fine soil wastes, the water wash cannot be used, since the
fines are contaminated and will carry out the contaminant if a water
wash is used. The occurrence of soils with water-wet fines also has
been experienced. The sequence of steps therefore, needs to be estab-
lished on a site-specific basis using bench-scale tests
Solvent Extraction
A countercurrent solvent extraction is employed next to remove the
bulk of the oily material from Che oil-wet portion of the waste and to
solvent-wet any dry material that is present, using a solvent selected
to form MLM. If the oily part of the resulting mixture is in the high
ppm concentration range, such as in an oil sludge, a solvent may be
selected to optimize the solvent extraction step and not be chosen neces-
sarily for the MLM formation, which would then be done in the next
step.
Second Extraction
A halogenaled solvent, such as methylene chloride, or a suitable com-
pound already present at the site in large quantities, or p-xylene, together
with a small amount of an extract of Athabasca or other Bitumen, are
then used in a counter-current solvent extraction. When the oil and sol-
vent concentration reaches the level where MLM oil-to-water exchange
is possible, as determined from bench testing, water is added. This point
is determined in the laboratory, where the solvent color showing the
residual contaminant density and the dwell time for full penetration
of the solvent, are used to determine the initial conditions for the water
exchange to take place. The MLM, as it is being formed, then strips
and isolates the oily compounds, from the surface, replacing them with
a film of water.
As the mixing proceeds, the oil-to-water exchange process irreversi-
bly coats all of the material with water, even if it was originally dry
and had no oil on it, as would be found in a typical Superfund Site
where some uncontaminated material is mixed in with the oily frac-
tion of the contaminated material. The solvent is recovered and the con-
taminant appears as a concentrate. The extract of Athabasca Bitumen
is not recovered and goes with the contaminant, since its cost does not
justify recovery.
Water Wash
A water wash is employed to flush away all of the MLM that remains
with the waste after it has been weakened by further exposure to sol-
vent, if necessary. This material is collected by the water treatment sys-
tem and recovered for disposal. The cleaned material is water-wet and,
after being de-watered, can be returned to the site or disposed of in
a landfill.
The bulk of the oily residue will be carried in the solvent used for
the preliminary extraction. If this oil is a usable oil, the solvent selec-
tion for the first extraction might be kerosene, naptha or some other
solvent that will not interfere with the subsequent use of the residue.
in which case, the solvent need not be recovered.
The use of a halogenated compound for the solvent extraction step
will make it difficult for a refinery to accept the oil for processing.
since chlorine is not compatible with the catalysts used in the refining
process. p-Xylene is acceptable, however and, in addition, will perform
about as well as melhylene chloride. Its main drawback is that is very
difficult to handle in the pure state since it can convert large amounts
of pumping turbulence energy into electrostatic energy resulting in an
extreme fire hazard. Methylchloride also must be used alone and not
in the presence of the other two isomers, since only one of the three
isomers will form MLM, while the other two inhibit the formation of
MLM, even in the presence of p-xylene.
Problem Contaminants
If the oil is dioxin- or PCB-contaminaied, the recoverable oil is almost
certainly not commercially usable. The solvent selection will then be
optimized for best Target Material extraction, i.e., for best MLM for-
mation. Some of the oily material will appear in the solvent that will
form the MLM for the waste involved; the rest will appear in the MLM
sludge. This solvent will be selected from methylene chloride, p-xylene,
tetrachloroethane 1,1.1 or any other halogenated solvent. Work is
proceeding to find other non-halogenated solvents.
The solvents arc always used either under a water blanket or in a
closed container. The selection of a solvent must take into account the
vapor loss from leaks in the solvent recovery system, with a much
smaller amount of a solvent leaving the process with the solid residue.
322 HAZARDOUS MATERIALS TREATMENT
-------
We are very conscious of the need to select the solvent so as to be neither
phototropically active nor present a health hazard. The MLM-extract
from Bitumen is a high molecular weight material that is used in low
mg/L quantities in the MLM forming solvent. It is retained with the
oily material from which it can be extracted if necessary, but this is
very unlikely since it is relatively insoluble.
MLM WATER TREATMENT
Considerable portions of the MLM and free oils are likely to show
up in the wastewater stream. Also, a large amount of water is often
present with the oily waste material and is carried into the MLM process
with it. Conventional water treatment for the resulting wastewater is
difficult and expensive.
An MLM-based water treatment system for broad spectrum oily
materials is under development with a Phase II and Phase m SBIR
project for industrial laundry and other industrial wastes. This process
is an adaptation of the MLM process for solids to allow the formation
of large amounts of MLM, in the presence of salts such as sodium chlo-
ride, or calcium chloride and to form a floe. Adding air to this mixture
allows standard flotation techniques to be used for the separation of
oily materials.
Pilot plant tests have shown that chlorinated solvents, such as perch-
loroethylene, were reduced to concentrations near non-detectable in
water; additionally, the heavy metal content of the industrial laundry
waste was reduced to less than local municipal discharge limits in a
single pass. The heavy metal reduction was not expected until it was
realized that the oily contaminants were particles that included the bulk
of the heavy metals and these were removed en masse from the waste
water, without solid and oil separation.
The MLM Water Treatment Unit is rated at 20 gpm, but the hydraulic
limit is well over this, depending on the contamination level. The
capacity of the water treatment system is approximately in balance with
the solids processing system at the nominal flow ratings. The water solu-
ble components will not be extracted by the MLM Water Treatment
system, but will be treated in a following stage using a different method.
Wastewater containing oily contaminants is pumped to a static mixer
where a foam consisting of CaCL,, water, solvent and MLM Extract
is created. The resulting mixture is fed to a second static mixer, to which
a metered quantity of air is added. The resulting mixture is then passed
to a separation tank where the oily fraction is floated off and the clear
water, with its dissolved components, is fed to an outlet pump. The
water level in the tank is controlled by an automatic control valve on
the discharge side of pump, and water flow rate is controlled by an auto-
matic control valve on the discharge side of the drive pump.
The MLM water treatment unit will handle water containing over
20% oil and will reduce the oil concentration into the range of 1 to
10 mg/L, leaving water soluble materials in place.
ADVANTAGES OF THE MLM TECHNOLOGY
The goal was to design an Oil Production process that was environ-
mentally and economically sound. To meet both economic and environ-
mental needs, the process yield must be high. The most important
contribution that the MLM Process and related technology offers, is
that it allows the development of a realizable remediation process at
reasonable cost. The major cost of an oily waste remediation is not
the process cost, but is the handling cost of the materials. Excavation
and return of contaminated soil is expensive, but not nearly so great
as reburying or incinerating a hazardous material.
The MLM Process is based on a new principle, not previously known
or used. It provides an enhancement to the separation and treatment
option that has not heretofore been available. Working at the molecular
level, it introduces an impermeable barrier to the return of the con-
taminant that makes separation easy for all wastes that are oil soluble.
The separation step is a potentially low-cost extraction which is a com-
plete separation of soils and organic materials.
The MLM Process at it's presently developed stage will remove any
material that is soluble in the solvents selected, including the solvents
themselves.
WASTES AMENABLE TO THE MLM PROCESS
Oil wastes encountered in oil production, refining and transport will
be particularly amenable to treatment by the MLM soil washing process
as demonstrated by laboratory tests. It is expected that the MLM process
will also be effective in the treatment of halogenated and polycyclic
organic materials. Soils which are present together with contaminated
material can be either removed or processed through the MLM water
exchange step, without performance penalty.
SUMMARY OF MLM RESULTS TO DATE
A process evaluation procedure was developed for the rapid deter-
mination of yields greater than 99.9%. A mass balance was performed
on each sample tested to determine, by weight, the amount of solvent
residue lost on the sand or soil. An upper limit of 50 ppm of solvent
plus oil on a solid phase, was established, above which the sample would
be rejected and further process development would be performed. This
criterion was selected primarily as a commercial yield limit and it was
recognized that it was far better than the .then best mining practice. A
second test was established using a colorimeter with the reference cell
filled with clean solvent and the test cell filled with known amounts
of oil plus solvent. The upper limit of 50 ppm of residual oil was used
for the screening tests, but this value was never reached. Selected
samples were tested for solvent plus oil residue at an independent labora-
tory. The results were as follows:
• e a
• a a
•
• « 9
•
mm
mm
mm
"3
9
mm
mm
• •
W
i
)'
)[
3:
>•
1
n . •
» u
Figure 10
MLM Water treatment Unit
HAZARDOUS MATERIALS TREATMENT 323
-------
• No oily material was found that could not be processed to belter than
99.9 % removal after process adjustment.
• The sample size varied from 10 to 1800 g. Two groups of solvents
were used, one for preliminary stripping and the other for MLM
stripping of the oil plus the first solvent residue. Many solvents were
tried, both Halogens and non-halogens; with the exception of p-
xylene, no non-halogenated solvent would work.
• Solvents which would not form MLM were always found to be re-
moved to the oily side and none of them appeared as interferences,
so that oils could always be extracted without halogen residue with
an appropriate choice of solvent.
• Some of the materials tested are shown below. Sample 3 (athabasca
Tar Sand), which was the most intensely studied of the oil group,
was also tested to see if sea water could be used instead of fresh water:
as it turned out, sea water could be used. This result then led to tests
on Sample 12, the Ohio Brine, which demonstrated the existence
of the very strong partitioning effect between water-soluble salts and
other dissolved materials and non-soluble, or partly soluble male-rials
such as oils and solvents. The test also led to the approach which
was successful, for our industrial laundry wastewater cleaning project,
with the Phase II Project, for the SBIR Program of the U.S. EPA.
CONCLUSIONS
The MLM technology represents an innovative soil washing/contaminant
recovery process which can achieve significant reduction of various heavy organic
materials. The process satisfies ihe goals of the SITE emerging technology pro-
grant since its functional characteristics are applicable to a wide variety of wastes
in a variety of soils. The process also has the potential of being an effective
resource recovery technology which will substantially reduce the amount and
Contaminated Materials Tested
OILS
No. Source
1. Athabasca
2. Athabasca
3. Athabasca
3. Peru. S.A.
4 Columbia. S.A.
5 Utah. US
6. New Mexico
7. Alabama
8. California
API WASTES
9 Montreal
K). Calgary Can.
DIOXIN
II Weslon
BRINES
12, Ohio
Type
Tar Sand
Tar Sand
Tar Sand
(oxidized)
Heavy Oil
Heavy Oil
Tar Sand
Tar Sand
Tar Sand
Light Oil
API Waste
API Waste
sample dioxm
Brine
*OH
14%
10%
8%
14%
13%
8%
8%
6%
19%
25%
20%
-
6%
ppmRcmoval
140,00099.9 +
WOJ00099.9 +
80j00099.9 +
MOJXH99.998
130J00099.9 +
80J00099.9 +
80J00099.9 +
60/30099.9 +
I90J00099.9 +
250fl0099.9 -•-
200J000999 +
I090X)
60J00099.9 -t-
costs of environmental control requirements because of the uniquely high effi-
ciency of stripping oily and heavy organic materials from soils.
The process can be used effectively as a concentration step yielding cleaned
water-wet soils and low volume concentrated wastes which could be further
processed by other means such as incineration The process operating parameters
in ambient conditions offer the advantage of low cost simple and reliable operating
systems which should enhance operational acceptability
324 HAZARDOUS MATERIALS TREATMENT
-------
Bioremediation of Pesticides and Chlorinated Phenolic
Herbicides - Above Ground and In Situ - Case Studies
Harlan S. Borow
John V. Kinsella
ECOVA Corporation
Redmond, Washington
ABSTRACT
Remediation of hazardous waste sites requires the integration of
science, technology and engineering to cost-effectively cleanup complex
contaminants. Bioremediation, long recognized as an effective tech-
nology for treating petroleum hydrocarbons, is also effective treating
more complex compounds such as pesticides and chlorinated phenolic
herbicides. This paper will discuss the application of bioremediation
technology to more complex waste sites, the various bioremediation
technologies and actual case histories of complex site cleanups of soil
and groundwater using bioremediation in above ground and in situ
processes.
INTRODUCTION
Hazardous waste disposal represents one of the major environmen-
tal problems in the world today. Numerous methods and techniques
have been proposed and tried for treating and disposing chemical wastes
and their by-products to render them harmless to man and his environ-
ment. In spite of all the effort and money spent, no single technology
has evolved which is economically and technically satisfactory for all
waste constituents and matricies. One of the most promising technolo-
gies for this enormous problem, however, is the application of biotech-
nology.
Biodegradation, the microbial transformation of organic compounds,
has long been recognized as an effective process for the removal of toxic
chemicals from the environment. Biodegradation offers a relatively in-
expensive yet highly efficient method of removing toxic chemicals from
contaminated soils and groundwater.
Bioremediation, the controlled use of biodegradation to remove toxic
chemicals from soil and groundwater, is an effective and efficient
remedial technology for many complex sites, utilized alone or in
combination with other physical and chemical treatment strategies. The
purpose of this paper is to address some of the principles of biological
degradation of toxic chemicals and to demonstrate through actual field
experiences the effectiveness of bioremediation of contaminated sites.
DESCRIPTION OF BIOTREATABLE WASTE
Biological processes have been used for many years to remediate
petroleum hydrocarbons such as gasoline, diesel, crude oil and creosote.
More difficult compounds, such as pesticides and their derivatives (i.e.
phenoxyacetate herbicides, carbamates, and organophosphates);
chlorinated solvents (i.e., methylene chloride, trichloroethylene (TCE)
and vinyl chloride); and halogenated aromatic hydrocarbons (i.e., penta
chlorophenol, chlorinated benzenes and even some PCBs also can be
biodegraded successfully. Microbiological processes can also poten-
tially be used to transform and recover metals such as lead, cadmium,
mercury and chromium (Table 1).
The biochemical pathways for many contaminants found at hazardous
waste sites have been extensively studied in a range of
microorganisms'-2. The complexity of the environment and the
complexity of organic material substances often preclude easily predic-
table biodegradation results in hazardous waste sites. However,
Table 1
Examples of Superfund Wastes and Hazardous Constituents
Which Can be Treated by the Solid-Phase Processes
Spent halogenated solvents
and compounds from the
manufacture of chlorinated
aliphatic hydrocarbons
Wastes from the use and
manufacture of chlorinated
phenols, benzenes and
their pesticide
derivilives
Spent non-halogenraed
solvents
stet plating and
leaning wastes
Petrochemical products
and wastes, straight and
branched chain-alkanes,
gasoline, diesel, crude
oil, creosote and
refinery waste
1,1,2-Irichloroethylene CTCE
Chloroform
1,1,1-Trfchloroethane (TCA)
Tetrachloroethene (PCE)
1,2-Trans-Dichloroethylene ([
Methylene Chloride
1,1-Dichloroethane
1,1-Dichloroethene
Vinylchtoride
Carbon tetrachloride
1,2-Dichloroethane
-------
accumulated data and an understanding of mitrobial biochemistry make
it possible to generalize to some extent the relative rates of biodegra-
dation of compounds found at the site. Treatability studies are conducted
to establish degradation potential rates for specific site contaminants.
An understanding of the metabolic pathway(s) is required in order
to control and manipulate the environment to bring about optimum bio-
remediation at complex sites. The controlled biodestruction of com-
plex hazardous materials by natural microorganisms is not an accident
of nature but the systematic interaction of scientific knowledge (ecology,
physiology, genetic, chemistry and hydrogeology) with sound
engineering principles to maximize the desired metabolic reactions in
environmental cleanups.
Biological processes can be used to remediate water, soil, sludge,
sediment and other types of materials contaminated with organic and
inorganic constituents. A prerequisite to the development of effective
bioremediation processes for the aforementioned types of media is the
design of materials handling and engineering systems which ensure that
the contaminated material is processed into a form which is amenable
to bioremediation. Clays can be particularly difficult to treat because
of material handling problems and the tendency of the clay to keep con-
taminants away from the microbial cultures. This latter problem, one
of limited bioavailability, is one of the more challenging problems to
overcome from an engineering standpoint.
BIOLOGICAL TREATMENT TECHNOLOGIES
Biological treatment technologies for contaminated soils and ground-
water fall into four main categories: (1) solid-phase biotreatment (land-
farming); (2) slurry-phase biotreatment; (3) insitu biotreatment; and
(4) combined technologies with chemical or physical treatment. The
selection of a specific treatment process is a function of the physi-
cal/chemical nature of the contaminant, the contaminant concentrations.
the waste matrix and economic considerations (i.e.. overall cost, treat-
ment time-frame, etc.).
Solid-Phase Biotreatment
Solid-phase biotreatment relies on principles applied in agriculture
in the biocycling of natural compounds. The conditions for biodegra-
dation are optimized by aerating the soil with regular tilling and by
the addition of nutrients and water. Naturally indigenous microbial popu-
lations are diverse and often contain the appropriate microorganisms
for degradation of many site contaminants found in the contaminated
soils.
The rates of bioremediation of contaminated soils are enhanced by
optimizing oxygen levels, moisture content, available nutrients such as
nitrogen and phosphorous, pH and contact between the appropriate
microorganisms and the contaminants. This technique has been
successfully used for years in the managed disposal of oily sludge and
other petroleum refinery wastes through a process called landfarming.
Solid-phase biotreatment of contaminated soils is probably the most
widely used and cost-effective biotreatment technology applied today.
Typically, the process, illustrated in Figure 1, is used for petroleum
and creosote-contaminated soils. Typical costs for this type of treat-
ment are $40-$90/yd' but are highly dependent on conditions at the
(TO I
OPTUIAI
(TO UAJNTAJN
. UOnTURC)
- rotcf.
NUTRJOm CULTURC
(» MOLWttO)
«»OUN[ CMUIC'U *
AMA1TIIS
(A/tW RCGULATOHf APPROVAL)
Figure 1
Typical Solid-Phase Biorcmcdialion Diagram
site and materials handling costs. These costs compare favorably with
disposal costs that typically range between $250-$300/yd' for Class I
disposal. Recent federal regulation (RCRA, Land Ban) may even
prohibit disposal of some wastes due to fugitive emissions and leaching
of organics and metals, thus requiring treatment prior to disposal.
Generally, solid-phase bioremediation can be conducted without
extensive engineering of a treatment unit. As shown on Figure 1, con-
taminated soils are spread over an area of the site and treated using
landfarming techniques. When leachale collection is required, because
the waste is highly teachable and/or there is ground water near the sur-
face, leachatc collection systems using liners, trenches and/or wells can
be employed.
An example of a highly controlled solid-phase bioremediation system
was engineered and constructed by ECOVA to control volatilcs and
leachale The System consisted of a treatment bed which was lined with
a 80-mm high-density liner with heal welded seams on top of which
was placed clean sand. The sand provided protection for the liner and
proper drainage for con,laminated water as it leached from contami-
nated soils placed on the treatment bed. Lateral perforated drainage
pipe was placed on top of the synthetic liner in the sand bed for collec-
tion of soil leachale. The lined soil treatment bed was completely en-
closed with a modified plastic film greenhouse. An overhead spray
irrigation system contained within the greenhouse provided moisture
control and a means of distributing nutrients and microbial mocula (as
needed) to the soil treatment bed.
Contaminated leachale that drained from the soil was transported by
the drain pipes and collected in a gravity-flow lined sump. Leachatc
was then pumped from the collection sump to an on-site bioreactor for
treatment. Treated leachate was used as a source of microbial inocula
and reapplied (o the soil treatment bed through the overhead irrigation
system, after adjusting for optimum nutrients and environmental
parameters.
Volatile organic compounds which were released from the soil during
processing were controlled by an air management system, which was
attached to the soil treatment facility enclosure. As the volatile com-
pounds were released from the soil, they were drawn through the struc-
ture to the air management system.
Biodegradable volatile organic com,pounds can be treated in a vapor
phase bioreactor. Non-biodegradable volatile organic compounds can
be removed from the effluent gas stream by adsorption on activated
carbon. The design of choice will depend on the nature, concentra.tion
and volume of the air emissions, regulatory controls and cost-
effectiveness.
Soil heap bioremediation is a modification of solid-phase treatment
used when available space (area) is limned. In soil heap bioremedia-
tion, contaminated soil is excavated and stockpiled into a heap on a
lined treatment area to prevent further contamination. Microbial
inoculum (as needed) and nutrients are applied to the surface of the
stockpile and allowed to percolate down through the soil. The pile can
be covered and an air emissions recovery system installed as described
above. A leachate collection system is used to collect the fluid, which
is recycled. An internal piping system may be installed to blow air up-
wards through the soil and thus accelerate the biodegradation process
through the addition of oxygen. During operation. pH and moisture
content are maintained within ranges conducive to optimum microbial
activity. Typical costs are similar to conventional solid-phase treatment.
Composting processes are another modification of solid-phase treat-
ment in which the system is operated at higher temperature due to
increased biological activity. This technology would be used for highly
contaminated soils, treatment of poorly textured soils and in areas where
temperature is critical to the sustained treatment process. Contaminated
soils are mixed with suitable bulking agents, such as straw, bark or
wood chips, and piled in mounds. The bulking agent improves soil
texture for aeration and drainage. The system is optimized for pH,
moisture and nutrients using irrigation techniques and can be enclosed
to contain volatile emissions. Care must be taken for leaching control,
for volatile emissions control and that the bulking agent does not inter-
fere with the biodegradation of the contaminants (preferential carbon
source).
326 BIOREMEDIATION
-------
Slurry-Phase Soil Bioremediation
The biotreatment of organic waste in bioreactors has been an effec-
tive treatment system for wastewater, groundwater and other waste types.
Several commercially available bioreactor treatment systems have been
used for hazardous wastewater treatment3. These systems will not be
described here except to emphasize the need to optimize the advan-
tages of the system from both microbiological and operational perspec-
tives regardless of the choice of system. Some of the advantages of using
a bioreactor include:
• Greater process management and control
• Increased contact between microorganisms and contaminants (less
heterogeneity)
• Use of specific cultures or inoculum
• Decreased acclimation times and faster biodegradation rates
Slurry-phase bioremediation is a process where contaminated soils
are treated as an aqueous slurry in large, mobile bioreactor tanks. This
system maintains intimate mixing and contact of microorganisms with
the hazardous compounds and creates the appropriate environmental
conditions for optimizing microbial biodegradation of target con-
taminants. One disadvantage is the additional excavation and material
handling of the contaminated material that is often required. The slurry
retention time may be varied, as required, and the bioreactor has the
potential to operate in batch or continuous modes. Treated soils are
dewatered and the water containing high populations of acclimated
microorganisms is recycled. This process greatly reduces the acclima-
tion and treatment times for subsequent batches. The general schematic
of slurry-phase treatment is shown on Figure 2.
Figure 2
Typical Slurry-Phase Bioremediation Diagram
The first step in the treatment process is to create the aqueous soil
slurry. During this initial step, all stones and rubble greater than 0.25
in. in diameter are physically separated from the soil, and the soil is
mixed with water to obtain the appropriate slurry density. The water
used to make the slurry may be contaminated ground.water, surface
water or another source of water. A typical soil slurry contains approx-
imately 40% solids by weight; the actual percent solids is determined
in the laboratory based on the concentration of contaminants, the rate
of biodegradation and the physical nature of the soils. The soil is
mechanically agitated in a reactor vessel to keep the solids suspended
and the appropriate environmental conditions for enhancing biodegra-
dation maintained. Inorganic and organic nutrients, oxygen and acid
or alkali for pH control may be added to maintain optimal conditions.
Microorganisms may be added initially to seed the bioreactor or ad-
ded continuously to maintain the correct concentration of biomass neces-
sary for rapid biodegradation. The residence time in the bioreactor varies
with the soil matrix, physical/chemical nature of the contaminant (in-
cluding concentration) and the biodegradability of the contaminants.
Once biodegradation of the contaminants is completed, the soil slurry
is dewatered.
Depending on the nature and concentration of the contaminants, and
the local regulations, volatile emissions may be released to the at-
mosphere or treated to prevent emission. Because the soil is treated
in a contained process, a remediation system can be designed for soils
contaminated with a complex mixture of hazardous compounds. The
design of the slurry-phase bioremediation process can be modified to
treat soils that are contaminated with biodegradable semi-volatile and
volatile compounds, as well as some heavy metals.
The cost of treatment using slurry-phase bioremediation is higher
than other biotreatments per unit cost but is substantially lower than
incineration or direct disposal. Costs are influenced by materials
handling, retention times, equipment needs, volume of waste and reactor
designs. Slurry-phase reactors can be as simple as lined ponds or lagoons
engineered for slurry-soil treatment. Costs for slurry-phase treatment
typically range from $75-$150/yd3.
In Situ Bioremediation
In situ bioremediation is the biological treatment of contaminated
soils and groundwater without excavation, usually where contamina-
tion is deep in the subsurface or under buildings, roadways, etc. In situ
treatment involves the controlled management and manipulation of
microbial processes in the subsurface. This process requires an under-
standing of both microbiological processes relative to biodegradation
of the target contaminants and the soil physical and chemical environ-
mental effects on the microbial processes. Typically, these systems utilize
aerobic processes and involve the addition of oxygen as air, oxygen gas
A. IN SITU BIORECLAMATION USING
RECHARGE WELLS OR TRENCHES
MICROBES.
NUTRIENTS.
OXYGEN SOURCE
BIOLOGICAL
TREATMENT
BIOREACTOR
MAKEUP
WATER
RECHARGE
RECOVERY
B. IN SITU BIORECLAMATION
USING INFILTRATION
MICROBES.
NUTRIENTS. BIOLOGICAL
OXYGEN SOURCE TREATMENT
CLARIFIER
BIOREACTOR
RECOVERY
Figure 3
Typical In Situ Bioremediation Diagrams
BIOREMEDIATION 327
-------
or hydrogen peroxide and small amounts of inorganic nitrogen and phos-
phorous. Recent evidence has shown anaerobic processes may be ef-
fective in the biotreatment of, at least, BTX (gasoline) type
contamination.
The in situ bioremediation system usually is accompanied by a surface
bioreactor for treatment of the recovered groundwater which can then
be reinjected to enhance the subsurface active microflora. The design
and engineering of any system site is highly dependent on types of con-
taminants, permeability of the soils, regulatory constraints, contami-
nation of vadose zone, etc. Therefore, a design is highly dependent upon
the site and costs vary greatly. Two adaptations of in situ bioremedia-
tion are schematically shown in Figure 3.
Biological degradation of subsurface contaminants can be accom-
plished through delivery of an oxygenated nutrient solution to the zone
of contamination to stimulate natural microbial activity. Areas of
contaminated soils above the water table can be treated by artificially
raising the groundwater table. Water is cycled through the subsurface
using a series of recovery and recharge trenches or wells. Figure 3 sche-
matically depicts an in situ biorcclamation system using recovery and
recharge wells. Water recharge (using recovered groundwater and
supplemental makeup water) causes groundwater mounding in the water
table where the bulk of the contaminants are located. The water is reco-
vered in a downgradieni trench or well and is pumped to a surface bio-
reactor where it is treated to remove residual contaminants, amended
with nutrients and oxygen, and reintroduced into the subsurface. Water
may be oxygenated by sparging with air or pure oxygen or by adding
hydrogen peroxide. The system can be cycled by reversing the role of
the recharge and discharge wells to target zones if contamination is
located above the existing water table.
CASE HISTORIES
In Situ Bioremediation of Chlorinated Phenolic Herbicides
Shallow groundwater contamination was detected beneath a herbi-
cide formulation facility in 1981. The aquifer consisted of 35 ft of glacial
ourwash deposits; 25 ft of silly sand and clay overlaying 10 ft of coarse
sand and gravel which rested on shale bedrock. The major contaminants
were identified as chlorinated phenols, primarily
4-chloro-2-methylphenol (4C2MP). A pump and treat system (consisting
of 11 extraction wells feeding two activated carbon units) was installed
in 1983. Effluent from the system was returned to the aquifer via eight
injection wells. To achieve a more rapid reduction in contaminant lev-
els, an in situ program was evaluated in 1987.
Aerobic laboratory culture techniques were used to assess 4C2MP
biodegradation potential in the site groundwater. High 4C2MP biodegra-
dation potentials were observed in groundwater samples obtained from
three site wells (Table 2):
Table 2
4C2MP Concentrations In Aerobic Cultures
Bill Ł„ U
1-4 X - 1133 X • 1133
P-4 X - 3400 X • 3800
P-8 X - 710 X - 710
C, % Removed
X • <41 >%
X - 1380 60
X - <4I •'»
X - Average 4C2MP concentration (3 replicates).
C, » Initial Concentration.
Ck - Final Control Concentration (7 days).
C, - Final Tut Concentration (7 days).
High 4C2MP biodegradation potentials were observed with no
nutrient adjustment. This study showed that only aeration was needed
to reduce 4C2MP concentrations in the groundwater.
In 1988, the number of recovery wells was increased to 19 and two
additional injection wells were installed. Airlift pumps were placed in
the recovery wells, thereby increasing the oxygen concentration in the
injected effluent. Initial results are promising:
• Significant reduction in off-site contaminant plume size was effected
by gradient control of the recovery system.
• Decreased dissolved oxygen concentrations were initially measured
in the injection wells; this suggested that phenolic degrading microbial
populations had been established adjacent to the injection wells.
• In the initial 3 mo of operation, the total phenol plume exhibited
a 25 to 35% reduction in size; after 6 mo a 50% reduction was
observed.
Bioremediation of Pesticide-Contaminated Soil and
Groundwater - North Dakota
ECOVA Corporation was responsible for the cleanup of soil, surface
water and groundwater at a site in North Dakota after it had been
contaminated during a fire at a pesticide storage facility. Water used
to put out the fire carried large amounts of insecticides and herbi-
cides into the soil beneath the warehouse facility and into a nearby
creek which carried contaminants downstream. The principal con-
taminants were 2.4-dichlorophenoxyace.tic acid (2,4-D) and
4-chloro-2-mcthy Iphcrvoxy-acctic acid (MCPA), with lesser amounts
of trifluralin. alachlor. carbofuran and others.
The remediation program involved extensive material handling, soil
and material segregation and the use of several bioremediation
techniques. The remediation techniques included solid) and slurry-
phase biological treatment of soil, above-ground biological treatment
of water and in situ biodegradation. Activities during this project
included the following:
• Decontamination of over 12X>00 yd' of soil, containing from K) to
2.000 mg/kg 2,4-and other pesticides, in slurry-phase and solid-phase
bioremediation systems.
• Surface and in situ bioremediation along with surface granular acti-
vated carbon (GAC) treatment of over 5.000,000 gal of groundwater
to a 100 ug/L cleanup criterion established by the overseeing regula-
tory agencies. Operation of the GAC groundwater treatment units
was conducted in temperatures as low as -20°F.
• Separation of approximately 650 yd' of riprap from the soils,
followed by decontamination and placement in municipal sewage
lagoons.
• Construction of a sandblasting containment facility to decontaminate
200 yd'3 of concrete.
Initial feasibility studies were designed to establish the effectiveness
of biological treatment of pesticide-contaminated soils and groundwater
and to identify the treatment conditions needed to maximize biodegra-
dation of the compounds present at the site. Three treatment systems
were studied: water treatment and both solid- and slurry-phase treat-
ment of soils.
Contaminated water from the site which contained 100 mg/L 2,4-D
was biologically treated to below 1 mg/L within 4 days in the laboratory.
Laboratory studies on soils showed moderately contaminated soils could
be treated in a solid-phase bioremediation system to meet regulatory
criteria (total MCPA and 2,4-D=10 mg/kg-l). Two site soil samples
were incubated as soil slurries to determine biotreatability. One soil
was a highly-contaminated sample from the center of the burn site
(14,000 mg/kg-l 2,4-D) and the other was a moderately contaminated
sample from the edge of the bum site (400 mg/kg-l 2,4-D).
The inoculum used was strain JMP-134. At 4-day intervals, 0.5 mL
of a washed suspension containing approximately 10" cells/mL was
added to each inoculated slurry. Nutrients (nitrogen and phosphorus)
were added at the beginning of the incubation.
When moderately contaminated soil was treated in a soil-slurry bi-
oreactor. the contaminant concentrations declined on average from 390
to 15 mg/kg over 16 days.
In the highly-contaminated soil, the 2,4-D concentrations declined
from 13,200 to 2,610 mg/kg in nutrient amended soil and to 2,220 mg/kg
in soil which had been inoculated and nutrient-amended. The results
showed that even highly contaminated soil could be rapidly treated in
a slurry-phase reactor. Further, soil with contaminant concentrations
328 BIOREMEDIATION
-------
near the expected average for the burn site area (200 mg/kg of 2,4-D)
could be treated to achieve the regulatory criteria in approximately
2 wk. The systems became nutrient limited after initial growth and
nutrient additions produced the lowest final contaminant concentrations,
particularly evident in the highly contaminated soil. Inoculation with
2,4-D degrading bacteria had no effect on the rate of biodegradation.
Using the treatability data, solid-phase biological treatment techniques
were implemented to remediate 10,000 yd3 of soil contaminated with
the complex mixture of herbicides and insecticides as illustrated in
Figure 4. To treat this soil, ECOVA designed and constructed a soil
treatment area approximately 5 ac in size. The treatment area was con-
structed with an engineered clay liner 12 in. thick and a drainage system
to control water movement both inside and outside the facility.
Figure 4
Integrated Bioremediation Treatment Approach
Following construction of the solid-phase treatment facility, 10,000
yd3 of soil were removed from the burn site and contaminated creek.
The soil was spread on the treatment bed to an average depth of 15 in.
During the 3 mo of field operations, soil conditions were optimized
for biological activity by daily tilling and by maintaining the soil moisture
content between 8 and 15% by weight. The combined 2,4-D and MCPA
concentrations decreased from 86 to 5 mg/kg during the 3 mo of
operation of the solid-phase treatment facility (Table 3). In addition,
over 1,000,000 gal of contaminated water were treated biologically. The
water was treated in on-site bioreactors and then either discharged or
applied to the solid-phase bioremediat,ion facility to maintain moisture
content.
Table 3
2,4-D and MCPA Concentrations During Field Operations
o
7
25
33
55
77
2,4-D Cone. MCPA Cone.
Mean S.D. Mean S.D.
mg/kg-
41.8 63.5
17.8 16.7
4.6 5.1
4.0 4.9
25 2.4
4.0 3.5
44.2 315
32.3 24.5
23.0 12.5
16.0 15.2
<5.0' -
1.2-0.6
Total
86.0
50.1
27.6
20.0
7j
5.2
* Not detected in all samples, detection limit of 5.0 mg/kg.
•• Maximum value, since only 7 of 24 analyzed samples had detectable MCPA at a detection limit of 1
mg/kg. For samples with no detectable MCPA, a value of 1.0 mg/kg was used for calculations.
Slurry-phase soil bioremediation techniques were used to treat more
than 750 yd3 of soil contaminated with up to 1,500 mg/kg 2,4-D and
MCPA. Three slurry bioreactors capable of treating 26,000 gal of fluid
were mobilized to the site along with equipment to slurry the soil and
optimize the biodegradation process. Material was withdrawn from a
stockpile of highly contaminated soil and added to a trommel unit that
slurried the soil and separated out stones and rubble greater than 1/4 in.
in diameter. The slurry was then pumped into 26,000 gal bioreactors.
Each reactor was capable of holding approximately 60 yd3 of soil.
Temperature, pH and dissolved oxygen were optimized to increase the
rate of degradation.
Biodegradation of pesticides in the soil slurry reduced 2,4-D and
MCPA levels from 800 mg/kg (400 mg/kg in the slurry) to less than
10 mg/kg in 13 days (Table 4). The estimated 2,4-D half-life was 2.1
days over this period, again similar to that observed in the treatability
study with moderately contaminated soil. Upon completion of the bio-
logical treatment, the slurry was spread onto the solid-phase treatment
facility.
Table 4
2,4-D and MCPA Concentration During Operation of the
Soil/Slurry Bioreactor
2,4-D Cone. MCPA Cone.
Day (mg/kg) (mg/kg)
0
2
4
6
13
204
104
10
5
3
186
177
246
51
ND
ND = Not Detectable. Detection Limit = 15 mg/kg.
Groundwater treatment continued during the winter months of
1988-89. A hydrogeological assessment revealed subsurface ground-
water contamination in three areas: (1) the burn site, (2) a subsurface
location 1,800 ft downstream of the burn site and (3) the impoundment
(an area blocked off to contain the run-off water from the fire). The
contaminant level at the burn site has been reduced so that only
monitoring is necessary. At the latter two areas, long-term recovery
systems were designed and built to recover and treat over 5,000,000
gal of groundwater by carbon filtration and in situ biodegradation which
reduced the treatment time by half.
An upgradient injection gallery was established to flush treated water
and nutrients, as required, through the contaminated plume. Downgra-
dient recovery well and trenches recover treated groundwater. During
treatment, the groundwater was monitored to guard against off-site
migration.
Additional groundwater treatment was accomplished by using auto-
mated GAC treatment units. These units, consisting of sand filters, GAC
filters and automated control systems, successfully treated groundwater
at the site through sub-zero temperatures.
This case history demonstrates that microbiological processes can
be used to develop cost-effective, onsite remediation systems for
hazardous waste sites.
The initial laboratory treatability studies showed that both water and
soil at the site were amenable to bioremediation. These studies also
provided reasonable estimates of degradation rates in the three treat-
ment systems used at the site: above-ground water treatment, solid-phase
treatment of moderately contaminated soil and slurry-phase treatment
of highly contaminated soil.
When the project was complete, over 12,000 yd3 of soil, riprap and
concrete and 6,000,000 gal of water had been decontaminated for less
than half the cost of off-site disposal. The site has been restored to it*s
pre-contaminated state and can be redeveloped by the owner.
Solid-Phase Bioremediation of
Contaminated Soil - California
A former manufacturing facility producing heavy equipment for over
65 yr had soil contaminated with volatile organic compounds and
hydrocarbons. Motor oil, diesel fuel and cleaning fluids had been stored
at the site during its operation. During demolition of the plant, soils
in two areas of the plant were found to be contaminated. These con-
taminated soils (approximately 16,000 yd3 were excavated and stock-
piled for remedial action. ECOVA's objective was to biotreat the
BIOREMEDIATION 329
-------
stockpiled soils using solid-phase bioremediation to reduce contaminant
concentrations to a target level that would allow disposal of the treated
soil in a Class III landfill (100 mg/kg TPH)
Preliminary chemical evaluation (35 samples) detected TPH concen-
trations ranging from detection limits (lOmg/kg) to 16.000 mj; kj; with
an average concentration of 1,275 mg/kg. Aerobic microorganisms were
relatively abundant ranging from 10 to 10" cells/g (well. Bench-scale
biotreatability evaluations indicated that biodegradation of the petrole-
um contamination could be stimulated relaiivel\ rapidly; h> the fourth
week of treatment, the TPH concentration was reduced to below 100
mg/kg (Fig. 5). This biodegradation occurred with the addition of
nutrients (nitrogen and phosphorous) and aeration of the soils Inocu-
lation with microorganisms which degrade diesel fuel constituents was
not necessary.
Hydrocarbon Contaminated Soil
Lab Biotreatability
TPH/IR
2000
1500
1000 -
500-
ULL
34567
Time/Weeks
I Active mm Abiotic Control
Figure 5
Bench-Scale Evaluations
For ECOVA's solid-phase bioremediation of the site, the contaminated
soils were spread over the treatment area. Due to the limited treatment
area, the contaminated soils required treatment in two lasers (30 in ).
Over-sized material and debris (concrete, etc.) were removed from the
treatment zone during spreading. Because little rainfall was expected,
no liner was installed. Treatment operations consisted of daih opera-
tion of the upper lift (18 in.) with the soil stabili/er ihg 6)
6ASQJNC O4OACA1.
WICMOftM. AMAIYSS
^"^ NUTKCHTS \ CULTUHt
' (f «ŁOU«CO|
| ) LTIl-
/ /~^~' «.axiV«~'W
r /- •'! -rt"
fi 15- un /
12 is" un •/
J3U,,
0 P^lv.
OCAJWUK CU.TAM..
^
Mil -M >:>UW'CA, fc
1
It VJl ~ ~ ~ ~
.x— j
l»Al
CHCUlCAl AMAi TVI
(AFTER RCOULATOAY APPROVAL)
Figure 6
Process Mow Diagram
Moisture and nutrients (N & P) were added using l-.COVA's terraga-
tor which tills the soil and adds nutrients in a single pass A nutrient
mixture developed in KCOVA's laboratory was applied at a rate deter
mined in the laboratory and monitored periodically during operation
Remediation monitoring included soil sampling for TPH from 35
separate cells, air monitoring samples for volatile contaminants,
microbial enumeration and nutrient (NH'4) concentration.
Two applications of nutrients were required during treatment (Fig. 7)
to maintain optimum microbial activity but not overload the system with
nutrient*. The microbial population increased rapidly and stayed at a
high level throughout the remediation process (F-ig 8). Treatment of
the second lift was completed within 4 wk to target concentration.
Ammonia Concentration During
Bioreclamation of Contaminated Soil
mfl NH3/kg Soil
70 T
60
SO -
40 ', '•
30 ji
20 <
10 • \
•
0 ''
10
20 30
Days
* NH-, -*
Figure 7
Nutrient Applications
Enumeration of Microorganisms
Solid-Phase Field Remediation
Colony Forming Unita/gm Soil
1 UUUfUB
1 OOOE«07
1 OOOE-08
1 OOOE-05
1 OOOE-04
1 0006-03
1.000E-02
1 OOOE-01
(
^
* 7i^~^i^ ^ t
^ _. .
/
/
/
_^-Non-d«l»cllt>l« (PlOt B~5)
• '
) 10 20 30 40 50 60 70
Days
PCA
• LSB/Succmata
Constitutive
I
Figure 8
Microbial Population
This case study exemplifies the utility of biological technology in
the rapid treatment of petroleum) contaminated soils. The cost of treat-
ment was $65.00 yd'.
CONCLUSION
Bioremediation is a technical!) feasible and cost-effective treatment
for a wide range of wastes Effective design and implementation of bio-
remediation systems relies on a detailed understanding of the ph\si-
cal.'chemical nature of the contaminants and the site soils and
groundwater. Much of this information can be obtained by conducting
carefully designed bench- and pilot-scale (rcatabilit) studies. Selection
of (he most effective process, whether it be solid-phase, slurry-phase
or in situ systems, depends upon analysis of this type of information
by experienced professional scientists, microbiologists. chemists.
330 BIOREMEDIATION
-------
hydrogeologists and engineers. The case studies presented here are but REFERENCES
a few of the examples of successful applications of bioremediation sys- L Gibson DT _ Microbial Degradation of Organic Compounds, Marcel Dek-
tems to a wide array of hazardous wastes. ker, Inc., New York, NY, 1980.
2. Atlas, R.M., Petroleum Microbiology, Macmillan, New York, NY, 1984.
3. Grady, C.P. L., Jr. and Lim, C, Biological Wastewater Treatment, Marcel
Dekker, Inc., New York, NY, 1980.
BIOREMEDIATION 331
-------
Bioremediation Of Hydrocarbon-Contaminated Solids
Using Liquid/Solids Contact Reactors
Hans F. Stroo, Ph.D.
John R. Smith, Ph.D.
Michael F. Torpy, Ph.D.
Mervin P. Coover, M.E.
Randolph M. Kabrick, Ph.D.
Remediation Technologies, Inc.
Kent, Washington
ABSTRACT
The use ofliqu id/sol ids contact reactors (LSCs) for hioremediation
is increasing because these treatments offer rapid on-site treatment with
low land area requirements and little risk of off-site contamination. LSCs
are biological treatment systems operated to maximize mass transfer
rates and contact between contaminants and microorganisms capable
of degrading the contaminants. The purpose of this paper is to provide
an overview of the concepts and current results from using LSCs for
the cleanup of soils and sludges contaminated with hydrocarbons. The
data show that LSCs can sustain high numbers of microorganisms and
rapid respiration rates in comparison to unmanaged soils or simulated
land treatment. The results also show that microbiological monitoring
of LSC reactors is critical for optimizing performance during treatment.
Representative data from laboratory studies and Held applications
demonstrate the usefulness of LSCs for destruction of hydrocarbons
in both oil refinery sludges and wood preserving wastes. Results of pilot-
and full-scale testing show that LSCs can achieve contaminant removal
rates much greater than those typical of land treatment.
INTRODUCTION
Bioremediation is a proven cleanup technology for soils and sludges
contaminated with heavy hydrocarbons such as creosote or oil (1). The
use of liquid/solids contact reactors (LSCs) for bioremediation is
increasing because it offers rapid on-site treatment, with low land area
requirements and little risk of off-site contamination. The purpose of
this paper is to provide an overview of the concepts and practical aspects
of using LSCs. After a brief discussion of the rationale behind the
development of LSC technology and the basic concepts involved, the
results of laboratory and field testing of LSCs will be presented. Results
of microbiological monitoring will be discussed first, both to demon-
strate the increases in numbers and activity which arc achievable, and
to introduce the critical operational parameters which must be controlled
to maximize the effectiveness of LSC treatment. Finally, representa-
tive data from laboratory studies and field applications are presented
to demonstrate the usefulness of LSCs for destruction of hydrocarbons
in different organic wastes.
BASIC CONCEPTS
Rationale for LSCs
Bioremediation has become recognized as an effective and cost-
efficient approach for on-site cleanup of a wide variety of hazardous
organic wastes'. Historically, land treatment has been the principal
bioremediation method for contaminated solids, and it is still the least
expensive and most widely used alternative2•'. However, recent restric-
tions on land disposal (Federal Register, 40 CFR Part 268). driven by
concerns over off-site migration, will require that some form of con-
tained bioremediation be used in mam cases, either as an alternative
form of treatment or as a prelreatmcnt before land application in many
cases. Contained bioremediation alternatives include enclosed land treat-
ment (in lined reaction cells with emission controls if needed).
composting and LSCs Of these alternatives. LSCs are capable of
producing the fastest and most effective cleanup in many cases, and
field experiences have shown that the technology can be readily
implemented and completed at reasonable cost.
Besides the ability to contain wastes and thereby reduce the risk of
off-site contamination, LSCs also have the important advantages of rapid
reaction rates, low land area requirements and a high degree of flexi-
bility The rapid reaction rates achievable in LSCs result from the ability
to maximize mass transfer rates and provide optimum conditions for
microbial activity. Because solids can be treated rapidly in contained
reactors, much less area is needed for on-sitc remediation than during
land treatment. Also, bench-scale LSCs can be used in laboratory test-
ing to establish the feasibility of biorcmediation more rapidly than simu-
lated land treatment.
The flexibility derives from the fact that the wastes are contained
in a relatively homogenous form in an engineered system designed for
precise and rapid control of environmental conditions. This flexibility
allows for the use of a variety of different biological treatment options
as well as the use of biological treatment in conjunction with chemical
and physical treatment procedures, such as soil washing or phase sepa-
ration. The variety of options available for mixing and aeration allows
treatment of a wide variety of materials w ith differing handling charac-
teristics.
Technology description
Liquid/solids contact treatment is analogous to conventional biological
suspended growth treatment (e.g., activated sludge). LSCs are designed
to relieve the environmental factors commonly limiting microbial growth
and activity in soil (principally the availability of carbon sources.
inorganic nutrients and oxygen). To achieve this goal of maximizing
biological activity, the wastes are suspended in a slurry and are mixed
to maximize mass transfer rates and contact between contaminants and
the microorganisms capable of degrading those contaminants.
Aerobic treatment in batch systems has been the most common mode
of operation, but LSCs are sufficiently flexible to allow anaerobic treat-
ment at a variety of redox potentials, or aerobic/anaerobic cycling. LSCs
can be operated in single batches, in sequenced batch reactors or in
semi-continuous or continuous feed. LSC treatment can be performed
in contained mobile reactors or in lined in situ lagoons (Fig. 1).
A principal goal of the mixing and aeration is to supply sufficient
oxygen throughout the slurry matrix to prevent oxygen transfer limita-
tions to activity which generally occurs when oxygen must be supplied
by diffusion over even short distances. Mixing can be provided by
332 BIOREMEDIATION «
-------
aeration alone or by aeration and mechanical mixing. Aeration can be
provided by floating or submerged aerators or by compressors and
spargers.
Chemicals added to LSC reactors include nutrients and neutralizing
agents to relieve any chemical limitations to microbial activity. Other
materials, such as surfactants, dispersants and cometabolites (compounds
supporting growth and inducing degradation of contaminant compounds)
can be added to improve materials handling jcharacteristics or increase
substrate availability or degradation.
Since the overall goal of LSC operations is to maximize microbial
numbers and activity, microbiological monitoring can be used as an
inexpensive monitoring parameter to provide rapid feedback on per-
formance. The next section of this paper gives typical data on microbial
numbers and activity as well as examples showing the use of these data
to optimize LSC performance.
Figure 1
Schematic Diagram of ReTeC's Mobile LSC Reactor,
Showing Various Options for Mixing and Aeration
MICROBIOLOGICAL MONITORING
Microbial numbers
Biodegradation of hydrocarbons can be viewed as proceeding by the
following stoichiometry, in which biomass is produced initially (Eq. 1)
and then is eventually completely degraded (mineralized) to carbon di-
oxide and water4:
C7H,2
C5H702N
502 + NH3 -» C5H7O2N + 2CO2 + 4H20
50
5CO2 -I- 2H2O
NH3.
(1)
(2)
The first phase can occur much more rapidly than the second, so
that biomass can increase 3 to 4 orders of magnitude during the initial
stages of degradation in batch reactors. Under typical operating condi-
tions, the microbial population densities increase dramatically at the
start of LSC treatment and then stabilize and slowly decline (Fig. 2).
Additionally, the organisms responsible for PAH biodegradation are
capable of rapid increase both in total numbers and in relative
abundance. Microbial population densities remain high for a longer
time in sludge with higher levels of contamination, reflecting the fact
that substrate depletion occurs after most of the readily-degradable
material has been used.
As indicated in Figure 2, the lag period commonly observed when
hydrocarbons are added to a pristine environment is generally brief
during LSC treatment. The lack of an apparent lag period results from
the rapid growth rate in LSC reactors and the fact that appropriate
acclimated organisms are generally present in these materials, since
the sites have a long history of exposure to the contaminants. As a result,
microbial inoculation usually is not beneficial5, although it can be use-
ful in some cases, such as the cleanup of highly concentrated wastes
or extremely recalcitrant compounds6.
To some extent, the success of enhanced bioremediation can be gauged
by monitoring microbial numbers. Generally, unmanaged soils and
sludges have population densities on the order of 106 cells/g solids
(Table 1). The numbers typically increase to between 107 and 10s
cells/g during land treatment of soils or sludges. Numbers of recovera-
ble cells (which represent 1 to 10% of the total number of cells present)
rarely exceed 108 cells/g in soil, even when high levels of organic
matter are present7. This upper limit to population densities during
land treatment is probably a result of the diffusion-limited, nutrient-
and oxygen-supplying capability in soils. However, during LSC
operations, the numbers typically range between 108 to 109 cells/g
when soils are treated, while for sludges, which have relatively high
organic carbon contents, microbial populations generally range from
109 to nearly 10'° cells/g solids. These results reflect the success of
LSCs in maximizing microbial growth and demonstrate that LSCs are
particularly useful for highly contaminated sludges with high oxygen
demands.
r--
«
o
UI
o
400
300
200
100
High contamination
Low contamination s . Jotal c»lls
10 20 30 40
TIME (DAYS)
50
60
Figure 2
Number of microorganisms in LSC reactors over time during batch operation.
Open circles represent total aerobic heterotrophic microorganisms and closed
circles show numbers of PAH-degrading microorganisms. Results are shown
for creosote-contaminated soils with high or low PAH concentrations.
Respiration
Direct measurements of microbial activity are a more important
measure of performance than microbial counts or biomass estimates,
because the amount of activity per cell or per gram of biomass can
vary widely. For example, oxygen uptake rates (OURs) were measured
at various times during operation of the LSCs described in Figure 2.
The results (Fig. 3) show that the specific OUR varied over time, so
that microbial numbers were not necessarily correlated with activity.
This was true whether the specific OUR was expressed as a function
of the microbial biomass as measured by plate counts (Fig. 3A, assuming
10'2 cells/g) or as a function of the volatile suspended solids (Fig. 3B).
The pattern seen in the low contamination reactor is the most common,
in which the organisms are relatively active initially and specific OUR
then declines with time.
Because respiration should be directly correlated with overall bio-
degradation, OUR measurements in an aerobic reactor are the best
measures of performance and the effects of amendments. Equations 1
and 2 suggest that complete biodegradation will require nearly 4 g
O2/g Carbon (C) mineralized, so that the progress of biodegradation
can be estimated by measuring cumulative oxygen uptake as a proportion
of the total organic carbon.
For example, the data in Figure 4 are taken from LSCs operated with
creosote-contaminated material from three separate sites. The solids
were all creosote-contaminated impoundment sludges, similar in physical
and chemical characteristics, except that the sites had varying concen-
BIOREMEDIATION 333
-------
trations of the wood preservative pentachlorophenol (PCP), which is
highly toxic to microorganisms. The ratio of cumulative O3 uptake:
TOC content over one month varied from 1.9 (half the theoretical
maximum) in the sludge with low PCP to 0.7 in the sludge with the
highest PCP concentration (near 3000 mg/kg). Figure 4 also shows that
respiration in the site C material during LSC treatment was approxi-
mately four times faster than that measured during simulated land
treatment.
Table 1
Microbial Population Densities in Samples from Unmanaged Sites
(Time 0 samples before treatment) and In samples taken during
steady-state operation of either land treatment demonstrations
(LTD) or liquid/solids reactors (LSC) with the same
wastes used in laboratory Instability studies.
Site No. Sample Type Treatment Total Aerobic Heterotrophs
(CPU/g dry weight x 10"1
Creosote
1 Soil Time 0
LTD
LSC
2 Soil Time 0
LTD
LSC
3 Sludge Time 0
LSC
Oil Refinery Wastes
4 Soil Tine 0
LTD
LSC
S Soil Time 0
LTD
LSC
6 Soil Time 0
LSC
7 Sludge Time 0
LSC
8 Sludge Time 0
LSC
Coal Gasification Hastes
9 Sludge Time 0
LTD
LSC
10 Sludge Time 0
LTD
LSC
0.
8.9 -
22 -
0.1
2.0
22
0.2
440
1.
1.1 -
59
0.
0.3 -
5.3
0.
3.5
0.
S5 -
0.
44
0.
4.0 -
380 -
0.
8.4 -
130
3
15.3
94
4.7
9.6
57
0.6
1600
4
9.7
107
07
0.6
8.8
4
10.2
4
895
8
157
04
25
1450
09
17.3
400
OPTIMIZING PERFORMANCE
Microbiological monitoring can be used to assess the impact of
alternate operating practices and to ensure adequate performance during
operations. Obviously, any aerobic biological treatment process must
be operated to maintain adequate pH (generally from 5.5 to 8.0),
dissolved oxygen (in excess of 2 mg/L) and salinity levels. However,
there are operating considerations that are unique to LSCs or deserve
extended discussion.
Nutrients
Equation 1 suggests that nitrogen (N) availability can be an impor-
tant factor controlling biodegradation rates, and that the demand for
N during the initial stages of LSC operations can be very high. This
N will be recycled eventually (Eq. 2), so that N must be supplied in
large amounts initially but in lesser amounts as treatment progresses.
To estimate N demand, we can use the microbial numbers presented
earlier. A typical cell density during LSC operations may be roughly
10' cells/mL, or approximately 1 g/L dry weight of cells. Since
microbial cells are approximately 50% C and have a C:N ratio of
roughly 5:1', this means that 100 mg/L of N will be required for the
initial rapid population increase.
However, the demand for nitrogen, and other nutrients (especially
P), is hard to predict and can represent a major cost in LSC opera-
tions. It is therefore generally necessary to empirically determine the
nutrient demands. Respiration monitoring can aid in this determina-
tion. For example, in the case shown in Figure 5, 100 mg/L of nitrate-
N were added to all reactors at start-up, and 50 mg/L were added at
the indicated times to simulated LSCs containing varying amounts of
solids with a TOC content of 24%. Calculations based on the responses
observed at varying solids loadings indicate that the inorganic N required
was equivalent to a C:N ratio between 120 and 240:1. Using the lower
value as the supply rate resulted in substantial cost savings over the
commonly-assumed target C:N ratios of between 10:1 and 50:1'
26
20
1C
in
u
o
o
N
O
O 16
S
i
o
I 10
o
III
n.
High contamination
Low contamination
10 20 30 40
TIME (DAYS)
60 60
cr
\
in
in
a
-------
increased from 5 to 20% dry weight of solids. However, it is apparent
from the data that the relative respiration rate (g O2/g solids) decreased
as the solids content increased, suggesting some inhibition resulting
from either less effective mixing and oxygen transfer or from toxicity
of the contaminants.
10 20
TIME (DAYS)
30
Figure 4
Oxygen uptake per gram total organic carbon in three PCP and
creosote-contaminated sludges during LSC treatment.
PCP concentrations increase from sites A through C.
Uptake during simulated land treatment of the
site C sludge is shown for comparison.
nutrients were added until no response was observed. However, the
increased respiration rate was associated with visibly larger average
floe size in the mixed reactor. It therefore appears that mechanical mixing
can impair LSC performance in some cases by promoting agglomera-
tion of oily materials and therefore reducing the surface area available
for dissolution and microbial colonization.
in
5
2
HI
o
h-
80 -
60
40
20
••NP
..-- Mixed
15
30
TIME (DAYS)
45
60
Figure 6
Cumulative oxygen uptake during LSC treatment of ail refinery
sludge in reactors with aeration only or aeration plus
mechanical mixing. Arrows indicate times of supplemental
nutrient (N and P) additions.
NADDED N ADDED NADDED
DAYS
Figure 5
Oxygen uptake during LSC treatment of creosote-contaminated
sludge at three loading rates (5, 10 or 20% dry weight of
solids). Arrows show times of supplemental nitrate additions.
Mixing
One of the more surprising findings has been that high-energy
mechanical mixing does not necessarily improve performance. Typical
mixing energy requirements needed to keep solids in suspension range
from approximately 0.1 to 1 hp/1000 gal depending on the solids con-
tent and the physical characteristics of the solids. Higher mixing ener-
gies are not only more expensive, but they also can be detrimental.
For example, Figure 6 shows data from two reactors containing the same
sludge but operated either with or without mechanical mixing (in
addition to identical mixing through aeration). The faster respiration
in the unmixed LSC reactor was not due to nutrient availability, since
CONTAMINANT REMOVAL
LSCs have been used predominantly in two types of situations:
(1) wood preserving wastes, especially impoundment sludges and the
surrounding soils contaminated with creosote oil and PCP; and (2) oil
refinery wastes, principally sludges from storage and treatment lagoons.
In both these cases, the wastes have high concentrations of oil and grease
and relatively high levels of higher molecular'weight hydrocarbons,
including the potentially carcinogenic PAHs.
This section presents examples of the degree of contaminant destruc-
tion achievable in LSCs used for bioremediation of these wastes.
Wood preserving wastes
LSCs have been used by ReTeC for the remediation of creosote- and
PCP-contaminated soils and sludges in at least 10 full- or pilot-scale
field treatment systems. Significant reductions in both solids mass and
contaminant concentrations have been achieved. Successful cleanup
resulting in closure has been achieved in full-scale remediation efforts.
Table 2 presents typical data on the solids and mass loss of total PAHs
during LSC treatment of sludge contaminated with creosote at two solids
loadings. The results show that very litde of the material is in the aqueous
phase, since the solubility of the compounds is low and degradation
of the compounds in the aqueous phase is rapid. Almost 30% of the
solids were lost in both cases. This loss can reduce the costs for eventual
disposal or further treatment, if either is necessary. Also, the solids
mass loss must be known to calculate the true amount of contaminant
destruction during treatment. The analytical results indicate that
approximately 90% of the total PAHs were degraded over 55 days of
operation.
It is also important to note that there are differences in the degrada-
bility of different compounds. Representative data on the loss of various
PAHs is shown in Figure 7. As typically observed, biodegradation of
the 2- to 4-ring PAHs was much more rapid than degradation of the
more carcinogenic 5- and 6-ring compounds, although all compounds
were degraded to some extent. The differences in degradability reflect
BIOREMEDIATION 335
-------
the lower solubilities and greater inherent resistance to degradative
enzymes of the 5- and 6-ring PAHs* However, the removal rates
observed during LSC treatment were substantially faster than those
typical of land treatment"x".
Table 2
Concentrations and total masses of solids and total PAH.s
before and after 8 weeks of simulated LSC reactor operations
with two samples from a wood-preserving site contaminated
with amounts of creosote oil.
Mtctor MM! PAH Cone. PAH MM! KM* PAH Cono. PAH IUII Hill PAH
(«) (my/kg) (9) (9)
II: 101 Solid!
Solid! «7.« 56.9S3
Liquid
12: 251 Solid!
Solid! 274.« S*,»SJ
Liquid
j|3: 5t gnllda
Solid! 191.J 129,151
Liquid
14; 251 Solid.
Solid! HI. 2 125,251
Liquid
Low ContiKtfwtlon
5.0 74.• «,tO«
1000 «.«
15.6 U7.1 7,3»»
1000 12.4
Rlgb Contuiiwtlon
61.5
1000
100.3
1000
57,201
*.5
0.5) 14.4 <0
g.007
1.41 2t.J »>
0.012
i.«3 10.4 <•
0.00?
97,744 t.12 27.) 4)
9.3 0.010
organic material consisting of relatively degradable compounds. Thus,
it is not surprising that these wastes are excellent candidates for treat-
ment in LSCs. Oxygen uptake measurements show that, on a unit reactor
volume basis, treatment of oil refining wastes produces extremely high
rates of metabolic activity (Fig. 8). The oxygen uptake rates were almost
twice those of creosote sludges, which have somewhat lower TOC
contents and a higher proportion of resistant organic compounds, and
10 times faster than contaminated soils.
300
s
01
CM
O
e
c
o
o
c
250-
200-
150-
100-
50-
Oll Refinery
Sludge*
DAF
Sludge
Weathered
Oil Sludge
Creosote
/ Contaminated Soil
10
20
30
40
Figure 8
Cumulative oxygen upuke (normalized 10 mass of solids)
during bench-scale LSC treatment of a variety of contaminated solids.
a
«
c
^3
«S
X
a
800
6OO -
400 -
200 t.
2&3 Ring
4 Ring
546 Ring
Time (weeks)
Figure 7
PAH concentrations by ring number during LSC treatment
Of creosote-contaminated sandy soil,
LSCs have been used for the successful full-scale treatment of wood
preserving wastes in aerated on-site lagoons. In one representative case,
100 yd3 of impoundment sludge containing PCP and creosote were fed
into an on-site sequenced-batch LSC weekly. Closure criteria were based
on the concentrations of PCP and the combined concentration of the
PAHs phenanthrene and anthracene. These criteria were exceeded during
operations, with an average reduction of PCP concentrations from 2600
to 32 mg/L and an average reduction in the target PAH concentrations
from 1200 to 86 mg/L.
Oil refinery wastes
Oil sludges have extremely high oxygen demands, with much of the
In addition to several laboratory-scale treatability tests of oil sludges
in LSCs, ReTeC has performed several pilot-scale demonstrations of
LSCs for biological treatment of oil refining sludges. Removal rates
for oil and grease have been rapid in comparison to land treatment data,
but results have varied widely, principally because the sludges differ
in the proportions of readily-degradable and recalcitrant hydrocarbons.
Typical land treatment half-lives for similar sludges range from 6 to
IS mo (C**C\2, 13D.--D). Assuming first-order kinetics for our studies,
half-lives for oil and grease generally ranged from 2 to 4 wk for lagoon
sludges and 6 to 14 wk for more recalcitrant stockpiled sludges in sludge
ponds and pits (Fig. 9).
120
100
eo
Z 00
GC
O
O 40
#
20
Sludge ponds
Lagoon sludges
10
TIME (WEEKS)
Figure 9
Oil and grease losses during LSC treatment of oil Sludges from
wasiewater lagoons or from sludge storage pond.
336 BIOREMEDIATION
-------
Losses of PAHs were also relatively rapid, again varying depending
on the nature of the waste and loading rate. In one study, the losses
of carcinogenic PAHs (principally the 5- and 6-ring PAHs) ranged from
30 to 80% over 2 mo, while virtually all of the non-carcinogenic PAHs
were degraded (Fig. 10). The total PAH reductions ranged from 70 to
95 %, again well in excess of typical losses during land treatment of
oil sludges over a similar time period14.
1500
Lagoon Sludge Pit Sludge
Reactor Treatment
Figure 10
Total concentratioins of carcinogenic and noncarcinogenic PAH compounds
before and after 60 days of LSC treatment of oil sludges
at differing solids loading rates.
CONCLUSION
LSC technology represents a method for rapid biological treatment
of contaminated solids in a contained reactor. The technology has proven
highly effective for oil refinery sludges and wood preserving wastes.
LSCs are operated to maximize microbial activities by encouraging rapid
mass transfer and maximum contact between contaminants and micro-
organisms capable of degrading the contaminants. Microbiological
monitoring demonstrates that the technology effectively enhances
microbial numbers and activity and provides the rapid feedback needed
to optimize performance. Pilot- and full-scale applications have shown
that LSCs can provide highly effective on-site bioremediation, with con-
taminant removal rates much greater than those typical of land treatment.
REFERENCES
1. Bartha, R., Biotechnology of petroleum pollutant biodegradation, Microbial
Ecology, 12 1986, 155-172.
2. Loehr, R.C. and Malina, J.F., Land Treatment - A Hazardous Waste Manage-
ment Alternative, Water Resources Symposium Number Thirteen, Center
for Research in Water Resources, Bureau of Engineering, The University
of Texas at Austin, Austin, TX, 1986.
3. Atlas, R.M., Stimulated petroleum biodegradation, Critical Reviews in
Microbiology, 5, 1977, 371-386.
4. McCarty, P.L., Stoichiometry of biological reactions, Progress in Water Tech-
nology, 7, 1975, 157-172.
5. Atlas, R.M. Microbial degradation of petroleum hydrocarbons: an environ-
mental perspective, Microbiology Review, 45, 1981, 180-209.
6. Deutsch, D.J., Waste treatment boosted by bacterial additions, Chemical
Engineer, 86 (9), 1979, 100-102.
7. Alexander, M., Introduction to Soil Microbiology, 2nd ed., John Wiley and
Sons, New York, NY, 1977.
8. American Petroleum Institute, Manual on disposal of refinery wastes, API,
New York, Chap 7, 1980, 1-3.
9. Gibson, D.T., Biodegradation of Aromatic Petroleum Hydrocarbons, in Fate
and Effects of Petroleum Hydrocarbons in Marine Ecosystems and Organisms,
ed. D. Wolfe, Pergamon Press, New York, NY, 36-46.
10. McGinnis, G.D., Borazzani, H., McFarland, L.K., Pope, D.F. and Strobel,
D.A., Characterization and Laboratory Soil Treatability Studies for Creo-
sote and Pentachlorophenol Sludges and Contaminated Soil,
EPA/600/52-88/055, U.S. EPA, Washington, DC, Jan., 1989.
11. Sims, R.C. and Overcash, M.R., Fate of polynuclear aromatic compounds
in soil-plant systems, Residue Reviews, 88, 1983, 1-68.
12. Kincannon, C.B., Oily Waste Disposal by Soil Cultivation Process, U.S. EPA
Publication No. R2-72-110, U.S. EPA, Washington, DC, 1972.
13. Loehr, R.C., Martin, J.H., Neuhauser, E.F., Norton, R.A., and Malecki,
M.R., Land Treatment of an Oily Waste - Degradation, Immobilization, and
Bioaccumulation, U.S. EPA Rept. No. 600 FLQ 2/85, 009, Feb., 1985.
14. American Petroleum Institute, The Land Treatability of Appendix VIII Con-
stituents Present in Petroleum Industry Wastes, API, Washington, DC, 1984.
BIOREMEDIATION 337
-------
Oxygen Sources for In Situ Bioremediation
Richard A Brown
Jill R. Crosbie
Groundwater Technology, Inc.
Chadds Ford, PA
INTRODUCTION
It is well recognized that microorganisms play prominent roles in
the transformation and degradation of organic chemicals. Microbial com-
munities in nature exhibit a truly impressive biochemical versatility
in the number and kinds of synthetic organic compounds that they are
able to metabolize12 Microbial metabolism is virtually the only
natural transformation of organic contaminants that can result in
complete mineralization.
However, there are limits to the metabolic versatility of micro-
organisms. Many substrates, even those that are known to be highly
biodegradable, are often transformed so slowly in nature that they cause
some degradation of environmental quality. This resistance lo
biodegradation is primarily a function of: (a) the existing environmen-
tal conditions; (b) the structure of the particular contaminant; and (c)
the physiology of the requisite microorganisms". Of these, the
environmental limitations are the most common and the most easily
rectified.
In order to grow, microorganisms need a suitable physical and
chemical environment. The nature of the limiting environmental factor(s)
can be classified as general environmental quality or metabolically
dependent. In the first case, extremes of temperature, pH, salinity and
contaminant concentrations will markedly influence the rates of
microbial growth and substrate utilization. In the second case, micro-
organisms must have the basic requirements for growth and metabolism.
Like all other forms of life, microorganisms are primarily composed
of C, H, O, N. P and S, although a variety of other elements are also
found in trace amounts. These substances must already be present or
be supplied in the proper form and ratios for the requisite micro-
organisms to proliferate and degrade organic substrates.
In most cases, the organic pollutants themselves are able to supply
the carbon and energy required to support heterotrophic microbial
growth. However, the introduction of carbonaceous materials to soils
and groundwater aquifers can cause an imbalance in the natural
biodegradation processes, limiting the microbial transformation of the
organic pollutant. For example, when labile carbon is introduced to
an aerobic aquifer, the microorganisms consume oxygen along with the
carbon substrate. An anaerobic aquifer can be expected whenever the
rate of aerobic respiration exceeds the rate of oxygen input lo the site.
To sustain aerobic microbial growth, oxygen, therefore, must be sup-
plied to the subsurface microorganisms.
IMPORTANCE OF OXYGEN
The importance of oxygen supply to in situ biodegradation was well
documented recently in a study of a wood treating site in Conroe,
Texas'. A downgradienl portion of the contaminant plume was
characterized by low levels of organic pollutants and dissolved oxygen,
while inorganic contaminants (i.e.. chloride), which were associated
with the organic wastes, remained al elevated concentrations. The
authors suggested that oxygen was consumed during the aerobic
metabolism of (he organic contaminants by the indigenous
microorganisms. Hydrocarbons persisted in areas of the plume where
oxygen levels were insufficient to support aerobic biological activity.
Artificially increasing the oxygenanon of subsurface environments
will dramatically increase the growth of heterotrophic bacteria. In a
study of petroleum hydrocarbon degradation, sand columns were used
to determine the effect of oxygen supply on bacterial growth and
degradation of gasoline.
Several columns were prepared under identical conditions using 50 ml
of wet sand sieved to 40 lo 60 mesh. A total of SO ml of gasoline was
added to each column and allowed to drain through. An average of
4.3 mL of gasoline was retained in the column. The columns were then
washed with 2 L of nutrients made up in groundwater. Different levels
of oxygen were supplied to the columns by using air, oxygen and/or
hydrogen peroxide dissolved in groundwater. The columns were kept
at design oxygen levels for 2 wk. At the completion of the experiments,
the columns were drained and analyzed for gasoline content, total
organic carbon (TOC), total bacteria and gasoline utilizing bacteria.
Bacterial counts in the interior of the column showed a very strong
dependence on the oxygen level:
Table 1
Dependence Of Bacterial Growth On Available Oxygen
Bacteria. Colony Forming Units (CPU) / gram dry soil
Available Oxygen
(mgl, ar)
8
40
112
200
Correlation w/D.O.
Hcierotrophic Bacteria
(x 10")
0.05
5.5
75
207
0.979
Gasoline Utilizing
Bacteria (x 10*)
0.0001
0.7
27
31
0.933
338 BIOREMEDIATION
-------
As can be seen from the data, the bacterial counts increased
dramatically with increasing available oxygen. Gasoline-utilizing
bacteria are even more sensitive to oxygen levels than are general
heterotrophic bacteria.
The biodegradation of gasoline in the columns also was affected by
the oxygenation:
Table 2
Dependence of Gasoline Degradation on Oxygen Levels
Available
Oxygen
ppm (Ave.)
8
40
112
200
Gasoline Bio-
degraded
grams
.388
.508
.773
1.272
of contamination or the amount of the contaminant within a phase,
measured as either total weight or concentration. The following table
gives a representative phase distribution for a gasoline spill in sand
and gravel:
Table 3
Phase Distribution of Gasoline in Sand and Gravel
Extent of Mass
Contaminations'! Distribution
impacted Sediments of Hydrocarbons
Phase
Volume,
cu. yd.
Cone.
pom
Free phase1 780
Adsorbed (soil) 2,670
Dissolved (water) 11,120
5.3 26,800'
18.3 11,500
76.3 390
* of
Total
69.3
29.7
1.0
Corr. w D.o.
1 Actual value recovered fron site
Two conclusions ban be drawn from these data. First, the more oxygen
that was supplied, the more gasoline was biodegraded. Second, the rate
of biodegradation under highly oxygenated conditions was greater than
the rate of physical removal/dissolution.
These sand column studies demonstrate that bacterial growth and
metabolism are very dependent on oxygenation. As a result, an impor-
tant part of the biological treatment of organic contaminants is oxygen
supply.
METHODS OF OXYGEN SUPPLY
There are basically two approaches to oxygen supply—physical and
chemical. Physical oxygen supply involves forcing air and/or pure
oxygen into the contaminated matrix. Chemical oxygen supply involves
the addition of substances which can be converted to oxygen (such as
hydrogen peroxide)6 or substances which can act as terminal electron
acceptors directly (such as nitrate)7-8. All of these methods have been
used to treat contaminated soils and aquifers. This paper will review
five methods of oxygen supply: air sparging; injection of aerated/oxy-
genated water; venting; injection of hydrogen peroxide; and injection
of nitrate.
The choice of an oxygenation method depends on several factors.
Basically, one wants to achieve maximum efficiency in oxygenation.
The principle is to balance oxygen supply with oxygen demand. The
factors that must be considered in achieving this balance of supply and
demand are:
• Contaminant load and location
• Oxygen mass transfer, (Ib per unit time) supplied by each method
• Ease of transport/utilization
CONTAMINANT LOAD
The first factor to consider in choosing an oxygen source is the con-
taminant load and location. Contaminant location is important because
vent systems require unsaturated environments and will, therefore, be
excluded in treating contaminants below the water table. Contaminant
load, on the other hand, impacts all means of oxygen supply in that
it determines oxygen demand. What drives contaminant load is the phase
distribution.
Petroleum hydrocarbons exist in the subsurface as three condensed
phases: mobile free product (phase separated); residually saturated soil
(adsorbed phase); and contaminated groundwater (dissolved phase).
The distribution of hydrocarbons into these different phases, while a
result of dynamic transport, is ultimately a function of their physical
and chemical properties and the hydrogeological and geochemical
characteristics of the formation. One must examine the phase distri-
bution by two means: (1) by the areal extent of contamination or the
volume of the subsurface impacted by a phase and (2) by the severity
There are several generalizations that can be made from the above
data concerning the distribution of petroleum hydrocarbons between
the different phases. First, groundwater flow is the primary long-term
mechanism for dispersion of the contamination once the free phase
product layer has achieved flow equilibrium. Thus, the areal extent of
dissolved phase hydrocarbon contamination is typically greater than
that for other phases. However, the amount of material in the ground-
water is small compared to that retained in the soil matrix less than
5%. The residually saturated soil (i.e., adsorbed phase), if untreated,
is a continuing source of groundwater contamination.
In looking at the contaminant load, the presence of and the distribu-
tion between the different phases is an important factor. Table 4 gives
the amounts of contaminants in lb/yd3 of water for the dissolved phase
and lb/yd3 of soil for the adsorbed phase contamination. The calcula-
tion assumes a dry soil bulk density of 2700 lb/yd3. From these data,
it is obvious that contaminated soil drives the contaminant load. One
cubic yard of soil contaminated at only 100 ppm contains as much con-
taminant as 45 yd3 of contaminated water (dissolved phase)
contaminated at 100 mg/L. Thus, knowing whether or not there is a
high contaminant load, adsorbed phase or a low contaminant load,
primarily dissolved phase, is important in choosing an oxygenation
method.
Table 4
Comparison of Contaminant Loading Dissolved
and Adsorbed Phases
Dissolved
1 ppm
10 ppm
100 ppm
Water
fib/yd3)
6xlO"5
6x10"*
6xlO"3
100 ppm
1,000 ppm
10,000 ppm
Soil
(Ib/yd3)
.27
2.70
27.00
OXYGEN MASS TRANSFER
The second factor to consider in choosing an oxygen supply method
is oxygen mass transfer. It is easy to calculate the amount of oxygen
supplied by the different methods. The more oxygen supplied per unit
time, the greater the contaminant load that can be treated.
Air Sparging
Air sparging, one of the simpler techniques of oxygen supply, pro-
vides oxygen by diffusing air/oxygen into a well bore. This supply pro-
cess is accomplished by supplying air (or oxygen) to a porous stone,
BIOREMEDIATION 339
-------
sintured metal or fitted glass diffuser beneath the water surface. The
water in the well bore is saturated with oxygen and diffuses out into
the formation. The amount of oxygen supplied is a function, therefore,
of the rate of water flow by the well bore. The rate of water flow, in
turn, is a function of the hydraulic conductivity, the groundwater gradient
and the surface area of the formation affected by the well bore. The
matrix in Table 5 shows the amount of oxygen an air sparger provides
per well per day for different hydraulic conductivities and gradients.
The table assumes a 30-ft saturated thickness and that the lateral
influence of the well is 3 ft.
Table 5
Oxygen Supplied
By Sparging, Single Well lib/day)
Hydraulic Gradient
(ft/ft)
(high)
0.1
(medium)
0,01
(low)
0.001
(air) (oxygen) (air) (oxygen) (air) (oxygen)
6
0.06
6x10 '
JO 06
0.3 6*10 '
3x10" 6x10 *
03. 0.06
3x10 2 6xK) -'
3xKTs 6x10 •'
0.3
3x10 '
Hydraulic Conductivity
gals / day / f
I04 (gravel)
Vf (medium sand)
K)-' (silt)
As can be seen, air sparging supplies a limited source of oxygen.
Sparging pure oxygen instead of air will increase the amount of disolved
oxygen supplied by a factor of five so that the maximum oxygen delivered
would be 30 Ib. oxygen per day instead of 6 Ib/day.
Saturated Hater
A second system is to pump air/oxygen saturated water into a con-
taminated aquifer. The amount of oxygen supplied is a function of
injection rate (Table 6).
Table 6
Oxygen Supplied
By Aerated/Oxygenated Water Injection
Single Well (Ib/day)
Oxygen Supplied (Ib/day)
aerate
Injection Rate, (gpm)
I
10
too
Air Vent Systems
Air vent systems are an efficient means of supplying oxygen through
unsatu rated contaminated soils. This technique is used to treat vadose
zone contamination or to treat excavated soil piles. Air can be added
by either injection or by withdrawal. In vadose zone treatment, the
common method is vacuum withdrawal. This method has the added
advantage of physically removing volatile contaminants in addition to
supplying oxygen. The amount of oxygen supplied is a simple function
of the air flow rates. The following table uses a 20% oxygen content
for air to calculate air supply
Table 7
Oxygen Supplied
By Venting System (Unsaturated Soils)
Single Well
(scfm) (Ibs/day)
aerated water
(H) mg/L D.O.
0.12
1.2
120
oxygenated water
(50 mg/L D.O.
0.60
6.0
60.0
I
5
10
20
50
100
23
117
233
467
1170
2330
Chemical Supply
Finally, there are two chemical carrier systems - hydrogen peroxide
and nitrate. While both of these materials are highly soluble, their
common use rate is about 1000 mg/L. The number of oxygen equivalents
supplied is dependent on the chemistry involved. Hydrogen peroxide
is converted through decomposition to oxygen:
H;0} ----- Hp + 1/2 02
Each pound of hydrogen peroxide supplies 0.47 Ib of oxygen. Nitrate
is. on the other hand, directly utilized as a terminal election acceptor.
Its oxygen equivalents can be calculated by comparing the amount of
nitrate required to oxidize a substrate versus the amount of oxygen.
Take, for example, the oxidation of methanol:
Oxygen: CH.OH + 3/2 O, - -» CO, + 2H,O
Nitrate: NO, + 1.08 CH,OH + H" -•» QMS C,H,NO, + 0.47 N2
+ 0.76 CO, + 2.44 HjO
Based on the above equations, I Ib of nitrate is equivalent to 0.84 Ib
oxygen.
The oxygen equivalents supplied by these two chemical carriers are
a simple function of injection rate.
TaMe8
Oxygen Equivalents Supplied
By Chemical Carrier*. Single Wrll
(at NOOmg/L)
Oxygen Supplied (Ib/day)
Injection rate
H,O,
NO.
(047 equiv 0, pan H.Oj (084 equiv 0,/pan NO,)
IjO 6 X)
50 28 50
10.0 56 KX)
200 112 200
50 280 500
EASE OF TRANSPORT
The third factor in considering an oxygen source is the ease of
transport and utilization. This factor involves the mode of application.
the maintenance of the system and the rate and/or degree of utilization.
An air sparger system uses a small compressor able to deliver 1 cfm
per well. The sparger itself is either a porous stone, a sintured metal
diffuser or a fritted glass diffuser. Power consumption is minimal. The
transport of the aerated water is limited by the rate of groundwater flow.
The most significant operating cost in an air sparger system is
maintenance of the compressor and of the diffuser and well screen.
Biofouling or inorganic fouling of the diffuser and well screen can be
significant and therefore require a high degree of maintenance. Bacterial
utilization of the dissolved oxygen is very high.
Injection of aerated/oxygenated water is a relatively simple system.
The simplest approach is to use an air stripper absorber to aerate the
water. Often in treating a contaminated aquifer, groundwater is recovered
and air-stripped to achieve hydraulic control of the contaminant plume.
Reinjection of the stripped groundwater can therefore be accomplish-
ed at a relatively low cost. The main cost of operation is controlling
fouling of the injection system. Transport of the oxygenated water is
dependent on the geology (hydraulic conductivity). Bacterial utiliza-
tion of the injected dissolved oxygen is very good.
Venting systems, while limited to unsaturated soils, are very efficient
means of oxygen supply. The primary capital cost is the vacuum pump(s)
needed to drive the system. Maintenance of the pumps is fairly simple
and power consumption is minimal. The efficiency of the vent system
is enhanced by volatile chemical removal from the soil. The largest
potential cost with a vent system is treatment of the vapor discharge
which can be accomplished by using disposable carbon, regenerate
carbon or catalytic oxidation.
340 BIOREMEDIATION
-------
A hydrogen peroxide system is generally a low capital cost, easy to
maintain system. The use of hydrogen peroxide does have a fairly high
operating cost due to the cost of the purchased hydrogen peroxide which
is dependent on the volume used. On a per pound of oxygen basis, the
cost will range from $1.50 to $2.50. The greatest cost factor involved
with hydrogen peroxide is how quickly it decomposes. There are two
mechanisms of decomposition - biological and metal catalysis. Ideally,
one would like minimal metal catalyzed decomposition. However, in
some soils containing high levels of iron or manganese, metal catalyzed
decomposition can be severe. In such cases, the solubility limit of oxygen
in the water is rapidly exceeded and the water phase degased, losing
available oxygen and drastically reducing the efficiency of the system.
Finally, nitrate systems are a potential electron acceptor alternative.
Operationally, these systems have not been proven. Capital costs for
a nitrate system would be fairly low, consisting of a supply tank and
metering pump (similar to hydrogen peroxide). Chemical costs for nitrate
are $0.60 to 0.70/lb oxygen equivalent. The issue with nitrate, however,
is neither the cost nor the ease of addition, but instead the biochemistry
of utilization and the regulatory issues. In a recent test of nitrate utiliza-
tion, it was found that even with an extremely labile substrate such as
sucrose, there was a significant lag phase in the utilization of the nitrate
when oxygen was also available at low levels. It would appear that nitrate
utilization requires low oxygen requirements. If the biochemistry of
nitrate is complicated, the regulatory issues become significant. Nitrate
levels in groundwater are regulated at 10 mg/L. If nitrate is not rapidly
utilized, injection would have to be tightly controlled and may be
precluded or the nitrate would have to be removed.
COST ANALYSIS
To put the above analyses into perspective, the costs and effectiveness
for the different oxygenation systems will be compared for a high degree
of contamination (significant adsorbed and dissolved phase) and for
a low degree of contamination (primarily disolved phase only).
The analysis for the high degree of contamination assumes an area
of contamination of approximately 250 x 100 ft with a loss of approxi-
mately 500 gal of a petroleum hydrocarbon fuel in a permeable sand.
The example also assumes that the majority of the contaminant is
adsorbed phase and is at, or above, the water table. Based on these
assumptions, Tables 9 to 11 were constructed to compare the various
oxygen systems.
Table 9
Operating Cost Comparison High Degree of Contamination
Bvatea
Air Sparging
Hater Injection
Venting (vpr Ctrl) $88,500
Capital
$35,000
$77,000
Operation
$800/mnth
$1200/mnth
$1500/mnth
Hydrogen Peroxide $60,000 SlOOOO/mnth
Nitrate Injection $120,000 $6500/mnth
Treatment Total
Maintenance Time Cost
1716 d $150k
1580 d 5194k
132 d $101k
330 d $187k
335 d $210k
$1200/mnth
$1000/mnth
$1000/mnth
$1500/mnth
$looo/mnth
Several things should be noted in this table. First, the nitrate capital
costs are high because of a projected need for tight off-site control of
nitrate due to groundwater regulation of nitrate levels. Second, the vent
system includes a vapor phase control system - a catalytic oxidizer which
costs approximately $60,000. If vapor phase controls are not necessary,
then the capital and total cost would be reduced significantly for the
vent system.
Table 9 gives the gross operating and capital costs for the different
oxygen systems. It does not, however, take into account the effectiveness
of treatment. The different systems will not equally treat all phases of
contamination. For example a vent system is ineffective in treating
contaminated groundwater and in treating adsorbed phase contamina-
tion below the water table unless the water table drops naturally or is
artificially lowered. An air sparging system is ineffective in treating
vadose zone contamination unless the water table rises. The following
table takes into account these factors and other efficiency factors and
estimates a cost-effectiveness for the different systems.
Table 10
Cost-Effectiveness Comparison
High Degree of Contamination
Flow
Site
System
Air Sparging
Water Inject
Venting
Peroxide
Nitrate
Rate Oxygen
(lb/day)
15 wells 6
@2cfm
70 gpm 8
160 cfm 4000
70 gpm 190
70 gpm 211
(120 gpm recovery)
Treat*
(%}
41
85
72
95
85
System Treatment Contaminant
Utilization Treatment
Efficiency Time Cost
<%) (days) ($/lb)
1 (sparg)
70 (0.0.)
50
5
15
13
1716
1580
132
330
335
90.3
100.2
13.4
65.1
77.2
As can be seen from Table 10, there is a wide variance in both cost-
effectiveness and in treatment-effectiveness. In terms of cost per-
formance, the order is:
venting * •» peroxide * nitrate •» air sparger * water injection
In terms of treatment effectiveness, the order is:
peroxide * nitrate = water injection •» venting * •» air sparging
While venting is a very cost-effective method, it is limited to treating
the vadose zone. Consequently, its treatment-effectiveness is limited.
The above analysis is given for a situation with extensive contamina-
tion. If the degree of contaminantion is changed so that the soil con-
tamination is minimal, the analyses would change. Assuming that there
is no soil contamination above the water table and that the soil levels
are < 100 ppm, the performance of the different systems would be as
follows assuming all other factors, such as capital, operating and
maintenance costs, etc., remain constant
Table 11
Cost/Performance
Low Degree of Contamination
(Dissolved Phase Only)
System
Oxygen Delivered
°2
(lb/day)
Time of
(days)
Cost of
Treated
Air Sparging
Water Injection
Venting
Peroxide
Nitrate
6
8
Not Applicable
190
211
180
330
_..
180
240
117
314
134
166
Where the degree of contamination is less, simpler systems such as
air sparging become more cost-effective. Where the contamination is
only in the dissolved phase, an air sparger system is often the best choice.
The choice of an oxygen supply is dependent on the contaminant load,
the mass transfer and the ease of transport/utilization. Depending on
the degree of contamination, different systems will be most effective.
CASE HISTORIES
To examine the performance of different oxygenation systems, three
case histories will be discussed. All three case histories deal with
gasoline contamination. In the first case history, the oxygenation system
was an air sparger network. In the second case history, hydrogen
peroxide was used. In the third case history, a vent system was used.
Each case history will discus the degree of contamination, the installa-
tion and operation of the oxygenation system and the results attained.
Case History 1: Air Sparging Network
In this case history, the contamination problem occurred when an
undetermined amount of gasoline leaked from a below ground storage
tank. The area of the loss is underlain by approximately 6 to 7 ft of
BIOREMEDIATION 341
-------
red-brown, heavy silt loam which, in turn, is underlain by a fractured
red-brown shale and siltstone. Depth to groundwatcr is 20 to 25 ft below
grade within the bedrock system. Impact from the loss included a
dissolved phase hydrocarbon plume that extended approximately 250
ft in a north-south direction and 350 ft in an east-west direction with
concentrations ranging form 10 mg/L to 15 mg/L for gasoline-type
hydrocarbons. Ten domestic water supply wells were impacted in
addition to organic vapors within nearby residential basements, Free-
floating phase hydrocarbons were absent.
The remedial system designed and implemented at this site included
contamination plume and water table manipulation via pumping.
dissolved organic removal of the pumped water by air stripping, and
accelerated in situ bioremediation of adsorbed and dissolved phases
by the physical addition of oxygen and nutrients (Fig. 1). The physical
addition of these components to the original loss area was accomplished
through the re-infiltration of treated oxygen and nutrient-rich ground-
water into an infiltration gallery located in the former tank pit. An air
sparging system, consisting of mechanical air compressors, air lines
and down well diffusers, provided needed oxygen to peripheral areas
of the plume outside the infiltration gallery.
DISSOLVED HYDROCARBON CONTAMINATION
, . SERVICE
,« STATION
JL
RECOVERY WELL
OBSERVATION WELL
DOMESTIC WELL
MR SPAR6INC WELL
AIR COMPRESSOR
INFILTRATION «AU-ERT
AIR STRIPPING TOWER
AREA OF CONTAMINATION
— AIR.UNE
WATER OtSCHAACE UNE
Figure I
Schematic of Bioreclamalion System
The air sparging system effectively to partially effectively delivered
needed oxygen to the peripheral areas. Major limitations focused on
the maximum quantity of oxygen that could be induced into the ground-
water system (K) mg/L) at the sparging point and the fouling and plugging
of the sparging points by the development of thick biologic growths.
These conditions interfered with optimum oxygen transfer to the frac-
tured bedrock system and required frequent mechanical cleaning.
The first 11 mo of operations showed a general 50 to 85% reduction
in organic contaminants, despite the non-optimum conditions of the air
sparging system (Fig. 2). At 85% reduction, the treatment stabilized
indicating that the air sparging was a limited system. The residual
contamination was adsorbed phase trapped in the fractures of the
bedrock system.
A comprehensive program to accelerate oxygen transfer rates was
subsequently incorporated at the site in an effort to reduce the project
restoration time-frame. The program involved the delivery of increased
quantities of oxygen to the groundwater system via the trickle feed and
disassociation of dilute quantities of hydrogen peroxide (WO mg/L). This
case history demonstrates that air sparging is ineffective in treating
adsorbed phase hydrocarbons. It also indicates that maintenance of the
air sparging system is significant and therefore is not a desirable
application for long-term programs.
lift n ua. mn
• t I <
ill I
urr o
Figure 2
Total Hydrocarbon Concentrations for Air Stripping Tower
Case History 2: Hydrogen Ptroxid*
This case study involves petroleum leakage over a period of time from
underground storage tanks, pumps and lines at a service station. The
area in which the loss occurred has complex geology with a varied
hydrocarbon phase distribution within the water table in both the 7-
to 10-ft thick variable fill overburden and in the underlying fractured
limestone bedrock. The subsurface hydrocarbon contamination was not
only limited to the property on which (he loss occurs, but also migrated
with natural groundwater flow across a busy intersection to a commer-
cial building (Fig. 3).
Three general areas of interest were addressed by the remedial
program (Fig. 3): Area A was a lightly contaminated area with most
of the subsurface hydrocarbons being found in the fill material; Area B
was the location of the underground storage tanks and included signifi-
cant contamination in both the fill and the bedrock; and Area C was
the commercial building basement which had been impacted by phase-
separated hydrocarbons.
The general remediation program designed for this site involved the
in situ bioremediation of impacted groundwater and sediments through
the addition of aerated water supplied with nutrients and hydrogen
peroxide. Remedial response was strongly correlated to the product
distribution and the geology. A 95% reduction in dissolved hydrocarbon
levels was achieved in 5 mo of operation in Area A (Fig. 4). Response
in Area B was the least dramatic and most variable, with the greatest
reduction of dissolved levels (40 to 50%) achieved in a few months
representing treatment of the adsorbed phase in the more permeable
fill. Following this initial response, the remedial response slowed
representing treatment of contamination in the bedrock. Area C achieved
an 85 % reduction in hydrocarbon contamination in approximately 6 mo.
Hydrogen peroxide was added to the groundwater system at both the
infiltration gallery and former air sparging wells. The most recent results
show overall hydrocarbon concentration levels to have declined in the
core area with only five of the original 10 home owner wells still
contaminated.
Case History 3: Soil Venting
In this case study, a pre-closure site investigation of a former service
station facility in Massachusetts revealed low levels of both dissolved
342 BIOREMEDIATION
-------
f f
&
/
7
J
_
4
Cs
__l -^ •*»
=^\
^
MttHM
CNCLQMM*"***^
Win' X^
\ _
uw-« ^r
^,9 u
IXHAIHT nou
IAKMHT VCHTK.ATKM
MW-8
lUMVLflaM. BVIUlmv
AREA-C"
ew-4
a
MW-IZ
• cw-s
9
SITE MAP
PROJECT- ULTRAMAR CANADA INC.
DRAWING NO.' 911-ISO -8167-3
MONITORING DATE'
A RECOVERY WELL
• MOtdTORIMG WELL
LI ~1 UNDERGROUND TANKS
WELL IDCHTIFKATIOKi
CX.iUV-1-WILL ID.
- «AT« TULC CLCV.dNKOt
(B CORE WELL
O^^^JO
SCMJE IN FTET
UTI
lfl>-4« ARW-I
0MW-II
Mw-ia
•
•
'»KL20,«7
^
AREA "B
iMTcuctrrwi TOCHCH
(IIIXHMH UUJRY HOJ)
Figure 3
Site Map
and adsorbed phase gasoline contamination associated with the
underground storage tanks. The highest concentration of adsorbed phase
hydrocarbon contamination was found to be present at a depth of 9 to
11 ft in soils that are in a zone of seasonal groundwater fluctuation.
A soil venting system was installed at the site to address the adsorbed
phase hydrocarbons above the water table. The system designed for the
site included six soil vapor extraction points, two of which were placed
within the former underground storage tank locations. The points were
Table 12
Air Sample Result Sheet
CARBON ETHYL
OXYGEN DIOXIDE METHANE BENZENE TOLUENE BENZENE XYLENE TPH WATER
DATE ppmppnppmDpmppmppnppmppnppm
02/07/89 210000 11000 740 1.20 0.44 0.22 2.00 560 15000
02/08/89 210000 10000 750 0.00 0.52 0.22 1.80 88 15000
02/09/89 210000 10000 420 0.00 0.00 0.00 0.00 0 15000
02/13/89 210000 2900
02/14/89 210000 2800
02/27/89 210000 1700
03/09/89 210000 1300
04/12/89 210000 1300
04/25/89 210000 1400
55 0.00 0.00 0.00 0,00
230 1.40 0.00 0.00 0.68
170
100
0
0
0.36
0.36
0.00
0.00
2.50 4.20
0.29 0.00
0.00 0.00
0.00 0.00
3.90
0.00
0.00
0.00
51 15000
240 15000
130 15000
190 15000
46 15000
21 15000
4 6 B 10
MONTHS OF TREATMENT
12
14
46 8 10 12 14
MONTH3 OF TREATMENT
Figure 4
Bioreclamation Results Area "A" (MW-1,2,13,14)
constructed of 2-in. diameter schedule 40 PVC with 0.020-in. slot well
screen extending over the contaminated zone and placed using a 30-ft
radius of influence. Granular activated carbon for adsorption of volatile
organic carbons from the soil vapor was utilized for the soil vent blower
effluent.
Effluent air sample results of the system's operation over an approxi-
mately 1.5-mo period during the dry late winter and early spring are
presented in Tables 12 and 13. Up to 122 scfm of soil vapor-were drawn
through the system during the period. Figure 5 shows CO2 production
rates basically parallel vapor phase concentration of methane and total
petroleum hydrocarbons in the effluent soil vapor. This observation
Table 13
Air Sample Result Sheet
HYDROCARBONS REMOVAL RATE
DATE
02/07/89
02/08/89
02/09/89
02/13/89
02/14/89
02/27/89
03/09/89
04/12/89
VAPOR
prop
0.015
0.015
0.015
0.015
0.015
0.015
0.015
0.015
UEICHT
It*, per
Ib.pole
28.88
28.84
28.84
28.74
28.74
28.68
28.76
28.75
HEAD
In. water
0.15
0.11
0.15
0.15
0.15
0.10
0.10
0.17
TEMP
F
55
55
55
55
55
55
55
55
PRESS
In. mere
30.02
29.99
30.05
30.52
30.20
29.67
30.44
30.37
PRESS
in. water
0.00
0.00
0.10
0.00
0.00
0.00
0.00
o.oa
VELOC
ft/sec
21.4
18.4
21.4
21.3
21.4
17.7
17.4
22.7
RATE
cubic
ft/mln
113.8
97.4
113.8
115.0
114.4
92.7
93.7
122.1
Ibc per
day
a. 2
7.0
3.1
3.1
3.1
2.5
4.6
1.5
Ibs to
date
8.2
15.2
18.5
30.7
33. B
76.6
103.7
229.8
approx
equiv
gallons
1.3
2.4
2.9
4.9
5.4
12.2
16.5
36.5
AIL samples taken from effluent of soil vent blower
All samples taken Irom effluent of soil vent blower
BIOREMEDIATION 343
-------
TOTAL CARBON
INEFFLUENT(PPM)
veLOCITV MEAD
INCHES IN WATER
Figure 5
Gas Production Kale
demonstrates that significant biodegradation occurs even with highly
volatile compounds.
Normal CCX content of air is OXM%; the 11% CO2 observed initially
represents a 275-fold increase in CO2. This response demonstrates that
bacteria can readily use oxygen provided by a vent system.
REFERENCES
I. Alexander, M., "Biodegradation of Chemical!, of Environmental Concern."
SCIENCE. 211 pp. 32-138, 198!
2 Kobayaski. H. and Riltmann. B.E., "Microbial Removal of Hazardous Organic
Compounds." Environ Sci. Tech. 16, pp I70a-I83a, 1982.
3. Alexander, M . "Biodegradation Problems of Molecular Recalcitrance and
Microbial Fallibility." Adv. Appl Microbiol. 7 pp 35-80. 1965
4. Alexander, M., "Nonbiodcgradable and Other Recalcitrant Molecules."
Biotech. Biocng . 15. pp 611-647. 1973
5. Wilson, J T . McNabb, J F.Cochran, J.W . Wang, T.H., Tomson. MB. and
Bcdicnt, P.B., "Influence of Microbial Adaption on the Fate of Organic
Pollutants in Ground Water." Env. Toxicol. Chan. 4. pp 743-750. 1985.
6 Brown, R.A.. Norris, R.D. and Raymond, R.L. "Oxygen Tranvport in Con-
taminated Aquifers." Proc of the NWWA/Afl Confon Pnrokum Hydrocarbons
and Organic Chemicals in Ground tMuer-Preveniion. Detection and Restora-
tion Houston. TX. Nov.. 1984
7. Leprincc. Y Richard, "Use of Biotechnics in Water Treatment Feasibility
and Performance of Biological Treatment of Nilraics." Aqua Sci. Tech. Ret ,
pp 455-62. 1982
8. Andreoli. R. Reynolds, M. Banilucci. R Forgione. "Nitrogen Removal in
a Subsurface Disposal System." Hbier Sci Tech.. 13 (2). pp 967-76. 1981.
344 BIOREMED1ATION*
-------
Remedial Options and System Characteristics
of an Inactive Land Treatment Facility
Timothy R. Marshall, Ph.D.
Woodward-Clyde Consultants
Los Angeles, California
Joseph S. Devinny
Robert L. Islander
Civil and Environmental Engineering
University of Southern California
Los Angeles, California
ABSTRACT
Regulatory restrictions on land disposal of certain wastes have
prompted closure of land treatment systems. Remaining soils are often
highly contaminated. This paper compares some system characteristics
measured in soils at an intensely-loaded petroleum waste treatment
system when it was operating and when loading ceased. Results indi-
cate that microbial activity decreases rapidly as usable substrate be-
comes limiting, leaving high concentrations of inassimilable waste in
the soil. Remedial options are discussed for soils of former land treat-
ment systems.
INTRODUCTION
The technology of hazardous waste land treatment is evolving rapidly
in the face of regulatory restrictions and concerns about long-term lia-
bility and environmental damage. Systems presently are designed with
double-walled liners to protect against the migration of hazardous
leachate, atmospheric emissions collection and air delivery systems,
providing oxygen to maintain aerobic soil conditions. The treatment
pad and associated appurtenances are constructed at the remediation
isite. When the project is completed, the system is disassembled. These
systems are more carefully engineered than old landfarms, which
essentially took advantage of the natural soil assimilative capacity for
organics. The kinetics of waste degradation and degree of success varied
tremendously among different systems. The differences arose from geo-
graphic location and from management practices. These systems usually
were operated with minimal environmental controls and containment.
Most systems were designed to control rainwater run-on and potentially
contaminated run-off. The U.S. EPA mandated a monitoring program
for land treatment facilities that included installation of lysimeters for
soil pore liquid monitoring, groundwater monitoring wells and soil core
monitoring'.
Given the number and history of use of industrial waste land treat-
ment systems few reported episodes of environmental con-tamination
have occurred. Streebin, et al.2 have documented that metals are
generally immobilized in the top 25 cm of soil at land treatment sites.
They found, however, trace quantities of poly nuclear aromatics migrating
into the unsaturated zone and high levels of TOC and COD in soil pore
water. The American Petroleum Institute3 reports that poly-nuclear
aromatic hydrocarbons have sufficiently high organic carbon partition
coefficients and, as such, will be strongly adsorbed in soil and immobi-
lized in land treatment systems.
Remedial use of land treatment created many systems which are now
being closed. Other factors also are responsible. The 1984 Hazardous
and Solid Waste Amendments to the RCRA mandate a land ban on dis-
posal of specific waste streams. Two-thirds of listed wastes are now
regulated; the remainder are scheduled for regulation by May 8, 1990.
For some areas of the country such as southern California, tightening
air quality restrictions and land use pressures have contributed to the
closures.
Soils of treatment systems require remediation or disposal when waste
incorporation ceases. A moderate, continuous waste application rate
is needed for maximum stabilized performance of land treatment
systems4. When applications cease, micro-biological activity declines.
The petroleum industry historically has operated land treatment
systems for selected waste streams economically at moderate loading
rates. Petroleum waste materials are generally easily biodegraded by
an acclimated consortia of microorganisms. This paper reports a 3-year
monitoring study of an intensely-loaded petroleum waste land treatment
facility. System performance was evaluated by monitoring operational
variables such as waste loading rate and frequency, soil physical varia-
bles and microbiological parameters such as carbon dioxide evolution
and microbial population density. Additionally, aspects of the struc-
ture and function of the microbial ecosystem were inferred from the
monitoring results. Waste additions were more or less continuous until
November, 1987, when operations ceased completely.
METHODS AND MATERIALS
Representative soil samples were obtained from the land treatment
facility in pre-sterilized sampling jars. Samples used for analysis were
selected from a composite of soils from aparticular section. This sam-
ple compositing was done to reduce spatial variations introduced from
waste loading practices, a significant source of variability even in systems
that practice uniform waste application and incorporation4.
Total Viable Count
Soil samples were mixed with sterilized, distilled water and gently
swirled and sonicated under low power to break up the oily agglo-
merates. Growth media compositions for enumeration of microorgan-
isms were chosen on the basis of ability to select for
petroleum-degraders5. Brain-heart infusion agar was chosen as a base
to which amendments were added, described elsewhere5. Enumeration
of bacteria, fungi and actinomycetes was accomplished by plating 10-mL
drops of successively diluted suspensions on hardened agar. The proce-
dure, developed by Harris and Sommers6, is a modified most proba-
ble number determination. The number of organisms determined is a
function of the entire dilution series rather than the most dilute member
of the series.
Respiration
Degradative activity of the microorganisms was measured as carbon
dioxide evolution in 250-mL biometer flasks (Bellco Biotechnology).
Carbon dioxide is absorbed in alkali and is analyzed titrimetrically
BIOREMEDIATION 345
-------
following precipitation with BaClr Details of the procedure can be
found elsewhere3
RESULTS AND DISCUSSION
The land treatment system received no waste loading after Novem-
ber, 1987. The fortuitous closure allowed comparison of measurements
of system parameters during continuous waste loading and after loading
ceased.
Microbial Numbers and Activity
Active treatment system soils harbor a diverse assemblage of micro-
organisms. Figure 1 shows seasonal fluctuations of microbial numbers
for one section. Millions of organisms per gram of soil are recorded
from all seasons with peak numbers occurring in the hot summer
months. Bacteria, fungi, yeast and actinomycete groups were cultura-
ble from the treatment soils. Statistical analysis of the ecological rela-
tionships of the microbial groups have been presented'
12
c '0
J
s
° a
47 96 152 201 251 306 370 447 505 563 636 699 H3 840 899
Ooys of teof from Februory 1987 lo July 1989
Figure 1
Total Viable Count Measured in Landfarm Section A from
February 1987 to July 1989
Substantial densities of microorganisms are present in the treatment
soils of the closed facility. However, fewer kinds of colonies were
observed. The lower diversity of microorganisms suggests more
restricted resource partitioning. Organisms currently inhabiting the treat-
ment soils are viable, but presumably substrate-limited.
Respiratory activity, as measured by microbial CO, evolution, is
influenced by time of year, general environmental conditions and the
presence of a continuous supply of usable substrate. Figure 2 shows
respiration values in mg CO2/g soil/day for one section of the land-
farm from February, 1987 to August, 1989. The values indicate high
levels of microbial activity and substrate decomposition until about
July, 1988. Proper environmental, biochemical and physical conditions
were maintained in this period. Loading ceased in November, 1987.
In late summer of that year, approximately 400 tons of waste were
applied to each section. This substantial waste loading enabled degrada-
tive activity to continue for approximately 6 mo. The rate of degrada-
tion began to decline in late summer of 1988 and eventually leveled
off at low, constant activity.
Degradation Kinetics
Steady-state concentrations of 15 to 30% oil in soil by weight is an
assimilable loading when environmental conditions are not stressful and
a reasonable portion of the waste is usable substrate. Martin et al,'
showed that for representative land treatment units, activity in the system
was a function of the waste half-life and the weight percentage of oil
added with each application. Larger molecular-weight, more recal-
citrant, less soluble organics will biodegrade slowly or cometabolically.
Polynuclear aromatics, heteronuclear species or poly-substituted
molecules will degrade by different pathways than simple hydrocarbons.
16
1.4
I ,2
a
| ,
fo,8
I 06
v
a
0,4
0?
\r
'.'•*ir- o' '"0* Irorn f*t>fuory '987 fco July I9OT
Unit* O* We*p«roU*n
"*j CO2/Q
Figure 2
Microbial Respiration Rates Measured in Landfarm Section A from
February 1987 to July 1989
The degradation kinetics of these groups may differ from the overall
performance of the system.
Cometabolic substrate decomposition of a particular group may be
first-order (or some other order) with respect to the energy-yielding
substrate. Diauxic relationships, whereby one substrate is preferred over
another until that substrate becomes limiting, are also likely. Thus, while
overall system performance may be driven by an excess of usable sub-
strate, degradation kinetics for specific groups are likely more com-
plex. System performance is best described by a multiple-substrate
kinetic model. However, rate coefficients describing breakdown of in-
dividual waste components are presently unknown. The rapid break-
down of some of these groups is contingent on the supply of readily
assimilable substrate for maintenance of a stable microbial communi-
ty. This is the driving force for total system respiration.
Continuous applications of differing quantities and quality of
petroleum waste makes predictions of waste degradation kinetics difficult
in functioning systems. Martin et a/.' demonstrated the dependence of
the predicted stabilized weight percentage of oil in soil on waste half-
life. For a facility located in a moderate climate that can operate with
a waste application frequency of weekly for 52 weeks per year (0.33%
oil in soil per application), the predicted weight percentages correspond
with the following half-lives: t,, = 60 days, 5.7%. t., 1/2 = 125 days,
12.2%,t, = 146 days, M.2%,t, = 304 days, 30%'. Maintenance of
a reasonable waste degradation half-life is essential. Buildup of recal-
citrant organics will occur regardless of how a system is operated, but
the process may be slowed by repeated, modest waste applications.
Cessation of waste loading caused profound changes in system
dynamics. Respiration rates fell to low levels, remaining constant under
varying seasonal conditions (Figure 2). The decline in respiration rales
to present levels occurred as usable substrate presumably became
limiting.
Table 1 shows a comparison of the length of time to reduce oil con-
centrations to nominal levels. A concentration of 15% oil in soil by
weight (150,000 parts per million) was chosen as representative of present
conditions in the facility. The cleanup level chosen was 100 ppm (0.01%
oil in soil), current regulatory levels for petroleum hydrocarbons in soil.
The kinetic coefficients were estimated from CO2 evolution data cor-
responding to conditions of intense activity (1.0 mg CO,/g soil/day)
and closure conditions (0.10 mg CO3/g soil/day). The results of the
two calculations arc presented; time to reach the targeted cleanup level
using zero-order and first-order kinetic equations. A ten-carbon alkane
was chosen as a representative waste molecule for calculation of the
moles of oil per gram of soil, although the actual weight of material
present in the facility is likely heavier. The calculations presented in
Table 1 are likely extreme estimates. The time predicted by first-order
346 BIOREMEDIATION
-------
kinetics for natural remediation of the soils indicates the process is not
a practical alternative. The calculation assumes equilibrium conditions;
time to reduce a one-time application of 15% oil in soil is predicted.
However, the zero-order prediction is too simplistic; some materials
in the treatment soils have more complicated cometabolic (probably
first order for growth-supporting substrate) biochemical pathways. The
calculation assumes all substrate is completely assimilable by the
microbes. The indication is that the treatment soils will require reme-
diation. The community of microorganisms adapted for life in intensely-
loaded land treatment facilities require supplies of readily usable sub-
strate to maintain adequate degradation activity.
Table 1
Comparison of Times to Reduce Oil in Soil Concentrations
Initial Concentration = 150 mg oil/g soil (15% by weight)
Final Concentration = 0.
= 7.
Zero-order Kinetics
(1) Operating System
evolution rate =
t = 500 days
(2) Closed System
evolution rate =
t = 5000 days
1.056 X 10 moles oil/g soil
10 mg oil/g soil (0.01%)
042 X 10"fi moles oil/g soil
First-order Kinetics
(1) Operating System
• 1.0 mg CO2/g soil/day
t = 3650 days
(2) Closed System
: 0.1 mg CO2/g soil/day
t = 36,500 days
Remedial Options
The preceding discussion show that options and remedial strategies
are necessary for soils of closed land treatment facilities. The high con-
centrations of polynuclear aromatics limits the applicability of several
developing strategies for reuse of hydrocarbon-contaminated soils. Con-
taminated soils have been used in asphalt production, although limits
are set on hydrocarbon concentrations and clay content of the soils.
Fixation of soil hydrocarbons with chelating agents and polymer material
or batch chemical oxidation with subsequent reuse of the soil as fill
material have been used at sites with petroleum or gasoline contami-
nation. Again, it is unlikely that either process would be adequate for
the high concentrations of heavier molecular weight hydrocarbons in
the former treatment soils. Concerns about long-term stability of soils
with chemically-fixed hydrocarbons have also been raised.
The concentrations of petroleum waste in the treatment soils man-
dates that off-site disposal be in a Class I landfill. Costs for disposal
and transportation for 18,000 cubic yards of soil (conservatively esti-
mated for a 9 acre landfarm of 1 foot depth) could cost between $300
to $400 per cubic yard or 5.4 to 7.2 million dollars. Additionally, the
generator would maintain liability for the landfilled waste.
An aggressive in-situ remedial program re-stimulating the dormant
natural petroleum-degrading organisms should be investigated. Bench-
scale studies can determine an appropriate substrate addition to stimu-
late natural biodegradation and cometabolism of the larger-molecular
weight organics. The compound chosen would provide the driving force
for breakdown of the remaining waste materials. The substrate may be
a petroleum hydrocarbon or some other compound. Monitoring oxygen,
nutrients and soil water content would assure the correct environmen-
tal conditions. A surfactant carefully applied may facilitate the solubili-
zation of the more recalcitrant molecules. Insoluble or sparingly soluble
organics are readily adsorbed and act to bind soil particles together,
creating anaerobic zones.
Solubilization in pore or hygroscopic water promotes microorganism-
substrate contact. Microbial augmentation may be attempted, although
the natural petroleum-degrading organisms are highly adaptable and
ubiquitous. The oil in soil concentrations could be reduced to levels
where other remedial options could be utilized. Without substrate
additions, it is unlikely that biological remediation would be successful.
CONCLUSIONS
Soil characteristics of the closed treatment facility dictate that effec-
tive options be developed for management of the highly contaminated
soils. Biodegradation decreases to an unacceptably low level, in part
as a result of a high percentage of remaining high-molecular weight,
recalcitrant organics and a minimal supply of growth-supporting sub-
strate. Contaminated soils allowed to lie fallow for long periods restrict
land usage and increase the chance for undesirable environmental
impact. Soil remedial options currently in widespread use for gasoline
or waste oil contamination may have limited applicability because of
the nature and concentration of the petroleum residuals. An effective
in-situ remedial program would include development of a suitable
cometabolic substrate addition with maintenance of environmental con-
ditions conducive for biodegradation. Applications of such an approach
would include remediation of contaminated soils at former coal gasifi-
cation plants and abandoned oil fields slated for redevelopment.
REFERENCES
1. Morrison, A., Land Treatment of Hazardous Waste, Civil Engineering, May
1983, pp. 33-38.
2. Streebin, L.E., J.M. Robertson, A.B. Callender, L. Doty and K. Bagawandoss,
Closure Evaluation for Petroleum Residue Land Treatment, EPA Rept. No.
600/S2-84-162, U.S. EPA, Ada, Oklahoma, December 1984.
3. American Petroleum Institute, The Land Treatability of Appendix VIII Con-
stituents Present in Petroleum Industry Wastes, API Publication 4379, 1984.
4. Loehr, R.C., J.H. Martin and E.F. Neuhauser, Spatial Variation of Charac-
teristics in the Zone of Incorporation at an Industrial Waste Land Treatment
Site, Hazardous and Industrial Solid Waste Testing: Fourth Symposium, ASTM
STP 886, American Society for Testing and Materials, pp. 285-297, 1986.
5. Marshall, T.R., Biodegradation of Petroleum Wastes in Soil: The Microbial
Ecosystem and Optimization of a Treatment Process, Ph.D. Thesis, Univer-
sity of Southern California, 203 pp., 1988.
6. Harris, R.F. and L.E. Sommers, Plate-dilution Frequency Technique for Assay
of Microbial Ecology, Applied Microbiology, 16 (2), pp. 330-334, 1968.
7. Martin, J.P., R.C. Sims and J. Matthews, Review and Evaluation of Current
Design and Management Practices for Land Treatment Units Receiving
Petroleum Wastes, Hazardous Wfastes and Hazardous Materials, 3 (3), pp
261-280.
BIG-REMEDIATION 347
-------
Case Study of the Bench-Scale Solvent Extraction Iteatability
Testing of Contaminated Soils and Sludges from the
Arrowhead Refinery Superfund Site, Minnesota
Joseph A. Sandrin
Dorothy W. Hall
CH2M Hill
Milwaukee, Wisconsin
Rhonda E. McBride
US. EPA
Region V, Chicago, Illinois
ABSTRACT
(Solvent extraction is a separation process that has emerged as an
effective hazardous waste treatment technology.^It has been success-
fully applied to industrial wastewaters. soils and sludges contaminated
^m Bygrocarbons. petroleum products and heavy organic compounds.
Solvent extraction is being considered as an alternative treatment tech-
nology at the Arrowhead Refinery Superfund Site in Hermantown.
Minnesota, the location of a former waste oil recycling facility. Highly
acidic, metal-laden sludge bottoms and oil-saturated clay filter cake were
disposed of in a 2-ac lagoon. The peat layer underlying the lagoon and
the surrounding soils are contaminated with oil. metals and numerous
organic compounds.
Under subcontract to CH2M HILL. Resources Conservation Com-
pany (RCC) conducted bench-scale tests on sludge peat and soil wastes
from the Arrowhead Refinery site using its Basic Extractive Solvent
TfechoologyJB.E.S.T.*'). The results of the bencTT-scale lesl aM (he
applicability of the process to the wastes at the site are discussed.
INTRODUCTION
The Arrowhead Refinery Site occupies approximately 10 ac in north-
east Minnesota near Duluth. According to Minnesota Pollution Control
Agency (MPCA). milk cans were retinned at the site before 1945. From
1945 to February. 1977. the site was used as a waste oil recycling facility.
The waste oil was treated with sulfuric acid to deemulsify the oil/wafer
mixture. \\fcstewaier from the process was recovered and discharged
to the wastewater ditch. The waste oil was then filtered through a
clay/sand filter. The sludge from the deemulsification process and the
filler cake were disposed of in an uncontained 2-ac lagoon in a wet-
land on the site. The filter cake also was used as fill in the process
area adjacent to the lagoon (Figs I and 2).
Site Characterization
The U.S. EM and MPCA investigated the environmental effects of
on-site waste disposal from 1979 through 1984. The results of their in-
vestigations indicated that a variety of organic and inorganic con-
taminants are present at the site in the subsurface soil, sediment, surface
water, groundwater and sludge lagoon. The two major contaminant
sources defined in the remedial investigation were the contaminated
soils in the process area and the sludge and filter cake disposed of in
the lagoon (Fig. 3).
The surface soils consist of gravelly sand. silt, and fill material that
were deposited during site operations. Much of the soil is visibly stained
and saturated with waste oil. The lagoon contains a viscous, black oDy
liquid sludge and a black filter cake that consists of an oily clay and
Figure I
ArrdwhcacI Refinery Sludge I-ag
-------
SCALE IN FEET
LEGEND
•• ' ' EPA DITCH
' SITE BOUNDARY
NOTE: Arrows Indicate direction of flow.
Figure 3
Site Plot Plan
a silty sand and gravel fill layer. The entire lagoon is underlain by a
peat layer that appears to be persistent throughout the site and is also
highly contaminated. Contaminants detected at the site included poly-
cyclic aromatic hydrocarbons (PAHs), volatile organic compounds, lead,
zinc and small quantities of PCBs.
As documented in its ROD for the site, the U.S. EPA's selected
remedial action was thermal treatment of site wastes. The U.S. EPA
and MPCA are both interested in the application of alternative tech-
nologies that might achieve similar levels of treatment more economi-
cally than thermal treatment. As a result, the U.S. EPA agreed to fund
a treatability study of the refinery wastes using a solvent extraction treat-
ment process.
CH2M HILL had previously performed a remedial investigation and
feasibility study of the site for the U.S. EPA. Under contract to the
U.S. EPA, CH2M HILL subcontracted the treatability tests to Resources
Conservation Company (RCC).
SOLVENT EXTRACTION
Background
Solvent extraction technology has been used for many years as solvent
leaching to recover valuable minerals from ores, to remove unwanted
materials from coal processing operations and to de-oil quench waters
in refinery processing operations. More recently solvent extraction has
been used to treat sediments and soils contaminated with PCBs, wastes
generated by chemical manufacturers and oily hazardous and toxic
wastes.
Organic solvent extraction is particularly suited for treatment of oily
wastes, because the wastes can be separated into product oil, solids
and water fractions. Solvent extraction can effectively extract the oil
fraction of a waste, including PCBs. The remaining solids can some-
times be disposed of as non-hazardous wastes and the water discharged
to a wastewater treatment plant.
BIOREMEDIATION 349
-------
AIIOIIOCI
Oll/lolvinl ,
Ftnllcn.^-L liflpplng Colimn
Sltim
Conftimti
Oil PiodQd
Solid! Prodoel
Figure 4
B.E.ST. Process Flow Diagram
The successful application of solvent extraction depends on solvent
selection, process configuration, the nature of the waste and the con-
taminants, and the economic value of the recoverable compounds. Sol-
vent extraction can be an especially viable treatment alternative where:
(1) valuable products can be recovered from the waste; (2) the process
can yield non-hazardous residual solids; or (3) inordinate wastewater
disposal or air emissions problems are not encountered.
B.E.S.T.*
RCC is the owner of the B.E.S.T " solvent extraction technology.
a patented process that takes advantage of the peculiar solubility behavior
of certain aliphatic amines. Tnethylamine (TEA) has chemical and
physical properties that make it a good candidate for use in solvent
extraction.
At temperatures below 65°F. TEA is completely miscible with water
and is a good solvent for a variety of organic compounds such as PCBs,
PAHs, and petroleum products. The soluble organic and water compo-
nents of a waste can be separated from the solids component using TEA.
When TEA is heated, the solubility of water in TEA decreases to
less than 2%, separating the water fraction from the soluble organic
fraction. The TEA is then removed to yield an organic fraction. The
B.E.S.T. process separates the waste into three waste products: (1) a
solid with soluble organic contaminants removed. (2) a wastewater that
may require treatment before discharge and (3) an oil product that can
be recycled for energy recovery or incinerated (Fig. 4).
TEA is a basic compound that reacts with acids in the waste yielding
ammonium salts Excessive reaction of the basic solvent with an acidic
waste will result in the loss of expensive solvent. To minimize solvent
loss, the B.E.S.T. process includes the addition of caustic to increase
the pH of the waste above 11. The high pH has the side benefit of
precipitating low concentrations of metals into the product solids and
thereby potentially decreasing the teachability of metals in the EPA
toxicity test.
To evaluate the ability of the process to treat a given waste. RCC
conducts bench-scale instability tests at its laboratory facility. In the
trcatability tests, l-kg batches of waste are subjected to the same unit
processes a.s a full-scale operation. The performance of the process
is evaluated at each step of the process, and samples of the products
and process intermediates are analyzed to determine if the wastes are
being treated effectively.
Trtatabllily Testing
Samples of the contaminated soil, sludge and peat wastes were sent
to RCC. and bench-scale trcatability tests were conducted in May. 1989.
Samples of the raw waste and the treated waste products were analyzed
through the U.S. EPA's Contract Laboratory Program. RCC reported
its final results to the U.S. EPA in August. The final project report is
scheduled to be submitted by CH2M HILL in September for the US.
EPA and MPCA review. The results of the treautbility study and the
conclusions of the report will be available after the review of the project
report by EPA and MPCA. The results of the study are scheduled to
be reported in November.
350 B1OREMED1ATION
-------
Bioremediation Using Adapted Bacterial Cultures
Topic 1: Examination of Site Data and Discussion of
Microbial Physiology With Regard to Site Remediation
Ralph J. Portier, Ph.D.
Louisiana State University
Baton Rouge, Louisiana
INTRODUCTION
The number of organic compounds introduced into the environment
by humans has increased dramatically in recent years.' As a conse-
quence of this xenobiotic (i.e., man-made) introduction, the fate of these
compounds, such as pesticides, polycyclic aromatic hydrocarbons and
domestic wastes, in the environment is a very important issue. Of par-
ticular concern is disappearance, persistence, and/or partial transfor-
mation of such compounds and its potential hazardous effect. While
many are readily biodegradable, others have proven to be recalcitrant
and persistent in soil and water. In recent years, a great deal of research
has been done on the biochemistry and genetics of xenobiotic-degrading
microorganisms. Both the newer literature on biotechnology and the
older literature on industrial microbiology describe important commer-
cial processes in which microbial microorganism cultures play an im-
portant role23-
This discussion of microbial processes of importance to scientists
and engineers involved in an active remediation program on refinery
sludges/solids is presented as an overview of the subject only. The
primary focus is bacterial processes due to the considerable volumes
of information available. However, fungal and actinomycete contribu-
tions to soil biotransformation processes/productivity are of equal im-
portance and will be presented within the context of the discussion.
Although some bacteria and fungi can cause adverse effects, most
species are benign, and many are involved in processes of direct benefit
to man. Most of the adverse effects are subject to control and a relatively
limited number of species are pathogenic. Life on earth depends on
their activity, playing an important role in the biotransformation and
mineralization of organic compounds, such as the transforming of free
nitrogen molecules in the air for use by plants.
MICROBIAL FORMS
Microbe/Surface Interactions
The evolution of different forms of life has resulted in many large
groups which can be divided quite clearly into two categories, plants
and animals, showing a variety of well-established characteristics
specific to each one. Microorganisms have developed in a different way
in which the "plant-animal" relationships are not always well-defined
and the criteria for life has to be modified3. The majority of organisms
which comprise the major microbial group are of microscopical dimen-
sions, generally with no differentiation of tissue as in higher organisms
and living in an interrelated group in nature.
Bacteria include a great variety of unicellular microorganisms of dif-
ferent size and shape, present in almost all natural environments, often
in extremely large numbers. Bacteria are usually about 0.2 to 1.5 /mi
in diameter. The mean diameter is about 1.0 /mi. Bacteria have rigid
cell walls, as do plants, but some are motile and require organic
nutrients, as animals do.
Bacteria, along with molds, yeasts, viruses, and algae are allocated
to the vegetable kingdom. Bacteria have been assigned to the Protophyta
division, class Schizomycetes. The class Schizomycetes may be divided
into 10 orders. The orders Eubacteriales and Pseudomonadales con-
tain the largest number of species and include most of the bacteria im-
portant to man.
Morphology of Bacteria
Bacteria are procaryotic cells, that is they do not have a true nucleus.
The procaryotic nucleus has no membrane, does not undergo mitosis,
and its hereditary material is contained in a single naked DNA molecule.
The procaryotic cell has none of the specialized structures found in
eucharyotic cells, such as mitochondria for respiration, an endoplasmatic
reticulum as an extension of the cell membrane, lysosomes containing
hydrolytic enzymes, and a Golgi apparatus to transport metabolic
products4. They are usually not photosynthetic microorganisms. On
the basis of their shape, bacteria are divided into three conventional
groups: cocci, bacilli and spiral forms.
The cocci are spherical or nearly so. They vary in size from 0.5 /mi
to 1.0 /mi in diameter, and their arrangement depends on the order of
successive cell division. If this is random, the organisms may occur
in clusters and are called staphylococci from the Greek Word for grape.
When the division takes place in the same plane and the daughter cell
adheres to another, chains are formed, called streptococci.
Bacilli, by far the most common, are bacteria shaped-like rods or
cylinders. They are about 1.0 ^m to 10 /mi long and 0.3 /im to 1.0 /mi
wide. The end of the rod appears to be rounded or square, and some
tend to form chains. Spiral forms comprise a large variety of cylin-
drical bacteria which, instead of being straight like bacilli, are con-
voluted in varying degrees. Vibrios are curved rods, Spirilla are spirals
with their bodies relatively rigid, and Spirochetes are also spirals but
able to flex and wriggle their bodies.
Filamentous forms of bacteria are also found and these may be several
hundred /mi long but are usually only about 1 to 2 /mi in diameter.
The shape of bacteria is determined by its heredity, but some organisms
may show morphological changes depending on age and certain en-
vironmental conditions.
Structure of a Bacterial Cell
It is generally accepted that all kinds of living cells have some form
of outer wall or membrane, cytoplasm, and nuclear material, with each
component making its own contribution to the life of the cell. The outer
part of the bacterial cell is made up of three definite structures: slime
layer or capsule, cell wall, and cytoplasmic membrane.
The slime layer is the outside coating of the bacterial cell. It is a
BIOREMEDIATION 351
-------
jellylike layer and may vary in thickness even in different cells of the
same culture. When it becomes sufficiently thick and firm to have a
form, it is called a capsule. It is usually a polysaccharide (or a polypep-
tide) and continually produced by the cell as a result of its metabolic
activity. Its formation may depend on the presence of carbohydrate in
the environment. In many bacteria, this structure is a high-molecular-
weight polymer (a long molecule made up of repeating structural units)
of a simple hexose sugar, such as glucose (dextrans) or levulose (levans).
In others, the structure is chemically much more complex, being formed
of units of simple sugars (glucose, mannose, and galactosc). derived
sugars (amino sugars), and sugar acids (gluconic and glucuronic). The
slime layer is not an integral pan of the cell but a result of its meta-
bolic activity, and consequently it is greatly influenced by the
environment* The capsular material confers type specificity on the
organism; for example, the type of pneumonia which develops in a host
depends on the molecular composition of the capsule.
The chemical composition of the cell wall varies with different
bacteria. All bacteria are made up of proteins and complex carbohydrates
or polysaccharides, frequently with large amounts of fat or lipids. The
structural component of the cell wall is murein. The most commonly
studied cell walls most studied are Slaphylococcus aureus. a complex
polymer of N acetyiglucosamine (the basic structural unit in chitin from
insect exoskeletons). In general, bacteria cell walls appear to be double
or triple layered structures. The cell wall limits the volume occupied
by the cytoplasm, providing a strong rigid structural component that
can support the high osmotic pressure caused by high concentrations
of cytoplasmatic content in the cell. The cell wall also plays an impor-
tant role in cell division and a major role in regulating the passage of
various materials between the internal and external environment of the
organism.
Bacteria can be divided into two large groups on the basis of a dif-
ferential staining technique called the gram stain. These two groups
of bacteria differ mainly in their cell walls: gram-positive and gram-
negative cell walls. The gram-positive cell walls consist of 60 to
100 percent murein. Some have a glycerol type of teichoic acid located
between the cell membrane and the cell wall. Gram-negative cell walls
are chemically more complex, containing about 10 to 20 percent murein
There is a second structure outside of this layer composed of proteins
and fatty acids linked to polysaccharides.
The cytoplasmic membrane, located just inside the rigid cell wall,
is a semipermeable membrane composed mainly of proteins and lipids
acting as pan of the osmotic barrier between the external and internal
environments of the cell, regulating the permeability of substances
entering and leaving the cell. It contains many of the oxidation-reduction
enzyme systems concerned with energy metabolism. The cytoplasmic
membrane accounts for 8 to 10 percent of the dry weight of the entire
cell, and it is chemically composed of a molecule containing a
lipoprotein.
The cytoplasmic membrane always initiates division of the cell and,
because of its semipermeable nature, plays an important role in con-
trolling the passage of waste products out of the cell without permit-
ting the cell contents to escape. The cytoplasm is the internal environ-
ment of the cell. It is a colloidal system containing salts, sugars, pro-
teins, fats, carbohydrates, vitamins, granules, and other materials
characteristic of a particular organism. The major component of liv-
ing cells is water, which accounts for approximately 75 percent of the
total mass of the cell. It serves as the medium in which soluble com-
ponents are diffused, and it serves to hydrate large molecules whose
functions depend not only on their chemical composition but on their
configuration in space as well. The cytoplasm contains most of the en-
zymes necessary for metabolic processes of the cell and growth of the
organism.
Procaryotic cells, or cells restricted only to microorganisms, do not
possess a true nucleus The nuclear region is seen as a weakly con-
trasting area that contains thin fibrillar material of deoxyribonucleic
acid (DNA), the genetic material of the cell. Sometimes more than one
nuclear region is seen in a single cell, but each of these probably con-
tains only a single DNA molecule.
Various inclusions have been observed in the cytoplasm, such as
granules of starch-like compounds called ganulose, flat droplets,
pigments, and a polymer of inorganic phosphate called volutin. Under
deficient conditions these granules are broken down to provide useful
energy and building blocks to the cell. When free from inclusions, the
cytoplasm appears homogeneous.The cytoplasm also contains some
particles called ribosomcs which are part of the protein synthesizing
machinery of the cell composed of ribonucleic acid (RNA) and protein.
The bacteria nucleus differs from the nucleus in higher organisms
because it has no membrane with only one chromosome in the form
of a ring.
Many types of bacteria have the ability to move by themselves. Almost
all spiral bacteria and many of the bacilli are motile. Cocci are usually
nonmotile. The propulsive mechanism of motion is a threadlike
appendage called flagellum, arising from within the cytoplasmic mem-
brane, generally several times the length of the cell. Motility can be
observed most satisfactorily in young cultures.
A large number of bacteria have shod fibers, called pili or fimbriae.
attached to their walls. These filamentous appendages, usually shorter
than flagclla. arc composed of protein and have been found only in
gram-negative bacteria. Such bacteria have more tendency to stick to
each other because pili apparently are used for attachment to the sur-
faces. Pili can be dissociated into smaller identical subunits called pilin.
This accumulated mat is referred to as a glycocalyx.
MICROBIAL FUNCTIONS
Nutritional Requirements
Microorganisms can be classified into three major groups based on
the types of material used as energy sources: (1) chemoorganotrophs
that use the energy of organic compounds; (2) photoautotrophs that
utilize radian energy: and (3) chemotithiotrophs that oxidize inorganic
molecules. Most bacteria are chemooiganotrophs.
As do other forms of life, bacteria require water, minerals, vitamins.
and sources of carbon and nitrogen for much the same purposes, but
in relatively smaller quantities; tap water will often meet their mineral
needs. Necessary mineral ions include such trace elements as
molybdenum, manganese, and cobalt. Phosphates are frequently added
to a media, both as a source of phosphorus for the synthesis of nucleic
acids and as a buffer for the media against excessive acidity to neutralize
acids. Tables 1 and 2 list the major and minor bio-elements, respec-
tively, their sources, and some of their functions in metabolism.
T»Mt 1
The Ten Major Bio-Elements, Their Sources, and
Some of Their Functions in Microorganisms:
Adapted From Bacterial Metabolism (Gottschalk, 1979)
Element
ISource
"Function in Metabolism
C
o
II
N
S
organic compounds, CX>2
O2. HjO. Mjinic compounds. OO2 nuun constituents of cell material
H2,H2O, organic compounds
N1U4. HO-3. N2. organic compound!
SO2 4, 1IS-. SO, S2O2-3. constituent of cysieine, melhtoninc
organic sulfur compounds ihiunin pyrophosphaie, coenzymc A.
biotin. and lipoic acid
P HFO2 • 4 condiment of nucleic acids, phosphotipids
and nuclcoodes
K K« principal inorgank cation in the cell.
cofic lor of some enzymes
Mg Mg2-t coficior of many enzymes (e-g . Irinases);
present in cell walls, membranes, and phosphite
csten
Ca C«2+ cofactor of enzymes, present in eioenzymes
(amylascs, proteases); Ca-dipicolinate is an
important component of endosporcs
I'c Fe2», Fe3< present in cytochromes. fcrredojtins, and other
iron-sulfur proteins; cofactor of enzymes
(some dehydraiases)
352 B1OREMEDIATION
-------
The chief vitamin requirement for bacteria is B complex. Biological
assay methods based on these requirements for specific vitamins or
minerals by specific strains of bacteria have been developed7.
The source of carbon in synthetic media is usually glucose but other
carbohydrates can be used in a diagnostic test. Because few species
form lipases, or enzymes capable of hydrolyzing fats, generally they
do not use fats as such. However, many can utilize salts of the lower
fatty acids, especially acetic and butyric acids (and other organic acids),
required mainly for synthesis of cell protein7. For precise study of
bacterial physiology, a synthetic medium made from known constituents
is preferable.
The energy released by a catalyzed enzyme oxidation of carbon,
whether a PAH or glucose, is accumulated in the chemical bonds of
the adenosine-5- triphosphate (ATP) when formed from the addition
of inorganic phosphate H2PO44— (Pi) to adenosine diphosphate (ADP)
in a vital energy process required by all living cells called oxidative
phosphorylation. Cellular processes utilize the energy of biological ATP
changing back to adenosine-5-diphosphate (ADP) and inorganic
phosphate in a reverse reaction (Fig. 1). The formation of phosphate
bonds requires energy. The energy stored in ATP is released when the
bond connecting the last phosphate is broken. The principle involved
when ATP absorbs the energy given off during oxidation and transfers
it to the different processes of the cell is called energy coupling and
is applied to many metabolic reactions. The energy liberated in these
reactions is directed primarily toward biosynthesis of cell materials
(Fig. 2).
NH2
I
000 N
I-P-0-P-0-P-0-CH2
Q- 0- 0- I ^
C H
I XI
H C-
I
OH
H20
H C
I/ I
• C H
I
OH
Adenosine triphosphate ( ATP )
Cellular processes
-Q-P-O-P-O- [edenoslnel * -Q - P - OH + energy
0~ 0~ OH
ADP
PI
Figure 1
Reverse Reaction Between ATP and ADP
H20
energy
absorbed
+ Pi
energy
released
Figure 2
Relation Between Adenosine Triphosphate ATP and
Adenosine Diphosphate
Adenosine triphosphate (ATP) consists of a monosaccharide called
ribose to which a nitrogen heterocycle, adenine, and a triphosphate group
are attached. The triphosphate group has two phosphates bonds. They
work as excellent phosphorylating agents and are used as such in a large
number of reactions by all organisms, activating the intermediates of
cell metabolism to further reactions of condensation, cleavage, and
reduction.
The chemical change in the body of living organisms depends on
enzymes. The enzymes formed by a bacterial cell determine whether
a bacterium can digest a complex material and use it for food. These
compounds are biological catalysts that increase the rate of reaction
but are not used up in the process. Enzymes are very large protein
molecules that bind to the substrate, reacting chemically. Substrate is
defined as the compound on which an enzyme exerts its catalytic effect.
Inside each microbial cell about 4,000 to 10,000 different chemical
reactions take place in order for that to grow and function8. Each
enzyme has a special region on to which it binds, called an active site,
as well as specific activity for the substrate. Certain substances known
as inhibitors prevent or slow down the action of enzymes. Competitive
inhibitors are molecules similar to those of the substrate. They bind
reversibly at the active site, stopping the enzyme from catalyzing the
reaction8.
Molecules involved in a reaction must have a certain amount of energy,
called the activation energy. Enzymes decrease the activation energy
barrier of the reaction resulting in more product in a shorter period
of time. Some enzymes need an extra non-protein part essential for
their functioning, called a cofactor. If the cofactor is an organic molecule
it is called a coenzyme.
The utilization of O2 as an electron acceptor is called respiration and
can be measured by the uptake of oxygen gas by the respiring organism.
Many respiratory reactions are fundamental for almost all forms of life,
including bacteria. If free oxygen enters the reaction, it is called aerobic
respiration. Atmospheric oxygen functions as the final hydrogen acceptor
in the series of oxidation and reduction reactions and liberates the energy
from food in the metabolic process. When O2 accepts electrons, it is
reduced to H2O. Since enzymes are proteins and exist in living cells,
certain environmental conditions, such as temperature, pH, and salt
concentration, must be met in order for the enzyme to be active.
One of the most important factors affecting the rate of microbial
growth is environmental temperature. There is always a temperature
below which growth will not occur because of the deactivation of the
enzyme-catalyzed system, as well as a maximum at which heat
denaturalization will occur. Between these limits, there is an optimum
temperature for bacterial growth, resulting in a very rapid increase in
the rate of activation of heat-sensitive cell components, such as enzymes,
ribosomes, DNA, and membranes4 In general, an increase in
temperature produces increased molecular motion which promotes more
rapid bacterial growth. Most enzymes experience their optimum activity
at a temperature between 20 °C and 30 °C.
The hydrogen ion concentration (i.e., the acidity or alkalinity) of the
solution markedly affects the activity of an enzyme. Some enzymes
are active at rather low acid pH values, pH 3 to 4, while others may
be active at alkaline pH values as high as 11 or 12. The majority of
bacteria (whole cell) prefer a neutral medium neither markedly acid
nor alkaline, demonstrating a maximum activity in the range of pH 6
to 8.
Additional concentrations of sodium chloride have been shown to
increase bacterial growth by increasing the osmotic pressure up to an
optimal point. However, when the concentration is too high, osmotic
pressure is raised to a level that inhibits bacterial growth.
Reproduction of Bacteria and Population Growth
Sexual reproduction has been demonstrated in only a few bacteria;
they are, for the most part, asexual. Bacteria multiply by an elongation
of the cell, followed by a division of the enlarged cell into two cells
by a vegetative process called binary fission. Although bacteria can
vary in size, they retain their unicellular structure, and the primary
definition of bacterial growth is reflected in an increase in the number
of individuals. When a bacterial cell grows and divides, the final out-
BIOREMEDIATION 353
-------
come is two cells where there was one. Cell division requires the
doubling of all cell constituents and their orderly partitioning into two
daughter cells. For this process to be completed, therefore, every atom
and molecule in the parent cell has to be duplicated and then inserted
into its correct place in the developing structure that will eventually
become the mature daughter cell.
Under favorable conditions almost all bacteria are able to reproduce
very rapidly. Nuclear division is the first step, followed by the division
of the cytoplasm into two equal portions separated by an inward growth
of the cytoplasmatic membrane. A cross wall then divides the cell, pro-
viding each daughter cell with a complete cell wall, followed by the
final step which separates the sister cells, right after the cell wall is
formed. Some bacteria do not separate easily and form chains. When
long chains are formed, they appear to be rough or wrinkled because
of the resistance to their continued elongation'
Growth of a single cell with an orderly increase in the conMituems
of the cell is going on all the time, while multiplication is only occurring
at the instant of the division4. When bacteria are inoculated in a
suitable medium and incubated under optimum conditions, the popula-
tion of bacteria generally increases through several well defined steps
(Fig. 3) in a predictable manner.
Reproduction usually does not begin immediately. If conditions are
favorable, a period of adaptation to the new environment is required
by the organisms to begin their growth. This is called the lag phase
and its duration usually varies from an hour to several days depending
on the type of bacteria, the age of the culture, and the available nutrients
provided in the medium. This period is characterized by (he lag in multi-
plication only since the cells are very active melabolically (Fig. 3).
•.- L*g phn*
t>.- Log ph*M
c.- Stdlona
time
Figure 3
A Common Form of Bacteria Growth Under Favorable Conditions
Following the lag phase is the period of most rapid reproduction in
which the typical characteristics of active cells are observed. This is
the period of exponential growth. Individual cells grow approximately
linearly with time, while the population of cells grow exponentially,
doubling at each period of cell division. It is the phase of a constant
and rapid generation time, called the log phase.
When rapid growth is halted by exhaustion of nutrients, a deficient
supply of oxygen or an accumulation of toxic end product, growth
declines to a point where the number of cells remains constant. This
is called the stationary phase. During this phase the cells remain in
a slate of suspended animation. The length of the stationary phase
depends upon favorable conditions and the specific microorganism.
Unless the cells arc transferred into a new environment capable of sup-
porting continued growth, they will eventually die. This death phase
in old cultures often becomes exponential in a repeated process until
no cells remain.
Kinetics of Bacterial Growth
When the logarithm of the number of cells is plotted versus the lime
of growth, a straight line results. The rate of exponential growth is
usually expressed as the generation time, or doubling time, which is
the time it takes for the population to double". Bacterial cells can be
maintained in the logarithmic phase by continually transferring them
to a fresh medium of the same constitution. The process can be con-
tinued automatically by using a chemostat"
The generation time of an organism can be determined during the
log phase. At this period, some cells are just beginning to divide, others
arc half divided, and still others are finishing division. Each genera-
tion results in a doubling of the cell number. With this information,
the following can be used to calculate the generation time:
B = number of bacteria at beginning of time interval
B = number of bacteria at end of any interval of time (t)
g = generation time, usually expressed in minutest = time, usually
expressed in minutes
n = number of generations
B, = B > 2'
By taking the logarithms of both sides of the equation, we find
log B, = log B + n log 2
Solving for n yields
n =
log B, - log B
log 2
Since by definition
t
and
n =
g
Substitution generates the following equations:
( _ log Bt - log Bo
g
then
log 2
t log 2
log Bt - log Bo
Generation time depends on the type of organism, concentration of
available nutrients, temperature, pH, and oxygen. In general, species
multiply rapidly when provided with favorable conditions".
MICROBUL PROCESSES
The Concept Of Btotransfornwtion/Biodegradation:
Polycyclic Aromatic Hydrocarbon Degradation As A Case Study
intrvductitw
Microbial metabolism of hydrocarbons has been reported in the litera-
ture tor several decades. Some of the first investigations date back as
early as 1928, when Gray and Thornton" first reported soil bacteria
capable of decomposing certain aromatic compounds. In 1941, Bushnell
and Haas'-' documented microbial degradation of certain hydrocarbons.
Sister and Zobcll'-', in 1947 used microorganisms of marine origin in
their experiments to degrade aromatic hydrocarbons. They studied the
utilization of polycyclic aromatic hydrocarbons (BVHs) by mixed cultures
of marine bacteria. PAHs were introduced into seawater cultures ad-
sorbed to ignited sand. The amount of PAH metabolized by the bacteria
was determined by measuring the amount of carbon dioxide evolved
in hydrocarbon oxidation and subtracting carbon dioxide produced by
the control cultures. In these experiments, phenanthrene and anthracene
were metabolized more rapidly than naphthalene. benz(a)an(hracene,
and dibenz(a,h)anthracene''
In the present decade, it is well known that hydrocarbons are
ubiquitous in the environment and even found in relatively pristine areas.
Their sources are of natural as well as anthropogenic origin. Due to
354 BIOREMEDIATION
-------
the toxicity, mutagenicity and carcinogenicity that many of them exhibit
after undergoing metabolic activation, hydrocarbons in the environment
may pose a hazard to the biota and, ultimately, to human health13
Major environmental fate/transport mechanisms include:
evaporation (volatilization)
photochemical oxidation
sedimentation
microbial degradation
Of these, microbial degradation is the area most extensively studied
and commercialized as evidenced by the most recent developments in
biotechnology and genetic engineering. It is a major mechanism for
compound removal from sediments and terrestrial systems. Microbial
degradation of aromatic hydrocarbons by bacteria as well as fungi has
been documented in numerous publications. The degradation processes
are generally inversely proportional to the ring size of the respective
PAH molecule. The lower weight PAHs are degraded more rapidly, while
molecules with more than three condensed rings generally do not serve
as amenable substrates for microbial growth24, hence, the effectiveness
of creosote as a wood preservative.
Aromatic and Aliphatic Hydrocarbons
The capacity of microorganisms to grow in a given habitat is deter-
mined by their ability to utilize the nutrients in their surroundings14
Among the energy sources available to be utilized by soil heterotrophic
microorganisms are cellulose, hemicellulose, lignin, starch, chitin,
sugars, proteins, hydrocarbons and various other compounds14.
Numerous hydrocarbons, or their derivatives, are naturally synthesized
within the soil while others are added to the soil from various pollu-
tion sources. Their mineralization and formation by the indigenous
microflora are a fundamental component in the general carbon cycle14
The three major types of microbial metabolism are: fermentation,
aerobic respiration and anaerobic respiration15'16. Aerobic respiration
plays the most important role in the transformation of PAHs. Very lit-
tle anaerobic respiration of PAHs has been reported. However, anaerobic
biodegradation of PAHs has been observed where suited electron ac-
ceptors were supplied" Aerobic respiration initially involves the in-
corporation of molecular oxygen in the hydrocarbon molecules. The
hydrocarbons are then converted to more oxidized products. Energy
produced during these oxidation processes is partially used in the syn-
thesis of protoplasmic constituents15.
Definition
Hydrocarbons are compounds containing carbon and hydrogen.
Aliphatic hydrocarbons are straight or branched chain hydrocarbons
of various lengths. Aliphatic hydrocarbons are contained naturally in
waxes and other constituents of plant tissues as well as in petroleum
or petroleum products. Their transformations are therefore of great
significance in the terrestrial carbon cycle14. The rate of their decom-
position is markedly affected by the length of the hydrocarbon chain14
Aromatic hydrocarbons contain the benzene ring as the parent
hydrocarbon. Several benzene rings joined together at two or more ring
carbons form PAHs. The toxicity of these molecules is determined by
the arrangement and configuration of the benzene rings. The hydrogens
in the aromatic hydrocarbons may or may not be substituted by a variety
of groups. Some of the common substituents are -Cl, chloro; -Br, bromo;
-I, iodo; -NO2, nitro; -NO, nitroso; and -CN, cyano60
Sources and Formation
Most of the aromatic hydrocarbons are initially formed by the
pyrolysis of organic material15. In this process, the temperature deter-
mines the type of compound formed. For example, unsubstituted PAHs
are formed at high temperatures (2,000 C) whereas alkyl-substituted
molecules predominate at 80-150 C. The latter temperature range is
usually associated with the formation of petroleum15 Generally, PAHs
are formed when organic material containing carbon and hydrogen is
subjected to temperatures exceeding 700 C, which is the case in pyrolytic
processes and with incomplete combustion16 Some common sources
associated with incomplete combustion are cigarette smoke, automobile
exhaust, and industrial processes.
The higher the number of joined benzene rings, the lower the rate
of degradation. The very high molecular weight PAHs are less signifi-
cant in environmental pollution problems, due to their low volatility
and solubility16 The growth rates of bacteria on PAHs are directly
related to the solubilities of the PAHs16- Solubility and relative adsor-
bance are the most important physical properties that influence the rate
of transformation. Among the chemical properties, photochemical reac-
tivity is the most relevant. Tricyclic or larger PAH and related
heterocyclic systems show a very reactive photochemical behavior. They
have strong UV adsorption at wavelengths longer than 300 nm (present
in solar radiation) and most are readily photo oxidized. Photo oxida-
tion plays one of the major roles in the removal of PAHs from the
environment17'18'19 Adsorbed PAHs are photo oxidized more rapidly
than dissolved PAHs16.
The chemical structures of some of the major aromatic hydrocar-
bons are shown well known15 Biological activity of these compounds
depends on their inherent stereochemistry. The addition of another
benzene ring in a select position of the compound can result in the for-
mation of a powerful carcinogen, even if the parent compound does
not exhibit much toxicity20. The reactive sites of the molecules are
called "Bay-regions20." Such a Bay region is found in phenanthrene,
the simplest PAH. It resembles that of benz(a)-anthracene and
benz(a)pyrene, and is the region between an angular benzo ring and
the rest of the molecule21'22. If dihydrodiol-epoxides are formed in this
region, the molecule becomes very biologically reactive and is suspected
to be a ultimate carcinogen. The primary active carcinogen is usually
in the form of a diol epoxide21. Phenanthrene itself has been shown
to be inactive or only slightly mutagenic in Salmonella assays, but its
metabolites may be highly mutagenic and tumorigenic22.
Historically, it was believed, that a certain area, called the "K region"
was related specifically to the carcinogenic potential of a hydrocarbon
compound. Evidence now suggests that activation of PAHs is not likely
associated with this K region, but rather occurs via a two step oxida-
tion with the eventual formation of dihydrodiol epoxide20 Another
portion of the molecule, called the "L region" can increase the
carcinogenic potency of the molecule, if there are substituents on these
positions (i.e., the 7 and 12 carbons in benz(a) anthracene-20.
Microbial Metabolism of Hydrocarbons
There are various and controversial scenarios reported in the literature
as to the physical form under which the hydrocarbons are metabolized.
Some studies indicate the presence of large hydrocarbon droplets, others
mention micro-drops as small or smaller than the microbial cells, still
others suggest the importance of the water soluble fraction (WSF) or
the utilization of the hydrocarbons in a vapor phase23 There are also
reports on the importance of emulsifying agents for initiating hydrocar-
bon utilization. However, most reported microbiat hydrocarbon
metabolism processes are intracellular oxidation processes23.
Historically, most of the investigations of PAH biodegradation were
concerned with measuring the amount of CO2 produced or the frac-
tions of the toxicants (parent, molecule) converted into CO2. In these
early studies, CO2 production was the major focus of attention with
little consideration paid to the intermediates formed. Only recently has
it been recognized that there is a need to investigate these metabolites
and the ratio of polar compounds to CO2. The oxygenated polar com-
pounds may be highly mutagenic and/or accumulative in the aquatic/ter-
restrial environment and thus be dangerous to living cells. Recent ad-
vances in analytical techniques (such as Thin Layer Chromatography
and/or MS) have revealed the subtle complexity of biotransformation
intermediates and end products.
Bacterial Transformation (Biotransformation)
Bacteria are the dominant group involved in the degradation of PAHs.
The most widely occurring species are Pseudomonas, Myobacterium,
Acinetobacter, Arthrobacter, Bacillus and NocardiaH. Bacteria can
oxidize PAHs ranging from the size of benzene to benzo(a)pyrene. For
more highly condensed PAHs, there is little evidence of bacterial
oxidation15
The mechanisms used by bacteria for the introduction of hydroxyl
moieties into PAHs will depend on whether the substrate contains alkyl
BIOREMEDIATION 355
-------
substituents". The initial step of aromatic metabolism consists of the
modification or removal of substituents on the benzene rings and the
introduction of hydroxyl groups'7. The first metabolites of un-
substituted PAHs created by bacteria are m-dihydrodiols, formed by
the incorporation of two atoms of molecular oxygen. Fungi, in con-
trast, form f/wu-dihydrodiols. The enzymes catalyzing these processes
are oxygenases, known as cytochrome P4W enzyme complexes.
Bacteria use dioxygenases, a multi-componenl enzyme system consisting
of a flavoprotein, an iron-sulfur protein, and a ferrodoxin".
Although, initial phases of the degradation pathways differ, the
reactions proceed such that only a few common and key intermediates
are produced. These few are then metabolized by essentially similar
processes. Most common of these intermediates are catechol, proto-
catechuic acid, and to a lesser degree, gentisic acid" These three
molecules have in common the presence of two hydroxyls. The products
of these reactions, namely pyruvate, fumarate, and succinate may then
be incorporated in the TCA and other biochemical cycles. The degrada-
tion pathways involved are dictated by the site of cleavage of the aromatic
nucleus.
Naphthalene and its alkylated homologs are among the most water-
soluble and potentially toxic compounds in petroleum. The product of
bacterial oxidation of naphthalene is catechol. There are also different
pathways for the bacterial oxidation of phenanthrene" Oxidation of
this compound by fungi has not been reported. Special interest has been
paid by various researchers to the degradation of anthracene and its
derivatives. These compounds are not acutely toxic, bui possess a struc-
ture also found in other carcinogenic PAHs" Degradation of
anthracene has been reported by bacteria as well as fungi and follows
the general degradation pattern of the other PAHs.
Fungal Transformation (Biotransformation/biodegradaiion)
Many fungi cannot grow with PAHs as a sole source of carbon and
energy, but still have the ability to oxidize these compounds15 Fungi
carry out reactions similar to mammals in the degradation process.
Therefore, fungi are often used as model systems. Their enzyme systems
for the oxidation of PAHs differs from that of bacteria (e.g., mono-
oxygenases) and is similar to that of higher organisms. The cytochrome
P45o mono-oxygenase system catalyzes the initial steps in the oxidation
of these lipophilic PAHs. Many fungi add hydroxyls lo the ring struc-
tures without being able to open the ring, but subsequent ring opening
and cleavage of ether bonds can then be brought about through com-
etabolic conversions'*. Cometabolism is defined as the metabolism of
a compound by a microorganism that the cell is unable to use as an
energy source or source of growth" An example for a fungal
metabolic pathway quite similar to those in mammalian systems for the
oxidation of naphthalene is given by DoulP-'1 In contrast to bacteria,
fungi incorporate only one atom of molecular oxygen into naphthalene
via a cytochrome P4SO mono-oxygenase0.
BIOKINETICS OF PETROLEUM HYDROCARBONS
IN SLUDGES/SOILS IN A BIOLOGICAL
CONTACT UNIT (BCU)
Introduction
The following section reviews the results of data sets generated in
a laboratory/field pilot test of a biological contact unit for the semi-
continuous treatment of petroleum hydrocarbons. Investigations were
carried out in a multiple-task effort to achieve, by microbiological
methods, detoxification of contaminated soils at an abandoned
petrochemical facility along the Mississippi River. This facility,
designated a CERCLA/SARA site by state and regional environmental
agencies, presented particular difficulties using non-biological conven-
tional methods in accomplishing remediation due to close proximity
to the flood protection levee (dike) system of the river. Waste materials,
consisting primarily of aliphatic and polycyclic aromatic hydrocarbons
(PAH's) found in buried soil/sludges and lagoon wastes were examined.
Optimal toxicant loading levels were evaluated on the basis of
biodegradative potential tests and acute toxicity of leachate.
Microbiai ATP and microbial diversity were used in conjunction witfi
the Microtox™ Test to establish an acceptable land treatment experi-
mental design. The biodegradative potential of the microbial consor-
tium was evaluated using laboratory mesocosms (phase D) at a predeter-
mined optimal waste loading rate, based on percentage oil and grease
(O&G). mixed with a predetermined optimal soil mixture of river silt
and sandy clay (one part river silt: two pans sandy clay). Experimental
mesocosms were inoculated with an adapted indigenous microflora.
Microbial ATP, microbial diversity and the Microtox™ test were used
to establish the detoxification efficiency. Quantitative toxicant concen-
trations and transformations were documented by GC/MS methods.
GC/MS data in phase D studies (mesocosms) and phase ffl studies (field
verification studies) documented substantial biotransformation and
biodegradation of the wastes at these optimized loading rales.
The hazardous waste site investigated was located on the East bank
of the Mississippi River near Darrow. LA. The site was designated as
a priority site for Superfund assistance in April, 1982, scoring highest
out of the five qualifying sites in Louisiana, with 48.98 on the EPA
Superfund list. This abandoned oil reclamation facility, the Inger Oil
Refinery, was operated between 1967 and 1978. Wrete oils were brought
to the site by barge and truck, re-processed in cracking lowers by heating.
with produced final products being transported from the facility by truck.
As pan of plant operations, sludges were stored in large, open lagoons
and/or buried shallow pits. Some wastes were spilled into an adjacent
swamp in March 1978, contaminating a total of 16 acres of the surroun-
ding area. This spill was associated with the unloading of used oil from
a barge in the Mississippi River, A shut-off valve failure or human error
led to overtopping a tank and a containment area. Failure by the owner
to clean up the site resulted in the formal declaration by the Louisiana
Environmental Control Commission in June. 1981. that the site was
abandoned.
The site occupies about sixteen acres, including a 1.5 acre swamp.
The most highly contaminated wastes are found in the tanks, lagoons,
and diked containment areas. Contamination is found to a depth of three
to five feet in the areas of the closed lagoon and filled portions of the
swamp. Swamp sediments are less contaminated. The wastes identified
at the site were consistent with the nature of the oil reclamation plant.
They were mixtures of refinery oils, motor oils, and lubricating oils.
As is typical of waste oils, hazardous priority pollutants such as benzene,
toluene and PAH's were present. No PCB's were found; very low levels
of chlorinated hydrocarbons and low levels of heavy metals were found.
The site soil consists predominantly of silly and sandy clays, silts
and fine sands, to a depth of about US to 125 feel* Below this is a
substratum silly sand, a potential water supply source. The average ver-
tical and horizontal permeability is about 1 x K) ' cm/sec (K) ft/ year).
Ground water was encountered generally at a depth of six to twelve feel,
however rising to within a few feet of the ground surface. Trace amounts
of some hazardous compounds had migrated vertically through the site
soils to depths of 20 feet or more. Trace amounts (parts per billion)
were found in the groundwater ai the site to a depth of 75 feet. The
potential for continued vertical and horizontal migration of hazardous
compounds exists.
Pure wastes were classified as "buried waste" and "lagoon waste"
with the river silt (control) for all laboratory and field tests". All were
collected from the site and analyzed prior to waste application as deter-
mined by GC and GC/MS. About 24 polynuclear aromatics (PNAs)
were identified (F-2 fraction) and 22 aliphatic hydrocarbons (F-l frac-
tion)'. Quantitation was by external standard GC in all cases. Some
analyses were semi-quantitative due to problems in obtaining accurate
external standards. Large dilution factors necessary to prevent GC detec-
tor saturation also contributed to variability in waste analysis. Detec-
tion limits for both fractions was 10 ppm due to the targe dilution factors
and the lower response factors of the higher molecular weight com-
ponents. The detection limit for the control soil, however, was I ppm.
Optimal toxicant loading rates, determined in earlier screening tests,
were shown to be acceptable for inducing microbial biotransforma-
lion/biodegradation in laboratory mesocosms and field application plots
with minimal acute leachate toxicity. In the above studies, all compounds
356 BIOREMEDIATldSl
-------
Table 2
Minor Bio-elements, their Sources, and Some of their
Functions in Microorganisms
[Adapted from Bacterial Metabolism (Gottschalk, 1979)]
Element
Source
Function in Metabolism
Zn
Zn2+
Mh Mn2+
Na Na+
Q Cl-
Mo Mo2-4
Se SeO2-3
Co Co2+
Cu Cu2+
W WO2-4
Ni Ni2+
present in alcohol dehydrogenase, alkaline
phosphatase, aldolase, RNA and DNA
polymerase
present in bacterial superoxide dismutase;
cofactor of some enzymes (PEP carboxykinase,
re-citrate synthase)
required by halophilic bacteria
present in nitrate reductase, nitrogenase, and
formate dehydrogenase
present in glycine reductase and
formate dehydrogenase
present in coenzyme Bi2-containing enzymes
(glutamate mutase, methylmalonyl-CoA
mutase)
present in cytochrome oxidase and oxygenases
present in some formate dehydrogenases
present in urease; required for autotrophic
growth of hydrogen-oxidizing bacteria
Table 3
Residual Concentrations of Toxicants in Mesocosms
Mesocosm
ToxicantfLoad*)
Residual*
* expressed as mg/kg dry weight soil (based on GC/MS )
0 expressed in days
Adjusted: correction for acclimation to waste loading
Half-Life0
2.5% O&G
(autochthonous)
4.0% O&G
(autochthonous)
4.0% O&G
(autochthonous)
4.0% O&G
(commercial)
Acenapthalene(20)
Anthracene(97)
Phenanthrene(138)
Acenapthalene(46)
Anthracene(llS)
Phenanthrene( 167)
Acenapthalene(43)
Anthracene(154)
Phenanthrene(142)
Acenapthalene(57)
Anthracene(149)
Phenanthrene(202)
0.2 ( 1.0)
4.2 ( 1.0)
1.3(1.5)
3.6(1.1)
12.1 ( 1.0)
6.6 ( 1.5)
Adjusted
Adjusted
Adjusted
0.4 ( 1.0)
4.6(1.1)
5.5 ( 1.5)
6.94
5.31
4.67
19.72
18.91
14.36
7.14
4.98
5.01
6.47
4.65
4.32
Table 4
Residual Concentrations of Toxicants in
Field Verification Study
Field Plots
Toxicant(Load')
Residual*
Half-Life0
analyzed exhibited decreases in concentration over time for both
laboratory and field tests. The decreases were mostly attributed to
microbial activity by the indigenous soil microflora. However, undefined
abiotic losses were noted and need to be further studied. Both waste
types, the lagoon and buried wastes, at loading rates of 2.5% and 4%,
were degraded by the indigenous microflora. Microtox™ data suggested,
that time periods between sequential reloadings need to be carefully
evaluated and adjusted according to environmental parameters to pre-
vent downward leaching of organic constituents.
Analyses of mesocosm data provided indications of the biotic and
abiotic fectors affecting toxic chemical breakdown in field studies. Com-
parisons of toxicant half-life estimates of targeted waste toxicants in
mesocosm tests and field validation tests is shown in Table 3 and Table 4.
Addition of Commercial Inoculum
In addition to investigations of the biotransformation processes by autochthonous
microflora, it was of special interest to evaluate the use of a commercially available
blend of bacterial cultures. These commercial cultures are marketed for their
known ability to biodegrade polynuclear aromatics. Their application is refer-
red to by the supplier as "bioaugmentation." The inoculum used in the experi-
ment was purchased from Microbe Masters, Inc., Baton Rouge, LA. Three of
the 9 mesocosms were inoculated with the commercial bacterial blend at the
suggested rate of 0.01 lb/ 4.5 kg soil mixture and contained waste at 4% load
plus the inoculum.
The commercial inoculum showed an enhanced degradation rate for these com-
pounds over the first 14 days of the experiment. The rate of degradation was
almost linear for the observed time period. The autochthonous or adapted
mesocosms at the 4% load again exhibited an initial lag phase of biotransfor-
mation, indicative of some microbial acclimation to the waste loading. Minimal
degradation was observed over the first 14 days of the experiment. Following
day 14, however, there was an increase in toxicant degradation rates. Biotransfor-
mation rates, at concentrations at or approaching 40% residual of the original
toxicant addition for both autochthonous and commercial inoculum, were similar.
However, these rates were noted for commercial mesocosms on day 7 and
autochthonous microcosms at day 28. This was directly attributable to relative
viable biomass contributions. Commercial inocula exceeded autochthonous levels
during the first two weeks of the study. Final residual concentrations for both
inocula were similar. Half-life estimates for compound disappearance for all
mesocosms are summarized in Table 3.
At the conclusions of field investigations, noticeable variation in biotransfor-
mation/degradation by the commercially available mutated bacterial cultures over
the autochthonous microflora was evident. Residual levels for 4% O&G loadings
4.0% O&G
(autochthonous)
4.0% O&G
(allochtonous)
Acenapthalene(66)
Anthracene(235)
Phenanthrene(288)
Phytane(131)
Acenapthalene(48)
Anthracene(212)
Phenanthrene(290)
Phytane(144)
~ — j : — TT ., , , :
2.9(1.1)
11.8(1.0)
5.9 ( 1.0)
26.8(1.0)
1.9 ( 1.1)
6.8 ( 1.0)
0.9 ( 1.0)
13.7(1.0)
17.24
14.98
15.01
19.66
9.24
8.98
12.01
10.66
0 expressed in days
are shown in Table 4. Specific toxicants were biotransformed at different rates
and reflected not only loading rates (%O&G) but also inoculum source. Cor-
recting for acclimation by the indigenous microflora at 4% O&G, economic dif-
ferences in microbial populations are then seen. Thus, commercial inocula would
appear to be effective in site remediation from two perspectives: (1) the inocula
used in this study was technically viable in achieving acceptable rates of toxi-
cant biotransformation; (2) although autochthonous (adapted) populations were
equally effective, an acclimation period must be considered for initial waste
loading, i.e., the commercial inocula provides a commercially significant ad-
vantage in kinetics performance.
A decision to proceed with site remediation , using a modified biotreatment
approach, was approved by state and federal environmental agencies. Post-closure
monitoring of soils and leachate collected from the site was recommended for
a time period of 30 years after completion of soil biotreatment.
BIOKINETICS OF PETROLEUM HYDROCARBONS IN
SLUDGES/SOILS IN A LIQUID SOLIDS CONTACT REACTOR
Introduction.
The five-ring polynuclear aromatics and related compounds are known to exist
in many sludges, contaminated soils, and contaminated slurries of materials having
significant hydrocarbon content. Of particular concern to state and federal agencies
are the benzo(a)pyrene, benzo(a)anthracene, and chrysene found in chloroaliphatic
wastes such as creosote waste materials, particularly those materials containing
high oil and grease concentrations'4 To document to EPA biokinetic data
on these and other PNAs of concern, liquid /solids contact (LSC) reac-
BIOREMEDIATION 357
-------
tors, were constructed and placed in an environmentally closed laminar
flow hood and inoculated with acclimated microbial populations
developed in earlier pentachlorophenol studies'1 Specific
measurements included microbial ATP for the determination of
microbial biomass", parent compound disappearance, contribution of
incident UV light on photolytic decomposition processes, and post-
treatment residual determinations of dioxins using GC/MS. Particular
emphasis was on following the degradative pathway of ben/o(a)pyrene
in creosote sludges.
Each unit consisted of a 1,000-ml reaction vessel in which toxicants
or substrates were introduced via peristaltic pump. Temperature was
maintained by a heat lamp system regulated by a proportional indicating
temperature controller. The pH/Eh of the reaction vessel was main-
tained by a series of controllers connected to the peristaltic pumps or
gas regulators. Samples were withdrawn aseptically from the reaction
vessel by means of micropipet or syringe. Samples consisted of aqueous
slurries, grab-sampled from reactor vessels, at periodic intervals. Con-
tents of the agitated reactors were presumed to be homogenous suspen-
sions. However, all calculations of toxicant residuals were determined
on a dry weight basis.
LSC Creosote Studies Roughing Cell Reactor and
Biotreatment Reactor Tests
At the conclusion of initial abiotic and biotic tests on PAH wash waterv
contaminated creosote waste was suspended in LSC reactors over a
seven-day period followed by a 14-day biolreatment test. This seven-
day roughing step provided indications of mixing phenomena for
creosote, predominantly K001 constituents, and also provided indica-
tions of fate of percent ring PNAs associated with these wastes. After
seven days of high energy contact, the supernatant was transferred to
a polishing biological reactor cell where additional biological treatment
was again performed for a 14-day period. Over these time frames. GC
/MS determinations were made of the primary K001 constituents as
well as the chlorinated dioxin and chlorinated furan contaminants.
To identify microbial contributions to PNA degradation, two
approaches were considered. Sterile L/S contact tests were conducted
using the aforementioned laboratory approach. Antibiotics were used
to hinder microbial growth and kinetic response. Comparisons were
made between abiotic and biotic tests for targeted removal.
Biotransformation Of Creosote ttbsie K001 Constituents
Figures 5 and 6 provide information on the residual levels of key
KOOI constituents for all reactors for roughing cell and biological treat-
ment. The roughing step involves the actual resuspension and solubiliza-
lion of creosote and pentachlorophenol materials over a seven-day
period.
LSC Process Treatment: Creosote
•? 20000
21
TIME (DAYS)
Figure 5
KOOI Biolran&formation
This key initial step forces the solubilization of the KOOI constituents
as a result of the addition of surfactant (Triton XlOO, Sigma) and pH
adjustment (to 7.3) resulting in the increased availability of these
materials for biological attack.
The data presented show high concentrations of fluorene, phenan-
threne, and fluoranthrene for initial waste loading. Reactor cell ft\ had
the highest levels of these KOOl constituents in concentrations exceeding
7,000 ppm. Reactors cells tfl - #4 had concentrations approaching 8,000
ppm or less with the exception of phenanthrene, averaging 15000 ppm.
It is important to note that the initial concentrations varied in terms
of chemical content, however, they all represented a 20% loading rate
based on solids for all reactors. With combined microbial addition and
surfactant addition, the residual level for fluorene, phenanthrene,
fluoranthrcne, and pyrene were greatly reduced after seven days of con-
tinuous aeration and agitation.
Residual concentrations from this roughing cell step, which is then
normally transferred to a polishing reactor step, averaged in concen-
tration between 500 and 4.000 ppm with phenanthrene appearing to
be the most resistant to the continuous agitation over a seven day treat-
ment period. In reactor test D, minimal microbial levels were noted
as determined by direct plate counts and microbial ATP estimates. As
a consequence of this, minimal removal levels for all KOOI constituents
were noted. In particular, pyrene and chrysene resulted in negligible
biodcgradation. Phenanthrene was marginally reduced from 13,000 ppm
to approximately 10,000 ppm. Fluoranthrene and fluorene appeared to
be the most significantly reduced of the KOOI constituents.
Reactor H\, having the highest KOOI constituents loading rales, had
the greatest reduction in total hydrocarbon content In particular, pyrene
and crysene were more dramatically reduced in reactor cell K\ as com-
pared to reactor cell #3. High biomass levels were noted in reactor cell
H\. Microbial ATP levels exceeded 10* cells per ml for continuous
I real mem For final biological treatment, phenanthrene and fluorene
were both significantly degraded to below 100 ppm residuals within
21 days. Fluoranthrene and pyrene were reduced to levels below 500
ppm over (he same time frame.
Carcinogenic PAHs
Figure 7 provides information on the initial and final concentrations
of key five-ring polynuclear aromaiics found in creosote/peraa-
chlorophenol waste materials. Of particular interest is the
benzo(b)fluoranthrenc and benzo-(a) pyrene constituents of these wastes.
As in the previous data sets on KOOI constituents, reactors A, 12 and
#4 provided significant reduction in the five-ring polynuclear aromaiics.
Reactor celt tf\ . having the highest accumulated biomass. indicated
the greatest reductions to < 500 ppm for all constituents. Reactor cell
#3 which experienced incomplete mixing showed negligible reduction
in the benzo-(a) pyrene. Note in particular that benzo-(b) fluoranthrcne
was not significantly reduced during this 7-day mixing step. With
biological treatment, notable reductions were seen for all 5-ring PNAs.
However, as was noted previously, benzofb) fluranthrene was not as
significantly reduced as the other PNAv
E
CL
Q.
01
3
M
C
O
o
O
o
LSC Process Treatment: Creosote
12000
10000
7 14
TIME (DAYS)
Figure 6
KDOI Biotransformauom
Summary
Table 5 provides kinetic expressions for LSC biotreatment of highly
concentrated PAHs.
LSC Rationale
Optimal mixing of PAH waste materials such as creosote can result
in significant reductions in KOOI constituents. Reactor cell #3 had
minimal performance in KOOI reductions. This correlated with poor
358 BIOREMEDIATION
-------
LSC Treatment Process: Creosote
E
Q.
Q.
Q.
O>
c
0 7 14
TIME (DAYS)
Figure 7
Biotransformation of Carcinogenic PAHs
TableS
Biokinetic Rates of K001 Reductions in LSC Reactors
KOOl Constituent
Initial Conccntran'on(ppni)
Rate (mg/tg soil/day)
Phenanthrene
Fluorene
Fluoranthrcne
Pyrene
Benzo(a) pyrcne
Benzo(b)nuranthrene
Benzo(a)anthiacene
13,000(mcan)
7.200(mean)
8,000(nrean)
6,000(racan)
9,000(msan)
13.300(mean)
ll,000(mean)
584.8
316.6
367.4
261.9
366.7
595.2
521.4
Mean values are based on 4 replicate LSCs for each experimental and control test and 3 GC sample
analyse for each day sampled.
microbial performance and mixing. Thus, a key component in the ability
to biologically transform these materials rests with the ability to suffi-
ciently mix and suspend by wet weight the creosote materials in ques-
tion. In subsequent investigations not reported here, more optimal
mixing of the KOOl materials was achieved with a reconfiguration of
the reactor cell. This reactor cell includes baffle systems to prevent
settling and incomplete mixing. Rates of KOOl disappearance responded
to this improvement in reactor design.
BIOTRANSFORMATION OF POLYNUCLEAR AROMATICS.
GEOCHEMICAL INFLUENCES
IN REFINERY BIOREMEDIATION
The Requirement For Some Metals By Microorganisms.
Very low concentrations of certain metals are required by all
microorganisms for normal cellular functioning. These include
potassium, magnesium, manganese, calcium, iron, cobalt, copper, zinc,
and molybdenum. For example, copper zinc and molybdenum are con-
stituents of specialized enzymes. Cobalt is found in vitamin B|2 and
its coenzymes. Magnesium, iron, manganese, calcium and potassium
are also enzyme cofactors24. Of these metals, copper, iron, potassium,
and magnesium are required to a greater degree than the others, which
are usually required only in trace amounts. These micro-nutrients are
often, in fact, toxic at high concentrations25- Cadmium, however, has
no known metabolic role.
Metal Toxicity and Microorganisms
High levels of heavy metals in the environment are usually toxic to
microorganisms. Some microorganisms may even be affected by quite
low concentrations of particularly toxic metals. For the overall cell
population toxicity may manifest itself as a drop in cell numbers due
to cell death, bacteriostasis, or extension of the lag phase of the cell
cycle25. If bacteriostasis, or a lengthened lag phase occurs, cell
metabolism is interfered with, but not severely enough to cause cell
death. Heavy metal toxicity may also manifest itself in altered cell
morphology26 The toxic metal is likely to interfere with transport
systems within the cell27 This may be a result of interference with cell
function by protein denaturation28, disruption of enzyme structure, and
disruption of DNA25.
In general, the toxicity of a heavy metal is determined by its degree
of attraction to natural metal binding sites on and within the cell. The
similarity in chemistry of some heavy metals to other elements required
for cellular functioning may result in some being actively accumulated
within the cell. In general, the ability of a toxic metal to penetrate
through to the cell cytoplasm is a significant measure of its potential
toxicity25. However, metal toxicity is mediated by several factors. The
nutritional state of the organism may alter toxicity as cells in a nutrient-
depleted environment are often more susceptible to metal toxicity. En-
vironmental factors heavily influence heavy metal toxicity and some
of these are reviewed below.
Influence of the Environment On Metal Toxicity
The presence of metal-chelating compounds, other ions, and pH of
the environment all affect the toxicity of heavy metals to microorganisms.
Other cations, particularly those of similar ionic radii, can decrease
toxicity due to competition for binding sites25'28'29'30-
Low pH, (i.e., high hydrogen ion concentration) reduces metal
toxicity25'28, probably due to ionic competition between hydrogen ions
and metal ions31. High pH may enhance metal toxicity32 due to low
hydrogen ion concentration leading to less ionic competition, but for
some metals, increase in pH beyond a particular point may lower toxicity
because of precipitation removing metal from solution29-30. Agents
capable of chelation can affect toxicity by binding the metal. For
example, in nature Kaolinite and montarillonite clays can reduce heavy
metal toxicity by binding the metal. Humic, fulvic acids and proteins
can also have the same effect28. The presence of synthetic chelating
agents such as E.D.T. A. have been shown to reduce heavy metal toxicity
toward micro-organisms25.
Resistance To Metal Toxicity By Microorganisms
Microorganisms exposed to adverse environmental conditions may
soon produce strains capable of surviving in a hostile environment
through genetic modification. In many cases the evolved mechanisms
are highly specific. In bacteria this metal resistance is often plasmid-
linked26'33'34 and often associated with antibiotic resistance33'35 Two
general strategies exist for achieving resistance to toxic metals:
• Increase impermeability of the cell to the metal
• Biochemically achieved transformation of the metal.
The former process protects the cell from toxic elements in its environ-
ment. The latter detoxifies the immediate .environment of the cell by
eliminating the toxic metal from it or altering it to a non-toxic
form26'34.
Increased impermeability may be achieved non-specifically by pro-
duction of an outer protective layer around the cell. This allows some
metal to be bound at a distance from the cell wall with little damage
being caused30 This non-specific mechanism appears to be employed
by the bacterium Zooglea ramigera, a common member of sewage sludge
microbiota. Comparison of metal toxicity on strains of Z. ramigera
capable of producing extracellular polysaccharide around the cell with
that of a strain incapable of exopolysaccharide indicated that the former
fared better in metal contaminated solutions and also accumulated more
metal than the latter30. Encapsulated strains of Azobacter have been
found to survive better in lead-rich solutions than non-capsule produc-
ing Micrococeus luteus, due to the former's ability to immobilize lead
without the metal being able to exert toxic effects at the cell surface
or intracellularly. Some periphytic pseudomonads have been found to
take up copper predominantly in their extracellular polymer, with on-
ly a fractional amount actually reaching the cell.
Capsulate strains of Klebsiella aerogenes were found to survive in
10mg/l cadmium better than a strain that did not secrete extracellular
polysaccharide around the cell. Furthermore, when capsular polysac-
charide was separated from polysaccharide producing strains and added
to non-producing strains in cadmium solution, the survival of the latter
was enhanced25. A layer or matrix of extracellular polymer therefore
appears to enhance cell tolerance of toxic metals by immobilizing them
away from the immediate proximity of the cell where they cannot bind
to functional groups on the cell surface or within the cell. It should
BIOREMEDIATION 359
-------
be noted, however, thai extracellular polymer capsules and matrices
may not have evolved specifically to protect bacterial cells from toxic-
metals; they are also known to offer resistance to phagocytosis inges-
tion by amoebae or phagocytes, protect against bacteriophage and
dessication, and might also act as a food reserve1*
More specific resistance mechanisms to toxic metals are known in
which cellular permeability to the metal is decreased Some strains of
Staphylococcus aureus are more resistant to cadmium tox icily than
others, due to an alteration of the specific transport system responsible
for bringing cadmium into the cell. Some Eschtricia call strains arc
cobalt resistant due to a change in the specific uptake system responsi-
ble for translocation of cobalt"
The alternative strategy to increasing cell impermeability is transfor-
mation of a toxic metal into a non-toxic form. This may be achieved
intracellularly, but is more commonly achieved cxtracellularly. Alter-
natively, a toxic metal may be transformed into a form that is in-
assimilable by the microorganism. Toxic metals may be oxidized,
reduced or methylated to produce less toxic compounds. Mercury
resistance is often plasmid-linked via a plasmid-detcrmmed enzyme
which can transform mercury and organo-mercurials into volatile forms
which are soon lost from the environment. Another mechanism for
removing metals from solution is production of hydrogen sulphide by
microorganisms. As most heavy metals form insoluble sulphides, the
production of sulphide by the bacterium Desulphovibrio desulphuricans.
the fungus Poria vaiHantii and some strains of the yeast Saccharomyces
cerevisiae results in precipitation of the metal from solution1* Some
fungi are also capable of producing chelaiing agents which bind metal
away from the cell. Cormllus palustris, among others, can produce
oxalic acid to enhance its copper tolerance by this means'". Thus it can
be seen that many mechanisms exist by which microorganisms may
enhance their tolerance of toxic metals.
Accumulation of Hea\y Metals By Microorganisms
Several mechanisms exist by which microorganisms remove heavy
metals from solution. These may be divided into two general categories:
metabolism dependent uptake into the cell and binding of metal ions
to extracellular material (e.g., capsular polymer), or the cell wall which
is not an active process'6-'"* Some potentially toxic metal ions have
already been previously mentioned to be micronutrients at low con-
centration. Most are divalent meial ions (for example, Alcaligenes
eutrophus exhibits a growth requirement for nickel) and active uptake
systems exist to bind these ions.
These divalent cation uptake systems tend to be particularly specific;
however some do transport metals into the cell apart from those
primarily required. The magnesium uptake system of Ł. colt is suspected
also to accumulate Ni^-CO3* and ZnJ' The Mg;* transport system of
Saccharomyces cerevisiae is known to take up Co-'' Mn!' Zn:* and
W* Generally, ion uptake systems are specific for ions of a certain
ionic radius. Thus monovalent cation uptake mechanisms tend not to
take up divalent metal ions or metal ions of a higher valency, excluding
the toxic heavy metals. However, caesium and radio isotopes of caesium
and T" have been observed to be taken into the cell via the potassium
transport system".
Anion transport systems have also been implicated in carriage of toxic
metals into cells. Metals that exist as oxanions in solution may be ac-
cumulated by such systems. Chromate for example, has been
demonstrated to be competitive with sulphate ions for uptake via the
sulphate permease system of Neurosposa crassa. Many of the cases
of intracellular uptake of toxic metals kncywn are active processes, but
intracellular uptake of toxic metal by non-viable cells is also known
to occur"
The term 'biosorption' has been coined to describe the non-active
adsorption of heavy metal ions by microorganisms or biological
polymers. This process has been defined by Shumate and Strandberg"
as "the non-directed, physical-chemical complexation reaction between
dissolved metal species and charged cellular components, akin in many
respects to ion exchange. Such processes usually occur us interactions
between negatively charged ligands and metal ions and may occur as
ion-exchange or formation of complexes. The most likely components
of microbial polymers capable of ion exchange are carboxyl groups,
organic phosphate groups and organic sulphate groups. Chelation or
complex formation tends, to occur on biopolymers where neuiral divalent
oxygen, sulphur atoms, or trivalent nitrogen atoms are present. Examples
include amino- and heterocyclic nitrogen groups of proteins and nucleic
acids and also the carbonyl and hydroxyl oxygens of the same polymers.
The latter two groups are also found in polysaccharides,
polyheterocyclics and polyphenics". As previously mentioned,
extracellular polymers have been demonstrated to bind heavy metals,
such as the binding of metal to the extracellular polymers produced
by the bacteria Z. ramigera and K, aerogenes. Extracellular accumula-
tion of metals has also been demonstrated to occur with the extracellular
polysaccharides of the algae Mesotaenium kramstei and Mesotaenium
caidariorunf.
Accumulation of metal at or within the cell surface has been observed
to occur with many microorganisms. The bacteria Bacillus subrilu".
Bacillus lichenfnnni.t". and Escherichia coll" have been demonstrated
to bind heavy metal ions to their cell surfaces. Among the fungi Sac-
charomyces cerevisiae1', Neocosmospora vasinfecta", Rhiiopus
arrhizius (Tsczos and \folesky. 1981), Neurospora crassa and a
Penicillium species'1 have all bound metal to their cell walls. Con-
siderable diversity exists between the cell wall composition of bacteria
and fungi, yet all apparently contain groups capable of metal binding.
Bevcridge and Murray" and Doyle et. al." have identified the
predominant divalent metal ion binding group in Bacillus subtilis cell
walls as the glucamic acid carboxyl groups of the wall peptidoglycan.
Bevcridge and Koval41 proposed that the polar heads of the cell
envelope phospholipids of Ł. colt were primarily responsible for its
metal binding. For Bacillus licnenformis the predominant metal bin-
ding sites in the cell wall have been shown to be the tcchoic acids".
For Rhitopus arrhizius the chitm of the cell wall has been implicated
in uranium binding'" and thorium binding* Accumulation of metals
by microorganisms is widespread and occurs by a variety of mechanisms.
CONCLUSIONS
Biokineiic Data Bases
Microorganisms, whether in a constructed remediation cell or deep
ocean environment, are constantly laced with fluctuating environmen-
tal conditions. Tidal action, upwellings. storms, and solar radiation cause
changes in salinity, temperature. pH, and oxygenation. they can also
transport microbes to new environments. The most important parameter
is the availability and quality of nutrients. The majority of
microorganisms in stressed soil/sludge micro-environments are
oligotrophic, and their inhabitants must cope with the uncertain, and
often unsuitable, conditions for survival. The complexity and diversity
of these microorganisms, and of their environment, makes it virtually
impossible lor optimal conditions to exist for each organism. Therefore,
at any one time, most microbes are surrounded by waters lacking suf-
ficient energy-yielding substrates (nutrients from which ATP can be
produced). Without energy, can viability be maintained? The concept
of "starvation survival" deals with this particular situation, and the adap-
tations that organisms have evolved to deal with the problem.
This discussion has concentrated on heterotrophic microbial popula-
tions. Biodegradable organics are the "energy-yielding substrates" of
these genera When water or allochlhonous forces deposit a bacteria
in an area of nutrient deficiency, the organism must adapt or perish.
Inherent in any living thing is the necessity for the continuation of the
species. Microorganisms enter a transient state of dormancy until ex-
ternal conditions improve. Apparently, bacteria are very "patient" and
can maintain this state for many, many years. Starvation survival dor-
mancy is a physiologically complex occurrence.
Hach species has a characteristic threshold for utilizing nutrients. The
threshold may be lower in organisms that have a high affinity, and low
specificity for nutrient uptake. Below this threshold concentration, the
organism is unable to grow, and reproduce: it must take drastic action
to remain viable itself. Onset of the starvation survival condition is often
characterized by division of the bacteria, without concurrent growth,
to produce ultramierocells. In \
-------
been observed up to 400% as a result of introduction into a starvation
media. Miniaturization results in a larger surface to volume ratio, which
is an advantage in scavaging; the increase in cell numbers increases
the probability of survival of the species. Upon encountering an area
with utilizable nutrients, the ultramicrocells will resume normal size
proportions, indicating that dormancy is reversible, and is a function
of the availability/concentration of suitable energy yielding substrates.
Microbial uptake of nutrients is a competitive process, so dormancy
includes several mechanisms, that operate primarily under low nutrient
stress, to increase the inherent ability to compete. The capacity of an
organism to find an essential nutrient, "capture" it, and then bring it
into the cell is especially important in oligotrophic environments. The
oligotrophic organism must be able capture the substrate and then hold
it on its surface long enough for active transport to occur across the
cell membrane. Periplasmic binding proteins are the structural entity
that perform this task. It appears that some binding proteins also func-
tion as chemoreceptors. This chemotactic ability further increases the
efficiency of a bacterium in its search for energetically rich substrates.
Lab experiments have shown that there is an "optimum chemotactic
period." It is possible that if the organism has not been successful at
the end of that period, it may then enter its state of dormancy. During
dormancy, endogenous metabolism is reduced virtually to zero, although
laboratory studies have demonstrated that dormant bacteria maintain
high levels of RNA and amino acids, and a high energy charge. In vitro,
this build-up begins shortly after introduction to low nutrient condi-
tions. Although energetically expensive, this allows the organism to im-
mediately (and efficiently) utilize a nutrient when it becomes available.
Microbes can be considered as living catalysts. Technically, a catalyst
is a substance whose presence alters the velocity at which a reaction
proceeds; a catalyst can be recovered unaltered at the completion of
the reaction. Microbes often cannot be recovered from the reactions
in which they participate (much less do they remain unaltered). Bacteria
function to convert DOC to POC for higher order consumers. Other
microorganisms produce organic metabolites which serve as food for
other organisms. However, they themselves are eventually consumed,
as are the bacteria, by organisms of the next trophic level. Strictly from
a biomass perspective, the catalyst could be recoverable at each in-
termediate. Therefore, as the base of the food web, microbes do facilitate
the flow of organics through the system. By increasing the efficiency,
microbes affect the rate of carbon cycling through the ecosystem. In
this way they do serve a catalytic function, but are not "catalysts" in
the strictest sense of the word. But as catalysts, they are the central
focus in biotreatment effect in a bioremediation system.
REFERENCES
1. Pfaender Frederic K., and G.W. Bartholomev. 1982 Measurement of Aquatic
Biodegradation Rates by Determining Heterotrophic Uptake of Radiolabeled
Pollutants. Appl. Environ. Mlcrobiol, 44 (1):159-164.
2. Chibata I, T. Tosa, and T. Sato 1974. Immobilized aspartase-containing
microbial cells: preparation and enzymatic properties. Appl. Microbiol,
27:878-885.
3. Walter William G., R.A. McBee, and K.L. Temple. 1973. Introduction to
Microbiology. Litton Educational Publishing, Inc. Pp. 1-7, 59-71, and 121-141.
4. Wilkinson J.F. 1975. Introduction to Microbiology. Vol. 1. Second Edition.
Blackwell Scientific Publications. Pp. 11-91.
5. Poindexter IS. 1971. Microbiology an Introduction to Protists. Pp. 23-66,
and 149-199.
6. Walter William G, and R.H. McBee. 1962. General Microbiology. Second
Edition. D. Van Nostrand Company, Inc. Pp. 27-114, and 145-180.
7. Nester Eugene W, C.E. Roberts, N.N. Pearsall, and B.J. McCarthy. 1973.
Microbiology. Second Edition. Holt, Rinehart and Winston. Pp. 237-274.
8. Ucko David A. 1977. Living Chemistry. Storage and transfer of energy.
Academic Press, Inc. Pp. 388-419.
9. Caldwell Douglas E, and John R. Lawrence. 1986. "Growth kinetics of
Pseudomonas fluorescens microcolonies within the hydrodynamic boundary
layers of surface microenvironments." Microb. Ecol., 12 : 299-312.
10. Volk Wesley A, M.F. Wheeler. 1973. Microbiology. Third Edition. J. B. Lip-
pincott Company. Pp. 3-129.
11. Neff, Jerry M. 1979. "Polycyclic Aromatic Hydrocarbons in the Aquatic
Environment." Sources, Fates and Biological Effects. Applied Science
Publishers LTD, London. 61-73 and 102-149.
12. Sisler, F.D. and C.E. ZoBell. 1947. "Microbial Utilization of Carcinogenic
Hydrocarbons." Science, Nov. 28, 1947. Pp. 521-522.
13. Cerniglia, C.E. 1984. "Microbial Transformation of Aromatic Hydrocar-
bons, in: Petroleum Microbiology." edited by R.M. Atlas. Macmillan
Publishing Company N.Y., London. 99-129
14. Alexander, M. 1977. Introduction to Soil Microbiology. John Wiley & Sons.
N.Y. Chichester, Brisbane, Toronto. 115-225
15. Gibson, D.T. "Microbial Metabolism." in: Hutzinger, O.1980. The Hand-
book ofEnviro. Chem., Vol. 2 Part A, Springer-Verlag Berlin Heidelberg
New York Tokyo 1980. pp. 161-189 and 231-244.
16. Zander, M. 1980. "Polycyclic Aromatic and Heteroaromatic Hydrocarbons."
in: Hutzinger, O.1980. The Handbook of Enviro. Chem.,Vol.3 Part A,
Springer-Verlag Berlin Heidelberg New York Tokyo 1980. pp. 109-128.
17. Alexander, M. 1981. "Biodegradation of Chemicals of Environmental Con-
cern." Science, Vol. 211, 9. Jan. 1981. pp. 132-139
18. Mandelstam, J. and K. McQuillen 1982. Biochemistry of Bacterial Growth,
Third edition, Blackwell Scientific Publications, Oxford, London, Edin-
burgh, pp. 142-158
19. Hutzinger, O. 1985. The Handbook OfEnviro. Chem., Volume 2, Part C,
Springer Verlag, pp 117-147.
20. Casarett and Doull, J. 1980. Toxicology. The Basic Science of Poisons. Sec.
edition, edited by J. Doull; C.D. Klaassen and M.O. Amdur. Macmillan
Publishing Co, Inc. New York. pp. 91-93 and 632-661.
21. Harvey, R.G. 1985. "Synthesis of Dihydrodiol and Diol Epoxide Metabolites
of Carcinogenic Polycyclic Hydrocarbons." in: Harvey, R.G. Polycyclic
Hydrocarbons and Carcinogenesis. 1985. ACS Symposium Series 283. 19-35.
22. Cerniglia, C.E. 1983. "Microbial Metabolism of 4-, 7-, 10-
Methylbenz[a]anthracenes." in: Cooke, Marcus and Dennis, Anthony J. 1983.
Polynuclear Aromatic Hydrocarbons: Formation, Metabolism and Measure-
ment. Battelle Press, Columbus, Richland, pp. 283-293
23. London, S.A., C.R. Mantel and J.D. Robinson 1984. "Microbial Growth
Effects of Petroleum and Shale-derived Fuels." Bull. Environm. Contam.
Toxicol. 1984. 32: pp. 602-612. Springer Verlag New York Inc.
24. Strandberg, G.W.; Shumate, S.E.; Parrott, J.R. (1981) Appl. Environ.
Microbiol. 41:237.
25. Bitton, G.; Freihofer, V. (1978) Microb. Ecol. 4:119-125.
26. Ehrlich, H.L. (1978) In Microbial Life in Extreme Environments. Kushner,
D.J. (ed):p.381-408. Publ. Academic Press.
27. Maxwell, W.A.; Metzler, R.; Spoerl, E. (1971) J. Bacterial. 105:1205-1206.
28. Gadd, G.M., Griffiths, A.J. (1978) Micro. Ecol., 4:303-317.
29. Tuovinen, O.H., Kelly, D.P. (1974a) Arch. Microbiol., 95:153-164.
30. Tuovinen, O.H.; Kelly, D.P. (1974b) Arch. Microbiol., 95:165-180.
31. Friis, N.; Myers-Keith, P. (1986) Biotech. Bioeng., 28:21-28.
32. Babich, H.; Stotzky, G. (Wl)Appl. Environ. Microbiol., 33(3): .681-695.
33. Hardy, K. (1983) In: Bacterial Plasmids. p62-3, 71-2. Publ. Nelson.
34. Iverson, W.P.; Brinckman, F.E. (1978) In: Water Pollution Microbiology
pp.201-232. Mitchell, R. (ed) Publ. Wiley & Sons.
35. Nakahara, H.; Tomoaki, I.; Sarai, Y; Kondo, I; Mitsuhasi, S. (1977) Nature,
266:165-167.
36. Wilkinson, J.F. (1957) Bact. Rev. 22:46-72.
37. Shumate, S.E.; Strandberg, G.W. (1985) In: Comprehensive Biotechnology,
4:Ch. 13 p. 235-. Robinson, C.W.; Howell, J.A. (eds).
38. Kelly, D.P.; Norris, PR.; Brierly, C.L. (1979) In: Microbial Technology:
Current State, Future Prospects.
39. Hunt, S. (1986) In: Immobilization of ions by bio-sorption. p.11-15. Eccles,
H.; Hunt, S. (eds).
40. Mangi, J.I.; Schumacher, G.J. (1979) Am. Midi., Nature, 702:134-139.
41. Beveridge, T.J.; Murray, R.G.E. (1976) J. Bacterial. 727:1502-1518.
42. Gadd, G.M., (1986) In: Immobilization of ions by bio-sorption, p.149-159.
Eccles, H.; Hunt, S. (eds).
43. Beveridge, T.J.; Koval, S.F. (1981) Appl. Enviom. Microbiol., 42(2):325-335.
44. Beveridge, T.J.; Forsberg, C.W.; Doyle, R.J. (1982) J. Bacterial
/50:1438-1448.
45. Tsezos, M.; Volesky, B. (1981) Biotech. Bioeng., 23:583-587.
46. Corpe, W.A. (1975) Dev. Ind. Microbiol., 16:249-255.
BIOREMEDIATION 361
-------
TOPIC 2: Land Treatment Case Study
Biological Detoxification of a RCRA Surface Impoundment Sludge
Using Land Treatment Methods
John A. Christiansen
Tracey Koenig
Susan Laborde
Duane Fruge
Environmental Remediation, Inc.
Baton Rouge, Louisiana
BACKGROUND
Remediation of contaminated soil presents a major challenge to bus-
iness, scientists, and regulators. Remediation of solid and hazardous
waste sites containing soils were first required under the Resource Con-
servation and Recovery Act of 1976 (RCRA) and the Comprehensive
Environmental Response. Compensation, and Liability Act of 1980
(CERCLA). The authority of regulators to order cleanups preferen-
tially through permanent, on-sile remedies was established under the
Hazardous and Solid Waste Amendments of 1987 (HSWA) and the Su-
perfund Amendments and Reauthorization Act of 1986 (SARA). SARA
in subparagraph 121 (l)b says in principal part:
"Remedial actions in which treatment permanently and signifi-
cantly reduces the volume, toxicity or mobility of the hazardous sub-
stances, pollutants, and contaminants as a principal element, arc
to be preferred over remedial actions not involving such treatment.
The off-site transport and disposal of hazardous substances or con-
taminated materials without such treatment should be the least la-
vored alternative remedial action where practicable treatment
technologies are available. ."
This regulatory authority has laid the groundword for allowing bi-
oremediation and thermaJ treatments to be considered as permanent
remedies.
BIOLOGICAL TREATMENT FOR SOILS
A number of biological remediation technologies have been used in
demonstration or in full-scale on contaminated soils. Those technolo-
gies include:
• Land Treatment—or soil reactors
• Compositing—a type of land treatment system using well-mixed pile
material with chemical, nutrient, and biochemical amendments. The
piles may be aerated to enhance degradation rales.
• In-Situ Treatment -or initiation of biological action in the subsur-
face environment.
• Liquid Solids Contact (LSC)—A methodology using high energy,
suspended growth reactors capable of 10-20 percent solids suspen-
sion to treat the organics in contaminated soil This process resem-
bles a batch-activated sludge process.
• Biological Soils Conditioner (BSQ- a variant of the LSC pnvess
where the solids concentration is as high as 50% dry weight solids
LAND TREATMENT SURFACE IMPOUNDMENT CLOSURE
A Fortune 500 chemical company located in Plaquemine, LA. had
a 30,000-cubic yard surface impoundment called the "North-South
Pond." The impoundment was 180 feet by 250 feet by 25 feet deep and
contained sludge and soil arising from previous spill cleanups, includ-
ing a rail tank car spill cleanup. The impoundment was classified as
a RCRA facility. The principal chemical constituents shown in Table
I were aromatic chemicals. Table I depicts the contaminants and range
of concentrations.
Table I
Constituents of the RCRA Chemical Contamination
Chemical
Range of Contaminant
Concentration rug/kg
Phenol
Cumene
Acciophcnonc
Benzene
Benzyl Alcohol
Tan
Vinyl Chloride
Siyrene
10-3600
<1~48
500-2500
-------
NORTH-SOUTH
POND WASTE SITE
LEAK
DETECTION
SYSTEM
Figure 1
. AUXILLARY LOADING
RAMP (TYPICAL)
BOTTOM OF BCU = 380'
Table 2
Land Treatment Design
Volume Initial Concentration
Lift # c.y. T. Phenol1, mg/kg
Target Concentration
T. Phenol, mg/kg
Treatment Time
(Design) days
7,000
4,300
5,300
-500
-200
-200
TOTAL 16,600
90
75
75
240
'Total phenol was used as a design and QA/AC parameter. TCLP was used as guidance for
post closure care levels.
were used to control emission. During Lift One, five air monitoring
stations were set up. One was inside the ECU at the decontamination
pad, while four were outside the ECU. None of the units outside the
ECU showed detectable hydrocarbons of interest above background.
RESULTS
The replicates from the BCU Lift One initial treatment averaged
mg/kg phenol on Day 0. A hot spot was found in Quad 1, which had
phenol concentrations of 4,000 mg/kg phenol. This spot was diluted
by spreading it throughout the BCU to allow biodegradation. After regu-
lar addition of bacterial product and tilling of the soils, treatment tar-
get levels of < 1 mg/kg was achieved by Day 71. Phenol values of < 10
mg/kg in all quadrants were used as a guide for reloading events. Lift
One was accepted as complete by all parties on Day 71 of the closure.
Table III depicts the treatment results. The mean value of phenol dur-
ing the closure of Lifts One and Two is shown in Figure 3. Figure 4
depicts quadrant phenol levels versus time.
Table 3
Treatment Results (Actual)
Volume Initial Concentration
Lift # cu. yds T. Phenol, mg/kg
Final Concentration
T. Phenol, mg/kg
Treatment
Time, days
7,200
4,300
5,300
3,600
2,600
137
19
15
20
38
TOTAL 23,000
71
28
28
30
28
185
BIODEGRADATION CLOSURE UNIT, GRID
PLAN VIEW
80 mil HOPE LINER
WASTE
APPLICATION
CROSS SECTION DETAIL
ENLARGED FOR CLARITY
Figure 2
Biodegradation Closure Unit
hydrocarbons consisted of total phenol and a GC/MS analysis of Ta-
ble 1 constituents at lift beginning and end.
AIR MONITORING AND EMISSION CONTROL
During the field demonstration phase, air monitoring showed low
potential for hydrocarbon release outside the BCU. During full-scale
remedial design, all measures to prevent air emissions were considered,
Eventually high ring levees, moisture control covers, and masking agents
200
First
Bioremediation of Phenol
in Contaminated Soil
5/24/88
11/9/88 ~ 12/15/88
Treatment Time
Figure 3
Bioremediation of Phenol in Contaminated Soil
BIOREMEDIATION 363
-------
TiMlnwnt Time, day*
Figure 4
Soil Phenol Concentration VS Time
Due to excessive rain in the BCD area, reloading with 4.300 cubic
yards of material for Lift Two was not completed until Day 135. This
8-inch lift was mixed with up to 6 inches of the previous lift to allow
the microorganism population grown on Lift One to metabolize new
substrates. During Lift One, ATP and microoramsm concentration had
markedly increased as shown in Table IV.
Table 4
Microbial Activity
Estimated Organisms Soil Bioremedialion (Lift One)
Day
Microbes/gram Soil
(Mean Values)
0
8
15
36
43
57
64
2.4 x 103
1.3 x 103
5.0 x 106
4.0 x 106
4.9 x 106
6.3 x 10«
5.6X106
This data is based on individual quadrant microbial concentration depict-
ed in Figure 5. The relationship of the phenolic substrate to mean
microbial concentration is shown in Figure 6. Additional bacterial
product was added to enhance the exisimg microbial population. A com-
mercial microbial culture, Micro Pro "Cec." was used on Day 8 us
inoculating seed.
As a result of the mixing of the lifts, Lift Two had an initial phenol
concentration of 24.3 mg/kg. On Day 147 or 14 days after the Lift Two
application, all quadrants exhibited phenol values of
-------
TOPIC 3: Liquid/Solids Contact Case Study
J. Christiansen, RE.
T. Koenig, M.S.
George Lucas
Environmental Remediation, Inc.
Baton Rouge, Louisiana
A major petrochemical company operates a refinery and an olefins
plant in Houston, Texas. The refinery and petrochemical plant combined
to form the fifth largest chemical complex in the continental United
States. The refinery crude oil capacity exceeds 265,000 barrels per day.
The olefin plant contains two surface impoundments on site which are
part of the complex's NPDES permitted treatment facility. The
impoundment serves as wastewater surge capacity immediately after
an API separator. The two impoundments are depicted in Figure 1. OP1
had an initial sludge volume of nearly 4000 cubic yards while OP2 had
an initial sludge volume of 2600 cubic yards. The sludge was classified
by the refinery/olefins complex environmental staff and found to be
nonhazardous solid waste. An analyses of the sludges is shown in
Table 1. During chemical plant turnaround, a project was initiated to
clean the surface impoundments for future use. The alternatives
presented to the refinery complex were:
Table 1
Analyses of Olefin Sludges, Selected Parameters
T—
•* 99 ~-
25 HP
AERATOR
©
©
25 HP
AERATOR
©
15 HP
MIXERS
OLEFIN PIT NO. 1
©
25 HP
AERATOR
©
15 HP
MIXERS
25 HP
AERATOR
©
> WINDOW
H
SEPARATOR *" ^NX^
1(
i
JX.
OUTLET (9
DIVERSION CURTAIN
(99 FEET WIDE, 6 FEET WIDE)
NOTE: AERATORS AND MIXERS WILL MOVE TO PIT NO. 2 WHEN
COMPLETED WITH PIT NO. 1
Figure 1
Olefms Pit No. 1-Two Typical
Parameter
pH
Oil and Grease
Benzene
Toluene
Xylene
Ethylbenzene
Napthalene
Concentration
mg/kg
OP1
6.5
32.5%
15
4
0.93
17
360
OP 2
6.3
41.4%
28
5
17.30
33.4
448
Method
90711
82401
82401
82401
82401
82701
Phenanthrene,
Anthracene
Moisture
Solids
Ash
860
66%
34%
410
68%
32%
82701
209A2
209A2
209D2
iuSEPA SW-846, Third Edition
2Standard Methods for the Examination of Water and Wastewater, 16th
Edition.
• Sludge dewatering by belt or plate-and-frame press followed by off-site
disposal of solid waste,
• Sludge removal and stabilization followed by off-site disposal (without
volume reduction), and
• Sludge volume reduction through Liquid/Solids Contact
Bioremediation processes.
The third alternative was selected as a form of remediation which
offered volume reduction in place at competitive costs, while ensuring
that the residual would be low in objectionable organics. The residual
was to be removed from the impoundment at the end of biological
treatment and placed on the refinery complex's existing land treatment
facility. In order to meet the refinery's requirements regarding volume
of material applied to this land treatment facility each year, a minimum
sludge volume reduction of 50* was required.
PRELIMINARY LABORATORY ASSESSMENT AND STUDIES
An initial assessment of the sludge was conducted using composite
sampling techniques. A survey crew with a boat divided the pond into
grids and pulled sludge samples from at least 8 locations within each
surface impoundment. Samples were taken by inserting a 6-inch PVC
casing and pumping out free liquid. A 2-inch-diameter PVC pipe was
then inserted into the sludge portion and used to take a vertical section
of sludge. This step was repeated until all 8 locations within each pond
had been sampled. The sludge was composited in a 5-gallon bucket
BIOREMEDIATION 365
-------
and mixed with a paint mixer. The analysis of the constituents is shown
in Table 1. The waste contained low concentrations of volatile
hydrocarbons and higher concentrations of semi-volatile base neutrals.
Oil and grease ranged from 32 to 42 %. As free liquid disposal to the
wastewater plant and solids residuals to disposal to the land farm were
controlled by oil and grease. This was used as a target hydrocarbon
for the treatment program.
A Liquids/Solids Contact (LSC) simulation reactor, was set up to
determine feasibility of sludge reduction. Oil and grease samples were
also taken periodically throughout a study which lasted 14 days. The
study was run with indigenous as well as commercially available
microbial products in replicate. The study was run for 17 days and a
vigorous bacterial population was established with indicator protozoa
appearing in both reactors within 10 days. Sludge volume in the
augmented reactors was reduced 50% and oil and grease reduced 60%
by mass. Based on acceptable reduction of sludge and mass, a target
reduction of 50% volume reduction and 60% mass oil and grease
reduction was set for the performance portion of the project. Treatment
then proceeded to the field.
LIQUID/SOLIDS CONTACT REACTOR DESIGN
Each existing impoundment was set up as an in-situ Liquid Solids
Contact reactor. The reactor was designed to suspend sludge in liquid
in a 1:1 (v/v) ratio. In each impoundment, five 25-horsepower surface
aerators (modified to pump 14,000 gpm) and a 15-horsepower,
direct-drive floating mixer were placed to supply mixing and aeration.
The units were energized through a local power system controlled on-siie
by a field operator. The unit was energized in OP1 on October 1.
1988,and on OP2 on November 2. 1988. After 24 hours of mixing, liquor
samples were taken to ensure solids were suspended at at least 15%
dry weight solids. A chemical amendment consisting of surface active
agents, pH control chemicals, macro and trace nutrient amendments.
and an adapted microbial culture (Micro Pro Super "Cee") were added
to enhance microbial degradation. Sludge and liquid depth were
measured weekly throughout the impoundment. Composite samples
were analyzed for oil and grease content using EPA Method 9071.
Composite samples taken on a weekly basis from the sludge were
analyzed for oil and grease concentration in a similar manner. Mixed
liquor control samples were also taken weekly. These were analyzed
for pH, total Kjeldahl nitrogen, total phosphorous, adenosine
iriphosphate, and COD. Samples were also settled to determine
supernatant COD because ultimate disposal of free liquid to the
wastewater treatment plant would require a COD of less than 450 mg/1
and an oil and grease of less than 100 mg/1.
AIR MONITORING AND PERSONNEL SAFETY
Personnel at the site were trained in accordance with OSHA 29 CFR
1910.120 and outfitted in minimum level C personal protective
equipment. Air monitoring was provided in the vicinity of each pond
during start-up and on a daily basis during the first week of operation.
As part of the written health and safety plan contained at the site,
measurements exceeding 0.5 ug/n' resulted in the operator shifting to
level B personal protective equipment or breathing air. It was found
that breathing air was adopted during the first week of each treatment
operation when fugitive benzene emissions were at their height. During
this period of time, benzene measurements taken at the top of the reactor
levee measured as high as 2 ppm benzene. An exclusion zone established
at the bottom of the corresponding levee was another measurement site.
Benzene was not measured at any location outside the exclusion zone
during the entire treatment process.
MIXED LIQUOR AND SLUDGE SAMPLING
Mixed liquor and sludge were sampled on a weekly basis throughout
the project. To do this the reactor was de-energized and a crew (equipped
in level B personal protective equipment) entered the area with a boat.
One operator in the boat took level measurements. These were taken
using 1-inch-diameter PVC pipe marked off in 1-foot and 1/2-foot
increments. At the end of this pipe was a 12-inch square plate which
was coupled to the PVC pipe. The plate was thrust down and the operator
probed for resistance, first to the settled sludge layer, then to the hard
clay soils at the bottom of the impoundment. The operator doing the
work signaled another operator and engineer who recorded
measurements of liquid and sludge depth. After measuring sludge depth
of 8 stations throughout the impoundment, the operator returned and
took his sludge samples in accordance with the procedure described
previously These sludge samples were labeled and retained for analysis
with full chain-of-custody procedures. Preservation, transportation, and
analytical methods were in accordance with USEPA SW 846. The
reactor was re-energized and the crew took 4 mixed liquor samples
approximately 15 minutes later. The 4 mixed liquor samples were then
combined to perform a single liquor composite. The liquor was analyzed
for the parameters to determine microbial population (adenosine
triphosphatc (ATP) and nutrients). Sludge was analyzed for oil and
grease, acid extractable and base neutral compounds (Method 8270),
and moisture solids and ash.
RESULTS
Table 2 provides the sample dates and a summary of analysis and
calculations from settled sludge and the final supernatant sample in
in OP1. Initial oil and grease was 32.5% on 11/2/88. The final sludge
samples taken on 11/22 /88 showed a settled sludge oil and grease of
36.3%. The measurable sludge at that time was 1288 cubic yards or
a volume reduction of 68%. This met the performance standard of at
least 50% volume reduction. Table 3 shows the mass balance calculated
for OP1. This was calculated by taking the initial sludge volume and
multiplying it by the dry weight oil and grease to derive the mass of
oil and grease in the sludge on a dry weight basis. This amount was
tracked throughout the 21 operating days until adequate volume reduction
was achieved. On the last day, 11/22/88, samples of both mixed liquor
and sludge were analyzed to allow closing of the mass balance of oil
and grease. These values were added together to produce a total mass
on Day 21 of 299, 372 pounds oil and grease dry weight, or a 62%
removal over the 21-day period. Figure 2 depicts this mass removal for
OP1 sludges.
Tabfe2
LyondeU fVlrotaim Data OP1
Oil and G
Oat* O«y C*apl«
11/07* J«U *J Sl**»*
11/01 **tt XI *l«*j*
11/U I J«il *rt *1<*4«*
11/22 I t«tt Mt 11«*3*
11/72 -' »up« Mtuc
t-f mi ut>
lb>v diy vt ~ ((«.; Ji-j wt t O-*^
(%1 4«) (f/cf) (e« !•*•> Ill
« 5' i • •' Jtf; »otns
U » ' 1 I M«V U3CID
*1 ' t • !»»•' ICISi^
11 ) * i UM ?«UftI
c » it . « '. vv»).. n«
it * ••
Oat*
Day
TaMe3
OR Mass Balance
M»«a O « G lb», dry wt.
Sample
11/02
11/08
11/18
11/22
11/22
0
7
11
21
21
808,875
783,080
459,567
291,607
7.765
Settled Sludge
Settled Sludge
Settled Sludge
Settled Sludge
Supernatant
Total Mass Day 0
Total Mass Day 21
Removal
Time
- 808,875
- 299, 372 (291,607 + 7,765)
- 62%
•21 days
The average temperature during treatment was 18°C
The volume of OF1 was 1.33 million gallons
366 BIOREMEDIATION
-------
21
1
e-
SUPERNATANT
Figure 2
OP#1 Mass Balance
In a similar manner, calculations shown in Tables 4 and 5 depict the
mass balance calculation for OP2. Figure 3 depicts the mass oil and
grease reduction. OP2 was run a much longer period of time because
of treatment initiation late in the year. Actual treatment of OP1 operating
temperature for mixed liquor averaged 18 °C. Oil and grease was
calculated to have a half-life of 16 days based on the field data. During
OP2, average operating temperature was 14 °C or much lower. This
resulted in extended oil and grease degradation. The final mass balance
shows an 85% mass reduction of oil and grease in 61 days at 14 °C.
This is consistent with an oil and grease half-life of 40 days for OP2,
which can be converted to a 32-day half-life at 20 °C.
Table 4
Lyondell Petroleum Data OP2 - Oil and Grease
Figure 3
OP#2 Mass Balance
The final material which was disposed in the complex land farm was
characterized for parameters equivalent to those shown in Table 6. As
is indicated in Table 6, these parameters show the volatile hydrocarbons
to be stripped or biodegraded during the treatment process. It is
interesting to note that napthalene, phenanthrene, and anthracene
(significant base-neutral compounds) did not significantly increase in
the reduced-volume residual left over from the treatment process. This
indicates significant reduction of those hydrocarbons above the amount
identified in the volume reduction.
Table 6
Chemical Characteristics of Sludge Residual
Day Sanple
Moisture
(*)
Volume
(cu.ydo)
12/021
12/19
12/28
01/12
01/25
01/31
01/31
1Assumed
Date
12/02
12/19
12/28
01/12
01/25
01/31
01/31
0 Se tied Sludge 41.4 62
18 Se tied Sludge 49.0 29
27 Se tied Sludge 64.0 76
43 Se tied Sludge 56.0 551
56 Se tied Sludge 28.1 55
61 Se tied Slue
61 Supernatant
, not recorded
Day
0
18
27
43
56
61
61
ge 18.6 65
0.67 99.8
Table 5
OP1 Mass Balance
Mass O Ł G Ibs, dry
1,058,897
952,368
726,456
205,437
330,599
121,556
12,070
2590 1,058,897
1491 952,368
2576 726,456
1762 205,437
1424 330,599
1017 121,556
1.08 mgal 12,070
wt . Sample
Settled Sludge
Settled Sludge
Settled Sludge
Settled Sludge
Settled Sludge
Settled Sludge
Supernatant
Total 0 & G Mass Day 0 = 1,058,897
Total 0 & G Mass Day 61 = 133,626 Ibs (121,556 + 12,070)
Removal = 87.3%
Time = 61 days
The average temperature during treatment was 14°C
The supernatantvolume of OP2 was 1.08 million gallons on Day 61
Parameter
PH
Oil and Grease
Benzene
Toluene
Xylene
Ethlybenzene
Napthalene
Phenanthrene , Anthracene
Moisture
Solids
Ash
OP1
Residual
6.9
36.3
<0.1
<0.1
<0.1
1.2
423
620
65
35
13
OP 2
Residual
6.6
18.6
<0.1
<0.1
<0.1
6.1
117
406
65
35
18
SUMMARY
A Liquids/Solids Contact reaction technology was used to reduce
sludge volumes and oil and grease content in two wastewater treatment
lagoons at a major olefins refinery outside of Houston, Texas. In OP1,
a degradation time of 21 days was required to achieve 68% volume
reduction and 62% mass oil and grease reduction at an operating
temperature of 18 °C. In OP2, a treatment time of 61 days was required
to achieve 61% sludge volume reduction and 87.3% mass oil and grease
reduction in a lagoon containing 2590 cubic yards operating at 14 °C.
For sludges which have similar biodegradable characteristics, this
offers a major alternative to standard dewatering practices such as
plate-and-frame press. Selection of a method of treatment for individual
sludges should be based on site or laboratory treatability studies
conducted to account for losses from volatilization, absorption, and
other nonbiodegradable sources.
BIOREMEDIATION 367
-------
TOPIC 4: Modular Bioreactor Approaches For
Remediation Of Groundwater:
A Case Study With Volatile Chlorinated Aliphatics
David D. Friday, M.S., P.E.
Environmental Remediation, Inc.
Baton Rouge, Louisiana
Ralph J. Portier, Ph.D.
Institute for Environmental Studies
Louisiana State University
Baton Rouge, Louisiana
ABSTRACT
Most of the current efforts in biotechnology of waste management
have relied upon conventional genetic and microbial technology. The
genetic engineering of microorganisms has found very limited use.
mainly due to concerns regarding the question of environmental release.
In this manuscript, a technology will be presented which is based upon
the principle that natural populations of microorganisms are able to
adapt to biotransfbrm mixtures of refractory molecules. Information
on a field investigation in which an immobilized microbe bioreactor
was used to treat high concentrations of the chlorinated aliphatic ethylene
dichloride (EDC) will be provided. EDC is one compound in this class
of xenobiotics which have been implicated as the most common
contaminants in industry effluents and groundwater.
THE BIODEGRADATIVE POTENTIAL
People commonly associate bacteria and most microorganisms with
paihogenicity, but the majority are benign and are essential to the ecology
of our planet. They assimilate nitrogen for plant growth, and recycle
carbon (from plant and animal tissues, biological and chemical wastes)
for both aquatic and terrestrial primary consumers. It is this natural
ability to biotransfonm and mineralize organics that we harness, and
manipulate in biological remediation of hazardous waste. Bacteria either
feed directly on an organic pollutant, degrade it concomitantly with
another primary carbon source, or secrete enzymes to break down the
compound. Many biodegradation events proceed through a cometabolic
pathway. In the process of breaking down an abundant primary carbon
source, the pollutant is fortuitously catabolized. Bacteria have been
isolated to degrade a wide range of toxic and recalcitrant compounds.'
(Balthazor. 1986, Haley1, 1988, Roberts, 1987 and Portier, 1982) The bulk
of bioremediation research is being conducted with organisms whose
natural abilities have been enhanced, Timmis and his collaborators at
the University of Geneva (Timmis and Hurayama 1987) are dedicated
to Designing original catabolic pathways, locating the genes necessary
for the reactions, and engineering, by recombinani DNA, ihc complete
pathway into a host cell. However, recombinani DNA is still a new
technology and as fate-and-effect data regarding releases accumulates,
the regulations will become more consistent.
Nature has the ability to recycle and purify itself, but in recent years,
the demand placed on the environment by huge amounts of
anthropogenic pollution exceeds its capacity to recover. Bioremediation
technologies simply attempt to optimize the natural capacity of
microorganisms to degrade organic compounds by supplying essential
inorganic limiting reactants and minimizing abiotic stress.
Biodegradation techniques are versatile and can be utilized at various
stages of treatment. There are three basic ways that the above can be
accomplished: 1) Direct Release. Bacteria, or their extracellular products
may be released directly into the contaminated environment. 2)
Enhancement of Indigenous Microbes. Enhancement of the indigenous
population's dcgradalive potential may avoid the aforementioned
problems of predation. nutrient competition, and subsequent colony
inactivation. Enhancement is achieved primarily by supplementing the
natural supply of nutrients ai the site with additional oxygen, nitrogen,
phosphorous, essential vitamins, or an organic compound necessary
for comctabolism. 3) Microbes in Contained Reactors. Microorganisms
may be used in contained reactors to circumvent the problems of i
complex, and often unfavorable, natural environment. The methods to
be presented in this paper involve use of such specialized biological
reactors. In an enclosed bioreactor. parameters like pH/Eh, oxygenation,
nutrient concentration, temperature, and salinity can be controlled for
optimal biodegradation. Applications include removal of contaminants
from raw materials prior to processing; treatment of pipeline wastes
before discharge; treatment of effluent streams; and decontamination
of soils, sediments, surface water, and groundwatcr (Portier et al.. 1986)
IMMOBILIZED CELL BIOREACTORS
The technology to immobilize whole cells for the decomposition of
toxic organics has only been developed within the last decade. Bacterial
immobilization involves the entrapment of cells onto a matrix. Once
bound, the cells are then readily accessible to the surrounding substrate
(Portier et al . 1986). Chitm, cellulose, glass, and diatomaceous earth
have been tested for use as the solid support material. Bacteria adhere
to chilin, cellulose, and diatoms through covalent bonding and to glass
through adsorption. (Portier, 1987). Chilin and diatomaceous earth act
as sorptive surfaces for many organics and nutrients (Portier, et al, 1988).
There are threshold concentrations below which microorganisms cannot
scavenge nutrients. Chemical sorptkm creates a microenvironment about
the organism that is more copiotrophic than the surrounding medium—*
situation that greatly enhances growth and decomposition. Immobilized
column bioremediation technology is based on the theory that natural
populations of bacteria can be adapted to break down refractory
compounds. The purpose of the packed bed is to provide a large surface
area for microbial colonization.
GROUNDWATER BIOLOGICAL REMEDIATION OF
CHLORINATED ALIPHATICS
Due to the volatility and environmental persistence of low molecular
weight chlorinated hydrocarbons, a very large fraction of them simply
volatilize into the atmosphere when wastewaters contaminated with them
are discharged into conventional industrial aeration lagoons, discharged
into conventional industrial aeration lagoons has simply volatilized into
the atmosphere. The major route for their vapor phase abiotic destruction
is thought to be photo-induced tropospheric hydroxyl ion attack
(Pearson. 1982). When EDC is oxidized in this manner, the
368 BIOREMEDIATION
-------
intermediates are thought to be the mutagenic compounds
2-chloroacetaldehyde, formyl chloride, and 2-chloroacetate (McCann,
et al., 1975).
The challenges to building an effective aerobic biotreatment system
for volatile organic chlorinated aliphatics are creating conditions under
which aerobic organisms can account for a large fraction of the
compound disappearance rate and selecting/maintaining a biological
population with maximum degradation kinetics and minimal exogenous
production of dangerous intermediate compounds. A 75 L, continuous
flow, immobilized cell bioprocess system was developed specifically
for volatiles degradation and was tested at a chemical production site
having extensive halocarbon contaminated ground water. The
groundwater contains a variety of one- and two-carbon chlorinated
compounds. EDC is present in far greater concentration than any of
the other organics, and was routinely observed at concentrations in
excess of 1,800 mg/L.
A detailed discussion of the reactor design has been presented
elsewhere (Friday and Portier, 1989) and is summarized as follows:
the current system consists of three functionally distinct subsystems.
The first is a raw effluent conditioning system which removes foreign
materials via a lOOju filter, dilutes recovered ground water to the degree
required to achieve biologically acceptable toxicant concentrations, adds
nutrients, and adjust/maintains media pH and temperature. Biological
conversion occurs in the second subsystem (reactor vessel) which is
partitioned into two distinct volumes. In the first, air is sparged into
the feed waters to mix and aerate the influent water. Admixed air is
then separated from the water before it enters the second reaction stage.
In the second section, the water moves in plug flow through a bed packed
with porous biocarrier on which the selected chlorinated
aliphatic-degrading organisms are immobilized throughout the support.
Modular column units have been fabricated which can be mechanically
coupled to provide a desired packed bed volume and control the extent
of the bioconversion. Design considerations have included gas
sparging/gas distribution, maintenance of carrier integrity, gas/liquid
separation, and materials of construction. The reactor is instrumented
to allow pH, temperature, and dissolved oxygen levels to be continuously
monitored and controlled. In addition, a gas scrubbing unit is attached
to remove organic vapors from process off-gases prior to release into
the environment. A third subsystem provides final clarification of the
decontaminated water.
Adapted Microorganisms
Bacterial cultures which aerobically metabolize EDC as a sole source
of carbon and energy were adapted for continuous degradation of EDC
using protocols as discussed in detail in Portier, et al.1983. These strains
were adapted for detoxification applications using mechanisms outlined
in earlier aquatic microcosm studies. Particular efforts were made to
insure that no other sources of carbon were available for metabolic
maintenance and that volatilization losses were controlled to avoid
erroneous estimates for substrate availability.
Site Deployment
The reactor was deployed on site at the facility and connected to the
existing ground water recovery system to provide a continuous source
of contaminated water. Compressed air (oil-free), was introduced at
the base of the well-mixed section of the reactor at approximately 500
standard cc/min. The reactor operating pressure was regulated to 30.0
psig and temperature was controlled at 30 °C. The pH of the ground
water was automatically maintained between 6.5 and 7.5 by addition
of 1 M sodium hydroxide solution. Ambient temperatures ranged from
11-35 °C over the 30 day field trial. Approximately 12.75 kg of
diatomaceous earth carrier (Type R-630, Manville Filtration and
Minerals) was installed in the system for the initial pilot test. This carrier
is unique in that it has a controlled porosity for optimal colonization
of microorganisms, thus providing a considerable biocatalytic capability.
BIOTREATMENT OF EDC-CONTAMINATED GROUNDWATER
Contaminated ground water, diluted 33%, was treated during the
course of the field pilot study. Ethylene dichloride (EDC), the primary
waste constituent of concern in this process stream, was monitored for
microbial mineralization at dilute and elevated levels of contamination.
Both batch and continuous modes of operation were investigated. Batch
tests were initiated with initial concentrations of 1.5 to 2.5 mM EDC,
while continuous flow tests were run on influent streams with more
than double this concentration. Time zero concentration averaged 2.30
mM EDC (Molecular Weight EDC = 98.96 g/gmole) for Batch Tests
#3 (Figure 1). EDC concentration was undectable after 20 hours of
holding time. Thus, for batch #3, a mineralization rate of 0.14 mM/L/h
was realized. As reported elsewhere, with an influent flowrate of 2.85
L/h, steady-state removal rates for continuous flow mode were 1599
mg EDC/h (Friday and Portier, 1989). Influent feed concentrations
entering the system averaged 5.68 mM EDC. Effluents from the reactor
averaged 0.009 mM EDC (see Figure 2). A carbon trap in series with
the reactor off-gas sorbed volatilized EDC at the rate of 4.84 mg/h,
inferring that in excess of 99% of the observed removal rate was due
to biodegradation. GC/MS analysis of an off-gas sample collected
downstream of the carbon trap just prior to removing it showed
non-detectable levels of EDC, indicating that no organic break through
occurred.
Chlorinated Ethane Ground Water Study: Gas Chromatography Analyses
Batch Operation #3
§• 3.0-
E
C3
6
Q
LU
8 15.25
Time (Hours)
20.25
24.5
Figure 1
Batch Biotreatment of Ethylene Dichloride (EDC)
Using an Immobilized Bioreactor
(Adapted from Friday and Portier, 1989)
CD
jo
o
100-
10-
1
.1'
.or
.001-
.0001
EDC: Gas Chrom. Feed (mM)
EDC: Gas Chrom. Export (mM)
0 8.25 16 24.5 32 40.25 48.25 56.5 64.75 72.75
Time (Hours)
Figure 2
Reactor Influent and Effluent Ethylene Dichloride (EDC)
Concentrations in Continuous Flow Operation
(Adapted from Friday and Portier, 1989)
DISCUSSION
The technologies evaluated to date for the effective treatment of
BIOREMEDIATION 369
-------
contaminated ground waters and industrial effluents in industrialized
corridors have provided pragmatic, cost-effective solutions for the
removal of xenobiotics.
"Once developed and proven, biodegradation is potentially less
expensive than any other approach to neutralizing toxic wastes Such
systems involve a low capital investment, have a low energy
consumption, and are often self-sustaining operations" Office of
Technology Assessment (Nicholas ,1987)
Biological treatment of many groundwater contaminants will
significantly minimize the associated costs of excavation, transport and
incineration of these materials which are the current commercially
available technologies. Additionally, since many xenobiotics have been
effectively decomposed to nontoxic substances, a permanent solution
to the removal and disposal of such materials can be realized. Biological
solutions which involve treatment in place further reduce the risk to
the general public by minimizing the necessity of large scale excavation
and transportation from contaminated sites to U.S. EPA approved
disposal facilities. Future applications of these modular biorcactors in
treating waste streams associated with the manufacturing of high
technology systems such as circuitry, computers and advanced
metallurgical processes is anticipated. Additionally, the usefulness of
these systems as recycling devices in life-support systems is technically
feasible and, currently, under evaluation.
ACKNOWLEDG EMENTS
Research presented in this paper was supported by funds from NOAA.
Office of Sea Grant Development and the Louisiana State University
Sea Grant Program. Additional funding from the State of Louisiana
Board of Regents and from Manville Service Corporation is also
gratefully acknowledged.
REFERENCES
I. Balthazor, Terry M and Laurence E. Hallas, 1986. "Glyphosate Degrading
Microorganism!, from Industrial Activated Sludge" Applied and
Environmtnial Microbiology, vol 51 No 2 Feb 1986 p. 432-434.
2. Friday, David D. and Ralph J. Pomcr. 1989. Evaluation of a packed bed
immobilized microbe bioreactor for the continuous biodegradation of
halocarbon-conlaminated ground water* Proceedings of AWMAfEPA
International Symposium on Hazardous Wasit Treatment: Btosyslerra for
Pollution Control, Cincinnati, OH. Feb 20-23. 1989
3 Haley. Roger, and Howaid Simon. 1988. "Bacteria ihnve on phenolk wanes"
Chemical Processing. Feb 1988 p 156-158
4. McCann. J.. Simmon. V, Slrciiwicser. D . and Amcv B.N. Proc. Nail.
Acad. Sci. USA 72 (1975).
5. Nicholas. Robert B. 1987 "Biotechnology in Hazardous-Mule DttpotaJ:
An Unfulfilled Promise" ASM Newj 53. No. 3 (1987) p. D8-M2.
h Pearson. C R Cl- and C'2-halocarbons. In The Handbook of Environmental
Chemistry. U>l 3 1982
7 Ponicr. R.J . H.M. Chen and S P. Meyers. 1983 Environmental effect and
laic of selected phenols m aquatic ecosystems using microcosm approaches
Developments in Indust. Micmbiol, \bl 24. Pp 409-424.
8. Ponicr. R I and K Fujisaki. 1986. "Biodegradalion and cominuoui
dctmifkaiion of chlorinated phenols using immobilized bacteria" in
Toxicology Assessment. John Wiley & Sons (1986) vol I. 501-513.
9 Ponicr. R J . 1987 Enhanced biotransformaiion and biodegradation of
polychlonnatcd biphcnyU in (he presence of aminopolysacchande*. American
Society for Testing and Materials (Special Technical Publication 971, Aquatic
Toxicology — KXh Annual Symposium. Adarra. Chapman. Landis. Eds.).
pp 517-577
» Porucr. R J , ci al . I98R "Evaluation of a Packed Bed Immobilized Microbe
Bioreactor for the Continuous Btodegradalion of Contaminated Ground
Waters and Industry Effluents Case Studies" SAE Technical Paper Serin
No 881097. July 11-13. 1988
II. Timmis. K N and S. Harayama. 1987. "Potential for laboratory engineeraig
of bacteria to degrade pollutants." Paper presented al Reducing Risks From
Environmental Chemicals Through Biotechnology. University of Washington,
Seattle. WA. July 19-22.
370 BIOREMEDIATION
-------
TOPIC 5: A New Solid/Liquid Contact Bioslurry Reactor
Making Bio-Remediation More Cost-Competitive
Gunter H. Brox
Douglas E. Hanify
EIMCO Process Equipment Company
Salt Lake City, Utah
ABSTRACT
The reactor system described in this paper has been developed based
on slurry agitator technology used in the mineral processing industry.
The reactor has been modified to act as a vessel in which naturally
occurring biological degradation processes are enhanced. It provides
aeration, mixing, temperature control, nutrients, and in certain
applications, volatile emissions control. A bioslurry reactor approach
is recommended to biodegrade organic hazardous substances in a matrix
where in-situ land treatment often fails. A bioslurry reactor can also
be used in a soil-washing flow sheet for the fine particle fraction which
contains often the highest contaminant levels.
The bioslurry reactor presented in this paper can handle solids
concentrations in the 30-50 wt% range. Energy consumption is typically
only 25-50% of that needed in conventional liquid/solid contact (LSC)
reactors which use turbine mixers or surface aerators. The reactor is
presently being tested in RCRA and Superfund applications.
INTRODUCTION
Many of the organic substances listed by the U.S. Environmental
Protection Agency (U.S. EPA) as hazardous are biodegradable.1 On
most Superfund sites organisms have been identified which can biodegrade the
organic contaminants given the availability of oxygen and nutrients, and under
the right environmental conditions (soil pH, temperature, moisture). Since none
of these parameters are usually in the optimal range for the bacteria involved,
biodegradation in nature is often very slow.
Table 1 summarizes the four methods commonly used for bioremediation.
Most experience has been gained with land treatment, particularly in the oil
refining industries,2 and whenever hydrocarbon spills are being cleaned up. It
is the bioremediation technology of choice if land is readily available and time
is no constraint. In colder climate, bioremediation by land treatment often comes
to a virtual standstill durng the winter months as the top soil freezes. Clean-up
levels in a slurry reactor system are more predictable than land treatment units.
Composting, on the other hand, produces some heat and may become
a more widely used bioremediation technology, especially in colder
climates.
In-situ treatment is the only alternative when the contaminants have
reached deep subsurface levels or are primarily under buildings and
excavation is not possible. Hydrogeologists play a major role in the effort
to get nutrients and oxygen to the contaminated areas and stimulate
bacterial activity.
Liquid/solid contact systems have been used, primarily in lagoons
(in-situ) or where tanks are available on site. However, it has been found
that energy input has to be kept quite high in order to keep the soil
particles suspended. Solids concentrations often have to be limited to
10-20 wt% in order to keep the particles sufficiently suspended. Power
outages can cause significant operating problems as the materials settle
Table 1
Bioremediation
Land Treatment
Composting
Liquid/Solids Contact systems
In Situ Treatment
out and compact. Air is often provided through spargers which can clog
quite easily during a prolonged power outage. An LSC reactor is shown
in Figure 1. This features above ground tank construction, draft tube
with direct-drive mixing, and control of volatiles.
The EIMCO Biolifi™ reactor, shown in Figure 2, is basically a modified
slurry agitator that uses a dual drive design which EIMCO has
manufactured for its Reactor Clarifier™ for decades. Thisjlual drive
allows independent operation of the axial flow impeller and the rake
arms at two distinctly different 'speeds. In a large diameter tank, the
impeller, mounted on a separate shaft, typically rotates at 20-30 rpm
while the rake arms turn at less than 2 rpm. Diffuser pannels consisting
of vertically stacked diffuser tubes are mounted on the rake arms. The
diffuser tubes are of a special rugged design, allowing rotation through
an often viscous slurry without breaking at the point of connection to
the air manifold. The diffuser membranes typically consist of a slotted
elastomeric material which has been selected to chemically resist the
organic contaminants found in the soil slurry. Such diffusers are known
to be relatively clog free and to have superior oxygen transfer efficiency.
In addition, release of the rotating curtain of fine air bubbles keeps
most of the fine particles in suspension and creates the necessary
turbulence to enhance the mass transfer of oxygen, nutrients and
substrate molecules into the bacteria cell. The impeller turning at a
higher speed causes a downward flow and affects bulk blending.
Variability in contaminant concentrations in the feed stream is less a
problem in such a completely mixed reactor than it would be in a batch
reactor, where high substrate concentrations at the onset can be
inhibitory to the bacteria.
BIOREMEDIATION 371
-------
- COMPRESSED AH
HEADER
TYPICAL esc TANK REACTOR
PLAN VEW NOT TO SCALE
TANK DIAMETER . 45'
TANK HEIGHT . 48-
SECTION A-A'
NOT TO KALE
Figure I
Liquid Solids Tank Reactor*
BAFFLES
FINE
BUBBLE
DIFFUSERS
FIAKE ARM
l-igure 2
Eimco Biolifl™ Rcjicior
The tank is baffled to enhance mixing. Coarser particles which are
not kept in suspension by the fine bubble diffusers and have settled
to the tank bottom are raked to a central airlift which pumps them to
the top, where they are discharged into a specially designed slurry
removal system. A Y-shaped pipe with a vertical leg connected to a
funnel collects the slurry directly from the airlift. Since the airlift
transporting material from the tank bottom will contain a higher
concentration of coarse solids than the average slurry in the reactor,
it is possible to regulate the quantity of coarse solids within the tank
by means of this take'off device and pass a fraction of this material
on to the next reactor or out of the system. Control of coarse solids
is essential in order to minimize torque on the mechanism.
Depending on the application, any number of reactors can be arranged
in a cascading system to permit continuous feed and overflow. The more
stages are arranged in series, the more the system approaches true plug
flow conditions. At the same time, optimum biokinetic rate is achieved
in each stage. The bacteria population is fully acclimated to the organic
contaminants and biomass concentration has reached an optimum in
accordance with substrate concentrations available.
Alternatively, the EMICO Biolift™ Reactor can be run in a batch or
a semi-continuous feed mode. From a process engineering point of view,
such a mode of operation is more easily controlled, but kinetic rates
will be slower because of a lag phase in bacterial activity as a result
of acclimatization and biomass growth.
In order to use a slurry reactor effectively in a soil remediation project,
some pretreaonent will be required to remove all oversize material. A
proposed remediation flow wheel is shown in Figure 3. The excavated
contaminated soil is first moved through an attrition mill to slurry up
the material. After this, it passes through a trommel screen to remove
any gravel, debris, and other ovcrzize material. The soil passing through
the screen is then fed into a counter-current washing screw classifier.
Most of the sand will be clean after these three washing steps and can
be discarded. The finer materials and the excess wash water that can
not be recycled are then passed into a series of bioslurry reactors.
Total hydraulic residence time in these reactors will vary depending
on the nature of the organic contaminants, their concentration, and
clean-up level required. The soil slurry is finally dewatered in either
a pressure filter, vacuum filter, or centrifuge. The most efficient and
economical dewatering equipment is dependent on the soil characteristics
and the quantities of slurry to be processed. It must be evaluated on
a case-by-case basis.
The process shown uses bioslurry reactors as the primary treatment
step. Other flow sheets are possible as long as they achieve the
pretrcatment objectives of slurrying. washing, and classifying into
different size fractions.
VOLATILES EMISSION CONTROL
In many instances volatiles emission control is very desirable,
372 BIOREMEDIATION
-------
Figure 3
Bioremediation Flow Sheet
particularly if discharge of toxic air emissions would exeed applicable
air pollution control standards. Since a number of these volatile organic
compounds are readily biodegradable, but are air-stripped by the
diffusers before the bacteria have metabolized them, a mode of operation
was devised whereby the off-gas, collected in the reactor top, is
recirculated back into the slurry via the diffusers. The reactors are
gas-sealed and the compressor recompresses the off-gas. This gas stream
is continuously analyzed by on-line oxygen and carbon dioxide
analyzers. The flow schematics are illustrated in Figure 4.
02 Analyzer CO2 Analyzer Chrornalo
Carbon | Condensala
returned to
reactor
LEGEND
FCV . Ftaw Control Valve
SOV - Solenoid Operated Valve
Figure 4
Liquids/Solids Reactor With Volatilization Control System
The gas analyzers have control capability and can actuate solenoid
valves at pre-determined setpoints. For example, if carbon dioxide, due
to the bacteria's metabolic activities, increases above the setpoint, a
portion of the gas stream is passed through a scrubber until the carbon
dioxide concentration has been reduced to an acceptable level again.
Likewise, when the oxygen concentration due to bacterial uptake drops
below the setpoint, air or pure oxygen is admitted to the system until
ambient oxygen concentration has been restored. An equivalent volume
of air is treated through a carbon adsorption column to remove any
residual non-biodegradable organic volatile compounds. Operation in
the gas recirculation mode reduces the cost for expensive volatile
emissions treatment significantly. The reactor is always operated at a
slight vacuum of 1" to 2" W.C. to avoid any undesired emissions.
FIELD EXPERIENCE
The EIMCO Biolift™ reactor has been used to date in two applications.
In one application a RCRA refinery sludge with an oil and grease
concentration of approximately 40 wt% was aerobically digested. Total
solids concentration in the reactor was 25 wt%. The reactor was operated
in the batch mode and a 60 wt% reduction in the oil and grease was
obtained after 39 days. After all of the carcinogenic compounds of
concern have been removed to acceptable levels, the material can then
be further treated in a land treatment cell. In this application, gas
emission control was particularly important.
In a second application, the reactor is presently being used to treat
the fine particles stream, residue from a soil washing operation. The
contaminants are primarily PAH's and pentachlorophenol. Based on
preliminary results, a 90 to 95 % removal can be achieved in a three-stage
continuous flow system.3
ECONOMIC CONSIDERATIONS
The advantages of the EIMCO Biolift™ reactor are primarily related
to operating and maintenance costs. Energy consumption is typically
less than one half of what is required when turbine mixers or surface
aerators are employed. In a recent cost comparison between the two
technologies for the bioremediation of approximately 30,000 yd3 of
contaminated soil, capital costs were $76/yd3, and operating costs were
$60/yd3 for employing surface aerators and draft tubes. Using the
EIMCO Biolift™ reactor would result in the same capital costs but
would show operation cost savings of $13/yd3, primarily due to energy
savings. Because of the large size reactors required to meet the clean-up
schedule, all the bioslurry reactors would be depreciated over the life
of the project as reuse on another project would be difficult.
Capital costs are strongly influenced by the size of the project and
the time schedule in which it has to be executed. In order to achieve
further economies, it is important to standardize the bioslurry reactor
as much a possible. Presently it is envisioned to build four reactor sizes
from 70 m3 to 1,100 m3. The first size reactor would still be
transportable completely assembled and thus would require only minor
erection work in the field. After its use and decontamination on site
it would be shipped and reused on the next site. Any reactor larger
than can be transported by road in one piece will need to be assembled
and erected in the field. The rake and airlift mecheanism can be
constructed such that it can be dismantled into several pieces which
can be reconnected and reused. Tanks may or may not be reusable
depending on the circumstances.
OUTLOOK
EIMCO Process Equipment Company is presently engaged with
several process engineering firms in the proposal of pilot and full-scale
remediation projects intending to use bioslurry reactors. The issue of
scale up is being investigated in order to design large scale reactors
based on the kinetic data obtained at the bench scale. Several alternatives
to provide mixing and aeration in a more cost-effective manner are being
examined as well. It is believed that bioslurry treatment in large scale
reactors will one day be as common as Activated Sludge processes in
waste water treatment. To reach this point a concerted effort will be
required between process engineering companies and equipment
manufacturers.
REFERENCES
1. Nicholas, R.B. and Giamporcaro, David E. Nature's Prescription, Hazmat
World, June 1989.
2. ReTec, Effectiveness and Cost of Various Bioremediation Technologies, RCRA
Conference New Orleans, April 1989.
3. EPA SITE Demonstration, Biotrol Soil Treatment System, Sept. 1989.
4. David R. Hopper, Cleaning Up Contaminated Waste Sites, Chemical
Engineering, August, 1989.
5. C.H. Vervalin, Bioremediation on the Move, Hydrocarbon Processing August
1989.
BIOREMEDIATION 373
-------
Calcining Rotary Kiln For Detoxification
of Non-Autogenous Wastes
James F. Angelo
Universal Energy International Inc.
Little Rock, Arkansas
ABSTRACT
An advanced multiple burner system has been invented and is being
developed and tested. The multiple burner is designed to direct and
locus a plurality, usually 30 to 40 individual burner flames. direct!)
on the tumbling bed of contaminated feed stock to be thermally treated,
detoxified, and calcined in a rotary kiln combustion system Combus-
tion air which is pre heated, natural gas or LP gas and/oxygen is deli-
vered to each burner through a patented and proprietary free spanning.
air cooled, platform and manifold. The principal objects of this new
burner system are to increase throughput, improve destruction
efficiencies, decrease participate entrainmeni and fundamentally improve
the thermal destruction/treatment and calcination of feed stock.s. which
are contaminated with organic constituents and are non-autogenous in
nature such as soils, sludges, slurries and excavated landfills, for
example, in rotary kilns.
INTERACTION
For many decades, rotary kilns have been widely viewed as the "work
horse" of the calcination and toxic/hazardous/industrial waste inciner-
ation/thermal treatment/destruction industry. Globally, thousands of
rotary kilns are in use thermally treating/incinerating thousands of feed
stocks, many contaminated with various organic chemicals. It is widely
accepted that virtually any solid, sludge, slurry, etc.. or combination
thereof, can be processed in a rotary kiln. Many kilns calcine lime and
cement, as well.
In spite of rotary kiln's advantages and abilities to accommodate and
process almost all solid feedstocks, kilns have traditionally been an
inefficient process Rotary kilns have been si/,ed with very large com-
bustion volumes in order to reduce velocities of the gases as they exit
the kiln, in the effluent, in order to reduce paniculate cmrainment, Kilas
normally have to be fitted with large and expensive scrubbers to reduce
paniculate discharges to permittable levels. The relatively large volume
of kilns has increased their Capital costs as well as their operating costs,
particularly the replcement of refractories periodically, a time-consuming
and costly operation. Additionally, the inefficiencies of the heal loss
and dissipation of heat energy, through the kiln shell, which often is
a significant waste of energy, increases operating expenses. Tradition-
ally, kilns calcinating non-autogenous materials utilize large oil. gns
or combination burners which are mounted in the firing hood(s) or
breeching(s) of the kiln. Normally, the burner's flame pattern coven.
or contacts only a portion, often only a small /.one or section of the
tumbling bed of feed stock, an inefficient method. The poor contact
between the burner flame and the tumbling bed of non-autogenous feed
stock being thermally treated typically requires rotary kilns to be quite
large and particularly long.
MULTIPLE BURNER SYSTEM
The multiple burner system is typically 30 to 50 individual, small
burners, usually with thermal outputs of 0.5 to 1.000.000 BTU/HR
burner. These burners are distributed along an air cooled platform and
manifold which spam from head/breeching to head/breeching, generally
in an offset, axial location within the kiln. The offset location allows
for the individual flames, which are typically 3 ft in length, to be directed
at and on the tumbling bed of soil, sludge, slurry, lime, cement or
excavated landfill feed slock.
The thermal output of these burners can be regulated by adjusting
the combustion air, gas and oxygen flow rates. The combustion air is
also the ciKilmg air for the platform and is pre-heated due to the cooling
effect. The cooling effect enables the maintenance of the structural in-
tegrity of the air cix>led platform system. The flame temperature can
be regulated from 2.500T to 4,000°F with maximum oxygen
enrichment
The flame is positioned so that it is generally tangential to the kiln
shell. This tangential flame direction positioning has an additional
benefit in that a cyclonic, swirl, helical pattern of air. gas and panicu-
late is induced. Previously, tangential combustion air injection systems
and technologies developed by the author have repeatedly demonstrated
the ability to centrifuge paniculate out of the air/gas/particulate mixture
in rotary kiln incineration systems due to this beneficial tangential
injection of air and other gases.
US Department of Energy studies have documented the dramatically
reduced paniculate loading in a kiln's flue gases due to the centrifugal
effects derived from the tangential injection of combustion air. in swirling
patterns, throughout the entire length of rotary kilns via a plurality of
combustion air injection nozzles. Paniculate loadings of flue gases have
been consistently demonstrated and documented at 0.08 GR/SCF
utili/ing the earlier combustion air injection version of this free spanning
system. Paniculate entrainmem rates, in flue gases, below 0,08 GR/SCF
have been demonstrated with certain feed stocks, as well. This cyclonic
effect on paniculate entrainment can eliminate the need for scrubbers.
In other cases, gas cleaning systems can be down-sized and are subject
to less wear, maintenance and abrasion than is typically the case.
KILN DOWNSIZING
This new system is showing great promise with its ability to shorten
rotary kilns A great many very large rotary kilns are operating in
calcining modes, producing lime and cement. Most of these kilns are
150 to 400 feet in length. Our studies indicate that this system of multi-
ple burners, directed and focused on the tumbling bed of feed stock,
rather than one large flame, with its uneven heat transfer and hot/cold
rones, can dramatically reduce lengths of kilns from 150 to 400 ft down
374 INCINERATION
-------
118
---- « 86
Figure 1
to 60 ft in length while maintaining throughputs. Thermal processing
and efficiency are improved and fuel is saved.
An additional goal in the testing and demonstration of the process
is to improve destruction efficiencies of the organic contaminants in
the non-autogenous feed stocks.
Another feature of the process is its ability to divide a rotary kiln
into four or more independently controllable zones.
These systems typically are outfitted with thermocouples which are
distributed along the air cooled platform. There usually is one
thermocouple per zone and four usually are installed. The thermocouple
bases and wiring are protected in the air cooled, free spanning system.
RESEARCH AND DEVELOPMENT PROGRAM
A Research and Development/Demonstration project is being
implemented for this patented and proprietary system now assigned to
Universal Energy International, Inc. The system discussed herein will
be installed on a test center rotary kiln owned by Fuller Company's
Fuller Power Corp. of Bethlehem, Pennsylvania. Additional participants
are Air Products & Chemical, Inc.'s Applied Research and Develop-
ment group. A wide variety of non-toxic, non-hazardous materials will
be pyro-processed. Destruction of surrogate contaminants in feed stocks
16
8
62 60 121
U
120
52 120
50 120
Figure 3
?fc>^???^y^^
108
Figure 4
Figure 2
will be studied and destruction efficiencies will be established. Air
Products & Chemical's, Inc. is providing oxygen control and regula-
tion systems and technical input.
Certain versions of these systems and technologies are being offered
for immediate commercialization. The assignee of this technology has
offered proposals to sell two systems which are approximately 60 ft
long for commercial soils detoxification and incineration projects, both
portable and fixed sites and systems.
Grant applications are pending with the U.S. Department of Energy
and the National Science Foundation. Plans are being developed to
demonstrate this system under a U.S. EPA SUPERFUND program,
the innovative technology program. Research and development funds
have been allocated and set aside by the U.S. Department of Energy/
Pittsburgh Energy Technology Center for a similar test demonstration,
and R & D program where an air/sorbent injection/data acquisition
version of this proprietary will be operating during the parallel program.
INCINERATION 375
-------
o
z
m
H
5
z
OXXfDl
<;• A.
VAMT ^||O
t>«'^<
f'igurc ^
-------
Site Remediation Using Mobile Thermal Destruction
At the Electric Utilities Site in LaSalle, IL
James F. Frank
Richard M. Lange
Greg R. Michaud
Illinois Environmental Protection Agency
Springfield, Illinois
ABSTRACT
Electric Utilities Co. (EUC) was a manufacturer of capacitors in
LaSalle, Illinois until 1981. EUC left this site in 1981 and filed bank-
ruptcy in 1983. During operation, EUC had used PCB dielectric fluid
in the product and had used waste oils for dust control both at the facility
and in an adjoining residential neighborhood. Subsequently, wind
erosion and vehicle traffic transported PCBs up to 1.2 mi from the site.
In addition to PCB soil contamination, the local groundwater has been
impacted by chlorinated solvents.
Two phases of remedial action (RA) were planned: to (1) off-site soil
contamination in Phase I, and (2) on-site soils, groundwater and stream
and sewer sediments in Phase II. On-site thermal destruction is the
selected alternative to remediate this site. Phase I is in progress with
Phase II in the procurement process.
The Phase I RA which involved 1.2 of inner city state highway
required relocation of 25 families during excavation. After excavation,
extensive landscaping was required to restore the neighborhood to pre-
excavation conditions. This landscaping required replacement of
$120,000 in trees and perennial plants and 27,000 yd3 of sod.
The excavated material was segregated into two stockpiles based on
levels of PCB contamination. One stockpile contained less than 50 ppm
PCB-contaminated material while material with more than 50 ppm PCBs
was placed in the other. The total payable yardage excavated was 23,258
yd3 and was nearly equally divided between less than 50 and greater
than 50-ppm contaminated material. Thermal destruction is ongoing
at this time on the greater than 50 ppm waste under authorization by
IEPA, with concurrence by the U.S. EPA. Treatment of the less than
50 ppm material is complete. Thermal destruction services are being
provided by Westinghouse-Haztech utilizing an infrared unit originally
manufactured by SHIRCO.
The thermally treated soil has to meet a cleanup criterion of 2 ppm
total PCBs and originally was regulated as a State of Illinois Special
Waste. The Phase I RA treated soil is going off-site for disposal to a
landfill where it is permitted to be used as daily cover. The Phase II
treated soil will be used as on-site backfill where possible. This handling
of treated soil allows control of the material but does not consume valu-
able landfill capacity.
The unique feature of this RA is the extensive interaction with the
residential population due to the extensive excavation of lawns. This
project posed a major community relations challenge.
INTRODUCTION
The Electric Utilities Company (EUC) site in LaSalle, Dlinois is cur-
rently the subject of Phase I of a multiphased Remedial Action (RA).
The EUC manufactured industrial capacitors utilizing PCB as a die-
lectric fluid. In 1981, the company left this location and relocated their
operations to North Carolina. Soon after moving, EUC entered
bankruptcy and dissolved the company. During their final years in
LaSalle, EUC had been the subject of a number of regulatory com-
plaints and enforcement actions by both the Illinois Environmental Pro-
tection Agency (IEPA) and the U. S. EPA.
In 1983 and 1984, Immediate Removal actions by the U.S. EPA re-
moved some waste material and redirected surface water flow back onto
the site and into a pond for sedimentation and infiltration. An adjacent
off-site business parking lot and driveway were asphalted to limit access
to contaminated soil. In 1986, the IEPA conducted a followup Imme-
diate Removal action to dispose of 260 drums of waste and 735 gal of
trichloroethylene (TCE). Following in lEPA's tradition of reducing quan-
tities of waste for disposal, this solvent was analyzed and determined
to be of sufficient quality to appropriately re-enter the commercial
market.
The Remedial Investigation identified extensive on-site soil contami-
nation by PCBs on-site, certain soils contaminated by chlorinated
naphthalenes, on- and off-site contamination of groundwater by various
chlorinated solvents (predominantly TCE) and PCBs (including free
oil) and extensive off-site PCB contamination. The off-site PCB con-
tamination of soils unexpectedly included widespread contamination
in residential yards, business properties, agricultural fields and approxi-
mately 1.2 mi of street right of way. Some of the more unusual areas
contaminated with PCBs included the presence of PCB dust in the in-
teriors of homes and businesses including furnace ducts, storm and sani-
tary sewer sediments and stream sediments where the storm sewers
surface.
The measured quantities from Phase I of the RA and the Engineers'
estimate for the Phase II RA revealed the extent of contamination. The
following quantities and types of waste have or will be remediated: over
23,500 yd3 of off-site soil and 42,000 yd3 of on-site soil with PCB con-
centrations ranging up to 113,000 ppm; as much as 1000 gal of trans-
former oil with PCB concentrations expected in the 50 to 60% range;
up to 1000 ft of a remote off-site stream requiring excavation. Over
7500 ft of storm and sanitary sewer will be hydraulically and mechani-
cally cleaned and 3500 ft of passive groundwater collection will be
piping placed at depths to 25 ft to feed a water treatment plant which
will be constructed for remediation of the solvent- and PCB-
contaminated groundwater. This water treatment plant is expected to
generate an additional 188 tons of PCB-contaminated oil.
The factory building complex is so heavily contaminated with PCB
that demolition is the selected option followed by thermal destruction
of all amenable materials. Additionally, the factory buildings and
remaining process equipment harbor significant quantities of asbestos
contaminated with PCB. This material will require off-site disposal in
an appropriate secure facility.
INCINERATION 377
-------
The Feasibility Study (FS) evaluated the standard options to protect
the human health and environment including the No Action Alterna-
tive, Waste Consolidation and the construction of an On-site Landfill
and Off-site Disposal in a Landfill. Additionally, the FS evaluated the
long term permanent solutions of Off-site Thermal Destruction and On-
site Thermal Destruction. The landfill options seemed to meet the
criteria for appropriate alternatives but did not satisfy citizens' con-
cerns about removing the contaminated material from their town and
making the property a usable resource. Additionally, this option was
not fully acceptable to the State of Illinois because it failed to provide
the permanence of destruction, leaving the state a long-term operation
and maintenance responsibility. The alternative which was ultimately
selected utilizes on-site mobile thermal destruction for all treatable
material, with off-site landfill disposal held to a minimum. The only
on-site disposal will be for innocuous, thermally treated material
PHASE I REMEDIAL ACTION
During Phase 1 RA. 23,258 yd' of PCB-contaminated soil from off-
site locations was excavated and placed in stockpiles on the site pending
thermal treatment. The containable material was placed into two separate
stockpiles based on the level of contamination. One stockpile was
reserved for material with PCB concentrations of less a 50 ppm and
the other for material found to have concentrations greater than or equal
to 50 ppm. Thermal treatment of the less than 50 ppm material was
initiated on Nov. 29, 1988, under an interim Operating Approval Letter
issued by the IEPA. Completion of thermal treatment of this material
was accomplished on June 14, 1989, and treatment of the 50 ppm or
greater material began on that date. Treatment of the greater than
50 ppm material is also taking place under an Operating Approval Letter
issued by IEPA with U.S. EPA Region V concurrence.
The thermally treated soil from this RA was originally being trans-
ferred as Special Wiste to a local, IEPA- permitted Illinois Special Waste
Landfill under Illinois' Special Waste Manifest system (Special Waste,
as defined in Illinois, means any industrial process waste, pollution con-
trol waste or hazardous waste). After sufficient experience with the
treated soil was gained (both chemically and physically) by IEPA, the
RA Contractor and the landfill operator the following program was has
been approved: the thermally treated soil from the Phase I RA has been
delisted as Illinois Special Wasie and is simply considered waste; the
treated soil is no longer subject to manifesting requirements and the
treated soil is permitted to be utilized at the landfill as daily cover
material. This delisting resulted from a coordinated effort by all parties.
This delisting and daily cover use meets two needs. The material is
sufficiently innocuous to require no manifesting or special management.
Useablc as daily cover, the material is being removed to an appropriately
secure facility for public comfort but is not consuming valuable land-
fill capacity.
In addition to the excavation and thermal treatment of PCB-
contaminated soils, the Phase I RA included cleaning of the interiors
of 25 private homes and 2 businesses and the replacement of all land-
scaping material removed during excavation in their yards. This effort
involved (he laying of 27,000 yd! of sod and the replacemeni of over
$120,000 worth of landscaping.
PHASE II REMEDIAL ACTION
In the Phase II portion of thi.s RA, a contractor will demolish the
existing factory complex with the goal of decontaminating or thermally
destroying all possible materials, in order to reduce off-site disposal
to a minimum, thereby reducing disposal facility consumption and
reducing the State's long-term liability. This waste minimization effort
is being encouraged by the absence of various pay items in the bid
specifications and financial encouragement of thermal destruction, the
payment for certain decontamination efforts and the return of all salvage
dollars to the contractor. One example of these specifications will be
an extensive coring and sampling effort directed at over 68,000 IV of
concrete flooring; this concrete will be analyzed in an attempt to iden-
tify the depth of PCB penetration. Where the concrete overlays uncon-
taminated soil the contaminated surface of the concrete will be
mechanically removed and the collected material will be thermally
treated. Following decontamination, this concrete may go to a "Demo-
lition Debris Only" landfill or. in the absence of reinforcing steel, may
be used as clean fill in land reclamation or as rip rap in local surface
water projects (the town of LaSalle is located on the North bank of
the Illinois River).
Following demolition of the factory buildings, excavation of an esti-
mated 42.000 yd' of PCB-contaminated soil can proceed unencum-
bered. One significant difference between Phase I and Phase D is that
in Phase II the area of excavation will be under complete control of
the remediation contractor and the State of Illinois. This control will
allow the use of treated soil as backfill on site. In Phase I. the excava-
tion and backfill had to proceed rapidly to reduce impact on various
residential and business property owners, thereby requiring an imme-
diate source of backfill material; in Phase II, the excavation of coo-
laminated material can more closely follow the production capacity of
the thermal destruction unit. This item is not specifically required in
the specifications but is encouraged by the absence of both a Backfill
pay item and an Ash Disposal pay item in the contract documents. Ob-
viously, the contractor is financially encouraged to utilize treated soil
as on-site backfill. The chemical quality of the treated soil will be closely
monitored, and the treated material will not be used within I ft of final
grade; this requirement should assure rapid establishment of vegeta-
tive cover and reduce the potential of light tillage operations turning
treated material to the surface, thereby unnecessarily raising public
concern.
The removal of contaminated sediments, soils and debris from me
off-site stream will follow relatively standard cleanup methods as will
the sewer cleaning operations One exception is to the standard methods,
is that trees and brush must be removed to construct a temporary access
road to the stream area. All woody vegetation growing in unconiami-
nated areas, and vegetation not in contact with contaminated soil, must
to be mulched for landscaping use or destroyed in an Air Curtain Des-
tructor. None of this material will be allowed to consume landfill
capacity. All sediments, soil and potentially contaminated vegetation
must be collected and treated in the Thermal Destruction Unit
The remaining significant portion of the Phase II RA is the ground-
water treatment system As previously staled, the aquifer under the site
is contaminated with both chlorinated solvents and PCBs in both a free
oil and dissolved state. This treatment plant will be supplied by
approximately 3500 ft of perforated PVC pipe in a washed gravel bedding
with the bedding encased in a filter fabric outer casing. This piping
network will be placed al depths of up to 25 ft and will be placed in
such a manner as to gravity feed a single wet well collection point for
pumping to the water treatment plant. The treatment plant will consist
of an oil/water separator, a paniculate filler system, two air stripper
columns and a pair of carbon filters. The plant will duplicate air strip-
pers and carbon units to allow a higher initial flow rate by using these
units in parallel. Later, during normal operation, series operation will
be employed to obtain higher effluent quality. Finally, one unit may
be placed in standby status to allow operation to continue when a unit
must be removed from use awaiting service.
The water treatment plant will discharge its effluent 10 the City of
LaSalle wastewater treatment plant and be relatively maintenance free
to allow ease of operation. The operation of this plant will be turned
over to another party when the Phase II RA contractor exits the site.
The groundwatcr treatment effort is expected to be operated an additional
8 to 10 yr.
COMMUNITY RELATIONS
Three factors suggested that community interest would be relatively
high at this site: (1) location in a residential area with over 10,000 resi-
dents; (2) presence of PCB in high concentrations; and (3) location
in the hometown of a state legislator who is an active member of the
legislative committee which reviews the lEPA's budget.
Following a community assessment in the fall of 1983, the first com-
munity relations activity was a joint presentation with U.S. ERA offi-
cials at a City Council meeting in January, 1984. During the RI/FS.
personal interviews, telephone calls, fact sheets and "living room"
meetings were used to identify and respond to community concerns.
378 INCINERATION"
-------
During the RI, soil and ground water sampling revealed high con-
centrations of PCBs on-site and, unexpectedly, on adjacent property.
A literature review and discussion with state and federal potentially
impacted groups as a result of contact with PCB-contaminated soils:
(1) children when playing in yards and, (2) adults when gardening. This
discovery was presented individually to the City Council, LaSalle
County States Attorney, owners of the 26 affected properties and finally
the news media, in personal meetings conducted by two teams of com-
munity relations and technical staff. It was felt that releasing this
information only through a letter or news release would be inadequate
and could create confusion or panic in the community. None of the
residents reacted negatively.
Four risk communication guidelines were followed in designing this
public information effort to inform the community of these results. First,
the Agency wanted to explain what the numbers meant, with a special
emphasis on exposure and routes of exposure. In this case, a literature
review and discussion with the Illinois Department of Public Health
(IDPH) and the Centers for Disease Control (CDC) indicated two
exposure routes: (1) children playing in the dirt and, (2) to adults gar-
dening. Second, IEPA needed to coordinate both internally as well as
between agencies (U.S. EPA, IDPH, etc.) to prevent sending out mixed
or contradictory messages. Third, the IEPA followed a strict sequence
for releasing the results to prevent affected families from hearing about
this problem through the news media first and to enlist others, notably,
City officials and the States Attorney's office, to provide a calming effect
from within the community. Finally, a practice session was held in which
the IEPA developed and critiqued an approach based on a simple,
candid, low-key explanation using words easily understood by each
resident.
A year later the FS, describing proposed remedies, was completed
and presented to the community. At this time, IEPA preferred a remedy
which included mobile incineration. However, several issues existed
which threatened community acceptance of this remedy. A newly elected
Alderman was openly critical of the IEPA. A popular state senator from
LaSalle expressed reservations about incineration of hazardous waste.
A small portion of the community still doubted that the site posed any
health threat and 25 families would have to temporarily leave their homes
while their yards were excavated. Also, at the time this remedy was
being considered, a mobile incinerator had not been successfully used
to destroy on hazardous waste anywhere in the state.
Upper management agreed that if significant opposition from the com-
munity arose towards this remedy, another remedy would be selected.
A fact sheet summarizing the proposed remedies, their advantages and
drawbacks and explaining how to submit comments was distributed
through the mail and made available at the LaSalle City Hall. Small
group meetings with, interested citizens, city and county officials and
local news media were held to discuss the proposed remedies. Following
these meetings, a public hearing and a 3-wk public comment period
were scheduled. A list of anticipated questions was prepared and an-
swers were critiqued before the hearing. Verbal and written comments
received at the hearing and during the public comment period supported
the proposed remedy which included mobile incineration. Both the Al-
derman and state senator, who had previously expressed concern,
provided statements of support for the incineration project.
Yard excavations were conducted during the summer of 1988. The
excavation offer was voluntary. In addition to the 25 families which would
have to temporarily vacate their homes, approximately 80 more resi-
dents were offered partial excavations, primarily of the right of way
area in their front yards. Every affected resident provided access and
cooperated.
Nearly 8 mo of planning preceded the first yard excavation and
hundreds of hours of planning were devoted to identifying and preparing
for the multitude of details which were expected to arise. The families,
many of whom were lifelong residents of the area, faced considerable
anxiety at the prospect of moving out of their homes and seeing their
yards and lawns excavated to depths up to 4 ft.
Food and lodging for the families were provided through the Super-
fund program at no charge to the residents. Some of the special
accommodations arranged by the Community Relations staff included:
professional health care for the blind; sick and elderly; around the clock
security for the vacated residences; strongboxes at a local bank for
personal items; meals to meet different dietary needs as well as dif-
ferent eating arrangements for those on unusual work schedules; care
for pets; customized room arrangements for special family needs; and
schedule adjustments to meet business needs.
The project's Community Relations staff served as liaison between
residents, the contractor and a landscaping subcontractor to coordinate
landscaping changes and respond to differing aesthetic values. Drought
conditions reinforced the need for the Community Relations staff to
assist in advising homeowners about proper care of new sod and land-
scaping.
The safety of nearby residents, particularly children, was a major
concern. Community Relations and other Agency staff met with city
officials, state police officers and officials of the State Department of
Transportation to discuss traffic safety. Truck drivers hauling contami-
nated soil from the excavated yards to the storage area were instructed
to take special precautions as they drove through an adjacent neigh-
borhood where many grade school children resided.
The close working relationship developed with city officials over the
previous 3 yr proved to be very useful in the summer of 1988. City
officials helped with closing streets during excavation, posting new speed
limit signs and maintaining water service despite disruptions caused
by'the excavation work.
An important part of any effort to mitigate fear is providing timely,
accurate information. Tours were arranged for the news media, the com-
munity, government officials and other interested parties, to show and
discuss both the excavation process and the incinerator operation. A
time-consuming, yet worthwhile method of preventing fear is
maintaining regular contact with affected residents. This contact was
accomplished through visits, telephone calls and letters. More than 1,600
contacts were made by the Community Relations staff with the fami-
lies scheduled to have excavation done in their yards.
RESULTS
"Knowing that the PCBs are gone is a tremendous relief," one resi-
dent said. "It takes a lot of worry and fearful thoughts away, regarding
my children and how it would affect them in the future."
Phase I of the RA will be completed in the spring of 1990, and the
Phase II specifications require completion of activity in the summer
of 1993. Including the Immediate Removals, the RI/FS and the Phased
RA, this site will be fully remediated about 2001 to 2003. Although
this process will have taken over 18 yr by the time the RA is complete,
the site will be free of use restrictions and all waste will have been
destroyed or placed in facilities of the utmost environmental integrity
and away from the residents of LaSalle, Illinois. Property values are
already on the rebound in the adjacent neighborhood and the EUC site
can be returned to the local tax base or placed in use for the public
good. Although the process was lengthy, it will result in an effective
and permanent solution to the problem.
INCINERATION 379
-------
Dioxin Destruction on a Small Scale Adjustments
and Achievements
Ritu Chaudhari
Alexis W. Lemmon, RE.
Joseph Towarnicky, Ph.D
Metcalf & Eddy, Inc.
Columbus, Ohio
ABSTRACT
Dioxin has been called "the most toxic chemical known lo man."
As a result, incineration of dioxin-contaminated material requires strin-
gent preparation and extensive safety precautions to assure all involved
parties that operations and procedures are sale. These requirements must
be met irrespective of the amount of material that needs to be reme-
diated.
Reported here is a case study of a successful small-scale remedia-
tion of 190 tons of dioxin-contaminated materials at Fort A. P. Hill—a
job site that was scrutinized closely because it previously had been pub-
licized as a dioxin site in local, state and national media. All site proce-
dures were critically researched to satisfy the review of the U.S. EPA,
the U.S. Army Toxic and Hazardous Materials Agency, the U.S. Army
Environmental Hygiene Agency, the Commonwealth of Virginia
interested individuals and public participation groups.
The remediation was performed according to an engineering and
design report prepared and approved by the Army agencies and U.S.
EPA prior to the commencement of field work. The design report
detailed the plans, equipment, procedures, rationale and methodology
for each activity performed on-site during the remediation. The report
included an evaluation of the effectiveness of a mobile rotary-kiln
incinerator, the required performance criteria for the incinerator, the
necessary sampling, analysis and health and safety considerations and
the procedures necessary to effect the overall implementation of the
thermal treatment of dioxin-contaminated materials
More than 190 tons of dioxin-contaminated material were succesv
fully decontaminated despite numerous obstacles encountered. On such
a small-scale site, however, numerous adjustments were required to com-
plete the remediation. This paper describes how appropriate treatment
technologies were combined with effective site management, engineering
expertise and advance planning strategics to safely remediate a dioxin
contaminated site.
INTRODUCTION
Metcalf & Eddy (M&E) was contracted by O.H. Materials, on be-
half of the U.S. Army Toxic and Hazardous Materials Agency
(USATHAMA), for the Phase I (engineering/design) and the Phase II
(subsequent remedial action) programs at Fort A.P. Hill. Fort A.P. Hill,
a U.S. Army installation located in Bowling Green, Virginia, needed
to dispose of building debris and soils contaminated with acutely
hazardous organic materials—including 2.3,7,8-Tetrachlorodibenzo-p-
dioxin (2,3,7,8-TCDD). On-site incineration using a mobile rotary kiln
had earlier been recommended as the remedial method of choice in
the site feasibility study prepared for USATHAMA.
For Phase I, M&E researched and provided the specific engineering
and design plans needed to assure that each remediation task would
be properly performed. For Pha.se II. M&E implemented the remedial
action—the on-site thermal treatment of the contaminated material-
according to the Phase I Engineering/Design Report. All work was
performed in accordance with U.S. EPA guidance, in compliance with
the CERCLA
The U.S. EPA has mandated that thermal treatment (incineration)
is currently the only sufficiently demonstrated treatment technology
for dioxin-containing wastes (51 FR 1733). However, the successful
application of incineration to a dioxin cleanup at (he Fort A.P. Hill site
differed from the approach used for other dioxin remediatkms, such
as Denney Farms, because a much smaller quantity of material needed
to be cleaned up. Since the tola) volume of waste treated at the Fort
A.P. Hill site was only 190 tons, the remediation was a very temporary
operation. All activities (excluding residue disposal) were completed
in SI days,
A site involving a smaller volume of waste actually requires much
more advance planning (including anticipation of problems that may
arise and resolutions), effective site management and appropriate treat-
ment technology. Due to the short duration of the remediation, any
problem causing system down-time results in a major percentage increase
in the effort and time required for the completion of the project.
Regardless of the volume of material to be treated, implementation
of a dioxin remediation requires all of the same quality assurance and
health and safety safeguards and all of the same tasks as a longer disposal
project. At the Fort A.P. Hill site, M&E had to perform these tasks
under media scrutiny because the site had received national publicity
due to the Boy Scouts of America jamboree that is held at the Fort ev-
ery 4 yr M&E provided two tours of the remediation site for concerned
citizens and newspaper and television reporters. The Army had kept
the public informed throughout the planning for the remediation. As
a result, there was public support for the remediation of the dioxin-
contaminated material, in part because everyone was eager to 'close
this chapter of the dioxin saga' prior to the August. 1989. Boy Scouts
jamboree.
SITE BACKGROUND
The Fon A.P. Hill site is a 76,000-ac Army installation located in
Caroling County near Bowling Green. Virginia. The installation grounds
are now used for Army training purposes and for other events. Fort
A. P. Hill is trot known for the Boy Scouts of America jamboree held
at the site every 4 yr.
From 1962 to 1978, the Army stored the herbicides silvex
(2,4,5-trichlorophenoxy-propionic acid), 2.4-D (2.4- dichlorophenoxv-
acetic acid) and 2,4,5-T (2.4,5-trichlorophenol), 4 of which dioxin is
a known impurity, in Building #225, which has since been demolished.
The herbicide-containing containers corroded, allowing the contents
380 INCINERATION
-------
to leak onto the floor of the storage building and eventually onto the
ground below. The leaking containers were repacked in 1978 and
removed from the base in 1980.
In 1984, the vertical and horizontal extents of silvex, 2,4- D, 2,4,5-T
and dioxin contamination were defined. In 1985, an interim response
action was undertaken in which Building #225 was demolished and the
soils underlying the building were excavated. The excavated soils and
miscellaneous debris were containerized in 35-gal fiberpack drums
which were then overpacked in 55-gal drums and stored in a warehouse
(Building #P01288) inside a secure, fenced area on the Army base.
A feasibility study prepared for the site in 1987 recommended on-
site incineration of the contaminated materials and subsequent disposal
of the residues in a licensed hazardous waste landfill.
WASTE CHARACTERIZATION
Table 1 shows that the site was contaminated by ppm-levels of
2,3,7,8-TCDD and other related constituents. The 2,3,7,8-TCDD isomer
is recognized as the most toxic of the 75 possible dioxin isomers. The
U.S. EPA classifies 2,3,7,8-TCDD and related organic compounds,
referred to as dioxins and furans, as acutely hazardous materials. The
classification and the public awareness of the possible hazards associated
with dioxin, prompted the 1985 interim response action which yielded
the 1,138 drums of contaminated materials which were stored in the
warehouse at the base from 1985 to 1989. The contents of the 1,138
drums were categorized as: dirt, block, wood and miscellaneous, as
noted in Table 2.
Table 1
Chemicals Detected in Excavated Soils at Fort A.P. Hill
Compounds
Range of Concentrations
Detected (ppm)
*2,3,7,8-TCDD
Lindane
0,P'-DDD
P,P'-DDD
P.P'-DDE
0,P'-DDT
P,P'-DDT
Chlordane
PCB (Aroclor 1260)
2,4-D (2,4-dichlorophenoxyacetic acid)
*2,4,5-Trichlorophenol
*Silvex (2,4,5 Trichlorophenoxy-propionic acid)
Pentachlorophenol
ND-1.03
ND-0.008
ND-0.04
ND-0.04
ND-1.88
ND-0.66
ND-2.64
ND-0.10
ND-0.25
NO-0.74
ND-1.98
ND-1.57
ND-0.89
* Chemical constituents for which wastes are listed under F027
METHODOLOGY FOR SITE REMEDIATION
Remediation of materials classified as acutely hazardous must be well-
planned so that the thermal treatment technology selected for the job,
the proposed operating conditions and the data needed to verify com-
pliance with U.S. EPA requirements are all recognized as the best treat-
ment to meet all relevant cleanup and operating standards. These
planning requirements are not significantly reduced even if only a rela-
tively small volume of contaminated material requires remediation.
Table!
Contaminated Materials Stored in Drums in Building PO1288 at Fort A.P.
Hill
Number of Drums
Contents
767
186
120
65
Soil
Cinder block (broken up)
Wood (cut up)
Miscellaneous: VISQUEEN
-------
Everything necessary to do the job had to be trucked to the site. The
site preparation requirements included making provisions for equip-
ment installation, utilities, personnel, equipment decontamination areas.
institutional and containment controls and a support area. Figure 1 shows
the site layout. Utilities required for operations included electricity.
water, telephone service and propane gas as fuel for the incinerator.
Ample lighting was installed at the site to assure a well-lit area during
evening and night operations.
The remedial action site was cleared and covered with gravel. Selected
areas, on which the heavier equipment was to be situated, were surfaced
with asphalt. The site was secured with a chain link fence after the
incinerator and shredder trailers were placed in the exclusion zone.
The area outside the fence was organized into a support area. The
support area consisted of two decontamination trailers, three project
trailers, three water tankers and other support items. Personnel access
to and from the exclusion zone was controlled to require passage through
EXCLUSION
ZONE AREA
SUPPORT
AND
PARKING AREA
OCOCNtM«UTII> MUM ITtKMC MICA
DRAWING NOT TO SCALE
NOTES:
1.
A. TEMPORARY TELEPHONE AND ELECTRIC SERVICE WAS
INSTALLED FROM SOUTH RANGE ROAD TO STE.
B. STORAGE TANKS WERE BROUGHT TO THE STE FOR
SUPPLY OF NATURAL 6AS (INCINERATOR FUELS) AND
POTABLE WATER. WASTEWATER WAS DIRECTED TO STORAGE
TANK FOR OFF SITE DISPOSAL TO A WATER TREATMENT FACILITY.
2. LIGHTING:
A. ALL WORK AREAS WERE PROVIDED WITH TEMPORARY LIGHTING
wrm wTENsmr M ACCORDANCE wrm 29 CFR 1*10.120 (m)
TABLE H-102-1.
Figure I
Site System Layout
PffLANATlQN
• - ELECTRK/TREPHONE POLE
« ELECTWC/TCLEPHONE LME
M FENCE
EOCt OF STONE AREA
ENCLOSED AREA
Illlllllll - EXCLUSION ZONE
382 INCINERATION
-------
the decontamination trailers via the contamination-reduction zone. All
other entry and exit points were restricted.
A high-vertical-clearance wooden building with a wooden floor was
erected in the exclusion zone to house the outlet of the shredder, the
10 bins used to store the shredded feed material, the small forklift, the
weigh scale and the incinerator feed hopper. The building was kept under
slight negative pressure to contain any fugitive particulate emissions
that might be generated during the waste shredding and incinerator feed
operations.
A berrned decontamination pad was installed to accommodate the
staging of drums and decontamination of drums and equipment. The
high-pressure water spray used for decontamination was collected in
an area of the decontamination pad equipped with a sump.
Health and Safety
All personnel at the site were required to abide by the master health
and safety plan prepared for the remediation. All personnel were safety-
trained in accordance with CFR 1910.120 and participated in a medical
surveillance program.
The hazard posed to the workers performing activities at the Fort
A.P. Hill site was generally rated as low. This low hazard rating was
based on the fact that the dioxin-contaminated materials were already
contained in drums, the dioxin was already adsorbed on particulates
and the remediation was short in duration. The low-hazard ranking due
to short-duration operation was a major benefit associated with the small-
scale operation.
The remediation site was segregated into two distinct, fenced work
areas: the drum storage area and the exclusion zone. The fenced drum
storage area was used to store the drummed contaminated wastes, the
drummed incineration residues and the storage tank for the carbon-
treated contaminated water. All workers were required to be in Level
D protective gear when working in the drum storage area.
Workers in the exclusion zone (where the shredding, incinerating and
handling of the contaminated material were performed) were required
to be in Level C protective gear, which includes a full-face air purifying
respirator to protect against particulates. The standard Level C pro-
tective gear was modified to include a second layer of (TYVEK) pro-
tective clothing. This second layer of protection was added to reduce
any off-site migration of contaminants via underclothing and to eliminate
on-site washing of the clothing. All employees who entered the exclu-
sion zone to perform their duties were required to shower in the
shower/locker trailers prior to leaving the site. An M&E health and
safety officer, assigned for each of the three shifts, was responsible for
the well-being of the site workers.
Once the first waste-containing drum was opened, the safety levels
for each zone were formally in effect 24 hr/day. These levels were not
downgraded until incineration was completed, the temporary building
was demolished, the site was cleared and all site samples were found
to be free of dioxin.
Materials Handling Prior to Incineration
Drums were trucked to the incineration site in 31 trips between the
site and the warehouse. Drums were stored within the fenced areas.
Once the incineration began, the drums were staged on the decon-
tamination pad in the exclusion zone, opened and fed to the shredder.
The contents of the 55-gal overpack drums were dumped into the shred-
der hopper with a forklift equipped with a drum-handling attachment
(a grappler). The forklift and grappler replaced electronic arm hydraulic
drum-handling equipment because the latter equipment was bulky and
difficult to maneuver in the limited space of the drum-staging area.
The drums were fed in a pre-determined sequence according to waste-
type categories to assure a homogeneous feed. Drum handling posed
logistical difficulties due to limited space, poor weather conditions and
incorrectly marked drums, which markedly increased the amount of
time required to feed drums. Mislabeled drums caused non-
homogeneous feed—initially resulting in an 80% wood feed that caused
incinerator temperature maintenance problems. M&E adjusted to this
condition by improving quality control documentation measures and
modifying the predetermined drum feeding sequence.
Other difficulties were more difficult to overcome. The contents of
some of the drums were frozen. Record-breaking low temperatures for
the month of March in Virginia created significant difficulties with emp-
tying the drums. Some of the equipment was immobilized by snow and
ice. Fiberpacks leaked and their contents froze against the overpack
drum walls.
Time lost due to these difficulties was minimized due to rapid
adjustments and decisions made on site. Frozen contents of drums were
manually removed by hitting drums with a sledge hammer. Manual
equipment replaced the hydraulic equipment used to feed waste into
the shredder.
M&E planned to re-use decontaminated 55-gal overpack drums to
contain the ash and other residues. However, some of the overpack
drums had corroded and exhibited pinholes from the deterioration
resulting from 4-yr storage of wet materials. Therefore, new drums were
ordered, delivered the next day and used to contain the ash.
Once shredded, materials were collected in metal bins inside the
wooden building. A small forklift, dedicated to operations inside the
building, was used to move the bins from the outlet of the shredder
to the weigh-scale to the incinerator feed hopper.
The shredding initially operated 12 hr daily. This generated suffi-
cient shredded material for the entire 24 hr of operation during mild
weather conditions. Shredding operations were extended to 15 hr in
cold weather due to the drum handling problems noted above. Shredder
operation was extended to 18 hr when feed material had to be reshredded.
Reshredding was required by the incineration subcontractor to reduce
the shredded-feed dimensions to accommodate their ash-discharge con-
veyor requirements.
Incineration
The selected incineration unit had been proven to be capable of com-
plying with all the hazardous waste incineration technical standards set
by RCRA and TSCA. The primary combustion chamber (PCC) that
was used was a countercurrent, cylindrical, refractory-lined rotary kiln.
The non-combustible materials (ash) were discharged through the bottom
ash conveyor.
Combustion of the off-gases generated during the destruction of the
organic materials in the PCC (kiln) was completed in the SCC, a
cocurrent afterburner. The off-gases from the afterburner were cleaned
in a three-stage scrubber system to remove acid gases and other impu-
rities in the gas stream. Cleaned flue gas was exhausted to the
atmosphere.
M&E submitted, as part of the remedial design plan, an explanation
why a trial bum was not necessary at the Fort A.P. Hill site. The demon-
strated performance of rotary kiln systems at well-defined operating
conditions on materials of similar composition was sufficient to verify
that the incinerator would provide effective destruction of the hazardous
organic constituents. Operational controls provided better assurance of
contaminant destruction than a trial burn would have provided.
The start date for the incineration was delayed several times. The
first delay occurred because the truck driver delivered the incinerator
controls/equipment to Bowling Green, Kentucky instead of Bowling
Green, Virginia. It took several days to locate the instrumentation/con-
trols and have it delivered. A second major delay resulted when large
pieces of refractory broke from the kiln during controlled heating of
the unit to temperature. The system was cooled, the refractory was
repaired and the system was reheated to temperature before incinera-
tion could begin.
The incinerator requires as long as 48 hr of controlled heating to
reach operating temperatures. The unit's temperature had to be main-
tained 24 hr/day to avoid repeating the heatup period. Therefore, the
incineration of contaminated materials was performed in a 24 hr/day,
7 day/wk operation. Each subcontractor devised his own staffing
schedule, which allowed for personnel overlap. The staffing schedules
of each subcontractor were staggered to avoid crowding in the decon-
tamination trailer and the contamination reduction zone.
The short duration of the remediation allowed subcontractors to
INCINERATION 383
-------
operate with two or three shifts daily until all work was completed.
This schedule allowed smaller crews to complete the remediation than
would have been required for a long-term remediation.
Incinerated material totaled 190 tons. The feed rate for the system
ranged from 1400 to 2600 Ib/hr. The range of operating temperatures
for the unit is described in Table 4.
Table 4
Incinerator Operating Temperatures
Piraneter
Banoe of Qperttlno Condition!
Proposed Actuil
Primary Combustion Chuiber
Teapenture
Secondary Combustion Clumber
Temperature
1400 1500° F *857 13?3° f
20M - I $00° F ZZI4 2240° F
857° F In > counter-current kiln «as determined to correspond to 1490° F
In a co-current kiln
The minimum PCC operating temperature achieved during reme-
diation differed from the proposed operating temperature. The proposed
conditions, based on tests at other sites, applied to a cocurreni-fired
PCC system. A countercurrent PCC has a different kiln temperature
profile than a co-current kiln. M&E adjusted the discrepancy between
the proposed and actual temperature and verified adequate destruction
temperatures by placing five thermocouples across the outside of the
primary kiln and measuring surface temperatures. A temperature of
300°F on the outside surface of the kiln was calculated to correspond
to a PCC outlet temperature of 857 °F and an inside kiln surface tem-
perature of approximately 1490°F — which is above the proposed
minimum kiln operating temperature.
Another problem encountered during the incineration arose due to
an inconsistency between the two feed-weighing systems (a weigh-scale
in the enclosure and a weigh-belt feeder to the incinerator). The
measured amount of processed waste differed by more than 25%
between the two weighing systems. As the incineration subcontractor
was paid on a pcr-lon-incinerated basis, accurate feed weights were
essential. This inconsistency was resolved by recalibrating the scale
and weigh belt using a known amount of sand. The results of this exercise
determined that the weigh belt reading was 25% too high. Other major
problems included the periodic breakdown of the ash conveyor, pump
failure in the air pollution control system and a buildup of fly ash in
the SCC.
The shear pins on the ash conveyor failed numerous lime during opera-
tions. Each conveyor failure took several minutes to fix. At other times,
pieces of metal caught in the conveyor chain, rendering the conveyor
inoperative. This problem took as long as six hr to repair. An improved
conveyor design might have reduced these problems.
The system was shut down for several days due to excessive buildup
of fly ash in the SCC. The system was cooled, the combustion cham-
ber cleaned and the system was re-heated to temperature.
Even with the numerous difficulties encountered and the unavoid-
able delays, the 190 tons of material were incinerated in 18 days. The
total remediation effort (site preparation to site closure) was completed
in less than 2 mo (except for final residue staging and disposal).
Materials Handling After Incineration
The kiln ash was discharged through an enclosed conveyor and
deposited in re-conditioned 55-gal steel drums A sample was collected
from each filled ash drum as part of a daily composite. The filled drums
were sealed and labeled (date and time) and moved to a temporary area
designated in the hot zone. During the night shift, after shredding
operations had ended for the day, the drums were decontaminated and
moved to the storage area.
At the end of every 24 hr period, each composile drum was sealed
and thoroughly mixed before being re-opened and having a sample col-
lected for analysis. This drum was scaled, labeled and placed with the
other ash drums.
Sampling and Analysis
Daily composite samples of the incinerator residue were collected
and shipped to the analytical laboratory. The samples were analyzed
for polychlorinatcd dibenzo-p-dioxins (PCDDs) and poly-chlorinated
dibenzofurans (PCDFs) on a 48 hr turn-around basis, using U.S. EPA
Method 8280. as specified in 40 CFR 261. Appendix X. Results from
the analysis of the dtoxin congeners were weighted according to Toxicity
Equivalent Factors (TEF). These factors convert data on dioxin/furan
isomers into an equivalent toxicity of 2,3,7,8-TCDD. Toxicity equiva-
lents for dioxins and furans are shown in Table 5.
TableS
Detection Umils/Toxkity Equivalents for CDD* i
ICDFs
Coapouod
Croup i
Tottl TCOOs
Total PeCOOs
ToUl HiCDOs
Total HpCOOs
lotll TCOFs
ToUl PeCOFs
Totil HiCDFs
Totil HpCOfs
Detection
U*lts
S O.t pob
j 0.5 ppb
< ? s ppb
S 100.0 ppb
S 1.0 ppb
S 1.0 ppb
i 10.0 ppb
$ 100.0 ppb
ToUl
Hut*
Halt
Toitclty
Equivalence
2.3.7.8-TCDO
1
0.5
0.04
0.001
0.1
0.1
0.01
0.001
TCF Based m
• Detection
TEF Contribution tt
Klilu
Detection Licit
0.1 ppb
0.25 ppb
0.1 ppb
0.1 ppb
0.1 ppb
0.1 ppb
0.1 ppb
0.1 ppb
O.Hppb
2, 1.7,8- TOXI
Equivalent
MOTE: If each of these groups Is present at I ppb (as In the allowable TOP
requirements) the TEF Mould be 1.7 ppb 2.3.7.8-TCOO equivalent.
For the Fort A. P. Hill site, the acceptance criterion for the incinera-
tion process was set so that the ash or other residues of incineration
had to be proved to contain less than I ppb TEF of 2.3.7.8-TCDD. M&E
was prepared to re-incinerate all materials that did not meet this criterion.
None of the materials required re-incineration.
In addition to the PCDDs and PCDFs tests, the ash samples were
also subjected to the Extraction Procedure Toxicity (EP Tbx) test, as
specified in 40 CFR Part 261 and a Toxicity Characteristics Leaching
Procedure (TCLP) analysis, as specified in 40 CFR Pan 268. The EP
Tox leachate was analyzed for the eight RCRA metals and the pesti-
cides and herbicides detected in the samples of material at the site to
assure that the residues would not be considered toxic. The TCLP ex-
traction was analyzed for PCDDs, PCDFs and chlorinated phenols to
assure that the residues (F028 wastes) could be land disposed.
Dioxin-contaminated soils (F027 wastes) are considered acutely
hazardous wastes Residues resulting from incineration of dioxin wastes
are still considered hazardous because of toxicity and are classed as
F028.
Analytical Results
The residuals generated during the remediation that required disposal
were: ash, SCC ash (the fly ash collected from the SCC), air pollution
control system filter cake and the treated process water. The personnel
protective gear (Tyrek clothing) worn on-site was incinerated daily. The
384 INCINERATION
-------
carbon used to treat the process water and decontamination pad sump
sludge also was incinerated as part of site closure. The ash, the secon-
dary ash and the filter cake were tested for dioxins/furans, EP toxicity
and TCLP analysis.
The post-incineration residues that required disposal were: final
decontamination pad sump sludge, two drums of carbon, personnel
protective gear generated after the incinerator shutdown, shredded wood
and miscellaneous debris. The post- incineration residuals were tested
only for dioxins/furans. The analytical results for these samples are
shown in Table 6, along with the regulatory limits.
Site Closure
Preliminary decontamination procedures commenced soon after
laboratory confirmation that all ash samples from the drummed waste
had met the U.S. EPA dioxins/furans criteria. Heavy equipment, hand
tools and other miscellaneous items that were no longer needed were
decontaminated and removed from the exclusion zone. The secondary
wastes (decontamination pad sump sludge, carbon, vacuum filters, the
floor of the wooden building and miscellaneous debris) were incinerated.
This procedure insured that all of the contaminated materials had been
incinerated.
Table 6
Analytical Results of the Residual Samples at Fort A.P. Hill
Matrix
Soil sample
of general area
Site background
soi I/gravel
No. of Samples
Collected
Analyzed 2.3
1
1
Dtoxins/Furans
,7,8-TCDD equivalent
ND
ND
Chlorinated
Herbicides
NO'1)
ND
EP Toxicitv
Inorganics
(ppm)
-(2)
Pesticides/ Dioxins/
PCBs Furans
ND
ND
TCLP
Phenols
(ppm)
Ash from incineration '
of dioxin-contaminated
material
"Fly" ash from secondary
combustion chamber
Post-incineration (4)
"Fly" ash from secondary
combustion chamber
Filter cake
Waste water
after carbon treatment
13
ND
ND
0.19 ppb
ND
17.5 ppt
ND
ND
ND
ND
Barium ND
Chromium (0.083)
Nickel (0.049)
Barium (0.043)
Barium (0.165)
Cadmium (0.098)
Nickel (0.091)
ND
ND
ND 2,4,6 Trichlorophenol
(NO 0.007)
ND 2,4,6-Trichlorophenol (0.006)
ND ND 2,4,5-Trichlorophenol (0.05)
ND 2,4,6-Trichlorophenol (0.009)
Water after retreatment
with carbon
Carbon from initial
treatment of process water
Carbon from retreatment
of process water
ND
ND
ND
Trash
Wood building
Vood pallets
Final site background
soil/gravel
(1) ND - not detected
(2) (-) - not analyzed
(3) Chromium and nickel
1 0.48 ppb
8 ND
B ND
1 ND
were only detected in one sample.
M)
Regulatory limits:
Inorganics
Barium - 1 ppm
Chromium - 5 ppm
Nickel 16 ppm
Phenol
2,4,6-Trichlorophenol - <0.05 ppm
The ash was stabilized by the disposal facility prior to being landfilled.
Regulatory limit for cadmium is 1 ppm.
INCINERATION 385
-------
Site closure activities continued by dismantling the wooden building.
The shredder was then decontaminated and used to shred the wooden
building and the wood pallets used to store the 1,138 waste drums. The
building and the pallets were shredded separately, sampled, analyzed
and stored in their own roll-off container for disposal. A temporary
enclosure was constructed on lop of the shredder and feed conveyor
to contain fugitive emissions during closure activities.
Upon verification that all residue analyses met U.S. EPA standards.
it was determined that incineration was complete. The incinerator was
run for an additional 2 hr at elevated temperatures to assure complete
contaminant destruction. During this time, the shredder and the
remaining equipment were decontaminated. After all the equipment
was removed from the fenced area, the exclusion zone was thoroughly
cleaned.
The wastewater from personnel and equipmeni decontamination was
treated by carbon adsorption and stored in a tanker. The water was
sampled and found to contain traces of 2,3.7.8-TCDD (at pg/L level).
The water was retreated by carbon adsorption and resamplcd The
carbon used for retreating the water also was sampled. When re-icstcd,
the water and the carbon did not show detectable TCDD.
Samples of various areas in the exclusion zone were collected and
analyzed to verify that the site had not been contaminated during the
remediation.
Residue Disposal
M&E prepared the RCRA waste codes for each residue and the final
disposal destination. A residue disposal scheme, delineated in Table 7,
was approved by USATHAMA and concurred with by U.S. EPA.
Tfcble?
Waste- Code and Final Declination of Residues Generated
at Fort A.P. HOI
TVPC(I)
WASH CODE
OESTIIMTIOH(Z)
Ash
SCC Ash
f1H«r Cakt
Tnatcd Mater
T/veki, Hlu.
Carbon
Su«p Sludge
rote
F028
FOZB
Non Haiardoui
Non-Hazardous
Non -Hazardous
Non-Haierdout
USPCI
USPCI
USPCI
duPont
USPCI
USPCI
USPCI
(I) Shredded w>o4 and decontaminated overpack druei tflipotal MI arranged BJ
olneri,
(?) USPCI •• U.S. Pollution Control. Inc.. In Na/noka. Oklahoma
duPonl •- duPont facility In 0»tptr«t*r, He* Jersey
All residuals generated as a result of the remediation have been proper-
ly disposed at permitted treatment and disposal facilities. The off-site
disposal option is not available to large-scale dioxin sites. However,
off-site disposal allowed the Army to "close the chapter on the dioxin
saga" at Fort A.P. Hill by completely destroying all contamination and
shipping the residues off-site.
386 INCINERATION
-------
Transportable Incineration of
Industrial and Superfund Waste
Thomas F. McGowan, RE.
Consulting Engineer
Atlanta, Georgia
Grady R. Harmon
Williams Incineration Services, Inc.
Stone Mountain, Georgia
ABSTRACT
Williams Incineration Services owns and operates a 15 ton/hr trans-
portable rotary kiln incineration system that was used to burn 9,200
tons of creosote-contaminated soil at the Prentiss Creosote Site in Mis-
sissippi in February 1989. This incinerator will be used next at Bog
Creek Farm, New Jersey.
INTRODUCTION
There is considerable interest in incinerating organic wastes at Super-
fund sites because burning is a permanent solution for organic wastes
and costs are lower than for off-site treatment. The terms mobile and
transportable are frequently heard in connection with on-site incinera-
tion. While no clear delineation exists, the author defines transportable
systems as those systems with more than a 5 ton/hr soil treatment
capacity and construction times of less than 2 mo, with all components
shippable by road with only normal oversize/overweight permits. Mobile
technology is generally restricted to a capacity of less than 2 ton/hr
capacity, is shipped without need for special truck permits and can be
set up and ready to operate in less than a week.
This paper discusses the larger — a system with 15 ton/hr capacity.
A system of this size can compete with smaller systems at a site with
10,000 tons of contaminated soil and reaches an economy of scale above
20,000 tons.
PRENTISS SUPERFUND SITE
The Prentiss Creosote Superfund site was a wood treatment facility
for 19 yr, supplying treated forest products to wide variety of markets.
Plant lagoons containing wastewater and creosote sludge threatened to
overflow into Little White Sand Creek. In March 1987, U.S. EPA Region
IV initiated a cleanup and removal action which consisted of on-site
treatment of pumpable lagoon water, solidification of sludges and ex-
cavation of contaminated soils.
On Dec. 22, 1987, the U.S. EPA signed a contract with Envirite Field
Services (now known as Williams Incineration Services) to incinerate
the soil. The project duration was 14 mo and cost $1,831,642, including
insurance pass through and additional tonnage, for an average of $199/ton
of soil treated.
Operations commenced on-site in April, 1988. After equipment erec-
tion and checkout, incineration of soil began on July 27. The unit
achieved 100% capacity within 7 wk of that date. The trial burn was
completed Oct. 12, production burn finished Dec. 6 and the project
was closed out on Feb. 17, 1989.
In January, 1989, a contract was signed for incineration of soil at the
Bog Creek Farm site in New Jersey. This second project for the
incinerator will involve approximately 22,500 tons of sandy soil con-
taminated with solvents and paint sludge.
CONTAMINATED SOIL CHARACTERISTICS
The Prentiss soil was a soft, sandy, clay-like material with a strong
creosote odor. It contained moderate amounts of gravel, wood, metal
objects and moisture. The creosote sludge had been stabilized with
approximately 2,100 tons of cement kiln dust and fly ash.
Tests on a composite sample (made from over 80 core samples)
showed that the soil had a relatively high heating value of 1,148 Btu/lb
dry basis (Table 1). While the heating value was high, it did not on
the average exceed the heat release limits of the kiln at rated capacity.
Table 1
Proximate and Ultimate Analysis of Contaminated Soil
on a Wet Basis
Percent Weight
Water
Ash
Volatiles
Fixed Carbon
C
H
N
S
Cl
Oj (by difference)
Total
Proximate
10.07%
82.18%
6.95%
0.80%
100.00%
Ultimate
10.07%
82.18%
6.90%
0.46%
0.22%
0.13%
0.12%
100.08%
Using U.S. EPA SW-846 test protocols, the soil was found to contain
seven poly nuclear aromatic hydrocarbons (PAHs) (Table 2). These
organic compounds were consistent with the major creosote constituents
noted in the wood treating literature. No pentachlorophenol or arsenic
compounds were found. Small amounts of inorganic chloride were dis-
covered. Inorganic and organic sulfur were present in small quantities.
INCINERATION 387
-------
Table 2
PAH Analysis of Composite Core Sample
Compound
Phenanthrene
Naphthalene
Anthracene
Acenaphthene
Fluorene
Fluoranthene
Pyrene
mg/kq
1400
1100
1100
1000
900
520
220
SITE OPERATIONS
The site plan is shown in Figure 1. The incinerator was positioned
on high ground adjacent to the stockpile and away from the creek to
avoid flooding.
Figure I
Site Plan Premiss Creosote Site
The soil feed system is shown in Figure 2. Front-end loaders moved
contaminated soil to the staging area, which was a roofed, concrete
pad. A vibrating screen removed material larger than 2 in. Oversized
material (which was not soil) was stockpiled for final disposal by U.S.
EPA. Material less than 2 in size was stockpiled on the pad and then
fed to the hopper of the apron feeder, which in turn fed the wcighbelt.
A magnet over the weighbelt conveyor removed steel scraps which were
stockpiled separately. A sufficient amount of soil was screened and piled
on the pad to last until the next scheduled screening operation.
Although some soil blending occurred when the lagoons were dredged
and the soil was stabilized, the pile was not homogeneous and addi-
tional blending was required to maintain a stable feed system. This
occurred during soil removal from the stockpile, screening operations,
storage on the pad, placement into the apron feeder hopper and when
the soil moved through the apron feeder, weighbelt and rotary dryer/con-
ditioner.
DESCRIPTION OF THE TDU
The general process flowsheet of the thermal destruction unit (TDU)
is shown in Figure 3, and the general TDU equipment layout is shown
in Figure 4. The equipment was produced by Boliden Allis, Inc.
(formerly the Allis-Chalmers Minerals Division), an experienced kiln
and combustion system vendor.
388 INCINERATION
Figure 2
Soil Feed System
Figure 3
TDU Processing Flow Diagram.
Figure 4
TDU Layout
-------
From the apron feeder, the soil dropped onto a weighbelt which
recorded the soil feed rate to the rotary dryer/ conditioner. Integrated
weight totals were used to report quantities of soil processed.
The dryer/conditioner partially dried the soil, broke up large
agglomerated particles, homogenized the feed to the kiln and micropelle-
tized the fines fraction of the soil. Solids moved from the dryer/con-
ditioner in an enclosed conveyor to the rotary kiln, where drying was
completed and creosote compounds were volatilized and burned.
Solids exited the kiln and were conveyed by a chute into a rotary
cooler. Adding water moisturized the decontaminated soil to minimize
dust emissions and promote compaction.
The rotary kiln was fired with two burners. One burner produced
an intense flame (via a custom secondary air scroll) to rapidly dry the
solids and initiate volatilization of the organics. The other burner had
a long flame to burn the volatiles. The kiln was operated to maintain
an exit gas temperature of approximately 1600 °F.
The kiln exit gases passed through a cyclone dust collector, where
much of the entrained particulate matter was removed prior to entering
the secondary combustion chamber. A portion of the gases exiting the
cyclone was diverted to the dryer/conditioner to partially dry the soil,
while the dryer/conditioner exit gases were returned to the inlet of the
cyclone.
After the cyclone, the gas temperature could be increased to as high
as 2200 °F at a residence time of 2 sec in the secondary combustion
chamber. A more typical temperature level in the secondary combus-
tion chamber for this waste was 1700 °F. To ensure complete combus-
tion, a minimum of 3 % excess oxygen was maintained in the secondary
combustion chamber exit gas.
Exiting gases entered a quench tower, where they were cooled by
atomized water, and then entered the baghouse, where particulates were
removed. Dust collected from the secondary combustion chamber,
quench tower and baghouse was conveyed to pug mill where it was mixed
with water prior to discharge onto the belt conveyor. Use of a baghouse
eliminated the production of vast quantities of sludge which would be
produced by a wet scrubber (e.g., high pressure venturi particulate
scrubber). The baghouse also did a better job of removing fine salts
and metals, which can be formed by vaporization in the incineration
process.
After the baghouse, flue gases passed through the 350-hp induced
draft fan to an acid gas absorber where HC1 and SO2 were removed.
Scrubber blowdown water passed through an activated charcoal filter
before being used to cool processed soil. A flue gas sampling condi-
tioning system extracted gases from the stack and fed them into the
continuous analyzers for regulatory and process monitoring and control.
TDU PROCESS PERFORMANCE SPECIFICATIONS
The TDU design criteria meet or exceed the RCRA technical require-
ments of 40 CFR 264. Table 3 summarizes the design data. The burner
input rating of 82 million Btu/hr is the total design capacity (higher
heating value) of all three burners. Most of the fuel value for the system
was derived from the heating value of the creosote contaminated soil.
Process Variables Monitoring
Major variables monitored were the flow of solids and fuel, tempera-
tures, pressures and process gas stream constituents. A weighbelt located
in line with the feed conveyor monitored the feed rate of the soil to
the TDU. The readout in the control room gave the instantaneous feed
rate in tons/hr and integrated totals. The following data were
continuously recorded:
• Waste soil feed rate
• Combustion gas velocity
• Temperature at the exit of the kiln and secondary combustion chamber
• Stack gas carbon monoxide concentration
• Particulate level
• Absorber water flow rate
• Kiln draft
• Dryer draft
• Baghouse inlet temperature
These data were recorded by strip chart recorders and a 48-channel
Table3
TDU Process Performance Specifications
Item value
Waste soil rate, wet basis @ 15% 15 tons/hour
moisture
Solid residence time 45 min minimum
Kiln size, diameter x length 7.5 ft x 45 ft
Kiln outlet gas temperature 1200-2000*F
Secondary combustion chamber
outlet temperature 1500-1800°F
Secondary combustion chamber
residence time @ 2200°F 2 sec
Fuel for burners Propane or natural c
Burner rated capacity, maximum 82 million Btu/hr
data logger. Sheathed type K thermocouples, shielded from direct flame
radiation, sensed the combustion temperatures.
Emissions Monitoring
The TDU is equipped with several continuous gas analyzers. The
oxygen concentration was measured at the kiln exhaust and at the out-
let of the secondary combustion chamber. An extractive flue gas
sampling and conditioning system removed gases downstream from the
air pollution control system for analysis of O2, CO2, CO, TUHC and
NOx. A backup monitor was provided for CO monitoring.
Disposal of Processed Soil and Scrubber Blowdown
Processed soil was placed in conical piles with a volume equal to
24 hr of incinerator output. Samples were taken to ensure that the soil
was clean (less than 100 ppm PAH). Clean soil was moved to the final
disposal site after analysis.
The scrubber liquor and equipment wash water passed through a sedi-
ment filter and an activated carbon adsorber to be stored in a 25,000
gal tank. This water was used to cool the processed soil, eliminating
the need to discharge wastewater.
TRIAL BURN
The trial burn for the incinerator was performed on Oct. 11 and 12,
1988. Naphthalene was used to test overall incineration destruction
efficiency. Naphthalene was selected as the POHC (principal organic
hazardous constituent) because of its relatively high stability rating
(ranked 5th of 320 compounds) in U.S. EPA's Thermal Stability-Based
Incinerability Ranking (revised ranking issued Dec. 14, 1988). No
spiking was done, as naphthalene was present in ample concentrations
in the soil along with a variety of other poly nuclear organic compounds.
The natural soil concentration was measured and used to calculate
incinerator loading and DREs.
Two test conditions were used. The first test was at a kiln tempera-
ture of 1620 °F and a secondary combustion chamber temperature of
1670 °F. The second test used a kiln temperature of 1570 °F and a
secondary combustion chamber temperature of 1710 °F. For both tests,
the average waste feed rate was just above the 15 ton/hr design rate for
the incinerator system (Table 4).
The incinerator stack test results showed that during all tests and under
both test conditions, the incinerator achieved a > 99.998% destruction
removal efficiency (DRE). The DREs were unusually consistent. The
DREs for total PAHs were, without exception, higher than those for
naphthalene. This result suggests that naphthalene was a good choice
for the POHC, for it was more resistant to thermal decomposition than
the average PAH compound. The DRE data from the second test were
all "more than" values, since insufficient POHC was accumulated in
INCINERATION 389
-------
Dibit 4
Trial Burn Test Results Using Transportable
Incinerator at Premiss, Mississippi
Kiln Operating temperature, F
Secondary Combustion Chamber, f
PARAMETER
waste Feed- TPH
Naphthalene Feed. Ib/hr
Total PAH Feed. Ib/hr
Naphthalene DRE, »
Total PAH-DRE. %
Particulate Emleelon Rate
gr/dacf at 71 o*
TEST CONDITION NUMBER 2
Kiln Operating temperature, F
Seconder)* Combustion Chamber. F
PARAMETER BV1I
Matte Feed- TPM I •, »
Naphthalene Feed. Ib hi '4 •
Total PAN Feed. It hr 401
Naphthalene ORE. I ••>•> <
Total PAH-DRE. \ -99.<
Particulate Emmtion Rat*
qr d*c r «t '% o, 001-
F
RUH
1
li. 1
Il.t
140
99.9911
»99 . 999^
RUH
}
15.1
48 . 1
J8-.
99 . 99BB
• 99. »99l.
1670
RUN
'
14.4
9.. 7
411
99.9911
•99.999>
74.7)
14*
99 99i7
>99 . 999*>
RUt<
2
I •> 2
70 J
1170
1 1 10
RUD
1 4 . 6
4«. "
J«J
AYERAt.1
the XAD resin to quantify (he destruction and removal efficiency In
this case, the limits of detection were used to back-calculate DRE An
average paniculate emission rate of 0.012 grains/dscf (corrected to 1%
oxygen) was measured, some six times better than RCRA requirements
The DRE for the total PAH compounds was determined to be
Ł99 999* , at least 10 times better than required by RCRA standards
The test data for DRE and particulars from the trial bum are sum-
marized in Table 4
Hydrochloric acid gas concentration, determined from preliminary
tests, showed that the total HCI emission rate (as calculated from the
theoretical chlorine feed rale based upon soil analysis) was less lhan
4.0 Ibs./hr Actual stack concentrations were negligible, at less than
O.I Ibs/hr
Sulfur dioxide was generated by organic sulfur in (he coal tarv
Uncontrolled levels were expected to be in the 160 ppm range. Con-
tinuous emission monitoring data during the (rial burn showed con-
centrations from 0 to 10 ppm, well below (he stale of Mississippi limit
of 500 ppm.
Ash tests were performed during (he trial burn and on a daily basis
throughout ihe project. In all tests, the ash product contained less than
the required 100 ppm total PAH compounds and was below (he more
stringent land ban requirements. The PAH level was less than detec-
tion limit for each compound (minimum detection limit 0.05 ppm) on
35% of the tests Total PAH was below 5 ppm for 92% of (he ash tests
Maximum total PAH was 35 ppm. experienced on one test during star-
tup of the system.
Problems Encountered and Solutions Employed
The primary problem encountered during startup was the higher than
expected fines content of the soil The stabilization reagent (cement
kiln dust and fly ash) and local clays produced an extremely fine ash.
Approximately 40% of the ash output was from the air pollution con-
trol system (cyclone through baghousc). The conveyors on this system
were undersi/cd and were changed out The secondary combustion
Figure 5
chamber originally had no ash extraction system. A water cooled screw
conveyor was added to remove Tine solids which accumulated there.
Slagging occurred on two initial shakedown runs, resulting in
agglomeration of the ash This problem was solved by running at lower
temperatures and by relocating ihe kiln exit thermocouple which had
been reading low due to seal air leakage.
Initial tests showed inconsistent DREs This problem was due 10 a
duct which collected steam and dust from the product cooler being
vented into the baghousc This vent line was rerouted into the secondary
combustion chamber to prevent bypassing.
General mechanical problems occurred in the material handling
system, principally with the apron feeder. These problems were solved
by upgrading individual drive components and by consistent feeding
of the feed hopper by the front-end loader.
U.S. EPA ACTIVITIES AND
REGULATORY FACTORS
U.S. EPA Region IV has taken a leadership position in the use of
on-site destruction technology to remediate hazardous waste sites. The
Region has attempted to move away from land filling and other tem-
porary solutions and toward destruction and permanent remedies in
accordance with SARA which emphasizes permanent solutions.
Accordingly, incineration was the chosen method for remediation of
the Premiss site.
Since this was a Superfund site, permits, per se, were not required.
However, data requirements were essentially the same as those for a
formal Part B permit. These data were submitted in a work plan which
was reviewed and approved by U.S. EPA. Regulations were primarily
federal, with the major State of Mississippi concern being SO,
(limited to 500 ppm).
BOG CREEK FARM PROJECT
The second use for the transportable incinerator will be the inciner-
ation of approximately 22.500 tons of solvent and paint sludge con-
laminated soil ai the Bog Creek Farm Superfund Site in New Jersey.
The project is being contracted by the U.S. Corps of Engineers. U.S.
EPA Region II (New York) is in charge of the site. The contract was
signed (he first week of January, 1989.
The wastes were deposited in trenches by a past owner. Solvents and,
to a lesser degree, metallic contaminants are entering the groundwater.
Chemical Waste Management is the prime contractor on this $14 mil-
lion project and will coordinate the activities of the incineration, exca-
vation and water treatment subcontractors. As of July. 1989, all plans
390 INCINERATION
-------
were submitted to U.S. COE, U.S. EPA and NJDEP and approvals were
being issued. The project will take 400 days to complete from issuance
of the notice to proceed.
CONCLUSIONS
The Prentiss project marked the first field remediation action involving
the incineration of creosote wastes. The experience gained is directly
applicable to remediation efforts for sites which have soils containing
significant amounts of stable organic contaminants. The incinerator has
been designed to meet RCRA and TSCA regulations and brings state-
of-the-art technology to field remediation of all types of organic
hazardous waste.
DISCLAIMER
Because the preceding paper has not completed the U.S. EPA tech-
nical and administrative review, it does not necessarily reflect the views
of the Agency and no official endorsement should be inferred.
LEnvirite Field Services, Inc., was acquired in December, 1988, by
Williams Environmental Services and is now being operated as Williams
Incineration Services, Inc.
INCINERATION 391
-------
Mobile Thermal Volatilization System for
Hydrocarbon-Contaminated Soils
Gregory. J. McCartney, P.E.
George H. Hay
O.H. Materials Corp.
Findlay, Ohio
ABSTRACT
In view of the current major environmental concern about leaking
underground storage tanks and spilled fuel, O.H. Materials Corp.
(OHM) undertook the development of an innovative system for the ther-
mal treatment of petroleum hydrocarbon-contaminated soils. The
objective in the development of the Mobile Thermal Volatilization
System (MTVS) was to design a technically-sound and cost-effective
method for the on-site treatment of hydrocarbon-contaminated soils
This technology is applicable to the treatment of organic compounds
with boiling points of up to 800°F including diesel fuel, heating oil
and high boiling point anomalies. The equipment designed for the
cleanup process was limited to only non-halogenated hydrocarbon com-
pounds which eliminates the need for a high-temperature afterburner
and acid scrubbing equipment.
A prototype unit was designed in the winter of 1986 and construc-
tion was completed in August, 1987. The prototype unit has been
operated at several locations from which valuable production and
performance data were obtained and used in the design of the second
MTVS.
The second MTVS was designed to maximize production with a
design feed rate of 10 tons/hr. The air pollution system consists of a
hot cyclone, an afterburner (1400°F) and a veniuri scrubber. The tech-
nical basis of the design of this unit will be discussed.
INTRODUCTION
Leaking underground petroleum storage tanks are a major threat 10
groundwater supplies. When the tanks are removed and replaced, the
contaminated soil must also be addressed. Presently, these soils are
not regulated by the U.S. EPA and the authority for their treatment and
disposal has fallen to the individual slates. Several states require (he
disposal of these soils in hazardous waste landfills which can cost from
$100 to S220/ton plus the cost of transportation and future liability
Low temperature thermal treatment was identified by OHM as a
potential on-site treatment process for hydrocarbon-contaminated soils
The objective in the development of the Mobile Thermal Volatilization
System (MTVS) was to design a technically-sound and cost-effective
method for the treatment of hydrocarbon-contaminated soils The equip-
ment designed for the cleanup process was limited to only non-
halogenated hydrocarbon compounds which eliminates the need for a
high-temperature afterburner and acid scrubbing equipment.
The goal of the equipment design was to achieve maximum processing
throughput in a highly mobile, single trailer system of Icgal-si/cd load.
The prototype unit was designed in the winter of 1986 and the con-
struction was completed in August, 1987. The prototype unit has been
operated at six locations from which valuable operational and per-
formance data have been obtained and used in the design of the second
MTVS.
EQUIPMENT
The theory behind thermal volatilization consists of heating the soil
to the temperature at which the organic contaminant is vaporized and
removed from the soil. The vapors are then passed through an after-
burner which oxidizes the organic constituents to carbon dioxide and
water. The off-gases are then cooled and scrubbed to remove paniculate
matter before discharge to into the atmosphere. The technical specifi-
cations which were developed during the design phase of the project
are contained in Table I.
Table I
Prototype Votittalkm SvMem Technical Specifications
System Design:
Maximum feed rate
Particle size
Maximum hydrocarbon content
Moisture content
Soil discharge temperature
Primary heat capacity
Solids retention tine
Secondary temperature
Secondary volume
Secondary heat capacity
Secondary retention tine (Bin)
Vater requirements
Performance:
Design VOC destruction
Particulate emissions
Soil Cleanup Quality:
Hydrocarbon content
Benzene
Toluene
Xylene
12,000 Ibs/hr
up to 3 inches
5 percent
15 percent
400°P
6.0 MMBtu/hr
15 minutes
1,400°P
160 ft3
3.0 MMBtu/hr
0.6 seconds
6 gpm
99X
<0.04 gr/DSCF
<100 ppn
-------
The thermal volatilization process consists of a feed hopper which
regulates the flow of material into the primary chamber. The primary
chamber is directly heated to approximately 800 °F using natural gas
or propane. This 800-°F temperature results in a 300- to 600-°F soil
discharge temperature. The required soil discharge temperature is
dependent on the vaporization characteristics of the hydrocarbon
contaminants.
The conveyance system on the primary chamber of the prototype unit
is a 4 by 8 ft pugmill. The pugmill consists of two shafts with paddles
attached at a slight incline that rotate at approximately 60 rpm which
aids in conveying and mixing the soil. The burners are mounted over
the pugmill and directed down toward the soil.
The conveyance system on the second unit consists of a rotary drum
which improves the heat transfer efficiency in the primary chamber.
The rotary drum also allows for greater soil discharge temperatures
and throughput.
Both units have afterburners which have been designed for an
operating temperature of 1,400°F and a gas retention time of 0.6 sec.
This temperature was chosen based on the auto-ignition temperature
of the anticipated hydrocarbon compounds as listed in Table 2.
Table 3
Results Summary of Initial Testing of the
Thermal Volatilizing System
Waste
Description
Feed Rate
Discharge
Temperature
Test 1
Soil spiked
with 3.8 percent
No. 2 diesel
220 degrees F
Test 2
Soil spiked
with 1.9 percent
No. 2 diesel
and 1.9 percent
leaded gasoline
4 tons per hour 4 tons per hour
440 degrees F
Iable2
Autoignition Temperature of Some Common Organic Compounds
Compound
Benzene
Carbon monoxide
Cyclohexane
Ethyl benzene
Kerosene
Methane
Propane
Toluene
Xylene
Temperature
(degrees Fahrenheit)
1075
1205
514
870
490
999
974
1026
924
Hydrocarbon
Reduction
86 percent
99.3 percent
The second project took place at a service station in Cocoa, Florida,
where leaking underground gasoline storage tanks had contaminated
approximately 800 yd3 of fill. Before the tanks were removed, a new
set of tanks was installed in another location on the site. This enabled
the service station to continue operation throughout the remediation
process. The State of Florida required stack emissions testing of the
unit for paniculate and organic emission at the beginning of the project.
The results of this testing are contained in Table 4.
Table 4
State of Florida Required Emission Tests
Soil Contamination
Feed Rate
755 ppm total hydrocarbons
12 percent moisture
5 tons per hour
The air pollution control equipment consists of a hot cyclone which
is used to remove the majority of the particulate from the gas stream
before it enters the wet scrubber. The gases are quenched in a stainless
steel-lined duct before passing into a venturi scrubber which is followed
by a mist eliminator. The cleaned gases are then exhausted from the
system by an induced draft fan which maintains a negative draft on the
entire system. The fan also controls fugitive emissions from the system.
OPERATIONAL EXPERIENCE
The prototype unit has been in operation since September, 1987, when
it was first tested in the Fabrication Shop. The results showed that the
system achieved a significant reduction in soil hydrocarbon contami-
nation. A summary of these results is presented in Table 3.
A total lead analysis was also performed on the feed sample and the
concentration was found to be indistinguishable from background. No
lead was detected in the scrubber water at a detection limit of 10 /x/L.
Stack emissions testing was not performed during this preliminary
program due to schedule constraints.
The first field use of the equipment was for the moisture reduction
of a recyclable sludge. A metals fabrication facility was closing its
primary settling lagoon. The sludge in the lagoon contained a high con-
centration of titanium, which could be recycled. A mobile filter press
was used to the waste, producing filter cake. The prototype MTVS was
then used to reduce the moisture of the filter cake from 40% to 5 %.
Soil Discharge
Temperature
Particulate
(corrected to
7% oxygen)
Volatile
Emissions
(by VOST)
Opacity
Soil Quality
340 degrees Fahrenheit
0.011 gr/dscf average
0.31 Ib/hr
Benzene
Toluene
Ethylbenzene
Xylenes
0.0 percent
22.2 ug/m3
16.0 ug/m3
3.1 ug/m3
15.0 ug/m3
Total Petroleum <100 ppm
Hydrocarbons by GC
Aromatic Volatile <100 ppb
Organics
INCINERATION 393
-------
Based on this testing, a Florida statewide permit to install was issued
for the system in November 1988.
Using the data and operational experience gained with the prototype
unit, a second unit has been designed and constructed. This unit is simi-
lar to a rotary dryer which is thermally more efficient than a pugmill.
The construction of the MTVS II was completed in September, 1988.
PROCESS TESTING OF THE MTVS II
The new unit is designed with a 5-ft diameter rotary drum for con-
veyance of the soil. The use of this rotary drum improves the thermal
efficiency of the unit and increases the soil discharge temperature and
processing rate. A process flow diagram is shown in Figure I.
Cl«»n Combutllon
QM
Table 5
Technical Specification for MTVS II. Rotary Drum
Thermal Treatment System
Figure I
Mobile Thermal Volatilization System
The flow of gases from the primary chamber enters a high-efficiency
cyclone where the majority of the paniculate is removed. The gases
then flow into the afterburner which is followed by a wet scrubber. The
scrubber is mounted on a separate trailer and consists of a quench section
and pumpless venturi, The technical specifications for the second MTVS
are contained in Table 5.
The initial soil testing of MTVS II was conducted at the manufac-
turer's facility in Connecticut. The results of this soil testing are illus-
trated in Table 6.
The initial testing indicates that the technology will successfully
remove gasoline and diesel fuels from contaminated soils
After the unit was delivered, a State of Ohio compliance test was
conducted. The test consisted of three trial runs conducted with soils
spiked with a combination of diesel fuel and gasoline. Samples were
collected from the exhaust slack, feed hopper, soil discharge screw and
scrubber water. These samples were subsequently analyzed for total
petroleum hydrocarbons, benzene, toluene and total xylenes.
The results of the stack emissions testing are contained in Table 7.
These results demonstrated compliance with ihe State of Ohio Air Pol-
lution regulations. The results of analysis performed on the soil and
scrubber water are shown in Table 8.
CONCLUSION
The second generation MTVS has been successfully used to treat
hydrocarbon-contaminated soil at several sites in Ohio and Pennsyl-
vania. The treated soils have been placed back into the excavation areas
after analytical verification that the cleanup criteria were obtained.
The use of a low-temperature thermal treatment unit for hydrocarbon-
contaminated soils is now a viable alternative to off-site land disposal.
The remediation of underground storage tank leaks and transportation
spills can be completed on'site with minimal future liability.
System Design;
Max!mum feed rate
Particle size
Haximua hydrocarbon content
Moisture content
Soil discharge temperature
Primary thermal rating
Solids retention tine
Secondary temperature
Secondary thermal rating
Secondary retention time
Water requirements
Performance!
Design VOC destruction
Particulate emissions
Soil Cleanup Quality:
Hydrocarbon content
Benzene
Toluene
Xylene
20,000 Ibs/hr
up to 3 Inches
5 percent
15 percent
500-800 of
10 HHBtu/hr
10-60 minutes
1400-1600 oP
10 KMBtu/hr
0.6 seconds
12 gpm
99 percent
<0.04 gr/DSCP
<50 ppm
<0.1 ppm
<0.1 ppm
<0.1 ppn
Table 6
Preliminary Test Results of MTVS II. Treatment of
Hydrocarbon-contaminated Soil
TEST 1
TBST 2
TEST 3
Vaste Soil with Soil with Soil with
Description l.SZ gasoline 2.SX diesel 2.5* diesel
Percent 8.0 percent 12.5 percent 6.0 percent
Moisture
Peed
Rate
6.5 tons/hr B.2S tons/hr 6.0 tons/hr
Discharge 420 °P
Temperature
412 °F
550 °F
Discharge <100 ppb VOA <50 ppn TPHC <50 ppm TPHC
Soil
Quality
394 INCINERATION
-------
Table 7
Demonstration Test Stack Emissions Results MTVS II for the
Rotary Drum Thermal Treatment System
ELEMENT
TEST 1
Particulate 0.03
Collected
(gr/dscf)
Particulate 1.1
Collected
(Ibs/hr)
Benzene
(mg/m3)
Toluene
(mg/m3)
Xylenes
3
(mg/m )
<24
<24
<24
TEST 2
0.04
1.23
87
<3
62
TEST 3
0.04
1.14
79
Non-Methane 71
Hydro Carbons
(ppra)
Methane (ppm) <3
Carbon 34
Monoxide (ppm)
REFERENCE
1. Brunner, C. R., Incineration Systems Selection and Design, Van Nostrand
Reinhold Co., New York, NY, 1984.
<3
68
Tables
Demonstration Test Results for Tests on MTVS II
Rotary Drum Thermal Treatment Systems
TEST 1
Waste Feed:
Waste Feed Rate 12,380
(Ibs/hr)
Calculated 309
Hydrocarbon
Feed Rate
(Ibs/hr)
Moisture (%) 10.0
Benzene (ppm) 9.50
Toluene (ppm) 198
Ethylbenzene (ppm) 46.0
Total Xylenes (ppm) 349
Ash:
TPHC (ppm) 144
Benzene (ppm) <1.0
Toluene (ppm) <1.0
Ethylbenzene (ppm) <1.0
Total Xylenes (ppm) <1.0
Scrubber Water;
Benzene (ppb) <1.0
Toluene (ppb) <1.0
Ethylbenzene (ppb) <1.0
Total Xylenes (ppb) <1.0
TEST 2
16,203
405
10.2
6.47
73.1
20.0
129
382
TEST 3
17,500
437
9.5
2.96
48.4
18.1
128
505
INCINERATION 395
-------
Contaminated Soil Remediation by Circulating Bed Combustion
Demonstration Test Results
Brenda M. Anderson
Robert G. Wilbourn
Ogden Environmental Services, Inc.
San Diego, California
ABSTRACT
The Circulating Bed Combustor (CBC) is an advanced generation
of incinerator that utilizes high velocity air to entrain circulating solids
in a highly turbulent combustion loop. Because of its high thermal
efficiency, the CBC is ideally suited to treat organic wastes with low'
heat content, including contaminated soil. This paper discusses the
development of the CBC technology for the treatment of contaminated
soils and its application to site remediation. The CBC process, pilot
plant and transportable field equipment units are described and the
results of four recent tests are presented.
In March. 1989. a Superfund Innovative Technology Evaluation
demonstration test burn of McColl Superfund site soil was conducted
in Ogden Environmental Services' Circulating Bed Combustion research
facility. In Stockton. California, two performance tests of soil con-
taminated with fuel oil were conducted during February and July of
1989. A demonstration test of PCB-contaminated soil was performed
in September of 1988 at Swanson River. Alaska, resulting in the June.
1989 issuance of a U.S. EPA TSCA permit for operation. The results
of these tests demonstrate that the Circulating Bed Combuster meets
or exceeds all applicable California. Alaska and U.S. EPA criteria for
each of these projects.
COMBUSTION
CHAMBER
LIMESTONE
FEED
SOLID
FEED
STACK
ASH CONVEYOR
SYSTEM
Figure 1
Schematic Flow Diagram of Circulating Bed Combustor for
Soil Treatment
INTRODUCTION
The Circulating Bed Combuster (CBC) is ideally suited to treat feeds
with low heat content, including contaminated soil (Fig. I). Soil is
introduced into the combustor loop at the loop seal where it contacts
hot recirculating soil from the hot cyclone. Hazardous materials adhering
to the introduced feed soil arc rapidly heated and continue to be exposed
to high temperatures throughout their residence time in the ceramic
lined combustor loop. High velocity air (14- to 20-fi/s) entrains the feed
with circulating soil which travels upward through the combustor into
the cyclone. Retention tunes in the combustor range from 2 sec for gases
to 30 min for larger feed materials
A cyclone separates the combustion gases from the hot solids, which
return to the combustion chamber through a proprietary, non-mechanical
seal. Hot flue gases and fly ash that arc separated at the cyclone pass
through a convectivc gas cooler and on to a baghouse filter which
removes the fly ash. Filtered flue gas then exhausts to the atmosphere.
Heavier particles of purified soil remaining in the lower bed of the com-
bustor are removed at a controlled rate by an ash conveyor system.
AJS a consequence of the high turbulence in the combustion zone.
temperatures around the loop (combustion chamber, hot cyclone, return
leg) are uniform to within ±50°F over the typical operating range of
1450 to 1800 T The uniform low temperatures and high solids turbu-
lence also help avoid the ash slagging that is encountered in other types
of incinerators.
Acid gases formed during destruction reactions are rapidly captured
in the combustor loop by limestone that is added directly into the com-
bustor with the feed. HCI and SO, that are formed during the com-
bustion of chlorine- and sulfur-bearing wastes react with limestone to
form dry calcium chloride and calcium sulfate. Due to the high com-
bustion efficiency attainable in a CBC. an afterburner is not needed.
In more than 90% of the cases studied to date, post-combustor acid
gas scrubbing is not required. Emissions of CO and NO, are con-
trolled to low levels by the excellent mixing resulting from turbulence,
relatively low temperatures and staged combustion which is achieved
by injecting secondary air at locations ascending the combustor. Because
of the design and operating features, the CBC can attain required des-
truction and removal efficiencies (DREs) for both hazardous wastes
(ORE ^99.99%) and toxic wastes (ORE ^ 99.9999%) at temperatures
below those used in conventional incinerators which typically burn at
temperatures greater than 2000°F.
The Circulating Bed Combustion technology is well developed and
is being applied on two contaminated soil site remediation projects that
will clean over 80,000 tons of contaminated soils. OES and its predeces-
sors have pursued a systematic technology development and an appli-
cations approach comprised of the following elements:
• Definition of treatable soil contaminant waste types
396 INCINF.RATION
-------
fEEOIRANSPORl tlOWR
COUCH!.. »«;"
-trJ*-^XD
^^
m FAM
^—J
coaiiNG
VVA1ER OUT
tUCKll
ElEVAlUR
IKJUiOtllOrukirSIAllON
Figure 2
Research Facility CBC Process Schematic
• Fifteen years of fluidized and circulating bed pilot plant testing
• CBC performance demonstrations in private and governmental
programs, including the Superfund Innovative Technology Evalua-
tion (SITE) program
• Extensive permitting activities
• Design, engineering, fabrication, deployment and operation of
modular transportable CBCs for hazardous waste site cleanups
OES CIRCULATING BED COMBUSTOR UNITS
Research Facility
Ogden Environmental Services (OES) research CBC is the heart of
an integrated, highly flexible waste combustion demonstration facility
located in San Diego, California. Initial CBC-soils treatment develop-
ment and engineering studies were carried out in this 16-in., 2 million
Btu/hr CBC. The test data obtained were used to design the larger, trans-
portable CBCs. The configuration of the research CBC is shown
schematically in Figure 2. Figure 3 is a photograph of the 16-in. CBC
unit.
Transportable CBC's
The transportable 36-in., 10 million Btu/hr CBC consists of seven
structural steel modules that contain the process equipment and provide
the structural framework of the CBC. The modules do not exceed meas-
urements of 8.5 ft wide, 10.3 ft high and 35 ft long. As a result, the
modules can all be transported on single drop trailers that do not require
special highway transportation permits. The CBC cyclone and com-
bustor are mounted to the top of one of the structural modules. When
erected, the transportable CBC itself sits on a pad of 30 by 50 ft and
is approximately 60 ft in height. In field operations, the transportable
CBCs are incorporated in a complete system layout which includes
ancillary equipment units and transportable buildings, e.g., a control
room, a motor control center, an analyzer room and a chemistry sup-
port laboratory (optional). Figure 4 is a photograph of the OES Stock-
ton Project field assembled, transportable 36-in. (i.d.) CBC unit.
MEDIA TREATED
Circulating Bed Combustion is widely applicable to many hazardous
waste forms. Solids, including contaminated soils, liquids and sludges,
are treated with equal facility by using the appropriate feeding systems.
OES has conducted extensive pilot-plant and field-unit testing on soils
contaminated with hydrocarbons and chlorinated hydrocarbons.
PROCESS WASTE STREAMS
The CBC process typically produces solids (i.e., bed and fly ash)
and stack gas, as shown in Figure 1. The composition of the stack-gas
system effluent must meet U.S. EPA and other governmental require-
ments in accordance with permitted conditions. All the CBC incinera-
tion tests of contaminated soils verify that the purified soil treated by
the CBC is non-hazardous with respect to organic residuals. Since most
metals migrate to the ash during combustion, the disposition of ash
is specific to each waste feed case and must be determined on an in-
dividual basis. For most organic-contaminated soil sites, the ash
produced by the CBC meets the criteria for redeposition on-site. At
both Stockton and Alaska the purified soil is deposited on-site as non-
hazardous soil. It is expected that McColl ash will be non-hazardous
since the SITE program McColl ash leach testing found organic and
metal concentrations to be well below the regulatory limits. Post-
combustion fixation processes may occasionally be required if the ash
metals content or leachability exceeds permissible levels.
RECENT TESTS WITH CONTAMINATED SOIL FEEDS
Through the two large-scale site remediation projects that will treat
over 80,000 tons of contaminated soil and the pilot-scale operations
during SITE program testing, OES has proven the effectiveness of trans-
portable CBCs by locating and operating them cost-effectively in
demanding environments. Every regulatory requirement for site oper-
ations has been meet. The transportable CBCs have been operated in
weather as cold as -40 °F and as high as 110 °F. The ruggedness of the
units has been demonstrated by mobilizing and operating successfully
INCINERATION 397
-------
16-m CBC in (he
in a remote and ecologically sensitive wildlife refuge. OES has main-
tained high levels of availability through (he use of careful logistics plan-
ning that includes design factors, maintenance and supply planning.
A description of the projects is given below.
Superfund Innovative Technology Evaluation Program
In 1986 the CBC was selected by U.S. EPA for a demonstration under
the SITE program. Contaminated soil from the McColl Superfund site
in Fullerton, California was selected as the waste Iced for the demon-
stration project. Due to multiple delays encountered in (he securing
all of the required permits, it was not possible to conduct the planned
feasibility demonstration test until this year.
Figure 3
San Dicgu Research Facilit)
The treatabilily study was conducted during March, 1989. The demon-
stration approximately 31 hr over a 4-day period. The project was moni-
tored by (he U.S. EPA. the California Department of Health Services
and the San Diego County Air Pollution Control District. A total of
7,500 Ib of contaminated soil were processed through the CBC of which
4.700 Ib were actual McColl waste. The materials that were processed
included: unblended waste, waste blended with clean sand and
unblended waste spiked with carbon telrachloride. The materials were
processed without difficulty.
Samples of the waste feed, fly ash, bed ash and stack gas were taken
by a U.S. EPA contractor for analysis. The samples were analyzed for
organic compounds, (including dioxins and furans), metals, criteria poi-
398 INCINERATION
-------
Figure 4
36-in. Diameter Transportable CBC at Stockton Site Leaking
Underground Storage Tank Remediation Photo
lutants and physical properties.
Table 1 contains operating conditions and data on stack criteria pol-
lutant and acid-gas emissions from the triplicate testing. Test, 1 fed
325 Ib/hr of McColl waste blended with sand. Tests 2 and 3 processed
waste alone and waste spiked with carbon tetrachloride. The tests were
performed at the target temperature of 1700 °F at lower than maximum
throughput. While permit limits on this test precluded the evaluation
of feed rates higher that 200 Ib/hr of waste, the successful results indi-
cate that processing in a commercial CBC is feasible. The criteria
pollutant and acid-gas release data obtained are well within federal,
state and local requirements. Paniculate emissions were more than ten
times lower than the 0.08 gr/dscf corrected to 7% oxygen federal limit.
Combustion efficiency and ORE were consistently higher than the
regulatory limits.
The U.S. EPA has officially released preliminary data which has been
checked to assure that it meets U.S. EPA standards and the complete
demonstration test report will be available in late 1989.
The results show organic material was effectively destroyed as
exhibited by the ORE value (99.9937%) shown in Table 1. Complete
stack and ash analysis for volatiles, semi-volatiles and metals indicate
that no significant levels of hazardous compounds left the CBC system
in the stack gas. Ash analysis indicate that no significant levels of
hazardous organic compounds remained in the bed and fly ash material.
A Toxicity Characteristic Leaching Procedure (TCLP) test was per-
Table 1
McColl SITE Itests: Operating Conditions
Test Conditions
Combustor temp, °F
Residence time, sec
Soil throughput, Ib/hr
Carbon tetrachloride, Ib/hr
Flue gas oxygen, dry %
CO emissions, ppm
HC emissions, ppm
SO2 capture, %
NO, emissions, ppm
Carbon dioxide, dry %
HC1 emissions, Ib/hr
Particulate gr/dscf at 7% 02
Combustion efficiency, %
ORE, (%)
1721
1.54
325
0
11
30
5
>95%
49
9.9
<0.0090
0.0041
99.97
1726
1.52
170
0
9.9
30
1
>95%
58
11.9
<0.008S
0.0044
99.97
1709
1.55
197
0.22
11.8
26
2
>95%
48
9.2
<0.0098
0.0035
99.97
99.9937
INCINERATION 399
-------
formed on the McColl CBC ash. Arsenic, selenium, barium, cadmium,
chromium, lead, mercury and silver leachabilitics were found to be
well below the federal requirements (40 CFR Part 268).
While the McColl site waste averages 8% sulfur, the soil selected
for this testing ranged between 4 and 5% sulfur to comply with research
facility permit feed concentration limits. The efficiency of in-situ sulfur
capture using limestone was >95%. Further quantitalion is not possi-
ble as the sulfur dioxide continuous emissions monitor low range was
not performing to specification.
Waste, limestone, ash and flue gas were analyzed for the following
17 metals: arsenic, antimony, barium, beryllium, cadmium, chromium.
copper, lead, mercury, manganese, nickel, selenium, silver, thallium,
zinc, cobalt and tin. Table 2 lists the partitioning results for the six
metals found in all three waste samples at more than twice the minimum
detection limit for that metal. Waste concentrations for these six mctaK
were not high, ranging from 3 mg/kg cobalt to 211 mg/kg manganese.
making it difficult to trace their fate. Total metal mass exiting the process
further illustrate the small quantities. Mass balances around the me-
tals did not account for more than 91% of the input and typically ac-
counted for about half of the feed metal content. This is to be expected
when low concentrations of naturally occurring metals arc measured
in complex soil matrices.
Two figures are listed as " <" to indicate that the metal was not
detected. Detection limits were used to quantify these partition fractions.
Partition data is therefore to be used with caution. As expected, the
fly ash consistently showed both higher metal concentrations and metal
mass flows than the bed ash. Zinc data is not presented due to inter-
ference from zinc coatings in the equipment.
Table 2
McColl SITE Test: Metals Partitioning
Total Flyash
ng/nr Fraction
Bod ash
Fract ion
Flue Gas
Fraction
Test 1 Copper
Nickel
Cobalt
Chromium
Ba r 1 UB
Manganese
Test 2 Copper
Nickel
Cobalt
Chromium
Barium
Manganese
Test 3 Copper
Nickel
Cobalt
Chromium
Barium
Manganese
666
1350
226
3206
6110
15667
122]
1171
204
2932
6435
2074 1
874
VJ2
150
1630
4157
11682
0. 769
0.714
0.765
0.843
0.832
0.761
0.938
0.904
0.906
(J.948
0.937
0. 958
0.949
-------
Figure 5
PCB Remediation at Swanson River Alaska Project Site, Aerial Photo
liable 3A
Swanson River Tests: Operating Conditions Tests 1 through 3
Table 3B
Swanson River Tests: Operating Conditions Tests 4 through 6
Test Conditions
Combustor temp, °F
Residence time, sec
Soil throughput, Ib/hr
Soil PCB cone., ppm
Flue gas oxygen, dry %
CO emission, ppm
HC emissions, ppm
S02 emissions, ppm
NO, emissions, ppm
Carbon dioxide, %
HC1 emissions, Ib/hr
Particulate gr/dscf at 7% O2
Combustion efficiency, *
DUE, %
Test 1
1620
1.68
8,217
632
7.1
12
2
16
89
8.8
1.49
0.0072
99.980
>99. 99993
Test 2
1606
1.68
8,602
615
7.4
11
2
15
88
8.7
1.08
0.0065
99.990
>99. 99992
Test 3
1620
1.67
8,603
801
6.9
17.5
2
13
88
8.6
1.37
0.0093
99.985
>99. 99997
Test Conditions
Combustor temp, °F
Residence time, sec
Soil throughput, Ib/hr
Feed PCB cone . , ppm
Flue gas oxygen, dry %
CO emissions, ppm
HC emissions, ppm
SO2 emissions, ppm
NOK emissions, ppm
Carbon dioxide, %
HCl emissions, Ib/hr
Particulate gr/dscf at 7% O2
Combustion efficiency, %
ORE, %
Test 4
1701
1.52
8,194
289
6.2
8.7
2
27
82
8.8
1.42
0.0120
99.990
>99. 99996
Test 5
1693
1.47
9,490
608
6.1
10
2
21
90
8.9
1.57
0.0190
99.990
>99. 99994
Test 6
1686
1.53
9,555
625
8.1
12.5
2
20
95
8.8
1.21
0.0182
99.990
>99. 99993
INCINERATION 401
-------
Table 4
Transportable 36-in. CBC Monitoring Equipment
Naee/ Function
Plue qaa O2 Probe
Extractive Gee Analyala
oxyqen
CO
CO]
NO/HO,
SOI
HC
coabuetor Preeeure
varloua
Teaperaturea
Varloua
Soil reed Rate*
Principal or
Operation
tlrconla Call
Pa aaagnetlc
In rared
In rared
Ch • LuBina.
In rared
Fl Be ionliation
Olaphraov
Thermocouple
Correlation
Accuracy ( 1
of full
Ranga Sea la)
0-10% 1
0-JM
0-2SOppa>
0-19%
0-500ppe
0-500pp»
o-iooppa
varioua 2
o-aooo "r o.i
0-1001 10%
Saaipllno
Hathod
In altu
extractive
extractive
extractive
extractive
extract ive
extractive
In altu
in altu
n/a
* Soil f*Md rat*a ar* d«t«r»in«d from • corroUtlon of »otor «p*«d v«. f«*d
rat*.
TfcbieS
Swanson River Demonstration lest, Summan of
lest Results and ferformancc Calculations
rfsr on
i
'
PC* covcnmtAr
•no UH in
oj
PP* PP* 1.5
on OIOKIM oonTprr n.
ACR UH fTfcCt Al
U CAB
/•lo pyt ' ««/«.» I
FIUU. 0*7WT
ui rate*
CAJ
•frt w«/«in |
OXUMUOU »
110 rrikcit
•O»e «VO*«*
^^
> pel • p*rt* Mr crllllen
• vo • vol*cll« or^«tilc»
• IVO - S«ftl-voUcll« organ
laminated soil.
Swanson River DRE measurement was limited both by the size of
the stockpile of contaminated soil and by the concentration of PCBs
in the soil. OES was not allowed to bring any PCBs into the Kenai
National Wildlife Refuge for any purpose, including feed spiking, even
though the available soil concentrations and quantity were low. In all
6 tests, the DREs are based upon estimated maximum possible
concentrations rather than detection limits or measured quantities since
the measured quantities of PCBs in the flue gas were so low. The Swun-
son River DREs are the highest possible with these feed concentra-
tions and current detection limits. Had OES been allowed to spike the
soil to 10,000 ppm, as in the 1985 ICM at the research facility, higher
DREs would have been possible as they were then. Soil was incinerated
during four tests in August, 1985, three at I800°F and one at 1625 °F.
The feed concentrations ranged from 9,800 ppm to 12,000 ppm and
the DREs were all between 99.99998% and 99.999995%.
FUEL OIL SITE REMEDIATION
For more than 50 yr, a leaking underground storage tank at a cannery
in Stockton, California contaminated surrounding clay soil with No. 6
fuel oil. OES was contracted by the site operator to remediate the site
using one of its transportable 36-in. CBCs. OES developed and is now
completing a remediation plan that encompasses site characterization,
demolition of tanks and buildings, installation and operation of water
intercept wells, water treatment, soil excavation, stockpiling, CBC treat-
ment, placement of slurried purified soil and site and building
restoration.
The excavation and backfilling is complete and the CBC thermal treat-
ment of stockpiled soils is ncaring completion (August, 1989). Upon
completion of the project, over 11,000 tons of contaminated soil will
have been treated. Following restoration, the site will have its full cotn-
merciai value restored and it will be available for unrestricted use. Figure
4 is a photograph of the Stockton project site. Table 6A details the
February, 1989 source test operating conditions performed at Stock-
ton. The emissions are comparable to those from the other two CBC
units. In July, 1989 a demonstration test was performed using
naphthalene-spiked soil. Destruction removal efficiency data was the
only preliminary information available in August. Table 6B lists the
DREs from the three tests, all other results and operating conditions
were similar to those recorded during the February source test detailed
in Table 6A.
TREATMENT COSTS
Site remediation costs are divided into three categories. The first cost
category includes both direct and indirect costs for engineering design,
base equipment cost, materials, foundation and installation labor to erect
a mechanically complete unit. The second cost category includes labor,
materials, utilities, repair and maintenance and indirect costs. The thud
category includes material handling operations including excavation.
feed processing and ash disposal. Costs for all three CBC soil remedi-
ation cost categories combined typically range from $KX>-$300/ton of
soil depending primarily on soil moisture content and the quantity of
wastes to be processed.
IbMe&t
Stockton Source Test: Operating Conditions
Paraaiater
Coa\buetor taap, *F
Residence tin*, eec
Soil throughput, Ib/hr
Soil TPH cone., pp»
Flue gas oxygen, dry %
CO eBjiaaions, pp» at 71 Oj
HC «• tea lone, ppai at 71 o,
SO, eeiseions, Ib/day
SO| eniaaiona. ppat at 71 o^
NO, emissions, Ib/day
NO, e»iaalona, ppe, at 7% O,
Carbon dioxide, \
Particulat* gr/dscf at 7» o,
Combustion efficiency, t
Test 1
1588
1.8
4000
2130
U.6
28.0
<2
16.6
84
7.«
5J
7.0
0.045
99.989
Test 2
1588
1.8
4000
1160
13.6
2S.4
<2
12.0
61
7.3
52
6.6
0.046
99.990
Teat }
151?
1.1
400C
34 SO
13.6
23.6
a
24.2
123
6.7
47
6.»
0.015
99.990
lable 6B
Stockton Demonstration lest: Preliminary Results
Naphthalene cone., ppn
ORE, t
4314 47JO «IM
>99.9960 >99.999S6 >99.99»M
Conclusion
OES has developed a Circulating Bed Combustion waste treatment
technology and demonstrated its applicability in private- and
government-sponsored programs including the Superfund Innovative
402 INCINERATION
-------
Technology Evaluation program. Based on this development and testing four units, with two units now in operation and two ready for
program, modular CBC units have been designed, fabricated and deployment.
deployed. CBC treatment is being utilized in two large remediation
projects. REFERENCES
Treating contaminated soil in a CBC is cost-effective, highly effi- L Young DT .Trocess Demonstration Test Report for Trial Burn of PCB
cient and meets all performance and operation criteria established by Contaminated Soils, PCB Destruction Unit: Circulating Bed Combustor"
regulatory agencies. Ogden Corporation, OES' parent company, has paper GA-C18051, GA Technologies Inc., San Diego, CA, Submitted to Office
made a major commitment to the site remediation business by building of Toxic Substances, U.S. EPA, Aug. 1985
INCINERATION 403
-------
Site Program Demonstration Test of the CF Systems Inc.
Organics Extraction Unit
Richard Valentinetti
U.S. EPA
Washington, D.C.
ABSTRACT
The Superfund Innovative Technology Evaluation (SITE) Program
demonstration of the CF Systems organics extraction technology was
conducted at the New Bedford Harbor Superfund site in Massachusetts.
The demonstration was conducted concurrently with pilot dredging
studies managed by the U.S. Army Corps of Engineers, from which
samples of contaminated harbor sediments were obtained for use in
the demonstration. Several tests were conducted on a trailer-mounted,
pilot-scale unit to obtain specific operating, analytical, and cost
information that could be used in evaluating the potential applicability
of the technology to New Bedford Harbor and other Superfund sites.
The primary objective of this demonstration was to evaluate the
developer's treatment goals for extracting PCBs from harbor sediments.
Secondary objectives included an evaluation of (I) the unit's performance
in terms of extraction efficiency and a mass balance, (2) system operating
conditions, and (3) health and safety considerations.
The developer achieved an overall PCB concentration reduction of
89 percent for sediment samples that contained 350 ppm and 92 percent
for sediment samples that contained 2,575 ppm. The unit generally
operated within specified conditions for flow rates, pressures,
temperature, pH and viscosity. Results of the demonstration tests show
that the CF Systems technology is capable of reducing the PCB content
of contaminated sediment by greater than 90 percent without a risk to
operating personnel or the surrounding community.
INTRODUCTION
Through the SITE program, the U.S. Environmental Protection
Agency (EPA) is assisting technology developers in the development
and evaluation of new and innovative treatment technologies. The SITE
program's objective is to enhance the commercial availability and use
of these technologies at Superfund sites as an alternative to land-based
containment systems that are used most often at Superfund sites. Part
of the SITE program involves field demonstrations to gather real-world
data on a technology. The developer is responsible for the cost of
operating the equipment during the demonstration, while EPA is
responsible for the analytical costs and evaluation associated with the
demonstration. In most cases, the demonstration is performed at an
actual Superfund site that provides appropriate site and waste
characteristics for the specific technology to be tested.
PROCESS FLOW OF UNIT
CF Systems Inc., of Boston Massachusetts, developer of a liquefied
propane extraction technology, was selected to demonstrate their
pilot-scale system. New Bedford Harbor was chosen for the
demonstration site for CF Systems technology. The harbor sediments
are contaminated with polychlorinated biphcnyls (PCBs), a complex
organic substance amenable to extraction with CF Systems' process.
The developer's pilot-scale treatment technology is a trailer mounted
unit designed to handle pumpable soils, sludge, or sediments. The unit
operates in the six basic steps shown in Figure 1, that can cover
extraction, phase separations, and solvent recovery. A mixture of
liquefied propane and butane was used as the extraction solvent.
In step one, pumpable fsiurried) solid waste is fed into the top of
an extractor. Then (step two), the solvent, a propane/butane mix, is
condensed by compression and allowed to flow upward through the same
extractor. In the extractor the solvent makes non-reactive contact with
the waste, dissolving out the organics it contains. This is a somewhat
non-specific organic extraction process, though it is based on the
solubility of the organic waste in the extracting liquefied gas. Following
this extraction procedure, the residual mixture of clean water or
water/solids can be removed from the base of the extractor (step three).
In step four, the mixture of solvent and organics leaves the top of
the extractor and passes to a separator through a valve which partially
reduces pressure. The reduction of pressure causes the solvent to
vaporize out of the top of the separator. It is then collected and recycled
through the compressor as fresh solvent (step five). The organics left
behind are drawn off from the separator.
The demonstration tests devised by CF Systems for their pilot-scale
units (PCU-20 nominal capacity • 20 bbl/day) were designed to
demonstrate the treatability of New Bedford Harbor sediments and to
provide operating and scale-up data to assess potential commercial-scale
applications. The demonstration included equipment setup; a
"shakedown" stage to set process conditions; and daily start-up,
operation, and shutdown. When tests were completed, the demonstration
concluded with equipment decontamination and site closure. Thus, all
of the major components of a full-scale cleanup of New Bedford Harbor
were demonstrated.
TEST DESIGN
Sediments were dredged from five New Bedford Harbor locations
and stored in 55-gallon drums for processing by the pilot unit. Drummed
sediments were sieved to remove panicles greater than one-eighth inch
that could damage system valves. Water was also added to produce a
pumpable slurry. The drummed sediments were blended to provide
feedstocks for four tests. Measurements were also included on the
decontamination of the equipment.
A test consisted of a number of "passes." When the sediment was
treated or processed through the unit, the treated sediment became the
feed stock for another "pass." The variation in the number of passes
was to simulate a large full-scale unit, and to get additional design
parameters on such a unit. The following are the test, concentration
of the PCB's in the feed stock and purpose of the test:
404 U.S. EPA SITES .
-------
1. Test 1 was run as a shakedown test to set pressure and flow rates
in the PCU. The feed was a 50-gallon composite of sediments. The
feed had a PCB concentration of 360 ppm. Three passes were run
to gain experience with materials handling.
2. Test 2 was a 10-pass test. The feed was 350 ppm of PCB, and was
contained in a 511-pound composite of sediments. Ten passes were
run to simulate a high-efficiency process and to achieve treated
sediment levels less than 10 ppm. A 350 ppm PCB concentration
was chosen for this test since this represents an average, or typical,
PCB concentration in the harbor.
3. Test 3 was a 3 pass test. The feed was 288 ppm, and was contained
in a 508-pound composite of sediments. The purpose of this test
was to reproduce the results of the first three passes to Test 2.
4. Test 4 was a 6 pass test. The feed was 2,575 ppm of PCB, and was
contained in a 299-pound composite of sediments. The purpose of
this test was to reduce a high-level waste to a lower level waste such
as that used in Tests 1, 2, and 3. High-level wastes are found at several
"hot spots" in the harbor.
Decontamination of the system involved running toluene through the
PCU as a solvent wash. Samples were taken of the feed at the
commencement of each test. Treated sediment products and extracts
were planned for sampling at each pass. Additional samples were taken
of system filters and strainers. The amount of PCB contained in these
miscellaneous samples later proved to be small. The pilot unit's operating
pressures, temperatures, and flow-rates were monitored throughout the
tests. Field tests were conducted for feed viscosity, pH, and temperature.
TEST RESULTS
The objectives of this testing program were to evaluate: (1) the unit's
performance, (2) system operating conditions, and (3) health and safety
considerations.
SYSTEM PERFORMANCE
The evaluation criteria established for system performance were:
• PCB concentration in sediments before and after treatment
• PCB extraction efficiency with each pass of sediments through the
PCU
• Mass balances established for total mass, solids, and PCBs.
These criteria are discussed with respect to analytical results below.
PCB CONCENTRATION REDUCTIONS
PCB analyses for feed sediments and treated sediment, conducted
for samples collected at each pass, are shown in Table C-l. The data
show that treated sediment concentrations of 8 ppm are achievable and
that as much as 84 percent of the PCB contained in sediment can be
removed in a single pass. In Test 2, feed containing 350 ppm of PCB
was reduced to 8 ppm after 9 passes through the PCU. In Test 3, a
288 ppm feed was reduced to 47 ppm after just one pass. In Test 4,
a 2,575 ppm feed was reduced to 200 ppm after 6 passes. The percent
reductions in PCB concentration, based in a comparison of untreated
feed to the final pass, for each test were:
Test
2
3
4
Percent Reduction
in PCB Concentration
72%
Number of
Passes
10
3
6
The data for each test show general reduction trends based on
differences between initial feed and final treated sediment
concentrations. However, these trends are not consistent on a
pass-by-pass basis. For example, PCB concentrations in treated
sediments increase at Test 2, passes 4 and 10, and at Test 3, passes
2 and 3. These anomalies are not related to the extraction process.
Instead, they reflect cross contamination within system hardware, and
partially attributed to the limited analytical precision and accuracy. Since
the treated sediment collection tanks were under pressure, it was not
possible to clean out collection hardware and piping. Therefore, a
pass-by-pass mass balance could not be established.
Data for each test show the potential number of passes required to
reduce PCBs in harbor sediments to specific concentrations using the
Pit Cleanup Unit (PCU). If data from Test 2, 3, and 4 are displayed
side-by-side such that similar concentrations coincide, then a PCB
reduction can be plotted. Data are displayed in table C.I side-by-side
so that similar concentrations overlap.
Table - Cl
Pass-by-Pass PCB Concentrations
TEST 4
2,575
1,000
990
670
325
240
200
TEST 2
288
47
72
82
350
77
52
20
66
59
41
36
29
8
40
Based on the presentation of the data in Table C.I, it can be construed
that harbor sediments containing 2,500 ppm of PCB could be reduced
to 100 ppm after 6 passes through the PCU. A level less than 10 ppm
may be achievable after 13 passes.
EXTRACTION EFFICIENCY
For each test, the first pass results in efficiencies greater than
60 percent. However, at later passes efficiencies range from negative
values to 72 percent. This wide range is the result of cross-contamination
of solids retained in the treated sediment subsystem.
Data show that the system irregularly retained and discharged treated
sediments. For some passes, as much as 50 percent of the feed was
retained in the system. That feed was treated sediment that clung to
internal piping and tank surfaces. If discharged with a later pass, the
combined discharge could have a higher concentration than feed for
the later pass. For example, assume an extraction efficiency of
60 percent, a feed concentration of 350 ppm, and a carry-over of solids
from the first pass to the second pass of 25 percent. Then, the treated
sediment would contain 77 ppm, instead of 56 ppm if no cross
contamination occurred.
Cross contamination did affect the interpretation of each test, but
it does not invalidate the fact that treated sediment concentrations as
low as 8 ppm were produced. Furthermore, the decontamination
procedure snowed that PCB which accumulated in system hardware
was contained in the extract subsystem, not the treated sediment
subsystem.
OPERATIONAL ISSUES
System operating criteria were set during the shakedown portion of
the demonstration.
Extractor pressure was controlled at the unit's main compressor and
U.S. EPA SITES 405
-------
at the organics discharge from the extraction segment of the unit. Solvent
flow rate and the solvent to feed ratio are set after laboratory bench-scale
tests were run on various mixtures of solvent and feed.
The feed temperature represents the temperature of the material
pumped into the feed unit. Feed in the extractor was maintained above
60°F to avoid the possibility of hydrate formation, which could have
infered with the extraction process. If the feed is above 120°F, it must
be cooled to prevent vaporization of the solvent.
The feed flowrate represents the rate at which material is pumped
from the feed kettle into the unit. Operational flow rates above the listed
maximum can force segments of the system, such as decanters and
control valves, beyond their effective hydraulic capacity.
The viscosity and solids content must be such that the feed material
is pumpable. Reds with a viscosity above the listed range were slurried
with water to yield a pumpable viscosity. In order to prevent damage
to the process equipment, the pilot-scale unit has a maximum limit for
solids size.
OPERATIONAL MEASUREMENTS
Process controls, wastestream masses, and utilities were measured
at various intervals during each test. Listed below arc critical operational
parameters and measurement frequencies:
• Feed temperature, viscosity, and pH—measured at each pass
• Feed sediment and treated sediment mass—measured at each pass
• Feed flow rate—measured every 10 minutes
• Extractor pressure and temperature—measured every 10 minutes
• Solvent flowrate—measured every 10 minutes
• Extracted organics mass—measured each test
OPERATIONAL RESULTS
The unit generally operated within the specifications with only several
exceptions. Criteria were met for feed flowrates, solids content,
maximum possible size, viscosity, and pH as well as extractor pressure.
The solvent flow rate and solvent to feed mass ratios fluctuated above
and below criteria throughout the tests but did not have an observable
effect on pass-by-pass extraction efficiency. Temperature of the feed
sediments fell below the minimum temperature criterion during passes
6, 7, 8, 9, and 10 of Test 2.
Commercial-scale designs for application of the technology should
ensure that operating specifications are maintained. Feed materials are
likely to be well below 60F throughout winter months and this could
affect system performance. Therefore, heat must be added to sediments
fed to a commercial-scale unit (or the unit could be located in an
enclosed structure). Coarse solids removal will be required to maintain
feed sediment particle sizes below one-eighth inch. Wide fluctuations
in the feed to solvent ratio should be minimized. Extraction efficiency
is directly related to the amount of solvent available for solubilizing
organics contained in the feed.
HEALTH AND SAFETY ISSUES
The Health and Safety Plan established procedures and policies to
protect workers and the public from potential hazards during the
demonstration. Implementation of these procedures and health and safety
monitoring showed that OSHA level B protection is necessary for
personnel that handle system input and output, although only OSHA
level C protection is necessary for unit operators.
Combustible gas meters indicated that levels at approximately
20 percent of the lower explosive limit for propane were encountered
while samples were taken. Background air sampling and personnel
monitoring results indicate that organic vapors and PCB levels were
present at levels below the detection limit for the analytical methods.
Site spoil samples taken before and after the demonstration indicate
that demonstration activities did not result in increased PCB levels in
the staging area soils.
ComprtfKX
Figure 1
Simplified Flow Chan
DISCUSSION AND CONCLUSIONS
In the design of this treatability demonstration, and in many other
cases, the demonstration plan is fraught by the reality of the field
implementation. A perfectly good premise, even backed by previous
field data, can go awry, and this was the case in this demonstration.
The premise was that by recycling the material through the unit a number
of times it would stimulate a full-scale unit. Even though there was
a recognized deficiency in this method, it was felt that substantial data
could be obtained on extraction rates and efficiencies. But the apparent
cross-examination and retention of solids and PCB's in the equipment
flawed the basic premise. There was a lack of consistency in the data
as reflected in Table C.I, and there was an actual increase in PCBs in
the process from one pass to another in some of the runs. In fact, if
you review the last two passes of test 2 (the 350 ppm of PCB in toe
original feed with K> passes), there is an increase from 8 ppm to 40 ppm
of PCB's in the last pass. This last data point was in contrast to the
general downward trend of the other passes, but this outiler data point
was used for final determinant data point for this treatability test. If
the 9 passes were used in the determination of the extraction rate, the
total test extraction efficiency would have been 97% rather than the 89%
with the use of pass K).
Besides the possible issue of cross contamination, the precision and
accuracy of the sampling procedures and the analytical methods for
PCB's at the low ppm range, cause problems in the interpretation of
the data. Even within limits set by QA/QC there is a wide variant in
the 95% confidence range, thus making the data difficult to interpret
at these low levels.
PCB Analytical Method 8080 precision criteria established for this
project were plus or minus 20 percent and accuracy criteria were plus
or minus SO percent. Despite the occurence of the cross contamination
and its effect on each test, this does not invalidate the fact that treated
sediment concentrations were as low as 8 ppm. Furthermore, the
decontamination procedure (using toluene rinse) showed that the PCB's
which accumulated in the system hardware were contained in the extract
subsystem, not the treated sediment subsystem.
In contrast, the retention of solids in the pilot unit was a concern
on the operating ability of the full-scale unit. The developer has indicated
that the design of the full-scale unit will compensate for solid retention
issue. This claim will be validated by the review of data from a full-scale
unit (200 barrells/day) in December 1989. This evaluation should answer
many of the questions associated with the on-line capacity of the
full-scale unit.
406 U.S. EPA SITES
-------
A Field Evaluation of the UV/Oxidation Technology
To Treat Contaminated Groundwater
Norma Lewis, M.A.
U.S. EPA
Cincinnati, Ohio
Kirankumar Topudurti, Ph.D.
Robert Foster, RE.
PRC Environmental Management, Inc.
Chicago, Illinois
ABSTRACT
This paper presents the field evaluation results of the ultraviolet
radiation (UV)/oxidation technology developed by Ultrox International,
Santa Ana, California. The field evaluation of the technology was
performed at the Lorentz Barrel and Drum (LB&D) site in San Jose,
California, under the Superfund Innovative Technology Evaluation
(SITE) program from Feb. 27 through Mar. 10, 1989.
The UV/oxidation technology uses UV radiation, ozone and hydrogen
peroxide to oxidize organic contaminants present in water. At the LB&D
site, this technology was evaluated in treating groundwater contami-
nated with volatile organic compounds (VOCs). The Ultrox system
achieved VOC removals greater than 90%, and the majority of VOCs
were removed through chemical oxidation. However, for a few VOCs,
such as 1,1,1-trichloromethane (1,1,1-TCA), and 1,1-dichloroethane
(1,1-DCA) stripping also contributed toward removal. The treated
groundwater met the applicable discharge standards (NPDES) for
disposal into a local waterway at 95% confidence level. There were
no harmful air emissions from the Ultrox system into the atmosphere.
INTRODUCTION
The EPA is finding better solutions to hazardous waste remediation
through its Superfund Innovative Technology Evaluation (SITE)
program. The SITE program was created to demonstrate and evaluate
technologies that may destroy or permanently change the composition
of hazardous waste in the environment by significantly reducing the
waste's toxicity, mobility or volume. The SITE program also generates
reliable performance and cost data for these treatment technologies to
be used in evaluating alternatives under the Superfund site remedia-
tion process.
In 1988, Ultrox International's proposal for its ultraviolet radiation
(UV)/oxidation technology was selected by U.S. EPA's Office of
Research and Development (ORD) and Office of Solid Waste and Emer-
gency Response (OSWER) under the SITE program. This technology
was demonstrated at the Lorentz Barrel and Drum (LB5D) site in San
Jose, California, through a cooperative effort between Ultrox Interna-
tional, ORD, OSWER and U.S. EPA Region IX.
UV/OXIDATION TECHNOLOGY: EQUIPMENT
AND PROCESS DESCRIPTION
The Ultrox UV/oxidation treatment system uses UV radiation, ozone
and hydrogen peroxide to oxidize organics in water. The major com-
ponents of the Ultrox system are the UV/oxidation reactor module, air
compressor/ozone generator module, hydrogen peroxide feed system
and catalytic ozone decomposition (Decompozon) unit. An isometric
view of the Ultrox system is shown in Figure 1.
The UV/oxidation reactor used in the demonstration (Model PM-150)
CATALYTIC OZONE DECOMPOSER •
Figure 1
Isometric View of Ultrvox System
has a volume of 150 gal and is 3 ft long by 1.5 ft wide by 5.5 ft high.
The reactor is divided by five vertical baffles into six chambers and
contains 24 UV lamps (65 w each) in quartz sheaths. The UV lamps
are installed vertically and are evenly distributed throughout the reactor
(four lamps per chamber). Each chamber also has one stainless steel
sparger that extends along the width of the reactor. These spargers uni-
formly diffuse ozone gas from the base of the reactor into the water.
Hydrogen peroxide is introduced in the influent line to the reactor from
a storage tank. An in-line static mixer is used to disperse the hydrogen
peroxide into the contaminated water in the influent feed line.
The Decompozon unit (Model 3014 FF) uses a nickel-based
proprietary catalyst to decompose reactor off-gas ozone to oxygen. The
Decompozon unit can accommodate flows of up to 10 scfm and can
reduce ozone concentrations in ranges of 1 to 20,000 ppm (by weight)
to less than 0.1 ppm.
During the Ultrox system operation, contaminated water first comes
in contact with hydrogen peroxide as it flows through the influent line
to the reactor. The water then comes in contact with the UV radiation
and ozone as it flows through the reactor at a specified rate to achieve
the desired hydraulic retention time. As the ozone gas in the reactor
is transferred to the contaminated water, hydroxyl radicals (OH°) are
produced. The hydroxyl radical formation from ozone is catalyzed by
UV radiation and hydrogen peroxide. The hydroxyl radicals, in general,
U.S. EPA SITES 407
-------
are known to react with organics more rapidly than the oxidants ozone,
hydrogen peroxide and UV radiation. They are also much less selec-
tive in oxidation reactions than the three oxidants. Ozone that is not
transferred to the contaminated water will be present in the reactor off-
gas. This ozone is subsequently destroyed by the Decompozon unit be-
fore being vented to the atmosphere. The treated water flows from the
reactor for appropriate disposal.
LB&D SITE HISTORY
The LB&D site is in San Jose, Santa Clara County. California. This
site was used for drum recycling operations from about 1947 to 1987.
The drums contained residual aqueous wastes, organic solvents, acids,
metal oxides and oils. A preliminary site assessment report for the
LB&D site showed that the groundwater and soil were contaminated
with organics and metals'. In 1987, the LB&D facility ceased opera-
tion due to a restraining order issued by the California Department of
Health Services. U.S. EPA Region IX assumed the responsibility for
site remediation.
The shallow groundwater at the LB&D site was selected as the waste
stream for evaluating the UV/oxidation technology. Groundwater sam-
ples collected in December, 1988. indicated that several volatile organic
compounds (VOCs) were present in the shallow aquifer. VOCs detected
at high levels included trichloroethylene (280 to 920 /i/L), vinyl chloride
(SI to 146 /j/L) and 1,2-trans-dichloroethylene (42 to 68 /i/L). The pH
and alkalinity of the groundwater were approximately 7.2 and 600 mg/L
as CaCO,, respectively. These measurements indicated that bi-
carbonate ion (HCO.), which acts as an oxidant scavenger, was
present at high levels. Other oxidant scavengers such as bromide, cyanide
and sulfide were not delected.
TECHNOLOGY DEMONSTRATION
The objectives of the technology demonstration were to: (I) evaluate
the ability of the Ultrox system to treat VOCs present in the ground-
water at the LB&D site at different operating conditions; (2) determine
the extent of VOC stripping, if any, from the bubbling of ozone gas;
(3) evaluate the efficiency of the Decompozon unit to decompose reactor
off-gas ozone; (4) determine the operating conditions needed for the
effluent to meet applicable discharge standards (NPDES) for disposal
into a nearby waterway; and (5) develop the information required to
estimate operating costs for the treatment system, such as electricity
consumption and oxidant doses.
Testing Approach
Eleven test runs were performed to evaluate the Ultrox system under
various operating conditions. After these runs, two additional runs were
performed to determine if the system's performance was reproducible.
The operating conditions for the runs are summarized in Table I. All
13 runs were performed over a period of 2 wk.
The study was designed to evaluate the Ultrox system by adjusting
the levels of five operating parameters: (1) influent pH, (2) retention
time, (3) ozone dose, (4) hydrogen peroxide dose and (5) UV radiation
intensity. The initial operating conditions (Run 1), given in Table 1.
were based on the treatability study conducted by Ultrox on LB&D
site groundwater.
During the demonstration, a preliminary estimate of the Ultrox
system's performance in each run was obtained based on the effluent
concentrations of three indicator VOCs. The VOCs selected for this
purpose were trichloroethylene (TCE, a major volatile contaminant at
the site), 1,1-dichloroethane (1,1-DCA) and 1,1,1-trichloromethane
(1,1,1-TCA). 1,1-DCA and 1,1,1-TCA were selected because Ullrox's ex-
perience indicated that these VOCs are relatively difficult to oxidize.
In the first three runs, the influent pH was adjusted by adding sul-
furic acid to evaluate the system's performance and to determine the
"preferred" influent pH ["preferred" operating conditions are those
conditions in which: (1) effluent concentrations of indicator VOCs are
below NPDES limits and (2) the relative operating costs are the lowest].
Once the "preferred" influent pH was determined, it remained at that
level for the remaining runs. The Ultrox system performance was then
studied by varying other parameters, one at a time, as shown in Table I,
Table I
Operating Parameter* Matrix for the Ultrox SyBem Demonstration
Run Ho. Retention
TiM
Of on*
Dose
Influent pH
X-
X
X
l.SX
O.SX
Preferred
Preferred
Preferred
Preferred
Preferred
f f
t I
1 I
t z
1 I
1.9V I
B.»Y t
Preferred i.«
Preferred 0.11
All OK
All ON
All OH
All Oil
All OH
All 00
All Of
All OX
All Oil
Uie first
three chAoben
Unadjusted
(Unadjusted - l|
(Unj.dju.ted - 2)
Preferred*
Preferred
Preferred
Preferred
Preferred
Preferred
Preferred
II'
II'
Preferred Preferred preferred Only OH in
the list
three chejMbers
Preferred Preferred Preferred Preferred
preferred Preferred Preferred Preferred
Preferred
Preferred
HotelI
X - 40 •inutes.
1 - 7» mq/L.
I - 3S ml/L.
(X. V, end I veliMi vere determined by Ultro« Internetlonel to be On
op tine condition* for treetlnq qround wtter in the (Testability etntf
•t the LftiD >lte.)
•Preferred* operetlnq condition* ere Uioee condition* to which. (1) tb>
concentretlone of effluent Indlcetor VOCs ere below their respective
HPOES Halts end (1) the reletlve operetlno coeu ere tne lowest.
Verification runs performed to check the reproduclblllty of the Gltm
eystea's performance et the "preferred* ape-retina, conditions.
to determine the "preferred" values for those parameters. The criteria
were the same as those used in determining the "preferred" value for
the influent pH. After the "preferred" values were determined for all
five operating parameters, two runs (12 and 13) were performed to verify
the reproducibility of the Ultrox system's performance at the "preferred"
operating conditions. By duplicating the "preferred" operating condi-
tions determined during the previous 11 runs, the two verification runs
served to ensure that the results could be based on repeated observa-
tions, with comparable findings.
Sampling and Analytical Procedures
Air and water samples were collected from the Ultrox system at the
locations shown in Figure 2. For the critical parameters in this study
(VOCs in water), six replicate samples were collected. Duplicate samples
were collected for other parameters listed in Table 2. Sampling at the
influent port began approximately IS min after each run was started.
At other locations in the reactor, sampling began after three retention
times to allow the system to reach steady-state. All the air and water
samples for off-site laboratory analysis were preserved as required before
being shipped to the laboratory.
The analytical methods followed in this study are listed in Table 2.
To obtain reliable data, strict QA/QC procedures were followed. Details
on all aspects of the QA/QC procedures are presented in the Demon-
stration Plan and the Technology Evaluation Report2-'.
RESULTS AND DISCUSSION
This section summarizes the results of the Ultrox system demonstra-
tion and also presents an evaluation of the UV/oxidation technology's
effectiveness in removing VOCs from the groundwater at the LB&D she.
Summary of Results for VOCs
The purpose of the test runs was to evaluate the effectiveness of the
Ultrox system in removing 44 VOCs present in the groundwater at the
LB&D site. The removal efficiencies and concentration profiles of all
VOCs are not presented in this paper. Instead, a summary of the results
is given.
The mean concentration profiles and the discharge standards (NPDES)
for the three indicator contaminants (TCE 1,1-DCA and 1,1,1-TCA) in
408 U.S. EPA SITES'
-------
OWOM """ Ł
NMA
Hyp
(to- _
sr-4-
r
VWn
tori)
— >
F
(lyplotf] Oion* MtnHoM
41 "n ii n ft.'r
^<:
0
C
*fl-V
O Of
0 0
>J
W)IPH*
H o^
•1
O
o
1*
• •
J-.
-^-4-
O c
O C
°(
1
L__
each run for
5. The VOC c
to the mid-po
3. In Run 3, th
that in the inl
tion or the in
decrease in cc
radiation prov
the increase i
port. Additioi
comparativel)
ozone dose d
80 -
70 -
60 -
B «-
20 -
•N.
H
i
0*
Ul
each san
oncentra
nt and ft
e concen
luent. It
fluent cc
mtainme
ided in t
n the ret
lally, the
' high in
uring thŁ
! '
•
/ ^
/
'', \
• ^
;, s
Or
_^L_r^
^
>
3^.
0
0
^^ —
o
•5?
0
• ^ - y^ -
r?
0
Table 2
Analytical Methods
Onm l
IJO UM i Method Method
JB mumim m Analyte Matrix Type Reference
ft
0 tf
.
•
cb-J
{t"*>in FromSunw.
1 SwwInTwUt
F— ^
., a 1
r Alkalinity Liquid Field MCAWW 310.1™
Arsenic Liquid Lab SW-846 7060*
BNA Liquid Lab SW-846 8270"
171 HI ""- (Semivolatiles)
' 1 U-EMMm
»"•» Chromium Liquid Lab SW-846 7195V
*J """ (Cr-)
| Chloride Liquid Lab SM 429™
^ EffluMt v
s«nt*»Ti0 Chromium Liquid Lab SW-846 7191
Conductivity Liquid Field Manual*
" "" B I J
to^^m Hydrogen Liquid Field Boltz et al.
FMT..IIIUOM Peroxide (1979)"
„. , Metals Liquid Lab SW-846 6010*
Mgure 2 (Barium, Cobalt,
trox System Sampling Locations Iron, Manganese,
r Nickel, Zinc,
Potassium, Calcium,
ipling location are plotted in Figures 3, 4 and sodiuno™' and
tions progressively decreased from the influent
om the mid-point to the effluent except for Run Ozone Liquid Field ^dgnea982)™
tration of 1,1-DCA at mid-point was higher than . . ,. „ . . „
' r . . ° Ozone Air Field 40 CFE Part 50"
is believed that either the mid-point concentra-
ncentration is just an outlier. This progressive pH Liquid Field Manual"
nt concentration is due to the ozone and the UV pesticides/pcBs Liquid Lab sw-846 aoao"
le last three chambers (after the mid-point) and smca Liquid Lab sw-846 eoio"
ention time from the mid-point to the effluent suifate Liquid Lab SM 429™
effluent and mid-point VOC concentrations are Temperature Liquld Field Manual»
Run 7, which appears to be due to the decreased
.. , Total Organic Liquid Lab SM 505A™
it particular run. carbon
'S
^
:a :
^3 >.
*M *•
•
«
' ^
' \
' ^
n
' V
^ S
t
fi
^
'L
\
4
>a
4
Volatile Organics Liquid Lab sw-846 8010
and 8020V
Volatile Organics Liquid Lab SW-846 8240"
Volatile Organics:
Vinyl Chloride Air Lab NIOSH 1007"
' 1,1-Dichloroethene Air Lab NIOSH 1015"
11 s i,l-Dichloroethane Air Lab NIOSH 1003X
" " 1,2-Dichloroethene Air Lab NIOSH 1003"
" ' 1.1,1-
v " Trichloroethane Air Lab NIOSH 1003"
, '
s s Trichloroethene Air Lab NIOSH 1022"
' n ' _ Benzene Air Lab NIOSH 1.500*
',* ^i- ^3-. 1.1.2.2-
1 ' ' 456789 10 01213' Tetrachloroethane Air Lab NIOSH 1019X
RUN NUMBER Acetone Air Lab NIOSH 1300X
Figure 3
TCE Concentrations in Different Runs
The average effluent concentrations (determined during the demon-
stration by analyzing only two of the six replicates) for each indicator
VOC with the discharge standard (NPDES) showed that the effluent
met the discharge limits in Runs 8 and 9. Since a lower hydrogen
peroxide dose was used in Run 9, compared to Run 8, Run 9 was chosen
as the "preferred" operating run. However, based on a complete analysis
of the six replicates performed after the demonstration, the mean con-
centration of 1,1-DCA was found to be slightly higher than 5 /*/L, the
discharge standard for the VOC. Since Run 9 had the "preferred"
operating conditions during the demonstration, the verification runs
(12 and 13) were performed at those conditions.
A comparison of 95 % upper confidence limit (UCL) values for the
effluent VOCs in Runs 9, 12 and 13 with the discharge standards
(NPDES) is presented in Table 3. The UCL values were calculated using
the one-tailed Student's t-test. Table 3 shows that the effluent met the
discharge standards for all regulated VOCs at the 95 % confidence level
in Runs 12 and 13. In Run 9, the mean concentrations for 1,1-DCA and
1,2-DCA exceeded the discharge standards. Although 1,1-DCA and
1,2-DCA were present at levels slightly greater than the discharge
standards, the difference in performance among the three runs is
negligible.
The mean concentration profiles for total VOCs are given in Figure 6.
A comparison of the VOC concentrations presented in Figure 6 with
those in Figures 3, 4 and 5 indicates that the concentration profiles for
total VOCs are similar to those for the indicator VOCs. For example,
the peaks present at the mid-point and effluent for indicator VOCs are
also present in the total VOC concentration profiles.
The percent removals for the indicator VOCs and total VOCs are
presented in Figure 7. The figure shows that the removal efficiencies
U.S. EPA SITES 409
-------
14 -
D -
a -
D -
10 -
*"
7 -
8 -
4 -
J
a
D
3 J R R
'! '3 ' '5
: :; :i III \\ ;i
v; ; • ; • :; •; ; -;s v ;
•^ >H *iS 'H *H ^:
' | ; ' | ; ' ; ; ' : ; ' ] ; • : ;
R
\\ ;
! : :g :a ;a : :
! I . ' W S / Si* S ' S J
! « ' s / si* i / v i* » '
:; • :; •;; -!a ; ! : ' • a | | !
i k . . 1 • . . ( » . . ( k . . i iL . i jJ-Li-i . 1
I234S«789B
RUN t
Figure 4
1,1-DCA Concentrations in Different Runs
4 -
7 -j
( J
rj
!5
!!
' S
' S
,
/ ^
' S
•• s
s
s
!a *
;s -
2
>
<3
; s;
' ' ' 3
' ' S
' S ' N
1 S ' \
\\ ;:s
J S J S
3 4
^
>;
/
' t S '
i '
' * 7 " '
; , j s;
' s! - :
' 3 ' I s '
ihljii
567
0 %
s s
,{ J
:||:
i
D •
'
:1 ^
9
' s
'
i:
D
•j
' v
'
:a v
:3 ,
n
'
i
D
i.
D
Figure 5
1,1,1-TCA Concentrations in Different Runs
BO -
ISO -
170 -
160 -
ISO -
DO -
00 -
no -
no -
90 -
so -
70 -
SO -
40 -
30 J
XI -
10 -
i
; ^
^
1
^
a'
2
-1
•1
3
> s
; ^
; „
III
4
'
^
'$
:
s
nu
n
s
s
s
'
^
!?p
0
N NUM
'
;
;
'
;
' S
' s
' «.
i\
7
DEA
1
a
^
|
1
9
p
3
I
»
q
' N
' N
n
N
D
'
;
i«
D
Table 3
Comparison of Effluent VOC Concentrations in Runs 9, 12 and 13
991 UCL,»g/L RT.pg/L Conclusion
Bun numberi I
1,1,1-TCA
1,1,2,2-PCA
1,1-DCA
1,1-DCe
1,2-DCA
1,2-DCPA
Benzene
Chloroethane
Chloroform
PCE
7-1,2-DCE
TCE
vinyl Chloride
Run numberi 1J
1,1-TCA
1,2,2-PCA
1-DCA
i-Dce
2-DCA
2-DCPA
Beniene
Chloroethsne
Chlorofom
PCE
T-1.2-DCE
TCE
Vinyl Chloride
Run
11
1,1,1-TCA
1,1,2,2-PCA
1,1-DCA
1,1-DCE
1,2-DCA
1,2-DCPA
Benzene
Chloroe thsne
Chloroform
PCE
T-1.2-DCE
TCE
Vinyl Chloride
Notes:
99% UCL:
RT:
OK:
N:
0.7S
0.045
S.I
0.000
1.)
3.)
0.021
0.000
1.1
0.24
0.000
1.2
0.11
0.43
0.04i
3.8
0.000
0.92
2.6
0.023
0.000
0.74
0.19
0.000
O.S5
0.11
0.48
0.04}
4.2
o.ooo
1.0
2.9
0.45
0.000
0.81
0.091
0.000
0.63
0.12
1.0
0.049
t.S
0.000
1.4
3.4
0.02C
0.000
1.2
0.63
0.000
1.3
0.11
0.48
0.045
4.2
0.000
1.0
2.9
0.026
O.OOO
0.82
0.18
0.000
0.69
0.11
0.44
0.049
4.5
0.000
1.0
3.1
0.52
0.000
0.87
0.17
0.000
0.73
0.12
5
5
5
5
1
5
5
9
5
9
5
5
1
5
5
t
9
1
5
5
5
5
9
5
5
2
5
5
5
5
1
5
5
5
5
5
5
5
2
OK
OK
M
OK
*
OK
OK
OK
OX
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OK
OX
OK
OX
OK
OK
Uppor 95% Confidence Limit
Regulatory Threshold
Effluent »et the regulatory threshold
Effluent did not Beet the regulatory threshold
Abbreviations:
1,1,1-TCA: 1,1,1-Trichloroethane; 1,1,2,2-PCA: 1,1,2,2-
Tetrachloroethane; 1,1-DCA: 1,1.-Dichloroethane; 1,1-DCE:
Dlchloroethylene: 1,2-DCA: 1,2-DichloroeUwne; 1,2-DCPA:
1,2-Dlchloropropane; PCE: Tetrachloroethylene; T-1,2-DCS:
Trans-l,2-Dlchloroethylene: TCE: Trichloroethylene.
1,1-
for TCE were higher than (hose for 1,1-DCA and 1,1,1-TCA which is
consistent with the rationale used in selecting the indicator VOCs. The
percent removals for total VOCs and the indicator VOCs decreased con-
siderably in Run 7, which appears to be due to the decreased ozone dose.
no
90
ao
TO -
80
I J345«7»9lonDD
RUN NUMBtn
ESS ™ S ••«» ES '"-'» ^ i—«»
Figure 6
Total VOC Concentrations in Different Runs
Figure 7
VOC Removals in Different Runs
410 U.S. EPA SITES «
-------
Since ozone gas is bubbled through the groundwater treated by the
Ultrox system, the VOC removal could be attributed to stripping in ad-
dition to oxidation. To determine the extent of stripping within the treat-
ment system, VOC samples were collected from the reactor off-gas.
Twenty-five samples were collected during the demonstration. Although
1,1-DCE, 1,2-DCE, benzene, 1,1,2,2-tetrachloroethane and acetone were
present in two samples at concentrations close to the detection limits,
TCE, vinyl chloride, 1,1,1-TCA and 1,1-DCA were detected more fre-
quently. To determine the extent of stripping, the emission rates in the
reactor off-gas for these latter four VOCs were compared to the VOC
removal rates (estimated by difference between the VOC input rates
at the influent and output rates at the effluent ports of the Ultrox system).
The results are summarized in Table 4.
liable 4
Extant of VOC Stripping in the Ultrox System
Air flow rate
Run
Ho.
1
2
3
4
5
6
7
8
9
10
11
12
13
Water flow rate
2.
2.
2.
2.
2.
4.
1.
4.
4.
4.
4.
4.
4.
1
3
1
0
1
5
0
5
5
3
6
4
3
Percent striooina
Contribution
For
1,1-DCA TCE 1,1,1-TCA VC"
0.0043° 0.0091' 0.014" 0.082*
7.4
9.1
9.9
7.4
17
16
4.9
23
16
27
44
34
37
2
3
2
3
3
1
1
7
6
9
24
7
26
.0
.4
.7
.0
.5
.2
.2
.5
.6
.4
43
34
31
29
29
65
12
85
58
73
>99
.0
76
75
0.
0.
0.
0.
1.
0.
3.
1.
0.
1.
13
8.
1.
013
95
013
01
7
072
1
2
04
1
9
8
Notes:
a
b
VC: Vinyl
chloride
Henry's law constant
of the VOC,
atm-mVmol .
Since the extent of stripping for any particular VOC is expected to
be proportional to the ratio of the air flow rate to the water flow rate,
this ratio is presented in the table. The ratio for Runs 1 to 5 is approxi-
mately 2; for Run 6 and Runs 8 to 13, it is approximately 4.5; and for
Run 7, it is 1. If stripping contributed to the total removal of the four
VOCs, the extent of stripping should be the least in Run 7 and the most
in Runs 6 and 8 to 13. The data presented in the table follow this trend
for three of the four VOCs (except for the vinyl chloride in Runs 6,
7 and 9). However, a quantitative correlation of the extent of stripping
cannot be made because the operating conditions were different in each
run. For example, at a given air to water flow ratio, when oxidant doses
are varied, the extent of oxidation also varies. Therefore, the extent
of stripping will be indirectly affected.
Table 4 also presents the Henry's law constants for the four VOCs".
By comparing these constants for the VOCs, their volatility is expected
to increase from left to right:
1,1-DCA a-»a TCE a->a 1,1,1-TCA a-»a vinyl chloride
However, a significant removal fraction for 1,1,1-TCA and 1,1-DCA
were observed to be due to stripping. Conversely, the extent of stripping
was low for vinyl chloride and TCE. This difference in stripping rates
is because it is easier to oxidize vinyl chloride and TCE than 1,1-DCA
and 1,1,1-TCA because there are double bonds between the carbon atoms
in TCE and vinyl chloride. In other words, in the UV/oxidation process,
stripping is a significant removal pathway for compounds that are
difficult to oxidize.
Performance of the Decompozon Unit
The ozone concentrations in the influent to and the effluent from the
Decompozon unit were analyzed in each run. These concentrations are
presented on a semi-log plot in Figure 8. The effluent ozone concen-
trations were low (less than 0.1 ppm) for Runs 1 to 8, approximately
1 ppm in Runs 9 and 10 and greater than 10 ppm in Runs 11, 12 and
13. The high ozone levels (greater than 1 ppm) in the effluent are
attributed to the malfunctioning heater in the Decompozon unit. The
temperature in the Decompozon unit should have been 140 °F for the
unit to properly function, whereas the temperature for Runs 11 to 13
was only approximately SOT. The ozone destruction efficiencies greater
than 99.99% were achieved in Runs 1 to 10.
\
\
\
\
\
\
\
\
\
s
\
1
s
s
\
s
s
s
s
s
s
s
s
s
\
I
\
\
\
\
\
s
s
\
s
s
\
s
1
^
s
s
\
\
s,
\
1
1
-«;
S,
\
\
\
S
V
\
S
S
s
s
\
s
\
\
\
s
1
1
\
\
\
\
\
1
^
\
\
\
\
\
\
s \ \ r^q r^
\ \ \ \ \
s \ \ \ \ ^.
\ \ \ \ ^3 \ ^
\ \ ^^^^^$1
\ s \^ \^q \^
\ \^ \^ \ ^J \^
!
S 6 7
RUN NUMBER
Figure 8
Ozone Concentration in Different Runs
Although the primary function of the Decompozon unit is to remove
ozone, the data presented in Table 5 indicate that significant VOC
removal occurred when the unit functioned as designed (Runs 1 to 8).
Table 5
VOC Removal in the Decompozon Unit
un
0. 1
l
2
3
TCE, ppm 1,1-DCA, ppm 1,1,1-TCA, ppm Vinyl chlorida, ppn.
n u«n n
5 <0.15 0.1 <0.
5 <0.15
5 <0.15
5 <0.15
5 <0.1S
15 <0.15
15 <0.1S
.15 <0,15
.15 <0.15
.55 0.325
.15 <0.15
.45 <0.15
1 <0.
<0.
<0.
<0.
<0.
5 <0.
<0,
<0.
5 0.
<0.
2 <0.
0.15 <0.
0.1 <0.
0.1 <0.
0.1 <0.
0.1 <0.
0.1 <0.
0.1 <0.
0.1 <0
0.1 <0
0.1 <0
0.2 <0
0.1 <0
0.1 <0
0.002 <0.002
0.070 <0.002
0.002 <0.002
0.002 <
0.150 <
0.003 <
0.517 <
0.041 <
0.002
0.064
0.570
0.271
0.420
002
002
002
002
.002
.061
.078
.189
.006
.004
Summary of Results for Noncritical Parameters
In addition to the critical parameters (VOCs), many non-critical
parameters also were measured. The non-critical parameters for organics
included changes in semi-volatiles, PCBs/pesticides and total organic
carbon (TOC); the non-critical parameters for inorganics included
changes in pH, conductivity and alkalinity. Temperature, turbidity,
residual oxidants and electricity consumption also were measured.
No semi-volatiles or PCBs/pesticides were detected in the influent
U.S. EPA SITES 411
-------
or effluent. TOC removal was achieved only at trace levels indicating
that complete oxidation of organics to carbon dioxide and water did
not occur. However, since no new VOCs were found by GC/MS analy-
sis and GC analysis of the effluent, the oxidation products were not
VOCs.
Metals such as iron and manganese were present at low concentra-
tions in the influent and no significant metal removal occurred. No
changes in alkalinity and conductivity were observed after the treat-
ment. However, the pH increased by 0.5 to 0.8 units after the treat-
ment. The increase in pH is probably due to the reaction between
hydroxyl radicals and bicarbonate ion (the predominant form of alka-
linity at the groundwater pH, which is 7.2) in which hydroxyl ions are
produced13.
Turbidity increased by 1 to 4 units after the treatment. This increase
may be due to the insignificant amount of metal removal by oxidation
and precipitation. The temperature increased by approximately 4 to 5°F
after the treatment and was due mainly to the heat from UV lamps.
The efficiency of ozone gas transfer to the groundwater was over 95 %,
with 5 % remaining in the reactor off-gas. After the reaction, the residual
ozone and hydrogen peroxide concentrations in the effluent usually were
less than 0.1 ppm. The average electrical energy consumption to operate
the Ultrox system was approximately 11 kwh/h of operation.
CONCLUSIONS
The groundwater treated by the Ultrox system met the discharge
standards for disposal into a nearby waterway at the 95% confidence
level at a hydraulic retention time of 40 minutes, an influent pH of 7.2
(unadjusted), an ozone dose of 110 mg/L, a hydrogen peroxide dose
of 13 mg/L and with all 24 UV lamps operating.
There were no VOCs detected in the air emissions from the treat-
ment unit into the atmosphere.
The ozone destruction unit (Decompozon unit) destroyed reactor off-
gas ozone to levels less than 0.1 ppm (OSHA Standards) with destruc-
tion efficiencies greater than 99.99%.
The Ultrox system achieved removal efficiencies as high as 90% for
total VOCs present in the groundwater at the LB&D site. The removal
efficiencies for TCE were greater than 99%. However, the maximum
removal efficiencies for 1,1-DCA and 1.1,1-TCA were approximately 65%
and 85%, respectively.
The removals of 1,1-DCA and 1,1,1-TCA are due to both chemical
oxidation and stripping. Specifically, 12 to 75 % of the total removals
for 1,1,1-TCA and 5 to 44% of the total removals for 1,1-DCA were due
to stripping. However, stripping for TCE and vinyl chloride was ob-
served to be less than 10%. For other VOCs, such as 1,1-dichloroethene,
benzene, acetone and 1,1,2,2-tetrachloroethane, stripping was found to
be negligible. VOCs present in the gas phase within the reactor at levels
of approximately O.I to O.S ppm were removed to below detection levels
in the Decompozon unit.
Based on the GC and GC/MS analyses performed for VOCs, semi-
volatile organics and PCBs/pesticides, no new compounds were dis-
covered in the treated water. The organics analyzed by GC methods
represent less than 2% of the TOC present in the water. Very low TOC
removal occurred, a result which implies that partial oxidation of
organics took place in the system but not complete conversion to carbon
dioxide and water.
The Ultrox system's average electrical energy consumption was
approximately II kwh/hr of operation.
ACKNOWLEDGEMENT
The authors sincerely thank Dr. Gary Vfelshans, PRC Environmen-
tal Management, Inc. for managing the field demonstration and for
reviewing this paper.
REFERENCES
I CH2M Hill, Preliminary Site Assessment Report for iHe Lorrntz Barrel ant
Drum Silt, 1986.
2. PRC Environmental Management. Inc.. and Engineering-Science, Inc.,
Demonstration Plan for the Ultra* International UV/Oadation Process, pie-
pared for U.S. EPA. Fcb 1989
3. PRC Environmental Management. Inc.. and Engineering-Science. Inc., Tech-
nology Evaluation Report SITE Program Demonstration of the Ultna
International UV/Oiidation Technology, in preparation for U.S. EM.
4. Methods for the Chemical Analysis of Wuer and Wastes. ERMJOO/4-79-02flt
U.S. EPA Environmental Monitoring and Support Laboratory, Cincuuitti.
OH. 1983.
5. U.S. EPA Test Methods for Evaluating Solid Haste. Volumes IA-1C: Labora-
tory Manual. Physical/Chemical Methods; and Volume II: Field Manual,
Physical/Chemical Methods. SW-846. Third Edition. Office of Solid Waste.
U.S. EPA. Document Control No 995-001-00000-1. 1986.
6. Boltz. D.F., and Howell. J.A.. Hvdrogrn Peroxide, ColorimetricDetermi-
nation ofNonmetals. John Wiley A Sons. New Ybrk. NY. 1979. 301-303.
7 APHA. AWWA. and WPCF. Standard Methods for the Eaumnation ofWaer
and Hbstewaer. 16th Ed . 1985.
8. Bader, A., and Hoigne, J . Determination of Ozone in Water by Indigo
Method. Ozone Set. and Eng . 4. 169. 1982.
9. The National Primary and Secondary Ambient Air Quality Standards, 40
CFK Pan 50. Appendix D— Measurement of Ozone in die Atmosphere.
10. N1OSH. Manual of Analytical Methods. Third Edition. U.S. Department
of Health and Human Resources, DHHS (NIOSH) Publication No. 84-KW,
1984.
11. Operating instructions provided with the instruments.
12. U.S. EPA Superfund Public Health Equation Manual, EPA 540/1-86/060,
Office of Emergency and Remedial Response, Washington, DC 1986.
13. Hoigne. J.. and Bader. H.. Ozonation of Water: Role of Hydroxyl Radicals
as Oxidizing Intcrmediales, Science. 19. pp. 782-784. 1975.
412 U.S. EPA SITES *
-------
Evaluation of the Soliditech SITE
Solidification/Stabilization Technology
Walter E. Grube, Jr.
U.S. EPA
Cincinnati, Ohio
Kenneth G. Partymiller
PRC Environmental Management, Inc.
Chicago, Illinois
Danny R. Jackson and Debra L. Bisson
Radian Corp.
Austin, Texas
ABSTRACT
The Soliditech technology demonstration was conducted at the
Imperial Oil Company/Champion Chemicals Superfund Site in Mon-
mouth County, New Jersey. The primary contaminants at this site include
PCBs and lead. Oil and grease and other metals are considered
secondary contaminants.
The Soliditech process consists of mixing the waste material with
proprietary additives, pozzolanic materials and water in a batch mixer.
Methods used to evaluate the effectiveness of the process include:
(1) batch extraction and engineering test, (2) long-term extraction and
leaching tests, (3) petrographic examination and (4) structural integrity
observations.
Three different waste types were treated: contaminated soil, waste
filter cake material and a filter cake-oily sludge mixture. Pure sand
was substituted for waste in one solidification batch to provide samples
to evaluate the chemical constituents of process reagents.
The analytical results did not indicate the presence of PCBs and vola-
tile organic compounds (VOCs) in the TCLP extracts of treated wastes.
Metals concentrations were reduced significantly in TCLP, EP Toxicity
and BET extracts of treated compared to un-treated wastes. Low con-
centrations of phenols and cresols were detected in some post-treatment
TCLP extracts. Negligible release of contaminants was observed from
all extraction and leaching tests performed on solidified samples. The
pH of treated waste was near 12. Unconfined compressive strength of
treated wastes was high; permeability was very low. Weight loss of
treated samples after repeated wet/dry and freeze/thaw cycles was very
low.
INTRODUCTION
The U.S. EPA's Office of Research and Development has been
carrying out the Agency's formal program to accelerate the develop-
ment, demonstration and use of new or innovative technologies which
can provide permanent cleanup solutions for hazardous waste sites.
The Soliditech, Inc.'s waste solidification/stabilization process was
the seventh technology to be demonstrated within this Superfund
Innovative Technology Evaluation (SITE) program.
In cooperation with U.S. EPA's Office of Solid Waste and Emergency
Response (OSWER), the Imperial Oil/Champion Chemical Superfund
site in New Jersey was selected as the location to demonstrate the
Soliditech SITE technology. This site currently is partially occupied
by a private company involved in blending and packaging oil products.
Technical staff of the New Jersey Department of Environmental
Protection (NJDEP) provided data describing the characteristics and
extent of contamination at this site and assisted U.S. EPA in public
relations aspects of the demonstration.
This technology demonstration was conducted in early December,
1988. A batch-mixer, a supply of portland cement, Urrichem reagent,
other additives for their formulation and accessory equipment were
provided by Soliditech, Inc. The U.S. EPA's support contractor provided
a sampling team. The demonstration was completed over a five-day
period, resulting in nearly 14 yd3 of solidified material and over 300
individual samples for analyses of the numerous parameters applied
to evaluate this technology.
PCBs and lead were the primary contaminants of concern on the
Imperial Oil/Champion Chemical site. These contaminants were
determined in TCLP and EP Toxicity extracts of untreated and treated
wastes to assess chemical stabilization by the Soliditech process.
PURPOSE
The primary goal of the SITE program is to evaluate the effective-
ness of a technology by conducting a field-scale demonstration of each
technology, collecting samples of treated waste materials and analyzing
data from a variety of laboratory tests.
TEST METHODS
The Soliditech SITE technology evaluation was based on the results
of laboratory tests on samples of waste material before and after treat-
ment. Physical tests included particle size analysis, water content, un-
confined compressive strength', bulk density of treated waste,
permeability of treated waste and wet/dry and freeze/thaw tests on treated
waste2. Extraction tests included TCLP extraction, EP Toxicity, Batch
Extraction Test, American Nuclear Society 16.13 and Waste Interface
Leaching Test4. U.S. EPA SW-846 methods were applied for pH, Eh,
total dissolved solids, total organic carbon, oil and grease, VOCs, semi-
volatile organic compounds, PCBs and metals5.
Methods used in the evaluation of the Soliditech process were
described in the Demonstration Plan, which was written and peer-
reviewed prior to initiation of field activities6. This Demonstration
Plan also included an approved Quality Assurance Project Plan which
described all planned sample acquisition and analytical methods.
APPROACH
Contaminated soil was excavated from a pit approximately 5 ft wide,
3 ft deep and 8 ft long in off-site Area One of this Superfund site. Filter
cake waste was collected from the open face of a waste pile (Fig 1).
Oily sludge was scooped from an abandoned storage tank with a bucket
and stored in steel drums until the waste was processed. A filter cake/oily
sludge mixture was prepared for processing by mixing equal parts of
filter cake and oily sludge. All waste feedstocks were screened to pre-
vent large objects such as rocks, roots, bricks or other debris from being
incorporated into the treated waste. Although this debris would not have
interfered with the Soliditech process, it was removed to prevent inclu-
U.S. EPA SITES 413
-------
sion within samples taken for analytical testing.
A - Proprietary additives N
B - Portland Ceawnt supply S
0 - Onus containing oily sludge U
F - Forms for treated waste Monoliths H
Mixer
Satpte preparation
Urrlcheai supply
Filter cake waste pile
Figure I
Soliditech Technology Demonstration Operations
Waste materials were mixed in a ribbon-blender shown in Figure I.
Water was added to the waste within the mixer to provide the proper
mixing consistency. Portland cement, other specific additives formu-
lated by Soliditech staff and Urrichem were then added and mixed.
The mixture was discharged from the mixer into l-yd' plywood forms
(Fig. I). Aliquots of the slurried mixtures were taken from the forms
and poured into waxed cardboard and PVC cylindrical forms, of several
different sizes, to provide samples for various physical and chemical
analyses.
All materials were allowed to set for 28 days inside a heated ware-
house. Cylindrical samples were transported to the storage area of an
analytical laboratory. Nearly 14 yd' of treated waste were contained
within the plywood forms to form the treated waste monoliths (TWM).
These monoliths were placed in a two-tiered stack and covered with
a plastic sheet for subsequent long-term examination.
Figure 2 illustrates the approaches used to evaluate the effectiveness
of the Soliditech process. The Quality Assurance Project Plan, within
the project's Demonstration Plan6, specified the details of sample col-
lection and preservation, analytical protocols, matrix and surrogate spike
procedures, blanks, replicate analyses and statistical procedures to be
applied to data evaluation. Triplicate samples were provided for all ana-
lytical determinations on the treated materials. The Demonstration
Report' presents the complete data resulting from this technology
evaluation.
EVALUATION PARAMETERS
SHORT-TERM
TESTING
LONG-TERM
EXTRACTION AND
LEACHING TESTING
PETROGRAPHIC
EXAMINATION
Extraction Ttsli
TCLP
EP
ANSIS 1
BET
TCUP
EP
WILT
Cut Cytndtrt
Cut Sllbl
Thin Stclloni
Powd»(
Cha
Phyiical An.iyi.i Figure 2
Evaluation Parameters Used to Evaluate the Solidtcch Process
RESULTS
Table 1 shows the compositions of the waste treatment mixtures. The
reagent mixture includes clean sand as a substitute for waste. The filler-
cake/oily-sludge consists of filler-cake and oily sludge because Soiidtedi
preferred not to treat the oily sludge in its original liquid form.
IkMe I
SoHditcch Treatment Formulation
mu> Ut* ffllutn.*.
H
e «
M
H
II
II
Table 2 shows that, after treatment, the bulk density increased and
water content decreased in all cases. These results are attributed to the
effects of cement in the treatment process. In fact, bulk densities, penne-
abilities and UCS were directly related to the amount of type D cement
added in the process (Table I). The permeabilities of treated waste were
very low with values below I x K>-8 cm/sec. The unconftned compret-
sive strengths ranged from 390 to 860 psi.
Table 2
ffc iii t
rroperoes of
PU4M C4**Aflr
tl t 4M 14 I
n > u.i
Total chemical analyses, shown in Tables 3 through 5, for uanofed
and treated wastes indicate a variable effect from adding process ragafti
to untreated waste. PCBs varied from no observable change lo am
one-third less in the treated waste. Analyses of pure sand solidified wih
the Soliditech process showed that arsenic was present at 59 tug/kg.
Chromium, copper, lead, nickel and zinc were noted to the extent of
a few tens of mg/kg in this sand plus reagent mixture. A few rag/kg
of phenols and cresols were delected in analyses of the treated MACS
for semi-volatile organic compounds. Although the origin of (best
Table 3
Chemical Analysis of Untreated and Treated finer Cake
Vfeste* and Tbelr TCLP and EP Extracts
MAI
II
I't.OOT
II.»
4 I
It
I.IN
II
4 4 |t t
• I' •
• I I
1.4 44
t.m m
• I M
t 04 t «
« i ttn
t M •
n ».•
• H «•«
•.« *•
414 U.S. EPA SITES
-------
Table 4
Chemical Analysis of Untreated Filter Cake/Oily Sludge
Wastes and Their TCLP and EP Extracts
Total Amivi 1.
Untraitad Tr**t«d
PH
VOCi
SVOCl
PCBi
Oil and Cr««i«
U*d
Zinc
0.55
0.86
0 002
0.04
0.015
-------
system. After the equipment operator gained familiarity with waste
materials at this site, the process mixed all components into a homo-
geneous solidified product.
ACKNOWLEDGEMENTS
This SITE demonstration project was conducted under U.S. EPA
Cooperative Agreement No. CS-815494010 with Soliditech, Inc.,
Houston, Texas. Carl Brassow is the Soliditech Project Manager. The
U.S. EPA was assisted by PRC Environmental Management, Inc. in
conducting the technology evaluation under Contract No. 68-03-3484.
Bob Soboleski. Site Manager in the New Jersey Department of Environ-
mental Protection, provided valuable support in demonstration site selec-
tion and public information in New Jersey. Mr. George C. Kulick, Jr.,
Vice-Presidem of Imperial Oil Company, provided access to the property
on which the demonstration was conducted.
REFERENCES
I. ASTM. Annual Book of ASTM Standards, Vol. 4.03. American Society for
Testing and Materials, Philadelphia, PA, 1987.
2. Cote. P. (Draft) "Investigation of Test Methods for Solidified Waste Charac-
terization (TMSWC)," Wastewatcr Technology Centre, Burlington, Ontario.
Prepared for RREL. U.S. EPA, Cincinnati, OH, 1986.
3. American Nuclear Society. ANS 16.1 Laboratory Test Procedure. American
Nuclear Society. UGrange Park, IL . 1986.
4. Jackson, O.R. "Comparison of Laboratory Balch Methods and Large Cohimra
for Evaluating Leachftie from Solid Wfestes." Prepared for RREL, U.S. EPA,
Cincinnati, OH. 1988
5. US. EPA. Test Methods for Evaluating Solid Vfeste (SW-346), VbU. IA, IB,
1C and D, Third Edition. U.S. EPA Doc. Control No. 944-001-00000-1, 1986
6. PRC Environmental Management, Inc. "Demonstration Plan for the
Soliditcch. Inc., Solidification Process." WA 0-5, Contract No. 68-03-3484,
US. EPA. Cincinnati. OH, 1988.
7. US. EPA Technology Evaluation Report SITE Program Demonstration Tea.
Solidilcch, Inc., Solidification Process. US. EPA/S40/x-89/xx. RREL. US.
EPA. Cincinnati. OH, (in press).
416 U.S. EPA SITES
-------
Concurrent Application of RCRA and CERCLA
at a Unique Federal Facility:
The Hanford Site
Paul T. Day, MPH, RS
U.S. EPA
Richland, Washington
Emily M. Pimentel
Planning Research Corporation, Environmental Management
San Francisco, California
ABSTRACT
A goal of the U.S. EPA is to integrate RCRA and CERCLA at
hazardous waste sites where both laws may apply. On May 15, 1989,
the U.S. EPA, the Washington State Department of Ecology (Ecology)
and the U.S. Department of Energy (DOE) entered into an Interagency
Agreement to provide a legal and procedural framework for cleanup
and regulatory compliance at the numerous hazardous waste sites at
DOE's Hanford Site. This document is entitled the Hanford Federal
Facility Agreement and Consent Order. Hereafter, it is referred to as
the Tri-Party Agreement or the Agreement.
The objective of this paper is to describe a creative approach to in-
tegration of the RCRA and CERCLA programs and to explain the
development of an efficient, productive working relationship between
the joint regulatory agencies; the U.S. EPA, Ecology and the owner
of the Hanford Site, DOE.
INTRODUCTION
The Hanford Site is the largest CERCLA site in the nation, encom-
passing 560 mi. in Southcentral Washington. The site is bordered to
the north and east by the Columbia River and is adjacent to the northern
boundary of the city of Richland, Washington (Fig. 1). Four general
areas of the Hanford Site were proposed for inclusion on the U.S. EPA's
NPL on June 24,1988. EPA anticipates that the proposal will be finalized
in late FY-89. These four areas include over 1000 inactive waste dis-
posal and unplanned release sites, ranging in scope from minor spill
areas to burial grounds up to 100 ac. in size. The areas also contain
55 RCRA treatment, storage or disposal (TSD) groups which contain
over 300 individual RECRA units that will be closed or will be per-
mitted to operate in accordance with RCRA.
The areas include significant amounts of contamination. Estimates
of the extent of soil contamination exceed a billion cubic yards, and
there are known plumes of contaminated groundwater totaling over
230 mi. The contamination is in the form of RCRA hazardous waste,
radioactive mixed wastes (hazardous waste mixed with either high-level
or low-level radioactive waste, the hazardous component of which is
subject to RCRA regulations) or CERCLA hazardous substances (such
as radioactive waste which is not regulated under RCRA).
The State of Washington has received authorization from the U.S.
EPA to implement the state's dangerous waste program in lieu of the
federal RCRA program. In addition, the state has received authoriza-
tion to implement the U.S. EPA's radioactive mixed waste program.
The state currently is planning to apply for authorization to implement
the Hazardous and Solid Waste Amendments of 1984 (HSWA). There-
fore, an argument could be made that all of the hazardous or mixed
waste units could be investigated and remediated under either
RCRA/HSWA authority, eventually to be delegated its CERCLA
Hanford Site
Figure 1
Proposed Aggregate National
Priorities List (NPL) Areas
authority to the state. There also was a concern about how to deal with
contaminated groundwater plumes which contained contaminants from
both CERCLA and RCRA regulated units. The DOE was very con-
cerned that only one regulatory agency direct the investigation and
remediation at each unit and that cleanup standards be consistent under
the RCRA and CERCLA authorities. These potential sources of conflict
and confusion were recognized early in the process of the Tri-Party
Agreement negotiations and were primary topics and underlying themes
throughout the negotiations.
U.S. EPA SITES 417
-------
SCOPE AND OBJECTIVES OF THE
TR1-PARTY AGREEMENT
Section 120 of CERCLA requires the U.S. EPA to enter into Inter-
agency Agreements with Federal Facilities which are listed on the NPL.
The U.S. EPA encourages its state counterparts to be involved with such
agreements since, in many cases, the states' cleanup standards will be
applicable or relevant and appropriate to be a CERCLA action. In this
case. Ecology had an ongoing RCRA program at the Hanford Site, and
the need for an active state role in the Interagency Agreement was even
more evident. The formal negotiations for the Tri-Parly Agreement
began in February, 1988, resulting in a draft document which was issued
for public comment in February, 1989. The final Agreement was signed
and became effective on May 15, 1989.
The three parties recognized the need to incorporate the CERCLA
program, the federal RCRA/HSWA program and the state's Dangerous
Waste program into the Tri-Party Agreement. As such, the scope of
the Agreement includes all actions leading up to CERCLA remedial
actions and RCRA/HSWA corrective measures. The Agreement also
includes activities related to RCRA interim status compliance, RCRA
permitting and RCRA closure activities—all of which apply to TDS
units that last received waste after Nov. 19, 1980.
There were numerous specific objectives that the parties intended
to meet through the Tri-Party Agreement. A major objective was to
bring the Hanford Site into full RCRA compliance and to achieve full
cleanup within 30 yrs. The panics considered this a reasonable period
of time based on the extent of contamination, complexity of the site
and wastes involved, need for development of new technology and realis-
tic expectations for funding. Another objective was to create a clear
picture of the work that needs to be done by specifying detailed schedules
and milestones. This type of planning is necessary to support the large
amounts of money that DOE will have to request over the next 30 yrs.
Another specific objective, as noted above, was to provide specific roles
and a plan of interaction between the regulatory agencies. All three
parties considered this to be an essential element in order to minimize
potential conflicts and disputes as the Agreement is implemented over
the years. Another objective focused on a coordinated RCRA-CERCLA
public involvement process in order to maximize available resources.
to avoid duplication of effort and to provide a consistent format for the
public.
RCRA - CERCLA INTEGRATION
Because of the large number of sites or units to be investigated and
remediated at Hanford, the CERCLA "operable unit" concept was
deemed necessary. The parties agreed to divide the site into 74 opera-
ble units (Fig. 2) plus four groundwater operable units. Each operable
unit will undergo a separate investigation and remediation process on
a priority basis. The criteria used to assign specific waste management
units to operable units are identified in the Tri-Party Agreement, as
are the criteria used to prioritize operable units for scheduling purposes.
Figure 2
RCRA / CERCLA Integration
There has been a recent effort by the U.S. EPA to provide better
coordination between the RCRA and CERCLA programs, specifically
in regard to remedial actions or corrective measures. Some of the
primary examples of this effort are the requirement to adhere to ap-
plicable or relevant and appropriate requirements as part of CERCLA
remedial actions, elimination of RCRA permitting requirements for
certain activities during CERCLA remedial actions, significant enhance-
ment of quality assurance provisions to RCRA laboratory protocols
(SW-846), the U.S. EPA's corrective action rule which contains sig-
nificant parallels to the CEDRCLA approach and the U.S. EPA's evalua-
tion of a RCRA "deminimus rule" (yet to be proposed) which would
consider cleanup standards for listed wastes as something other than
background concentrations. In short, there is a recognition of the need
to draw these two statutes closer together, whenever possible, to
eliminate conflicting procedures and requirements. The U.S. EPA's
general approach to the private sector (i.e.. non-federal facilities) is
that if RCRA applies at a facility, the U.S. EPA will not pursue that
facility through the CERCLA NPL ranking process. All of the cleanup
or corrective actions would be taken under RCRA authority. At federal
facilities, both (he RCRA and CERCLA statutes apply and, therefore,
a rational approach to integration is necessary.
The parties also have integrated certain administrative elements of
RCRA and CERCLA in addition to technical elements. For instance,
a single administrative record is being maintained by DOE and its coo-
tractors. As a federal facility, DOE is required to maintain the adminis-
trative record under CERCLA. Since many of the CERCLA activities
are closely tied to RCRA work, the parties decided that DOE would
maintain one overall administrative record, to include both RCRA and
CERCLA. The system in place allows sorting of the data base in i
number of different ways, allowing the user maximum utility by
reviewing the entire record or by extracting specific components.
Another task that was viewed as a cost-saving, practical step was the
consolidation of the public involvement activities under RCRA and
CERCLA. A significant amount of time was spent developing the joint
Community Relations Plan to merge the requirements of both programs
into a single process. This joint Community Relations Plan will simplify
the process for both the parties and the public and will maximize the
efficiency of available resources used for public involvement.
RCRA AND CERCLA AUTHORITIES FOR
PAST PRACTICE UNITS
The parties reached agreement that any of the operable units could
be managed under either RCRA or CERCLA authority. This was a
major step during negotiations. Accordingly, each of the first 20 operable
units has been assigned to either the RCRA past-practice program or
the CERCLA program for investigation and remediation. Additional
assignments will be made annually, as the work schedule is updated.
The Tri-Party Agreement requires the U.S. EPA and Ecology desig-
nate the regulatory process to be used at these additional operable units.
Most of the past-practice activities involved mixed waste. Therefore,
the first area of agreement between the parties was that, in general,
the radioactive component of mixed waste would be addressed as part
of a RCRA corrective measure. This does not extend RCRA or state
Dangerous Waste authority to regulate radioactive wastes; rather, it pro-
vides an understanding that DOE has agreed to address radioactive
wastes as part of a comprehensive investigation and corrective action
at an operable unit, whether the operable unit is being managed under
RCRA or CERCLA. The Agreement states that "the corrective action
process selected for each operable unit shall be sufficiently compre-
hensive to satisfy the technical requirements of both statutory authori-
ties and the respective regulations" (Fig. 3). It is important to note that
the authority for radioactive wastes remains under CERCLA. This
agreement eliminates the potential for a worst-case scenario—the
application of requirements of both programs at the same unit, a situa-
tion which would not serve the best interest of any party or the public.
The success of this approach requires flexibility in interpretation of
the statutes and regulations by the U.S. EPA and Ecology and is predi-
cated on certain assumptions and requires some concessions on the part
of all parties. It also provides a solid framework under which the parties
418 U.S. EPA SITES *
-------
can work cooperatively toward cleaning up the Hanford Site.
INCLUSION OF RCRA LAND DISPOSAL UNITS
IN OPERABLE UNITS
Fourteen of the operable units contain significant RCRA land disposal
units that received hazardous waste or mixed waste after Nov. 19, 1980.
All of these TSD units are scheduled for closure under RCRA and,
therefore, operational activities at these units will not be included in
the Hanford RCRA permit. In some cases, the units will be covered
in the Hanford RCRA permit for post-closure activities. The remaining
41 TSD groups contain only storage and treatment units. These storage
and treatment groups have not been assigned to operable units, since
the level of investigations required for storage or treatment Part B per-
mit applications and closure plans is less comprehensive than that re-
quired for land disposal units (Fig. 2). Accordingly, the schedule for
submittal of Part B applications and closure plans for these groups is
separate from the operable unit schedule. The need for RCRA - CERC-
LA integration obviously centered around those operable units which
contained the 14 RCRA land disposal groups. The parties agreed to
the basic approach that the RCRA land disposal groups would be in-
vestigated concurrently with the past-practice sites within the operable
unit and that the overall priority and schedule for the operable unit would
drive the schedule for submittal of the closure plans and post-closure
Part B applications. For this approach to succeed, the parties had to
agree that a CERCLA RI/FS for an operable unit would yield a suffi-
cient level of detail to develop a closure plan or post-closure Part B
application. As with the integration of past-practice units, the worst
case scenario, from an efficiency standpoint, would be a duplication
of effort by the U.S. EPA and Ecology, using their different authorities.
In some cases, identically designed units located side-by-side may
have received the same RCRA regulated waste streams, differing only
in the date on which waste receipt ended. If that date was after Nov. 19,
1980, the unit would be a RCRA TDS unit. If the date was prior to
Nov. 19, 1980, the unit would be regulated as a past-practice unit under
either RCRA/HSWA or CERCLA. The parties concurred that a single
investigation and coordinated timing for a remedial action and closure
activity would be the most efficient method of dealing with this issue.
For this reason, the parties agreed that only one investigative process-
either RCRA or CERCLA—would be used within an operable unit.
As stated earlier, this approach required agreement that the investigative
procedures of CERCLA and RCRA as implemented at the Hanford Site
would provide results that could be used to support technical decisions
under either program.
LEAD REGULATORY AGENCY CONCEPT
The design of an efficient and comprehensive regulatory compliance
and cleanup program for implementation under the Tri-Party Agree-
ment incorporated numerous factors. One major factor, the integration
of RCRA and CERCLA authorities, has been discussed above. Before
this system could begin to work, the parties had to come to agreement
on another major element—the roles of the two regulatory agencies.
One can envision numerous logistical and efficiency problems that would
be encountered if both regulatory agencies were to insist on full in-
volvement with their respective authorities.
It became apparent early in the negotiations that a work-sharing
approach for the regulatory agencies would be necessary. This approach
was carefully crafted in the Agreement so that responsibilities were
shared and clearly spelled out, but that authorities could not be trans-
ferred arbitrarily between the U.S. EPA and Ecology. In this way, the
regulatory agency with the responsibility for oversight can fulfill its
obligation to keep the projects running as efficiently as possible,
obtaining the co-signature of the agency having authority, when
necessary.
The concept of a lead regulatory agency was developed for the regula-
tory oversight of each operable unit. Its definition and use is restricted
to that level. The U.S. EPA and Ecology will decide which agency will
be assigned as the lead regulatory agency in each case. Such assign-
ments have been made for the first 20 operable units, and additional
assignments will be made during each annual update of the work
RCRA Facility
Assessment
(RFA)
Preliminary
Assessment/
Site Investigation
(PA/SI)
Identify
Releases
Needing Further
Investigation
RCRA Facility
Investigation
(RFI)
Remedial
Investigation
(Rl)
Characterize
Nature, Extent,
and Rate of
Release
Corrective
Measures
Study
(CMS)
Feasibility
Study
(FS)
Evaluate
Alternatives and
Identify Preferred
Remedy
Draft
Permit
Modification
Proposed
Plan
Propose
Selected
Remedy
Public
Comment
Public
Comment
Public
Participation
RCRA
Permit
Record of
Decision
Authorize
Selected
Remedy
Corrective
Measures
Implementation
(CMI)
Remedial
Design/
Remedial Action
(RD/RA)
Design and
Implement
Chosen
Remedy
Figure 3
RCRA / CERCLA Comparison
schedule. The regulatory agency not designated as the lead regulatory
agency will automatically be designated as the support agency. The roles
are defined below.
Lead Regulatory Agency Responsibilities
The lead regulatory agency is responsible for overseeing all activi-
ties that are related to a given operable unit. This may include a com-
bination of RCRA TSD and CERCLA work, RCRA TSD and RCRA
past-practice work, or CERCLA work without any RCRA activity. The
lead regulatory agency serves as the primary contact for DOE, the sup-
port agency or the public rfegarding any questions or issues at the
operable unit.
Ecology may serve as the lead regulatory agency for an operable unit
that has been designated under either the RCRA past-practice program
or the CERCLA program. Likewise, the U.S. EPA may be the lead
regulatory agency for such operable units. Ecology and the U.S. EPA
have agreed to certain general criteria in the Agreement for designating
the lead regulatory agency. Much of this agreement centers around
whether significant TSD units are present in the operable unit. Such
operable units generally would be assigned to Ecology, and the RCRA
past-practice authority would be used. Since there are only 14 of these
situations, as discussed earlier, this criterion will have no effect on the
majority of the assignments. For those operable units involving only
radioactive waste, the U.S. EPA generally would be the lead regula-
tory agency and the CERCLA process would be designated.
U.S. EPA SITES 419
-------
One important criterion for the designation of the lead regulatory
agency is the availability of each agency's resources at any point in time
to provide adequate oversight of the activities at the operable unit. Main-
taining the proper balance of resources will be an ongoing effort by
both regulatory agencies.
Support Agency Responsibilities
Both the U.S. EPA and Ecology believe that it is important that the
support agency stay informed of the progress at every operable unit.
In some cases, the lead regulatory agency and the support agency roles
may be reversed at adjacent or nearby operable units. These situations
will require close coordination of Held activities and data, since tech-
nical information obtained at one operable unit may overlap to another.
Certainly, the level to which the support agency can become involved
will depend upon available resources and the issues at hand. The sup-
port agency may submit comments on work plans or other documents
submitted by DOE for review. In such cases, the support agency will
submit its comments to the lead regulatory agency in order to maintain
a single point of contact and to avoid the potential for DOE to receive
conflicting comments from the regulators.
CONCLUSIONS
The two processes described above for designation for the regula-
tory process and of the lead regulatory agency at each operable unit
form a basic structure on which the Tri-Party Agreement is imple-
mented. While the approach may seem simple from an overall view
of efficiency and what makes sense, the construction of this approach
into a working document was a complex task. It required a substantial
amount of initial technical work to accurately identify the universe of
waste sites and to design and prioritize the operable units. From the
point, it required significant negotiations between the U.S. EPA and
Ecology to determine appropriate regulatory processes and lead agency
responsibilities for the operable units. This type of Interagency Agree-
ment has been referred to as a "carve out agreement," since much of
the workload distribution, has been determined prior to signature of
the document. By expending a large amount of effort in initial planning,
the parties believe that the total resource needs for this project have
been established with some degree of accuracy. This makes it much
easier for each party to identify and justify its resource needs over both
the short and long-term.
One must keep in mind that this approach was developed specifically
for the Hanford Site, due to its size; the number of units; the state's
authorization status and its involvement and commitment to regulatory
compliance and cleanup; and the number of situations which would
require the integration of RCRA and CERCLA. This approach may
not be appropriate for all sites at which RCRA and CERCLA integra-
tion is an issue. It can only work when all of the panics negotiate in
a cooperative manner and when the U.S. EPA and the state are willing
to place a significant amount of trust and confidence in each other.
The U.S. EPA and Ecology have included a dispute resolution process
section in the Tri-Party Agreement that can be implemented in the event
they can not come to agreement on certain integration issues. The DOE
is not a party to that dispute resolution process since it involves only
decisions between the regulatory agencies.
Present Status
The bottom line of any methodology can be simply stated as "Does
it work?" The parties to this Agreement are now 3 mo. into implemen-
tation. At this point, we are still hiring staff and developing some of
the detailed procedures necessary for efficient implementation. To dale,
work plans have been submitted for the first five operable units. Three
of these are under the CERCLA process with the U.S. EPA as the lead
regulatory agency and two are under the RCRA process with Ecology
as the lead regulatory agency. One of the Rl/FS work plans has been
approved and field work has been scheduled. Overall, the process is
running relatively smoothly, and we are optimistic that it will get belter
as to get over some of the hurdles for the first time.
Future
Many federal facilities currently are negotiating cleanup and com-
pliance agreements with the U.S. EPA and/or state agencies. As these
facilities are added to the NPL, some are faced with the potential
conflicts of concurrent application of CERCLA and RCRA. To the extent
that it may apply, the approach used by DOE, the State of Washington
and the U.S. EPA could be used as a framework or model for negotia-
tions between such federal facilities and the regulators. The experience
gained at Hanford can be used to foresee and eliminate many of the
conflicts and redundancies of the two regulatory programs, resulting
in a streamlined approach to cleanup and compliance.
420 U.S. EPA SITES
-------
Hazardous Waste Decontamination
With Plasma Reactors
Laurel J. Staley
U.S. Environmental Protection Agency
Cincinnati, Ohio
ABSTRACT
The use of electrical energy in the form of plasma has been con-
sidered as a potentially efficient means of decontaminating hazardous
waste. Only a few attempts have been made to actually treat hazardous
waste with plasma, however. This paper discusses both direct and in-
direct waste heating with plasma. Direct heating involves the direct in-
jection of liquid waste into the plasma plume. Indirect heating involves
using the plasma to create a bath of molten solid material which is used
to heat and decontaminate solid hazardous waste. This paper summarizes
the experience to date with plasma based hazardous waste treatment
and discusses the implications of the limited data available.
INTRODUCTION
A plasma is created when gases are ionized by passing through an
electric field strong enough to strip electrons from the molecules of
the gas. Even though the aggregate gas remains electrically neutral,
this occurs only because it is made up of equal numbers of positively
and negatively charged particles. These charged species contain a high
level of energy. When the ionized species in the plasma recombine with
the stripped electrons, significant amounts of energy are released. This
energy can be used in a variety of ways. Plasma torches have been used
in the metals industry and have been considered for use in wood gasi-
fication, glass manufacturing and in radioactive waste nitrite
reduction1'2'3. Because of the large amounts of energy that can be deli-
vered by plasmas, plasma torches have been considered as a possible
means of decontaminating hazardous wastes.
Relatively little data are available on the use of plasma to treat
hazardous waste. Basically, there are two ways in which plasma can
be used to decontaminate hazardous wastes. One way is to inject the
waste directly into the plasma. In this way, plasma energy is used to
break apart molecules of various hazardous substances into their con-
stituent atoms. The other way is to feed waste into either a molten metal
bath or a bath of molten soil. While direct heating has been shown to
treat only liquids and gases, indirect heating can also treat solids. The
bath heats the waste feed, volatilizing the waste contaminants. Once
volatilized, these waste contaminants are thermally destroyed in the hot
atmosphere of the reactor. The molten material solidifies into a vitri-
fied mass which, if containing heavy metals, is non-leachable.
This paper discusses the relative advantages and potential disadvan-
tages of the use of both direct and indirect heating with plasma as a
means of treating hazardous waste. Since there are only limited data
available on this use of plasma, more questions will be asked than
answered. Hopefully, asking questions will stimulate discussion on this
topic. The U.S. EPA is interested in obtaining as much information as
possible on the use of plasma technology because the Agency currently
is evaluating it for its potential use in hazardous waste decontamination.
The potential advantages of the use of plasma in this application
include the following.
Plasma may be able to deliver high levels of energy to the waste.
When injected directly into the plasma plume, hazardous wastes are
directly subjected to the high intensity plasma energy. This energy is
believed to be sufficient to break the molecular structure of the individual
waste compounds into their atomic constituents and is fer in excess of
what is possible with conventional incineration.
Upon recombination, carbon dioxide, water and other common and
relatively innocuous end products of combustion are formed. Products
of Incomplete Combustion (PICs) are not believed to be formed in sig-
nificant quantities. PIC formation can be a problem with conventional
incineration and the use of plasma could eliminate it.
Plasma may be able to treat metal contaminated solids.
When plasma is used to create a molten bath of soil, metals or glass,
a very uniform, high temperature environment is created. While the
temperatures achieved are far below plasma temperatures, they are hot
enough to ensure the thermal destruction of organic waste constituents
treated in this way. Cold spots where PIC formation may be exacer-
bated are eliminated in this environment. In addition, it may be possi-
ble to entrap metal contamination in the melt. Upon cooling, this would
result in a non-leachable solid residue which would not require further
treatment. Fluxing agents could be added to the melt to adjust its proper-
ties (i.e., melting point and residue leachability).
If enough energy is provided through the plasma, the process could
be very versatile and would be able to treat waste with any physical
characteristics including entire waste-filled drums. This capability would
eliminate the need for waste pretreatment. Combined with the produc-
tion of non-leachable organic-free residues, the use of indirect plasma
heating could eliminate the need for either pre- or post-treatment. This
capability would reduce the overall costs associated with plasma treat-
ment and may compensate for the likely additional energy costs for
this process
Both oxidizing and non-oxidizing atmospheres can be used while still
achieving very high temperatures.
Since the only requirement for creating a plasma is that the gas used
be ionizable, plasma can be used in the absence of oxygen in situations
in which high temperature pyrolysis is desirable. Since the ionization
potential of gases varies, torch efficiencies will vary with the torch gas
used2'7.
Two instances in which plasma is being used or is being developed
for use in hazardous waste treatment will be discussed. One instance
involves direct heating by plasma and the other concerns indirect heating.
U.S. EPA SITES 421
-------
DIRECT HEATING WITH PLASMA
Waste decontamination through direct healing by plasma has been
studied more than indirect heating. For direct heating, non-transferred
torches are used. Figure 1 is an illustration of ;i non-transferred plasma
torch in which both positive and negative electrodes of the torch are
contained in the body of the torch itself. An electric arc is created
between the two electrodes. Gas passing between the terminals passes
through the electric arc and is ionized, thus forming the plasma.
Rear
Electrode
Arc Gas
Hot Gas
Figure 1
Non-transferable lurch
The Westinghouse Pyroplasma unit, originally developed by Pyrol-
ysis Systems Inc. of Ontario, Canada, was tested for possible use at
Lave Canal by the New. York Department of Environmental Consecu-
tion (NYDEC) and the U.S. EPA. A schematic diagram of the Mobile
Pyroplasma Unit is provided in l-igure 2'. The entire system,
including the analytical laboratory, is contained in one trailer and
operated as follows.
i
Scrubber
Torch QM
— ^Torchj -
DbentUquU * '
Feed
Emergency
Carbon
Filer
Water
toDralnf
Figure 2
Mobile Pyroplasma Unil
Sample Uw
Mobto
Laboratory
Up to 1 gpm of solids-free liquid feed was injected into the plume
of a 350-kw non-transferred electric torch. Air was used as the torch
gas. Exhaust gas from the plasma torch was treated by aqueous scrubbing
and water separation prior to being released into the atmosphere via
.1 flare. Exhaust gasses were sampled upstream of the flare and were
analyzed on-site through the use of continuous emission monitors and
an on-line gas chromatograph*.
Two tests took place from 1982 to 1986. Both tests involved the treat-
ment of simulated liquid wastes consisting of chemicals diluted in a
mixture of meihanol and methyl ethyl kctone. The first lest treated car-
bon tetrachloride diluted in MeOH/MEK. The second test treated a
mixture of PCBs also diluted in MeOH/MEK. These liquids contained
no suspended solids and were free of water. Table 1 shows the Des-
truction and Removal Efficiencies (DREs) achieved.
Table 1
ORE* Achieved During (he CC14 and PCB Trial Burns4
Chemical Test 1 Test 2 Test ?
CC14 99.99995 99.99996 99.99996
Monodecachlorobyphenyl 99.99999 99.99994 99.9999
TMdecachlorobyphenyl 99.999999 99.99997 99.999999
A few parts per trillion of Dioxins and Furans were discovered in
the stack gases during the PCB trial burns.
In use. the Mobile Pyroplasma Unil was quite sensitive to changes
in waste feed or operating conditions. Virtually no solids could be
present in the feedstream without causing operational problems. Cost
data are not available from either study Torch efficiency was 80%'.
Recent data made available by Westinghouse Environmental Systems
and Sen ices confirms the ver> high DREs achieved earlier (and shown
above). These tests involved the treatment of 300 gal (at 1 gpm) of trans-
former oil containing 70-80% PCBs by weight1 These data are shown
in Table 2
Table 2
Recent PCB Destruction Test Results'
ORE X
HC1 Ib/hr
PartIculate Gr/DSCF
Test 1 Test 2 Test 3
99.999999 99.999999 99.999995
0.941 0.972 0.343
0.00837 0.00845 0.00441
The results of the operating experiences with the Mobile Pyroplasma
Unit raise several questions.
Question 1.
Is injecting waste directly into a plasma hazardous waste "overkill"
since the energy delivered is far in excess of that normally needed to
thermally destroy most waste compounds'.'
Question 2.
Is injecting waste directly into a plasma really worth all of the precau-
tions necessary to assure solids-free and uniform feed since PICs can
still form from the recombination of fragmentary molecules during the
decay of the plasma?
Even though the levels of PIC material produced were relatively low,
the presence of PICs in the exhaust suggests that the waste chemical
molecules are not completely broken down into their constituent atoms
when injected into a plasma plume. The free radical chain reactions
of conventional fossil fuel/air combustion result in the formation of some
422 U.S. EPA SITES
-------
PICs when the free radicals and/or fragmentary molecules recombine.
Is it possible that these recombination reactions still occur even in a
very high temperature plasma environment?
Evidence that this can happen is provided by Drost, et al.5, in a
study of the effect of recombination reactions on the formation of
products from reactions at plasma conditions in shock tube experiments.
Conditions in the pressure waves of shock tubes are similar to plasma
conditions and so were used to study the effects of hydrogen ions on
the formation of products produced when hydrogen was used as a plasma
gas. Hydrogen decreased the formation of acetylene from methane,
presumably by recombining with methyl radicals initially created in
the plasma. Hydrogen also decreased the formation of soot from these
reactions as a result of recombining with methyl radicals. These results
contradict the previously held notion that the torch gas used to create
the plasma acted only as a medium for the transfer of energy and did
not react with the waste. They also suggest that PICs can be formed
from the use of plasma gas just as they can be in conventional com-
bustion.
These studies, in combination with the observed results, suggest that
not even the intensive energy of plasma is sufficient to guarantee that
waste compounds will be completely oxidized without forming any un-
desirable side products. If this is true, yet another question is raised.
Question 3. Is the type of torch gas used very important in deter-
mining the types and levels of any PICs formed? Or, is it merely neces-
sary to use air or oxygen?
Question 4.
Finally, can real waste streams be treated in this device given the
need to filter out all solids and the overall sensitivity of the process
to the properties of the waste materials fed?
INDIRECT HEATING WITH PLASMA
For indirect heating, a transferred torch is used. Figure 3 is an illus-
tration of a transferred torch. As the name implies, transferred torches
strike an electric arc between the torch and a conductive body external
to the torch. That body can be the heat conducting medium used in
the case of indirect heating.
Electrode
Insulator
Nozzle
Melt (O Ground Potential)
There is less experience with indirect heating. The Centrifiigal Reactor
developed by Retech Inc. of Ukiah, California, uses a transferred torch
to melt soil and debris. Figure 4 is a schematic diagram of the Cen-
trifugal Reactor8. The system operates as follows. Waste is fed through
a screw feeder and enters the rotating tub in the upper chamber. There
the rotating tub retains the solid waste for sufficient time to allow the
500-kw torch to melt and vitrify the soil. The torch fuses the solid matter
into a slag, presumably trapping less volatile metals. The hot environ-
ment helps to oxidize the organic material volatilized from the slag.
Air pollution control devices downstream of the reactor's secondary
chamber remove paniculate and acid gases from the exhaust gas stream.
FCEDCR
(sot'd mjltrill)
PIASMA TORCH
Figure 3
Transferred Torch
Figure 4
Centrifugal Reactor
The reactor currently is being evaluated under the Superfund Innova-
tive Technology Evaluation (SITE) program at the U.S. Department
of Energy's (DOEs) Magnetohydrodynamics Component Development
Integration (CDIF) in Butte, Montana. The demonstration of this device
will begin as soon as development work is completed which will
optimize the performance of the reactor. DOE is interested in evaluating
the reactor for its potential use in consolidating Transuranic waste
currently stored at the Idaho National Engineering Laboratory (INEL).
DOE has planned a 6-mo study to evaluate its potential usefulness in
this application. The study will take place at the CDIF and will occur
in conjunction with U.S. EPA's performance evaluation under the SITE
program.
Even though there are no results yet, several questions can be raised
about the applicability of indirect plasma heating to the treatment of
hazardous waste.
Question 1.
How energy intensive is indirect plasma heating relative to conven-
tional incineration? Is added energy input worth it in order to form a
non-leachable solid residue that will require no further treatment? Or,
is it cheaper and less risky to incinerate the organics and treat the ash?
Question 2.
How sensitive is indirect plasma heating relative to changes in waste
properties such as water content and heating value. Indirect heating
with plasma is presumed to be less sensitive to changes in waste proper-
ties. Since plasma is capable of melting rocks etc. the belief has been
that treatment processes based on indirect heating with plasma were
omnivorous and required very little, if any, waste pretreatment. Is this
.true or just a myth? Given the potential difficulty of providing suffi-
cient energy to these processes, so much energy might be used to heat
rocks and water that not enough would be available to thermally des-
troy hazardous wastes.
U.S. EPA SITES 423
-------
Question 3.
How efficient is the use of plasma in the heating of solid material
when compared to the use of other forms of electrical heating?
Question 4.
Is plasma, therefore, useful only for certain types of waste such as
high BTU low water content wastes?
Question 5.
Can fluxing agents be used to enhance melt properties and residue
quality? How would the use of such agents affect the economics and
practicality of the process?
CONCLUSIONS
Only limited data are available on the use of plasma to decontaminate
hazardous waste, although the idea of doing so has been considered
for a number of years. Although some of the results achieved thus far
are promising, a number of questions remain about the usefulness of
plasma in this application. Until more data become available, it will
be impossible to answer these questions and to determine how plasma
might best be used in this application. Information from future tests
on the Retech Centrifugal Reactor and the Westinghouse Pyroplasma
Unit will provide needed information to further assess the use of plas-
mas in hazardous waste treatment.
REFERENCES
1 Joseph, M.F., Barton, T.G. and Vomdran, SC. "Incineration of PCBs by
Plasma Arc," Proceedings oflhe 33rd Ontario Industrial Wale Conference.
June, 1986. Toronto, Ont. 201-206.
2 Johnson. A.J.. Arnold, P.M.. Deitcsfeld. C.A. and Morales, L.M. "Wane
Generation Reduction-Nitrates FY 1984 Status Repon US. Government NTIS
Report PBDE8SI02067. Apr.. 1985.
3. Dolcnko, A J., Research on Gassificalion ofHbod in a Plasma Pyrofyrit Unit.
Canadian Forestry Service NTIS PB 84901566. Jan.. 1984.
4 Lee. CC and Huffman, G.L., "Update of Innovative Thermal Destruction
Technologies" ERV500/225 PB89U8541/AS US. EBV Cincinnati. OH, I98&
5. Drost. H., Klotz, H , Schullz, G. and Spangenberg. H , "The influence of
Hydrogen on the Kinetics of Plasmapyrolytk Methane Conversion" Plasma
Chemistry and Plasma Processing. I Mar. 1985 55-65.
6. Lee. CC and Huffman. G.L., "Innovative Thermal Destruction Technolo-
gies." Handbook on Hazardous Htistr Incineration. CRC Press, Boca Raton,
PL. July. 1988.
7 Reed. WS . Sales Manager. Westinghouse Environmental Systems. Madi-
son, PA. Personal Communicator. Aug I, 1989.
a Eachenbach, R.C. Hill. R.A. and Sears, J.W. "Process Description and Initial
Test Result* With (he Plasma Centrifugal Reactor" Forum on Innovative
Hazardous Waste Treatment Technology. Atlanta, GA. June. 1989.
424 U.S. EPA SITES
-------
Evaluating the Cost-Effectiveness of
SITE Technologies
Gordon M. Evans
U.S. EPA
Cincinnati, Ohio
ABSTRACT
The goal of the U.S. EPA's Superfund Innovative Technology
Evaluation (SITE) Program is to develop reliable performance
and cost data for unique and commercially available hazardous
waste treatment technologies. A major challenge which faced the
SITE Program was how best to insure that the cost evaluation
process produced cost projections which would be useful to
Superfund decision-makers. In this evaluation process, several
impediments to the collection and analysis of cost data were iden-
tified. This paper discusses the four most important problems en-
countered and then offers a set of five cost guidelines which
address those problems.
INTRODUCTION
Among the programs created by Congress through the passage
of SARA was the Superfund Innovative Technology Evaluation
(SITE) Program. The goal of the SITE Program is to help the
Superfund decision-making process through the formation of re-
liable performance and cost data for unique and commercially
available hazardous waste destruction and treatment technol-
ogies.
Interestingly, of all the language contained within SARA, the
following section is the only one which specifically requires the
Agency to collect and report cost data for those technologies be-
ing demonstrated. Section 31 I.e. states that the SITE Program
will prepare an annual report for Congress, in which shall be
"...an evaluation of each demonstration project..., findings
with respect to the efficacy of such demonstrated technologies in
achieving permanent and significant reduction in risk from haz-
ardous waste, the cost of such demonstration projects, the poten-
tial applicability of, AND PROJECTED COST FOR, such tech-
nologies..." (emphasis added). While other language within the
legislation indirectly speaks to the cost issue, either by specifying
the need to select technologies for the Program that "are likely to
cost-effective and reliable" (Section 311.b.7.B.), or by stating
that the demonstrations will determine "whether or not the tech-
nologies used are effective and feasible" (Section 311.b.5.A.v),
no specific guidance is offered on the scope or content of the
economic analysis.
Within the Agency, implementation of the SITE Program is
handled jointly by the Office of Research and Development and
the Office of Solid Waste and Emergency Response. As stated in
the SITE Program's first "Report to Congress"1, it's goals are
fourfold:
• To identify and, where possible, remove impediments to the
development and commercial use of alternative technologies
• To conduct a demonstration program of more promising inno-
vative technologies to establish reliable performance and cost
information for site characterization and cleanup decision-
making
• To develop procedures and policies that encourage selection of
available alternative treatment remedies at Superfund sites
• To structure a development program that nurtures emerging
technologies
In order to participate in the SITE Program, interested tech-
nology developers first are asked to submit a detailed proposal to
the Agency. This proposal should highlight the innovative aspect
of their process, offer any preliminary test results and provide
proof that they can commercialize the process. From those pro-
posals submitted, the Agency selects roughly 10 technologies per
year. Those accepted are invited to enter into a cooperative agree-
ment with the U.S. EPA. Under the terms of that cooperative
agreement, the government's primary financial commitment is to
cover those costs involved with the collection and analysis of
data. The developer, on the other hand, is responsible for all costs
associated with the actual operation of the equipment. After the
demonstration results have been analyzed, the engineering and
cost evaluations for each SITE technology are then presented
within one of a series of outputs; a report entitled "SITE Tech-
nology Application Analysis." This document is designed to pro-
vide the reader with an in-depth overview of the process includ-
ing a report on the demonstration, an analysis of the test results,
cost projections, case studies and comments by the developer.
Draft versions of the cost analysis are prepared by the U.S. EPA
Project Manager and his support contractor in consultation with
the SITE Program's staff economist. Before publication, this
draft version undergoes an extensive review. In the case of the
cost projections, this review is performed in order to examine the
soundness of the analytical approach and to insure conformity to
the generalized cost protocol described later in this paper.
Thus, one effect of the SITE Program is to create a limited
partnership between the U.S. EPA and the technology devel-
oper; between the public and private sectors. Nonetheless, the
Agency has a significant responsibility to provide the public with
an impartial analysis of each technology. The ultimate challenge
to the SITE Program is to balance the need to remain neutral
while encouraging the adoption of promising new technologies.
As the details of the SITE Program were being formulated, the
special nature of this public/private sector interaction suggested
the need to design a method for projecting the costs of new haz-
ardous waste treatment technologies. This paper examines four
U.S. EPA SITES 425
-------
issues which limit the ability of the SITE Program to establish
protocols for the conduct of each technology's cost analysis.
After this paper has explored these issues, it will describe the cost
methodology currently in use by the SITE Program.
PROBLEMS
The first problem that limits the ability of the Agency to pro-
vide Superfund decision-makers with accurate cost projections
(or new technologies confronts everyone involved with Superfund
cleanups. Every cleanup operation represents a mix of factors
unique to that site; these variables include the waste matrix, the
amount of waste to be treated, the physical characteristics of
the site and the cleanup treatment goals. The way that these fac-
tors combine will vary from site to site. Thus, any data (engi-
neering or cost) collected during a given demonstration speak
directly to the conditions found on a particular site at a given
point in time. The ability to extrapolate those data to other haz-
ardous waste sites is constrained in large part by the degree to
which similarities exist between the demonstration site and other
sites. The more similar the conditions, the more confidently we
can predict the outcome.
Layered on top of these individual site variations are those vari-
ables directly related to the operation of the SITE Program.
These variables include regulatory restrictions, programmatic
and budget constraints, the level of the developer's experience
(both working with the technology and operating within the haz-
ardous waste field) and the expertise of the SITE Project Man-
ager and his support contractor. Once again, the data collected
during a demonstration are a reflection of the interaction of all
variables. The lesson to be learned here is that the costs observed
during a demonstration represent nothing more than one of many
scenarios possible under different operating conditions. The
problem confronting the cost analyst is how best to capture and
portray the most likely costs; to generate a base-case cost pro-
jection which will have broad appeal among Superfund decision-
makers.
The need to collect and analyze engineering and performance
data under rigorous QA/QC conditions presents the second limi-
tation to the Agency's ability to project future technology costs.
When the research objectives of each demonstration are coupled
with a finite demonstration budget, the ability of the Agency to
collect economic data is reduced. There are many reasons why
this is so.
Each developer needs to finance all costs associated with the
operation of his equipment during a demonstration. Contracting
his services to a third party is a good way to do that. It is likely
that the actual demonstration will be conducted during an on-
going site remediation. The developer's primary responsibility is
to meet the dictates of his contract; to treat the waste. From his
perspective, the SITE demonstration activities are at times an im-
pediment to that remediation. The Agency's sampling and analy-
tical plan will specify the number and nature of the samples to be
taken. Typically, the collection of samples will occur during a
tery small subset of the equipment's total operating time. The
benefits derived from this "snapshot" view of the process are
that it limits both the sampling costs and the interruptions to
the process. The downside is that this sampling period may be the
only opportunity that the Agency has to closely observe the oper-
ator and the equipment, thus virtually eliminating the Agency's
ability to gather long-term economic data.
It may be impossible to directly observe other operations and
record the cost of those activities. These items range from peri-
odic maintenance to the average on-line utilization rate. The cost
implications of these activities must be obtained from secondary
sources or be estimated. Even in those cases where the developer
is prepared to provide regular access to the equipment, the ability
to collect variable cost data can be hampered by normal sampling
requirements. Continuous process operations may need to be rou-
tinely interrupted as sampling occurs. When this happeni,
observed costs must be adjusted to account for such activity.
Finally, one-time factors that are inherent in the operation of any
new equipment (particularly in equipment incorporating innova-
tive designs features) further limit the ability of the Agency to
collect real-time economic data. These include the need to sep-
arate startup and shakedown problems from normal operations,
including unplanned field modifications, materials handling ad-
justments and scale-up problems.
Commercialization
One of the stated goals of the SITE Program is to encourage
the commercialization of innovative technologies. As such, the
SITE Program has had to cope with both the impact that market
forces have on each participating firm as well as the impact the
SITE Program itself has on the market. This concern is the gen-
esis for the final two cost problems which confronted the Pro-
gram.
Each vendor accepted into the SITE Program must have
demonstrated the potential to commercialize his technology. To
put this requirement in economic terms each vendor has repre-
sented himself as profit maximizer willing (and prepared) to oper-
ate within a competitive marketplace. As a profit maximizer, each
developer has formulated a unique strategic view of that market;
one which he believes will ultimately sustain long-term profitabil-
ity. The SITE Program needs to be sensitive to each developer1!
strategic viewpoint, which is colored in large part by the under-
lying condition of his balance sheet.
Company Financial Strength
The SITE Program has seen great diversity in the financial con-
dition of vendors accepted into the program. Many firms enjoy
advantages made possible by some form of long-term financial
commitment. With this, they have the ability to withstand cash
flow problems inherent in the conduct of research-oriented engi-
neering work. These firms, and their backers, recognize the ben-
efit in deferring short-term profits in order to gain an opportunity
to establish their presence in the market and position themselves
for long-term benefits. At the other extreme are those developers
who enter the program with pilot-scale equipment, a promising
idea, but limited financial resources. Participation in the SITE
Program provides a wonderful opportunity for a firm to estab-
lish a presence in the marketplace. However, the precarious finan-
cial position of some firms often limits the time they have avail-
able to enter the market. Engineering, operational or regulatory
delays of any sort may severely hamper a company's ability to
successfully commercialize its technology, much less remain a
viable corporate entity. The firm's corporate strategy likely dif-
fers greatly from that of its better-financed counterparts.
Since one stated objective of the Program is to "identify and
remove impediments to the development and commercial use of
alternative technologies," it is important for the Agency to be
sensitive to these corporations' financial positions. Each firm wiD
use a different method for apportioning its research and develop-
ment expenditures. Each firm will have its own marketing strat-
egy which places emphasis on exploiting some niche within that
market. Each firm's growth potential will be constrained by its
ability to raise capital and attract (and retain) a competent tech-
nical staff. Each firm will have established target levels of profit-
ability, with a pricing strategy to reflect this. These points com-
bine to form a framework which guides corporate decision-
making.
Finally, each firm understands the important distinction be-
tween "cost" and "price," a point which is easily lost on those in
the public sector unconcerned with the profit motive. In simple
terms, cost reflects expenditures by the firm and is inherently a
function of accounting. Price, on the other hand, is the end pro-
duct of negotiations between the firm and those wishing to obtain
426 U.S. EPA SITES
-------
its services. It is much simpler for the outside observer to project
costs than it is to project price. In the end, the price which the
developer charges becomes a direct function of both the firm's
strategic view of the market and the interaction of supply and
demand forces.
Confidential Data
Armed with an understanding of the central role that a firm's
business strategy plays, and given the highly competitive nature
of the hazardous waste treatment market, it is easy to see why
any firm would have a strong incentive to withhold its cost data
from the public record. Such information, if placed in the hands
of a competitor, could severely undermine the firm's chance for
long-term success in the market. In those cases where cost data
are offered by a firm, one must ascertain the motivation of that
firm in releasing such data. Are the data accurate? Do the data
truly represent the firm's actual costs or do they represent costs
which the firm would like the market to believe are true? The
Agency must accept the proprietary nature of each firm's cost
data, despite the problems it creates in trying to project future
costs. Even SARA acknowledges the sensitive nature of a firm's
cost information when it states in Section 311.b.8. that all data
collected during a demonstration shall be made available to the
public except for "trade secrets or other proprietary informa-
tion." This secrecy provision leads to the unhappy conclusion
that SITE technology cost projections may end up being con-
ducted without input from the developer
The final cost issue confronting the SITE Program is the Pro-
gram's own impact on the hazardous waste treatment market. For
better or worse, the U.S. EPA's evaluation of demonstrated tech-
nologies will carry significant weight among decision-makers in
both the public and private sectors. Merely participating in the
Program confers a special status to those who are in it. The judg-
ments offered by the Agency on a technology's effectiveness are
likely to be viewed by the public as a U.S. EPA "Seal-of-
Approval," regardless of the Agency's intention to remain im-
partial. Opinions offered by the U.S. EPA regarding the engi-
neering effectiveness of a given technology can be supported
through reference to vast amounts of QA/QC data generated dur-
ing the demonstration. Not so with the cost projections. By con-
trast, those projections are supported for the most part by the
quality of the underlying economic analysis.
Cost Projections
As the SITE Program's ability to influence the market grows,
the real danger for all parties concerned is to discover that the
Agency's cost projections have been overly optimistic or pessi-
mistic. If it turns out that the cost projections end up being signif-
icantly lower than true costs, Superfund decision-makers will be
misled into concluding that the technology is exceptionally cost-
efficient when compared to other alternatives. In turn, other tech-
nologies under consideration may be rejected out of hand for
appearing to be too costly. Eventually, the developer may be
faced with the difficult problem of trying to negotiate a fair price
with a buyer who harbors false price expectations. At the other
extreme, if the cost projections end up being much higher than
true costs, potential users conducting a preliminary screening may
exclude the technology as being too expensive. Rather than help-
ing to promote new technology, the Agency will have inadvertent-
ly limited the developer's market. In either case, making a signifi-
cant error in its cost projections is the best way to endanger the
SITE Program's long-term credibility with Superfund decision-
makers.
In review, there appear to be four problems which significantly
limit the ability of the SITE Program to generate useful cost pro-
jections for the technologies it demonstrates:
• Each field demonstration represents a mix of unique factors
• The research and development aspects of each demonstration
will impact observed costs
• Each developer is a profit maximizer operating within a com-
petitive marketplace
• The SITE Program creates unique interactions between pub-
lic and private sector forces
SOLUTIONS
After reviewing the four issues presented above, it became
clear that a single, rigid cost protocol would not serve the goals of
the SITE Program. With the potential for several dozen demon-
strations to be conducted over the life of the SITE Program, the
sheer number of independent variables involved with each
demonstration made the usefulness of such an effort suspect.
What was possible, however, was to establish broad rules to guide
Project Managers and their support contractors as they worked
through the cost projections. The idea was to create a high degree
of uniformity among all the SITE cost analyses while allowing
the conditions of the demonstration to dictate the basic approach
used in each cost projection. Most importantly, insuring that all
cost projections follow the same basic rules should enhance the
ability of Superfund decision-makers to make relative cost com-
parisons between technologies. The remainder of this section will
highlight four major cost guidelines which, when taken together,
address the concerns set forth in the previous section.
Cost Categories
The first and most critical step was to establish a set of cost
categories which would serve as a common framework for a base-
case cost analysis. These 12 categories are listed in Table 1. While
the descriptions of each category have been omitted from this
paper, they are intended to encompass the range of activities
which could occur during a demonstration or cleanup. It is recog-
nized that these categories are but one combination of activities
and, in the long run, other ways of classifying these activities may
be more appropriate. Under the most ideal conditions, each
demonstration would provide enough information to make cost
projections for each of the 12 categories. However, each demon-
stration is a mix of unique factors, reducing the likelihood that
any final cost estimate would be based on the sum of all 12 cate-
gories. When assigning costs to these categories, it is incumbent
upon the analyst to leave empty those categories for which data
are unavailable. In other words, if data are unavailable for three
of the 12 cost categories, then the report should clearly state that
fact so that the reader fully appreciates the underlying basis for
the cost projection.
Aside from providing a common framework for all SITE cost
projections, use of these categories should help reduce the temp-
tation many have to compare the cost of technologies when the
bases for each of the cost projections are not equivalent. For ex-
ample, if technology A's cost projection is based on the sum of
eight categories while technology B's cost projection is based on
costs incurred in all 12 categories, comparing their projected costs
without first compensating for the difference in then- bases would
lead one to reach a false conclusion about the relative cost-effec-
tiveness of one technology over the other.
Order of Magnitude Estimates
The second rule simply requires all cost projections to be pre-
sented as "Order-of-Magnitude" estimates, a precision level
established by the American Association of Cost Engineers
(AACE). The expected accuracy of "Order-of-Magnitude" esti-
mates is within +50% and -30%. The AACE defines this level
of precision as being those estimates generated without the bene-
fit of detailed engineering data.2 AACE suggests that this type of
estimate is appropriate for feasibility studies or to aid in the selec-
tion of alternative processes. This analysis is the intended use of
U.S. EPA SITES 427
-------
Tiblcl
Coil Categoric*
SITE Application Analyib Report.
1. Site Preparation
2. Permitting & Regulatory Requirements
3. Capital Equipment
4. Start-Up
5. Labor
6. Consumables & Supplies
7. Utilities
8. Effluent Treatment & Disposal
9. Residuals/Waste Shipping i Handling
10. Analytical Services
11. Maintenance & Modifications
12. Demobilization
the SITE cost projections. While decision-makers may desire
greater precision in the cost projections, doing so would require
the preparation of much more detailed design work than is cur-
rently possible.
Base Cost Projection
The third rule provides that each cost analysis will generate a
base-case cost projection which presents the reader with a full dis-
closure of all assumptions and calculations. The key idea here is
full disclosure. There should be no question as to how final cost
projections are derived. Providing full disclosure not only means
clearly stating the assumptions, but it also means providing the
source of that information. Were cost figures based on direct
observation of the process or were they derived from secondary
sources? The Agency has an obligation to clearly indicate where
data points were obtained, allowing the reader to make an inde-
pendent judgment on their worth. Regardless of whether the data
are taken from the developer, a standard reference source, a cost
curve or are arrived at through an educated guess, the analysis
should be forthright and state the source. Formulas must also be
presented, and where calculations are complex, each step should
be outlined. Using the 12 cost categories will help to insure that all
relevant assumptions are covered.
Adherence to this rule will provide several benefits. At the min-
imum, it places the burden of proof upon the reader to examine
the assumptions used to generate the cost projections and insure
they are appropriate, given the details of his cleanup problem.
Unfortunately, experience suggests that the tendency is for many
decision-makers to seek out and focus upon a single unit-cost
estimate. The result is that these unit-cost projections often are
taken out of the context of their assumptions and, as noted,
understanding the nature of the assumptions is critical to the use-
fulness of the projection. While this rule cannot hope to stop the
inappropriate use of cost data, it will insure that if and when
questions arise surrounding a cost projection, the answers will be
readily available.
Full Disclosure
The full disclosure of assumptions and calculations also will
produce an end-product which the reader can replicate on his
own. Suppose the reader finds the basis for a cost projection to be
inappropriate. It is very likely to be the case. As was stated earlier,
each Superfund site represents a mix of factors. Any attempt to
set forth a "standard" Superfund site upon which to base each
cost projection is an exercise in futility. Thus, the base-case cost
projection offered by the U.S. EPA can only hope to represent
the most typical outcome of an infinite number. By affording the
reader the means to recreate the cost analysis it becomes a
straightforward matter for him to recalculate the projection, sub-
stituting any assumptions deemed inappropriate with others
more in line with his own situation. In this way, each reader ii
offered the limited ability to tailor the cost projections to fit the
needs of his problem.
Engineering Parameter Variations
Having generated a base-case cost estimate, the next rule spells
out the need for the analyst to examine the effect on cost from
changes in key engineering parameters. In other words, what dev-
iations from the base-case assumptions will lead to significant
increases or decreases in final costs? At the minimum, this analy-
sis should offer the reader a short narrative which details the
effect these alternative assumptions can have on the base-cast.
When the opportunity presents itself, the analyst is encouraged to
conduct a numeric sensitivity analysis which will demonstrate the
degree to which these changes can impact costs. The identifica-
tion of these key parameters is a task best left to the judgment
of the Agency's Project Manager. His goal is to apply the insights
gained from the demonstration to the question of cost so that the
reader will have enough information to ask intelligent questions
concerning cost.
Market Forces
The final guideline attempts to address the problems that are
created by the Program's interaction with market forces. As
earlier portions of this paper have pointed out, the Agency needs
to be sensitive to the fact that corporations will employ different
strategies as they pursue the goal of commercializing their tech-
nologies and maximizing profit. The U.S. EPA cannot (and
should not) factor these strategies into its cost projections. How
does the Agency assist the developer's attempts at commercializa-
tion when it needs to remain at arms' length from that developer?
The solution is to provide each developer with a forum to present
his own cost analysis. This process is accomplished by setting
aside a chapter just for the vendor's comments within the "SITE
Technology Application Analysis."
In practice, the developer is given a chance to review and com-
ment on the draft versions of the "Application Analysis." White
the developer is free to offer criticisms regarding any of the Re-
port's findings, the Agency is under no obligation to change the
results of its evaluation. Instead, the developer is asked to prepare
a chapter for the "Application Analysis" in which he is able to
state his case, free from the U.S. EPA's editorial control. This
means that the vendor has an opportunity to present a cost analy-
sis which should implicitly account for all the market forces he
perceives to be significant. In other words, the strategic view-
point from which the vendor approaches the market will form the
basis for his cost projections. In the end, the "Application Analy-
sis" presents the reader with two different perspectives on the
technology's cost-effectiveness. By comparing the developer's
cost projections with the U.S. EPA's, the reader should be in a
better position to determine the true range of future costs.
Cost Analysis Summary
In review, the five rules which govern the conduct of each
SITE cost analysis are as follows:
• Place each base-case cost analysis within a common framework
of 12 cost categories
• Present each base-case cost projection as "Order-of-Magni-
428 U.S. EPA SITI-.S
-------
tude" estimates (+ 50% and - 30%)
• Provide full disclosure of all assumptions and calculations used
in the base-case analysis
• Identify key operating parameters which are likely to have sig-
nificant cost implications beyond the base-case
• Offer developers the opportunity to present their own cost
analysis
CONCLUSIONS
No methodology will insure that projected costs can be cal-
culated with the same degree of precision as engineering or chem-
ical data can be. When one combines the imprecise nature of cost-
estimating with the heterogenous condition of Superfund sites
and the unanticipated problems one is likely to encounter work-
ing with new technologies, one must be prepared to accept the
fact that the cost projections will be imperfect. However, cost
data which will give decision-makers meaningful insights into the
relative cost-effectiveness of new and innovative Superfund tech-
nologies can be prepared.
REFERENCES
1. The Superfund Innovative Technology Evaluation Program: A
Report to Congress, EPA 540/5-88/001, U.S. EPA, Washington,
DC, February, 1988.
2. Humphreys, K.K., Project and Cost Engineers' Handbook, 2nd Ed.,
pp. 51-53, Marcel Dekker, New York, NY, 1984.
U.S. EPA SITES 429
-------
Use of a Geographic Information System in Selecting
Residential Properties for Remediation
at the Bunker Hill NPL Site
Ian H. von Lindern, P.E., Ph.D.
Kara Steward
Margrit von Braun, P.E.
TerraGraphics Environmental Engineering, Inc.
Moscow, Idaho
Sally Martyn
U.S. EPA, Region X
Seattle, Washington
ABSTRACT
A Geographic Information System (CIS) was used to rank and select
residential properties for remediation in a removal project at the Bunker
Hill NPL Site. This site encompasses a 21-mi22 area surrounding a
defunct primary lead-zinc smelter in northern Idaho. Approximately
5.000 people live in the area. More than 1,000 home yards are con-
taminated with soil lead levels exceeding 500 ppm. More than 75%
of these home yard soils exceed 1,500 ppm.
The CIS served an initial inventory function. For each of 3,000
individual properties, basic data were encoded to a relational data base.
The primary information included: (1) legal and ownership data obtained
from county tax records; (2) childhood health and census data and
(3) sampling data.
COEUR
D'ALENE
Figure 1
Bunker Hill NPL Study Area
Each property attribute was location-coded and a base map of all
properties was created from the tax records. Specific uses of the CIS
in the removal ranking process included: notifying owner/resident and
verifying data using mail-merge options; identifying populations at risk;
ranking risk according to sampling results; characterizing neighbor-
hoods according to aggregate risk and population characteristics; identi-
fy ing candidate properties for removal; and preparing exhibits for public
meetings and discussions. The CIS proved to be an efficient tool in
performing a variety of tasks related to selecting and ranking proper-
ties for remediation.
INTRODUCTION
Site Background
The Bunker Hill NPL Site encompasses a 21-rn? area surrounding
a primary lead/zinc smelting complex in Northern Idaho. Smeller opera-
tions shut down in 1981. The industrial complex has since been sal-
vaged through unregulated activities and is rapidly deteriorating. Large
waste piles, dilapidated buildings and defunct industrial process equip-
ment litter the 365-ac smelter complex site. The study area is located
in a deep narrow, sub-alpine river valley in the Northern Rocky Moun-
tains (Fig 1). Years of sulfur dioxide abuse have left many of the hill-
sides denuded and subject to severe erosion. On the valley floor, massive
impoundments of mine wastes and major deposits of unconfined tailings
dominate the flood plain and major hydrologk drainage system. Poor
incorporated cities, home to a population of more than 5,000, are found
within the site boundaries.u
This area was the scene of epidemic lead poisoning in children during
the 1970s. More than 75 % of the area's children exhibited excess blood
lead absorption in 1974" These poisonings were largely associated
with environmental lead contamination resulting from uncontrolled
smelter emissions. More than 1,000 children experienced lead levels
in excess of current Centers for Disease Control (CDC) health criteria
during the 1970s1. In 1983, 2 yr after smelter closure, communitywide
testing revealed that 25% of the preschool children in the most con-
laminated residential areas continued to have blood lead levels above
the CDC criteria*.
Subsequent studies linked this excess absorption to contaminated soil
and dust exposures in the community. More than 1,000 homes lave
yard soil lead levels exceeding 500 ppm. Seventy-five% of those yards
exceed 1,500 ppm, with 47% greater than 2,500 ppm. Housedusts as
high as 52,700 ppm lead have been measured and average 3,400 ppm
in the most contaminated residential area.
As a result of these studies, the Bunker Hill Site was placed on the
NPL in 1983. In 1985 a large multi-phase RI/FS commenced and several
expedited response activities have been undertaken.
430 SITE REMEDIAf ION
-------
RI/FS and Response Actions
The structure of the RI/FS and Response Action activities reflect the
complexity of this site. The overall project is managed by Region X,
U.S. EPA, with major investigation responsibilities delegated to the State
of Idaho in the populated areas of the site and to the PRPs in the non-
populated areas. The U.S. EPA exercises oversight responsibilities in
both cases. The major portions of the PRP effort include the smelter
complex, denuded hillsides, waste piles and tailings impoundments,
groundwater problems and the river and floodplain system. The State's
responsibilities encompass the health-related and private property-
ownership issues. These include two major efforts: (1) an RI/FS to in-
vestigate contaminated soils, homes and features in the populated areas
and (2) a health intervention program to reduce lead absorption through
a combination of testing, followup and education response actions un-
til source control measures can be implemented.
Several expedited response actions also have been implemented to
reduce exposures. In 1985, an aggressive public health intervention pro-
gram was undertaken to reduce excess absorption among young children
in the community. The program included door-to-door testing of chil-
dren for elevated erythrocyte-protoporphyrin (EP) levels, follow-up
testing for blood-lead home visits and parental counseling for the families
of children who had elevated blood leads. Additionally, public educa-
tion programs were instituted with schools, community service organi-
zations and health professionals. These programs stress the preventative
hygiene, behavioral and home environment modifications that can
effectively reduce lead absorption in young children.
In 1986, a removal action was instituted to reduce soil and fugitive
dust exposures on publicly owned and accessed areas of the site7. Soil
removal and replacement, seeding, sodding, cover and dust control
efforts were instituted in parks, playgrounds, schoolyards and street
berms. These efforts substantially reduced exposures in common areas
accessible to community children. The combination of testing, educa-
tion program and remedial measures was quite successfully reduced
the prevalence of excess absorption. The percentage of children exhi-
biting excess absorption declined from 25% in 1983 to 2% by 1986.
However, area participation rates for the important testing portion of
the program had decreased by one-third in the 3 yr from 95 % coverage
in 1985 to 65% in 19878. There was significant concern that the
effectiveness of the screening program was compromised by the low
participation rate. Informal community surveys revealed that the suc-
cess of the program (i.e., individuals believing the problems had been
solved) and public frustration regarding the pace of cleanup were
resulting in a growing complacency in the community. Of greater con-
cern, was the fact that many of those who were dropping out of the
program were from socioeconomic groups at higher risk of lead
poisoning.
As a result, a decision was made to expedite cleanup of private proper-
ties where there was a high risk of lead poisoning to young children.
A removal project was scheduled to begin in the summer of 1989. In
addition to designing the remedial action, there were great logistic and
informational challenges in initiating the removal. Those challenges
included determining cleanup criteria; informing the public; contacting
owners; determining which homes to remediate; securing access agree-
ments with owners and residents; and implementing the project in the
most health protective, efficient and cost-effective manner.
The Geographic Information System (GIS) data base management
strategy developed for this project assisted the various agencies involved
in accomplishing these tasks. This paper discusses the development of
the populated areas data base and its use in helping to rank and select
properties for expedited response actions during a 1989 soil removal
project.
DATA BASE MANAGEMENT/THE GIS
In implementing such an involved project structure on this complex
site, the U.S. EPA recognized the need for effective information manage-
ment. At the beginning of the project, a determination was made to
employ a GIS-based strategy. That system was briefly described in an
earlier HMCRI conference9. This system has been used extensively in
the populated areas to integrate health, population and property-related
data bases for risk assessment, inventory and notification purposes.
Methods/Data Base Development
The GIS Data Base contains two principal components. The Base
Map is the vector data base that provides spatial reference for each piece
of information. The Attribute Files contain the sealer data or descrip-
tive information about a location. Attribute files are maintained in a
relational data base with a location-specific reference to the Base Map.
Base Map
The Overall Base Map encompasses the site as a 3 X 7 mi rectangle
centered on the smelter complex (Fig 2). The populated portions of
the study area are maintained as a series of sub-unit base maps repre-
senting each town. Each sub-unit serves as a'n inlay to the overall site
base map. These maps were digitized from the County property tax
inventories. Shoshone County properties are tracked by parcels assigned
Figure 2
Bunker Hill GIS Base Map
SITE REMEDIATION 43!
-------
unique identification numbers per u sub-division, block and lot hierar-
chy. All local government information pertinent to that parcel is in-
dexed to the Property ID Number.
Attribute Files
Attribute files are maintained in ASCII formal indexed to the Property
ID Number. Three types of attribute files were developed for this effort.
Those are County Tax Records. Health Census Data and Sampling
Regime Results.
County Tax Records" pawide ownership details and legal descriptions
of the property. The County maintains these files by Property ID
Number. Health Census Data" have been collected as part of the
Health Intervention Program described above. All homes in the health
surveys are coded by a Block*Lot* Unit hierarchy analogous to federal
census techniques. These files were cross-referenced to the Property
ID Number and converted to a property-specific data file
Several Sampling Regimes"•'•' have been undertaken at this site The
most extensive was the 1986 Residential Soil Survey effort in which
surface soil sampling was offered to each property owner m four of
the site's five residential areas. This survey was designed using the GIS
data base.
A base map of each property was generated showing the Property
ID Number. Fig 3 shows the map for the City of Smeltervillc. These
maps were sub-divided into block maps (sec inset Fig 3) used by field
crews when interviewing residents and securing permission to sample
yards. These data were merged with the County Tax Records and used
to provide sampling assignments and track field crew progress. Infor-
mation secured through questionnaires administered and protocols com-
pleted during the survey were then indexed with sample results to the
Propert) ID Number. This resulted in attribute files containing the
information shown in Table 1.
Other, less extensive, sampling regimes have been accomplished at
the site. These are indexed to the Property ID Number and can be ac-
cessed by the GIS and include: deep-core soil profiles; houscdusts;
historical health, environmental and garden vegetable surveys conducted
in 1974. 1975. 1977 and 1983.
Table 1
Data Baae Attribute KUe Summary
on*
DMtrlotlon
Couity Tex Record!
(1988 updete)
Keetth Ceneue Dete
(1974, 1975.
1983. 1986.
1988)
Seepte lecorde
(1974, 1975,
19S3, 1984,
1989)
Property 10 mafcer, okneriftlp nee» end
•ddrMf, lien holder, property eddreu,
toning, lenduee end leflel description
Property 10 Muter, re*ldent'i nee», eddretf
end telephone rueber, nueber end eoee of
children, reeulti of Mepllng (blood leed
If, Zn tP), houiehotd end yerd condition*,
perent educetlon end Income, Maker*, dilld
behevlor (outdoor pley tlM, dletery vlteelnt,
oret behevlor) end echool ettended
County tn
Pertundl,
«Mtlk
Olitrtet
Property 10 fcxfcer, re«ldent ne>» end eddreu,
retulu of toll, litter end houuduit evteU *e*(tk
lecpllng (Pb. U, Zn, U, (e, Nn, Cu, «b), District
teeplt crw, rttident raponee, uipl> dete,
berth oett. leb uepte mater. Meple type,
end leb trenefer
CHS Capabilities
The combination of base maps and attribute files provides a com-
plete and comprehensive summary of the information available for each
property on the site GIS utilizes these data to perform four key (unc-
tions common to data ba.se management: inventory, tracking, analysis
and display. The principal advantages inherent in GIS are that each piece
of information is systematically indexed in space and time. That re-
quires that all data be reduced to a common format and meet minimum
quality control criteria That function, alone, is valuable for data
inventory purposes in a project involving five major agencies and a
dozen contractors compiling data and performing analyses.
The structure also provides for ease of updating and tracking. New
data can be added by substituting or addending attribute files. For ex-
ample, the County Tax Records are updated each year as the new lax
roles are prepared. Residents are tracked annually through the Health
Surveys and the data base is updated by substituting new information.
r ~^**^'-,'iY-- "' A , ,>»'V UL.'-.'i-LAgJ/
r-<-fT^fvvr, VM »'' ' -Vfiv-TrsEDr-*,—^— ^•-•-,-.mr^-^-vi.>i., ^^IM^-^---,
^'.•^'Ti- ••V^-'^"v-'^Ą^n^^^xVr^:!!!! ! !1 i'7''^^
,, ^x--;--^f^!^K^^^^^ :=ry'
•;:^;^^j^mo^^^c^^ ^^
!: :. lfV^<^\ '•\v^\^'^^^i^!^:r^ ^e^°'
Figure '
C'ny of Smcllsvillc Resident Property Map
432 SITE RKMF.DIATION
-------
New sampling regimes are appended as separately accessible files.
The greatest advantages associated with GIS, however, are analyti-
cal capabilities. The GIS offers a "tool box" of map analyses which
may be grouped into the four categories described below:
• Reclassification junctions create new maps by assigning new values
to existing maps. Reclassification values can be based on the position,
size, shape, or initial value of the original map's categories. For
example, properties with soil lead levels less than 500 ppm could
be reclassified into a map showing areas of "acceptable risk."
• Overlay functions are used to create new maps based on point-by-
point or area relationships between independent maps of the same
area. For example, a map delineating private properties with;
(1) excess lead levels and (2) young children can be generated from
individual maps representing each of the categories.
• Distance and connectivity operations include measurements of sim-
ple distance, perimeters, areas and volumes. These procedures are
also expanded to include the concepts of proximity and connectivity
(e.g., identifying equidistant zones of proximity to a smelter).
• Neighborhood characterization involves creating maps as a function
of an independent value within a specific area or cartographic neigh-
borhood of a location. The first step is defining the cartographic neigh-
borhood. For example, the neighborhood might be all "downwind"
locations within .5 mi of a smelter. Numerous maps then describe
categorical values within the defined neighborhood such as mean
childhood blood lead levels or number of children to exceed CDC
health criteria.
By organizing these four general analytic processes sequentially,
higher techniques of map analysis, called cartographic modeling, can
be developed to perform more complex analyses. Cartographic modeling
provides flexibility, "what if analyses, rapid simulation of various
strategies, optimization and effective communication through flow
charting, while documenting the factors and assumptions used in the
decisionmaking process.
GIS USE IN THE 1989 RESIDENTIAL REMOVAL
GIS analyses served five basic functions in the 1989 Home Yard Soil
Removal Project. Each of those functions is briefly discussed below.
Data Verification and Owner/Resident Notification
Health agencies had a responsibility to notify homeowners and resi-
dents of the data collected, provide interpretation of sampling results
and communicate the risk involved. Many of the data are confidential,
requiring individual summaries and notification letters. As more than
1,500 homes were involved, this was an onerous task. The relational
data base aspects of the GIS were exploited to prepare individual
property summaries containing the ownership, childhood census, sample
results and risk indices obtained for that property. Fig 4 shows a sample
Property Summary Sheet.
Data were also extracted from the data base to provide name, ad-
dress and key variable files for input to conventional mail-merge soft-
ware. Individual data were substituted into mailing labels and master
letters that explained the form and risk indices, asked recipients to pro-
vide updated information and invited them to attend public informa-
tion forums.
Providing Master Lists and Maps for Project Managers
Each notification letter and summary form was tailored to the
individual recipient. Several confidentiality issues were involved. For
example, only owners and residents could obtain sample results, property
owners did not receive confidential information about their tenants,
individual data were not released publicly, etc. Owner-occupied resi-
dents received a different letter than renters. Owners, whose tenants
had refused to have samples collected, were similarly notified.
Project managers and health response personnel, on the other hand,
require complete summaries and maps of all results. Confidential
Property Summary Sheets containing all data for each property were
prepared for select project personnel. The reclassification functions of
GIS were used to develop a number of maps for confidential project
CONFIDENTIAL: MOT TO BE RELEASED1
BUNKER KILL SITE 5UPERFUND REPORT
PROPERTY SUMMARY
DATE COMPLETED 01NOV88
PROPERTY, ID # F-0100-007-013-0 14223
OWNERSHIP / LEGAL INFORMATION
OWNER: JOHN SMITH
ADDRESS: 100 MAIN STREET
CITY: DENVER, CO 55555
LOT 13 INST 322789
6LK. 7 S.l. 35 FU 0 0 0 0
SMELTERV11LE 1ST ADD RES 00203 HILL
SMELTERV1LLE MKS 07/16/86
LEGAL DESCRIPTION
CHILDHOOD CENSUS
RESIDENT
YARD SAMPJLIMG
RESIDENT
COMMENT
DATE OF SAMPLING
RESIDENT
LAST RESIDENCE CONTACTS
NR 8/86
BOB DOE
203 HILL
SMELTERVfUE ID 83868
# CHILDREN 2
# PRESCHOOL 1
BOB DOE
203 HILL
SMELTERVULE ID S3&&S
WANTS GARDEN TEST
RESPONSE
SAMPLE COLLECTED
YES
10/01/86
SAMPLING INFORMATION
10/01/86
BOB DOE
203 HILL
SHELTERV1LLE ID 83868
SAMPLE TYPE SOU
RESULTS LEAD
MB/KG 5839
ZINC
746
CADMIUM
19
SAMPLE TYPE SOIL DUPLICATE
RESULTS LEAD ZINC CADMIUM
MG/KG 3780 629 15
LAB * T 4321
ARSENIC ANTIMONY COPPER PH
57 W.6 96.5 S.3
LAB # T 4322
ARSENIC ANTIMONY COPPER PH
64 17,2 106.8 5.1
SAMPLE TYPE LITTER
RESULTS LEAD
MG/KG 4070
LAB # M 5432
ZINC
1390
CADMIUM
48
ARSENIC
62
SAMPLE TYPE LITTER DUPLICATE
RESULTS LEAD ZINC CADMIUM
MG/KG 3410 1277 42
LAB # H 5433
ARSENIC
70
PH
5.6
PH
4.6
SUB-CHRONIC HAZARD INDEX
FOR LEAD IN YARD SOIL
Fictional data for display only
5.8
Figure 4
Example Property Summary Sheet
use. These maps were similar to Fig 5 except sample concentration
and health survey attributes were substituted for lot ID #s. These maps
included such items as:
• Top-inch Soil Metal Levels - As, Pb, Zn, Cd, Sb, Hg, Cu
• Litter Metal Levels As, Pb, Zn, Cd, Sb, Hg, Cu
• Soil and Litter Lead Hazard Indices (Color Coded)
• Children's Blood Lead Levels
• Sample Status (i.e., whether sites had been sampled, owners con-
tacted, etc.)
These series of summary sheets and maps allowed project managers
to quickly access and evaluate individual data when dealing with resi-
dents and parents.
Public Display and Risk Communication
For public presentation it was necessary to use maps and displays
that contained no identifiable individual results. GIS neighborhood func-
tions were used to prepare non-confidential maps for risk communi-
cation purposes in public meetings. Neighborhoods were defined and
summary statistics were developed. Fig 5 shows the results for Smelter-
ville. Table 2 shows one of the inset summary tables from this map.
There are 88 homes in this sub-division; 70 of these homes were sam-
pled. A Sub-chronic Hazard Ranking (SHR), (soil lead level divided
by 1,000), was developed to describe to residents how their soils com-
pared to proposed national criteria. The average for this area was 3.7
SITE REMEDIATION 433
-------
(i.e., 3.700 ppm). Comparing data in Table 2 to the risk criteria shown
in Table 3 shows that 3% of the homes had acceptable soil concentra-
tions, 7% were recommended for individual consideration and 90%
were candidates for remediation.
Table 2
Summary Table for SmelUvlllc
Table3
Sub-chronic Hazard Rank Criteria for ttrd Soil Lead Levels (ppn)
F-0100
Smelterville 1st
Addition
f resid BB
i sampled 70
Average SHR 3.7
SHR < 0.5 3X
SHR 0.5 -1.5 n
SHR > 1.5 90X
Ranking Properties for Remediation
The 1989 removal suggested that resources were available to remediate
about 100 homes. Rectification, overlay and neighborhood functions
were used to help select which properties should be remediated.
SHR
< .5
.5-1.5
> 1.5
Soil
Lead (DOT)
< 500
500-1500
> 1500
Risk
Acceptable
Marginal
Unacceptable
Pre-school children and pregnant women are those groups at greatest
health risk from lead absorption. Rectification functions substituting
health census data were used to develop maps of homes where young
children or pregnant women resided. Using overlay functions, these
maps were then intersected with the risk indices maps to yield output
maps of "high risk residences."
Neighborhood functions were then used to assess "cleanup zones."
The original removal strategy was a zonal approach, where cleanup
would be accomplished in particular areas of towns. Construction tech-
niques that isolated entire blocks, kept equipment in single areas or
accomplished block-long removals followed by replacements were
among the several logistic considerations that made cleanup zones a
favored approach.
All sites — summary
§ resid
| sampled
Average SHR
SHR < 0.5
SHR 0.5 - 1.5
SHR > 1.5
271
202
3.7
5%
14%
81%
F-OOOO, F-0050, F-01SO
F-0300. F-0350
Smeltervllle Town sit*
Eichels Addition
Sweeny Townslle
resid
§ sampled
Average SHR
SHR < 0.5
SHR 0.5 - 1.5
SHR > 1.6
F-0200, F-0250
Miller Amd. Addition
Slier Addition
t resid
sampled
Average SHR
SHR < 1.6
SHR 0.5 - 1.5
SHR > 1.6
Figure 5
Ciiy of Smcltsvillc Summary of Sampling and
Sub-Chronic Hazard Ranking (SHR)
434 SITE REMEDIATION
-------
GIS neighborhood functions are particularly efficient in such analyses.
Potential cleanup zones were defined as any 100 contiguous lots and
summary statistics were produced for each possible combination. These
analyses showed that no single 100 unit area would address more than
11 homes having preschool children. No more than 21 homes would
be affected if the definition of the risk group were extended to children
under 9 years.
Because of the small number of children impacted, zonal strategies
were rejected. The second evaluation examined all possible combina-
tions of non-contiguous blocks. This methodology resulted in a best
combination including 20 homes (out of 100) having pre-school chil-
dren present.
Based on these findings, the cleanup "zone" strategy was abandoned.
U.S. EPA, Agency for Toxic Substances and Disease Registry (ATSDR)
and State Health officials opted for a house-by-house strategy that tar-
geted young children and pregnant women. This was precisely the ''map
of high risk residence" developed above. However, the problem of
ranking these homes remained. In this case, that was accomplished by
overlaying the high risk residence map over the soil lead concentration
map and ranking the properties by lead concentration. An output file
containing the information shown in Fig 6 was produced for all high
risk homes. These data were used to contact homeowners and residents
and initiate removal activities.
ProDertv ID
D-1010-001-005-0
D-0010-018-002-A
F-0550-001-015-0
(1) Fiction
Owner Owner Owner
Name Address City
John Doe U80 Main Kellog
Bill smith 510 Howard Kellogt
Joe Willis 012 F St Boise
Blood Lead
Property ID (ua/dll
D-0100-001-005-0 40
D-0010-01S-OOZ-A 33
F-0550-001-015-0 30
al data for display only
esident Resident
ohn Doe 1480 Haln
red Jones 115 Hill St
arc 1 Miller 817 Main St
Sanple Results Pb (ppnO
Soil Litter Dust
13400 9660 5240
9B20 8370 2460
5330 10500 4550
Resident
Citv
Kellogg
Kellogg
SmelterviUe
Figure 6
Sample Listing of Prioritized High Risk Homes'
Tracking Remedial Progress
Several steps are involved in accomplishing remedial activities on
these properties. Both homeowners and residents must be contacted
and permission must be obtained. The remediation must be negotiated
and completed and there is provision for continued monitoring of both
the environment and the residents. All of these functions can be easily
tracked as attribute files in the GIS. This will aid in recordkeeping,
providing progress maps and logistic assistance in future remediations.
CONCLUSIONS
GIS proved to be a valuable tool in several aspects of the 1989 yard
soil removal project. The overall data base was accessed to provide a
mechanism for contacting the nearly 3,000 affected homeowners and
residents. These contracts served a dual notification and data verifica-
tion purpose. The system was then used to produce non-confidential
data displays for risk communication in public forums and detailed con-
fidential maps and summaries for project personnel to use in individual
consultations.
Cartographic analysis techniques were then used to rank properties
for remediation based on land use, susceptible populations and soil con-
taminant levels. These results were used to assess and select remedial
strategies based on health risk and logistic criteria. The GIS will also
be used to track remedial progress. These multiple tasks and inventory
functions demonstrate the utility and flexibility of GIS in projects of
this type.
REFERENCES
1. Woodward Clyde Consultants and TerraGraphics, Interim Site Characteri-
zation Report for the Bunker Hill Site, EPA Contract No. 68-01-6939, Walnut
Creek, CA, Aug. 4, 1986.
2. Dames & Moore, Bunker Hill RI/FS: Data Evaluation Report, Mar. 16, 1988.
3. Yankel, A.J., von Lindern, I.H. and Walter, S.D., The Silver Valley Lead
Study: The Relationship Between Childhood Blood Lead Levels and En-
vironmental Exposure, JAPCA 27, pp. 763-767, 1977.
4. Wegner, G., Shoshone Lead Health Project Summary Report, Idaho Depart-
ment of Health and Welfare, Boise, ID, Jan. 1976.
5. Jacobs Engineering Group, Inc. and TerraGraphics, Final Draft Endanger-
ment Assessment Protocol for the Bunker Hill Superfund Site, U.S. EPA Con-
tract No. 68-01-7531, Feb. 1988.
6. Centers for Disease Control, Kellogg Revisited—1983 Childhood Blood Lead
and Environmental Status Report, 1986.
7. Roy F. Weston, Federal On-Scene Coordinators Report Bunker Hill Initial
Removal Action Kellogg, Idaho, U.S. EPA Contract No. 68-01-6669, May
28-June 25, 1986.
8. Panhandle Health District, Summary Report 1988 Lead Health Screening
Program, Silverton, ID, Sept. 1988.
9. von Lindern, I.H. and von Braun, M.C., The Use of Geographic Informa-
tion Systems as an Interdisciplinary Tool in Smelter Site Remediations, Proc.
Natl. Conf. on Management of Uncon-trolled Hazardous Waste Sites,
Washington, pp. 200-207, HMCRI, Silver Spring, MD, pp. 200-207, 1986.
10. Shoshone County Tax Assessor, 1988 Update of Property Owner Files, Wal-
lace, ID, June 1989.
11. TerraGraphics, 1986 Residential Soil Survey Status Report, Contains Con-
fidential Data, Dec. 31, 1986.
12. TerraGraphics, Bunker Hill Site RI/FS, Soils Characterization Report, Dec.
31, 1986.
SITE REMEDIATION 435
-------
Application of the Observational Method to an
Operable Unit Feasibility Study—a Case Study
David L. Mark, R.G.
Larry A. Holm, P.E.
CH2M HILL
Santa Ana, California
Neil L. Ziemba
U.S. EPA
San Francisco, California
ABSTRACT
The observational method, developed for application in the
field of geotechnical engineering by R.B. Peck, has been applied
to the Whitiicr Narrows Operable Unit Feasibility Study (OUFS).
The observational method, a means of engineering under uncer-
tainty, presents opportunity for potential savings in project time
and costs. Recent U.S. EPA guidance has proposed implement-
ing this "Streamlined Approach" as a way to improve the Super-
fund Rl/FS process. The key components to the observational
method are described and illustrated by case example in this
paper.
The Whittier Narrows OUFS is pan of the San Gabriel Basin
Superfund site Rl/FS. Groundwater is only known to leave the
approximately 200 mP San Gabriel Basin, located in northeast
Los Angeles County, California, through the 1.5-mi-wide Whit-
tier Narrows. The purpose of the Whittier Narrows Operable
Unit is to control the migration of contaminated groundwater out
of the San Gabriel Basin. Remedial alternatives presented and
evaluated in the Whittier Narrows OUFS incorporate technolo-
gies for groundwater extraction, treatment, treated water use and
monitoring.
Data from a limited site investigation were used to formulate a
working hypothesis of the most probable site conditions and the
maximum credible deviations to those conditions. Alternatives
were developed to address both potential site conditions. In addi-
tion, general response actions to potential deviations are pre-
sented for each remedial alternative. Cost ranges based on the
most probable case and the maximum credible deviation case are
presented for each alternative.
The greatest challenge in applying the observational method to
the Whittier Narrows OUFS revolved around developing response
plans for the maximum credible deviation case. The maximum
credible deviation case is not a "worst case" scenario. It is based
on an evaluation of the uncertainty in the extent of contamina-
tion, the types of contaminants and their concentrations, and the
hydrogeologic parameters that govern contaminant transport.
However, if potential deviations to every parameter that affects a
remedial alternative are considered to occur simultaneously (i.e.,
compounding uncertainty), the required response is unrealistic.
For the Whittier Narrows OUFS, potential deviations to the
three-dimensional extent of contamination and respective con-
taminant concentrations (the parameters with the greatest uncer-
tainty and the greatest effect on potential remedial actions) are
used as the basis for the maximum credible deviation case. That
is, by developing response plans to address potential deviations to
the nature and extent of contamination, it is expected, in this
case, that deviations to the other parameters that affect contam-
inant migration can be managed with the same response.
In applying the observational methods to an OUFS involving
several parameters that could affect a remedial action (ground-
water flow conditions, contaminant types and concentrations,
extraction rates, etc.), it became apparent that a high number of
possible combinations of deviations could occur. Thus, it is not
practical to define specific responses to deviations for each al-
ternative. Instead, general response actions for the main com-
ponents of each alternative (e.g., extraction, treatment, water dis-
posal and monitoring) are presented along with design considera-
tions to facilitate modification. In the design phase, however,
specific plans for monitoring to detect potential deviations and
for subsequently modifying the remedial action will be developed.
INTRODUCTION
The Whittier Narrows Operable Unit is pan of the San Gabriel
Basin Rl/FS. The San Gabriel Basin, a 170-mP groundwater
basin, is located in northeast Los Angeles County (Fig. I).
Groundwater is the primary source of drinking water for the more
than 1,000,000 residents of the San Gabriel Valley. Extensive vol-
atile organic compound (VOQ contamination prompted the U.S.
EPA to place the San Gabriel Basin on the NPL.
Whittier Narrows is a 1.5-mi-wide gap in the hills which serves
as the boundary between the San Gabriel Basin to the north and
the Central Basin to the south (Fig. 1). Groundwater is only
known to flow out of the San Gabriel Basin through Whittier
Narrows. VOC contamination in and up-gradient of Whittier
Narrows prompted the U.S. EPA to designate the Whittier Nar-
rows area as an Operable Unit. The primary objective of the
Whittier Narrows Operable Unit is to control the migration of
contaminated groundwater from the San Gabriel Basin, through
Whittier Narrows and into the Central Basin.
In accordance with the NCP, a Draft Operable Unit Feasibility
Study (OUFS) for Whittier Narrows has been prepared. The
Draft Whittier Narrows OUFS was released for public review in
fall of 1989. A ROD is expected in early 1990.
APPROACH—OBSERVATIONAL METHOD
An approach to remediation that is demonstrated in the Whit-
tier Narrows OUFS to be more efficient and timely than the cur-
rent process for Superfund site remediation is proposed for the
Whittier Narrows Operable Unit. This approach, referred to as
the observational method (recently coined the "Streamlined
Approach" by the U.S. EPA), has been adapted from similar
436 SITE REMEDIATION
-------
SAN GABRIEL MOUNTAINS
SAN GABRIEL BASIN
FIGURE 1
LOCATION MAP
Figure 1
Location Map
methods developed for engineering under uncertainty in the geo-
technical field. This application of the observational method is
demonstrated in the Whittier Narrows OUFS to be consistent
with CERCLA/SARA, the NCP and U.S. EPA guidance for re-
mediation. The method provides a logical and consistent integra-
tion of the RI, FS, ROD, Remedial Design (RD) and Remedial
Action (RA) process.
The current Superfund process of remediation is based on the
traditional "study-design-build" engineering project sequence.
This process assumes that after the RI/FS is complete, residual
uncertainties at a site are reduced to manageable levels. The
observational method recognizes that while considerable time, ex-
pense and effort can be devoted to attempting to characterize
the complex subsurface, residual site uncertainties can be signifi-
cant; and monitoring and modifications to the remedial action
are to be expected. Using this approach, remedial action activities
may be initiated more quickly than with the traditional approach.
The complete application of the observational method embod-
ies eight general ingredients. The term "ingredients" is used be-
cause they are not necessarily followed in a sequential manner.
And, in fact, several of the ingredients are conducted iteratively.
The eight ingredients of the observational method (according to
Peck1) are as follows:
• Evaluate existing data and conduct investigation sufficient to
establish the general nature, pattern and properties of the phys-
ical setting and contamination conditions. The level of site
characterization depends on the site and the expected general
response actions.
• Assess the most probable site conditions and maximum credible
deviations from these conditions. The most probable site con-
ditions are working hypotheses based on interpretation of
available data and are not necessarily based on a statistical eval-
uation. The maximum credible deviations from the most prob-
able conditions do NOT represent worst-case scenarios or max-
imum conceivable conditions, but credible conditions based on
interpretation of existing data. If a reasonable working
hypothesis of the most probable site conditions cannot be
developed, additional remedial investigation may be required
(i.e., the ingredient above).
• Evaluate alternatives and establish a remedial design based on
the hypothesis of the most probable site conditions.
• Calculate or estimate the physical and chemical conditions ex-
pected to be observed during implementation and operation of
the remedial action, given the most probable site conditions.
• Calculate or estimate the same parameters for the remedial
action given maximum credible deviations to the most prob-
able conditions.
• Select a course of action based On the most probable con-
ditions, and prepare contingent design modifications for fore-
seeable maximum credible deviations.
• Construct and operate the selected remedial action, monitor the
selected parameters and evaluate the observed conditions with
SITE REMEDIATION 437
-------
Figure 2
Application of Obterv»uon*l Method to
Superfund RemeditJ Process
respect to the working hypothesis of the most probable con-
ditions and credible deviations.
• Modify the remedial action through the predetermined course
of action to suit actual conditions, as required.
The general application of the observational method to the
Superfund remedial process is illustrated in Figure 2.
The observational method offers distinct advantages to the
timely and effective implementation of remediation in the pres-
ence of substantial uncertainty. In addition, by developing re-
sponse plans, design modifications for deviations and a flexible
initial design, required RA modifications resulting from observed
deviations can be expedited.
The reamifications of using the observational method occur in
almost every section of the Whittier Narrows OUFS. The site con-
ditions and the nature and extent of contamination are described
for most probable conditions and maximum credible deviations;
and the uncertainty in developing these hypotheses is clearly
spelled out. For the baseline risk assessment, a range of potential
exposure point concentrations based on most probable conditions
and credible deviations is estimated.
Remedial alternatives presented in the OUFS incorporate four
main components: groundwater extraction, treatment, treated
water use and monitoring. In this OUFS, these components are
based on probable conditions. Also, modifications to the initial
remedial action, in response to an observed deviation, are also
presented for each component. For example, extraction options
incorporate the estimated number and location of wells required
to control contaminant migration given the most probable extent
of contamination. Additional wells and pumping rates are iden-
tified to respond to deviations, if they occur. For the treatment
options, a required facility is determined from most probable
conditions; and modifications to that facility to respond to poten-
tial deviations are identified. For example, in response to an ob-
served deviation in influent VOC concentrations, a larger blower
on an air stripping facility may be retrofitted to increase the air-
to-water ratio and enhance the VOC removal rates. Such mod-
ifications are included in the cost estimate range.
Measures to facilitate implementation of modifications in re-
sponse to deviations are also presented for the major compon-
ents. For example, it is recommended that treated water distribu-
tion pipelines be sized and constructed with the capacity required
for the maximum credible deviation case. It is expected to be
more cost-effective in the long term to install the oversized pipe-
lines rather than construct a smaller pipeline and have to tear up
streets, remove the original pipeline and reinstall a larger pipe-
line, in, very possibly, the near future.
Remedial alternatives presented and evaluated in the OUFS are
based on both most probable conditions and maximum credible
deviations. Alternatives include initial remedial actions and mod-
ifications that may be required if maximum credible deviations
occur. Thus, a range of cost estimates is presented for each altern-
ative. As much flexibility as possible is incorporated into remed-
ial alternatives.
The final section of the Whittier Narrows OUFS presents a gen-
eralized strategy for implementing the observational method.
Specifically, a generalized approach for responding to observed
438 SHI, REMI.DIATION
-------
deviations, and flags that indicate modification to a remedial
action component may be required, are presented. An imple-
mentation strategy is crucial to executing the observational
method for several reasons. Primarily, deviations are likely to
occur incrementally; and, thus, incremental modifications to
some or all of the remedial action components would be expected
to be required.
HOW THE OBSERVATIONAL METHOD IS
INCORPORATED INTO THE MAJOR REMEDIAL
COMPONENTS
There are four main components to potential remedial actions
in Whittier Narrows:
• Groundwater Extraction
• Treatment
• Discharge of Treated Water
• Monitoring
How the Observational Method is incorporated into the evalua-
tion of these components for the Whittier Narrows OUFS is dis-
cussed in the following sections. Applying the observational
method, given the uncertainty in the numerous aspects of remed-
iation, provided a significant challenge.
Groondwater Extraction
Factors affecting the groundwater extraction rate required to
meet the remedial objectives are summarized as follows:
• Nature and Extent of Contamination
Types of contaminants
Horizontal extent
Vertical extent
• Groundwater Flow Hydraulics
Hydraulic conductivity
Specific yeild (storage coefficient)
Porosity
Hydraulic gradient
Aquifer thickness
Pumping
Recharge (natural and artificial)
Basin boundary conditions
(e.g., groundwater outflow)
• Contaminant Transport Parameters
Dispersivity
Retardation
Degradation
A working hypothesis of the most probable conditions with re-
spect to the parameters listed above has been determined based on
available data and numerical modeling.
A dilemma was encountered in estimating maximum credible
deviations to the most probable conditions. Credible deviations
are to account for uncertainty. However, compounding the un-
certainty in a few of the parameters above, let alone all of them,
results in conditions that are unrealistic (i.e., no longer credible).
For example, the amount of extraction required to control con-
taminant migration, given combined maximum credible devia-
tions to the extent of contamination and hydraulic conductivity,
would result in extracting every single drop of groundwater that
flows through Whittier Narrows (e.g., up to approximately
40,000 ac-ft/yr, or 25,000 gpm of continuous pumping 24 hr/day
year-round).
Credible deviations to the parameters that have the greatest un-
certainty and the greatest effect on required groundwater extrac-
tion schemes were evaluated individually (i.e., assuming most
probable values for the remaining parameters). These parameters
are as follows:
Nature and extent of contamination
Hydraulic conductivity
Hydraulic gradient
Storage coefficient
Aquifer thickness
Porosity
Of these parameters, the nature and extent of contamination
and hydraulic conductivity have the greatest uncertainty and
greatest effect on required groundwater extraction in Whittier
Narrows. Deviations to contaminant concentrations, lateral ex-
tent and vertical extent have been evaluated. Hydraulic conduc-
tivity, which has the greatest effect of groundwater flow hydraul-
ics, was evaluated using the mean absolute deviation to the over
100 calculated values of hydraulic conductivity for the area. Re-
quired groundwater extraction under the various conditions was
evaluated using numerical modeling.
The estimated required pumping given maximum credible devi-
ations to the nature and extent of contamination is approximate-
ly the same as the estimated pumping required under maximum
credible deviations to hydraulic conductivity. And, as previously
discussed, combining deviations to both parameters is unrealistic.
For the Whittier Narrows OUFS, the required pumping given
credible deviations to the nature and extent of contamination is
used for a maximum credible deviation case. This increase in
pumping due to deviation consideration represents up to a 41%
increase in the required extraction over that required under most
probable conditions.
Groundwater extraction schemes proposed in the Whittier
Narrows OUFS are based on most probable conditions, with
modifications (additional wells and/or higher pumping rates)
for observed deviations. Observations that indicate modification
to an extraction system may be required are discussed later.
Treatment
The following treatment technologies were incorporated into
potential Whittier Narrows remedial alternatives:
• Stripping (packed tower, rotary and steam)
• Granular Activated Carbon (GAC) Adsorption
• Advanced Oxidation with Ozone/Peroxide
Influent flow rates (i.e., groundwater extraction) and contami-
nant types and concentrations were estimated for most probable
conditions and the maximum credible deviation case. Each of the
proposed treatment technologies would require modification for
deviations. However, some technologies would require substan-
tially more modification than others. A few examples of devia-
tions, required modifications and design considerations to facili-
tate modification are briefly described below.
Treatment
Technology
Modification Design Considerations
Higher VOC
tration
Increase oi;
•rater ratio
blower, or allow foe
replacement of blower
Higher VOC
concen-
tration
Onsite carbon Oversize carbon beds
regeneration on off-gas system,
for off-gas design for possible
addition of facilities
Higher Add stripping Design for additional
influent, tower* and towers to be added,
flow rate carbon beds adequate area at
for off-gnu treatment plant site
Higher
vinyl
chloride
concen-
trations
Replace GAC Design alternate
system (unable treatment system
to adequately
adsorb vinyl
chloride)
Advanced
Oridation
Higher
•ethylene
chlcride
and/or
carbon
tetra-
chloride
concen-
tration*
Add post
•tripping
Design for additioi
facilities, adequat
treatment
(netfaylene
chloride and
carbon tetra-
chlorlde not
adequately
oxidized)
site required
SITE REMEDIATION 439
-------
Observations that indicate modification to a treatment facility
may be required are discussed later.
Water Discharge
Pipelines are proposed to distribute treated water to local dis-
tribution systems and/or points of recharge. Deviations to flow
rates would be handled by expanding water distribution to addi-
tional local systems and/or by constructing injection wells.
Design considerations to facilitate modification to treated
water distribution systems include oversizing pipelines where
additional future capacity may be required. Constructing an over-
sized pipeline (i.e., sized for the maximum credible deviation
case) would be less expensive than replacing a pipeline in the near
future, especially in an urbanized area where pipeline construc-
tion involves digging up streets.
Monitoring Program
In applying the observational method, a monitoring program
is crucial. In addition to performance monitoring of the RA, the
monitoring program must be designed to verify most probable
conditions and detect deviations. By initiating the monitoring
program during the design process, an early indication of base
conditions, relative to the most probable case or the maximum
credible deviation case, is provided.
Parts of the monitoring program that provide information to
the RD are highlighted in the Whittier Narrows OUFS. Those
activities that are proposed for performance monitoring, but are
not crucial to the RD, are not proposed for implementation as
part of the RD.
As with the other components to remedial alternatives, devia-
tions to expected conditions may require modifications to the
monitoring program. Modifications to the RA may require mod-
ifications to the monitoring program.
Estimating potential modifications to the monitoring program
that may be required is difficult. For the Whittier Narrowi
OUFS, it is estimated that 33% additional monitoring wells may
be required for the maximum credible deviation case.
Indications that modification to the monitoring program may
be required are discussed below.
HOW THE OBSERVATIONAL METHOD IS
MANIFESTED IN THE OUFS
The observational method is manifested throughout the
Whittier Narrows OUFS. Description of the physical setting,
which includes the nature and extent of contamination, is pre-
sented in terms of most probable conditions and marin^im cred-
ible deviations. For the baseline risk assessment, two risk calcula-
tions are presented: one risk calculation was based on the most
probable nature and extent of contamination and one was based
on the estimated maximum credible deviation case. Remedial
alternatives are developed for most probable conditions, and
modifications are identified that may be required in response to
deviations. Cost estimates for both conditions are included. The
description of alternatives includes discussion on the action
necessary to respond to the deviations and design considerationi
to facilitate the timely modification if deviations are observed.
The detailed evaluation of alternatives in the OUFS addresses
the following criteria (per U.S. EPA Guidance for Conducting
an RI/FS Under CERCLA, October, 1988):
GRQUNDWATER EXTRACTION
TREATMENT
OBSERVATION
DEVIATION
CONTINGENCY
PLANS
OBSERVATION
DEVIATION
CONTINGENCY
PLANS
c
f
oesr*-* j
OfOtBiJ^ ^ OBSERVE
J
^— -
U2< IOADWCS MICHES )
\THAN EKPECTED /
*s
^
I/A«ECHŁMICAI.S \
/ PREStMt IHAt \
/ CAN NOT BC \*O
I AotouArut r—
\ TRtATEDBTTMC /
\ IECMNOLOCT /
CQ*fCENTRATION r\.OW RA
1
TŁ
/CXNOBOUtf»OE\ __ /SMVU
( CRJTEMABE \_IiiJ /(XOWR/
\ MET / / CRtATO
MO
r
r~
CMAMCC *EU
PUUPINC RATES
— —
^H —
l
\PLAMTC
1 1
•
MOW" CONSTRUCT AOOfflONAl.
TREAHCItT TBCATUENT f AOUTtCS
SYSTEM
D.T \
t \ w
ITKAH V— '
*' /
jftarrl
t
WATER USE
OBSERVATION
DEVIATION
CONTINGENCY
PLANS
GRQUNDWATER MONITORING
OBSERVATION
DEVIATION
CONTINGENCY
PLAN
/ IS MOMTOR'W
PROGRAM AOIOUATE
TOIOCNTTY
DEVIATIONS PRIOR
TO RtOUIRED
UOOinCATIONS TO
T»t REMtCKAJ. ACTION .
CAN THE
MONITORING
PROGRAM OETJRUWE
THE EXTENT AND
CAUSE OF AN
OBSERVED
DEVIATION
Figure 3
Obiervctlonal Method Implementation Strategy for the Major
Components to Remedial Action Alternatives
440 SITE REMEDIATION
-------
• Overall Protection of Human Health and the Environment
• Compliance with ARARs
• Long-term Effectiveness and Permanence
• Reduction of Toxicity, Mobility or Volume Through Treatment
• Short-term Effectiveness
• Implementability
• Cost
A remedial alternative may undergo modification in response
to observed deviations. Differences between the alternative de-
signed for most probable conditions and that alternative as mod-
ified for the maximum credible deviation case are addressed with
respect to the criteria listed above. For example, if a treatment
plant is modified for on-site carbon regeneration, consideration
is given regarding ARARs (e.g., requirements regarding trans-
port and disposal of condensate) and cost.
Two cost estimates are presented for each remedial alternative:
one estimate is based on most probable conditions and one is
based on the maximum credible deviation case. For the Whittier
Narrows OUFS, estimated costs for the maximum credible devi-
ation case are based on the assumption that initial construction
of all facilities is required. This assumption was made for ease of
preparing cost estimates, and because it is possible that the max-
imum credible deviation case currently exists. Most likely, the cost
of gradual or incremental modification would vary from this
estimate.
Cost estimates for the most probable case include costs for
measures that would facilitate modification (e.g., oversized pipe-
lines). The greater up-front cost could be expected to result in a
net future cost savings if deviations occur, as are expected.
The final section of the Whittier Narrows OUFS presents a
summary of the general strategy for responding to observed dev-
iations from: (1) the most probable site conditions and (2) the
expected performance of an implemented remedial action. Given
the high number of possible combinations of deviations that
could occur, it is not practical to define specific responses to devi-
ations for each alternative at the OUFS level. In the OUFS, gen-
eral strategy for dealing with observed deviations, and indicators
of a deviation that may require RA modification, are presented.
For the major components of the remedial alternatives
(groundwater extraction, treatment, water use and monitoring),
a general strategy for dealing with observed deviations is shown
graphically in Figure 3.
Water quality parameters in Whittier Narrows to be observed
in the monitoring program are presented in the OUFS. Table 1
summarizes the ranges of expected contaminant concentrations
for most probable conditions and deviations to the nature and ex-
tent of contamination in the Whittier Narrows area.
Table 2 summarizes expected contaminant concentrations
observed during performance monitoring.
Table 3 summarizes indicators of deviations and responses or
modifications to the treatment technologies that may be required
if deviations to treatment plant influent or effluent are observed.
Table 4 summarizes options for dealing with deviations in dis-
charge rates. Indications that modification to the treated water
distribution system may be required are straightforward in that
the receiving distribution systems either do or do not have the
capacity to take additional water.
An important point discussed in the Whittier Narrows OUFS
is that careful reevaluation of the RA will be required prior to
modification.
As more data on the nature and extent of contamination, aqui-
fer properties and flexibility of the chosen treatment technology
become available during remedial design and pilot testing, plans
for response to observed deviations can be refined and specific
modifications can be designed.
CONCLUSIONS
Applying the observational method is expected to result in
Table 1
Ranges of Expected Contaminant Concentrations Representing Most
Probable Conditions and Deviations to the Nature and Extent of
Contamination In Whittier Narrows
Well Number
or Group
Central Basin
production wells"
Proposed clusters near
basin boundary11
Bartolo Well6c
Proposed new wells in
upgradlent areas of
Whittier Harrows'1
Wells located between
the two large areas of
contamination'
01900001
01900094
01900331
01901749
01902579
08000004
Ranges Expected For Most
Probable Conditions(uR/l)
All contaminants MCL
TCE- >8, any other MCLs
Any contaminant >MCL
PCE- >25, TCE- >10, any other
> HCL
PCE, TCE- >10, any other >MCL
Any contaminant >MCL
PCE- >10, any other >HCL
PCE- >200, TCE- >10, any
other >MCL
Any contaminant >MCL
PCE, TCE- >10, any other >MCL
PCE- HCL to 40, TCE- MCL to
15, all others 40, TCE- >15, any other
>MCL
PCE, TCE- >15, any other >MCL
PCE- >10, any other >MCL
PCE, TCE- >10, CTC- >4, any
other >MCL
PCE, TCE- >10, CTC- >4, any
other >MCL
08000049
11900095
41900745
81902525
81902635
MP-i and WC-1 PCE, TCE- to 15 (to PCE, TCE- >15 (to 300 feet)
300 feet) PCE, TCE MCLa below
below 300 feet, all others 300 feet, any other >MCL for
30, TCE- >8 (to
(to 200 feet) PCE, TCE 200 feet) PCE, TCE >MCLs
MCLs for any depth.
MP-2 PCE- 40 to 100, TCE- MCL to PCE- >100, TCE- >20 (to
20 (to 300 feet) PCE, TCE 300 feet) PCE, TCE >MCLs
MCL at any
proposed extraction all depths depth
wellsb
Proposed clusters near All contaminants MCL at any
basin boundaryc all depths depth
Central Basin Discharge below
-------
Table 3
Effect of Treatment Technologic* on Varying Influent Condition*,1
Whlttier Narrow* OUF8
iK (ncntri royin iU
to* vlll llk.lf l»ng* b*tv»»n lOll to 'V
>ab*bl» dvilRn to cr»4IbU delation ewtdl
•w r*t>» vlll i»qi>lr* •ddKIaiMii ifRlt • onl
n.r (Kan ;.<-M-VO t.. i.OOO gpa rh.n«» !•*
4t«t r(b«i Ing t h» |nrr**»»d load •<-[»•> **
« ion l**«* c-t
carton r#pl«r**wnt
plng *UT b* f»»OO ()• *k*r r*^wl'
HCU CC14
AB«ia*Hn,( «11
; (-••*•• to flo« t*r* altowld b* k*>tkil*4 by t b* b«»» d*alg^ t«*rf tor* «bo
c«p«bl> ol (r*atln« flow* a* high *• 1)0 p*re*ot ol I!M 4«*lca *««h (1 7 r*Bir*l b*|i
fexbarn*
1,404
i. too
*,*•
10.*.
>.2M
11,400
Alt*rtut !*• t, awx of ib* WM flow c*p*clty
1.400*
i4 lor 4
more timely imptemenution of remediation for the Whittle;
Narrows Operable Unit, and possibly at a reduced cost. Gather-
ing data during the RD and RA as part of the monitoring pro-
gram is expected to be efficient and timely. The observational
method provides an acceptable means to expedite remediation
and logically manage and minimize risk.
REFERENCES
I. Peck, R.B. "Advantage* and Limitation* of the Observational
Method in Applied Soil Mechanic*." in Milestones In SoUMedimia,
TV First Ten Rankine Ltcttirts. Thomai Tel/ord Ltd., Edinburgh.
pp. 2*3-279. 1975.
2. U.S. EPA. Dnfi Whlttler Narrows Operable Unit Feasibility Stvfy,
San Gabriel Batiti. Lot Angela County, California. Prepand by
CH2M HILL, Santa Ana. CA. July 1989.
442 SITF-. REMEDIATION
-------
Advantages for the Regulated Business Community
Through Compliance with SARA Title in
John R. Stamatov, C.E.
Judith A. Barber
Susan W. Stoloff
Clean Harbors Environmental Engineering Corporation
Braintree, Massachusetts
ABSTRACT
SARA Title HI has revolutionized the approach to managing
chemicals within our communities. No longer can a company
operate "behind closed doors." With the establishment of the
State Emergency Response Commissions (SERCs) and the Local
Emergency Planning Committees (LEPCs), the general public
now has access to a wealth of information pertaining to a com-
pany's use of hazardous chemicals. Not only are the provisions
of Sections 302, 304, 311, 312 and 313 demanding in their own
right, but the Act also provides the LEPC with extremely broad
authority to influence and implement the development of com-
munity emergency planning actions.
Fortunately, this Act puts the business community on equal
footing with the rest of the community. Through the LEPC, the
business community can play an integral role in the formulation
of emergency response planning and LEPC policy. With a voice
on such a committee, business leaders can help establish a con-
structive rapport between business, media, public servants and
the general public. Involvement of the business community can
also add a greater degree of technical competence to such a
forum.
The focus of this paper is how industry's pro-active involve-
ment with the community in the emergency response planning
process can result in tangible benefits for industry as well as the
community at large.
INTRODUCTION
The need for comprehensive emergency response planning was
firmly established after the tragic chemical release in Bhopal,
India. Since that time, efforts by the Chemical Manufacturing
Association (CMA), public interest groups and concerned com-
munities have resulted in the formation of awareness programs
and local, state and federal laws and regulations, all addressing
the concerns of chemical hazards hi the community and the need
to protect the public and environment from these hazards. In
1985, CMA developed the Community Awareness and Emer-
gency Response program (CAER), designed to inform the public
of hazards inherent in the chemical industry and ways to pro-
tect the public from these chemicals in the event of a spill or re-
lease. The U.S. EPA then introduced voluntary programs which
included elements based upon the CAER program. These pro-
grams were the precursor to the Emergency Planning and Com-
munity Right-to-Know Act, also known as SARA Title III.
The intentions of SARA Title III are clear; to provide a com-
prehensive community emergency response plan to protect life,
property and the environment in the event of a chemical spill or
accident. This act requires by law the formation of and participa-
tion in the LEPC by community officials, public agencies and
community industries. Although the law provides for the levying
of severe penalties on industries if they fail to participate in the
LEPC, there are no provisions within the law that qualitatively
measure the results of LEPC efforts.
In essence, the ultimate goal of SARA Title III is no different
than that of CMA's CAER program. SARA Title HI, however,
now legally joins the community and industry at the emergency
planning table. For the chemical companies that have already en-
dorsed the CAER program, SARA Title III does not present new
regulatory requirements except for inventory and emissions re-
porting. Many companies not previously involved in the CAER
program are now faced with what they perceive to be just another
regulatory burden absorbing more resources and requiring addi-
tional manpower. To the contrary, however, industry's involve-
ment in the LEPC can provide industry with many distinct and
tangible benefits.
ADVANTAGES OF INVOLVEMENT
Liability
Non-compliance with the minimal LEPC participation require-
ments of SARA Title III can result in the imposition of penal-
ties of up to $25,000 per day. Pro-active and energetic involve-
ment by a company that goes beyond the minimal requirements,
however, will not only avoid such penalties, but also will enable
the company to realize significant advantages that are critical to a
business' continued prosperity. One obvious advantage is the re-
duction in risk of liability a company may realize if it has invested
the time and effort to ensure that its community's emergency
response plan works. A workable plan can be achieved only
through efforts on the part of the company that go beyond those
steps required by law.
It is an undisputed fact that the liabilities associated with major
spill response efforts are prohibitive. Cleanup costs alone can be
staggering. Costs associated with property damage and economic
losses add further to an already expensive bill. If there are in-
juries or even death as a result of an accident, the compensatory
costs can exceed the physical cleanup costs. In addition, the poor
public perception a company will receive if a spill response is mis-
handled is only too vivid in most Americans' minds. Damage con-
trol and recovery from these mishaps can be absorbed only by
the largest conglomerates.
With the erosion of sovereign immunity, there could be some
liabilities on the part of individual town officials or towns hi the
event of damages arising out of actions taken during an emer-
SITE REMEDIATION 443
-------
gcncy response. To avoid liability in the event of a spill or acci-
dent, a town must be able to defend that it has responded to the
best of its ability. To do a good job, towns must have the re-
sources, expertise and active cooperation of industries in their
area. In instances where the town may be negligent or at fault, it
is hard to imagine that some, or a good part, of the blame will not
fall on industry's shoulders. Therefore, it behooves industry to
ensure that their towns have the most effective and qualified
emergency response capability possible.
Unfortunately, accidents and spills will occur, but the actions a
community takes immediately following an event will dramatical-
ly affect the overall personal and environmental damage. If indus-
try and the community together develop effective emergency re-
sponse plans, the damage from accidents can be greatly miti-
gated. In effect, a competent response to spills or accidents can
directly reduce industry's risk of liability.
Improved Public Relations Equals Smooth Operation*
A poorly informed public is usually fearful, and this fear can
significantly impact operations, perhaps forcing the closure of a
facility. The public is becoming more educated through the media
about the hazards of chemicals. They have a right to know what is
going on in the chemical plant or warehouse down the street...
and they know it! Industry must understand the public's fear and
realize that it can be translated into operational difficulties.
The CMA recognized that this fear in people's minds concern-
ing the hazards associated with chemicals could lead to over-reg-
ulation of the chemical industry. Thus, through the formation of
CAER in 1985. the CMA took steps to attempt to reverse the
public's perception of the chemical industry. This program
stressed cooperation and interaction with the local communities.
The fundamental goals of CAER were to make the public aware
of the hazards and to take the protection necessary to minimize
or eliminate these hazards. Large chemical industries, with plenti-
ful resources, generally embraced the goals of CAER. Unfortun-
ately, it proved to be very difficult to establish a rapport be-
tween the companies and towns. People still believed that the
companies were trying to hide something and were not sincere.
In addition, the plant managers did not have the knowledge and
skill to communicate with the community and the media.
By the time SARA Title III was implemented, many companies
had emergency response plans in place but had not integrated
these plans effectively with the community. Following in the foot-
steps of CAER. SARA Title III was designed to open the door to
effective communication between industry and the community
through the establishment of the LEPC.
Through the LEPC, industry now has a legitimate conduit for
access and dissemination of information to the community. The
LEPC can be used a.s a forum to educate the public, reduce
mounting fears associated with industrial practices and activi-
ties, and ultimately increase industry's profitability through the
establishment of improved public relations and perceptions. At
LEPC or similar meetings, industry groups can provide the pub-
lic with information about the products they produce and the
chemicals needed to produce such products in response to public
demands. Through education, the public hopefully will under-
stand that, as consumers of these products, they are partly re-
sponsible for the hazardous chemicals needed to produce these
consumer goods. Once the public gains this awareness and sense
of responsibility, the bridge between the public and the industry
is established and effective communication between the two enti-
ties can take place.
Although the benefits to industry from better communica-
tions are improved public perceptions and relations, industry
must take care to ensure that this effort is, in fact, a sincere
effort. This effort on the part of industry must be focused upon
educating and ultimately protecting the public while maintaining
receptiveness to demands for consumer products. This effort
must not be, nor be perceived to be, a public relations campaign
designed to serve the company's needs only. Public relations
campaigns will lack credibility and must be avoided.
An industrial company with amicable community relationships
may choose to expand or develop in that same community. When
a good neighbor relationship develops, industry will receive its
share of subtle benefits. The permitting process for new construc-
tion or development can be shortened because the public will sup-
port the endeavor or opposition to the project will be fragmented
at best. Day-to-day operations also will be much smoother with-
out unnecessary local interference.
The same public support can be beneficial when one-sided en-
vironmental groups attack a reputable community industry.
Without support from the public, campaigns against the industry
will be short-lived. In Springfield, Massachusetts, for example,
local industries have put a tremendous amount of effort into help-
ing their community develop an effective and workable emer-
gency response plan. Through the development of this plan, the
citizens of Springfield have been educated as to what industry
does and why. The relationship between industry and the publk
is so strong in Springfield that consumer advocacy groups have
been unable to justify to the local citizens efforts to campaign
against industry operations.
Pre-emption of Powlbfe Future CosU to Industry
During the implementation of SARA Title HI, the chemical in-
dustry lobbied to make the Act a workable piece of legislation
that would share the burden of responsibility between the towns
and industry. This piece of legislation could have been far more
burdensome for industry, and, in fact, it may become more
burdensome if a method is not found to adequately fund the
LEPC's efforts to implement emergency planning and provide
adequate training. Industry is in a position to pre-empt the im-
position of fees and additional regulation by filling this void.
The biggest impediment that most LEPCs must overcome is
that of little or no funding. While some industries and states pro-
vide this funding for training first res ponders, funding is sorely
lacking in many states. Bill Kremer of the Federal Emergency
Management Agency (FEMA) explained that the federal govern-
ment, through FEMA, has supplied some funding and expertise
to assist emergency planning efforts. However, this funding is
far less than is needed. In Massachusetts, for example, only
$70,000 of federal funds was available in 1989 to train all first re-
sponders throughout the entire state. In addition, seven emer-
gency response vehicles which the federal government provided
to the state cannot be used because no one is trained to use them.
The situation in Massachusetts is not atypical. Without adequate
training and equipment, the best emergency response equipment
and plans are useless.
To combat this funding void that was not addressed in the Act,
states such as Maine and New Jersey have levied a fee on indus-
tries that participate in emergency response planning pursuant to
the Act. Public pressure will no doubt encourage the imposition
of fees as long as chemical spills and emergencies continue to
occur. To prevent levying of fees in states where the fee system is
not yet established, or to prevent further increases in states where
this fee already is in place, industry must share in the massive task
of training both community firefighters and other responders and
in-house personnel. Joint training exercises with industry, police
and fire departments are essential. Joint exercises help alleviate
the heavy burden on fire departments. In addition, they improve
emergency response operations and help cement relations be-
tween the emergency responders.
Industry Resources
To reach the intended goals of SARA Title III, industry and
the community should share resources. Communities typically
have large equipment resources, town departments and evacua-
444 SITF RI-.MI-DIATION
-------
tion capabilities. Industry must be willing to provide the neces-
sary expertise to complement town resources, as most local offi-
cials do not have the background required to address the com-
plex issues associated with chemical management and safe, effec-
tive emergency response procedures. Industry also has many re-
sources that are not available to the LEPCs or the local emer-
gency response agencies.
In small towns, fire departments may not have the personnel or
equipment to field a complete emergency response team. Equip-
ment needed to evaluate and assess chemical dangers may not be
available to the town departments. In these situations, industry
should support the LEPC and the community by making avail-
able these resources and supporting them in the use of these re-
sources. Equipment resources and an emergency response plan
are useless in the hands of inexperienced responders.
Industry is a Member of the Community
A strong emergency response plan can reduce a company's risk
of liability in the event of a spill or accident. However, the fact
that a company helps prepare a worthwhile plan does not guaran-
tee that the many benefits resulting from a good neighbor atmos-
phere will materialize. Good public relations and a good neighbor
relationship will develop only if industry accepts that they too are
a member of the community.
The industrial community must recognize that LEPC involve-
ment must not serve only the public relations or marketing de-
partments. Instead, industry must approach the emergency plan-
ning task with the same professionalism and expertise that they
would devote to any internal endeavor.
As a member of the community, industry has the responsibility
to protect its citizens. Industry does not have the unalienable
right to operate. Industry must prove to the community that it
can operate with minimal risk to the citizens of the community.
Companies that fail to accept this responsibility face the risk of
hostile public outcry and the detrimental ramifications it en-
genders.
Industry's pro-active involvement with the community can
yield long-lasting and positive results. When industry does final-
ly embrace the open door policy, it often is met with serious
skepticism and concern from the public. A high level of cooper-
ation and mutual trust does not develop overnight, but the re-
wards to both parties involved are well worth the effort.
CASE STUDY
The City of Springfield, Massachusetts, which received one of
the first CAER awards for its superior efforts in emergency plan-
ning, provides us with a good case study. Jim Controvich, the
Springfield Civil Defense Coordinator, has spent the past 5 yr
developing a viable emergency response plan. Mr. Controvich
said his program would not have been successful without the help
of industry.
An example of this cooperation is help supplied by Monsanto
Chemical Company of Springfield, Massachusetts. Monsanto
provides many resources to the Springfield Hazardous Materials
Response Team (Haz Mat) including Haz Mat responders, chem-
ists and equipment as needed.
When George Lemos, the Environmental Operations Manager
of Monsanto, was asked how he justified such a commitment to
the local Haz Mat team, he responded, "It is a symbiotic rela-
tionship. In many instances, the City does not have the chemical
expertise and experience that we have. That's our business. On
the other hand, we don't have the firefighting experience and
equipment available to the police and fire departments. Take, for
example, the heavy fire fighting trucks. Considering what is at
stake in even the smallest incident, it would be downright fool-
hardy not to recognize each other's strengths and weaknesses,
and then form a strong relationship to build a synergistic haz mat
response capability."
Despite the strong support from Monsanto and other com-
panies in the City, Jim emphasized that all of the planning and
equipment is no substitute for practice. For this reason, Spring-
field has conducted monthly tabletop exercises. Explains Jim,
"Only when personalities and response teams have worked to-
gether in a time of stress can they gain each other's confidence
and recognize each other's capabilities. The tabletop creates a
stressful situation in which we learn a great deal about our respon-
sibilities and capabilities during an incident. And we made mis-
takes. Of course if we never make mistakes there would be no
need to practice!"
This community is prepared! Others are not.
In some cases, emergency response planning is confounded by
the public's belief that if industry cannot plan or is prevented
from adequately planning for emergencies, industry should leave.
This is evident in northeastern Massachusetts where several towns
are protesting the operation of the Seabrook Nuclear Power
Plant. Towns in this area have refused to submit emergency evac-
uation plans and cooperate with planning agencies. One com-
munity even dismantled a warning system in their town that was
paid for and installed by the utility. Seabrook had to develop an
emergency response plan independently of the communities that
the plan was designed to protect. The fact that this plan was not a
product of mutual cooperation raises serious doubts as to its
effectiveness in the event of an emergency. The towns' failure to
cooperate leaves their communities vulnerable to a disaster. Such
actions also greatly increase industry's risk of liability in the event
that an accident occurs.
To mitigate public opposition, industry must actively and sin-
cerely attempt to develop a relationship with their community. It
could take years for a relationship to develop between industry
and the community which fosters effective planning and trust.
Such a relationship, however, is absolutely crucial to protect the
public and to minimize industry's exposure to liability in the event
of a spill or accident.
WHAT INDUSTRY CAN DO
Chuck Losinger from HMM Associates, Inc. assisted CMA
during the early development of the CAER program. Although
Mr. Losinger agrees that the initial skepticism of the community
toward industry is difficult to eliminate altogether, a sincere
effort by industry will quickly destroy many misconceptions and
counterproductive concerns. Mr. Losinger has highlighted the
following items that industry can do that will help lead to mutual-
ly supportive and beneficial relationships between industry and
the community.
• Industry must supply an interested and capable person to par-
ticipate in the LEPC. The LEPC designee should not view the
appointment as an assignment but rather as a challenging, ex-
citing and worthwhile endeavor. The designee must appreciate
the importance of the LEPC charter. This person should, of
course, be trained in emergency planning.
• A company should be honest and forthright about its hazards
and should be pro-active and reach out to community groups
before being required or forced to do so. Conducting plant
tours is a great way to facilitate community outreach programs.
• On a periodic basis, company emergency response training
should be conducted in concert with local responders.
• If local responders do not have the necessary resources to effec-
tively respond to emergencies, companies should help to aug-
ment these resources by providing emergency response equip-
ment as required.
• A strong rapport with the local fire department should be
established. Fire departments should tour the faculty frequent-
ly and be aware of the company's operations and emergency
response capabilities.
• Companies can help to make the paperwork burden on towns
SITE REMEDIATION 445
-------
more manageable by supplying emergency planning informa-
tion (i.e., chemical inventories) in an organized format. The
CAMEO program and/or a minicomputer can be supplied
which will store data that can be recalled when necessary and
aid a community in emergency planning. The expertise needed
to operate the program also should be provided.
Mr. Douglas Forbes, from the Massachusetts Civil Defense
Agency, also has ideas on what industry can do to facilitate effec-
tive emergency response planning. Some of these ideas include the
following:
• Industry can increase awareness through the distribution of
printed public information pieces (calendars, pamphlets, etc.)
which address the need for various types of emergency plan-
ning.
• Industry can encourage emergency planning by hosting lunch-
eons for tabletop exercises where industry and the community
work together to respond to mock chemical emergencies.
• Industry can lend its management capabilities to emergency
planning and practice efforts.
• Industry can offer services to the disabled public such as assis-
tance in notification of the hearing-impaired.
CONCLUSION
SARA Title III provides an opportunity for industries and
communities to work jointly to create a mutually beneficial, safer
environment. Progress has been most effective where localities.
states and industry have worked closely together. States with
strong county governments provide a natural structure to deal
with the demands of emergency planning. On the other hand,
states in the northeast have weak county governments and strong
local governments which can impede regionalization and shared
resources. These areas are particularly in need of industry's lead-
ership and resources.
Industry is a part of the community. Let's all accept the chal-
lenge and make emergency planning a corporate goal.
ACKNOWLEDGEMENTS
We wish to thank the following people for their help and sup-
port in the preparation of this paper: Mr. Douglas Forbes, Direc-
tor of Planning, Massachusetts Civil Defense Agency; Mr.
William Kremer, Federal Emergency Management Agency; Mr.
George Lemos, Environmental Operations Manager, Monsanto
Chemical Company; Mr. Terry Nelson, Personnel Superin-
tendent, Monsanto Chemical Company; Mr. Robert Rusczek,
Safety and Hygiene Superintendent, Monsanto Chemical Com-
pany; Mr. Hank Nowick, Monsanto Chemical Company; Mr.
Jim Controvich, Springfield Civil Defense Coordinator, Spring-
field, Massachusetts; Mr. Chuck Losinger, Senior Vice Presi-
dent, HMM Associates, Inc., Concord, Massachusetts; Chief
Hobart H. Boswell, Jr., Foxboro Fire Department, Foxboro,
Massachusetts, and Chief Jackson Macomber and Fire Officer
Justin Cronin (Civil Defense Director and LEPC member), Avon
Fire Department, Avon, Massachusetts.
446 SITE REMEDIATION
-------
Superfund in Action:
A Case Study in Planning a Successful Project
Mary Ann Croce LaFaire
United States Environmental Protection Agency
Chicago, Illinois
Daniel M. Caplice
Geraghty & Miller Engineers, Inc.
ABSTRACT
Community relations (CR) activities have long been an important
part of the Superfund program. The interactive process normally begins
in a local community once a potential hazardous waste site is identi-
fied and targeted for investigative work and continues until the site has
been remediated. During the course of the many phases of work at the
site, the CR activities are quite varied and usually range from the simple
preparation of informational materials to the time-consuming prepara-
tion and conduction of public hearings.
Many times the success or failure of a Superfund project depends
on how well the CR activities are implemented. Successful remedia-
tion of the Superfund site in LaSalle, Illinois (a residential community
with extensive PCB contamination of the soil) was totally reliant on
the success of the CR activities that were planned and implemented
in the area. Without the full cooperation and support of the residents
and businesses in the neighborhood, complete remediation of the site
was impossible.
This paper is a case study of the CR planning and implementation
activities, both obvious and intangible, that are necessary at most Super-
fund sites and which were conducted at the LaSalle site. The paper
has been prepared in conjunction with a video documentary that was
produced in order to document the positive characteristics of the Super-
fund program and the role that CR activities play in that success. The
goal of both the video and written documentaries was to capture the
positive side of the Superfund program and to show that with careful
planning, projects can be successfully implemented even under extreme-
ly difficult conditions within a community.
INTRODUCTION
The LaSalle Electrical Utilities (LEU) site in LaSalle, Illinois, was
a former manufacturing facility of electrical equipment. The plant began
operations prior to World War II, and in the late 1940s it began using
PCBs in the production of capacitors. This manufacturing process was
continued until approximately October, 1978. By May, 1981, operations
at the LEU facility had ceased, and by September, 1983, the company's
last operating facility in Farmville, North Carolina had filed for
bankruptcy. In December, 1982, the LEU facility was included on the
first NPL.
The now abandoned LEU facility is located on the northern outskirts
of the City of LaSalle bordering a small residential community. Approxi-
mately 70 homes are located in this area, and nearly 190 people occupy
those homes.
Information concerning the waste handling and waste management
practices of the LEU company is limited. However, based on conver-
sations with local residents and former employees, it appears that the
company regularly engaged in the practice of applying PCB-
contaminated waste oil to parking lots, roads and alleys at the plant
and in the adjacent areas to suppress dust. Following the Federal regu-
lation of PCBs, LEU company manifests document the legal disposal
of the contaminated material.
PROBLEM
Even though the LEU company altered its PCB disposal practices
in order to comply with the new Federal regulations, its historic opera-
ting practices had already released PCBs into the environment both
on the company property and in the surrounding residential area. The
extent of contamination varied, but extensive investigations conducted
by the Illinois Environmental Protection Agency (IEPA) defined the
limits. The IEPA data revealed that the contamination did not stop at
the company's gates. Rather, the results showed that the residential com-
munity directly adjacent to the plant contained extensive soil contami-
nation at depths up to 3 ft. In addition, wipe and dust samples taken
from the interiors of the residential homes revealed the presence of low-
level PCB contamination.
REMEDY SELECTION
Based on the results of the extensive remedial investigative work at
the site, the U.S. EPA and the IEPA determined that all soil in the residen-
tial area with PCB concentrations greater than 5 ppm should be
addressed in the remediation project. The feasibility study evaluated
numerous technical alternatives for alleviating the problem. Possible
solutions ranged from capping the contaminants in place to excavating
the affected soil and destroying the material in a permanent off-site
incinerator.
Unlike most feasibility studies conducted at Superfund sites, this
project was entirely within a residential area which meant that a great
deal of weight had to be placed on resident concerns. The evaluation
conducted during the project's feasibility study very seriously consi-
dered these concerns and regarded meeting them as an integral aspect
of successful implementation. For this reason, many technically sound
alternatives such as capping in-place or excavation and disposal in an
on-site landfill were readily dismissed.
The final alternative chosen for implementation at the site called for
excavating the contaminated soil and destroying the material in a mobile
incinerator which would be temporarily located on the LEU property.
In addition, the selected alternative included a thorough cleaning of
all affected homes after excavation of the contaminated soil. Although
this alternative best met the technical, financial and protective criteria
established during the review, it was not without numerous potential
barriers to implementation such as community acceptance of an on-
site incinerator and residential consent for entry onto and into the
affected property and homes.
SITE REMEDIATION 447
-------
COMMUNITY RELATIONS OBJECTIVES
In order to successfully implement the selected alternative, a thorough
and comprehensive community relations plan was necessary. The
primary objective of the plan was to aid project implementation through
extensive communication. This meant sharing the details of the selected
alternative with the residents, knowledgeably and patiently answering
individual's questions to ease their fears and ultimately obtaining the
full consent of the neighborhood so that the remedy could be imple-
mented. Without the full participation and cooperation of the residents.
the selected alternative could not be successfully implemented.
IMPLEMENTATION PLANNING
For the remediation project to be a success, all governmental agen-
cies as well as all elected officials had to work jointly to reach the project
objective. Government agencies included both the U.S.EPA and the
IEPA, the City of LaSalle, the Illinois Department of Transportation
and the Illinois Department of Public Health. Other interests included
the prime contractor Westinghouse/ HAZTECH. the IKPA oversight
contractor—Ecology & Environment and the local labor unions.
Before beginning any site work and any on-sile community relations
activities, the primary contacts with both the U.S.EPA and the IEPA
met on numerous occasions in both Springfield and Chicago to evaluate
the community's needs and concerns and to plan activities that could
be undertaken to address them. In addition, in order to optimize the
evaluation and planning efforts, all primary contacts at both the U.S.
EPA and the IEPA participated in 3 days of extensive community rela-
tions training in Dallas. Texas. That training focused on how to gain
consent in difficult situations when a common concern is the objec-
tive. In this case, the common concern of the project was eliminating
the PCB contamination.
Through the course of the planning meetings, countless problem areas
and needs were identified. These included basic items such as the need
for equal treatment of all homes and security in the area during the
actual work, as well as more complicated matters such as maintaining
access to the residential community while at the same lime adequately
protecting the health and safety of the residents and the workers.
Since it was necessary to temporarily relocate residents during (he
soil excavation and cleaning of the homes, the list of problems that could
be encountered during implementation multiplied rapidly. During the
planning meetings that were held, a great deal of time and effort went
into brainstorming and generally trying to determine what the needs
of the residents \wxild be when the actual work commenced. The project
coordinators in essence tried to picture themselves in the residents'
position. Through this role reversal, the coordinators were able to
determine what they would be concerned with and what they would
like to see happen during the relocation and remediation period if they
themselves were the homeowners.
As a result of the brainstorming sessions, many items which initially
might have seemed trivial were identified as being issues that could
ultimately determine the success or failure of the implementation. For
example, the contractors' work schedules during the cleanup had to be
carefully coordinated so that residents could be given adequate notice
when cleanup activities would commence at their homes and so that
the costs and the inconveniences associated with the relocation could
be minimized. Arrangements had to be made at nearby hotels to ac-
count for the special needs of those who were being temporarily moved
out of their homes. Special arrangements had to be made for transpor-
tation of children to and from school, of the elderly or sick to and from
doctors and of the residents who normally did not depend on cars to
get around the area. Normal day-to-day activities, such as bringing in
the daily newspaper or mail, which are not generally given much thought
also had to be addressed.
Ultimately, the issue that was the basis of the CR planning activities
was the temporary separation of the residents from familiar settings
and routines. The residents were being asked to hand over their homes
and their possessions, to forgo their day-to-day security and to trust
people that they had only recently met to take care of a part of their
lives. This traumatic interruption in the lives of area residents required
special consideration and respect during both the planning and im-
plementation.
The demographic makeup of the community was widely varied and
as a result, no simple plans could fit all the residents involved. Even
though meals were provided at the hotel, some elderly and infants had
special dietary needs that had to be addressed. Some people also had
pets that had to be relocated along with their owners. As part of the
initial security plan, no residents were to be allowed back into their
homes until all work was completed. However, since many of the resi-
dents had extensive collections of plants that needed special care and
since some people ran small businesses out of their homes, arrange-
ments had to be made lo allow for daily entry into the homes without
breaching the security that had been established.
Since the residents would be vacating their homes for extended periods
of time and allowing virtual strangers into them to clean the structures
while they were away, liability was a primary concern of both the resi-
dents and the contractors. Plans had to be drawn up to protect the resi-
dents' property while at the same time limiting the contractors' potential
liability.
IMPLEMENTATION
Once most of the problems and their solutions were determined, the
mam issue was implementation of the plans and education of die
community. Without the complete understanding, cooperation and
consent of the community, (here was no project.
The first step in the implementation process was to establish personal
contact with all the residents who would be affected. This was accom-
plished initially through door-to-door visits to each home in the area.
After sending out notices to all (he homeowners, groups of two or three
representatives from the IEPA and the U.S. EPA sat down with the resi-
dents in their homes and discussed the proposed project with the people
on a one-on-one basis. This informal and small atmosphere allowed
the residents to get to know the agency representatives, to learn about
the project in the relaxed setting of their own homes and to feel
comfortable asking questions and conveying concerns.
The pcrson-to-person contact was a visible sign that the bureaucrats
planning the project, a project that vrould totally upset resident's daily
living, had enough concern to take the time to talk. It allowed the resi-
dents the opportunity to meet the bureaucrats, to get to know them as
real people and concerned planners and to judge both the people and
the project for themselves. The personalized discussions gave the resi-
dents an option to decide whether or not (hey (bought the planners were
confident and knowledgeable and whether or not these virtual strangers
could be entrusted with their homes and their well-being.
In addition to the in-home meetings, large community meetings were
held at various times to begin specific aspects of the project and to allow
the residents to meet as a group and see that their neighbors had simi-
lar concerns. Prior to these meetings, information or fact sheets and
letters were prepared (o briefly summarize key aspects that were
important and to reinforce information that was passed on during the
meetings.
In order to successfully implement the project, both the residents
and the local city and county officials had to concur. Therefore, in
addition to formal and informal meetings with the residents, planned
meetings were also held with the local government officials.
Another tool used to successfully aid implementation was a survey
of the residents' needs during the project. Prior to (he actual start of
the remediation work, the IEPA distributed a detailed questionnaire to
all the residents with a cordial cover letter explaining that its purpose
was (o gather specific information from each home so that arrange-
ments could be made in advance of the relocation, thus minimizing
the degree of inconvenience. The survey asked the people what needs
had to be met in order to make them feel (he most comfortable with
the move. Additionally, the letter stated that all information collected
would be treated as confidential and would be used solely for the pur-
pose of implementing the project.
Once the project was actually underway, the needs of the residents
became much more important. It was decided in advance that a com-
448 SITE REMEDIATION
-------
munity relations contact person from the IEPA would actually stay with
the residents in the hotel throughout the course of the project. This
person acted as both a counselor and a concierge who answered all
questions and concerns and also made necessary arrangements for any
specialized needs. The decision to station a CR contact in the hotel
during the relocation period proved to be one of the most valuable
decisions made during the planning and implementation processes. Not
only did this decision act as a bridge and continuation of the initial
contacts made, but it also physically showed the residents the high level
of dedication to their needs.
VIDEO DOCUMENTATION
The LaSalle project was unusual because it directly involved a residen-
tial community. In addition, the proposed solution to the problem
involved not only direct intrusion and interruption of the day-to-day
lives of the area residents, but it also involved the temporary place-
ment and operation of a hazardous waste incinerator directly adjacent
to the community, a difficult proposal in itself to implement. The project
also was unusual because of its high degree of success. For all these
reasons, the U.S. EPA project coordinators felt that this project was
ideally suited for documentation.
The medium selected to capture the project was a videotape documen-
tary. The 22-min tape chronicles the project through the eyes of the
affected people. It gives the residents a chance to say what they felt
about the project and it records a successful implementation from their
standpoint. The documentary also allows the planners to step out of
their decision-making role to see how their ideas were accepted and
integrated. This is accomplished through community feedback which
shows, from the residents' perspective, what their fears and concerns
were and how these were addressed.
While the documentary focuses only on the LaSalle project, it can
be used as an example for all Superfund projects to illustrate the neces-
sity for thorough planning prior to the implementation of any remedia-
tion project.
The documentary was produced by the U.S. EPA, but it could not
have been completed without the help and cooperation of all the people
involved in the project, including the cleanup contractors, the local offi-
cials, the IEPA project planners who provided the lead role in implemen-
tation and who also provided copies of video footage from the lEPA's
own tape library and, most importantly, the people of the City of LaSalle.
CONCLUSION
This project was successful because it was carefully thought out and
implemented. The feet that all of the affected residents gave their consent
for the work to be done on their property attests to this fact. While
no two Superfund projects are alike, the fact of the matter is that regard-
less of the project and the associated degree of impact that the project
may have on the nearby community, any project has a higher chance
of successful implementation if careful pre-planning is undertaken and
if the needs and concerns of the community are adequately addressed.
REFERENCES
1. Black & Veatch, Final Report, Phased Feasibility Study For Remediation of
PCB Contamination of the LaSalle Electrical Utilities Site, Prepared for the
Illinois Environmental Protection Agency, Aug. 13, 1986.
2. Remedial Action Selection and Record of Decision for the LaSalle Electrical
Utilities Site, U.S. EPA, Aug. 19, 1986.
Appendix I
Community Relations Responsiveness Summary
August 19, 1986
*"*N Illinois -Environmental Protection Agency 2200 Churthill Road Springfield ILtiJT
^?
August 1986
COMMUNITY RELATIONS RESPONSIVENESS SUMMARY
ELECTRICAL UTILITIES COMPANY
LASALLE, ILLINOIS
The Illinois Environmental Protection Agency (IEPA) conducted the
cormunity relations program at this site. Community relations activities
continued throughout the remedial investigation and feasibility study.
During the phased feasibility study, a three week public comment period
(July 8 — July 29) was established to receive public comment about remedies
for managing contamination found in residential areas. A public hearing was
held on July 17 to discuss these remedies. This responsiveness summary
documents citizen concerns expressed during the comment period and lEPA's
response to those concerns.
Another public hearing and public comment period will be held to discuss
remedies for managing contamination found on the Electrical Utilities
property. The additional hearing and comment period will be held after the
feasibility study for the EUC property is complete. A separate responsiveness
summary will be prepared and distributed following that coranent period.
Introduction
Polychlorinated biphenyls (PCBs), used in-the manufacture of electric
capacitors, are present in the soil of a portion of the residential area east
of the EUC plant, in the commercial property south of the plant, and in a
small portion of the farm field to the north. In addition, PCBs are in the
soil north and south along St. Vincents Road.
Five remedies are proposed for managing this PCB contaminated soil:
landfill; on-site incineration; off-site incineration (outside LaSalle
County); temporary storage; and no-action.
Community sentiment is virtually unanimous in support of the remedy
preferred by IEPA—mobile incineration. A question expressed by several in
the community, including city officials, regards the level of noise that will
be generated when the incinerator is in operation. Residents are not opposed
to off-site incineration, but are skeptical that USEPA would approve this
remedy because of the high cost.
Two of the other remedies, landfill and "no action," received no support
and would not be accepted by the community judging from verbal comments from
residents during the remedial investigation and feasibility study.
Temporary storage was not supported by the community. However, this
remedy might be acceptable if storage did not exceed six months. Primary
coimujnity concerns are for residential property values and for attracting a
new business to the EUC site. Landfill and "no-action" conflict with these
comnunity concerns.
Community Involvement
Community relations commenced with a joint presentation by IEPA and USEPA
officials at a City Council meeting in January, 1984. Through personal
interviews, "living-room meetings," and public meetings, the following issues
were identified as concerns of the citizens during the remedial investigation.
Effect on business—Several small businesses are located in the immediate
vicinity of the EUC site. Business owners are concerned about how present and
potential customers are reacting to the news that PCB contamination exists in
the area.
One businessman has been refused liability insurance. Insurance companies
are citing PCB contamination and underground storage tanks as the reason.
Property values - According to residents, residential property values have
diminished in one area near the EUC site. Residents feel that once cleanup at
the EUC site is completed, property values will- increase. Removal of
approximately 260 fifty-five gallon drums containing PCB and drainage of a
tank containing trichloroethylene in February, 1986, did not affect property
values according to residents.
SITE REMEDIATION 449
-------
H«llth efftcll—PCB contamination In residential ylrdl raised questions
about potential health effect! to both prestnt and future generations
Offtcllll fro. the Illinois Department of Public Heilth ind tie IEPA -*nt
door-to-door when PCB level i Mre Identified to discuss the lapict of th* PCBs
tn resldentnl yirds.
During th* public cement period, t public heirlng Ml held to discuss
eich of the proposed reaedies. The heiHng MS held on July 17. at Us* Howard
Johnson Motor lodge, Bout* SO » 51, In LiSilli beginning it ' 00 p •. A
•rllt«n stiteaent MS prel«nt*d fro" the City of LlSlIU Approxlaiuly 15 of
the 35 In ittendance liked questions. Beginning two wtexs before this
hearing. jl« saill group wetlngs Mre Held .ith residents, «i«ted offlcll't,
ind tne news aedli to discuss specific questions about V* proposed remedies
Suroary of Contents and UP* Response
Issue: Superfund Progria
QUCSTIOH: Are tne residential /ards considered pirt of the (UC nu luted
on tne national Priority list'
RESPONSE. Ves
OUCSTIOH: who piys for th* rmedies iaplea*nt*d is part of cue Cut c'einup'
RESPONSE: Th« federal government »ner vho his significant concentritlons o' K.t
•nil be offered yird excavation and Inumil nousecleaning is in
option. 1E?A cncourigts ill iffect*d hoevovnert to Ukt
advintige of Uiis offer.
QUESTION WMt is the airiest concentntlon of PCI thit *i H be r«aoved
fro» residential yirds?
«ŁS?0«St The I EPA and USCPA considered four different concentrltlons
Etcl concentration Is Misured in parts per •Illlon. These
concentrltlons ire SO, 25, 10 and 5 ppa. PCB concentrltlons
that equil or exceed S pod will be excavated.
QUESTION Hot* euny tons of contieilnlted soil fill be reaoved frora
resldentlll jrlrds?
BC70NSE Appmileitcty 29,000 cubic yirds lequtvilent to 36,000 tons! «i II
be excivited.
QUESTION- Do my hoaes south of Z3rd Street hive excessive concentrltlons
of >CI
RESPONSE HO
QUESTION: Hox did PCI contailnitlon rticK residential yIras'
RESPONSE: PC8s were probably trinsporwd on th* bottoa of shoes Morn by
EUC employees and on tires of vehicles leaving th* EUC
property. Sow PC8s Merc carried by the «inds froei th* EUC
pirtlng lot HoMver, vrlnd deposition alone does not account
for the concentrations found In resldentlll yirds. Th* IEPA Is
not ure ho> ill the PCis reiched residenttil yirds. If oil MS
spriyed on roids. this uy n\* contributed to th* proble*.
QUESTION: To vhit depths n*re r*s1d*nt1*1 yirds SMpled to det*r*atn* th*
extent of cont*elnition7
RESPONSE: Five feet, but virtually ill msurible concentntloni of Kis
v«re found In the tap 10 Inches of toll.
QUESTION HO* Mr* residents first notified atwut th* rtultl of PCB
sibling In yirds?
RESPONSE Off Kills '
IEM *nt
tn* results of PCI
In; In yirds?
Ills fro* th* Illinois 0«pirt»*nt of Public H.ilth did tncMoroetnylen* ITCtl g*t Inu th* jroundnlter?
RESPONSE TCE -at either spilled or du*p*d onto tne ground «hen the EUC
sit* MS In operitlon. A tar* on tn* CDC property contained ICE
•Men MS drlined md Muled of'-SIU In feOrulry. 1586, U
prevent idditloMl quintitfes of fCC froa reaching groundwittr
Out STI ON If Our drilling MUr la'l*
XSPONU fes tlrtullly ill th* residents in tne vicinity of the EUC
sit* hive hookups to th* itSitt* puoi'c otter supply uklck is
not emunoered by TCE or PCI froa this sit*. « survey conducted
by uw UP* lo*ntifl*d ) Mils at EdMrds Street still us«d for
dr lot Ing «lt*r, ntes* Mils, and my other vtlls «i»1n 1/2
•n* of th« OK sit*. »lll be stapled. »*sidents '*v1ng uftnln
1/2 ail* of the CUC sit* «*c still use tn*lr pnvite Mil ire
urged to contact IM IEPA.
QUtSTION If * **«o*r Mn is drilled. Mvl« it be sife froai TCE
contailnitlonf
RESPONSE: A grvMKd«*t*r study Is Ming conducted by Heck I VeiUk. When
th* study Is coaplet* •• should kno» if de*p*r aquifers art
protected or connected to sMHo» graunoMter in Oils v1ci«ltj
Reaalnlng Concerns
IEPA intlclpaus • viriety of «testlo«s end concerns to iHs* daring yard
excavation. Soat of these questions inn concerns «uy be resolved before
excavation begins, for Motale. i pirtlcaUr bush a*y kir* santlawiul nlue
or speclil ippeil, therefore, the IC*A should caaaunicite this concern to the
contractor so chit plus cat be a»d« ID protect this bush during excintlon.
In in if fort to Identify thes* special concerns, th* I ETA Is pi inning I
door.u-door aeeting «iw iffected property oMiers during the design phis*
Soa* questions ind concems -nil iris* unexpectedly during *«civit1on
These include access to hoaes. wither delays, transporutlon routes, end
accoaaodations. N*M iclon a*y wtend beyond
LaSil'» County, md is likely to involve video-up ing for television stitlons
•s »eil is still photognphs for neospepers. A news conference will be
pllnned for the stlrt of ucivltlon. During tfi* excivctson. coaauni ty
relations stiff -ill be tn liSilie to assist residents >nOi special needs A
news release and oe-sonai letter «/i 11 be distributed which provides the
location ind telephone nuaber .her* IEPA stiff cin be reiched In laSall?
Growndirtter w|i| continue to be siajpled to determine the extent of
contadnition. The results of this additional wort will be available it the
laSalle Courthouse after inilyses of groundMter saapltng Is coaplet*. ind
-ill b* discussed it the n«xt public heirlng ibout rexaedles for on-site
contmlnitlon.
450 SITE REMEDIATION
-------
Appendix II
Household Relocation Survey
Household Relocation Survey
US EPA iiul It? A
LiSalle Project
If Staying Special
Eltevhert Dietary
Vttertl MeeJiT
Aftf Oth
Spec) 1 1 M
ROOM ir* ifilliblt with 1 double btd. t double bcdt, 1 king bed.)
Do /M haw «nr Ptti t(ut need McaeodttloMt If 10 how mtry, ind «ti*t type. If applicable have they K*d their
Appendix III
Consent Form For Entry and Construction
llmois Environmental Proitction Agency P.o Box 19276. Springfield, IL 6279-J-9276
ZONE A
CONSENT FOR ENTRY AND CONSTRUCTION
The undersigned property owner ("property owner") is the owner of the property,
including a residence, ("property") commonly known as .
(address)
LaSalle, Illinois.
A. The property owner, in consideration of having soil contaminated with
polychlorinated biphenyls ("PCBs"} removed from his property, does
hereby authorize and consent to the Illinois Environmental Protection
Agency ("IEPA") and the United States. Environmental Protection Agency
("USEPA") and their respective representatives, employees, agents
and contractors'to enter upon the property and into the residence
described above to conduct the following activities:
1. Excavation and removal of soils contaminated with PCBs from the
front, sides and back yard of the property, if necessary.
2. Cleaning of the inside and outside of garages and the inside
and outside of the residence, including commercial removal,
cleaning and return of draperies to the residence.
3. Placement of clean soils in excavated areas and restoration of
vegetation.
4. Soil sampling prior to, during, and after the above described
work.
8. The IEPA agrees that the following activities will be undertaken through
its duly authorized contractors and the property owner recognizes
and acknowledges the following activities in connection with the above
described work:
1. Soils located at the front, sides and back yard of the. property
which are contaminated with PCBs in the amount of 5 parts per
million ("ppm") or more within the top 12 inches of soil and
10 ppm or more below the top 12 inches of soil shall be removed
by excavation and replaced with clean soils, provided that the
Illinois General Assembly and federal funding source shall appro-
priate or otherwise make available funds sufficient for the IEPA
to undertake such activity.
2. Residential lawns shall be replaced with sod.
3. Shrubs and ornamental vegetation shall be removed and replaced
with average size nursery stock of like kind or quality.
4. Trees under 6" in trunk diameter measured 6" above the soil line
shall be removed and replaced with nursery stock of like kind
or quality having a trunk with a diameter of between 2" and 4"
measured 6" above the soil line.
5. Trees of trunk diameter over 6" shall not be removed wherever
possible and care will be taken to minimize any adverse impact
or damage to them. If replacement is necessary, the tree shall
be replaced with nursery stock of like kind or quality having
a trunk diameter of between 2" and 4" measured 6" above the soil
1 ine.
6. Permanent structures, such as driveways, sidewalks, steps, and
patios, shall, wherever feasible, not be disturbed or removed.
If removal is necessary, the replacement shall be made of like
kind or quality.
7. Fencing or similar items removed shall be reinstalled, wherever
feasible, or replaced with comparable items of like kind or quality.
8. In the event the IEPA determines it necessary that for health
and safety or logistical reasons that the property owner vacate
the property, temporary accomodations and meals shall be provided
for the property owner and the household members at the expense
of USEPA/IEPA while the work described in paragraph A above is
occurring at the property owner's property.
9. The property owner shall be notified seven to ten days prior
to the anticipated date that excavation of the property will
begin and relocation to a motel will be necessary.
C. The property owner agrees that any claims which arise against the
IEPA or other agency or department of the State of Illinois, or its
respective officers, employees, and authorized representatives, or
against any contractors for the IEPA or other agency or department
of the State of Illinois shall be brought before the Illinois Court
of Claims pursuant to the Illinois Court of Claims Act (111. Revised
Statutes 1985, Ch. 35 Section 439 et seq., as amended).
D. If the property is occupied by a party or parties other than the legal
owner such as a tenant or contract for deed purchaser, please provide
signature of the tenant, contract purchaser, or other, in addition
to the property owner's signature.
E.
The undersigned property owner agrees that this consent shall become
effective from date of signature for a term of one year.
Dated this 5th day of November
1987.
ILLINOIS ENVIRONMENTAL
PROTECTION AGENCY
Federal -Site Management Unit
Remedial Project Management Section
Division of Land Pollution Control
Owner(s)
Signature(s)_
Home Address
Non-Owner
Resident(s)
Signature(i)
Appendix IV
Video Documentary
Copies Available Upon Request From:
Mary Ann Croce LaFaire
U.S. tPA Region
(312) 886-1728
1-800-621-8431
SITE REMEDIATION 451
-------
Communication Traps for Engineers
Melissa F. Shapiro
United States Navy
Washington, D.C.
George Hanley
U.S. Army Corps of Engineers
Kansas City, Missouri
INTRODUCTION
Engineers today are facing the complexities of hazardous waste
cleanup. The process is not simply a technical one. They are increasingly
aware that straight engineering, no matter how technologically advanced,
is not enough to win public confidence. What is needed, in addition
to traditional engineering skills, is an ability to communicate with the
public on an ongoing basis about the risk and uncertainties of hazardous
waste cleanup. There is a need to reach a consensus so that the remedy
proposed by these engineers will be accepted. This goal demands that
engineers be not only technological wizards, but also skilled commu-
nicators.
As we all know, engineers are not trained to be skilled communi-
cators. They are trained to be engineers. What happens, more often
than not. is that engineers find themselves walking into large, angry
public meetings and facing hostile media interviews; they are generally
confused and overwhelmed. Engineers have discovered it isn't enough
just to be an engineer, and unfortunately they are not quite sure what
to do about it!
In fact, if we look at the Superfund program, whether at U.S. EPA
or at federal facility sites, we can see a emerging pattern of certain "com-
munication traps" that engineers fall into time and time again. These
"traps', which form the substance of this paper, in no way imply that
engineers are not capable of being good communicators. Rather, this
paper draws on the observations of community relations experts and
seeks to illustrate these traditional communication pitfalls, briefly dis-
cuss them and, by pointing them out. make engineers aware of them
so that they can develop strategies for coping with them in their site
remediation plans.
COMMUNICATIONS TRAPS
Communication Trap n
"Superfund/Installation Restoration is an engineering problem and
will be solved through engineering."
One of the painful realizations long-term Superfund Remedial Project
Managers have made is that factors impact on cleanup that have nothing
to do with engineering. Factors such as concerns about health, property
values, fear of the unknown (which comes into play when a new cleanup
technology is proposed), and even fears about loss of control over aspects
of community life and decision-making have, at one time or another.
all been issues at Superfund sites. A technical solution may not address
any of these factors. Yet each one of them, individually or collectively,
can slow down or stop a cleanup.
An engineer's basic task is to study a problem, make a recommen-
dation and take some remedial action. This task orientation by many
engineers excludes as non-essential anything that is outside the realm
of engineering. The phrase that comes to mind is".. .artists fall in love
with their models, engineers fall in love with their projects.''
An engineer, then, must know about these issues and take them into
account when he or she communicates. To ignore these issues is to
jeopardize the project and invite failure.
Communication Trap tfl
"Those people who don't like my technically sound solution are
extremists and I don't have to deal with them."
Almost nothing could be further from the truth. You do have to deal
with them! Today in America, we have something called an environ-
mental ethic. What this means is that everyone, regardless of back-
ground, income, education or politics, considers himself an
environmentalist. Hazardous waste is seen as a personal health and safely
issue and. therefore, have the highest priority to everyone.
Some pollsters have even gone so far as to suggest that the environ-
mental movement of the 1990s will be just like the civil rights move-
ment of the 1960s. What this means for engineers is that there is a much
broader-based constituency of those people calling themselves environ-
mentalists, and that same constituency represents a cross-section of the
American public, not a radical fringe group. To ignore those interests
in or in disagreement »ith the proposed alternative, may very well be
to ignore the heart and soul of the public, the very group needed to
build a consensus for remedy acceptance.
Communication Trap <3
"Everybody knows how somebody makes a reasonable, intelligent
decision. They base it on facts and data, and they weigh risks in a scien-
tific manner."
By virtue of the way engineers are trained, many of us have often
make the erroneous assumption that with the right information, and
enough of the right information, people will make the right decision.
Unfortunately, what has happened, time and time again, is that getting
the information out is only half the story. We have to find out what
happens to the information once it is out there. We cannot relax just
because we have inundated the public with facts and figures.
What comes to mind is the old joke about the child who asks his
parents where he came from. They go into a long discussion of human
reproduction. At the end the child says—"That's fine, but Johnny is from
Cleveland, where did 1 come from?"
People make decisions differently and they use facts differently. We
cannot be complacent just because we put out lots of information. Ws
also cannot assume that people will make decisions the way we want
them to or the way engineers do.
452 SITE REMEDIATION
-------
Communication Trap #4
"I am the expert. Therefore my judgment will be accepted."
Unfortunately, many engineers approach the public this way. What
always comes as a surprise is that expertise is not enough and that sound
engineering judgment will not necessarily be accepted. We have seen,
many times, that the risk assessments have very little to do with the
level of public involvement. As engineers, we often have told commu-
nities about what we perceive to be very minor threats to public health
or the environment—only to be besieged by calls from the news media
and from congressional offices asking if the rumors are true about
evacuation. At the same time, we have told communities that a very
real threat to public health does exist and we have been unable to arose
enough interest to get people to stop using their contaminated water
supply. Our engineering credentials will not necessarily be validated
by community support. We cannot expect background and training to
carry the battle and make our message understood and accepted.
Communication Trap #5
"Remedial action must be explained accurately and in great detail.
We cannot afford to simplify our explanations without sacrificing the
quality of our response."
This communication trap is referred to as the "Watchbulding Syn-
drome.'7 Ask any engineer what the time is, and he will tell you how
to build a watch! While it might be an oversimplification, it certainly
is true of many members of the engineering profession.
There are many arguments between the technical staff and the com-
munity relations staff on this subject. This dilemma, more than any
other, illustrates the tensions between the technical experts and the
public. Unfortunately, what gets lost in the battle is the ability to
understand the issues. If it is true, as have been suggested, that over
half the American people do not know the difference between astronomy
and astrology, why do we fight among ourselves over who can be most
precise when the message is lost. If we explain difficult concepts in
a non-technical way, using everyday examples, we may not get high
grades from engineering professors, but we will get understanding and
later reach a consensus.
Engineers have to find a middle ground—between being technically
accurate and being understood. It might mean developing a whole new
vocabulary for explaining some of these issues, a vocabulary that is
very different from the one we have been using. Engineers do
communicate well with each other, using the lingo of the trade, but
the public is not part of the trade. Those who may have served as combat
engineers know how far lingo can go. What they know as an E-tool,
or entrenching tool, is commonly known as a shovel.
Communication Trap #6
"The public wants lots of information, so I'll tell them everything,"
or—Til only tell them what I think they need to know."
Either one of these approaches is fraught with problems. The major
flaw is that this approach assumes what the public's need for informa-
tion is without validation. It is as though someone asks for a briefing
on the state of the world and the briefing is given without any clarifi-
cation or without asking any questions to ascertain what kind of infor-
mation is really wanted or needed.
The briefing is almost certain to miss the mark, either by being much
too detailed or by leaving out important pieces of information. Com-
munity relations at the outset of a study can be very helpful. The infor-
mation gathered from community interviews in developing the
community relations plan can be invaluable for determining future com-
munity concerns. Small group meetings with state and local elected
officials and with citizens can help those charged with providing infor-
mation deliver a message that is relevant and effective.
Communication Trap #7
"The public must know all the facts, that is the only way to deal with
the bottom line."
There is nothing more deadly for an engineer than to assume that
those listening to a presentation on hazardous waste cleanup want to
hear how one got from point A to point B to pint C to point D to the
bottom line. Too often the assumption is made that the public wants
to hear the various options for goundwater treatment and recharge, when
the central issue is—'Can I drink the water?"
This is another instance of the public wanting to know answers to
certain basic questions and the engineer taking that interest as a request
for highly detailed information. The problem again is that the engineer
is projecting his information needs and assuming that they are the
information needs of the public without validation. Again, unless it is
checked out, the engineer is taking an unnecessary gamble, with the
cards stacked against him. The engineer will probably lose not only
the audience, but also any chances of reaching a consensus on the
remedy.
Communication Trap #8
"Talking to the media about hazardous waste is just like talking to
anyone else."
We cannot avoid the media during hazardous waste cleanup. Nor
should we. We can, however, commit ourselves to making the media
equal partners, and we can scrutinize very carefully how we commu-
nicate with them so that we make the most of media opportunities. They
have a need to report information, and we have a need to reach the
public. It is a symbiotic relationship; we are interdependent on each
other.
With camcorders, as inexpensive as they are, we can practice press
conferences by asking our colleagues to pose very difficult questions
and hone your answers into short, sound bytes.
We can call on our community relations staff to make us go through
"dress rehearsals" so that we as well as our Public Affairs staff are
able to explain complex technical procedures. We also can be proactive
with the media, contacting them weeks before the first drill rig or
sampling crew arrives on the site. We also can contact them when we
do not have "hard news" but set aside time to explain the basics of
the Superfund process. Even the most hard-bitten New York Times
reporter is willing to listen when the word "Superfund" is mentioned.
Surprisingly, an informed media community will report more accurately
on proposed cleanup actions. Good news or bad news, reporters need
news. If you can give them a story, you have fulfilled their need.
Communications Trap #9
"Communicating risk is just like communicating anything else."
As stated earlier, one of the ironies of being a better engineer in
hazardous waste cleanup involves a lot of non-engineer skills. One of
these skills is the ability to communicate risk. But, before you com-
municate risk, you must communicate.
Where many engineers confront their first stumbling block is at the
large public meeting. For the first time, they are meeting with the com-
munity and not only releasing the results of a multi-million dollar study,
but also asking for the public's confidence in the results of that study.
In 30 min., an audience of 50 or a 100 people is asked to make a "leap
of faith" and wholly accept what they are being told as gospel truth.
It is unfair, not only to the public, but also to the engineer. All the time
and money spent on research and remedy selection go down die drain
if engineers ask for public confidence when they have no relationship
with the community.
Again, being proactive makes sense. Imagine how much easier it
would be to make a presentation to a community whose concerns are
known, with whom we have met often, whose profile we understand,
and where we have had the opportunity to practice making our
presentation both relevant and appropriate.
Communications Trap #10
"Everyone involved in this process—engineers, lawyers, the public,—
has shared goals."
Strange as it may seem, hazardous waste cleanup sites present a differ-
ent set of opportunities to everyone. While an engineer views cleanup
as an opportunity for study and problem-solving, others may see it as:
an issue around which to organize a community, an anti-growth initia-
tive, a serious threat to children's health, a reason to close a military
base, or perhaps even an opportunity for posturing various political
SITE REMEDIATION 453
-------
positions. The engineer needs to know that all these groups with different
interests are going to have different expectations about (he process, and
they are going to try to use the cleanup process for different ends.
Many of these players can be identified in developing the communi-
ty relations plan, and the earlier they can be singled out the better. If
an engineer has to communicate information to the public and is not
aware of the different actors, the presentation probably will not be suc-
cessful .
CONCLUSION
Engineers, by virtue of their training as engineers, have to recognize
that very human factors come into play on hazardous waste cleanup
projects that have nothing to do with their engineering expertise. As
project managers, they have to recognize that technical expertise is not
enough and that their stature as engineers does not buy them any extra
credibility or acceptance.
Engineers also need to remember that all communities are different,
and knowing (he community will be an important part of communi-
cating information effectively. Finally, becoming aware of false assump-
tions that are often made is the first step in developing communication
strategies for coping with site remediation.
454 SITE REMEDIATION
-------
Study, Design and Construction of an
On-Site Recoverable Storage Facility
Edward Patrick Hagarty, RE.
Robert M. Gruninger, RE.
C.C. Johnson Malhotra, P.C.
Silver Spring, Maryland
Mann A. Patel, RE.
City of Baltimore
Baltimore, Maryland
ABSTRACT
The City of Baltimore acquired a site for construction of a wastewater
treatment fecility along the Patapsco River, a tributary of the Chesapeake
Bay, in 1924. Since then, the Patapsco Wastewater Treatment Plant has
become a regional, publicly-owned treatment works (POTW) serving
the City and parts of three adjacent counties. Prior to the City's acqui-
sition of the property, land along the shoreline of Baltimore's Inner
Harbor, including part of the Patapsco site, had been filled with chro-
mium ore tailings from a nearby refinery.
Health effects from exposure to chromium were not known at the
time the original plant was constructed. However, new information is
available regarding the health effects of exposure to chromium. When
the regional fecility required expansion, the City was faced with a severe
problem. Special handling would be needed to excavate in areas of
chromium-contaminated soil to protect the environment and the health
and safety of construction workers, plant employees and nearby
residents.
Expansion of the POTW included the replacement of obsolete primary
settling tanks, the installation of additional biological treatment reactors
and clarifiers, additional chlorination capacity and new dechlorination
facilities. This expansion could not be postponed since the City was
required to add these facilities before any more users could be added
to the regional system.
The City and its consultant, C.C. Johnson and Malhotra P.C. (CCJM),
prepared a plan through which the contaminated soil could be safely
removed from the areas of proposed construction and stored in an on-
site, recoverable storage facility. State regulatory authorities were con-
sulted to determine the acceptability of placing the excavated soil in
recoverable storage on-site. Such a plan permitted construction of the
POTW improvements to begin without requiring an immediate deci-
sion on the treatment or permanent disposal of the contaminated soil.
After obtaining approval from the State, the remedial investigation,
including the preparation of the report, was completed in less than 6 mo.
Additional data were collected by drilling boreholes, installing wells
and collecting and analyzing soil and groundwater samples. The results
of the study were used to prepare the plans and specifications for con-
struction of an on[site recoverable storage facility area, excavation and
handling of contaminated soil and a pretreatment plant to reduce the
chromium concentration of groundwater and water from decontami-
nation.
The project, including project planning, the remedial investigation
and design were completed for bidding in only 15 mo. Construction
of the on-site recoverable storage facility and storage of the contami-
nated soil from construction of the primary settling tanks was completed
in August, 1989.
INTRODUCTION
Site History
In the late 1800s and early 1900s, residue from chromium ore refining
operations was used as fill along some of the shoreline of Baltimore's
Inner Harbor. In 1924, the City of Baltimore acquired a 65-ac site for
construction of a wastewater treatment facility along Baltimore's Inner
Harbor bordering the Patapsco River, a tributary to the Chesapeake
Bay. The site was located in an area that had fill material consisting
of chromium ore tailings. Nevertheless, the wastewater treatment fecility
was constructed and began serving the residents of the City.
Since its initial operation, the treatment facility has been expanded
to become a 70 mgd publicly-owned treatment works (POTW) serving
the City and parts of three adjacent counties. Throughout the initial
construction and subsequent expansions, no concern was expressed
about the POTW being constructed within an area containing chromium.
This oversight was primarily due to a lack of scientific knowledge
regarding potential hazardous characteristics of chromium and a lack
of regulation of such material.
In 1984, it was determined that the POTW would require expansion
from 70 mgd to 87.5 mgd to meet the anticipated population projec-
tions and to provide a higher level of treatment. To meet these needs,
the City prepared plans and specifications to replace obsolete primary
settling tanks, add new biological treatment reactors, provide additional
chlorination capacity and add new dechlorination facilities. After the
design for most of these facilities was complete, the prospect of ex-
cavating within an area contaminated with chromium ore tailings forcedi
the City to determine a course of action which would allow the neces-
sary construction to proceed without subjecting construction person-
nel, plant employees and the nearby public to high levels of chromium.
After discussion with the State of Maryland Department of the Environ-
ment (MDE), the City determined that the best solution was to con-
struct an on-site, recoverable storage facility. This option was selected
over other alternatives considered, such as on-site treatment of soils
and off-site disposal, on the basis of cost and the potential that the fill
material might be exempt from hazardous waste handling requirements
because it is a waste from the processing of chromium ore. An on-site,
recoverable storage facility allowed the City to construct the necessary
POTW expansion while retaining the flexibility to properly handle the
chromium-contaminated soil. The on-site storage facility could accept
chromium-contaminated fill incrementally during different phases of
the plant expansion and avoid the costlier on-site treatment until a long-
term management plan for the material could be determined.
Previous Investigations
Prior to 1986, no investigations had been conducted to determine the
SITE REMEDIATION 455
-------
nature and extent of contamination al the site. As part of the geotechni-
cal investigation for the POTW expansion, the City had samples
collected for analysis. The geotechnical consultant collected numerous
surface and subsurface soil samples and groundwater samples'.
Analyses included total chromium, hexavalent chromium and EP
Toxicity (for chromium only).
Results
Samples were collected throughout the entire 65-ac plant site and
showed a general pattern of chromium-contaminated soil within the old
bulkhead line and cleaner material located outside of this line. The old
bulkhead line was the limit of the site until 1976 when clean Till male-rial
was used to extend the shoreline to its current configuration d;ig. I),
Data review focused on the two area.s of proposed construe!ion: the
new primary settling tanks and the chlorination/dechlorination area
shown on Figure I. Chromium levels in the majority of the soil from
ground surface to a depth of 16 ft exceeded EP Toxicity Standards in
the area of the proposed primary settling lanks. Unfilterod groundwater
samples from this area had a maximum total chromium concentration
of ISO mg/L and a maximum hexavalent chromium concentration of
6.2 mg/L.
VFUTUHC PRIMARY
SCTTLING TANKS
I 1 I I HAUL KOAO—*^\. I I OH- «ITf~~». B) /
vj—' ^*s?sr.u/ ŁL.«
l^SLUOOt LI rtCIUTT / • ™*gf-
HOT TO SCAU
Figure I
Patapsco POTW Siie
The area of proposed additional chlorination and dechlorinaiion
facilities is located outside of the old bulkhead line and showed a dif-
ferent pattern of contamination. High concentrations of chromium in
this area are generally limited to the surficial soils. The only ground-
water sample collected from this area had a total chromium concentra-
tion of 130 mg/L and hexavalent chromium concentration of less than
O.I mg/L. The results of this investigation provided a good basis lor
gathering the additional information needed to design the excavation
of the contaminated soil and the on-site recoverable storage facility. The
additional site investigation, described in the following section, included
surficial and split spoon soil sampling and well construct ion for ground-
water sampling in the specific areas of proposed construction.
SITE INVESTIGATION
Investigation Objectives
This project was completed utilizing a standard engineering approach
to accomplish the project objectives. The overall objective of Ihc project
was to allow the construction of the POTW expansion to proceed as
quickly as possible in an environmentally safe manner. The standard
engineering approach, consisting of plan, design and construct, is similar
to that followed in the Supcrfund program.
The major phases of the project comparing the standard engineering
approach to the corresponding Supcrfund project phases are as follirws:
Project
Phase
Site
Investigation
Preliminary
Design
Design
Engineering
Approach
Plan
Plan
Design
Construction
Standard
Superfund
Ptuue
Remedial
Investigation
Feasibility
Study
Remedial
Design
Remedial
Action
Objective
Gather
Information
Select
Approach
Prepare
plans and
specifications
Construction Construction Remedial Implement
plans and
specifications
One major difference between the Patapsco POTW expansion project
and a Supcrfund project is the difference in objectives of the site in-
vestigation The usual objective of the Superfund remedial investiga-
tion is to fully characterize the nature and extent of contamination at
the cm ire site The site investigation for the Patapsco POTW expan-
sion had the following objectives:
• Confirmation that chromium was (he only contaminant of concern
• Determination of the extern of chromium contamination in the areas
of proposed construction
• Determination of groundwaicr contamination
Cleanup Criteria
A major problem of remedial actions at soil contamination sites is
the determination of cleanup criteria. Federal cleanup criteria exist
for drinking water sources; however, no such criteria exist for soils.
Some State regulatory agencies have developed site-specific cleanup
standards that are applied during real estate transactions of indus-
trial properties'. No such criteria were established for the Patapsco
POTW site. In lieu of such regulations, the State Hazardous and Solid
Waste Management Administration agreed with the City of Baltimore
to use the EP Toxicity Criterion for chromium as the guideline for
determining what soil had to be placed in the on-site, recoverable
storage facility.
The EP Toxicity test is one of the tests used to determine if a waste
material is considered hazardous under RCRA. Soil samples are
subjected to an Extraction Procedure using acetic acid intended to
simulate natural leaching in a landfill over a number of years. The
extract obtained is tested lor total chromium. The maximum allowable
value for non-hazardous material is 5.0 mg/L of total chromium in
the extract.
In addition to testing for EP Toxicity, a few selected soil samples
were analyzed for U.S. EPA's priority pollutants. This sampling and
analysis was performed to confirm that chromium was the only con-
taminant of concern and to establish minimum health and safety
requirements for the construction contractors. Groundwater samples
were collected in both areas of proposed construction and analyzed
for priority pollutants.
The City of Baltimore has an industrial pretreatment program which
has established maximum allowable contaminant concentrations for
industrial discharges to the POTW. These criteria were compared
to the results of the groundwater sampling to determine the need for
treatment prior to discharge to the plant influent.
Results
The results of the previous investigation were combined with the
results of the additional site investigation. The evaluation indicated
that chromium was the only contaminant of concern in either area
of proposed construction. In the area of the proposed primary settling
tanks, the majority of the soil to be removed during construction con-
tained chromium at levels exceeding the EP Tbxicity Standard of 5.0
mg/L. Therefore, all soil excavated from this area was recommended
for storage in the on-site, recoverable storage facility. Unfiltered
gmundwater samples from this area contained maximum concen-
trations of total and hexavalent chromium of 180 mg/L and 6.2 mg/L,
respectively.
456 SITE REMEDIATION
-------
The results of the investigation in the area of proposed construction
of the chlorination and dechlorination facilities differed from those
in the area of the new primary settling tanks. Chromium contamina-
tion in this area is primarily limited to soil at depths of less than
two ft. The only groundwater sample that showed any chromium con-
tamination had a total chromium concentration of 130 mg/L. This
sample was taken from an open boring and was probably high in
sediment content.
DESIGN PHASE
Following the site investigation, an on-site, recoverable storage facility
for contaminated soil was designed. During an earlier phase of
development of the site, a lagoon was constructed of native clay soils
having a permeability of 10* cm/sec. Various materials including
some chromium contaminated soil and some sludge were already
in the lagoon. The remaining capacity in the lagoon was estimated
to be sufficient to accept all of the contaminated soil associated with
the current expansion of the plant from 70 mgd to 87.5 mgd. It was
agreed that the existing lagoon should be lined, and when the soil
placement was complete, it would be covered with a high density
polyethylene (HDPE) material (Fig. 2). The bottom liner was
designed to be 80 mil thick and the top cover to be 40 mil thick.
-10 MIL HOPE
LINER CAP
EAN SOIL
W/ VEGETATION
NG
GROUND
NOT TO SCALE
DRAINAGE LAYER
SO MIL HDPE LINER
GEOTEXTILE
Figure 2
Section A-A'
Recoverable Storage Facility
The presence of contaminated material within the existing 4-ac lagoon
required that the design include construction staging so that half of
the lagoon was to be cleared of contaminated soil, the bottom sur-
face was to be shaped to provide sumps in which leachate could be
collected and the bottom liner was to be installed. When the first
half of the 80-mil liner was installed and tested, a synthetic drainage
mat and layer of soil was to be placed over the liner. This would
be covered with geotextile; then the contaminated soil in the lagoon
would be moved to the newly lined area. The liner construction
process was to be repeated for the second half of the lagoon.
Following the placement and testing of the liner and placement of
the drainage mat and soil, the contaminated soil from the new primary
settling tank construction could be safely moved to the on-site facility.
The new primary settling tank construction was to be done on a site
containing six old, low level tanks. These were a part of the original
plant construction and had been incorporated into the 1971 design
when the present plant was constructed. An Archimedes screw
pumping station was constructed in 1971 to lift the primary effluent
from the low level tanks to the biological treatment reactors. Because
the old tanks were no longer functional, they and the screw pumps
were to be removed under the current contract. Samples of the con-
crete from the walls and floor of the old tanks indicated no chromium
contamination in the concrete, so the plan was developed to brush
or scrape soil from the demolition debris and remove the concrete
debris from the Patapsco POTW site to a nearby rubble landfill. The
plan for demolition required removal of interior walls first, then
exterior walls and finally the bottom slab. In this way, exposure to
the contaminated soil was minimized.
Following the removal of the concrete rubble, soil excavation would
occur. A dedicated haul road between the excavation site and the
recoverable storage facility was constructed. The sites and the road-
way were fenced to deny access to the contaminated zone during the
removal of approximately 25,000 yd3 of contaminated soil. Including
the haul road in the contaminated zone also eliminated the need for
frequent equipment decontamination.
Dewatering of the site was necessary since the construction would
occur below the groundwater level. Based on pumping and sampling
the observation wells installed during design, it was determined that
7,000 gpd of chromium contaminated groundwater»would be pumped
and treated. A treatment system was designed to meet the total chro-
mium concentrations prescribed in the City's industrial pretreatment
ordinance. Additional water from equipment and personnel decon-
tamination and precipitation also required pretreatment. The total
capacity was specified to be 30,000 gpd. Because most of the
chromium contamination was associated with soil particles in the
water, a dissolved air flotation system with chemical addition was
designed and specified. Once treated to City standards, the effluent
from the portable pretreatment plant would be discharged to the
POTW influent for final treatment and discharge to the Patapsco
River.
The primary path for chromium contamination of the environment
and site personnel is airborne dust. The site health and safety plan
recognizes airborne dust as the major problem. Skin contact with
soil and contaminated water are also possible contaminant paths.
Therefore, the site health and safety plan required workers at the
site to keep the contaminated soil moist to minimize airborne parti-
cles. Also, the workers were required to wear washable cotton
coveralls, boots, gloves and hard hats. An emergency respirator was
part of each worker's standard gear. Dust monitors were specified
for the perimeter of the work zone with alarms set for the 5 ppm
dust action level. With the prescribed dust control procedures during
the soil excavation, haul and placement operations, alarms were not
expected to sound at this site.
CONSTRUCTION PHASE
Dust monitors were installed on the perimeter fence of the contami-
nated zone as specified. During the construction work, field inspec-
tors enforced the requirement that all exposed soil be kept moist.
Therefore, except for a single malfunction due to a spider, no dust
alarms were triggered during construction.
Several field changes were implemented during construction. At the
contractor's request, disposable Tyvek coveralls were used instead
of washable cotton coveralls provided the contractor assumed respon-
sibility for the disposal of the Tyveks apparel.
Unanticipated infrastructure was found during construction of the la-
goon liner. This included an abandoned, 42-in. diameter storm sewer
which had to be removed prior to installing the base liner. Also, it
was discovered early in the earth moving phase, that the existing con-
taminated stock pile extended several feet beyond the old lagoon
boundaries on the northern side of the area.
Construction of the 80-mil base liner occurred as planned without
other significant problems. However, placement of the top HDPE
cover presented some interesting problems. The second construc-
tion contract for chlorination and dechlorination facilities that was
planned to dovetail with construction of the primary settling tanks
was delayed. The City decided to reject bids and readvertise with
the result that less material was placed into the recoverable storage
facility than planned. Therefore, the top cover was placed on the slope
of the contaminated soil facing the empty northeast quadrant. A
problem of inadequate anchorage of the toe of the HDPE cover was
encountered. Wind got under the cover, lifted it and formed an air
bubble. The toe of the cover moved approximately 10 ft up the slope.
Before the cover could be adequately secured, the wind lifted almost
the entire cover. The cover was kept within the containment area only
by the anchor trench holding down two sides of the liner. The cover
had to be pulled back in place, slit, spliced and welded to repair
SITE REMEDIATION 457
-------
the damage. When sufficient sand bags for temporary anchorage were
placed on the toe of the cover, the problem was resolved.
Due to the high permeability of the soil and the proximity of the
Patapsco River, the flow rate required for the dewatering operations
was much higher than planned. However, analysis of groundwater
samples revealed that the water being pumped from the site was not
as contaminated as anticipated. Therefore, groundwater could be dis-
charged to the Patapsco POTW influent without treatment while
meeting the prescribed pretreatment standard for total chromium.
The groundwater quality was monitored daily for the initial two week*
and on a weekly basis thereafter. Any adverse changes in quality
would have resulted in increased frequency of testing followed by
pretreatment to reduce the chromium to concentrations within the
pretreatment discharge standard.
COMMUNITY RELATIONS
Two meetings were held with the concerned public regarding the
Patapsco contaminated soil containment project. One meeting was
held with the staff of the POTW to explain the nature of the problem
and the project, the precautions taken in preparing the design and
the requirements imposed on the construction contractor and the plant
staff. The plant staff was denied access to the contaminated area for
the duration of the project. The practical aspects of contaminant trans-
fer and the mitigation measures were explained to the staff. Following
the meeting, questions were answered by the consultants and City
engineering staff. It was agreed that results of monitoring would be
made known to the POTW staff.
A meeting for the general public, including residents living adjacent
to the POTW, was held several days following the meeting with the
POTW staff. The consultant presented the information in a similar
manner. A fact sheet was distributed by the City to local residents
to acquaint them with the project. The meeting for the general public
received local media coverage. As expected, the initial public reaction
was adverse, but after careful explanation, the neighbors were satis-
fled that adequate mitigation measures were planned to protect them
and the environment from contamination.
CONCLUSIONS
The City of Baltimore was in the difficult position of being required
to expand the Patapsco POTW while controlling potential exposures
to chromium-contaminated soil. The City reached an agreement with
the Slate Hazardous and Solid Waste Management Administration
that an on-site, recoverable storage facility would be used for long-
term storage of the contaminated soil.
A streamlined site investigation was conducted so thai plans and
specifications to construct the facility could be completed. The
process from the start of the site investigation phase through the end
of the design phase was completed within IS mo. Construction of
the recoverable storage facility has allowed the City to continue with
its required POTW expansion while handling the chromium-
contaminated soil on-site in an environmentally sound and cost-
effective manner. This approach has not precluded future decisions
regarding treatment, disposal or reuse of the stored material.
ACKNOWLEDGEMENTS
The authors express their thanks to George G. Balog, Director of
Public Works for Baltimore City, for permission to present and pub-
lish this paper. The authors also express their thanks to Brigid E.
Kenney and Glen J. Salas, P.E. for their technical review and Brigid
E.R. Hagarty for her editorial review.
REFERENCES
I. Earth Engineering it Sciences. Inc . Subsurface Investigation ftaapfco Hfrar
Hhier Treatment Plant. Baltimore. MD. Nov.. 1986.
2 New Jersey Stale Legislature. Environmental Cleanup Responsibility Act,
Trenton. NJ.
458 SITE REMEDIATION
-------
The Observational Approach for
Site Remediation at Federal Facilities
R. Scott Myers
Pacific Northwest Laboratory
Richland, Washington
Samuel J. Gianti
CH2M Hill
Reston, Virginia
ABSTRACT
The current approach to hazardous waste site remediation is based
on the assumption that all important information about the site will be
known before remediation begins. This approach, based on the con-
ventional engineering paradigm of study, design and construct, leads
to the selection of a single remedial alternative with no contingencies
for variations encountered during construction. The design and con-
struction of remedial alternatives is approached much as the design and
construction of a bridge or a treatment facility.
This approach works well for traditional engineering activities, where
uncertainty can largely be eliminated by study and investigation and
by the existence of a large body of empirical evidence. However,
hazardous waste site characterization and remediation is dominated by
uncertainty. Variations in soil conditions, geo-hydrology, transport
mechanisms, waste source and chemical and physical characteristics
make it impossible to completely characterize and understand actual
site conditions. In an attempt to overcome this uncertainty, site charac-
terization all too often consists of excessive rounds of sampling. At best,
excessive sampling requires too many resources. At worst, excessive
sampling can lead to a false sense of confidence and a disregard for
reasonable variations that could disrupt the effectiveness of the selected
remedial action.
There is, however, another way to approach hazardous waste site
remediation. The observational approach, developed by geotechnical
engineers to cope with the uncertainty associated with subsurface con-
struction such as tunnels and dams, can be applied to hazardous waste
site remediation. During the last year, the observational approach has
gained increasing attention as a means of addressing the uncertainties
involved in site remediation.
In order to evaluate the potential advantages and constraints of
applying the observational approach to site restoration at federal
facilities, a panel of scientists and engineers from Pacific Northwest
Laboratory and CH2M Hill was convened. Their review evaluated
potential technical and institutional advantages and constraints that may
affect the use of the observational approach for site remediation. This
paper summarizes the panel's comments and conclusions about the
application of the observational approach to site remediation at federal
facilities. Key issues identified by the panel include management of
uncertainty, cost and schedule, regulations and guidance, public involve-
ment and implementation.
INTRODUCTION
The remedial process, as it typically is approached, has become
"bottlenecked" by the uncertainties associated with fully understanding
the nature of hazardous waste problems. As it typically is implemented,
the current approach to site remediation is based on principles from
conventional engineering. The assumption is that uncertainties in site
conditions can be effectively reduced to manageable levels during the
RI/FS phase of the process, thereby allowing the design and imple-
mentation phases to proceed routinely and predictably. This not has
proven to be the case for many sites; new information discovered during
design and implementation has often forced significant alterations in
planned remedies'.
The current process leading to site remediation follows a traditional
engineering paradigm of study, design and build. Following a series
of discussions about the scope of the project, U.S. EPA's objectives,
budget, operating assumptions and initial data, an RI is initiated,
followed by a FS that compares alternatives. A ROD declares the
preferred alternative, and a design is then prepared for remedial con-
struction. This process of site investigation, alternative evaluation and
remedy selection is described in U.S. EPA guidance2 and followed by
remedial managers and planners in all U.S. EPA regions. There is,
however, often a significant difference between the process as described
in U.S. EPA's guidance and regulations and the process as it typically
is implemented. For example, the current guidance and regulations
provide considerable support for many of the fundamental elements of
the observational approach. Nonetheless, many of these elements, such
as the early screening of general response actions or including
engineering considerations in early characterization efforts, are not
generally included in remedial investigations.
This evaluation of the observational approach used the process of
site remediation as it generally is conducted as a baseline for compari-
son with the observational approach.
Problems with the Current Approach to Site Remediation
The process of site remediation, or at least the RI/FS phase of that
process, has been going on long enough to establish a track-record.
Recent reports conclude that the site remediation process, as it generally
is conducted, often fails to provide effective, efficient cleanups3. One
of the fundamental problems with the current process or approach is
the failure to explicitly recognize the role that uncertainty plays in
virtually every aspect of site remediation. It generally is assumed, for
example, that more study will reduce uncertainty. But to date, it has
not been fully recognized that the marginal value of further studies at
Superfund sites declines rapidly. At some point, more study does not
lead to better information.
Another problem with the current approach to site remediation is
the emphasis on a "paper" product: contractors and managers tend to
focus their efforts on producing an RI/FS and a ROD. As a result, they
often obscure the important goal of protecting human health and the
environment in their rush to meet the milestones of the RI, the FS or
the ROD. The extensive uncertainty of site conditions, as well as the
SITE REMEDIATION 459
-------
complex nature of the process remediation process, tends to keep both
technical contributors and management focused on a series of short-
term objectives (paper studies) and lends to prevent "big picture" per-
spective of the ultimate goal (cleanup). This emphasis on producing
a paper product has contributed to excessive costs and prolonged
schedules. At complex sites, for example, millions of dollars may be
spent over many years to simply issue the ROD.
The problems with the site remediation process derive largely from
the ever-present element of uncertainty and the apparent inability of
the process to respond appropriately to that uncertainty. Many factors
contribute to uncertainty in site remediation. The subsurface environ-
ment is complex, heterogeneous and almost impossible to fully charac-
terize. Moreover, small subsurface features or changes in geologic
conditions can have substantial impact on contaminant movement.
Major uncertainty also plagues source characteri/ation; assessment of
chemical fate and transport in the environment; assessment of exposure
risks and health effects and; remedial action performance. Taken
together, these factors make uncertainty an inherent feature of ha/ardous
waste sites.
This stale of uncertainty should not lead to inaction. Uncertainty is
not unique to hazardous waste problems; geotechnical engineering has
had to respond frequently to similar situations. The engineering com-
munity now has the opportunity to bring a different and a generally
more appropriate, paradigm to Superfund site remediation.
The Observational Approach to Site Remediation
The observational approach is based on principles developed by geo-
technical engineers in response to the uncertainty of conditions encoun-
tered when constructing tunnels and other sub-surface structures. Instead
of trying to completely characterize sub-surface conditions before
beginning construction, the observational method, as it is referred to
in geotechmcal literature, requires only that the probable conditions
of a site be known. Once the expected conditions are defined, potential.
but reasonable, deviations to those conditions can be identified and
contingencies can be prepared to respond to those deviations. If the
contingencies for all reasonable deviations can be accommodated by
the projected construction techniques, construction is begun. If. however.
the projected construction techniques cannot accommodate all
reasonable deviations, then further characterization is required to more
precisely define the expected conditions and thereby reduce the number
of reasonable deviations.
A complete explanation of the observational approach is beyond the
scope of this paper. Such an explanation is included, however, in
Peck" The fundamental elements of the observational method have
been refined by CH,M HILL for use on hazardous waste sites as
follows:
• Define scope of work: establish goals and objectives, review existing
data, develop a conceptual model and identify data gaps
• Conduct an initial screening of general response actions
• Collect information on site conditions, including the nature and ex-
lent of contamination
• Use the information collected to construct a conceptual model of the
site to establish probable conditions and reasonable dcvutions
• Prepare a feasibility study: evaluate the remediation alternatives and
prepare conceptual contingency plans as a response to identified
deviations; recommend the most effective alternative, given probable
conditions at the site
• Design the chosen remedial action, select parameters lo observe and
prepare contingency plans
• Implemenl remedial action and measure responses
• Respond to deviations
These eight steps represent an outline of the observational approach
to site remediation, li is probably more useful, however, lo ihink of
the observaiional approach as a conceptual framework for site
remediation. The three basic tenets of this conceptual framework are:
(1) characterization should be undertaken for a specific purpose, not
just to find oul about the contamination at or the general characteris-
lics of ihe site; (2) more data do nol automatically lead to better infor-
mation; and (3) the process should converge on a general response action
as early as practical. Keeping these bask tenets in mind throughout
the RI/FS process should provide a better focus to the technical work,
thereby offering (he opportunity for lower costs, shorter schedules and
a superior technical product.
The observational approach also may lend itself to maintaining a "big
picture'' perspective throughout the remediation process. This perspec-
tive could help change the focus from producing a RJ/FS to determining
the problem and the best solution and could result in a higher quality
of work, lower costs and a shorter schedule.
Other advantages to the observational approach include:
• Providing an opportunity for decision-makers to prepare for events,
rather than merely respond to them.
• Establishing a more formal mechanism for evaluating the work done
during a previous phase before proceeding to (he next phase (his
crucial step in strategic planning generally is missing from most
current site remediation activities.
• Providing for specific contingencies to respond to potential devia-
tions from expected conditions.
Implementing the observational approach for site remediation at
federal facilities will involve a number of issues. The key issues, as
identified by a panel of scientists and engineers from PNL and
CH.M Hill, are discussed below
BETTER MANAGEMENT OF UNCERTAINTY
Uncertainties in site remediation exist regardless of bow the
remediation process is conducted. Most current site remediation
strategies, however, either ignore those uncertainties or assume that
sufficient study and assessment will essentially eliminate them. But
one of the harsh realities of site remediation remains constant:
uncertainty can be neither ignored nor studied away. The observa-
tional approach, by explicitly recognizing uncertainty, offers a credible
mechanism for dealing with thai uncertainly. The practical advan-
tages of such a mechanism include an improved technical under-
standing as well as a more honest presentation of the situation to
the public.
Unexpected conditions always will be a possibility in any she
remediation effort. Since no amount of study will eliminate the
possibility of surprise, planning for reasonable deviations is simply
good risk management. The observational approach provides a
mechanism for planning for the unexpected. One of the fundamen-
tal elements of ihe observational approach, the inclusion of
contingencies to deviations from expected conditions, can help with
planning for the unexpected. By identifying reasonable deviations
to expected conditions, planners can consider general responses to,
or contingencies for. those deviations. The deviations to expected
conditions could be identified in the FS together with their impli-
cations for each of the alternatives under consideration. Specific con-
tingencies to respond to those deviations would be developed in the
remedial design (RD) phase, after the ROD has been issued.
R\ establishing contingencies as a formal part of the process, the
observational approach offers an opportunity to identify potential
problems and responses to ihose problems, before they occur. Instead
of facing a crisis when something goes wrong, decision-makers have
(he advantage of being prepared with pre-planned contingencies for
possible problems. Consequently, when a problem occurs, the
response can be implemented more quickly, potentially saving time,
money and embarrassment. The more conservative alternative is to
over-design the remedial action so that any possible contingency can
be readily handled. The excessive cost of such a conservative ap-
proach, however, essentially precludes it as a meaningful response
lo uncertainty.
In summary, uncertainly will exist regardless of the method or
approach used for characterization and remediation. The obser-
vational approach simply offers a belter way of managing that
uncertainty.
REGULATIONS AND GUIDANCE
U. S. EPA regulations and guidance offer substantial support, both
460 SITE REMEDIATION
-------
direct and indirect, for the use of the observational approach for site
remediation. Direct support comes from an OSWER Directive issued
early in 1989 recommending several measures to reduce the time
and the cost of the RI/FS process'. This guidance included a direct
recommendation for each U.S. EPA region to use the observational
approach in at least one Superfund-lead site or operable unit in 1989.
Other recommended streamlining measures that support the concept
of the observational approach include:
• Identifying the use of data before obtaining those data
• Identifying probable remedial action(s) as early in the process as
possible
• Tailoring the level of detail of the alternative evaluation to the scope
and complexity of the action
Portions of the proposed NCP also reflect, or offer further support
for, elements of the observational approach. According to the proposed
NCP, a primary program management principle to be followed by the
U.S. EPA is a "bias for action." The observational approach offers a
structure for applying this principle by integrating action and study as
the RI and the FS proceed. This integration could include:
• Narrowing down the field of potential remediation technologies earlier
in the process
• Tailoring the level or detail of the analysis of evaluation criteria to
the scope and complexity of the action* Tailoring the selection and
documentation of the remedy based on the limited scope or
complexity of the site problem and remedy.
The NCP clearly advocated that action be taken to move forward with
the actual work of remediation as early as site data and information
make it possible to do so. The observational approach provides a credible
mechanism for doing so.
Although there is specific support for the observational approach in
the U.S. EPA regulations and guidance, it is nonetheless possible that
implementing the observational approach could be perceived as an "end
run" around the regulations. It will, therefore, be very important to
get early agreement from the appropriate agencies (and individuals)
about establishing a mechanism for dealing with uncertainty (i.e., the
observational approach).
NEPA Constraints: One cannot pre-judge the process
In addition to compliance with all federal, state and local regulations
applicable to all Superfund sites, federal facilities often must conform
to the requirements of the National Environmental Policy Act (NEPA).
One of the primary constraints of NEPA, which many federal facilities
are obliged to work with in site remediation, is not pre-judging the out-
come of the process. The potential issue here is that one of the goals
of the observational approach is to converge on a probable remedy early
in the process. To work within the requirements of NEPA, however,
it is imperative that any such convergence not be made to the exclusion
of considering a wider range of alternatives. In other words, even if
early in the process there is little question about what the final remedy
will be, the investigation cannot be narrowed down to looking for only
information that will support decisions about that remedy. On the other
hand, there will be different levels of detail associated with the con-
sideration of various alternatives. There is no requirement under NEPA
to apply the same level of detailed information to all potential alterna-
tives, so it is possible to pursue "less likely" alternatives to a lesser
depth than the more (or most) likely alternative. In all cases, however,
there must be enough information on each alternative to justify the
eventual decision to either eliminate or select it.
COST AND SCHEDULE
One of the potential advantages of the observational approach is the
opportunity to start sooner on the actual work of site remediation. Given
the bias for action of the NCP and the consequent U.S. EPA guidance,
a remedial project manager does not need to "use the observational
approach" to get started sooner on actual site remediation. However,
the observational approach does provide a conceptual framework that
allows remediation to begin as soon as the regulating agency(ies) allow.
There are two primary ways that actual cleanup work can begin on
an expedited basis. The first is to move through the RI/FS process more
quickly, primarily because of a more focused and therefore reduced,
sampling program. An appropriately focused sampling program will
provide much of the information necessary for engineering design during
the RI phase. Such an integrated sampling program also may reduce
the length of time it takes to produce the FS report. Although interim
response actions are possible under the current process, they are not
typically integrated with the conventional aspects of investigation and
alternative evaluation. The observational approach offers the opportu-
nity for integrating interim response activities with the longer-range
objectives of site remediation.
A second way that cleanup action can begin sooner is to incorporate
cleanup activity into the RI process. Examples of integrating cleanup
activities with remedial investigations include removing underground
drums and tanks and initiating soil vapor extraction to recover volatile
organics that would otherwise continue to disperse into the environment.
Although it is possible that the observational approach may lead to
faster action, it would be unwise to claim that the observational approach
will lead to faster cleanups. A conservative claim is that the observa-
tional approach provides the opportunity for faster action, but it is not
clear it will do so in all, or even most, situations.
Just as the observational approach provides the opportunity for faster
action, it also may provide the opportunity for lower cost. The poten-
tial for lower costs comes from the possibility that an observational
approach RI/FS will be "leaner and meaner" than the conventional
approach, as it is typically implemented, would allow. With specific
targets for information, sampling costs could be lower. And narrowing
down the set of alternative remedies early in the process could reduce
the time and therefore the cost, of producing the RI/FS report.
There is support from the regulations for this sort of "streamlining"
that could lead to lower costs. According to the proposed NCP, "The
RI should be focussed so that only data needed to develop and evaluate
alternatives and to support design are collected."5. The observational
approach, could provide a coherent framework for decisions about what
data to obtain. Under the observational approach the specific goals of
sampling and analysis would be more strictly defined and the quantity
of samples required to reach a given decision could therefore be reduced.
It is very important to realize, however, that there is no guarantee
that using the observational method will result in lower costs.
CONCLUSIONS
Hazardous waste site remediation has, to date, been conducted
following the conventional engineering paradigm of study, design and
build. The reality of hazardous waste sites, however, is dominated by
uncertainty, a condition for which the conventional engineering paradigm
is poorly suited. The observational approach offers a way of
acknowledging and dealing with the inherent uncertainty of hazardous
waste site conditions. Specific conclusions include:
• Uncertainty in site remediation activities cannot be eliminated by
further study at some point in the remediation process, uncertainty
must be confronted.
• The observational approach can provide a central philosophy for the
entire process of site remediation, from project planning through post-
closure monitoring; the potential advantages of incorporating such
a philosophy throughout the process include providing a framework
for various streamlining measures which can lead to a more efficient
use of resources to achieve a high level of remediation.
• The observational approach provides a better mechanism for
managing risk; uncertainties will always be present in site remedia-
tion work, but acknowledging those uncertainties and preparing for
deviations can minimize the consequences of the unexpected.
• There is nothing in the current regulations and guidance that precludes
the use of the observational approach for site remediation; in fact,
an aggressive remedial project manager can implement the obser-
vational approach based on support from various elements of the regu-
lations and guidance.
SITE REMEDIATION 461
-------
ACKNOWLEDGEMENT "«". Feb., 1989
_...., . . . . , ... ,.c r^^r,mnn, 2 U.S. EPA, OSWER Directive 9355.3-01, Guidance for Conducting Remedial
Pacific Northwest Laboratory ,s operated for the US Department legations and Ktalbiu,y Studle, Under CEKCLA. Oct.. 1988,
of Energy by Battelle Memorial Institute under contract 3 us senators Uutcnbcrg and Durenbergcr, Repo* on Super/andImpknen.
DE-AC06-76RLO 1830 uuion: Cleaning Up ihe Nation's Cleanup Program.
oc-EfDpxir'irc 4 Peckl R'B • Advuwagcs and Limitauons of the OfaervationaJ Method in
KtUJtfclNCIiJj Applied Soil Mechanics, Gtouchniaue, 19. pp. 171-187, 1969.
1. US. EPA Office of Solid Waste and Emergency Response (OSWER) Direc 5. US EPA. Proposed Rule: Proposed National Contingency Plan. 300.430,
live 9355.3-06, RlfFS Improvements Phase II. StrtamlMng Krcommrnda Fed. Keg. 53. (245). Dec. 21. 1988
462 SITE REMEDIATION
-------
New Capability For Remote Controlled Excavation
William P. Wohlford
B.D. Bode
ED. Griswold
Deere & Company
Moline, Illinois
ABSTRACT
Remote controlled operation of construction equipment has been state-
of-the-art for some years. The availability of remote controls which have
been designed and developed for general use on commercial machines
is a recent development and is the subject of this paper. The John Deere
teleoperated excavator represents a new capability that is now availa-
ble to the construction industry for use on construction sites that preclude
the on-site presence of human operators. This paper will describe the
basic machine, the controls, vision system and integration of the remote
control adjunct to the operational system. Much of the development
of the initial capability was done with the cooperation of Vectran
Corporation of Pittsburgh, Pennsylvania.
THE BASE MACHINE
The John Deere 690 Excavator is a commercially available production
machine. The first teleoperated unit was fielded on this variant (Fig. 1).
The base machine is a 41,000-lb excavator, modified for the Air Force
to include a wheeled undercarriage, a dozer blade, stabilizers, a
hardening package and variable boom geometry. The 690CR and now
the 690DR are the mainstays of the Air Force rapid runway repair fleet.
The machine has 125 net hp, 31-ft reach and 20-ft dig depth. It is sup-
plied to the Air Force with a bucket, hydraulic breaker and tamper that
enable it to perform the functions needed to repair craters on damaged
runways. The repair of runways is currently a manned operation.
However, the Air Force needed an additional unmanned capability to
deal with unexploded bombs at an Air Force test range. This capability
was provided by the Teleoperated Remote Controlled Excavator
(TORCE).
The TORCE excavator will transport and perform all work functions
from a distance of 5,000 ft on radio command and 1,000 ft on coaxial
cable. The Air Force has used this machine with success since its
delivery in March, 1987. The conversion of the base machine to remote
controlled operation involved the integration of servo hydraulic controls,
vision and audio feedback, remote operator's station and data links.
ELECTRONICS
The remote adjunct has three basic subsystems. The simplified block
diagram shows the operator's console, the on-board package and the
data link (Fig. 2). The console includes the video monitor and
audio/video receivers and the decoding electronics needed to process
incoming signals. It also includes the control devices and encoding elec-
tronics to generate and broadcast commands. The on-board package
T
CONTROL STATION
T
VIDEO ( AUDIO
TRANSMITTER
i — H
i
1
1
AUDIO
:l«ci
L
COLOR
CAMERA
—
1
COLOR
CAMERA
PAN.T1LT.2OOM
RECEIVER A
DECODER
ON OFF
CONTROLS
RELAY
BOARD
— 1 II
ON-OFF
FUNCTIONS
(PROPORTIONAL
CONTROLS
DRIVERS
I
1
SERVO
VALVES
(ENGINt: STAKTJ
STOP. DttfLK UK/
DOWN, k 1C)
(DIG AND PkurtL
Figure 1
John Deere 69OCR Excavator
Figure 2
690 Remote Control Block Program
SITE REMEDIATION 463
-------
receives incoming commands and converts them to electrical signals
for valve and camera control. It also processes video and audio data
and sends these out to the operator.
The remaining element is the link that joins the on-board electronics
to the operator's console. This data link can be a coaxial cable, radio
waves or optic fiber.
The 690C on-board package consists of two separate subsystems for
the present version (Fig. 3 and 4). The digital receiver package pro-
vides the functional interface with the machine while the video trans-
mitter provides the sensory feedback data.
Figure 3
Digital Receiver
This environmentally sealed container (Fig. 3) is mounted on the
rotating house of the excavator. The system includes two receiver* capa-
ble of 9600 baud. Its function it to receive a data string of digitally
encoded commands on RS232. decode and interpret the command sig-
nals and then relay them to hydraulic servos and actuators. The relayed
commands are analog signals that provide proportional control capa-
bility to the boom. arm. bucket, swing and transport hydraulics. This
unit also generates signals that command discrete functions for eagiae
stop/start, high/low speed select, blade up/down, engine speed, road
speed, auxiliary tool, stabilizers and bucket clamp. Fail-safe MOOR
in this unit will shut down the engine in the event that clear commands
are not received.
This second container is similar in size and shape tothe first (Fjg, 4).
Its function is to code and transmit video and audio data from the
machine to the operator and to control the power pan/tih/zoom frac-
tions of the roof mounted camera. This unit is also enviroemeataOy
sealed and mounted on the rotating house.
The operator's console is designed for adverse weather and hamffiag
(Fig. 5). It weighs under SO IDS and can be accompanied by a battery
pack (Fig. 6) for 8 hr of isolated continuous operation. The 8-in. mentor
provides viewing from either the fender mount or roof mount camera.
A camera select switch, pan/tilt/zoom and manual iris override cjptnfs
are mounted on the panel. The four joysticks are operated ia the same
manner as the cab mounted controls.
Figure 4
Video Transmitter
Figure 5
Operator's Console
This feature maintains the continuity of similarity with the cab and
aids the operator in quick and errorless operation of all machine
functions. The console can be operated from a 60Hz, UO v source or
the batteries can be charged from that source.
The RF data link has a SjOOO ft range operating at 5 w on the com-
mand link and 10 w on the video. The Air Force system operates in
the UHF frequency band with 12 Khz bandwidth on the command link
and 6 Mhz on the video link. RF communications continue to be a
problem in the United States and abroad due to the heavy demand for
military and commercial use of the air waves. Video transmission, which
is essential to remote operation, requires wide bandwidths which are
increasingly difficult to obtain from the Federal Communications Com-
mission (FCC). The ideal frequency range for teleoperation lies in the
low end of the spectrum to achieve omni directional flexibility and
maximum penetration of interposed ground features. The frequencies.
464 SITE REMEDIATION
-------
Figure 6
Battery Pack
however, tend to be preallocated or available only in narrow bands. One
solution to the dilemma is to operate at higher frequencies and adapt
to the limitations. Deere has addressed this issue as noted in the following
paragraphs.
The heavy duty coaxial cable is provided for teleoperation with stand-
off distances up to 1,000 ft. This secondary data link enables fast
response to areas where the RF link is not approved or appropriate.
It may also be used on occasions where the suspect hazardous materials
may be affected by RF energy or in locations where the RF transmis-
sion is blocked by geological features or metal structures. The cable
was provided in coil form on the first unit for manual payout.
VALVES
John Deere 690 excavators are equipped with pilot operated hydraulic
valves. The remote control system is superimposed on the pilot pres-
sure system with this valve manifold assembly (Fig. 7). Proportional
functions are controlled through commercially available servo valves
while the discrete functions apply solenoid valves. The remote control
valve assembly is designed and integrated to be transparent to an oper-
ator seated in the cab who has the machine under manual control. Elec-
tromechanical actuators are mounted on engine fuel control and speed
selector controls in a way that does not interfere with manned operation.
VISION
Vision is provided to the remote operator with two cameras on the 690C
model. The first camera is fixed focus auto iris, fixed but manually
variable mount, located on the front right side. The camera has a wide
angle lens directed to the area swept by the excavator linkage (Fig. 8).
Experience has shown this to be an essential view for remote operator
inspection of details in the work area. The camera is color as is the roof
mounted camera. This camera has power pan/tilt/zoom with auto iris.
Additional manual iris override permits adjustment for improved vision
in dark excavations. The roof mount with remote controlled aiming and
zoom results in a narrower field of view with full operator discretion
of the viewing target. The camera also provides visual operating feedback
when manipulated to look through the cab roof at the instrument panel.
This is a patented feature of the John Deere TORCE 690C. The sensory
feedback includes an in-cab microphone which transmits operating system
audible warnings and engine and hydraulic system operating sound levels.
This audio feedback is a valuable link of operator to machine as he seeks
to optimize performance by loading the engine and hydraulics to capacity
without creating stall or relief valve opening. The antennae that are needed
to receive command signals and transmit sensory data are mounted on
the cab roof.
Figure 7
Electrohydraulic Valves
Figure 8
Cameras
MOUNTING
Modifications to the production excavator are needed to provide
mounting points for the on-board hardware and electromechanical
actuators. These brackets and components are designed to be mounted
in less than 8 manhours using only simple tools. They also provide
for on-board storage of the operator's console. The design for super-
position of the teleoperation subsystems over the existing manual sys-
tems gives the user the option of removing the remote control
components for storage during long-term manual operation or moving
the remote control capability to any similar excavator equipped with
an adaptor kit. This feature is expected to be particularly valuable to
commercial users who may have multiple machines at widely separat-
ed sites.
SITE REMEDIATION 465
-------
THE PRODUCT IMPROVEMENT UPGRADE
John Deere has introduced the next evolution of the 690 excavator
called the 690D (Fig 9) It incorporate!, a number of performance and
reliability upgrades including a new closed center hydraulic system
The new model has the same operating features as the previous model
A major difference is the addition of a third camera inside the cab
This offers the operator an instant view of his machine instrument panel
at the flip of a selector switch rather than repositioning and refocusing
the roof mounted camera The third camera is a color, fixed focus, fixed
mount unit.
Figure 9
John Deere 690D
Finally, the RF data link is being modified to operate partially in
the microwave frequency regime. This change has the advantage that
FCC approval is more easily obtained The wide bandwidth is more
readily available at microwave frequencies and therefore more appro-
priate for commercial uses The disadvantage of operating at these high
frequencies is the requirement for line-of-sight communication between
the sending and receiving antennae Present experience has not shown
this to be a problem for the ordnance disposal and cleanup tasks that
have been accomplished to the present time. Commercial uses, presently
envisioned, should be equally insensitive to broadcast frequencies. The
RF link that is now in development has the added feature of selected
bands within the available frequency range. This enables control of up
to five systems simultaneously at the same site Control of multiple units
from the same console is an option if only one machine is being worked
at any given time If multiple machines are in operation, then multiple
consoles would be required
The hardwire data link also is being upgraded with the addition of
a cable reel that will simplify the payout ret neve task for the coaxial
cable Fiber optic links have not been ordered up to this time, but they
also arc readily adapted to the system The advantages of fiber optic
ImLs lies in their resistance to electromagnetic interference which
particularly concerns the military
FIELD EXPERIENCE
The production excavator evolved over a number ol years to employ
manual valve actuation of functions using both hand and fool controls
Each succeeding generation was an improvement over its predecessor
in the man/machine interface The design of the remote operator's station
leveraged the previous design evolution by duplication of the hand con-
trol motions and relative locations. The propel controls could not be
incorporated as foot controls but did conform to the control response
patterns that are common to hydrostatic transmissions. The removal
of the operator from the cab. nevertheless, results in loss of sensory
inputs from tactile and vibration sources, and it considerably reduces
visual inputs The result of this is reduced productivity when engaged
in benign or conventional carthmovmg hut with substantial increases
in productivity compared to the alternatives when operating in hazardous
environment* Learning to operate a remote controlled excavator from
a remote station appears to be readily accomplished by totally imfcmafr
operator* and slightly more time-consuming for operators accustomed
to a full range of sensory information. In either case, very high leveb
of productivity can be achieved with practice.
The TORCE I, adaptation of the JD690CR wheeled excavator to the
Air Force rapid runway machine, has been in operation at an Air Fonx
base since March. 1987. The Explosive Ordnance Disposal (EOD)leam
stationed at the site has used the system on a routine basis to excavate
unexploded munitions and recover them for inspection. Their objec-
tive is to retrieve live pretriggered explosive devices intact for failure
analysis while remaining safe from harm. The EOD team has changed
personnel through the period, but has found that new people are readi-
ly trained. The excavator has functioned reliably and effectively in in
assigned role.
The EOD team with its remote excavation capability was enlisted in
the summer of 1997 to evaluate its use in cleanup activities at the Milan,
Tennessee. Army Ammunition Plant. Since its opening in 1941, the plan
has buried a variety of explosives and obsolete munitions in trenches
around the area The exact locations and contents of the burial sites
were unknown, but the Corps of Engineers is concerned about ground-
watcr contamination and identification of the buried materials. The
Corps of Engineers engaged the EOD team to remotely excavate 55
sites At the conclusion of the operation, the team had excavated 64
sites to an average depth of 18 ft in 84 machine hours or about half
of the time originally scheduled to complete the project. It had recovered
over 300 items of ordnance and provided soil samples for analysis. The
engineer in charge of the project estimated that 30 to 40% cost savings
could be realized using remote controlled excavation as compared to
using manned excavators Onboard operators at hazardous sites could
be required to wear fully encapsulated life support systems and then
for only short working intervals The reduced capacity, downtime and
multiple crews needed to support a single excavation are all unneces-
sary with remote control It also was noted that the Milan task was
only a survey, that real cleanup work was yet to be done and that there
are 12 other similar plants in the United States. Clearly, remote
controlled excavation is here to slay.
Another major change is the upgraded on-board electronics package
which includes miniaturized relays and compact circuitry. This
eliminates one of the sealed on-board containers. The reduction in
package si/e and weight simplifies the mounting design and results in
location which is more immune to the rigors of construction machine
environment. The new operator's console will incorporate an 8-in.
monitor and sufficient electrical power to complete 8 hr of operation.
Figure 10
Air Force Automatic Hxoavalor
THE KlTt'RE
The An 1-oac Engineering and Sen ices laboratory (AFESO is con-
cerned with rapid runway repair and with the availability ol trained
466 SITh REMEDIATION
-------
personnel to operate repair machines as the time to repair becomes more
critical. The Engineering and Services Laboratory initiated a program
that was headed toward full automation of the runway repair process.
The first task in the program was to provide automatic tool change.
Deere, University of Florida and Westinghouse worked with the Air
Force to produce the machine shown here (Fig. 10). With the added
expertise of an on-site contractor, the system can now change tools,
dig trenches, dig pie-shaped or circular holes, level blade, tamp and
break concrete all automatically by calling up the desired task on a com-
puter menu. The operator need not be on board while the machine is
working. This particular machine is a proof-of-concept system and
normally prone to the reliability problems that engineers and labora-
tory technicians often find in prototypes. The manager of R&D systems
at the AFESC has reported that the machine has logged 780 hr of
operation with the sensors and computers on board.
Where do we go from here? It is possible to remove operators from
construction machinery cabs. It is certainly a necessary thing in opera-
tions like the Milan ammunition plant and any job where hazardous
materials or dangerous conditions are likely to exist. Whether or not
it becomes commonplace in day-to-day construction work depends on
its cost-effectiveness. Can a contractor achieve a return on his invest-
ment by replacing manpower with computer power? Today, the answer
is yes only when conditions exclude human beings. Tomorrow's answer
will depend on the cost of labor and on the availability of low cost elec-
tronics.
SITE REMEDIATION 467
-------
Capture of a Groundwater Contamination Plume in
Fractured Bedrock by an Artificially Produced
Fracture Zone Created Through Controlled Blasting
Kristen Franz Begor
Rodney W. Sutch
Dunn Geoscience Corporation
Albany, New York
Michael A. Miller
General Electric Company
Fairfield, Connecticut
ABSTRACT
Recovery of contaminated groundwater in a fractured bedrock
system presents some unique problems. Typically, the most com-
mon problem occurs from the hydrogeologist's inability to ade-
quately characterize the discrete fractures through which con-
taminants may be migrating. Without adequate characteriza-
tion, difficulties arise in properly positioning recovery wells and
verifying the performance of the system. To overcome these dif-
ficulties at a site in Upstate New York, an innovative approach
was developed involving the creation of an artificial fracture zone
through controlled blasting to intercept contaminated ground-
water flow.
Site investigations delineated the extent of a groundwater con-
tamination plume migrating within a fractured bedrock aquifer
(Medina sandstone) which underlies approximately 15 ft of glacial
till. A 72-hr aquifer test involving one recovery well resulted in a
low yield (3.5 gpm with 20 ft of drawdown). Data collected from
adjacent observation wells indicated poor interconnection among
the naturally occurring fractures. Although the response of some
observation weUs mirrored that of the recovery well, others
showed little or no response to pumping. Therefore, achieving the
corrective action objectives (i.e., preventing further contaminant
migration and removing and treating contaminated groundwater)
would be difficult using a traditional, multiple recovery well sys-
tem. It was decided that controlled linear blasting could provide
the enhanced fracture interconnection necessary to successfully
intercept the contaminated groundwater plume, which would
then be captured and removed by judicious placement of recovery
well(s) installed within the fracture zone.
Using a carefully controlled single line pattern blasting tech-
nique, a 6-ft wide, 300-ft long fracture zone was created in the
upper 25-ft of the bedrock aquifer perpendicular to the center-
line of the plume. Following fracturing, a second 72-hr aquifer
test was conducted at the same location and under conditions sim-
ilar to the first test. The second test indicated that the single re-
covery well located in the newly created fracture zone should be
fully capable of recovering contaminated groundwater and pre-
venting further migration of the plume. The recovery well pro-
duced a substantially higher yield of 18.5 gpm with only 11.2 ft of
drawdown. Furthermore, all of the nearby observation wells
showed significant response to pumping. Success at this site is
promising and the approach may prove useful at other sites in-
volving contaminated bedrock aquifers.
INTRODUCTION
A manufacturing facility in Upstate New York operated a series
of surface impoundments used to treat wastewater from plating
operations and various other metal finishing processes (Fig. 1). A
comprehensive groundwater quality assessment program con-
ducted at the facility identified contamination of the ground-
water by volatile organic compounds (VOCs) within both the
overburden and bedrock aquifers. A corrective action program
was implemented upon completion of the groundwater assess-
ment program.
This paper focuses on the corrective action measure that was
developed at this site to prevent further migration of the contam-
inated groundwater. Background information is included on the
nature and extent of the contamination, site hydrogeology and
conceptual development of the fracturing technique. Also pre-
sented are a description of the fracturing process, results of pre-
and post-fracturing aquifer tests and a discussion of the effec-
tiveness of the technique.
SITE HYDROGEOLOGY
Unconsolidated deposits at the site consist of 5 to 20 ft of Late
Woodfordian sandy glacial till overlying approximately 50 ft of
Medina sandstone (Grimsby member) of early Silurian age.
Underlying the Medina are several hundred feet of Upper Ordo-
vician Queenstone shale. Regional bedrock dip is to the south at
approximately 50 ft/mi.
Groundwater at the site is presently under unconfined con-
ditions. The water table is typically 4 to 8 ft below ground surface.
Although the overburden and bedrock units art discussed below
as two separate aquifers, they are hydraulically interconnected.
The basis for discussing the two aquifers separately arises from
the inherent differences between the two units with regards to
the geologic material and nature of groundwater flow.
Groundwater flow within both the overburden and bedrock
aquifers is predominantly to the northwest across the site with an
increasing gradient to the north in response to the topography'
There is generally a downward gradient between the two aquifers
and within the bedrock. Based on slug and bail tests performed at
the site, the average hydraulic conductivity (K) of each of the
aquifers is roughly the same, 10 cm/sec (0.28 ft/day).
Groundwater flow within the till is assumed to be predominant-
ly through intergranular pores. Based on hydraulic conductivity
468 REMI.OIAL ACTIONS
-------
values, water level data and an estimated effective porosity of 10
to 20^0, the average linear rate of groundwater flow within the
overburden aquifer ranges from 0.04 to 0.26 ft/day. Ground-
water flow within the Medina sandstone occurs predominantly
through secondary porosity openings such as fractures, joints
and bedding planes. Intergranular flow is judged to be minimal.
Due to the nature of fracture flow, the true groundwater flow rate
varies considerably between individual fractures, making accurate
calculations of flow velocities and travel times almost impossible.
However, based on hydraulic conductivity values, water level data
and an estimated effective porosity of 5 to 15%, the average linear
rate of groundwater flow within the bedrock aquifer is expected
to range from 0.04 to 0.31 ft/day.
In an attempt to better understand the nature of groundwater
flow within the bedrock, several studies were performed. These
studies included a fracture trace analysis utilizing historic aerial
photographs, a joint analysis based on a nearby outcrop, corre-
lation of rock core data and the evaluation of geologic tunnel
data collected approximately 20 mi from the site. The information
indicated that two major sets of nearly vertical fractures existed
within the bedrock: a northwest trending set and a northeast
trending set. Although a great deal of generalized information
had been gathered, insufficient site-specific data were available to
determine the spacing of the fractures or to identify the existence
of major fractures into which recovery wells could be installed.
CONTAMINANT PLUME DELINEATION
A groundwater quality assessment program was implemented
to delineate the nature and three-dimensional extent of the con-
tamination. Many of the monitoring wells on-site were installed
as pairs, with one well monitoring the overburden and the other
well monitoring approximately the upper 10 ft of bedrock. Sev-
eral bedrock monitoring well clusters, located along the northern
boundary of the site, were installed to monitor the upper 25 ft of
bedrock. These wells were screened to monitor discrete 8-ft zones
within the bedrock (Fig. 2).
Utilizing this approach, the vertical and lateral extent of con-
tamination was evaluated at the property boundary.
Contamination by VOCs associated with the degreasing activ-
ities at the site was determined to be greatest within the bedrock
aquifer. VOCs identified included trichloroethylene (TCE) and
associated daughter products: trans- and cis-dichloroethylene
(DCE) and vinyl chloride. 1,1,1-trichloroethane (TCA), which re-
placed TCE around 1975, also was found in the groundwater.
The TCE contamination plume within the bedrock aquifer is
shown in Figure 1. TCE concentrations were much higher than
concentrations of the other compounds. The bedrock contamina-
tion is the result of a non-active source located southeast of the
manufacturing building. The resulting plume is migrating in a
northwesterly direction in response to groundwater flow.
255,^
26S t
26B
00° GROUNDWATER
' ROW DIRECTION
SC*LE IN FEET
0 100 200
o Recovery Well Locotion
• Monitoring Well Locotion
S=0verburden Well
B=Bedrock Well; Upper 10' of Rook
BS=Shollow Bedrock Well; 0'-8 Below Top of Rock
Bl-lntermedlote Bedrock Well; B'-16' Below Top of Rock
BD=Deep Bedrock Well; 16'-25' Below Top of Rock
Surfoce Impoundments (RCRA Units)
NOTE: TCE' Concentrotlons ore In ug/L
Figure 1
Location of Monitoring Wells; Trichloroethylene Concentrations
(in ug/1) in the Bedrock Aquifer
REMEDIAL ACTIONS 469
-------
furn ir, Section
EAST
S30 -
510 -
WRST NORTH
SOUTH
V'
GLACIAL TILL
SANnSTO'.T
VKl
«80
47ff
HORIZONTAL SCAtŁ
Figure 2
Geologic Crou-Section
PRE-FRACTURING AQUIFER TEST
Remedial measures were deemed appropriate after assessing the
magnitude and extend of groundwater contamination. Current
remedial alternatives were evaluated and it was determined that a
recovery well system would be an effective approach to mitigate
further groundwater degradation. In order to design an effective
recovery well system, it was necessary to further investigate the
hydrogeologic characteristics of the bedrock aquifer. Therefore, a
72-hr aquifer test was performed.
A recovery well was installed downgradient of the facility in the
centerline of the plume. The optimal location for the recovery
well was determined from in situ permeability test results on six
preliminary test borings. The recovery well was installed 25 ft into
bedrock at a total depth of 40 ft. The overburden material was
cased off and the bedrock section was left as an 8-in. diameter
open hole. The well was installed to a depth of 40 ft as no signifi-
cant contamination was detected in any of the bedrock monitor-
ing wells below this depth.
Static water levels were measured at all monitoring wells on
site during the 3 days preceding the test to identify background
water levels and trends. During the aquifer test, water level read-
ings were obtained for 54 wells installed within the bedrock and
overburden. Sixteen wells were continuously monitored using
pressure transducers and associated data loggers. These included
the pumping well, bedrock monitoring well 23B and well clusters
28, 29, 30, 31 and 32. The remaining well water levels were meas-
ured using electronic water level indicators. Water levels also were
obtained in selected wells during a 4-hr recovery period after
pumping ceased.
The pumping rate for the aquifer test was set at 3.4 gpm and
this rate was maintained throughout the test by monitoring the
rate at 30-min intervals. The water generated during the test was
treated using an air stripper and carbon adsorption unit in the
series to remove VOCs. In accordance with a temporary SPDES
(State Pollution Discharge Elimination System) permit, the ef-
fluent, with total VOC concentrations at the non-detectable level,
was released to a nearby canal.
The water level in the pumping well dropped approximately 20
ft during the test. Rapid response to pumping was noted in moni-
toring wells 32B1 and 31BD, with the water level in these wells
essentially mirroring the water level in the pumping well. Other
wells within these two clusters showed little response to pumping.
With the possible exception of wells 29BD and 30BI, water levels
in clusters 28, 29 and 30 did not appear to have been influenced
by the pumping. Semi-log drawdown curves for the recovery well
and clusters 29, 31 and 32 have been included as Figure 3. Wells
within clusters 28 and 30 responded similarly to wells in cluster 29
(i.e., little or no response).
The irregular responses of individual wells within clusters 31
and 32 reflect the complicated three-dimensional capture zone
created by pumping within the fractured bedrock aquifer. This
effect is particularly troublesome when realizing that verification
of the recovery well's capture zone would be essential in determin-
ing the effectiveness of the corrective action. Based on existing
data, a meaningful mathematical prediction of the capture zone
associated with the recovery well would be both exceedingly diffi-
cult and costly.
The system had not reached steady-state by the end of the 72-
hr test. The possibility exists that additional drawdown would
have occurredc at some wells under continued pumping. Al-
though budgetary, regulatory and logistical restraints precluded
extending the pumping period, the length of the pumping was suf-
ficient to develop an adequate understanding of the bedrock
hydrology.
470 REMEDIAL ACTIONS
-------
Based on the results of the pre-fracturing 72-hr aquifer test, the
following observations and conclusions were made:
• Variable response to pumping (i.e., drawdown) in monitoring
wells, even within clusters, indicates that monitoring wells are,
in general, hydraulically poorly interconnected
• No response to pumping was observed hi any monitoring well
located upgradient of the recovery well
• Delineation of the capture zone is extremely difficult due to the
irregular responses observed in the monitoring wells
• The single recovery well installed and tested would not ade-
quately prevent further migration of the contaminant plume
• The installation of additional recovery wells would not be a
particularly cost-effective approach to creating a well-designed
capture zone
CONCEPTUAL DEVELOPMENT
In order to create an effective capture zone, the influence of the
pumping must extend to all of the fractures that were transport-
ing the contaminated groundwater. Initial ideas aimed at meet-
ing this objective revolved around methods of increasing the num-
ber of fractures intersected by individual recovery wells. Options
explored included angle drilling or "frac-ing" wells. Angle drill-
ing is particularly effective if the fractures are relatively vertical
and closely spaced. The existence of such a fracture geometry was
not evident at the site. Frac-ing of wells is performed by using
either explosives or high pressure water in an effort to artificially
enhance existing fractures or create new fractures around individ-
ual wells. Both the shallow depth of the recovery well(s) and the
variable effectiveness of the frac-ing procedure warranted explor-
ing other alternatives.
An ideal solution to the problem would consist of a method
that would interconnect and drain all of the fractures transport-
ing the contaminants. The creation of a single, artificial fracture
oriented perpendicular to the direction of groundwater flow was
considered as an option. Such a fracture could be produced using
explosives positioned in a shot line similar to the pre-splitting
technique used to produce the neat face in road cuts. After creat-
ing the fracture, one or more recovery wells could be installed in
the fracture to remove the contaminated groundwater for treat-
ment. Two major concerns arose from this option. First, the pos-
sibility existed that complete interconnection along the fracture
might not occur. Second, the resulting fracture might not have
sufficient cross-sectional area to allow the drawdown necessary to
capture the plume.
To overcome the concerns of insufficient flow area and
hydraulic interconnection, a method was designed to create a
thoroughly fractured zone, several feet in width, within the upper
25 ft of rock. The rock within this zone would essentially be trans-
formed into rubble, thereby creating a highly interconnected
"drain" capable of transmitting substantial amounts of ground-
water. The fracture zone would be positioned perpendicular to
the direction of groundwater flow near the leading edge of the
contaminant plume. One or more recovery wells would be in-
stalled into the fracture zone to produce the desired draw-down.
The plume would be prevented from migrating further and con-
taminated groundwater downgradient of the zone would be
drawn back into the fracture zone for removal and treatment.
The concept of creating a fracture zone offered some major
advantages over a conventional recovery well network approach.
Foremost, verifying contaminant capture, often a difficult task in
fractured bedrock, would become much easier as the recovery
<
UJ
LU
>
I
SIS-
SOS-
495 H
485-
PUMPING RATE=3.4 GPM
RECOVERY WELL
(19.9')
10
-2
10
-1
—T~
10
10-
ELAPSED TIME (MINUTES)
o
t—
>
LJ
_J
%
QL
LJ
510-
soo H
490-
480-
(1.61)
WELL CLUSTER 31
1 1
10
-2
10"
1
10
10-
ELAPSED TIME (MINUTES)
ui
Wb-
515-
505-
1
Bl
(O.I1)
(02')
BD
(09'J
WELL CLUSTER 29
1 ' ' ' '
Q-2 10"1 1 10 102 105 104
ELAPSED TIME (MINUTES)
510-
O
t—
>
cr
LJ
t—
500-
490-
480-
WELL CLUSTER 32
10
-2
10
-1
10
10-
ELAPSED TIME (MINUTES)
Figure 3
Pre-Fracturing Aquifer Test Drawdown Curves for Selected Wells
REMEDIAL ACTIONS 471
-------
weU(s) would be directly connected to fractures along the entire
cross-section of the fracture zone. Assuming that the fracture
zone extends to the lowest depth of contamination, verification
would be reduced to assessing the extent of the capture zone
downgradient and on either end of the fracture zone. Because
fewer recovery wells are required, another advantage would arise
from the savings in operation and maintenance costs (e.g., well
redevelopment and pump replacement) that would be expected
with a recovery well network. Finally, higher pumping rates
would be possible and could result in faster remediation.
FRACTURE ZONE CREATION
Using a carefully controlled single line pattern blasting tech-
nique, a 6-ft wide, 300-ft long fracture zone was created in the
upper 25 ft of the bedrock aquifer perpendicular to the center-
line of the plume.
Prior to any blasting, utility companies were contacted and
plant diagrams were reviewed to determine any potential blasting
restrictions due to buried underground water mains, sewers,
cables, etc. The existence of a water main and a sewer line did re-
strict the length of the fracture zone to 300 ft. which included a
safety margin of 25 ft from each of the buried lines. The fracture
zone, as depicted in Figure 4, was positioned perpendicular to the
direction of groundwater flow and centered near the leading edge
of the contaminant plume.
It was necessary to restrict fracturing to the upper 25 ft of the
rock as significant contamination was not observed belows that
depth. It was estimated initially that approximately 30 Ib of ex-
plosives placed in 3-in. diameter shot holes would produce the de-
sired degree of fracturing. However, during the actual field activ-
ities, the amount of explosives loaded into each hole varied
according to the vibrationaJ impact of the previous blast. To re-
duce the potential for damage, shock waves resulting from each
blast were recorded by a seismograph that was positioned next
to the manufacturing building's foundation at the closest point
to the blast site. The maximum readings at the building did not
exceed 1.4 in./sec peak particle velocity which was well below die
normally accepted 2 in./sec. Maximum pounds/delay ranged
from 22 to 44 Ib.
Due to the high water table and relative instability of the un-
consolidated material, it was necessary to case each hole. In order
to accomplish this, two air track rigs were employed. A smaller
air track initially drilled a 5-in. hole into the top of rock and set a
4-in. OD. 3.5-in. ID steel casing. A larger air track then set op
over the hole and drilled a 3-in. hole 25 ft into rock. Austin
Powder Co., 2-in. by 16-in., 40* Gel Extra was lowered to the
bottom of the hole using premeasured cord with an electric cap
inserted into the bottom stick. The cord was used for purposa
of safety and to insure a full column shot. Each hole was drilled
and blasted before the next adjacent hole was drilled. A spacing
C3jl Jj
08 «^~S '
•*---^ 3.
° ««cov«ry W«l LOCOllon
• Monitoring w«li Locution
S-0>»rburd>n Well
B-B.drock Wtil; Uppv 10' of Rock
BS-ShollO" Btdrock Well; 0'-8' B«lo« Top of Rock
Bl-lnlwm>dlot< B.iJrock w.ll; B'-lfl' B«low Top of Rock
BD-D»«p Bedrock Will; I6'-2S' B«lo» Top of Rook
[__JSurtoc« Impoundment! (RCRA Unlti)
NOTE. TCE Concentrotlont or« In ugA
tou/ n rrn
Figure 4
Location of the Fracture Zone; Trichloroethylene Concentration*
(In ug/l) In the Bedrock Aquifer
472 REMEDIAL ACTIONS
-------
of 4 to 5 ft between holes was determined to be appropriate. If
fractured rock was not encountered at the next drilling location,
a new hole was drilled closer to the previously blasted hole until
fractured rock was encountered. The flexibility in spacing tke
holes associated with this "drill and blast" method allowed
immediate verification of the effectiveness of the blasting.
Once the explosives were in position and the hole was back-
filled with stemming stone, the lead lines were connected to a 450
VME condenser discharge blasting machine. The charge was
detonated from the bottom of the hole upwards using an electric
millisecond delay blasting cap. Little, if any, permanent surface
displacement occurred due to the blasting. Groundwater spouted
from the previously blasted holes for several seconds after each
blast. This spouting demonstrated the high degree of hydraulic
interconnection that had been created between blast holes.
The blasting program took 2 wk to complete and required 60
shot holes and approximately 2000 Ib of explosives. Extensive
fracturing is expected to extend several feet radially from each
shot hole with hairline cracks possibly extending as much as 10 to
15 ft. Fractures are not expected to extend below the bottom of
the shot holes due to the positioning of the explosives and the
detonating sequence. A cross-section of the fracture zone is
shown in Figure 5. The reduced depth of fracturing near each
end of the fracture zone is due to reductions in the amount of ex-
plosives used near the underground utilities.
POST-FRACTURING AQUIFER TEST
A second 72-hr aquifer test was performed approximately one
month after the completion of the blasting program. An effort
was made to simulate, as closely as possible, the conditions of
the pre-fracturing aquifer test.
Prior to the blasting, the steel casing of the recovery well used
in the first aquifer test was removed and the boring filled with
coarse sand. Following the blasting of the fracture zone, which
passed through the recovery well, the coarse sand in the boring
was reamed out to its original depth (40 ft). A 12-ft long, 6-in.
diameter, 0.060-in. slot stainless steel well screen was installed at
the bottom of the boring with the remainder of the well con-
structed of steel riser pipe.
Three 2-in. observation wells were installed at the ends of the
fracture zones to monitor water levels. Two of these wells (OW-1
and OW-2) were installed at the east end of the fracture; OW-1
was screened in the upper half of the fracture zone and OW-2 was
screened in the lower half of the fracture zone. This pair of wells
was necessary to verify that the entire vertical section of the rock
was thoroughly fractured.
If the anticipated degree of hydraulic interconnection was
attained, the response to pumping should be essentially identical
in these two wells (OW-1 and OW-2). OW-3, located at the west-
ern edge of the fracture zone, was installed to monitor the draw-
down at the opposite end of the fracture. Only one observation
well was positioned at this location due to the reduced depth of
fracturing at the west end of the zone.
Based on the response during development of the replacement
recovery well, an anticipated well yield of 20 gpm was determined.
This represents more than a five-fold increase in yield over the
first aquifer test (3.4 gpm).
As hi the first aquifer test, a portable treatment system consist-
ing of an air stripper and carbon adsorption tank in series was
utilized. Pressure transducers were again installed in the same
wells monitored during the pre-fracturing aquifer test as well as in
wells 30S, 30BD, 32S, OW-1, OW-2 and OW-3. Water levels were
again recorded throughout the test at all other wells on-site using
electronic water level meters.
A conservative pumping rate of 18.5 gpm was selected for the
second aquifer test. The water level in the recovery well dropped a
total of 11.2 ft during the 72-hr pumping period. Nearly identical
drawdowns were observed in wells OW-1, OW-2 and OW-3.
This "bathtub effect" emphasizes the high degree of interconnec-
tion created by the fracturing.
NORTHEAST
SOUTHWEST
520'
470'
460'
STATIC WATER LEVEL
500'
1 APPROXIMATE
DRAWDOWN
• 490' AFTER 72 HOURS
• 480'
• 470'
HORIZONTAL SCALE
0 80'
Figure 5
Fracture Zone Cross-Section
REMEDIAL ACTIONS 473
-------
Semi-log drawdown curves for the recovery well; observation
wells OW-1, 2 and 3; and well clusters 29, 31 and 32 are shown
in Figure 6. A comparison of results from the pre- and post-frac-
turing aquifer tests is presented in Table 1. Significant draw-
down was noted in all bedrock wells in clusters 28, 29, 30, 31 and
32 with more than 3 ft of drawdown occurring in 12 of the 15
wells. Drawdowns ranged from a minimum of 1.6 ft in 28BS to a
maximum of 11.2 ft in 31BD. In contrast, during the pre-frac-
turing aquifer test only three of those same 15 wells exhibited
drawdowns greater than 3 ft.
I
525-
515H
505H
tr
495-
PUMPING RATE-18.5 CPU
RECOVERY WELL.
OW-1.0W-2 tt OW-3
10~2 10~'
ill
1 10 I02
ELAPSED TIME (MINUTES)
103 104 I4 I4
Trtfel
Co«p*ilMM of Prt- ud PMl-Aqalfcr TM! RmUto.
Recovery Hell
OW-1
OW-2
OM-3
28BS
28 BI
28BO
Pra-Praeturino
Aquifer Tgat
3.4 apm
Maxinus Drawdown
(in f««t)
19.9
0.0
0.3
0.4
Amilf«r T««t
Haximun Drawdown
(in feet)
11.2
11.1
11.3
10.9
1.6
4.6
4.6
525
a 515-
u
*
a 505H
4
* 495-
WELL CLUSTER 29
10
-2
10
-1
10
10*
ELAPSED TIME (MINUTES)
10-
10
444
29BS
29BI
29BD
30S
30BS
30BI
30BO
31BS
31BI
31BD
32S
32BS
32BI
32BO
0.2
0.1
0.9
0.3
0.6
1.8
2.3
1.6
2.6
19.4
0.7
0.8
19.0
5.0
6.0
3.8
9.4
1.3
2.4
6.3
5.1
7.8
9.9
11.2
1.8
1.8
11.0
5.9
525
a 515-
505-
* 495-
WELL CLUSTER 31
10~2 10
I I I 1
'1 1 10 102 10:
ELAPSED TIME (MINUTES)
10
4 ) t
An excellent response to pumping was again observed at 32BI
with a drawdown of 11.0 ft. Increased responses over the pre-
blasting aquifer test occurred in wells 32BD and 32BS. The water
levels in 32BI and 32BD dropped to within 1 ft of the water level
elevation in the fracture zone. Additionally, over a two-fold in-
crease in drawdown occurred in 32BS when compared with the
pre-fracturing aquifer test results.
Dramatic increases in drawdown occurred at wells 31BI and
3IBS as well as continued excellent response in 31BD. The sud-
den drop in the water level in 3IBS at approximately 600 min into
the test is believed to be due to a fracture "cleaning out" in re-
sponse to the pumping.
Drawdowns at cluster 29 ranged from 3.8 ft in 29BI to 9.4 ft
in 29BD as compared to 0.1 to 0.9 ft during the pre-fracturing
aquifer test. Similar drawdowns were experienced in clusters 28
and 30. As observed in the pre-fracturing aquifer test, no other
wells on-site were influenced by the pumping.
525-
^ 515H
Ul
a 505-1
or
* 495-
BS
(•.a-)
Bl
WELL CLUSTER 32
10~2 10~'
i 1 1 r
1 10 1(
ELAPSED TIME (MINUTES)
2 ,o3 io4
Pigurefi
Post-Fracturing Aquifer Tat Drawdown Curves for Selected WeUi
474 REMEDIAL ACTIONS
-------
CONCLUSIONS
A thorough understanding of the hydrogeology and nature and
extent of the contamination is necessary before creating an arti-
ficial fracture zone. Limitations of the technique are expected to
revolve around the thickness of the overburden, proper position-
ing of the explosives and the ability of the nearby buildings and
other structures to withstand the vibrational impacts caused by
the amount of explosives required to sufficiently fracture the
rock. With appropriate guidance, these limitations can become
manageable and the technique a viable alternative to other exist-
ing technologies.
The groundwater system had not reached equilibrium by the
end of the 72-hr pumping period. Additional drawdown is ex-
pected to occur when the permanent pumping and treatment sys-
tem is placed on-line and allowed to run for an extended period
of time. Some degree of dewatering of the overburden aquifer
is expected to occur over time. Continued monitoring of water
levels in the surrounding monitoring wells as well as groundwater
quality analyses will be necessary to evaluate the long-term effec-
tiveness of the remediation system.
The coupling of existing blasting technology with site-specific
groundwater remediation needs has produced an innovative re-
medial alternative. Through blasting, a selected zone of bedrock
has been essentially transformed into a conduit which directly
drains the individual fractures. A single recovery well should
prove to be fully capable of preventing further migration of the
groundwater contamination plume as well as capturing contam-
ination that has traveled downgradient of the fracture zone. As
only one recovery well was required by this technique, substantial
savings are expected in operational and maintenance costs.
Furthermore, the very nature of the fracture zone alleviates the
concerns associated with determining if individual recovery wells
are successfully intercepting all of the fractures transporting con-
taminated groundwater. This method should prove to be
applicable to many sites with contaminated fractured bedrock
aquifers.
REMEDIAL ACTIONS 475
-------
Successful PRP Remediation of the
Pepper's Steel and Alloys Site
Leslie R. Dole, Ph.D.
QUALTEC, Inc.
Knoxville, Tennessee
ABSTRACT
This study established the feasibility and the performance charac-
teristics of the on-site. in situ immobilization technology used for the
remediation of the 120jOOO yd' of heavy metal and PCB-contaminated,
trans-former-oil soaked soUs at the Pepper's Steel and Alloys (PSA)
Superfund site in Medley. Florida (N.W. Miami). Until its June, 1988
remediation, the PSA site was on the NPL. This 30-ac site, used for
a junk yard and scrap recovery operation, had used transformers
purchased from the Florida Power and Light Company (FPL). Subse-
quently. FPL became the major "deep-pocket" principal responsible
party.
The PCB-oil and heavy metal contamination on the site extended from
2 to 8 ft into the soils and into the Biscayne aquifer that is the sole
source of drinking water for the Miami metropolitan area. This site
is up-gradient within 6 mi of Miami's well fields
With full disclosure to U.S. EPA Region IV. FPL planned and con-
ducted a formulation, testing and modeling program that demonstrated
the safety of the monolithic grout formula poured directly into this
critical regional aquifer.
INTRODUCTION
This study established the feasibility and performance characteris-
tics of the proposed on-site stabilization/solidification treatment
technology1 using cement-based, pozzolanic monoliths for the oily Till
and peal at Pepper's Steel and Alloys (PSA) Superfund site in Medley,
Florida, which was contaminated with both heavy metals and PCBv
This preliminary grout development and leach-testing program was
conducted on behalf of the Florida Power and Light (FPL) Company.
The treatment of these contaminated soils by on-site solidification with
cementitious grouts was identified by FPL as a potential remedial action
alternative in its Final Report of its RI/FS in September, 1983 (Alter-
native Four). In FPL's Feb 20, 1985, letter to the U.S. EPA. this remedial
action was outlined in detail in a Draft Scope of Work Design Program
to accomplish the on-site stabilization or fixation of the PSA soils. In
the report, Energex R85-003. the interim status of this program was
reported to Region IV of the U.S. EPA on July 29, 1985 in u letter from
F. Mullins, FPL, to J. Orban, U.S. EPA Region IV
BACKGROUND
FPL evaluated several options before selecting the in situ solidifi-
cation as the remedy at the PSA site'. The cost of hauling the con-
taminated material 850 mi to a hazardous waste landfill was over $43
million, and several thousand trucks would have had to traverse one
of the most heavily traveled interstate corridors in the eastern United
States. Incineration was opposed vehemently by the local citizens, and
there are no incinerators which could meet air quality standards with
the high lead concentrations (to 100.000 ppm) in the site soils. Incinera-
tion would have cost close to $25 million.
Solvent washing of the soil cost approximately $16 million, but did
nothing for the vast quantities of heavy metals at the site. Dole then
proposed a plan to develop an in situ disposal option using cement-
based pozzolans to treat the site soils and to form large impermeable
monoliths, an option that cost $7 million.
Cement-based and pozzolanic materials are the most widely used
materials for (he stabilization of chemically hazardous and radioactive
wastes because they result in: 11) low-cost waste forms that are processed
with standard "off-the-shelf" equipment. (2) waste forms that resist
leaching and degradation in many geochemical settings and (3) high-
waste loadings with minimum waste volume increase when the grout
formulas are tailored to the specific waste streams'.
REMEDIAL INVESTIGATION
The soil collection plan, developed and conducted under the direc-
tion of Dr. Mason, is summarized in Reference 1. The results of the
analyses of these soils by RMT are also included in Reference 1. This
phase of the work plan was completed, and the collected PSA samples
were forwarded to Canonic Engineers to be used to develop stabiliza-
tion/solidification formulas.
The locations for the soil samples were selected on the basis of the
results of previous soil analyses for the presence of oil and the PCB
concentrations in the oil. The classes of soils to be collected were: (1)
dry Till, (2) oily Till and (3) oily peat. Some of the oil collected from
the PSA site contained up to 2,000 ppm of PCBs (Aroclor 1260). These
results are summarized in Table I.
Table 1
Summary Analytical Results for PSA SoUs
% Solid
Dry Fill Oily Fill Oily P«»t
Water 17.0 45.0 107.0
Oil & Grease 1.2 2.6 3.8
PCB*, ppm 42.0 116.0 44.0
Lead, ppofl.6980.0 1030.0 836.0
* PCBs were Aroclor 1260
476 REMEDIAL ACTIONS
-------
TREATABILITY
The success of the proposed in situ monolith at PSA was based on
the establishment of performance objectives which included: (1) mixing
and emplacement characteristics, (2) curing rates and the timely
development of adequate physical properties and (3) leaching of con-
taminants at rates that protected the public.
A detailed formulation and testing program to screen materials and
verify that the performance objectives were achieved. The formula
screening and treatability study tasks5 included:
Task 1 Screening available materials from South
Florida, based on their availability, cost, worka-
bility and physical properties
Task 2 Testing compressive strength penetration
resistance and permeability
Task 3 Leaching and Durability Testing:
(1) EP-TOX (U.S. EPA SW-846)
(2) Modified MCC-1 Static Leach Test
(3) Modified ANS 16.1 Multi Extraction
The completion of Task 1 was reported in the interim Status
Report2, and the results of the remaining tasks are reported in the
Final Report1. These studies selected the dry-solids blend that is
summarized in Table 2.
Table!
PSA Soil-Grout Formulation
Component
wgt (wgt %)
1. Soil Solids 1,680
2. Soil Water 340
3. Cement, Portland-I 300
4. ASTM Class F Ash 450
5. Mix Water 260
55
11
10
15
9
Solids Blend:Soil Ratio 0.45
Volume Ratio of Fixed Soil <1.1
The soil-grout's overnight penetration resistance and 28-day uncon-
fined compressive strength were < 500 and < 21 psi, respectively, and
were sufficient to allow unrestricted traffic and construction over the
buried monoliths.
The constant-head permeabilities, using a modified triaxial appara-
tus, on 28-day cured specimens of spiked dry-fill and oily-peat grouts
are summarized in Table 3.
Table 3.
Soil-Grout Permeabilities
Soil-Grout
Darcy Hydraulic
Permeability Conductivity
(cm/s)
(cm/s)
Dry Fill
Oily Peat
1 . 6E-8
6.5E-8
1.5E-11
6.2E-11
Since the permeabilities of the soil components at the PSA site ranged
between K>2 to lO^cm/sec, the monoliths range of permeability was at
least 10,000 to 1000,000 times lower than PSA soil. Therefore, these
grouts will be relatively impermeable; groundwater or precipitation can-
not percolate through these stabilized masses6.
LEACHING
Then, two series of leach-test specimens were prepared with spiked
PSA soils. A composite PSA soil sample was prepared from the known
heavy metal "hot-spots." The soils were spiked with oil collected at
the PSA site in the summer of 1983. To this oil sample, FPL's labora-
tory had added more Aroclor 1260 in order to increase its PCB con-
centration to 3000 ppm. Based on the soils' oil and grease analyses
and using this spiked oil sample, the total oil concentrations of the soil
for the leach-test specimens were adjusted to 10% (PCBs to 490 ppm)
based on the soil solids (see Table 4.). This spiking was done in order
to ensure that the treatability tests were done with samples that exceeded
any expected oil and PCB contaminations at this site.
Table 4
Content of Oil-Spiked PSA Soil
Water
Oil & Grease
PCB**
Lead
15.6 %
10.0* %
490.* ppm
31,490. ppm
(216 original)
Two series of right-circular leach specimens prepared with spiked
soil, having surface areas of 100 cm2 and 30 cm2 respectively, were
analyzed for the organic and heavy metal in separate leaching tests.
These specimens then were leached by the MCC-1 and ANS 16.1
methods in PSA groundwater. Also, the 40 CFR 261 structural-integrity
and EP-TOX tests were performed.
Both the Modified MCC-1 and ANS 16.1 methods can measure an
effective-diffusion coefficient (De, cm2/sec) that conservatively esti-
mates the maximum credible release rates of contaminates from the
monolith7.
The effective diffusion coefficients (De) for the soil-grout, used in
the in situ monolith at the PSA site, are summarized in Table 5.
Table 5
ANS 16.1 teachability Indices
Element
-LOG[De]
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
PCBs*
15.9
>13.0 **
>11.9 **
>12.8 **
13.5
> 9.2 **
> 9.8 **
> 8.4 **
>14.0 **
* PCB Aroclor 1260
** leachate concentrations below detection
limits
The leach tests showed that only arsenic and lead were above the
detection limits in the leachates. In the fixed, spiked-fill soil-grout,
cadmium, chromium, mercury, selenium and silver were below the
detectable limits after 28 days of leaching the cured solid. More im-
portant, these low effective-diffusion coefficients predict infinitesimal
source-term of potential contaminates diffusing from large in situ
monoliths.
REMEDIAL ACTIONS 477
-------
Figure 1 describes the general case of diffusion from a semi-infinite
solid and will conservatively predict the maximum credible release rates
from a submerged monolith1".
Model
F := 2 •-•
De-
af i
F = Cumulative Fraction Released
S = Spec!Men Surface Area
U =- Specinen Uoluiw
De = DIffuss Ion Coefficient of 1
t - Leaching Interval
Figure I
Semi-infinite Slab Diffusion Model thai
Conservatively Over-estimates the Cumulative Fractions Released
RESULTS
The PSA soil fixation blend development and testing program has
achieved the goals of the fixation/stabilization work plan by success-
fully identifying a formula to fix the U.S. EPA. priority metals and PCBs
in place at Medley, Florida. Using materials from South Florida, this
study selected an initial 60/40 fly-ash/cement blend, shown in Table 2.
based on its engineering properties of cost, mixability, set time, com-
pressive strength and permeability.
Then, two series of leach tests verified that the monolith's in situ
performance was adequate to protect the public's health. For example,
based on Equation 1 and the effective-diffusion coefficients. Figure 1
shows the leach fraction released over 1,000 yr.
When the small fractional releases from Figure 1 are integrated into
the PSA site hydrology and annual tropical rainfall for the first 1.000
y". the resulting maximum credible groundwater concentrations for
lead (Pb), PCBs and Arsenic (As) are very low. (Table 6)
PEPPER'S STEEL MONOLITH PD, PCBS. & As
Figure 2
Maximum cumulative fractional losses of Lead.
PCBs and Arsenic from the PSA monolith over 2.4 millenia.
laMe6
Resulting GroundwBler Concentrations After 1,000 yr
•t Stabilized Site.
ELEMENT
Concentration, ppm
Lead (Pb)
PCBs
Arsenic (As)
0.001
0.0004
0.00005
These worst-case concentrations are below current standard analy-
tical methods and below any known thresholds for health effects. There-
fore, based on these conservative overprojections of the maximum
credible concentrations. FPL was permitted to treat and solidify the
PSA metals and PCBs into a monolith that was poured directly into
the Biscayne aquifer without requiring a liner or a cap for the trench.
This monolith was located within 5 and 7 mi upgradieffl from the well
fields for the cities of Medley and Miami, respectively.
The Pepper's Steel and Alloys Site is the largest superfund site yet
to be closed. It was the first ROD to be signed after SARA was passed
in 1986, and it contained an innovative alternative technology for the
permanent disposal of non-volatile organks.
REFERENCES
I Florida Power and Light, Fuabon/SttbUuation Final Report: Pepper's Steel
and Alloys Site. Medley, Florida. Volumes I and 2. Florida Power & Light
Company. Juno Beach. FL. Nov.. 1985
2. Dole, L R , "Interim Slams Report Soils Fixation and Stabilization and
Remedial Action Alternative for The Pepper's Steel and Alloys Sue located
at Medley. Florida," Energex R85-003. Energex Associates. Oak Ridge, TN,
July. 1985
3. Florida Power and Light. Final Report: Remedial Investigation/Feasibility
Study. Florida Power ' Light Company. Juno Beach. FL, Sep., 1983.
4. Dole. L. R.. "Overview of (he Application of Cement-Based Immobiliza-
tion Technologies at US-DOE Facilities." Volume 2 of the Proc. of Mat
Management 85. Ed. Roy Post. Tucson. AZ. pp. 455-463, March, 1985.
5. Florida Power and Light. Remedial Alternative Fintuon/Stabiiizabon V*rk-
plan, Florida Power & Light Company. Juno Beach, FL, July, 1981
6. Atkinson. A. "The Influence of Wasteform Permeability on the Release of
Radionuclides from a Repository." Nuclear and Chem. Hiutr Management
5. pp. 203-2W. 1985
7. Gilliam. T M . Dole, L R and McDaniel. E. W.. "Waste Immobilization
in Cement-Based Grouts," Hazardous Solid Hbste Testing and Disposal: Sob
tolume, ASTM SIT 933. D. Lorenzcn, R. A. Conwty. L. P. Jackson, A.
Hamza. C. L. Per to. and W. J. Stacy. Eds.. American Society for Testing
Materials. Philadelphia. PA. pp. 295-307.
8. Godbee. H. W. and Joy, D. S. "Assessment of the Loss of Radioactive Iso-
topes from V&stes Solids to the Environment," ORNL/TM-4333, Oak Ridge
National Laboratory, Oak Ridge, TN, 1974.
9. Landreth. R E' Guide to the Disposal of Chemically Sutbiliird and Solidi-
fied Htisie.- EPA/SW-872, EPA Municipal Environmental Research Labora-
tory. Cincinnati. OH, September, 1982, Revised Edition.
10. Dole, L. R. "Leach Testing of In Situ Immobilized Soils Contaminated with
PCBs and Lead," Papers Presented at the 194th National Meeting of the
American Chemical Society, Symposium on Leach Testing for Radioactive
and Chemically Hazardous Wastes: Mass Transport and Chemical Reac-
tions, New Orleans, LA, pp. 283, August, 1987.
478 REMEDIAL ACTIONS
-------
Remediation at the Verona Well Field Superfund Site
Joseph P. Danko, RE.
William D. Byers, RE.
James E. Thorn
CH2M HILL, Inc.
Corvallis, Oregon
ABSTRACT
The Verona Well Field (VWF) supplies potable water to most of the
City of Battle Creek, Michigan, three townships, and another small
city. The combined service area equates to a population of approxi-
mately 50,000 people. In 1981, the well field and surrounding area were
found to be contaminated with vola tile organic compounds (VOCs),
and principally with chlorinated solvents. The contaminant plume
extended throughout an area of approximately 0.5 to 1 mi. Two facili-
ties run by a local solvent wholesaler were identified as major sources
of contamination. VOC concentrations as high as 1,000 mg/L were found
in groundwater and soil on the facilities' properties. The U.S. EPA chose
groundwater extraction with treatment in combination with enhanced
volatilization using soil vapor extraction (SVE) to clean up contami-
nated groundwater and soil at the site.
The groundwater extraction system included nine extraction wells with
associated instrumentation and controls, extraction force main piping
from extraction wells to an existing air stripper and a GAC pretreat-
ment system. From March, 1987 to August, 1989, approximately 11,000
Ib of VOCs were removed. Groundwater concentrations initially were
as high as 19,000 ug/L total VOCs; by August, the concentration had
decreased to approximately 2,000 ug/L. An extensive monitoring pro-
gram provided analytical data to evaluate compound-specific perfor-
mance in the air stripper and carbon adsorption pretreatment system.
The SVE system consists of 23 vapor extraction wells, two blowers
(30 hp and 40 hp) and a vapor-phase carbon emission control system.
The full-scale SVE system has been operating since March, 1988,
resulting in the removal of about 40,000 Ib of VOCs thus far.
INTRODUCTION
The Verona well field supplies potable water to residents and com-
mercial establishments in Battle Creek, Michigan. In August, 1981, it
was discovered that a number of private and city wells in the well field
were contaminated with volatile organic compounds. Subsequent testing
revealed that nearly half of the city's potable water wells were con-
taminated.
In the fall of 1983, a remedial investigation was initiated to deter-
mine the extent and potential sources of the well field contamination.
The investigation revealed a contaminant plume with VOC concentra-
tions varying from 1 ug/L to 356 mg/L in the area of the well field.
Monitoring revealed the plume was steadily moving towards less con-
taminated wells. The investigation also revealed three major potential
sources of contamination; two of them are sites operated by a solvent
distribution center, and the third is a railroad car repair shop (Fig. 1).
The lithology at the site consists of fine- to coarse-grained sand with
trace clay, silt and pebbles. The water table is approximately 25 ft be-
low grade level and the hydraulic gradient is to the northwest.
GRAND
TRUNK WESTERN
RAILROAD
MARSHALLING
YARD
THOMAS
' SOLVENT
/ , ANNEX
' /
Figure 1
Vicinity Map: Verona Well Field, Battle Creek, Michigan
REMEDIAL ACTIONS 479
-------
Remedial Measures
In May, 1984, the U.S. EPA signed a ROD lo implement an Initial Remedial
Measure (IRM). As part of the IRM, a series of potable wells was converted
to blocking wells to prevent further migration of the contaminant plume.
An air stripping system to remove VOCs from the contaminated groundwater
also was designed and built. In addition, three new potable walcr wells were
installed to supplement the city's water supply system.
In 1985. the U.S. EPA signed another ROD that addressed the major source
of contamination. The ROD specified a corrective action thai included a net-
work of groundwater extraction wells to remove contaminated groundwaler, the
treatment of groundwater via air stripping, and a soil vapor extraction (SVE)
system to remove VOCs from the unsaturatod zone.
Site Characteristics
The facility addressed in the ROD was an industrial site that had been used
for the storage, transfer and packaging of chlorinated and non-chlorinated sol-
vents from 1970 to 1984. As shown on Figure 2, there arc 21 under ground storage
tanks at the facility, 19 of which were confirmed to be leaking in a 1984 invcsii-
gation. These tanks are surrounded by heavily contaminated soil. Direct exca-
vation and removal of the tanks was not an option since that process would
seriously violate stale air quality criteria. This problem is being avoided by using
the SVE system to remove the majority of VOCs before removing the under-
ground tanks.
SVE PROCESS BUILDING
OFFICE (UILOINO
Figure 2
Location of Underground Tank*
Verona Well Field, Battle Creek, Michigan
At the time of the 1984 investigation, vadosc zone contamination
extended over approximately 40,000 fV, including the area around the
leaking underground storage tanks, in the tank truck loading/unloading
area and near the warehouse (now demolished).
In addition to the vadose zone contamination, there was a floating
product layer in the vicinity of Extraction Well 8 (EW8). This well is
a product recovery well, combining groundwater extraction with
intermittent removal of floating product as it accumulates.
GROUNDWATER EXTRACTION SYSTEM DESCRIPTION
The groundwater extraction system specified in the ROD removes
VOC-contaminated water from the aquifer in the vicinity of the most
contaminated source area. The system consists of nine groundwater ex-
traction wells, associated instrumentation and controls, approximately
5,200 ft of extraction force main (EFM) and a carbon adsorption system
that served temporarily as pretreatment for the existing air stripper.
Sampling ports are located at various points all along the system.
A flow schematic of the groundwater extraction system is shown in
Figure 3. Eight of the nine extraction wells discharge between 30 and
70 gpm of contaminated groundwaler; at one time, the ninth well (EW1)
discharged 5 to 7 gpm, but it currently is not operating.
Figure 3
Groundwaler Extraction System Flow Schematic:
Verona Well Field. Bailie Creek, Michigan
Contaminated groundwater is piped by the extraction force main from
the source area to the air stripping system (installed by the U.S. EPA
in 1984). Flow from the extraction force main discharges directly to
the air stripper pump station (wet well).
The carbon pretreatment system was installed in March, 1987 and
removed in January, 1988, when the total VOC concentration was low
enough for the air stripper alone to meet NPDES permit requirements
for discharge. When it was operational, the carbon adsorption system
consisted of three pressure carbon units (one in parallel with two in
series) located adjacent to the air stripper. Following treatment from
the carbon adsorption units, the water was discharged to the air stripper
pump station (wet well).
VOC-contaminated groundwater from the well field blocking wells
(approximately 1,700 to 2,000 gpm) also discharges into the wet well
along with the extraction well flow (approximately 300 gpm). From
the wet well, the water is pumped to an air stripper, which removes
more than 95 % of the VOCs from the water. VOCs removed from the
water by the air stripper are adsorbed from the stripper off-gas by a
vapor-phase activated-carbon system.
To date, the treatment system samples generally have been analyzed
for U.S. EPA Methods 601 and 602 purgeable organic target compounds,
either by the NUS Mobile Laboratory or the NUS Laboratory Services
Group facility in Pittsburgh, Pennsylvania. Periodic analyses were per-
formed for naphthalene (U.S. EPA Method 610). acetone (U.S. EPA
Method 656). and methyl ethyl ketone and methyl isobutyl ketone
(Methods 8015/8030). Samples were analyzed semi-annually for NPDES
Priority Pollutants. Treatment system sampling was done on a schedule
determined by the Michigan Department of Natural Resources (MDNR).
The sampling and analytical techniques used to monitor the opera-
tion lacked some of the sophisticated quality control measures normally
used in U.S. EPA CLP protocols. However, this potential limitation
on the analytical quality should not have a significant impact on the
overall data analysis or the evaluation of system performance.
Description of the Extraction Well System
The locations of the nine wells making up the groundwater extrac-
tion system are shown on Figure 2. The extraction wells are screened
480 REMEDIAL ACTIONS
-------
from approximately 20 to 37 ft below grade in the unconsolidated glacial
overburden unit. All extraction wells are 8 in. in diameter with the
exception of EW8, which is a 24-in.-diameter dual extraction well (re-
moves nonaqueous phase liquids and groundwater separately).
Performance of the Extraction Wells
By August, 1989, more than 375,000,000 gal of groundwater containing
approximately 11,000 Ib of TVOCs had been extracted through the
groundwater extraction system. This estimate of TVOCs removed is
probably low, since analyses were run only for priority pollutant VOCs
(see the "Glossary of Compound Abbreviations" at the end of this report
for a list of com pounds included in TVOCs). No analyses for total
organic carbon (TOC) or total petroleum hydrocarbons were made on
any samples.
The predominant contaminants by total mass are PCE, CIS/TRANS,
TCE, 1,1,1-TCA and TOL (for full names, see "Glossary of Compound
Abbreviations" at the end of this report).
Figure 4 shows the change in concentration of TVOCs for the com-
bined flow from the extraction wells (sampling point WS1, from
Figure 2). Note that the figure uses both dates and days from startup
(0 to 900) to identify points in time for the extraction system. Figure 5
shows the cumulative amount of TVOCs removed by the extraction well
system.
100 200 300 400 500 BOO 700 BOO BOO
i i i I S I i ! I
S S S s 5 § S 5 i
DAYS FROM STARTUP
Figure 4
Concentration of Total VOCs from Combined EW Flow
200
i
DAYS FROM STARTUP
Figure 5
Total VOCs Removed by GW Extraction
Description of the Carbon System
The temporary carbon adsorption system consisted of two parallel
trains, one made up of two units in series, and the other a single column
adsorption unit. This arrangement provided more flexibility than housing
three units in parallel and resulted in less pressure drop than three units
in series.
Each unit was 10 ft in diameter, and 12 ft high and contained 20,000
Ib of granular activated carbon. This amount of carbon was estimated
to be sufficient for the entire period that the carbon system would be
needed. Thus, no on-site carbon storage was needed and none of the
units had to be taken out of service for carbon replacement. Flow was
distributed to provide approximately one-third of the total flow to the
single-unit train and the remaining flow to the train with two units in
series.
Performance of the Carbon System
An estimated 1,830 Ib of TVOCs were adsorbed in the single-unit
train and 4,340 Ib in the two units in series train, for a combined total
of 6,170 Ib of TVOCs adsorbed.
Several compound began to desorb as the carbon beds began to load
with VOCs. Desorption occurs primarily as a result of competition
between compounds. Every compound has a different capacity for
adsorption. When the carbon is new, the differences in adsorption
capacity among the different compounds are barely detectable because
competition for adsorption sites is minimal. However, as the carbon
begins to reach capacity, various compounds begin to compete for the
available adsorption sites. As a result, weakly adsorbed compounds are
desorbed by competition from the more strongly adsorbed species.
Several of the compounds desorbed at some point in the operation of
the system; methylene chloride, vinyl chloride and 1,2-dichloroethane
were the only compound that did so at a substantial rate.
In January, 1988, the VOC concentrations in the groundwater appeared
to be low enough to bypass the activated carbon system and go directly
to the air stripper without violating effluent standards.
Description of the Air Stripper
The air stripper is composed of a single 10-ft-diameter tower con-
taining 40 ft of 3.5-in. pall ring packing in two 20-ft sections. The tower
is made of fiberglass-reinforced plastic with stainless steel internals and
polypropylene packing.
Water enters the top of the tower through a 12-in. header to a Norton
wier-trough distributor. Atmospheric air is pulled upward through the
tower counter-current to the direction of water flow. A demister removes
entrained droplets from the air at the top of the tower prior to discharge
to the vapor phase carbon adsorption units.
The stripper was designed for a nominal water flow rate of 2,000
gpm with a maximum flow of 2,500 gpm. The air flow system is sized
to deliver 5,000 to 6,000 acfm.
The air stripper is equipped with a recirculation system to permit
periodic addition of acid or disinfectant used to control accumulation
of inorganic scale or biological growth on the packing material and
internals.
Performance of the Air Stripper
Air stripper performance was monitored as part of the data-taking
program. Most of the data were taken while the carbon adsorption
pretreatment system was operating. However, data also were taken when
the pretreatment system was being bypassed and after its removal in
January 1988. This set of data is presented under "Performance Without
Pretreatment," after the discussion of the larger data set.
Performance With Pretreatment
Air stripper performance generally is reported as percent contaminant
removal efficiency. The efficiency is determined by taking the difference
between the influent and effluent concentrations and dividing it by the
influent concentration.
This definition of performance poses some computational problems,
particularly when the influent and effluent concentrations are below
the detection limit. For data where the influent and/or effluent was be-
REMEDIAL ACTIONS 481
-------
low detection limits, removal efficiency is reported as "NA" (not
available).
Table 1 summarizes those data points where both influent and effluent
concentrations were above detection limits. Results are reported as the
average of such data points for each compound and are compared against
results predicted by an air stripper model created by CH2M HILL. The
number of data points used in each average is also reported.
Table 1
Air Stripper Performance
With Pretraataant
Jtlihaul,
Compoundt
CCL4
CCL3
1,1-DCA
1,2-DCA
1,1-DCE
CIS
TRANS
MECL
PCE
1,1,1-TCA
TCE
vum.
IE!
E1EI
TOL
o-m.
Predicted
Renoval
99.«
(0.2
91.1
37.4
99.6
62.9
98.4
72.1
99. J
99.7
97.7
99.9
90.9
94.6
93.1
15. «
Actual
Renoval
U)
66.7
NA
NA
31.7
NA
• S.9«
72. »
83..
73.9
97.6
57. S
HA
NA
95.0
NA
Number
of Dete
Point l
1
HA
NA
33
NA
27«
37
2
13
1
1
NA
NA
1
NA
"^"•"»" * — * AAft
Actual
Renewal C
Jll
HA
NA
NA
46.0
NA
86. J
68.1
97.9
94. S
96.0
NA
61.2-
77.7
9S.o
91.8
Influent
one ant rat ton
Ippb)
NA
NA
NA
16
NA
11
8
7
3
12
NA
J
2
13
4
•Reaulta for CIS and TRANS combined.
•One «« 3.61.
•ORSi Air atrlppar vatar (lav - 2,400 gpo.
HA - lot available.
Conpound
oethylene ,
oroethylene '
Carbon letrachlorIde
Chlorofom
1,1-Dlchloroethene
1,2-Dichloroethane
1,1-Dichloroe t hy1ene
Ci«-l,2-Dichloroethylene
Tren>-l,2-Dlchloroethyle
Methylene Chloride
Tetrechloroethylene
1,1,1-Trlchloroethene
Trichloroethylene
Vinyl Chloride
Benzene
Ethylbenzene
Toluene
o-Xylene
Abbreviation
CCL4
CCLJ
1,1-DCA
1,2-DCA
1,1-DCE
CIS/TRANS
MECL
PCE
1,1,1-TCA
TCE
VINYL
BEN
EBEN
TOL
0-XYL
Three compounds (1,2-DCA. CIS/TRANS and MECL) had more
than 25 data points that could be used to compute an average removal
efficiency. All of these averages showed reasonable agreement with
predicted results. Thirteen data points were available for 1,1,1-TCA,
which showed performance much lower than predicted (74% versus
99.7%). Other results were based on only one or two data points and
showed mixed results.
Performance Without Pretreatment
Computed average removal efficiencies for CIS/TRANS, PCE and
TOL agreed reasonably well with predicted values for these compound,
which all had seven or more data points (see Table 1). 1,2-DCA (16
data points) and MECL (8 data points) both had computed removal
efficiencies greater than predicted values. All other compounds had
less (nan five data points.
SOIL VAPOR EXTRACTION (SVE) SYSTEM
Description of the SVE System
The SVE system was installed to remove VOCs from the vadose zone
in the vicinity of the most contaminated source area. Figure 6 snows
a simplified schematic of the SVE system. The system consists of a
network of 4-in. diameter PVC wells with slotted screen from approxi-
mately 5 ft below grade to 3 ft below the water table. The wells are
packed with silica sand, sealed at the screen/casing interface with ben-
tonite, and then grouted to existing grade to prevent short circuiting.
The extraction wells are connected by a surface collection manifold.
Each wellhead has a throttling valve, sample port and vacuum pres-
sure gauge. The surface manifold is connected to a centrifugal air/water
separator followed by a carbon adsorption system. The outlet of toe
carbon adsorption system is piped to a vacuum extraction unit (VEU),
which induces a flow of air from the subsurface into the extraction wells.
The vacuum not only pulls vapors from the unsaturated zone, but also
decreases the pressure in soil voids, thereby causing the release of
additional VOCs. After passing through the carbon adsorption system
and vacuum extraction unit, air is discharged through a 30-ft stack.
Note! Above compound* are priority pollutants tmtc-d at the «it».
482 REMEDIAL ACTIONS
Figure 6
Schematic of Soil Vapor Extraction System:
verona Well Field, Battle Creek, Michigan
The carbon adsorption system is operating with four parallel primary
carbon units (PCU) connected to four secondary carbon units (SCU),
also in parallel. The PCU are used for the majority of VOC adsorp-
tion, while the SCU act as a backup in the event of breakthrough in
the PCU. The canisters each hold 1,000 Ib of vapor-phase granulated
activated carbon and are connected to header piping with flexible hoses
and quick-disconnect couplings. A sample port, vacuum pressure gauge
and temperature probe are installed upstream, downstream and between
the carbon units, respectively. A cartoon monoxide meter also is installed
between carbon units to provide early detection in the event of com-
bustion in the primary carbon unit. If the CO meter reaches its set-
point, the VEU will automatically shut down until it is reset.
-------
The carbon system was installed under negative pressure to make
sure that VOCs would not leak. Preliminary testing determined that
carbon adsorption efficiency was equivalent under negative and posi-
tive pressure.
During operation, the PCU is monitored for breakthrough by an
in-line HNu (organic vapor detector). The setpoint of the HNu was
established after determining the relationship between total VOCs as
measured by the HNu and compound-specific concentrations as
measured by an on-site gas chromatograph. The monitored compound,
their detection limits, and their breakthrough criteria are listed in
Table 2.
Table 2
Discharge and Test Result
Primary Carbon Unit Discharge
at the wellhead. From March, 1988 through August, 1989, approxi-
mately 37,000 Ib of VOCs were removed from the soil, bringing the
total amount of VOCs recovered to approximately 40,000 Ib.
PCE
TCE
MECL
BEN
0.0024
0.0073
0.0406
0.0057
355
1,360
11,654
1,783
When the breakthrough concentration is exceeded, the PCU is
changed out. The carbon change consists of placing the backup carbon
system into primary service and installing an unused standby carbon
canister into the backup position. By installing fresh carbon in backup
service at each PCU changeout, the chances of breakthrough on the
backup system are minimized.
In addition to monitoring primary and secondary carbon outlet con-
centrations, concentrations also are monitored at the wellhead and the
combined inlet. Results are used to quantify the VOC loading and help
predict the rate of carbon breakthrough. Each time a sample is collected,
the following process variables are logged:
• Wellhead vacuum
• Wellhead flow (as measured with a rotameter)
• Vapor/water separator water level
• Pressures and temperatures throughout the system
Performance of the SVE System
A pilot-phase SVE system was started up in November, 1987. Figure 7
shows the location of the SVE wells and the piping layout. Individual
extraction wells were operated first to determine their radius of influence,
flow rate and initial extraction rate. All gas stream analyses were gener-
ated by the on-site gas chromatograph. Figure 8 plots SVE performance
at the Thomas Solvents Raymond Road facility.
The radius of influence was measured by recording the vacuum in
nearby SVE wells and in vacuum piezometers. A 1.25-in. water vacuum
was recorded 60 ft from an extraction well. Since the vadose zone con-
sisted of homogeneous fine- to coarse-graded sand with trace silt and
clay, the extensive radius of influence was not unexpected. In addition,
vacuum piezometers located between tanks to analyze the effect of tank
shielding showed at least a 2-in. water vacuum. In spite of the vacuum
between tanks, fullscale SVE wells were installed at the end of tank
clusters to further enhance axial flow between tanks.
The total mass of VOCs removed in the pilot test was measured by
gas stream analyses and verified by analyzing the carbon. After operating
the system intermittently over 15 days (total run time of 69 hr), gas
stream analyses predicted 2,866 Ib VOCs removed, and the carbon
analyses showed an average loading of approximately 16.7%, or 3,006
Ib removed. The 5% difference between the two methods of analysis
can be attributed to the uncertainties inherent in the analytical
procedures.
The average of VOC concentrations measured in the stack was 0.0666
mg/L. At an average stack flow rate of 500 cfm over the 69-hour pilot
phase program, approximately 4.6 Ib of VOCs would have been released
through the stack (indicating a 99.8% removal efficiency).
The SVE system began full-scale operation in March 1988. The
average SVE extraction well flow rate is 70 scfrn at 2 to 3-in. Hg vacuum
f FENCE LINE
«INCH HEADER PIPING
Figure 7
SVE Wells and Piping Layout
/
I
I
\
\
/
\
\
r^~
\
\
\
"^
,
•^
"X^
\,
a « 7 a 11 13 IR 17 i
I ? I 1 S 8 1 1 I
1 ' t I i j | 1 1
Operating Month
Figure 8
Thomas Solvents Raymond Road SVE Performance:
Verona Well Field, Battle Creek, Michigan
REMEDIAL ACTIONS 483
-------
Conclusion
The loading rate of total VOCs has decreased from an initial high
of approximately 45 Ib/hr to less than 5 Ib/hr. As shown in Figure 8,
the concentration of total VOCs has dropped from a high of 23 mg/L
to about 1.5 mg/L after a total run time of 117 days. The apparent NAPL
layer has not been present since October, 1988. Operation is expected
to continue into 1990, with the potential for removal of the 21 under-
ground storage tanks.
ACKNOWLEDGEMENTS
These remedial efforts have been completed as part of U.S. EPA
Superfund Operable Unit Remedial Actions under Contract No.
68-01-7251. This paper has not been subjected to the Agency's peer and
administrative review. Therefore, it does not necessarily reflect the views
of the Agency, and no official endorsement should be inferred. Simi-
larly, any use of specific names in the paper should not be viewed as
an endorsement. The authors would like to thank Loughney Dewatering
(the groundwater extraction contractor) and Terra 'vac. Inc. (the soil
vapor extraction contractor), for the sampling and analytical work and
for their cooperation in making this paper possible, and Alan Amoih
of CH2M HILL for his assistance in preparing this paper. Additional
thanks are extended to Margaret Guerriero, Remedial Project Manager
for the U.S EPA.
484 REMEDIAL ACTIONS
-------
Using Bar Code Inventory Control at a Drum Site
Steven D. Warren, Ph.D.
Elise E. Allen
MK-Environmental Services
Cleveland, OH
ABSTRACT
Ample technical guidance documentation is available to help plan
and implement safe and cost-effective drum removal from hazardous
waste sites, and techniques and equipment for handling drums are
becoming well established and readily available.
An additional necessity at most drum sites where inordinate numbers
of drums are encountered is a system of record-keeping that includes
all information about the physical characteristics and handling/dispo-
sition of each drum. This system should be computerized, if possible,
and allow for easy data retrieval and search programs. Information col-
lected would include a description of each drum, its contents, a listing
of any labels, locations where found and ultimate disposition (for both
drum and contents when disposed of separately).
For a removal action in Pennsylvania, a system of bar coding was
developed for a site containing approximately 45,000 drums. Using this
system, each drum was assigned a unique bar code number that was
used to track its characterization, movement, on'site and final disposi-
tion (treatment and/or disposal). This bar coding system was developed
to reduce the amount of information needing manual transcription, there-
by saving time and reducing errors. The data were entered into hand-
held computers in the field and automatically transferred into a PC data
base at the end of each day or shift. This system represents the innova-
tive use of a well-established technology (bar coding) that greatly reduces
the effort needed for information control and also eliminates errors
inherent in manually transcribing data at each step.
INTRODUCTION
This paper discusses a PC-compatible system of bar coding and
automated data acquisition that was developed by MK-Environmental
Services to inventory and track characterization information on approxi-
mately 45,000 drums to be removed from a site in Pennsylvania. An
accurate characterization of each drum on-site was important for two
reasons. First, the cost of the removal action would be apportioned based
on the number of drums identified belonging to individual Potentially
Responsible Parties (PRPs). Secondly, it allowed for cradle-to-grave
documentation verifying disposal of each drum.
In order to eliminate the time-consuming and error-prone task of
manual transcription of data, a unique bar code number was assigned
to each drum and menus of bar code choices were generated that con-
tained drum characterization information of a repeating nature. This
system allowed most data to be entered by wanding bar codes. Infor-
mation collected in the field was then automatically downloaded into
a PC data base at the end of each day.
EQUIPMENT
Data were collected in the field using D. A. P. Technologies' Microflex
PC 1000, environmentally-sound (sealed to keep out dust, water and
common industrial solvents), hand-held microcomputers with an
attached Ricoh ProScan bar code reader wand. This microcomputer
has 640 K of memory, 896 K of data storage memory, an 80C88
microprocessor running at 4.9152 MHz and operates on rechargeable
NiCad battery packs. Its operating system is MS-DOS, version 2.25.
The keyboard has 47 programmable keys with audible feedback: 4
preliminary keys and 43 multi-function keys. The LCD (Liquid Crys-
tal Display) emulates a one-quarter section of a Color Graphic Adapter
(CGA) Monitor, showing 16 lines of 16 or 21 characters and acts as
a window that can move over the entire CGA screen.
Sequential pairs of self-adhesive bar code labels were purchased for
application on each drum. Menus of bar codes for characterization were
developed using Tharo Systems Inc.'s EasyLabel Plus bar-code/label-
generating software. The data base was maintained using Paradox 3
(Borland International) software on an IBM PS/2, Model 70, micro-
computer connected to a Hewlett Packard LaserJet Series II printer.
DISCUSSION
Initially, drums were removed from their original locations (mostly
in stacks of several thousand drums) and staged in a manner that allowed
for easy access and efficient characterization. This process was accom-
plished by aligning the drums in rows approximately 18 in. apart (mini-
mum distance to allow people to move between rows). At this site there
was enough open space to permit staging and characterization of all
drums before beginning disposal operations. Where space is limited,
groups 01 drums could be staged, characterized and restacked or di-
sposed of immediately.
Secondly, unique bar code identification numbers were affixed to each
drum. This proved to be a problem, however, due to the variability in
the condition of the drums. Rust and oil or grease on the drum surface
made it difficult for a bar code label to adhere. Several methods of
surface preparation were tried, but the most efficient was to mechani-
cally wire brush an area of the drum large enough for the labels. Two
labels were applied to each drum (usually one on the side and one on
the end) to assure that handling activities which might render a single
label unreadable would not make identification impossible. In addi-
tion, identification numbers were printed numerically on each label
(in addition to being printed as a bar code), so that they could be entered
manually when they were not readable by the wand. Using this system,
no drums became unidentifiable due to both labels becoming unreada-
ble. It should be noted that open-top drums with unsecured lids should
have both labels applied to the side of the drum, rather than one on
the top, to avoid any problems associated with tops and drums becoming
separated. It may be cheaper and equally time-efficient to physically
write numbers on the drums rather than apply bar code labels. However,
this necessitates manually entering drum identification numbers into
the data base and would certainly increase the occurrence of entering
incorrect numbers.
Several methods for wire brushing drum surfaces were tried. Using
hand-held brushes was slow and ineffective at removing rust. Recharge-
able drills with brushes attached were fast and effective for surface
preparation, but only lasted 0.5 hr before becoming discharged. The
best method was using air-driven grinders equipped with brass brushes.
This unit was fast, did an excellent job of surface preparation, could
be operated indefinitely and eliminated any spark hazard.
EMERGENCY RESPONSE 485
-------
SITE LOCATION:
ZONE 1
OWNER:
ZONE 2
ZONE 3
ZONE 4
AXIS CHEMICAL
XYZOILCO
ACME SUPPLY
NONE
DRUM COLOR:
BLUE BLACK
MARKINGS:
RED
YELLOW
GRAY
ALCOHOL
CAUSTIC
FLAMMABLE
NONE
SURFACTANT
TOLUENE
WASTE
RCRA EMPTY?
YES
Figure I
Sample Barcode Menu Choices
Next, each drum was characterized and the information was entered
into the data base. Information to be recorded included: location where
the drum was found; size, color and type of drum; condition of drum;
any markings that related to the original contents of the drum; all
information on any hazardous waste labels found on the drum; iden-
tification numbers of any samples pulled from the drum; and the amount
and type of any contents found in the drum. Most of this information
was entered into the data base using menus of bar code choices of the
type shown in Figure I. If the appropriate information was not availa-
ble on the menus, it was entered manually using the keyboard of the
Microflex unit.
Characterization of empty drums consisted of recording a physical
description of the drum, all information written on the drum and that
it was RCRA-empty. For drums that were not empty, characterization
included the above plus a description of the contents and the volume
nt liquid contained by the drum. At the time of disposal, manifest
numbers were recorded so that each drum and its contents could be
traces from the site to the disposal facility.
486 EMERGKNCY RESPONSF.
-------
Three problems were encountered and resolved using this system in
the field. First, the internal batteries of the Microflex microcomputer
were not designed to power bar code wands. Normally, the micro-
computer uses minimal power other than to illuminate the screen and
can operate for several days. However, during sustained use with the
wand attached, the batteries were completely discharged in approxi-
mately 3 hr. Therefore, it was necessary to add to the system an addi-
tional 5-amp camcorder battery, worn on a belt, to power the wand.
With the added battery, the system operated for as long as 24 hr.
The second problem was associated with operating in direct sunlight.
Bright sun made it difficult, and sometimes impossible, to operate the
wands. When the bar codes were in direct, bright sunlight, the wand
did not receive enough contrasting light to function. There was also
a problem with sunlight "burning" the LCD screens of the Microflex
micro computers to the point where they were no longer readable. Other
units are available with screens that are less sensitive to light. To
eliminate both of these problems, beach umbrellas were used to protect
the bar coding operation from direct sunlight. Two-person bar coding
teams had little difficulty maneuvering the umbrella to keep the units
in the shade. It also had the added advantage of keeping personnel cooler
and increased the time between breaks in hot weather.
The third problem was that the wands were not environmentally
sealed. When working in rainy or wet conditions, water accumulated
inside the wands. The umbrellas helped to alleviate this problem, but
at times it was too wet to use the wands. At these times, all data were
entered by keyboard, and the operating program was altered to produce
choice menus similar to the bar code menus. However, this method
is more time-consuming due to the limited amount of the screen visi-
ble at one time, which necessitates extensive scrolling to view all of
the choices.
CONCLUSIONS
This method of drum characterization and information control is ef-
ficient, cost-effective and less prone to errors than manual transcrip-
tion. For sites at which a large number of drums (or any other objects)
need to be individually characterized and tracked, this method is ideal;
there is no paper to get wet, soiled or blown about, and there are fewer
opportunities for manual inaccuracies. When conditions make wanding
bar codes difficult or impossible, the system can be easily modified
so that prompt menus are displayed on the Microflex screen.
ACKNOWLEDGEMENT
This paper would not have been possible without the efforts of Jeffrey
Smith in initializing the bar coding system and offering helpful review
of the initial draft.
EMERGENCY RESPONSE 487
-------
A Discussion of the Use of a Computer Data Base
Management Program to Categorize Hazardous Waste Data
Patricia Chadwick
Ecology & Environment, Inc.
San Francisco, California
Roman Worobel
Blymyer Engineers, Inc.
Alameda, California
ABSTRACT
The hazard categorization system currently employed by the Technical
Assistance Team (TAT) at Ecology & Environment (E & E) has been
upgraded by the development of a computer data base management pro-
gram that categorizes waste streams quickly and accurately and
according to RCRA regulations.
The use of this program allows both chemists and other hazardous
waste management personnel to facilitate the categorization of hazardous
materials. In an emergency removal, with a portable computer at the
site, compatible materials can be grouped quickly with this program.
Large removals involving hundreds of drums can therefore be carried
out much more efficiently with this program.
The hazard categorization computer program is menu-driven and
requires very little training to use. It alerts the user when an invalid
or inconsistent entry has been made. Data entered previously or on
the current sample are easily corrected. The program has a number
of reporting capabilities including error reporting it, can be used at
multiple sites and it has menus for selecting different printers and disk
storage options.
This program has been distributed to and used by all U.S. EPA regions
across the country. The software program, STREAMLINE, has been
uploaded to the U.S. EPA Office of Solid Waste and Emergency
Response (OSWER) bulletin board for use by U.S. EPA staff and U.S.
EPA contractors. The types of waste sites on which the program has
been used include plating facilities and drum recycling companies.
INTRODUCTION
The hazard categorization system was developed to help field inves-
tigation teams characterize chemical wastes at the time of sampling.
The hazard categorization system is based on chemical testing procedures
which permit a qualitative determination of waste characteristics
according to RCRA (40 CFR, Part 261) and Department of Transpor-
tation (DOT) (49 CFR, Parts 171 and 172) specificatioas. The charac-
teristics include ignitability or flammability, corrosivity, reactivity and
EP toxicity'
Following the initial determination of the physical and chemical
characteristics, waste streams can be determined. In an emergency
removal, this determination allows manifests to be completed and the
wastes to be legally transported to a disposal facility. At larger removals,
involving several hundreds of containers of different materials, the hazard
categorization information facilitates bulking of the compatible materials.
Small, bench-scale testing ensures that the bulking is conducted safely.
The bench-scale testing also minimizes laboratory costs involved when
accurately quantifying the waste streams for disposal, recycling or
treatment.
Previously, the consolidation of the waste materials could only be
performed by a chemist. The process was initiated following comple-
tion of the field testing for hazard categorization. Data sheets used tot
the collection of data were manually segregated based on the field testing
results. Not only was this procedure very laborious but also it requited
the full-time on-site presence of a trained chemist. It was at this time
that the E & E TAT upgraded the hazard categorization system by
developing a computer data base management program that categorizes
waste streams quickly and accurately and according to RCRA and DOT
regulations. The name of the computer data base management program
is STREAMLINE.
COMPUTER SYSTEM
Description
The software, STREAMLINE, was developed using physical and
chemical principles which categorize the material as a hazardous waste
according to RCRA and/or DOT. Information retrieved from field testing
is input into an on-screen computer data sheet (Fig. 1) which contains
attributes of the container, results of the field testing for hazard categori-
zation and the processed hazard designation of the hazardous waste.
STIUAMUKH DATA EHTUT
SAMPLE ID |OOO*A| (A)ll/(T)op/(»)ollo«
Container Attribute*:
TTPE (V/DVC) |0|
SIZE (Gallons) I IS. 00|
TOP (0)pen/(l)ung |0|
AMOUNT (1-Full/O-Bapty) j >i|
NATMI |S|
((S)olld/(L)lquld/(G)as)
OXIOIZKX (Y/N) |N|
CTAMIOB (Y/N) |N|
SAMPLE TAHHT |T|
|S|
CONDITION |P|
<(P)oor/(F)«lr/(C)ood)
SOLUBLE (Y/G/L) |T|
((Y)es/(G)real«r/(L)ess)
SULFIDB (Y/N)
BIC (Y/N)
CHUXUNK (Y/N)
INI
(HI
LABRL IFenic chloride. S»pl* taken off floor of aobile I
CONNKNTS jhoM near dru« described above. Tan sandy solid. 1
LOCATION (Mobile Ho»| ACT. TAKEN I I
HAZARD CLASS (AS I
Figure I
Data Entry Screen for Hazardous Waste Samples
The processed hazard designation is either a two or three letter code
which corresponds to a specific hazard class, e.g., "AOL*% references
an acid oxidizing liquid. Once this step is completed for all the samples,
the data can be sorted by hazard class and groups of samples with similar
physical and chemical properties can be determined.
Software Design
The STREAMLINE program is written in the dBASEIIH- data base
488 EMERGENCY RESPONSE
-------
management programming language and compiled by Clipper. The com-
piled version stands on its own and does not require dBASEIII+ to
run. However, all the data are stored in standard dBASEIII+ files to
make them easily accessible and transferable if desired. The software
system includes one executable file, 14 report form files (.FRM), six
system data base files (.DBF) and a data base file and two index files
for each site for which a categorization is run.
The main system file, HAZSYS.DBF, contains the disk drive and
code of the site categorization that is currently in progress.
HAZPRNT.DBF contains printer numbers and names and the printer
codes for compressed and regular printing. The data glossary and classi-
fication code definitions are stored in the HAZCLAS.DBF file. Table
1 contains a list of the contents of this file which is printed with every
sample data report.
Table 1
Report of Hazard Class Codes and Data Glossary
Class
Code Hazard Class Description
AL Acid Liquid
AOL Acid Oxidizing Liquid
AOS Acid Oxidizing Solid
AS Acid Solid
BL Base Liquid
BOL Base Oxidizing Liquid
BOS Base Oxidizing Solid
BS Base Solid
CLG Chlorinated Gas
CLL Chlorinated Liquid
CNG Cyanide Gas
CNL Cyanide Liquid
CMS Cyanide Solid
PG Flammable Gas
FL Flammable Liquid
FS Flammable Solid
NCG Non-Characteristic Gas
NCL Non-Characteristic Liquid
DCS Non-Characteristic Solid
NFL Non-Flammable Liquid/Oil
NS No Sample Taken
OG Oxidizing Gas
OL Oxidizing Liquid
OS Oxidizing Solid
SG Sulfide Gas
SL Sulfide Liquid
SS Sulfide Solid
The system can handle and store categorization data for an unlimited
number of sites. Each time a new site is begun, a new data file is created
specifically to store data for that site. Each data file has a unique name
created from the code that identifies that site. At the time the file is
created, the user specifies the letter of the disk drive on which the data
will be stored. This technique allows the files to reside on floppy or
hard disks, or both, and to be easily transferable between computers.
The site code and name and the disk drive letter are added to the site
master file, HAZSITE.DBF. A list of the all sites that have been entered
into the system can be displayed or printed (Table 2).
Table 2
Report Listing All Sites For Which Data Has Been Entered
List of All Sites in STREAMLINE Database
class, and H2CA0314.NTX which indexes by sample number. See Table
3 for a list of the fields in this data file and the field attributes.
Tables
Contents and Description of the STREAMLINE Sample Data File
Data Glossary
SAMPLE ID NO. :
CONTAINER TYPE
SIZE : Size of
AMOUNT : 1
0
0
(material that
the container
is made of)
CONTAINER COND.
CONTAINER TOP :
MATRIX :
CONTAINER MAT. :
SOLUBLE :
A = All material
T = Top portion
B = Bottom portion
• V
D
C
cont
.00
.75
.00
G
P
F
: P
Vat
Drum
Container < 55
iner in gallons
Full
3/4 Full
Empty
Glass
Poly
Fiber
Poor
F Fair
G Good
0 Open
B = Bung
S Solid
S = Steel
L Liquid
G = Gas
Y Soluble in H20
L « Floats in H20
G = Heavier than
PH - 15 if material is insoluble in
vater such that
soluble = L or G
Site ID No. Site Name
AZ0055 Bugfree Pesticide Co.
CA1000 ABC Recycling & Salvage
CA1052 Aladdin Barrel & Drum Company
NV0206 Nevada Auto Parts
NV0777 Phoenix Barrel & Drum
Data
Drive
C:
C:
A:
C:
A:
HZXXOOOO.DBF is the boilerplate data file from which new site data
files are created. The new data files names are determined by the site
code that is entered by the user. The site code consists of two letters
and four digits. The letters can be used to identify the state in which
the site resides; for example, "CA0314" identifies a California site. The
data file name would then be HZCA0314.DBF. There are two index
files associated with this data file: HZCA0314.NTX which indexes by
Field Name
SAMPLEID
TAKE SAMP
CONTTYPE
CONDITION
TOP
SIZE
AMOUNT
MATCONT
LOCATION
CLASS
MATRIX
SOLUBLE
PH
OXIDIZER
CYANIDE
SULFIDE
BIC
CHLORINE
LABEL
COMMENT
ERR_FLAG
ACTION
Field Type
Character
Logical T/F
Character
Character
Character
Numeric
Numeric
Character
Character
Character
Character
Character
Numeric
Logical
Logical
Logical
Logical
Logical
Character
Character
Logical
Character
Width
5
1
1
1
1
5
4
1
10
3
1
1
2
1
1
1
1
1
50
50
1
25
Dec
0
2
0
Description
Identifies the individual sample
Was sample taken? (Y or N)
Container Type: Vat, Drum, <55 ga]
Condition of container
G=Good, F=Fair, P=Poor
Was top open or bung?
Size (in gallons) of container
Amount of material in container
by dec. fraction (0 - 1.00)
Container material: Steel,
glass, poly, or fiber
Location on site of container
Hazard classification code
Is material solid, liquid or gas?
Soluble yes, greater, less
pH of material in container
Is the material an oxidizer?
Does material contain cyanide?
Does material contain sulfide?
Does material contain bicarbonate
Does material contain chlorine?
Label to be affixed to container
Comments/Description
Possible inconsistency in data
Action to be taken
Hardware Requirements
STREAMLINE can be run on any IBM-PC compatible computer
on either a floppy or hard disk. Because the program is compiled, no
additional software is required to run it. Currently, there are print settings
in the HAZPRNT.DBF file to run the program with either a HP Laser-
Jet, a Panasonic KX or an Okidata u93. If the HP option is selected,
the reports will print out in the landscape mode with compressed
printing. The dot matrix printers will print compressed (portrait mode)
on 8.5- x 11-in. paper. The print codes in the HAZPRNT.DBF file can
be modified for other printers through dBASEHU- if required. Approx-
imately 300K bytes of disk space are required for the program and stan-
dard system files and report forms. The additional amount of space
needed for the specific site files depends on the number of samples
collected and entered for that site.
Procedure for Using STREAMLINE
STREAMLINE is menu-driven and easy to use. The program is
started by typing HAZ at the DOS prompt. After displaying a title
screen, you are asked to select a printer from a menu. The program
then displays the code and name of the last site for which data were
entered or edited (Figure 2).
Pressing Fl at this point displays a list of all sites that have been set
up in the system. You then have the option to continue with the site
currently displayed on the screen, enter a new site code, edit the current
site's name or disk drive designation, or quit the program.
If you enter a new code, the system will check to see if the site is
already in the master file, or if data files exist for that site. If it is a
new site, you will need to enter a site name, and a new empty data
EMERGENCY RESPONSE 489
-------
file will be created. The program main menu is (hen displayed
(Figure 3).
STREAMLINE
SITE
I.D. (XX9999)
SITE NAME
DATA DRIVE
NV1006
Nevada
A
Barrel and Drum
Set Up Current/New Site/Bdit/Delete/Qult (S/N/B/D/Q) i
Press for List of Current Sites.
Figure 2
Site Selection and Update Screen
1 \ Select Another Site
2 \ Update/Edit Data
3 \ Re-Classify All Samples
4 \ Generate Reports
Q \ Quit to DOS
Figure 3
Main Menu of the STREAMLINE Program
The first option on the main menu is to select another site. Use this
if you are working on one site, then wish to enter or edit data for a
different site. Use the second option, Update/Edit Data, to add data
for new samples for the site or to edit data for samples that have al-
ready been entered in the system.
The screen format for entering the data into the computer is almost
identical to the data entry sheet from which the information is taken.
At the sample data entry screen, first type the sample ID number. If
the sample has already been added, the current information is displayed
and can be edited. Each item of information entered is checked so that
only allowable values can be entered. After all data for the sample have
been entered, the program determines into which hazard class the sample
falls. The code for the hazard class is displayed at the bottom of the
screen.
If inconsistent data are entered so that the system is unable to deter-
mine a hazard class for the sample, an error message is displayed, and
the user has the option to correct the data or leave it flagged for later
editing. For example, this might occur if the sample is marked true
for both cyanide and oxidizer. At this point the sample would be added
to an error file, and an algorithm would be used to determine both a
primary and a secondary hazard class. A list of all samples flagged
as errors can be printed through the report menu.
The third option on the main menu, Re-Classify All Samples, need
only be used if the program has been modified to change the way the
hazard classes are determined. The hazard class usually is determined
when data for the individual sample are entered or edited. If the pro-
gram algorithm for determining classes is changed, only those samples
that have been edited or newly added would use the new algorithm.
This option provides the opportunity to run the modified classification
program on each sample in a file automatically and re-classify it, if
necessary.
STREAMLINE Reports
The program provides a standard set of reports that can be selected
from the report menu. If different reports are desired (and the user
has dBASEIII+), additional report formats can be easily created at the
dBASEIH+ dot prompt or in the dBASEHI+ assist mode. The reports
available through the STREAMLINE report menu include: lists of all
the data for all samples on a site, listed either by sample ID number,
or grouped by hazard classes (Table 5); an error report showing all
the samples from a site that were flagged because of inconsistent data;
and a list of all the sites that have been set up in the STREAMLINE
site data master file. The hazard class code and data glossary are printed
with each report of sample results, as well as a report tided "Classifi-
cations of a Material Having More Than One Hazard as Defined in
Title 49" (Table 4). This report is used if the sample you entered falls
under more than one class. The primary class is the one that is higher
up on the list.
Table 4
Hierarchy Died for Determining Hazard Class Code
Classification of • Material lavlnc Nor* Than One Bazard
As Defined IB Title 49
Basard Ho. Description
1 Radioactive M«rUl (except • Halted quantity).
2 Potion A.
3 FlaiMble gas.
* Hon-f Iambic fas.
5 Plsanble liquid.
6 Oxldiier.
7 Plaseuble solid.
8 Corrosive nat«risl (liquid).
9 Poison ».
10 Corrosive aaterlal (solid).
11 Irritating aaterlals.
12 Combustible liquid (In containers having capacities > 110 f)
13 OKH-B.
14 OHM-A.
IS Combustible liquid (in containers bavin* capacities <-110 f)
16 OM-E.
CASE STUDIES
Aero Quality PUting Company
The initial site where STREAMLINE was implemented by the E & E
Emergency Response Section was the Aero Quality Plating Company
(APC), located in Oakland. California1. APC was an electroplating
facility which operated from 1958 to 1985. In October, 1985, APC filed
for bankruptcy under Chapter I) and later converted the filing to Chap-
ter 7. In April, 1987, the State of California of Department of Health
Services requested U.S. EPA assistance in the stabilization of the
hazardous wastes at the site.
The stabilization efforts were initiated by collecting over IjOOO sam-
ples and performing field testing for hazard categorization1. The field
testing results were then processed using STREAMLINE. The processed
information aided stabilization efforts by assisting in the consolidation
of on-site materials from both structurally unsound and sound containers.
The computer-generated report, grouping the samples by hazard classes,
simplified enforcement sampling. The report was used to zero in on
drums that were both structurally unsound and contained very hazardous
materials, i.e., those drums that were prime targets for enforcement
sampling. The stabilization efforts at APC were completed in June, 1987.
Lorentz Barrel and Drum Company
Another site where the STREAMLINE program was used was the
Lorentz Barrel and Drum Company (LB&D) located in San Jose,
California4. LB&D had been reconditioning used steel drums for
approximately 40 yr, since the 1940s. As a result of the operation, about
800 full drums of hazardous waste had been accumulated. In July, 1987,
the State of California Department of Health Services (DOHS) inves-
tigated LB'D and discovered that several State of California laws had
490 EMERGENCY RESPONSE
-------
been violated during the handling and disposal of hazardous waste. Oper-
ations at the facility completely ceased in July, 1987.
In September, 1987, the U.S. EPA, assisted by the E & E TAT, con-
ducted a site assessment at the facility as a result of the threat posed
to human health and the environment by leaking drums. U.S. EPA as-
sistance was requested by DOHS due to the lack of proper resources
needed by DOHS and local agencies to stabilize the site.
A total of 687 samples were taken. Field testing for hazard categori-
zation revealed the presence of such hazardous wastes as acids, caustics,
oxidizers, flammables and cyanides. The testing results were entered
into STREAMLINE and the data were then processed, assigning a
hazard class to each sample. The processed information later was used
to bulk compatible wastes and composite samples for laboratory analysis.
categorization, a chemist familiar with STREAMLINE can perform
both functions.
Reporting
The hazard categorization results are entered into a computerized data
sheet, processed and then assigned a hazard class by the computer. The
processed information can be sorted alpha-numerically and/or by hazard
class. The printed output varies with the user and purpose, but the most
common output is by hazard class. This output is most helpful when
bulking and/or compositing.
An error report is another type of STREAMLINE-generated report.
This report alerts the user to invalid entries or possible stratification
in the sample container. The user can then re-test the sample for a hazard
Table 5a
Report of All Data for All Samples-Sorted ? v Hazard Class
Sample Cont Container
ID Ho. Type Size Amount Volume Mat. Cond. Top Locale
Hazard
Class Matrix Soluble
Fh
Ox Cn
Sample
Sulf Bic Cl Taken?
** DATA FOR HAZARD CLASS : AOL
0017A D 55 0.75 41.25 S F B Semi tr. AOL L Y
0018A D 30 0.50 15.00 S F B Semi tr. AOL L Y
.T.
.T.
.F.
.F.
.F.
.F.
.F.
.F.
.T.
.T.
** DATA FOR HAZARD CLASS : AS
0004A D 15 1.00 15.00 S
0005A D 15 1.00 15.00 S
0010T
55 1.00 55.00 P
P O Mobile horn AS S
P O Mobile horn AS S
P O Lg cluster AS S
.F.
.F.
.F.
.F.
.F.
.F.
.F.
.F.
.F.
.F.
.F.
.F.
.T.
.F.
.F.
.T.
.T.
.T.
** DATA FOR HAZARD CLASS :
0001A D 55 1.00
BL
55.00
B Hr lab tr BL
13
.F. .F. .F.
.F. .F. .T.
** DATA FOR HAZARD CLASS : FL
0019A D 55 1.00 55.00 S
B Semi tr.
FL
.F. .F. .F. .T. .F. .T.
** DATA FOR HAZARD CLASS : NCL
0006T D 55 1.00 55.00 P G
0020A D 55 1.00 55.00 S F
O L? cluster NCL L Y 10
B Semi tr. NCL L Y 5
.F. .F. .F. .F. .F. .T.
.F. .F. .F. .F. .F. .T.
Hazard Data - Sample Conents - By Hazard Class
Sample Sample Hazard
ID Taken? Class Label/Comment
Action Taken
** Samples for Hazard Class : AOL
0017A .T. AOL Algae-green, clear liquid. EM Cu-0. HNU»3.
0018A .T. AOL AuS, Au Sol
Dark green clear liquid. EM Cu=0. HNU=2.
** Samples for Hazard Class : AS
0004A .T. AS Ferric chloride. Sample taken off floor of mobile
home near drum described above. Tan sandy solid.
0005A .T. AS Ferric chloride, anhydrous
100 1 Ib plastic bags in each drum. No hazcat done
BENEFITS
Cost Minimization
As a result of using STREAMLINE, the costs at a hazardous waste
site removal can be minimized. The cost savings can be attributed to
the bulking of compatible materials and the compositing of samples
for laboratory analysis. Cost control can also be achieved by minimizing
the need for personnel. The size of the field investigation teams can
be decreased by using individuals who can perform a variety of tasks
including STREAMLINE. For example, rather than assigning a chemist
and a data entry clerk to conduct an investigation field testing for hazard
categorization to prevent a chemical reaction during bulking or com-
positing.
Resource Utilization
As mentioned previously, chemists familiar with STREAMLINE can
perform a variety of tasks at a hazardous waste removal site. Another
aspect of resource utilization lies in the fact that other hazardous waste
management personnel who are familiar with field testing for hazard
categorization can be trained to use STREAMLINE. This allows other
personnel to become diverse in their duties and distributes the respon-
sibility among a number of personnel.
EMERGENCY RESPONSE 491
-------
Table 5b
Report of All Data for All Samples-Sorted by Sample I.I). Number
Haiard
Clatt Matrix Soluble
Ph
0001A
0002B
0003A
0004A
0005A
0006T
0007T
0008T
0009T
0010T
0011T
0012T
0013T
0014T
0015A
--!*--»-
D
V
C
D
D
0
D
D
D
D
D
D
D
D
C
55
3500
2500
15
IS
55
55
55
55
55
55
55
30
55
0
1
0
0
1
1
1
1
1
0
1
0
0
1
0
0
00
01
01
00
00
00
00
00
75
00
75
75
00
25
.00
55.
35.
25.
15.
15.
55.
55.
55.
41.
55.
41 .
41.
30
13
0.
00
00
00
00
00
00
00
00
25
00
25
25
00
75
00
S
r
r
s
s
p
s
s
s
p
p
p
s
r
s
r
p
p
p
p
a
r
r
p
p
p
p
p
p
p
B
0
0
0
0
o
0
0
o
0
0
0
0
0
Nr lab tr
NC lab tr
Mr lab tr
Mobil* horn
Mobil* ho»
Lg cluster
Lg clutter
Lg clutter
Lg clutter
Lg clutter
Lg clutter
Lg clutter
Lg clutter
Lg clutter
Under teal
BL 1
DCS
HCS
AS
AS
HCL
HCS
HCS
NCS
AS
AS
AS
HCS
RCS
RCS
L, I
t
r
Y
t
. T
T
T
t
T
T
T
T
T
*
13
10
S
2
2
10
9
7
6
1
1
1
10
10
10
ox
Saaple
Cn Suit Bic Cl Taken?
.F. .1
.F. .1
.r.
.r.
.F.
.r.
.F .
.F.
.F.
.r.
.f .
.r.
.F.
.F.
.F. .1
•
'.
.
.
.
.
.
.1
.1
•
.F.
.F.
.F .
f
f
f
f
f
f
f
.F.
r. .F.
', .F.
r . .F.
'. .F.
.F.
.F.
.F.
.T.
.F.
.F.
.F.
.F.
.F.
.T.
.r.
.r.
.r
.r
.T.
y
f
.T.
.T
.T.
.T.
.T.
.t.
Hacard Data Sejsple CoaMnta By Staple ID Bo.
Saaple Staple Haiard
ID
Taken? Clatt
Label/Coment
Action Taken
0001A .T.
00028 .7.
0003A .T.
BL
NCS
NCS
0004A
.T.
0005A .T.
0006T .T.
NCL
"Gold Stripper C Concentrate Alkaline Liquid H.O.S
Potattiua Hydroxide NA 1719 40I6607M* Red/org opaq
Beige,gray.orange tludge ' sandy toilid. HNU-0.5
Green tandy,chunky tolid. KKU«0. EM Cu»100 pp>.
Ferric chloride. Staple taken off floor of aobile
hoae near drum detcribed above. Tan sandy solid.
Ferric chloride, anhydrout
100 1 Ib plattic bagt in each drua. Ho hazcat done
25% liquid, 75% tludge. Brown opaque liquid. HNU-4
FUTURE APPLICATIONS
A possible future application of STREAMLINE is to integrate it with
a hazardous waste manifest generation program. The data base generated
by STREAMLINE would be combined with a data base thai contains
transporting and safety information, 10 automatically print the appro-
priate hazardous waste information on the manifest form. This proce-
dure would eliminate mistakes and lessen the workload of the on-site
personnel.
The STREAMLINE data base could be linked to an Alternative Treat-
ment Technologies data base to determine the appropriate treatment
standards for the hazardous waste on any particular site.
Another possible application would be to use the information entered
through the STREAMLINE program to search a computerized data base
for disposal sites that are designated for the types of hazardous waste
that need to be removed from the site.
CONCLUSIONS
When dealing with an emergency removal situation, it is important
to have tools in hand that help make the removal process expedient and
efficient. Because STREAMLINE is easy to learn and easy to use, it
also can make a field investigation team more productive. It is advan-
tageous to have the daia computerized, particularly in the widely used
dBASEHI+ format, not only for hazard categorization by STREAM-
LINE, but also because the information can be easily transferred to
a word processing program for inclusion in a report, to Lotusl23 or
to other data analysis software.
REFERENCES
1. Worobel, R.S . The Response Kit, U.S. EPA Kept. No. 198702-017, US ErA.
San Francisco, CA. Dec., 1987.
2. Wjrobel. R.S., Aero Quality Hating Compam'. Oakland, CA, US. Efi*. Kept-
No. T98704-010, U.S. EPA, San Francisco, CA, May, 1987.
3. Floyd. O.A., KCRA Quantification cfU.S. EPA Enforcement Samples Col-
lected at the Aero Quality Plating Facility StabUuotion Effort, May 19-Jiau
15, 1987. Oakland. CA, US. EPA Rept. No. T98705-008. U.S. ErW, San Fran-
cisco. CA. Apr, 1987.
4 Wtorobel, R.S., Site Assessment of the Lorrntz Barrel and Drum Compam,
San Jose. CA. U.S. EPA Repi. No. T98708-015, U.S. EPA, San Francisco,
CA, Dec., 1987.
492 EMERGENCY RESPONSE
-------
Remediation of Underground Explosives
Contaminated Sewer Lines
James L. Dapore, M.S.C.
Charles Lechner
O.H. Materials Corp.
Findlay, Ohio
ABSTRACT
O.H. Materials Corp. (OHM), under contract to the United
States Army Toxic and Hazardous Materials Agency, was tasked
with the decontamination of approximately 7 mi of buried sewer
line contaminated with 2,4,6-trinitrotoluene (TNT) at the former
West Virginia Ordnance Works. The sewer line had removed the
industrial wastewater effluent from ten TNT manufacturing
lines used during World War II for production of explosives.
An innovative, safe and cost-effective approach was needed.
Since thermal destruction of explosives is the only assured method
of decontamination, OHM focused on this technology.
In situ methods could not be certified as effective. The age and
unknown condition of the pipe and its contents precluded such an
approach.
Excavation and verifiable decontamination of the pipe was
necessary. Several approaches to decontamination were then
available. Rotary kiln technology was judged not only too expen-
sive, but also might require crushing the pipe with the attendant
risk of detonation. Stationary furnaces could be used, but the
logistics of such an operation in a remote area presented formid-
able obstacles.
A handheld flamer torch technique was adopted. This
approach was judged the safest, most cost-effective approach.
Verification of the effectiveness of the decontamination was
achieved using Certipaks, which are small ceramic beads impreg-
nated with TNT and DNT. The Certipaks, placed in the pipe dur-
ing flaming, are retrieved after flaming and analyzed using a
field colorimetric method to verify the effectiveness of the pro-
cess in decontaminating the explosives.
To illustrate the effectiveness of this approach, the decontam-
ination of over 7 mi of sewer line, including excavation, flaming,
certification and backfilling, was accomplished, accident free, in
less than 2 mo at a cost of approximately $1.4 million.
INTRODUCTION
The former West Virginia Ordnance Works (WVOW) encom-
passes an 8,323-ac parcel of land in Mason County, West Vir-
ginia. It is located approximately 6 mi north of Point Pleasant,
West Virginia, on the east bank of the Ohio River (Fig. 1). The
property is now owned by various state, local and private con-
cerns. The largest portion of the site is occupied by Clifton F. Mc-
Clintic State Wildlife Station (MCCLINTIC) operated by the
West Virginia Department of Natural Resources (WVDNR).
During World War II (1942 to 1945), the site was used to pro-
duce TNT. Contaminants from those operations were present in
the sewer lines which conveyed process wastewaters. The con-
tamination was in the form of TNT and its associated process by-
products and environmental degradation products.
In 1946, the property was declared excess by the government
and portions were sold. Most buildings have been removed; how-
ever, some foundations remain. In 1949, the U.S. Army deeded
the process and waste disposal property to the state of West Vir-
ginia for use as a wildlife refuge.
In 1981, evidence of contamination was found in one of the
ponds of MCCLINTIC, and the WVDNR and the U.S. EPA
were notified. The U.S. Army was notified of the contamination
in 1983.
An RI/FS was begun in 1984 to study the contamination prob-
lem. As a result of the study, it was determined the cleanup pro-
ject should be conducted in two remediation phases (operable
units). The first unit (i.e., TNT Manufacturing Area, Burning
Grounds and Industrial Sewer Lines) is described in this paper.
This project was almost entirely contained within the McClintic
Wildlife Station.
In 1987, a formal ROD was agreed to by the U.S. EPA and the
U.S. Army (in concurrence with the WVDNR). The ROD pro-
vided for excavation, flaming, and backfilling to clean-up the
explosives-contaminated sewer lines.
SEWER LINE EXCAVATION, FLAMING
AND BACKFILLING
MCCLINTIC was overgrown with trees, brush and weeds.
Prior to beginning sewer line excavation, trees and underbrush
had to be cleared. Manholes were used for location reference
points and a 60- to 100-ft wide path was created.
There were 10 manufacturing area sewer systems at
MCCLINTIC which has to be remediated. The 10 areas all drained
into a common sewer system, which also had to be remediated
(Fig. 2). Initial excavation of the sewer lines began near the first
manufacturing area in the common sewer system and proceeded
through each manufacturing area, cross-country and eventually
to three settling basins.
At the same tune as the common sewer line excavation, sewer
lines from each of the 10 manufacturing areas were excavated,
remediated and backfilled. Two independent crews were used to
accomplish the sewer line remediation in less than 2 mo.
A unique method was used to decontaminate the explosives in
the sewer lines and verify the decontamination. The cleaning
method is called flaming. Verification was accomplished using
Certipaks.
EMERGENCY RESPONSE 493
-------
WIST
VIRGINIA
.WEST VIRGINIA
ORDNANCE WORKS
WEST VIRGINIA
ORDNANCE WORKS
SITE
OHIO
CLIFTON F. McCUNTIC
STATE WILDLIFE
SCALE
1013 MILES
101 2 KILOMETERS
Figure 1
Location of ihe WVOW Site
494 EMERGENCY RESPONSE!
-------
Figure 2
Manufacturing Area & Sewer-Line Locations
After the sewer lines were excavated, they were brought to
ground level and stacked in a triangular pile (Fig. 3). Each clay
tile piece was 3 ft long and each pile of tiles contained approxi-
mately 10 clay tile pieces. Thus, about 30 linear feet of sewer line
were stacked in each pile.
FLAKE THROVER
IHSEXTED INTO EACH
TILE ABOUT 10 SECONDS fr
EACH V
•ONE CERTIPAK
RANDOMLY PLACED
FOR FLAMING
OFERATIOH TO*.
EVERT 60 FEET
OF TILE
Figure 3
Piling of Sewer Tile for Flaming
All loose explosive pieces were removed from each piece of tile
using sparkless shovels or plungers. These explosive pieces were
collected in plastic lidded buckets and transported to a storage
magazine located on MCCLINTIC.
Flaming
The tile pieces then had only a thin residue of explosive powder
remaining on the inside walls. The residue was so firmly attached
to the walls that it could not be removed by any conventional
method without breaking the clay tile pieces. If tile pieces were
broken, all loose explosives would have to be collected by hand
and taken to the on-site magazine. The tile pieces would then
have to be burned separately to destroy the explosives remaining
and then backfilled into the trench. Aside from the explosion haz-
ard in using this approach, it was not considered a desirable
method of remediation because it was not the most cost-effective
approach.
Certipak Analysis
The cleanup method chosen, as noted above, was to flame the
tiles at the excavation site using handheld flamers. To verify that
EMERGENCY RESPONSE 495
-------
the tiles were completely cleaned, a Certipak was randomly placed
on the side of the tile opposite the name throwers. After a period
of "flaming", the Certipak was removed from the stack of tile
pieces and the enclosed ceramic bead removed. Using a field col-
orimetric method, the bead was checked for remaining explosive
residues. If the bead has no explosive material, the flaming oper-
ation was considered successful in destroying the explosive residue
and the clay tile pieces could be backfilled in the trench. If the
bead had remaining explosive residues, the flaming and Certipak
placement were repeated, the field colorimetric method again
was applied and the results were noted. This process was repeated
as many times as necessary to obtain a "clean" Certipak. Very
few times did the flaming process require a second pass and only
once, a third pass.
A Certipak is a foil packet with an enclosed ceramic bead with a
length of wire attached for handling. The ceramic bead has a
standard amount of explosive impregnated in it. The standard
amount and placement of explosives on the bead are accom-
plished in a laboratory under controlled conditions. In this pro-
ject, the method for preparing the beads was written by an
approved laboratory and submitted to, and approved by, the
U.S. Army Toxic and Hazardous Materials Agency
(USATHAMA) prior to the Ccrtipaks being used in the field.
The method used to determine the presence (or absence) of
residue on the ceramic bead was a colorimetric process. When
the bead is removed from the Certipak, it is contacted with re-
agents and the color noted. Basically, a colored bead indicated
that the bead still had explosives on it. Thus, the flaming had not
successfully removed all the explosives and had to be repeated.
On the other hand, a white bead indicated that the bead was free
of explosives and the flaming process had been successful.
UM of Certipak in Flaming Process
Certipak beads are impregnated with explosive residues in the
laboratory as noted above. The explosive material is then driven
off by the application of heat on a few test beads and the field
colorimetric method is tested to assure that it is working properly.
This testing is done daily.
The beads are put in a foil packet with an attached wire. The
packet assures that the bead is subjected to the temperatures of
flaming without being soiled with any smoke residues which
make the developed colors hard to see. The attached wire makes
it easier to retrieve.
When the Certipak is placed in the stack of clay tile pieces, it is
put on the opposite side of the pile from the flame thrower. This
placement assures that the Certipak is being subjected to the
minimum temperature achieved within the clay tile pieces. Then,
if the ceramic bead has been successfully treated to remove all
explosive residues (i.e., subjected to enough heat), the clay tile
pieces, which had a higher contact temperature, also would be
clean.
The system of stacking and flaming tile pieces was used
throughout the project. Initially the length of time it took to
properly flame clay tile pieces was determined. This same time
period was then used on subsequent clay tile piles.
In all, over 7 mi of sewer tiles were decontaminated this way.
The work was accomplished with no accidents.
DESTRUCTION OF TNT PIECES FROM
STORAGE MAGAZINE
The pieces of TNT recovered from the excavated sewer lines
were taken to an on-site magazine. Subsequently, they were in-
cinerated. The procedure used was as follows.
A burning pad was constructed. It consisted of a semi-cylindri-
cal vessel with a cap on each end. TNT pieces were placed in the
vessel in a single layer; no more than 50 Ib were burned at one
time. A handheld flamer was used to burn the TNT in the vessel.
The flame configuration was arranged so the flame did not im-
pinge on the TNT until all personnel were a prescribed distance
away from the burning vessel. After each successful burn, a cool-
down period was required before more TNT could be loaded in
the burning vessel. Burning and cool-down had to be completed
during daylight hours.
In total, nearly 1 ton of TNT pieces was destroyed using the
burning vessel approach. Safety standards were very high and no
injuries occurred.
INTERESTING SIDELIGHTS
Along the cross-country sewer line route there was a section of
pipe running uphill. Because this was a pressurized system (i.e.,
pumps were being used at that point), the pipelines had been con-
structed of steel and wood staves rather than the clay tiles used
elsewhere. These pipelines, although buried for nearly SO yr, were
in excellent condition, almost Like new. Our expectation was to
find pipe badly deteriorated in spots, possibly even corroded
through. That was not the case. These lines also were clean of any
explosives contaminants as one would expect of pressure pipe.
Another item of interest was the on-site magazine used to store
explosive materials until destruction could be accomplished. The
magazine used was found to be almost like new with little musti-
ness, almost no cracks in the concrete liner and a workable door
—after 45 years!
CONCLUSION
The project successfully decontaminated over 7 mi of under-
ground sewer lines. Also accomplished was the installation of
nearly 14 ac of soil cover and the creation of a sizeable wetland
area (pond) in the process. There was virtually no disturbance to
existing wetlands as a result of remedial activities.
496 EMERGENCY RESPONSE
-------
Characterization and Remedial Assessment
of DNAPL PCB Oil in Fractured Bedrock:
A Case Study of the Smithville, Ontario, Canada Site
William J. Mills
Peter Beukema
Douglas White
The Proctor & Redfern Group
Toronto, Ontario, Canada
Tom A. Mclelwain
Golder Associates
ABSTRACT
In 1986, the Ontario Ministry of the Environment commissioned
hydrogeological investigations at the site of a former PCB transfer and
storage facility in Smithville, Ontario. Previous studies had identified
widespread soil contamination by PCBs and the potential occurrence
of PCB-bearing DNAPL oil within the fractured dolostone bedrock
underlying 6 m of clay beneath the site. The subsequent investigation
characterized the physical and chemical hydrogeological conditions at
the site and identified the sources, pathways and occurrence of DNAPL.
At the present time, DNAPL is identified as covering an area approxi-
mately 150 m in length and 70 m in width in the upper 5 m of bedrock.
The estimated volume of the DNAPL plume is 30,000 L. The DNAPL
contains approximately 50% PCB and lesser amounts of other chlori-
nated organic compounds.
Extensive rock coring, monitoring well installation and pump testing,
coupled with intensive groundwater quality monitoring, has allowed
the delineation of the DNAPL plume and two associated dissolved con-
stituent plumes of trichloroethene and trichlorobenzenes within the
bedrock which necessitated the closure of a municipal well serving the
Town of Smithville.
A thorough understanding of the microstratigraphy at this site is
critical to the technical assessment of remedial alternatives. A short-
term remedial action plan has been implemented to control the migra-
tion of the dissolved plumes from the source area by means of a system
of recovery wells, water treatment and discharge. Longer term remedial
options are being assessed, with the objective of cleaning up the con-
taminated bedrock using in situ techniques. The hydrogeological
investigations are continuing concurrently with ongoing site decom-
missioning activities, which include the proposed on-site incineration
of nearly 180,000 L of PCB-bearing oil, contaminated soil and other
solids which are in secure on-site storage.
INTRODUCTION
In 1978, a private waste management firm was issued a Certificate
of Approval from the Ontario Ministry of the Environment (MOE) to
operate a PCB transfer and storage facility at a location in the Niagara
Peninsula region of south-central Ontario, Canada. The site, as shown
in Figure 1, is located in an industrial park in the northern outskirts
of Smithville, Ontario, and originally consisted of approximately 0.8 ha,
of which 0.25 ha were used for the transfer and storage activities.
Between 1978 and 1985, the site reportedly received approximately
434,000 L of liquid waste including approximately 266,000 L of PCB-
contaminated wastes. The remainder of the waste inventory included
organic solvents, resins, acids, alkalia, inorganic liquids and inert
sludges. In early 1985, the site was effectively closed when the Certifi-
cate of Approval was revised to permit only the storage of wastes then
located on the site. Since that time, the site owner/operator has not been
involved in any activities and MOE has assured ownership and respon-
sibility for the site.
Subsequent to the closure of the site, testing by MOE disclosed the
presence of PCB-contaminated soil and water in a retention lagoon
located near the southeast corner of the site. During the fall of 1985,
a short emergency cleanup was performed. Approximately 72,000 L
of PCB-contaminated sludge, oily water and soil were removed from
the lagoon by a specialist decontamination contractor retained by MOE,
and these materials were placed in secure containers for on-site storage
until their ultimate disposal could be addressed.
In January 1986, MOE retained Proctor & Redfern Ltd. as the over-
all Project Manager for the decommissioning of the site. The initial
task was to secure and subsequently arrange for the disposal (destruc-
tion) of the PCB wastes then in unsecured storage at the site.
The work of cleaning up the site progressed during 1986, and all of
these wastes are currently in on-site secure storage within a specially
designed and constructed warehouse. An area of near surface contami-
nated soil (including the original retention lagoon area) was temporarily
secured by covering it with a synthetic membrane.
In 1985, MOE began to investigate the potential for off-site migra-
tion of contaminants from the facility. Due to the presence of 6 to 10 m
of clay overburden beneath the site, it initially had been assumed that
any migration of contaminants would be via surface and/or near-surface
pathways. However, in early February, 1987, Dense Non-Aqueous Phase
Liquid (DNAPL) PCB oil was detected within the dolomitic limestone
bedrock underlying the clay. This discovery significantly expanded the
scope of the proposed decommissioning work and, as a result, MOE,
through Proctor & Redfern Ltd., retained Golder Associates Ltd. to
act as a specialist subconsultant to Proctor & Redfern with a mandate
to investigate the extent and advise on the remediation of the subsur-
face contamination.
Since May, of 1987, investigations have characterized the physical and
chemical hydrogeological conditions at the site and identified the
sources, pathways and occurrence of the plume of DNAPL which con-
tains up to 50% by weight PCB and lesser amounts of trichloroben-
zenes (TCB), trichloroethylene (TCE), trichloroethane (TCA) and other
organic constituents. In addition, two distinct but interrelated dissolved
contaminant plumes exist within the bedrock which forms the water
supply aquifer for the Smithville vicinity. In late 1987, one of the two
municipal wells serving the Town of Smithville was ordered closed due
to its location within about 600 m of where the PCB contamination
had been detected in the bedrock aquifer. Primarily because of this
situation, MOE in co-operation with the Regional Municipality of
Niagara undertook to construct a water main to connect the Smithville
distribution system to that of the Town of Grimsby, located below the
EMERGENCY RESPONSE 497
-------
escarpment on Lake Ontario, 20 km away.
RESULTS OF INVESTIGATIONS TO DATE
The investigations completed to date at the Smithville site consist
of a detailed evaluation of the regional geology, hydrosiratigraphy and
groundwater chemistry of the original waste management site and
approximately 60 h of adjoining property. The scope of the study to
date has included the drilling of 78 cored boreholes to various depths
within the bedrock, the installation of 120 monitoring wells, aquifer
testing to determine relevant hydraulic parameters and detailed ground-
water chemical monitoring for the organic constituents associated with
the site. While the studies at the site are continuing, the focus has
changed to one of assessing the long-term remedial alternatives and
designing, implementing and monitoring the performance of the interim
control measures required to secure the site while the evaluation of the
most appropriate remedial action proceeds.
DNAPL SOURCES AND OCCURRENCE
Investigations at the site have identified DNAPL containing PCB,
TCB and TCE within the Overburden and Shallow Aquifer beneath the
Smithville site. Complete characterization analyses of the DNAPL
indicate that the oil contains between 35 and 55% PCB and 5 to 8%
TCB. The laboratory analyses of the PCB congeners suggest (hat a mix-
ture of Aroclor 1242, 1254 and 1260 comprises the DNAPL. Analyses
for organic solvents including TCE and TCA indicate that these con-
stituents are present in the DNAPL in concentrations ranging from the
high hundreds of ppm to approximately 1.8%. It is probable that the
solvents are present as contaminants in the PCB oil, arising from co-
storage or disposal at the site. Other principal constituents identified
in the ppm range include benzene, chloroform and mono- di- and tetra-
chlorobenzenes.
It has been demonstrated through drilling and coring of the overburden
on the site that DNAPL oil migrated vertically downwards through the
bottom of the former retention lagoon. DNAPL oil was observed in
weathered fractures in the clay 6 m below ground surface at the lagoon
site.
Investigations of other potential sources and migration pathways,
including the vicinity of two vertical storage tanks which formerly held
up to 160,000 L of PCB oil, a Quonset hut formerly used to store wastes
and the original geotechnical borings put down at the time of construc-
tion of the facility, have indicated that the former lagoon is the only
probable significant source of the DNAPL, and that vertical migration
under gravity via the weathered fractured clay is the only probable path-
way. The implications of this finding may have significant impact on
the way in which existing and proposed chemical storage and handling
facilities situated on clay deposits are evaluated in the future.
Estimates of the volume of DNAPL that could be resident within
the bedrock range from the low thousands of litres to as much as
30,000 L, based on drilling evidence and observation of recovered oily
rock core. Borehole drilling and sampling in the vicinity of the former
lagoon have delineated a kidney shaped plume of DNAPL centered
around the area of the former lagoon, extending down dip to the
southeast within the Shallow Aquifer for a distance of 160 m, achieving
a maximum observed width of approximately 70 m and a depth of
penetration into the bedrock of approximately 5 m.
Free oil has been observed to migrate rapidly within the open bedding
partings over distances of several metres under imposed gradient con-
ditions. As the lateral distance from the area of the former lagoon
increases, the depth of the first occurrence of DNAPL within the Shallow
Aquifer increases, strongly suggesting that the migration under gravity
occurred in a step-wise vertical fashion through vertical fractures in
the bedrock. However, monitoring suggests that the DNAPL plume
presently is stable and no longer is expanding in the horizontal plane.
The observed extent of DNAPL in the bedrock is shown in Figure 2.
CONTAMINANT PATHWAYS AND MIGRATION
WITHIN THE BEDROCK
The DNAPL plume within the upper bedrock zone provides the
source of an elongated dissolved contaminant plume within the Shallow
METROPOLITAN
TORONTO
LAKE
ONTARIO
ST.
GRIMS8Y CATHARI
Figure I
Location of Smiihville Site
In Ihe Niagara Peninsula Area
of Ontario, Canada
Aquifer. The movement of groundwater within this zone is governed
by the open bedding partings. The porosity of the open bedding partings
is expected to be quite viable. Observations of plume migration suggest
a contaminant velocity of 50 to 100 m/a for TCE, one of the relatively
more mobile constituents of the plume.
As shown in Figure 2, the shallow dissolved plume extends at least
600 m downgradient from the former lagoon where the TCE concen-
tration exceeds 5 ng/L. Lesser concentrations may extend beyond
600 m, but the results of additional drilling and testing to define the
leading edge are not yet available. The groundwater within the Shallow
Aquifer is inferred to discharge to Twenty Mile Creek approximately
2 km south of the Smithville site.
The advance of the TCB plume within the Shallow Aquifer appears
to be only slightly retarded with respect to the TCE plume, but the PCB
plume is significantly retarded, as contamination to the Ontario Drinking
Water Objective of 3 jtg/1 extends only 50 m beyond the leading edge
of the DNAPL plume.
An extensive network of off-site monitoring wells shown in Figure 3
has been installed between the site and all privately owned farm and
domestic wells in the vicinity to provide early warning of any expan-
sion in the area of groundwater impact. The majority of the private
wells are located to the north, east and west, hydraulically upgradient
from the site, and not downgradient to the south. However, there are
498 EMERGENCY RESPONSE
-------
Figure 2
Observed Configuration of DNAPL and
Dissolved Contaminant Plumes Within the
Shallow Aquifer Beneath the
Smithville PCB Site
a few wells located approximately 1.5 km south of the site between the
leading edge of the plume and Twenty Mile Creek.
The chemical analytical results of groundwater samples collected from
wells installed within the Deep Aquifer are less conclusive than those
drived from the Shallow Aquifer. However, there is definite evidence
that dissolved contamination plume exists at depth. No direct evidence
of DNAPL within the Deep Aquifer has ever been observed, but the
relatively high concentrations of TCE and TCB in some wells (several
hundreds of mg/1) located only 70 m south of the leading edge of the
DNAPL plume contained within the Shallow Aquifer suggests that
DNAPL may be present at depth.
The introduction of DNAPL and/or dissolved phase contaminants
to the Deep Aquifer is thought to have occurred by downward migra-
tion via the regularly spaced vertical fractures and joints in the rock.
In these local quarries where the Lockport Formation is exposed, these
features occur at intervals as close as 3 m within individual members,
and up to 20 m where they penetrate the entire Lockport sequence.
A significant downward gradient exists across the minor aquitard
between the Shallow Aquifer and the Deep Aquifer, and contaminated
groundwater could, therefore, migrate across the aquitard where these
joints and fractures occur. Once present within the Deep Aquifer, the
dissolved constituents migrate downgradient within the open beddings
partings.
Based on well monitoring results, contaminated groundwater reached
the Smithville Municipal Well No. 2, located 600 m south of the former
lagoon at the site, commencing in mid-1988. Assuming that the DNAPL
entered the bedrock sometime between 1978 and 1985, the average linear
velocity of the dissolved contaminant plume within the Deep Aquifer
actually lies in the range of 50 to 165 m/a.
SPRING CREEK ROAD
LEGEND
A RECOVERY WELL
• SMITHVILLE WELL HO.2
^ EXTENT OF DNAPL
• FORMER LAGOON
»—''ZONE OF CAPTURE
GROUNDWATER FLOW
DIRECTION
SO 0 SO 100
SCALE. METRES
SMITHVILLE
INDUSTRIAL
PARK
L
I
LONDON ROAD
SMITHVILLE
— i
Figure 3
Shallow Aquifer Hydraulic Containment
System Showing Conceptual Operational
Effects at the Smithville PCB Site
INTERIM CONTAINMENT STRATEGY
The hydrogeological investigations completed to date at the Smith-
ville site have identified a large-scale occurrence of bedrock contami-
nation by DNAPL and two separate dissolved phase plumes. A DNAPL
occurrence of this nature or magnitude has not been successfully
remediated, and the case histories of attempts to deal with smaller scale
and more easily accessible occurrences are not encouraging. Neverthe-
less, every effort will be made to remediate the bedrock at the site if
technically feasible. The expenditure of considerable time and finan-
cial resources will be required to assess the ultimate feasibility of com-
plete remediation. In the interim, the shallow and deep dissolved
contamination plumes identified beneath the site would continue to
migrate further downgradient off-site, expanding the scope of the existing
problem, unless control measures are implemented.
SHALLOW AQUIFER HYDRAULIC CONTAINMENT
In early 1989, MOE directed the project team to design and imple-
ment hydraulic controls to contain the dissolved contaminant plume
within the Shallow Aquifer, where clear evidence of off-site migration
of relatively high concentrations of TCE and TCB existed. Data derived
from a controlled pumping test carried out within the Shallow Aquifer
in the vicinity of the DNAPL source of the dissolved plume were used
to design a recovery well system.
The system consists of eight 20-cm diameter pumping wells, each
equipped with submersible pumping equipment which maintains the
hydraulic head within the aquifer at a designed elevation. The well net-
work creates a hydraulic trap, essentially preventing groundwater which
contacts the DNAPL from escaping the groundwater sink. The net-
work controls groundwater flow to a point 50 m downgradient from
the leading edge of the DNAPL plume by pumping a cumulative total
of up to 100 L/min. The recovered water is treated using activated carbon
treated on site, and then discharged to the sanitary sewer.
The portion of the dissolved plume in the Shallow Aquifer beyond
the zone of capture has, so far, been allowed to migrate uncontrolled,
EMERGENCY RESPONSE 499
-------
but another network of recovery wells can be installed beyond the leading
edge of this plume in the event that monitoring results indicate a
requirement to do so. The recovery system became fully operational
in July, 1989, and the early monitoring results indicate successful con-
tainment of the shallow plume. The system is shown in Figure 3.
DEEP AQUIFER CONSIDERATIONS
The working hypothesis adopted during the hydrogeological investi-
gations carried out to date at the site, uncertainties remain concerning
the occurrence, concentration and mobility of the dissolved plume in
the Deep Aquifer.
Planning is underway for a directed investigation to determine the
optimum locations (vertical and in plan) for recovery wells to control
the deep dissolved plume, should ongoing monitoring results indicate
a requirement to do so.
MONITORING
An extensive network of monitoring wells (Fig.3) was installed at
the site between 1987 and 1989. The wells are sampled regularly (as
often as weekly in some cases), and the samples arc analyzed for the
suite of constituents associated with the site, including PCB, TCB, TCE,
TCA, benzene and chloroform. Plume tracking and concentration trend
analysis are undertaken routinely to monitor the three-dimensional con-
figuration of the shallow and deep dissolved plumes.
Monthly groundwater elevation measurements are obtained from all
monitoring wells, and five automatic water level recorders collect con-
tinuous groundwater elevation data. The monitoring results are used
to modify the shallow plume recovery system and to review the need
for a deep plume containment system.
LONG-TERM REMEDIAL STRATEGY
A long-term remedial strategy for the site has not been final ized.
However, the long-term strategy includes the elimination at the earliest
practical time of the estimated 180.000 L of PCB waste currently in
secure storage at the site.
This activity is well advanced, and a thermal destruction contractor
has been selected to incinerate the PCB-contaminated material. The
contractor is preparing for hearings under the Ontario Environmental
Protection Act to obtain approvals to construct and operated a mobile
PCB incinerator at the Smilhville site. Destruction of the existing stock-
piled waste material is expected to be completed by the end of 1990.
Any remedial strategy for the subsurface contamination must include
the detailed assessment of potentially applicable technologies to deter-
mine the feasibility of application at the site. Assuming that such a tech-
nology can be developed, testing of the process at the bench-and
field-scales would be required, and, if successful, pilot-scale trials might
proceed. At the present time, the only potential remedial technique de-
veloped to the point at which a field trail is feasible is excavation, and
this option may only be feasible for the Shallow Aquifer. A program
of shallow bedrock shaft excavation and testing has been developed
which could be implemented at the site.
The costs associated with any effort at remediation, whether in situ
or by excavation, treatment and disposal, will be very high, and no
precedent for (he envisioned scale of remediation exists in North
America. Therefore, it will be extremely important to carefully review
all possible remediation options and to obtain as much data as may be
necessary to arrive at (he most appropriate solution to this problem.
Despite the best of intentions and technical efforts, it may not be possi-
ble to fully remediate the Smithville site in the foreseeable future, and
a longer period of secure containment of the site than is currently
anticipated may become necessary. In that event, in situ physical con-
tainment of the DNAPL plume possibly employing some combination
of cut-offs, grout curtains and low permeability covers in association
with some level of ongoing groundwater recovery and treatment may
have to be considered.
The challenge and opportunities for technical advancement associated
with the Smithville site are considerable, and the task of pursuing an
effective remedial solution leading to secure site decommissioning con-
tinues.
ACKNOWLEDGEMENT
The authors would like to acknowledge the support and assistance
provided to the study team by our many colleagues within Colder
Associates Ltd., Proctor & Redfern Ltd. and the Ontario Ministry of
the Environment.
500 EMERGENCY RESPONSE
-------
Panel Discussion:
Effectiveness of Groundwater Extraction—Technical
Considerations, Field Experience, Policy
Implications
Bill Hanson
Jennifer Haley
U.S. EPA
Washington, D.C.
Carl Enfield
U.S. EPA
Ada, Oklahoma
John Glass
CH2M Hill, Inc.
Reston, Virginia
PANEL DISCUSSION
Groundwater contamination is found at over 70% of the sites
currently on the National Priority List. The most common
method for addressing contaminated groundwater is extraction
and treatment. Groundwater extraction can effectively reduce
contaminant concentrations in the groundwater where contam-
inants are primarily present in the dissolved phase. However, re-
search and field experience indicate that it may be more difficult
than is often estimated to achieve cleanup concentration goals in
portions of the groundwater, particularly those zones near the
original source of contamination. Factors limiting the effective-
ness of extraction systems include: the presence of non-aqueous
phase liquids which lodge in the subsurface and create a contin-
uing source of groundwater contamination as contaminants with-
in the non-aqueous phase dissolve and as the non-aqueous phase
itself dissolves into the groundwater, and sorption of contam-
inants to the soil within the saturated zone resulting in a contin-
uous source of contamination to clean groundwater drawn into
the contaminated zone by extraction systems.
In an effort to determine whether these factors are influencing
the performance of extraction systems currently in operation,
OERR initiated a study to assess the effectiveness of several on-
going groundwater remediation sites. After reviewing data from
19 case studies, it was concluded that groundwater extraction can
effectively contain contaminant plumes and that significant con-
taminant mass can be withdrawn from the subsurface by extract-
ing groundwater. However, in most of the cases, contaminant
concentrations in the extracted groundwater tended to level off
after an initial decrease, at concentrations that were still above
cleanup goals. In many cases, it appeared the factor identified
by researchers were playing a role in the performance of the
groundwater extraction systems.
As a result of the study, some modifications to the current re-
sponse approach for contaminated groundwater are warranted.
The basic goal of returning groundwater to its beneficial uses,
however, will not change. Recommended modifications include:
considering containment early to prevent further migration of
contaminants and collect information on aquifer response to ex-
traction, providing flexibility in selected remedies to allow for
system modification based on data gained during operation, and
improving data collection during the remedial investigation to
identify situations and processes that may affect extraction per-
formance. Further study of extraction systems is warranted to
identify the signals that indicate cleanup goals cannot be attained
and to evaluate the point at which alternate goals should be estab-
lished.
I. Theoretical Background—Carl Enfield
A. Factors Affecting Groundwater Remediation
1. Hydrologeologic—diffusion through varying geologic
material, fractures
2. Chemical—sorption
3. Multi-phase Fluids
B. Need for Improved Data Collection/Methods
1. Characterize Vertical Variations in Geologic Materials
2. Evaluate Contaminant Partitioning in the Saturated
Zone
3. Identify Presence of Non-Aqueous Phase Liquids When
Practicable
C. Other Cleanup Options/Status of Research
1. Biorestoration
2. Vapor Extraction
3. Solvent Flushing
4. In-Situ Steam Stripping
II.Practical Experience—Jennifer Haley, John Glass
A. Description of Study
1. Identified Groundwater Extraction Sites
2. Selected 19 for Case Studies
3. Evaluated Performance of Extraction
B. Findings
1. Containment Generally Successful
2. Significant Contaminant Mass Removed
3. Contaminant Concentrations Level Off After Initial
Decrease
4. Several Factors Limited Effectiveness of Extraction
a) Hydrogeologic
b) Contaminant
c) Adequacy of Source Removal
d) Design of Extraction System
m. Implications for Superfund Response Approach to Contam-
inated Groundwater—Bill Hanson
A. Maintain Overall Goal
B. Initiate Response Early
1. Contain Plume
2. Collect Information on Aquifer Response to Extraction
GROUNDWATER TREATMENT 501
-------
C. Provide Flexibility in Remedies 2. Contaminant Sorption to Soils in the Saturated Zone
1. Contingency E. Guidance Needs
2. Interim Remedies 1- Signals Indicating Cleanup lo Health-Based Levels Not
D. Collect Better Data During Remedial Investigation Practicable
1. Vertical Variations of Hydraulic Conductivity and 2. Alternate Goals Where Health-Based Concentrations
Contaminant Concentration Cannot Be Attained in the Groundwater
502 GROUNDWATER TRI-.ATMENT
-------
CERCLA Sites Affected by RCRA
An Overview of the Corrective Action Process
Marvin Unger
Michael Trojan
Gordon B. Evans, Jr.
David C. Anderson
K. W. Brown & Associates, Inc.
College Station, Texas
ABSTRACT
An overview is provided of the corrective action program (CAP) de-
veloped under the RCRA. A description is given of both RCRA and
CERCLA program objectives. Basis of the CAP is discussed along with
its major components. The proposed codification of the CAP (Subpart
S Regulations) are also discussed.
INTRODUCTION
When the Hazardous and Solid Waste Amendments (HSWA) were
enacted in 1984 as part of the reauthorization of the RCRA, they set
in motion a regulatory vehicle in the form of the Corrective Action
Program (CAP), a program designed to address the nation's ever growing
concern centering on continuing releases of hazardous wastes or
hazardous constituents to sensitive environmental pathways. Since it
is not uncommon to encounter RCRA facilities which have or currently
are involved with CERCLA activities, under the realm of the CAP, these
CERCLA sites would be considered as part of the CAP.
In order to provide timely and appropriate responses when such
releases are identified, the complexity of such a comprehensive effort
must initially be approached in a systematic or phased manner.
Accordingly, the CAP, designed to impact all facilities that received
or processed hazardous waste after July 26, 1982, is comprised of three
phases:
RCRA Facility Assessments
RCRA Facility Assessments (RFAs) involve the identification of poten-
tial hazardous waste or constituent releases requiring further investi-
gation. This is the initial data gathering phase of the CAP, incorporating
a comprehensive Preliminary Review (PR) of available facility infor-
mation and data, a Visual Site Inspection (VSI) of all Solid Waste
Management Units (SWMUs) and Areas of Concern (AOCs) within
the contiguous boundaries of the facility, and an optional Sampling Visit
(SV) of potentially impacted areas.
The information collected during the RFA is considered in making
release determinations. This information, therefore, must, at a
minimum, include the history of the facility site, the type and design
of waste management units, the type and condition of potentially affected
soil, surface water, groundwater, subsurface gas and/or ambient air.
This RCRA phase parallels the CERCLA Preliminary Assessment/Site
Investigation (PA/SI).
RCRA Facility Investigation
RCRA Facility Investigations (RFIs) are designed to provide charac-
terization of releases identified during the RFA. The level of effort per-
formed as part of the RFI involves the comprehensive characterization
of suspect area, determination of the extent of releases into the specific
environmental media suspected of being impacted and examination of
the nature of the release as related to its impact on human health and
the environment.
In consideration of the many various sampling scenarios, the im-
plementation of an RFI may entail a broad range of sampling strategies
and techniques. This RCRA phase parallels the CERCLA Remedial
Investigation/Feasibility Study (RI/FS).
Corrective Measures
Corrective Measures (CMs) describe remedial measures to be used
in the impacted area. If, at any phase of the CAP, conditions are
encountered that suggest further action, then Corrective Measures may
be required. The nature of these measures depends on the Agency's
stance concerning how, in conjunction with the facility response to this
condition, they perceive the nature of the suspected release. As
mentioned, these Corrective Measures, designed to be developed under
RCRA, may pertain to facility CERCLA sites. The interrelation of these
phases are illustrated in Figure 1.
Strategies
There are technical strategies available to effectively address release
determination issues prior to and during the phases of the CAP. These
strategies are based on interpretation of release conditions based on
past and ongoing data and involve an active intercommunication be-
tween the owner/operators and the Agency. However, the ultimate im-
plementation of owner/operator's strategies depends on the site
conditions as perceived by the Agency, the level of supporting infor-
mation provided by the owner/operators, and the time-frames suggested
to address specific CAP phase requirements.
RCRA vs. CERCLA Objectives
It is not unusual for a RCRA facility owner/operator to have an area
or unit located on-site which has been associated in the past with the
CERCLA program. It is, therefore, important for the owner/operator
possessing a CERCLA site to be aware of the difference between RCRA
and CERCLA program objectives. Typical RCRA vs. CERCLA con-
cerns are described below.
CERCLA is designed as a response program to deal with environ-
mental contamination that already has been documented. Often, the
actions which caused the contamination were legal and non-negligent.
Since the facility personnel responsible for these activities did not an-
ticipate that a cleanup would be required in the future, associated cleanup
costs were not considered in future budgeting. However, people with
prior associations with a contaminated site may be imposed with retro-
active liability since the costs of doing so cannot be built into the trans-
actions associated with the disposal, as those transactions took place
years ago. Moreover, there is no regulatory/enforcement virtue to
GROUNDWATER TREATMENT 503
-------
PECULATORY AGENCY performi RCRA Facility Assessment (Rf A) (0
• Identify tolid wane management unui (SWMUi) and colled onung information
on contaminant releaiei.
• Identify releaiet or impeded releaiet needing further investigation
REGULATORY AGENCY specifies permit conditions or tiiuti enfofcemeni order to Utility
owner or operator to
e Perform investigation* on rtleaiei of concern; and/or
• Implement interim corrective meaiurei
OWNER OR OPERATOR perform* RCRA facility investigation (RFI) 10 verify the rtleatrtt). if
necessary, and to charaoenit the nature, extent and rate of migration for releaiei of
concern. Owner or operator reports results and contacts the regulatory agency
-------
ment with the broad-based, CERCLA corrective action approach (i.e.,
facility-wide), the regulatory triggering of the RCRA CAP investiga-
tions may bring about unanticipated conditions of compliance. For
example, new units or areas of concern may be identified through the
initial stages of the RCRA CAP; these new units may be considered
on a unit-by-unit basis, disregarding the facility-wide environmental con-
siderations already established under CERCLA.
For those RCRA owner/operators with CERCLA affected, on-site
areas, the following delineates the objectives and scope of the CAP.
Whenever appropriate, additional noteworthy RCRA-CERCLA com-
parisons will be identified.
BASIS OF THE CORRECTIVE ACTION PROCESS (CAP)
The basis of the Corrective Action Process (CAP) originates from
the Hazardous and Solid Waste Amendments of 1984. Currently, the
Agency has taken steps to have the salient conditions of the CAP codi-
fied, namely in the form of the proposed Subpart S ruling. The
Hazardous and Solid Waste Amendments (HSWA) The Corrective
Action Process (CAP) has evolved from its initial authority as issued
under the authority of the Solid Waste Disposal Act as amended by
RCRA, and finally, as amended by the HSWA of 19842. Providing a
more focused approach than the earlier environmental programs, the
primary objective of the RCRA corrective action program is to clean
up releases of hazardous waste or hazardous constituents that threaten
human health or the environment. The program applies to all operating
closed or closing RCRA facilities.
Although HSWA provides the authority to implement the CAP as
part of the current RCRA program, corrective action at hazardous waste
facilities is also considered under other authorities2. These addition-
al authorities include the following:
• 7003 of RCRA - The Agency has the authority to take action where
there is solid or hazardous waste that may present an imminent and
substantial endangerment to human health or the environment;
• 3013 of RCRA - The Agency has the authority to require investiga-
tions where there is the presence of hazardous waste or where releases
of hazardous waste that may present a substantial hazard to human
health or the environment; and
• 40 CFR Pan 264, Subpart F - The Agency has the authority to address
releases of hazardous wastes and hazardous constituents to ground-
water from units which are "regulated" under the RCRA program.
Nevertheless, HSWA has established new authorities that are even
more broad and far-reaching than these other authorities in the RCRA
program, that enable the U.S. EPA to accomplish their corrective action
objectives. The new authorities are:
• 3004(u) - Corrective Action for Continuing Releases - This authority
requires that for any permit issued to a RCRA treatment, storage
or disposal facility after Nov. 8, 1984, corrective action at that facility
is required for all releases from their solid waste management units
(SWMUs). This provision also requires that facility owner/opera-
tors must demonstrate financial assurance capabilities for any cor-
rective action which may be required. This provision also indicates
that schedules of compliance be used in permits where the required
corrective action cannot be completed prior to permit issuance.
• 3004(v) - Corrective Action Beyond Facility Boundary - This authority
directs the Agency to require corrective action beyond the facility
boundary where it would be necessary to protect human health and
the environment. This HSWA provision would not be invoked if the
owner/operator can demonstrate that, despite the best of efforts, the
necessary permission to perform these off-site corrective action
activities cannot be obtained. In cases such as this, the Agency still
has the authority to issue corrective action orders which, in turn,
require the necessary corrective action.
• 3008(h) - Interim Status Corrective Action Orders - This HSWA pro-
vision provides the Agency authority to issue enforcement orders or
bring about legal action when there is or has been a documented
release of hazardous waste or hazardous constituents at RCRA
facilities operating under conditions of interim status. These adminis-
trative orders or court action would compel either corrective action
or some other form of response measures in a manner that would
serve to protect human health and the environment. In addition, this
provision provides the U.S. EPA with the authority to take civil action
against facilities in order to obtain the appropriate relief. In providing
an interim status "response,", this provision more closely parallels
CERCLA objectives than the other HSWA corrective action
authorities.
Approach to Corrective Action in the
Proposed Subpart S Rule
Since the CAP was first developed in 1984, the U.S. EPA has had
the opportunity to implement the initial CAP stages, namely the PR
and VSI as part of the RFA. As a result, the Agency has experienced
a myriad of facility conditions where there have been or are releases
of hazardous waste or constituents. In order to provide an effective CAP,
proposed to be promulgated under the authority of Subpart S, the Agency
must first draw on this experience and establish their priorities and
management philosophy as they see appropriate for the implementa-
tion of the RCRA CAP. The Agency view of the types of RCRA facili-
ties and the noteworthy conditions involved with environmental impacts,
and, therefore, influencing the development of the proposed Subpart S
rules, are discussed below.
At some facilities, the type or level of contamination or the release
potential associated with the environmental setting may indicate to the
Agency that this facility is a high priority for the implementation of
corrective action. Most facilities where there are or have been con-
tinuing releases of hazardous constituents usually have point source
similarities in the conditions involved with the releases [e.g., unlined
impoundment(s) or landfill(s)]. Therefore, it would be in the best interest
of all parties that corrective actions be performed at the most environ-
mentally significant facilities (i.e., those with documented hazardous
waste or constituent releases) and on the most significant problems (i.e.,
documented continuing releases) at RCRA facilities. Experience has
indicated that U.S. EPA also might place a high priority on corrective
action performed at those facilities which have demonstrated an un-
willingness to provide timely and appropriate response to their environ-
mental problems in the past.
The Agency also has encountered RCRA facilities where the level
of contamination is either documented to exist over a wide-spread area
or the geographic location of the facility is such that there would be
little likelihood of a release causing significant impact to human health
or the environment. An example of the latter would be a hazardous
waste release overlying an already contaminated aquifer and/or located
many miles from the nearest town or residence. In cases such as this,
the Agency has, in the past, acceded to "conditional" remedies (e.g.,
immediate containment of the release). The Agency has recognized that
prompt action of this kind can reduce the risk to levels which would
be acceptable for the current related uses, or where final cleanup is
impracticable. Moreover, if the Agency intends to expedite the CAP,
then the types of investigative and remedial activities at all RCRA
facilities must be streamlined to focus on plausible concerns and likely
remedies. Therefore, if the proposed Subpart S rule is to provide the
Agency with a means to effectively manage the CAP, the Agency must
emphasize early actions and expeditious remedy decisions.
It would not be unusual for non-RCRA facilities, encountering
environmental problems similar to those encountered under RCRA, to
seek guidance in die remediation needed at their site. Although the non-
RCRA facility owner/operators would be expected to first explore
specific program objectives, lest the facility become burdened with an
overbearing regulatory yoke, it is becoming more apparent that the lia-
bilities and costs associated with an undefined corrective action approach
continue to increase. These increased costs often include the unneces-
sary corrective action tasks dictated by either facility personnel who
are unaware of, or unresponsive to regulatory requirements, or the dif-
fering corrective action approaches developed during the turnover of
Agency personnel, a common occurrence in both State and Regional
EPA offices. In such cases, it is often in the best interest of the facility
and the Agency to promote voluntary and independent action by the
facility.
GROUNDWATER TREATMENT 505
-------
PROPOSED SUBPART S REGULATIONS
In order to provide an effective regulatory vehicle for the implemen-
tation of the CAP, the Agency has proceeded to develop the final codifi-
cation of the HSWA amendments, namely the Subpart S regulations
as part of 40 CFR Parts 264'. Since numerous RCRA facilities cur-
rently are involved in various types of investigative and/or cleanup
activities, the corrective action guidance provided by the proposed legis-
lation would be both timely and appropriate.
The U.S. EPA's objective in providing such a rule would be expected
to establish a framework by which their RCRA corrective action
objectives may be implemented. This corrective action framework would
serve to provide the following:
• Protection of human health and the environment
• Control of the sources of hazardous waste or constituent releases to
reduce or eliminate, to the maximum extent practicable, any further
releases that may pose a threat to human health and the environment
• Development of standards, pursuant to the provisions of Subpart S,
involving acceptable levels of cleanup for environmental media
• Development of standards, pursuant to the provisions of Subpart S.
establishing specific waste management compliance criteria
Scope
There are varied scopes of corrective action which might be required
at an impacted RCRA facility. The Agency has recognized that the types
and degree of ongoing corrective action at a RCRA facility are based
on the complexity of the conceptual model, developed in response to
the CAP.
The proposed Subpan S provisions include the corrective action
remediation program objectives of the Agency. These objectives include
the environmental cleanup standards for remedies that represent a com-
bination of technical measures and management controls for addressing
the environmental problems at the facility. These objectives include:
• Reduction of toxicity, mobility or volume or wastes
• Provision of long-term reliability and effectiveness
• Provision of short-term effectiveness
• Implemenlability
• Realistic cost
In the case of CERCLA sites, the degree of corrective action (and
the associated remedial costs) usually is dictated by the Hazardous Rank-
ing Score (MRS), where, the level of required corrective action is based
on a numerical score. Generally, the U.S. EPA has encountered two
basic types of corrective action approaches to the many different types
of impacted RCRA facilities encountered under the auspices of the CAP:
• Streamlined or focused corrective action
• Complex or interdisciplinary corrective action
Under the RCRA CAP, the U.S. EPA has encountered scenarios where
it would be in the best interest of both the facility and the Agency to
develop a streamlined or focused corrective action plan. For example,
such facilities would be expected to include the following:
• Facilities considered to be a "low risk" - These facilities are typi-
fied by contaminant release problems which are relatively small, and
where releases present minimal exposure concerns. Often, facilities
such as these merely require a "band-aid" approach to remedial
action; for example, development of adequate secondary containment
or physical removal of a low volume of waste or contaminated soil.
• Facilities providing a high quality CAP remedy - Since more facili-
ties are realizing that the costs associated with an ineffective remedial
plan can escalate, high quality remedies are being considered in
response to corrective action problems. Certainly, it is not uncom-
mon to encounter situations where the final conceptual approach to
corrective action proposed by a facility would result in a remedy
which is highly protective. For example, many facilities have opted
for corrective action equivalent to a RCRA "clean-closure." It should
be noted that a high quality corrective action remedy need not neces-
sarily be cost-prohibitive, and the Agency will only accept a plan
which remains fully consistent with all other remedial objectives of
the CAP (reliability, etc.).
• Facilities where there are limited remedial options - The Agency hat
encountered several types of RCRA facilities where there may be
only limited remedial options. One example of such a type of RCRA
facility is one whose waste management practices preceded, and,
therefore, did not address the waste management operating require-
ments of RCRA. It is not uncommon for these facilities to possess
old fill or dump areas with appreciable volumes of uncontained waste
material (e.g.. a large unlined landfill). The associated remedial
approach to such a situation would be limited by few practicable
cleanup solutions.
Another type of RCRA facility where the types of remedial options
may be limited are those where the anticipated future uses of the
property, in turn, dictate a high degree of treatment to achieve very
low levels of residual contamination. An example of such a facility is
one in proximity to vulnerable environmental resources (e.g., wetlands,
human exposure); and, at a minimum, requiring cleanup to a level where
the contamination must be proven to ensure continuing protection for
human health and the environment.
• Facilities with straightforward remedial solutions Many RCRA
facilities have similar types of contamination problems and, there-
fore, require similar types of remedial approaches. In these cases,
the most effective remedial alternative considered acceptable by both
the facility and the Agency is one which applies standard engineering
solutions that have proven effective in similar situations. The equiva-
lent to "clean-closure" under RCRA. or the construction of a RCRA
protective cover or liner are examples of straightforward remedial
approaches which may be considered effective remedial solutions
to Agency personnel under certain conditions.
• Facilities providing well-developed, phased remedies - It is becom-
ing apparent that most environmental contamination problems
encountered during the CAP involve facilities where the nature of
the environmental problem often dictates a singularly focused cor-
rective action approach (e.g., cleanup of groundwater contamina-
tion). In these cases, the Agency has recognized that the most effective
remedial plan must consider the milestone information gathering
process (e.g.. assessment of ongoing monitoring data) developed as
part of the requirements of the CAP.
Another example of a facility requiring a phased approach is one
where there is one particular area of the facility that deserves imme-
diate measures to control further environmental degradation or exposure
problems. In these situations, it is in the best interest of all concerned
parties that the corrective action phases focus first on that specific ele-
ment of the overall remedy requiring immediate attention (e.g., providing
immediate and adequate containment of the contaminant source), with
follow-on corrective action developed as appropriate to deal with the
remaining lower priority remedial needs at the facility.
The other type of bask corrective action approach is one which would
likely need relatively extensive, interdisciplinary environmental studies
to be done to support sound remedy solution decisions. Facilities falling
in this category include the following:
• Facilities considered to be a "high risk" - These RCRA facilities
are marked by environmental conditions where the scope of the
anticipated corrective action is expected to involve complex remedial
solutions. These types of facilities typically have large volumes of
uncontained, concentrated wastes impacting any or all of the environ-
mental release pathways (e.g., soils, groundwater, surface water, soil
gas or air). Therefore, the most effective means to remediate such
a complex contaminant release scenario is to apply several different
treatment technologies in order to achieve the varying degrees of
remedial effectiveness (i.e., reduction of toxicity or volume) in each
of the affected environmental release pathways. In conjunction with
this effort, different types of containment systems must be considered
for each pathway for whatever residual contamination is expected
to remain during the treatment process.
• Facilities with various appropriate remedial concepts - These facili-
ties possess environmental problems for which there may be several
distinct technical approaches. all*of which are considered practica-
ble. While each of these remedial strategies may offer varying degrees
506 GROUNDWATER TREATMENT
-------
of long-term reliability, and would be implemented in a phased man-
ner over different time-frames, the costs associated with these sub-
stantially different remedial approaches also would be, in turn,
expected to be substantially different. In cases such as this, the final
remedial selection decided upon by the Agency and the owner/oper-
ators will necessarily involve a highly interactive information trans-
fer process, involving a balance of competing goals and interests.
Such decisions must be supported with adequate information.
Components
The proposed rule includes various information collection milestones
throughout the duration of the RCRA CAP process, most notably the
RFI and CM phases. In order to provide an effective corrective action
framework, these components would be expected to include the
following elements:
• Permitting procedures and permit schedules of compliance - As part
of the current RCRA permitting process, owner/operators are finding,
as part of their final operating permit, an attachment which calls for
specific corrective action measures. Most land disposal facilities
underwent the initial CAP investigative phase (i.e., RFA) prior to
issuance of the permit. The corrective action required as part of final
permit conditions usually entail the initiation of the RFI. These
attached permit conditions include specific milestones (e.g., reporting
requirements) and associated schedules of compliance.
• Trigger or "action levels" - During the investigation process (i.e.,
RFI), enormous amounts of media specific, environmental data are
likely to be generated. The ultimate interpretation of the RFI may
come down to the comparison of a single data point to another back-
ground, standard or reference number. Extreme care should be taken
at this stage of the CAP since a significant difference between the
two numbers may also represent the difference between costs
associated with no further action and costs stemming from develop-
ment of further corrective action (e.g., CM).
• Corrective measure study and remedy selection - If a trigger or action
level has been significantly exceeded, as the initial part of the CM
phase of the CAP, the owner/operators would have to conduct a
Corrective Measures Study (CMS). The recommendations of the
CMS (i.e., an evaluation of the potential cleanup remedies) should
allow the owner/operators to propose a single, acceptable remedial
alternative. However, the owner/operators of large sites with diverse
waste management operations, and hence, potentially more complex
environmental problems, may need to pursue several varied, remedial
alternatives.
• Cleanup levels - It is the goal of the CAP to clean up releases of
hazardous waste or constituents to levels determined to be protec-
tive of human health and the environment. In response to the "How
clean is clean?" question, the revised draft rule defined levels, specific
for each of the environmental release pathway medium, that are safe
for both current and future land use. Although media specific cleanup
levels remain a goal of the CAP, there may be cases, however, where
these cleanup levels are not achieved. Obviously, in cases such as
this, owner/operators must expect to be involved in continuing and
sometimes long-term management until the appropriate cleanup levels
are reached.
• Standards for management of corrective action waste - During the
implementation of the field tasks at an impacted RCRA facility, it
is expected that hazardous wastes will be generated as a result of
these various investigative tasks (e.g., wastewaters, contaminated
media). The revised draft rule has performance standards for
conducting proper waste handling during the CAP. Certainly, if cor-
rective action waste meets the RCRA regulatory definition of
hazardous, it would have to be managed as a hazardous waste. It is
anticipated that some facilities may elect to construct new waste
management units in order to achieve CAP cleanup goals. In cases
such as this, these new units also would be required to comply with
necessary performance standards (e.g., 40 CFR Part 264). In addi-
tion, only RCRA permitted Subtitle C facilities would be able to
receive off-site shipments of hazardous waste.
• Completion of remedy - In order to verify that remedial action at
a RCRA CAP site has been successfully completed, the Agency must
utilize a recognized approach. Similar to other closure operations
under RCRA, an independent engineer or other qualified professional
would have to certify completion of the remedy. However, in some
cases, cleanup goals as defined in the permit may not be achieved.
In cases such as this, the Agency has opted in the past for additional
investigation to determine if the key factors in the interpretive process
(e.g., validity or representativeness of the cleanup standard) are defen-
sible. If not, new standards as the result of subsequent CMS derived
data are a realistic consideration to the owner/operators who cannot
achieve each of the media specific cleanup standards. Certainly, if
the environmental contamination remained at levels unprotective of
human health and the environment, other long-term release controls
are likely to be considered in order to prevent continuing human and
environmental exposure.
RCRA FACILITY ASSESSMENT
The initial phase of the RCRA CAP is comprised of the RCRA Facility
Assessment (RFA). The objective of the RFA is to identify releases
or potential releases or hazardous waste or constituents requiring further
investigation.
Purpose of the RFA
The RFA is a three-stage process, the purpose of which is to provide
the following:
• The identification and gathering of information on hazardous waste
or constituent releases at RCRA facilities
• The identification and assessment of SWMUs and other areas of con-
cern for releases to all environmental pathway media; assessment
of regulated units for releases to media other than groundwater
• The development of preliminary determinations regarding releases
of concern and the need for further actions and interim measures
at the facility
• The determination of those SWMUs which do not prose a threat to
human health or the environment
During the RFA, Agency or Contractor investigators gather infor-
mation on SWMUs and other AOCs at RCRA facilities. They evaluate
this information and determine whether there are releases that warrant
further investigation or other action (e.g., structural integrity testing)
at these facilities. Following the completion of the RFA, Agency
personnel expect to have sufficient information to determine the potential
for the likelihood of release from any SWMU or other AOC. Conse-
quently, the completion of the RFA is an information milestone, whose
conclusions and recommendations are designed to indicate if there is
a need to proceed to the second phase (RFI) of the CAP.
The RFA has been developed as three distinct phases. All three phases
of the RFA require the collection and analysis of data to support initial
release determinations:
• The Preliminary Review (PR) This phase focuses primarily on
evaluating available existing information, such as inspection reports,
permit applications, historical monitoring data and interviews with
Agency personnel who are familiar with the facility.
• The Visual Site Inspection (VSI) - This phase of the RFA entails the
on-site collection of visual information to obtain additional evidence
of release. The VSI typically is comprised of personnel from the
Regional EPA office, the State office and supporting contractors.
• The Sampling Visit (SV) - This optional RFA phase is designed to
fill any data gaps that remain upon completion of the PR and VSI
by obtaining sampling and field data. This phase may be by-passed
in the RFA phase and reintroduced as the initial step in the RFI,
namely the "verification investigation."
Scope of the RFA
The scope of the RFA includes all areas of potential release at RCRA
facilities and includes the investigation of releases to all environmental
pathway media, namely:
• Soil
• Groundwater
GROUNDWATER TREATMENT 507
-------
• Surface waters
• Air
• Subsurface gas
However, as previously mentioned, groundwater releases from regu-
lated units are not addressed in the RFA.
The types of units requiring investigation under the RCRA CAP are
based on the HSWA 3004(u) provision which focuses on investigating
releases from SWMUs at RCRA facilities. SWMUs are defined as:
• Any discernible waste management unit at a RCRA facility from
which hazardous wastes or constituents might migrate, regardless
of whether the unit was ever intended for the management of solid
and/or hazardous waste.
The SWMU definition includes:
• Containers, tanks, surface impoundments, waste piles, land treatment
units, landfills, incinerators and underground injection wells.
including those units defined as regulated unit* under RCRA
• Recycling units, wastewater treatment units and other units which
the U.S. EPA has generally exempted from standards applicable to
hazardous waste management units
• Areas contaminated by "routine and systematic" releases from process
and product storage areas
It should be noted that the SWMU definition does not include
accidental spills from production areas and units in which wastes have
not been managed (e.g., product storage or process areas). Routine and
deliberate releases from process areas are defined under the RFA as
other areas of concern.
The RFA is not intended to routinely address releases that arc either
permitted discharges (e.g., NPDES) or required to be permitted under
other environmental programs. Where such discharges are of concern.
the investigators refer the case to the original permitting authority.
However, the RFA does address releases from SWMUs to media other
than the one covered by the unit's discharge permit. For example, where
there is a cause for concern, the U.S. EPA can use the HSWA authority
(and as proposed, the Subpart S rule) to control the release of volatile
organic compound from NPDES-permitted wastewater treatment units
where there is a cause for concern.
The U.S. EPA purposely designed the RFA to be limited in scope;
that is to say. determining the potential for only the likelihood of release
Nevertheless, the RFA framework emphasizes the need to focus data
collection and analysis efforts (i.e., historical documentation and/or field
sampling data) that are required to support specific permit or enforce-
ment order conditions. Typically, if the Agency encounters suspect areas
during the RFA (i.e.. PR/VSI) but cannot verify it even though visual
conditions supported the likelihood of a hazardous waste or constituent
release, then an SV is employed as a "final verification" to the RFA.
A broad-based, analytical contaminant list (e.g., priority pollutants)
is often requested by the Agency since the investigator must make a
strong case to compel owner/operators to conduct an RFI or to con-
vince the public that a SWMU does not pose a threat.
The information requirements needed to trigger an SV will differ on
a case-by-case basis. The type and extent of sampling will depend on
the amount and quality of information gathered in the PR and VSI and
the investigator's professional judgment regarding the amount of infor-
mation necessary to support an initial release determination. If an SV
is initiated, it is likely that the investigators will sample those ureas
most visibly affected (e.g., stained areas, areas of stressed vegetation)
As the CAP is currently set up, the U.S. EPA and/or the states arc
responsible for conducting RFAs. Because of the subjective nature of
these investigations, the Agency believes that it is appropriate for a
regulatory agency to conduct the RFAs. These initial release determi-
nations will provide the basis for requiring further action ranging in
scope, for example, from no further action to a multimillion dollar in-
terdisciplinary hydrogeologic investigation The U.S. EPA and the stales
have used contractors to assist them in conducting these investigations.
but the Agency has retained overall responsibility for the RFA deci-
sions. In some instances, however, the facility owner/operator has par-
ticipated in the SV (e.g., obtained split samples).
Technical Approach
The technical approach of the RFA requires the investigator to exa-
mine extensive data on the facility and specific units at the facility. These
data generally can be divided into the following categories:
Facility and unit characteristics
Waste characteristics
Environmental setting
Pollution migration pathways
Evidence of release
Environmental receptors
Regulatory history
Previous release events
Specific factors in each category that must be considered will vary
depending on which environmental pathway medium is most vulnera-
ble. For example, unlined. in ground units are more likely to have soil
and groundwater releases than lined, above ground units. Also, certain
wastes tend to volatilize and cause air releases, while other wastes are
soluble in water and lend to migrate via surface or groundwater. A
facility's environmental setting may determine which media are of con-
cern (e.g., shallow groundwater or fractured subsoils). In addition,
further investigation at a facility may be triggered by the facility's poor
compliance record or unwillingness to cooperate with the Agency.
The RFA is completed when the Agency has sufficient information
to make a determination regarding releases or likely releases at the
facility and the need for further investigations. Upon completion of
the RFA. a summary RFA report is prepared integrating the findings
from all three steps in the RFA. This report generally includes the
following components:
• A description of the facility, its waste management practices and
regulatory history
• Release information for all SWMUs or groups of SWMUs and other
AOCs
• Sampling plan and results
• Final release determinations and recommendations
This RFA report indicates those areas of the facility that require further
investigation during the RFI and contains the key information (e.g.,
contaminant characteristics) to be used to focus these investigations.
RCRA FACILITY INVESTIGATION
As already noted, the RCRA Facility Investigation (RFI) is generally
equivalent in scope to the CERCLA remedial investigation. Units or
areas of concern that are determined in the RFA to be a likely source
of significant continuing releases of hazardous wastes or hazardous con-
stituents may be selected for an RFI. The regulatory means of requiring
the RFA is either through RCRA permit conditions (operating or
closure/post-closure) or via enforcement orders [e.g., 3008(h)J. Because
of the HSWA statutory language, the agencies must focus the RFI
requirements on specific solid waste management units or known or
suspected releases that are considered to be routine and systematic. The
HSWA permit conditions or enforcement orders may include supporting
fact sheets, and they can range from very general (e.g., "characterize
the groundwater at...") to very specific (e.g., a specified number, depth,
location and frequency of samples analyzed for a given set of consti-
tuents).
Since the Agency, in the RFA, is not required to positively confirm
a continuing release, but merely determine that the "likelihood" of a
release exists, the scope of the RFI can range from a limited, specified
activity to a complex multi-media study. The investigation may be
phased, initially allowing for verification or rebuttal of the suspected
continuing release(s). If verified, the second phase of investigation con-
sists of release characterization. This second phase, much like an RI,
includes: (I) the type and quantity of hazardous wastes or constituents
within and released from the SWMU, (2) the media affected by the
relcase(s), (3) the current extent of the release and (4) the rate and direc-
tion at which the releases are migrating. Inter-media transfer of releases
(e.g., evaporation of organic compounds from contaminated soil to the
atmosphere) is also addressed during the RFI, where applicable.
508 GROUNDWATER TREATMENT
-------
In completing the investigative effort, the regulatory agency, in con-
cert with the owner/operators, interprets the release findings. The first
emphasis of the investigation is on the data quality (i.e., were sam-
pling and analytical data quality objectives defined and accomplished?).
The findings are then compared against established human health and
environmental criteria. These criteria or "action" levels are available
for each environmental medium and exposure pathway, taking into
account the lexicological properties of the constituent and standardized
exposure assumptions. At this stage, if the continuing release of
hazardous wastes or constituents is determined to present a potential
short-term or long-term threat to human health or the environment,
interim corrective measures or a corrective measures study may be
required. This evaluation is a crucial stage in the corrective action
process.
Identifying and implementing interim corrective measures may be
conducted during the RFI. This would occur in a case where, in the
process of conducting the investigation, a condition is identified that
indicates that adverse exposure to hazardous constituents is presently
occurring or is imminent. Where interim corrective measures may be
needed, both the owner/operators and the regulator agency have a con-
tinuing responsibility to identify and respond to emergency situations
and to define priority situations. If first identified by the owner/opera-
tors, the need for interim corrective measures should be communicated
to the regulatory agency at the earliest possible time. As indicated earlier,
the need for close interaction between owner/operators and the regula-
tory agency is very important, not only for situations discussed above,
but also to assure the adequacy of the data collected during the RFI
and the appropriate interpretation of those data. Of course the
owner/operators benefit from this exchange by allowing efforts to focus
on salient issues and minimizing costly misinterpretations or unneeded
characterization efforts.
General RFI Implementation Strategy
An investigation of releases from SWMUs requires various types of
information. This information is specific to the waste managed, unit
type, design and operation, the environment surrounding the unit or
facility and the medium to which contamination is being released.
Although each medium will require specific data and methodologies
to investigate a release, a general strategy for this investigation can be
described. This strategy can consist of two elements: one is "desk top''
in nature and the other focuses on the field:
• Conceptual Model Development - Collection and review of data to
be used in developing a conceptual model of the release that can be
used to plan and develop monitoring procedures. These data could
include existing information on the facility/unit or related monitoring
data, data which can be gathered from outside sources of informa-
tion on parameters affecting the release, or the gathering of new
information through such mechanisms as aerial photography or waste
characterization.
• Phased Held Investigations - Formulation and implementation of field
investigations, sampling and analysis, and/or monitoring procedures
designed to verify or rebut suspected releases (Phase 1) and to evaluate
the nature, extent and rate of migration of verified releases (Phase 2).
The latter phase can in turn be divided into logical technical steps.
Varying amounts of information will exist on specific releases and units
at the start of the RFI process. In some instances, suspected releases
may have been identified based on strong evidence that releases have
occurred, but with little or no direct data confirming their presence.
On the other end of the spectrum, there may be enough existing data
at the start of the RFI for the investigator to begin considering whether
some form of corrective measure may be necessary. This potentially
broad spectrum of situations which may exist at the beginning of the
RFI often calls for a flexible approach for the release investigation.
Thus, the steps given above allow a logical progression from general
knowledge of a unit and its potential for a continuing release toward
a detailed (or "adequate") knowledge of the situation.
The value and role of the conceptual model element of the RFI is
in providing a foundation upon which to design subsequent characteri-
zation efforts. The conceptual model may be as simple as a tabular and
graphical depiction of the perceived situation. On the other hand, this
model can include realistic and worst case fate and transport modeling
of known contaminants under the given site conditions. Regardless of
its complexity, the conceptual model consists of the following:
• SWMU or area description and an estimate of waste distribution in
that unit
• Estimated quality and quantity of waste present, including specific
constituents
• Environmental setting of the unit (e.g., soils, surface and subsur-
face hydrogeology and climate) and its vulnerable contaminant trans-
port pathways
• An estimation of how, how fast and where known or suspected con-
taminants would be transported and transferred between compartments
• An evaluation of what media would be most likely to be monitora-
ble for detecting any releases
The role of this model is, of course, to determine in broadest terms
whether significant release potential is present and, given a significant
potential, how to design an investigation/monitoring program capable
of release verification and/or characterization.
As already noted, the release characterization may be conducted in
phases, if appropriate, with each monitoring phase building on the
findings and conclusions of the previous phase. The overall level of
effort and the number of phases for any given characterization effort
depends on various factors including:
• The nature of the potential contaminants
• The level of data and information available on the site
• The complexity of the release (e.g., number of units, release path-
ways, affected media)
• The overall extent of the release
Field Investigation Strategies and Techniques
Entire books can be, and indeed have been, written on the topic of
field investigations for environmental characterization. Furthermore,
the colossal task of discussing RFI methods is multiplied by the fact
that any medium (e.g., soil, air, groundwater, etc.) might be involved
in a given RFI. Rather than give limited and clearly inadequate coverage
to these concerns, the focus here is on selecting an appropriate approach
to investigation in the context of corrective action program objectives.
The first effort following identification of the significant hazardous
constituents present in the unit or release area is an evaluation of the
likely compartment in which a given constituent will be found. Based
on chemical, physical and biological properties of the constituent rela-
tive to environmental media (e.g., air:water partition coefficient), the
evaluation of environmental compartment or medium helps to deter-
mine which media should be sampled to characterize whether a release
is occurring. A simple example would be to use soil gas monitoring
to detect migration of volatile organic compounds. The remaining aspect
that may be determined from the compartmental evaluation would be
the detection or analytical methods to be used, the expected detection
limits and the data quality objectives.
The next effort should entail selection of sampling and/or testing tools
or techniques. Various methods exist for obtaining acceptable samples
of waste and for each medium. The following criteria should be consi-
dered in choosing such methods:
• Representativeness The selected methods should be capable of
providing a true representation of the situation under investigation.
• Compatibility with Analytical Considerations - Sample integrity must
be maintained to the maximum extent possible. Errors induced by
poorly selected sampling techniques or equipment can result in poor
data quality. Special consideration should be given to the selection
of sampling methods and equipment to prevent adverse effects during
analysis. Materials of construction, sample or species loss, and a
chemical reactivity are some of the factors that should receive at-
tention.
• Practicality - The selected methods should stress the use of practi-
cal, proven procedures capable of being used in or easily adapted
GROUNDWATER TREATMENT 509
-------
to the given situation.
• Safety - The risk to sampling personnel and others, intrinsic safety
of instrumentation and safety equipmeni required for conducting the
sampling should be carefully evaluated.
Finally, the specific sampling/monitoring design may be chosen. This
amounts to selection of sample numbers, locations, depths and timing.
Because conditions present in the unit or in the contaminant release
will change both temporally and spatially, the design of the sampling
program or monitoring network should be developed accordingly. Spa-
tially, sufficient samples should be collected to adequately define the
extent of the contamination. Temporally, the plan should address spread-
ing of the release with time and variation of concentrations due to fac-
tors such as changes in unit operations, the environment surrounding
the unit, and the composition of the waste. For example, when possi-
ble, sampling and supplemental measurements (e.g., wind speed) should
be conducted when releases arc most likely to be observed.
It must be emphasized that investigations must consider and include
relevant physical and descriptive data and information associated with
the samples or the media sampled. This evaluation process is especially
critical where computer modeling of fate and transport is to be included
in the evaluation. Lack of sufficient pertinent physical and descriptive
data can render an investigation almost useless.
RFI Decision Points
As monitoring data become available, both within and at the conclu-
sion of discrete investigation phases, they typically are reported to the
regulatory agency as directed. The regulatory agency will compare the
monitoring data to applicable health and environmental criteria to
determine the need for
• Interim corrective measures
• A Corrective Measures Study
In addition, the regulatory agency will evaluate the monitored data with
respect to adequacy and completeness to determine the need for any
additional monitoring efforts. Notwithstanding this process, UK
owner/operators have a continuing responsibility to identify and respond
to emergency situations and to define priority situations that may warrant
interim corrective measures. For these situations, it is suggested that
the owner/operators obtain and follow the RCRA Contingency Plan
requirements under 40 CFR Part 264, Subpart D.
As a final note, the same is true of both the CAP and CERCLA
investigations. The owner/operators and the responsible parlies, respec-
tively, should maintain a significant presence throughout the process
and not rely on varied and inconsistent oversight from the agency. This
is especially true in providing realistic interpretations of findings, es-
pecially where transport and fate considerations may affect the interpre-
tation of what constitutes a "continuing release."
CORRECTIVE MEASURES
In addressing releases from SWMUs to the environment, the RFI
is followed by Corrective Measures. That is, a release, and hopefully,
a source of contamination, have been identified, and the owner/opera-
tor must initiate a remedial response. As in the Superfund program,
remedial action objectives of corrective action are site-specific and quan-
titative goals that define the level of cleanup arc required to achieve
the response objectives. These goals include any preliminary cleanup
levels for environmental media affected by a release, the area of attain-
ment and the remedial time-frame.
The mentioned objectives arc accomplished through the Corrective
Measures Study (CMS) and the Corrective Measures Implementation
(CMI) by identifying, designing and implementing the appropriate
remedial strategy, all in accordance with published CM guidance. The
CMS serves as a recommendation to the U.S. EPA or the State, while
the CMI is the allowed time frame for the actual corrective measures.
Corrective Measures Study
The first step in the CM phase is the development and implementa-
tion of the CMS to determine the most effective remedial option to
correct potential environmental impact and human exposure threats
posed by releases of hazardous wastes or constituents. Regardless of
whether the remedial response effort u conducted under CERCLA or
RCRA authority, the objectives of the CMS, or feasibility study, are
to utilize technical knowledge and propose actions to control the source
of the contamination (by preventing or mitigating the continued migra-
tion of contamination by removing, stabilizing and/or containing the
contaminants) and/or actions to abate problems posed by the migra-
tion of substances from their original source into the environment.
Through the CMS, the owner/operator must technically demonstrate
that the response action proposed effectively abates the threats to human
health and the environment posed by the releasc(s). This typically
requires the analysis of several remedial technologies in detail suffi-
cient to show that the recommended measures effectively remove (he
threats posed by the release. To do so, the owner/operators must assess
these alternatives in terms of their technical feasibility (including relia-
bility and requirements for long-term operation and maintenance), their
ability to meet public health protection requirements and their ability
to protect the environment and any adverse environmental effects of
the measures. The owner/operator also should consider any institutional
constraints to implementation of the measures, such as off-site capaci-
ty problems and potential public opposition.
The RCRA approach to assessing the level of remedial action required
for environmental media is similar to that of CERCLA W7 and generally
is based on the following criteria:
• Overall protection of human health and the environment
• Compliance with regulatory programs (e.g.. CERCLA or RCRA)
• Short-term effectiveness
• Long-term effectiveness and permanence
• Reduction of loxicity. mobility or volume of hazardous wastes and/or
waste constituents
• Implemcntability
• Cost
• U.S. EPA and/or State acceptance
• Community acceptance
The first two criteria are the basic regulatory requirements, while
the next five criteria are interactively used to analyze and compare the
options. The final two criteria are considerations in the overall
evaluation.
In some cases, it is possible for owner/operators to analyze and present
to the Agency or State only a single alternative that meets public health
and environmental requirements. This situation is often the case at
facilities that have taken "interim corrective measures" and thus have
had an opportunity to evaluate the remedial strategy and the associated
operations to determine their effectiveness. This solution is appropriate
when the U.S.EPA or the State agree that the remedial alternative the
owner/operator proposes is likely to effectively achieve corrective action
goals, including health and environmental requirements, and is techni-
cally sound. In most cases, however, given the array of feasible tech-
nologies, it may be necessary to analyze more than one alternative to
determine the appropriate response measure. For example, off-site or
on-site alternatives may be considered or there may be a difference of
opinion as to whether a particular alternative the owner/operator
proposes to analyze would be reliable or effective in abating threats
expeditiously. In such cases, the U.S. EPA or the State would require
the analysis of several alternatives to ensure that appropriate response
measures are completed on a timely basis and that response is not
delayed by a sequential analysis of a series of alternatives.
RCRA final remedies will be required to meet applicable, possibly
current, health and environmental standards promulgated under RCRA
and other laws. For example, at regulated units, groundwater releases
are subject to the groundwater protection standards, possibly consisting
of the following:
• Constituent specific maximum concentration limits (MCLs)
• The background level of that constituent in groundwater
• An approved alternate concentration limit (ACL) where approval
would be based on criterion set forth in the RCRA regulatory
framework
510 GROUNDWATER TREATMENT
-------
For soil, soil gas, surface water, groundwater and air emissions
problems that cannot be addressed by existing standards, the Agency
currently is assessing the appropriate technical approach. One possi-
ble alternative is to establish appropriate health-based standards on a
case-by-case basis.
Once the owner/operator proposes the remedial strategy(s) for
addressing releases to environmental media and the SWMU itself, the
U.S. EPA or the State will evaluate the owner/operator recommenda-
tion and approve or disapprove it. During the review process, the
owner/operator must be prepared to provide the technical support for
his or her proposition and must be open to negotiations. The views
of the public on the proposed measures and the financial assurance
demonstration also will be considered by the State and U.S. EPA in
making these decisions.
Corrective Measures Implementation
Once the U.S. EPA, the State, and the owner/operator agree on the
remedial approach, the owner/operators will design and construct the
selected response action. After construction, the appropriate measures
needed to operate, maintain and monitor the remedy will be taken by
the owner/operators. These activities will be required by permit con-
dition or compliance order and will be performed by the owner/opera-
tors with oversight by the U.S. EPA or State. Since the actual operations
serve to provide data concerning the effectiveness of the corrective ac-
tion, it is essential that these data are used as criterion in determining
whether the operations should be modified over time to meet the cleanup
objectives.
Effecting remedies (or interim measures) at facilities that do not have
RCRA permits will, in some cases, involve creating new treatment,
storage or disposal units. Rather than going through the actual process
of issuing RCRA permits to such new units, which could substantially
delay implementation of the remedy, the Agency is considering using
enforcement authorities and closure plan regulatory authorities to allow
those units to be constructed and operated without a formal RCRA
permit. The U.S. EPA may need to amend existing regulations to provide
for this proposed approach. Such new units would nevertheless be
required to generally comply with applicable Part 264 technical
standards, and appropriate public review and comment would be
provided.
REFERENCES
1. U.S. EPA, Interim Final RCRA Facility Investigation (RFI) Guidance, Vol.
I, Waste Management Division, Office of Solid Waste, EPA 530/SW-89-031,
May, 1989.
2. U.S. EPA, RCRA Facility Assessment Guidance, Permits and State Programs
Division, Office of Solid Waste, Aug. 14, 1986.
3. U.S. EPA, Draft RCRA Preliminary Assessment/Site Investigation Guidance,
Permits and State Programs Division, Office of Solid Waste, Aug. 5, 1985.
4. Porter, W. J., Memorandum, Assistant Administrator, Office of Solid Waste
and Emergency Response, Subject: National RCRA Corrective Action
Strategy, Oct. 14, 1986.
5. U.S. EPA, Guidance on Remedial Actions for Contaminated Ground Water
at Superfund Sites, Office of Emergency and Remedial Response, Dec., 1988.
6. U.S. EPA, Subpart S: Corrective Action for Solid Waste Management Units
40 CFR 264.500-264.560 (Revised Draft Sep. 12, 1988).
7. Stoll, R. G., Chapter 3: Comprehensive Environmental Response, Compen-
sation, and Liability Act (CERCLA or Superfund), 10th Ed., Government
Institutes, Inc. Rockville, MD.
GROUNDWATER TREATMENT 511
-------
Improved Methodology for Constructing Soil Liners
Using Highly Plastic Clays
Joseph M. Cibor, P.E.
G. Rai Mehdiratta, Ph.D., P.E.
McClelland Consultants (Southwest), Inc.
Houston, Texas
J.D. Martin, P.E.
Consultant
Port Lavaca, Texas
ABSTRACT
Use of highly plastic clays for construction of soil liners in haz-
ardous waste landfills has been discouraged partly due to their
adverse shrink-swell characteristics but primarily because it is dif-
ficult to eliminate macrovoids between soil clods and along inter-
lift boundaries. The macrovoids and lift laminations cause in situ
hydraulic conductivities to exceed the mandated maximum
hydraulic conductivity of 1 x 10-7 cm/sec.
Faced with no locally available source of low plasticity clay, it
was decided to experiment with various aspects of the compaction
process, including the effects of moisture conditioning, compac-
tion effort, clod size, equipment type, lift thickness, sequencing
of operations and penetration of compactor feet. This was done
to evaluate whether highly plastic clays could be compacted with
minimum of macrovoids, thereby increasing the probability of
successfully meeting hydraulic conductivity requirements.
This paper outlines an improved methodology for constructing
soil liners using highly plastic clays. The methodology was devel-
oped by varying eight aspects of the compaction process during
construction of two test pads and observing migration of dye.
The results of two successful SDRI tests (instrumented for swell
and movement of wetting front) also are presented along with
suggestions for improving the test procedure.
INTRODUCTION
Historically, below-grade landfills consisted of pits excavated
through clayey soils. The depth of the pits often was controlled
by the location of groundwater. At times, the landfills were lined
with imported clay installed using earthwork techniques similar to
those employed in the construction of embankments and road-
ways.
More recently, the influence of governmental regulations,
public sentiment and industry's concerns have led to improved
technology in design and operation of landfills. Use of composite
liners became widespread. However, research and development
of synthetics far outpaced advances in the construction of earthen
liners.
Monitoring the quality of clay liners centered around field
density testing and laboratory measurement of permeability. By
the 1980s, studies by Daniel4'6'8 and others2'3'9'10'11 suggested that
permeability of clay liners may be influenced by the presence of
macrovoids and laminations; development of in situ permeability
equipment3'7 demonstrated that field permeability of compacted
clay liners could be several orders of magnitude greater than
measured in the laboratory.
Many of the industrial facilities in southeast Texas are situated
within the Coastal Plain geological province. The near-surface
geologic units of this area consist mainly of Pleistocene Beau-
mont Formation clays. Derived from backswamp or overbank de-
posits laid down in quite environments adjacent to ancient river-
beds, the clays are typically highly plastic. Sources of low plastic-
ity, clay, preferred for construction of liners, are few and often
located considerable distances off-site.
McClelland's experience with the highly plastic clays of south-
east Texas indicated that their compaction to achieve low
hydraulic conductivity, particularly the strictly enforced, field-
tested permeability of 1 x 10X ? cm/sec as mandated by the U.S.
EPA and state environmental agencies, could be difficult. We
were concerned that it would be difficult to break down the highly
plastic clays into small clods and to adjust their moisture content.
Moreover, the clay's toughness would make it hard to eliminate
inter-clod voids and lift interfaces. Lastly, the compacted clays
would exhibit a high shrink/swell potential.
Our concerns were underscored by published case studies2-410
and discussions with colleagues in industry, which demonstrated
that field testing of clay liners constructed employing conven-
tional earthwork techniques failed to meet maximum permeability
requirements.
We realized the economic benefits of using locally available
sources of highly plastic clay, but we also realized how difficult
it would be to work with these clays to meet hydraulic conduc-
tivity requirements. Cognizant of the consequences in terms of
schedule delays and/or reduced waste containment capacity if
field testing failed to meet requirements, it was decided to inves-
tigate the effectiveness of additives to improve workability of the
clays by reducing their plasticity.
The laboratory study led to full-scale field observations of vari-
ous aspects of the compaction process, from which an improved
methodology of constructing soil liners was developed. The meth-
odology was employed successfully at several sites in southeast
Texas using clays of high plasticity.
CLAY MODIFICATION STUDY
Recognizing that elimination of macrovoids and inter lift lam-
inations hinged on the workability of clay, which in turn de-
pended on its plasticity, our laboratory study was aimed at mod-
ifying clay plasticity. An additive commonly used in the Gulf
Coast area to reduce plasticity is lime. Although relationships
between lime content, plasticity and shrink/swell potential are
well documented, data regarding permeability of lime-modified
512 BARRIhRS
-------
clay are limited.
Lime was added in varying concentrations to two different
sources of highly plastic clay. Pertinent properties of the two clays
are shown in Table 1 below.
Table 1
Properties of the Clays Used in this Study
Brownish Yellow
Clay
Liquid Limit, %
Plastic Limit, %
Plasticity Index, %
68-79
21-25
47-54
Dark Gray
Clay
96-120
29-32
67-89
Ume Content, %
%Lime
After addition of lime, the soil-lime mixture was allowed to
"mellow" for three days before initiating laboratory testing.
Samples subjected to strength and permeability testing were first
compacted to specified density at selected moisture content and
allowed to cure in a humid room for 7 days. The mellowing and
curing periods were selected to simulate actual field conditions.
The effects of Ume content on strength, plasticity, pH and
moisture-density relationships are shown for both clays on Fig-
ures 1 and 2. As expected, addition of Ume improved workability
and strength and reduced the shrink/swell potential.
100
6 9 12 15
Lime Content, %
6 9 12
Lime Content, %
15
Figure 1
pH and Plasticity Index Versus Lime Content
18
BROWNISH YELLOW CLAY
18
2.0
BROWNISH YELLOW CLAY
O
6 9 12
Ume Content, %
I
,.- 8.0
6.0 -
?
W 4.0
|
•n 2.0
DARK GRAY CLAY
36 9 12
Lime Content, %
Figure 2
Variation in Soil Properties with Lime Content
Addition of even small amounts of Ume increased the perme-
ability of the brownish-yeUow clay by one to two orders-of-mag-
nitude. The permeability of the more plastic dark gray clay in-
creased by as much as 10,000 times. The results of the perme-
ability tests are shown in Figures 3 and 4.
o
I.
o
.a
8
V
a.
10^
icr5
10'6
1C'7
10-8
t
10'9,
10'10
BROWNIS
Requirec
Permeab
A
H YELLOW C
[lity\ :
b,
fi
t
i
t
A
LAY
LEGEND
A Lime
O Sarr
41.7%
^40.6%
L38.0%
•37.2%
^
'41.6%
VMC
Me
Co
3 Treated Clay
pie With 5% E
A
Iding
isture
ntent
entonite
) 3 6 9 12 1
Lime Content, %
Figure 3
Permeability Versus Lime Content
BARRIERS 513
-------
10"'
,-7
Ł 10
If/1
10--
10'
10
DARK QR
Required
Permeabi
4
t
AY CLAY
iity\ A
\o
!
A 42.0%
A 47.4%
A 44. 7%
t
A 423%
A41 6%
\ M
Ą
X
A
i
nlriinn
x Moisture
Content
LEGEND
A LJITM
O Sarr
X Rek
A
> Treated Clay
pie Witti 5% Berrtonite
1 Compacted Samples
i
Table 2
Properties of dan Obtained al S«»d» Site
OM8ITE OFPSITE
6 9 12 15
Lime Content, %
Figure 4
Permeability Variation with Lime Content
A majority of lime-modified samples of the brownish yellow
clay met the required permeability benchmark of 1 x 10-7 cm/
sec, but by a narrow margin. This was not true for the dark gray
clay.
Of greater concern to us was the scatter (one to two orders-of-
magnitude) in measured permeabilities at a given lime content.
We hypothesized that this observed increase in permeability
upon addition of lime could be attributed to the "granulation" of
the mixture during mellowing and possibly be due to the forma-
tion of macrocracki in the more brittle lime-modified clay.
Our study showed that addition of lime significantly unproved
workability. However, considering the scatter in data measured
under controlled laboratory conditions, and realizing that varia-
tions in the field would be greater, we did not have confidence
that the lime-modified clays would successfully meet required
field permeability.
FIELD STUDY OF COMPACTION PROCESS
Over a period of 2 mo, we studied various aspects of the com-
paction process, guided by the goal of reducing the presence of
prominent interconnecting macrovoids and lift interfaces which
various investigators*' believed were responsible for the observed
high permeability.
Two different sources of highly plastic clay were used in the
study. The clay obtained at the study site, termed "On-site",
was more plastic than the "Off-site" clay obtained from a borrow
source located roughly 1.5 mi away. Pertinent properties of the
two clays are summarized in Table 2.
The following aspects of the compaction process were investi-
gated:
• Lift Thickness—the loose lift thickness was varied from 3 to
9 in.
• Moisture Content—moisture content was varied between 2°7o
below to 5% above optimum moisture content.
ClaiiKlcatloni
Liquid Limit, \
Plaittc Limit, %
Plattlclty Index, %
Sticky Limit, »
Specific Gravity
Percent Pas*Ing
No. 200 Slavs
OpilmujB Mnlature
Content, I
Ka • Imua Dry
Denalty, pcf
Brown and Yellow Clay
Mlth Silt Pocket* and
Gray to Light
Brown Sandy
Calcareous Nodulea Clay With Silt
70 to 80
21 to 24
48 to Si
26 to 27
2.72
95
21
SO to 60
IS to 17
35 to 43
22 to 23
2.70
(4
18
107
• Compactive Effort—three different compactors were used,
namely: (1) Caterpillar 815B, (2) double-drum sheepsfoot
towed by track-type tractor and (3) tamping foot compactor.
• Method of Reducing Clod Size—a disc and pulvimixer were
utilized. Number of passes of pulvimixer, depth of cut and
shield opening were varied.
• Coverage—number of equipment passes was varied to pro-
duce coverages of 100, 150 and 2004t.
• Effects of sequencing operations on uniformity of blending
and moisture conditioning.
• Location of Discing and Pulverizing—reducing clod size was
performed both on the trial pad and at an off-site location.
• Penetration of Compactor Feet—moisture content was varied
to produce compactor foot imprint of various depths.
Each of the eight factors described above was studied inde-
pendently and in various combinations. Variations in technique
were applied to both clay types. Careful observation and docu-
mentation, along with testing of density and moisture content,
were performed along each step.
The trial lifts were dissected routinely and inspected for mac-
rovoids and inter lift laminations. Absence of macrovoids and
laminations was considered to be paramount to the formation of
a successful technique.
It quickly became apparent, too, that compaction techniques
were bound by such factors as the "sticky limit" of the soil, a lit-
tle used Atterberg limit, the bearing capacity required to sup-
port compactors and the consistency of clay required for traffic-
ability of various other earthwork equipment.
After repeated trials, a method of placing, processing and com-
pacting the highly plastic clay, in a way that macrovoids and
prominent laminations were not observed, was developed. This
methodology is described in the following section.
METHODOLOGY DEVELOPED
The methodology developed by this study for construction of
liners using highly plastic clays significantly varies from conven-
tional earthwork techniques, yet follows an approach which is ex-
pedient, practical and verifiable. The methodology considers clod
size, moisture conditioning, lift thickness, compactive effort, cov-
erage and tamping foot penetration. Each of these six concepts is
described in detail below.
Clod Size
The effective clod size of the clay is broken down to an effec-
tive diameter of less than 3 in. using a pulvimixer, preferably
working off the pad. At least one pass of the pulvimixer is re-
quired. Discing alone was proven ineffective, but may be consid-
ered before pulvimixing. Maintaining proper shield opening
greatly affects both productivity and clod size.
514 BARRIERS
-------
Lift Thickness
The loose lift thickness is maintained at less than 6 in., produc-
ing a final compacted lift thickness of 4 in. As the plasticity of the
clay increases, consideration should be given to reducing maxi-
mum loose lift thickness to 4 in. Our experience suggests that use
of laser mounts on grader produces consistent and verifiable re-
sults.
Compactive Effort
A heavy-duty compactor (Cat 81 SB or equivalent with a gross
weight of 20 to 25 tons) is specified with a tamping foot projec-
tion of at least 7 in. and preferably 9 in. Sheepsfoot or medium to
light-duty compactors are considered ineffective in working with
the tough, highly plastic clays.
Coverage
Compaction should be controlled by the number of passes re-
quired for 150% coverage. Coverage, not density, should control
the number of passes (our experience indicates that minimum
density requirements usually will be met with the coverage and
compactive effort specified). Uniform coverage by the compactor
is essential.
Moisture Conditioning and Tamping Foot Penetration
The moisture content should be adjusted to above Compactor's
Optimum, defined as moisture at which tamping feet produce at
least 4-in. imprint following 150% coverage, but do not fully pen-
etrate (i.e., the drum is not in contact with soil). The desired
tamping foot penetration at initial pass and following 200% cov-
erage is illustrated in Figure 5. A rule-of-thumb which we found
useful in estimating the compactor's optimum moisture content
can be stated as follows: moisture content should be above opti-
mum moisture content (ASTM D 698) but below the "sticky
limit" (so that the clay does not stick to the drum). The "sticky
limit" is a little used Atterberg limit.1 one of seven limits devel-
oped by a Swedish soil scientist, A. Atterberg, in the early 1900s:
Figure 6 demonstrates the impact of varying moisture on penetra-
bility.
4"-6"
Loose Lift
Intermediate Lift
Fully Compacted Lift
INITIAL PASS
Loose Lift
Intermediate Lift
Fully Compacted Lift
AFTER 200% COVERAGE
Figure 5
Compactor Foot Penetration
Too Dry
l'-2" Imprint
vfc*^;
ft* I
Proper Moisture
4"6" Imprint
Too Wet
Clay Sticks
To 1
•-*&i& iMJ^^"^^ &Ł •
ST* ™"^'- • -^ ^ •J?^.^'
^y^y ••'. ^ "iai^f^
^^"'^^m' A'f|f|__
Figure 6
Effect of Moisture Content on Penetration
VERIFICATION OF RESULTS
Extensive field and laboratory studies were performed to check
BARRIERS 515
-------
if the newly developed methodology was successful. Field studies
included test trenches and dye penetrant testing. Laboratory
tests were performed to study overburden pressure versus swell
relationship.
Test Trenches
Test trenches were excavated through the compacted trial pads
to check for the presence of macrovoids, laminations, lift inter-
faces, homogeneity of the soil mass and any other construction
defects. To remove smearing caused by the backhoe during
trench excavations, portions of the trench were carefully trimmed
with a pocket knife. Typical trench cuts are shown on Figure 7.
The trenches revealed a homogeneous soil mass with extensive
mixing of soil colors and mosaic-like patterns. There was a con-
spicuous absence of lift interfaces and other construction defects.
These tests showed that the dye only penetrated a thin veneer of
soil (about 1/8 in.); the soil below this veneer was not stained.
Penetration along lift interfaces or interconnecting channel* alto
was not observed. The extent of dye penetration is typified on
Figure 8.
Figure 7
Cut Through On-Site CUy Pad
Dye Penetranl Testing
Dye penetration tests were made to evaluate the extent of large,
interconnecting macrovoids which may not have been apparent
under visual inspection. Four square prefabricated steel rings
measuring 6 ft by 6 ft were installed in the pads to observe vertical
defects: Seven 4-in. diameter PVC standpipes were installed in
boreholes to evaluate the extent of interconnecting horizontal
voids. The dye consisted of powdered methylene blue dissolved
in water at a concentration of 3 gm/gal. The rings and boreholes
were dissected approximately I wk after the dye was introduced.
Figure 8
Crow-Section—Dye Penetranl Te»t
IN SITU TESTING
The rate of infiltration of water into the test fills was measured
using the sealed double ring infiltrometer (SDRI) technology de-
scribed by Daniel and Trautwcin.'-7 Nine tensiometers at three
different depths were used to estimate the advance of the wetting
front and, thereby, to obtain a better estimate of the hydraulic
gradient during performance of the test. Four swell monitors were
used to estimate the quantity of water being held by the soil to en-
able calculation of the quantity of water flowing through the soil.
A schematic diagram of the testing apparatus is presented on
Figure 9.
Results of SDRI Tests
The computed hydraulic conductivity values are presented on
Figure 10 as a function of time for both the on-site clay and the
off-site clay. The dashed curves represent hydraulic conductivity
corrected for temperature, but uncorrected for swell. Hydraulic
conductivity corrected for both swell and temperature is plotted
as a solid line.
M6 HARRII-.KS
-------
Flexible
Water
BagS/Water Level Gauge
ling
Molded Berm
'^~I<<<<<^^^y^^--^--^-----?^ Compacted day Test Pad HX:-!-:
CROSS SECTION
Ancti
-
_ Water Level
Gauge
_ 2.3000 cc
Flexible Water Bags
Supported On
Masonary Blocks
The curves show decreasing permeability with time; this trend
is typical of other SDRI tests we have performed. We believe that
initially the computed permeability is high due to swelling of the
soils and disintegration of the upper few inches of the clay. As the
wetting front advances, the overburden pressure increases, reduc-
ing swell which we attributed to be responsible for fracturing of
the compacted mass, and permeability decreases. Figure 11 shows
the rate of swell and the rate of advance of the wetting front with
time.
40
30
E
Ł
= 20
10
ONSITE CLAYS
\
OJ
.c
o
c
3S
> f,^
i I I 11 i i i i I i W i i
1X1U""0 5 10 15 20 25 30 35 40 45 50 55 60
Elapsed Time, Days
OFFSITE CLAYS
1X10"
5 10 15 20 25 30 35 40 45 50' 55
Elasped Time, Days
Figure 10
Hydraulic Conductivity vs Elapsed Time
30
20
10
OFFSITE CLAYS
Swell
Wetting Front
I
12
18
8
0 5 10 15 20 25
Elapsed Time, Days
Figure 11
Swell and Movement of Wetting Front vs Time
Laboratory permeability tests were performed on compacted
samples prior to the start of the SDRI tests. These tests gave
permeability values on the order of 1 x 10~9 cm/sec or less.
Laboratory permeability tests also were performed on undis-
turbed thin-walled tube samples recovered from the test fill after
the completion of the SDRI tests. These laboratory tests gave
values ranging from 2.3 x 10-? to 1.2 X 10~9 cm/sec. These re-
sults agree well with the SDRI hydraulic conductivity data.
Limitations of SDRI Test
Although the SDRI test is useful in estimating the in situ
hydraulic conductivity of a clay liner or cap, the SDRI test has
several shortcomings. Further research and modifications to the
SDRI equipment should be considered to make it more useful and
representative of actual field conditions.
Swell and Overburden Considerations
A typical clay liner or cap will have some overburden. The
liners are generally overlain by a 1-ft thick leachate collection lay-
er; a cap may have top soil or other covering. However, the
BARRIERS 517
-------
SDRI setup does not account for any overburden. Lack of over-
burden results in swelling of high plasticity soils which destroys
clay structure in the upper few inches of the liner.
In addition, there is controversy as to how the swell correction
should be applied. We feel that the SDRI equipment should be
modified so that overburden pressure representative of actual
field conditions can be applied. Alternately, a thicker test fill
should be constructed and the upper portion (for example, 1 ft)
of test fill should be considered as the overburden.
Ideally, correlations should be established where SDRI data
with no overburden can be corrected for overburden effects.
Laboratory studies suggest that an increase in overburden pres-
sure reduces the hydraulic conductivity and a small amount of
overburden (for example, 1 psi) significantly reduces swell of soils
compacted wet of optimum moisture content.
Wetting Front and Hydraulic Gradient
Accurate methods are needed to estimate depth of wetting
front during performance of the SDRI test. We utilized tensio-
meters to estimate the depth of wetting front. Tensiometers only
indicate when the wetting front reaches the tip depth and do not
give a continuous measure of the depth of wetting front. More-
over, if not inserted and grouted properly, the tensiometer can
leak around its sides giving erroneous results. A continuous read-
out device based on moisture changes, perhaps soil resistivity,
should be considered. The final moisture content profile of the
soil, after the test is completed, is a reliable method but it is after-
the-fact datum and cannot be used during performance of the
test.
Cost and Duration of Test
The SDRI is a very expensive and time-consuming test when
compared to a laboratory permeability test. In addition, only a
small area which may not be representative of the test fill is tested.
We believe that consideration should be given either to laboratory
tests performed on large diameter samples or expediting the field
test.
Miscellaneous Factors
Miscellaneous factors influencing SDRI tests include the
growth of algae in the test equipment and the changes in volume
of the water and the ring due to temperature fluctuations SDRI.
Misuse of SDRI Test Data
The quality of a liner depends in part on the clay (i.e., whether
a particular clay meets permeability requirements as measured by
the SDRI) used in its construction. Of equal if not greater impor-
tance in determining liner quality is the compaction methodology,
the compaction equipment, the experience of construction per-
sonnel and the quality assurance program." Highly plastic clays
can meet the 1 x 10~7 cm/sec permeability requirement pro-
vided a proper compaction methodology has been developed for
the particular clay.
CONCLUSIONS
The following conclusions were drawn as a result of this study:
• Highly plastic clays are difficult to compact and their use as soil
liner material has been discouraged. Yet, these soils are the pre-
dominant near-surface geologic unit in southeast Texas and
their use presents economic benefits.
• Addition of lime greatly enhances the workability of highly
plastic clays but increases permeability by 10 to 1000 times.
• A methodology has been developed for construction of liners
using highly plastic clays. This new methodology differs from
conventional earthwork techniques, yet follows an approach
which is expedient, practical and verifiable. The methodology
considers clod size, moisture conditioning, lift thickness, com-
pactive effort, coverage and tamping foot penetration.
• The SDRI test is useful in estimating in situ hydraulic conduc-
tivity. However, the SDRI test has several shortcomings which
particularly affect the outcome of tests performed on linen
composed of highly plastic clays. Most importantly, the test
does not account for overburden or swell.
• The compaction methodology, the equipment and experience
of its operators, and the QA/QC program are factors as im-
portant to the quality of an earthen liner as the type of clay
used in its construction.
REFERENCES
I. ASTM "Annual Book of Standards: Silt and Rock, Building Stones;
Peau." Part 19, Philadelphia, PA. 1986.
2. Boynton, S.S. and Daniel, D.E. "Hydraulic Conductivity Tests on
Compacted Clay." /. Geotech. Eng., ASCE. /// (4), pp 465-478,
1985.
3. Brown, K.W., Green, J.W., and Thomas. J.C. The Influence of
Selected Organic Liquid* on the Permeability of Clay Linen,"
Proc., Ninth Annual Research Symposium on Land Disposal of
Hazardous Waslt. EPA-600/9-83-018. Cincinnati. OH. pp 114-125,
1983.
4. Daniel, D.E., "Predicting Hydraulic Conductivity of Clay Linen,"
J. Geotech. Eng.. ASCE. 110(4), pp. 285-300. 1984.
5. Daniel, D.E., "Hydraulic Conductivity Tetti for Clay Linen."
Proc., Ninth Annual Symposium on Geotechnical and Geohydro-
logical Asptctt of Waste Management. Feb. 1987, Fort Collins,
CO. 1987.
6. Daniel, D.E., "Earthen Liners for Land Disposal Facilities," Proc.,
Geotechnical Practice for Waste Disposal '87. ASCE, pp 21-39,
1987.
7. Daniel, D.E. and Traut*ein. S.J.. "Field Permeability Test for
Earthern Linen," Proc., Use of the In Situ Test in Geotechnical
Engineering. ASCE, New York. NY, pp 146-160. 1986.
8. Day. S.R. and Daniel. D.E., "Hydraulic Conductivity of Two Pro-
totype Clay Linen," J. Geotech. Eng., ASCE. ///(8), pp 957-970,
1985.
9. Elsbury, B.R.. et al. "Field and Laboratory Testing of a Compacted
Soil Liner," Unpublished document prepared for the U.S. EPA,
1988.
10. Gordon, M.E.. Huebner, P.M. and Kmet. P. "An Evaluation of the
Performance of Four Clay-Lined Landfills in Wisconsin," Proc.,
Seventh Annual Madison Waste Conference, University of Wiscon-
sin, Madison. WI. pp 399-460, 1984.
11. Herrmann, J.G. and Elsbury, B.R. "Influential Factors in Soil
Liner Construction for Waste Disposal Facilities," Proc., Geotech-
nical Practice for Waste Disposal '87. ASCE. pp 522-536.1987.
12. Lahti. L.R., King. K.S., Readea, D.W. and Bacopoulos, A., "Qual-
ity Assurance Monitoring of a Large Clay Liner," Proc., Geotechni-
cal Practice for Waste Disposal '87. ASCE, pp 640-654.1987.
518
BARRIERS
-------
Compatibility of Soil-Bentonite Slurry Wall Backfill Mixtures
With Contaminated Groundwater
Mark E. Zappi
Richard Shafer
Donald D. Adrian
U.S. Army Engineer Waterways Experiment Station
Vicksburg, Mississippi
ABSTRACT
The interactions of solutes found in leachates from uncontrolled land-
fills with the components of a soil-bentonite (SB) slurry wall are capa-
ble of causing swelling or shrinking of the SB backfill material which
alters the hydraulic conductivity of the slurry wall. The effect of solutes
in contaminated ground water from the Ninth Avenue Superfund Site
in Gary, Indiana, on the hydraulic conductivity of two SB slurry wall
backfill mixtures was evaluated using rigid-wall permeameters.
Groundwater samples taken from three observation wells from the
Ninth Avenue site contained solutes that could cause increases in the
hydraulic conductivity of a SB slurry wall. One groundwater sample
contained salt concentrations as high as 20,000 mg/L. A second sample
contained total priority pollutant volatile organic compound (VOC) con-
centrations as high as 2,300 mg/L. A third sample contained
approximately 50 mg/L of total base neutral-acid extractables (BNAs)
included on the Priority Pollutant List.
Free swell tests using organic solvents, salt and tapwater were
conducted on four commercially available bentonites. Sodium chloride
showed the most impact on the free swell capacity of the bentonite
samples by always reducing the free swell capacity of the samples as
compared to the control tap water samples. The organic solvents
produced variable results with the bentonite samples, sometimes
increasing their free swell capacity over the controls and sometimes
decreasing it. From the free swell testing, one bentonite was chosen
for use in preparing the SB slurry wall backfill mixtures.
Six clay borrow sources from the Gary, Indiana, area were screened
using Atterberg limits and grain size analysis. A high plasticity soil
(CH) and medium plasticity soil (CL) were chosen as borrow materials
used in the preparation of the two SB slurry wall backfill mixtures.
The backfill mixtures were prepared by adding enough 6.04 bentonite
slurry to the two clay borrow samples to achieve at least a 4.0-in. slump.
The water contents of the backfill materials were 49.5 and 41.1% for
the CH and CL clay backfill, respectively.
Each backfill mixture was loaded into eight rigid wall permeameters.
Sidewall leakage inside the permeameters was controlled by the appli-
cation of bentonite paste along the inside of the permeameter cell walls.
All 16 permeameters were run in Phase I with tap water; then in Phase
n, six permeameters for each backfill mixture were permeated in dupli-
cate with the three contaminated groundwater samples (i.e., two
permeameter per groundwater sample), while the remaining two
permeameters, or control cells, continued to be permeated with tap
water. The three permeants from the wells produced varied hydraulic
conductivity results. However, the solutes had little or no effect on the
hydraulic conductivities of the backfill mixtures.
INTRODUCTION
The Ninth Avenue Dump Site is listed on the NPL of hazardous waste
sites scheduled for cleanup under the Superfund Acts of 1980 and 1986
(CERCLA and SARA). The site is a 17-ac inactive chemical waste dis-
posal area located in Gary, Indiana.
The site is situated in an industrial area, although properties adja-
cent to the site are relatively undeveloped. The site topography is a
relatively flat area with poor drainage and is characterized by small
depressions and mounds from past disposal and/or cleanup activities.
Both solid and liquid wastes are reported to have been disposed on
the site. Solid wastes deposited there include industrial construction
and demolition wastes. Liquid wastes deposited there include oils, paint
solvents and sludges, resins, acids and other chemical wastes. Waste
disposal operations took place between 1973 and 1980.
The groundwater is contaminated with a variety of inorganic and
organic contaminants. Inorganic contamination is mainly in the form
of sodium chloride (road salt). Organic contaminants are present in
significant concentrations in the groundwater, with ketones, benzene,
ethylbenzene, toluene, xylene (BETX), polyaromatic hydrocarbons
(PAHs) and total chlorinated ethenes being detected in significant con-
centrations.
In order to eliminate the continual spread of contaminants through
groundwater transport and to facilitate site cleanup, a SB slurry wall
was proposed as a means of contaminant containment. Proposed place-
ment of the wall will key into the aquiclude and completely surround
the site.
SB Slurry Wall Construction
SB slurry walls typically are installed by first digging a 2- to 4-ft
wide trench, using either a drag-line or a back-hoe, around the area
containing the contaminated material(s) and aquifer. During the exca-
vation of the trench, bentonite slurry is pumped into the excavated area
in order to support the sides of the trench. Typically, the trench depth
reaches at least 2 to 3 ft into an aquiclude. As the excavation equip-
ment moves along excavating the trench, borrow material is mixed with
bentonite slurry to form a bentonite slurry/borrow material mixture
referred to as the SB backfill mixture. The SB backfill mixture is added
to the trench once the excavation equipment has moved far enough along
so that the addition of the backfill mixture does not interfere with
excavation activities. The final product is a wall of backfill material
that has a very low hydraulic conductivity. Typical SB slurry wall
construction methods are shown in Figure 1.
In most cases, the borrow material used during slurry wall construc-
tion is simply the soil that is excavated from the trench. However, the
soil to be excavated from the Ninth Avenue Site has relatively high
percentages of sand and gravel thus requiring trucking in an alterna-
tive borrow material with more suitable characteristics for use in slurry
wall construction.
BARRIERS 519
-------
Table I
Groundwater Sample* BtrmeanU Analytical Data
ObMrvation
Figure 1
Excavation and Backfilling Operations
Potential Compatibility Problems
Many of the contaminants found in the site ground-water samples
have been identified through past research efforts as potentially having
adverse chemical interactions with clays, resulting in an increase in
hydraulic conductivity14. Although these contaminants are present in
the site groundwater, their concentrations are not nearly as high as those
tested in the above mentioned research efforts. In fact, most of the
research efforts to date involving chemical interaction between con-
taminants and clay panicles have been performed using either pure or
highly concentrated solutions. The contaminant concentrations in the
site groundwater are high in terms of an environmental pollution
problem, but may not be significantly high in terms of possible chemi-
cal interaction between the contaminants in the groundwaier and the
SB backfill mixtures. Because little or no research in the area of chemical
interactions of moderately contaminated solutions with clay particles
has been documented, compatibility testing must be performed to assess
if the contaminants in the groundwater will adversely change the
hydraulic conductivity of the SB slurry wall.
Study Objective and Scope
The objective of this study was to determine through laboratory testing
if the contaminants in the site groundwater will increase the hydraulic
conductivity of a soil-bentonite slurry wall.
The scope of this study included permeameter testing of two SB back-
fill mixtures that were formulated as part of this study. The SB backfill
mixture formulations were based solely on technical considerations and
not on an analysis of the projected costs associated with the construc-
tion of a SB slurry wall at the site. Compatibility of the proposed SB
slurry wall with the contaminated site groundwater was determined
through permeability testing of two SB backfill mixtures with lest per-
meants consisting of tap water from the City of Gary, Indiana (CGI)
and contaminated ground-water from three site observation wells (X-l,
X-14 and X-25). The concentrations of the major chemical constituents
detected in water samples from each of the three wells arc presented
in Table 1.
Bentonite Selection
Bentonites used in the construction of SB slurry walls to contain con-
taminated groundwater should exhibit a hydration volume that will not
be significantly reduced when exposed to the contaminants present in
the site groundwater. Any significant decrease in hydration or free swell
volume could increase the hydraulic conductivity of the SB slurry wall.
Therefore, a bentonite that exhibits a significant decrease in free swell
Total KttoiMi
To til PlunoU
BtTX Compound***
Mchylwn Chlorld*
•odlin
Iron
CalcltiB
Koqnasiua
Na^aiMB*
Potaulu
Oil or Id*
MrtlMU
Alkalinity
1*
Total Organic Carbon
HD
MO
HO
MO
11000.0
K.f
ttl.O
117.0
l.S
13.7
32700.0
2410.0
42* .0
1.S
«.4
50). j
17. t
11*. 1
15.0
47. «
121. a
1040. o
110.0
• .•
11.1
14JO.O
17*0.0
401.0
9. a
17*4.0
1492.0
l.l
1.7
20.0
«»«.o
S74.B
• X.O
374.0
17. «
14.0
1200.0
1440.0
97$. 0
(.1
2*70.0
• All eonemtratlona In •?/!
• • Baniana. ttlqrl Mjuana, Tolu»n«. and Xylana
MO Hot Datactad
volume when exposed to the site contaminants should not be consi-
dered suitable for use in the construction of a SB slurry wall.
Bentonite samples from four commercial sources were evaluated for
use in the bentonite slurry used in the preparation of the two SB back-
fill mixtures. Samples from each source were labelled as Bentonite Sam-
ples B-l, B-2, B-3 and B-4. Evaluations were based on results from
free swell testing of the bentonites. The free swell test involves the
addition of 2 g of bentonite to WOmL of a test solution containing levels
of contaminant(s) at or above the levels found in the site groundwater.
Free swell tests usually are performed in 100 mL graduated cylinders.
The volume occupied by the bentonite or free swell (hydrated volume)
is measured at 2 and 24 hr. For this study, tap water and laboralory-
prepared solutions of tap water mixed with various solvents and sodium
chloride at concentrations greater than those found in the site ground-
water were prepared and used as hydration fluids.
The results of the free swell testing of the four bentonite sources are
presented in Table 2. The table lists the test solutions and the respec-
tive free swell volumes at 2 and 24 hr. The free swell tests using tap
water as test solutions were used as a test control. Table 3 presents the
percent of control for each solution and bentonite. The percent of control
is a comparative value that is calculated by dividing the free swell volume
of the bentonite for each test solution by the free swell volume of the
bentonite sample for tap water then multiplying by 100.
Bentonites Frtc Swell Data
contaminant
t Ła*MMnt rm 1. 1 on 1
Acatona (1000 BO/l)
Acatona (JOOO m/l)
Acatona (AOOO »9/l)
«K (1000 Kg/l)
NACL (4000 »9/l)
HACL (10000 «0,/1)
Tolauna (looo •?/!)
Tap Nttar**
(Unoontaa 1 na tad )
TlM
t houral
2
!«
3
24
1
24
2
24
2
24
24
24
2
24
2
14
S»pla
R-l
27.)
10
27
14.1
27.1
12
11.)
17.9
III
2).)
11. «
19.)
K.I
rraa Svall voliwa (ml}*
Supl* Caaiila Saapla
4-J B-S B-4
14.
19.
2*
11.
10.
11.
11.
11.
11.
1*.
11.
11.
It.
11.
"ill
10.
11.
11.
11.
27.
11.
11.
11.
17.
10.
11.
14.
11.
10.
11.9 19.
14. (1 1*.
1(.
14.
11.
27. • 1*.
29.7 10.
11.
11.
11
14.
14.
• Av*»9* of thr« r«pllc«t«i
•• T««t Control
520 BARRIERS
-------
Table 3
Summary of Percentage of Controls for Bentonite Sources
Contamipant
Acetone (1000 mg/1)
(1000 mg/1)
Acetone
(3000 mg/1)
Acetone
(6000 mg/1)
HEX
(3000 mg/1)
Nacl
(4000 mg/1)
Had
(10000 mg/1)
Toluene
(200 mg/1)
sanDle B-l
114
112
117
109
92
67
99
Samola B-2
121
105
107
112
100
77
108
Samole B-3
121
106 '
106
94
82
56
107
Sample B-4
95
87
99
87
77
53
99
Percent of control values were used to determine the degree of inter-
action, if any, between the various bentonites and test solutions. There-
fore, if a bentonite sample has a percent of control value (POCV) less
than 100%, then adverse interactions between the contaminants in the
test solutions and the bentonite particles are occurring. It is possible
to have POCVs greater than 100%. Some contaminants in solution at
lower concentrations will actually increase the swell capacity of some
bentonites. This phenomenon was observed by Hettiaratchi and
Hrudley5. They concluded that acetone solutions at concentrations of
less than 25 mole percent of acetone (approximately 52% acetone so-
lution) increased the free swell capacity of the bentonite-soil mixture
tested.
All three acetone concentrations (1,000 mg/L, 3,000 mg/L, and 6,000
mg/L) increased the POCVs for all the bentonites tested except for ben-
tonite sample B-4. MEK increased the POCVs for both bentonite sam-
ples B-l and B-2. The POCVs for B-3 and B-4 for the MEK tests were
94 and 87 %, respectively. All of the free swell testing using sodium
chloride (salt) as test solutions resulted in POCVs less than 100%, ex-
cept sample B-2 4000 mg/L NaCl test which had a POCV of 100%.
Sample B-2 performed the best with respect to the sodium chloride
free swell tests followed closely by the performance of the B-l sample.
Toluene did not have a detrimental effect on any of the bentonite sam-
ples tested. Bentonite samples B-2 and B-3 had toluene POCVs greater
than 100%, while bentonite samples B-l and B-4 had POCVs of 99%
for the toluene free swell tests.
Bentonite yield is a rough measure of the solids content based on
the viscosity and swell capacity of the bentonite. A comparison of the
yields of bentonite samples B-l and B-2 indicated that bentonite B-2
is an extremely high yield bentonite that was developed for use as a
liner material for industrial waste lagoons. Bentonite B-l, on the other
hand, is an average yield bentonite that is more suitable for use in for-
mulating bentonite slurries for SB slurry wall construction. Therefore,
based on the results of free swell testing and a comparison of the respec-
tive yields of samples B-l and B-2, bentonite sample B-l was chosen
as the bentonite source for use in formulating the SB backfill mixtures.
SELECTION OF BORROW MATERIALS
Six sources of borrow materials located within the CGI vicinity were
100
90
80
t-70
to
Seo
>-
m
Ł50
W
U.
5<°
Ł
Sao
20
10
0
sc
U.S. STAMJARD SIEVE OPENING IN INCHES U.S. STANDARD SIEVE NUMBERS HYDROMETER
6 43 2 la 1 i | | 3 4 6 8 \$ 16 §0 30 40 50 70 100 140 200
i
r
1
1
1 1 1
1
1
1
1
r
t
T
~r
\
t>
i r
\,
^6--
0 100 50 10 5 1 0.5
GRAIN SIZE IN MILLIMETERS
COBBLES
"• 50
GRAVEL
COARSE | FINE
SAND
COARSE 1 MEDIUM I FINE
PL 17 |PI 33 |GS 2.73 |N*™'"15.6
CLASSIFICATION
SANDY CLAY (CH) BROHN
GRADATION CURVE
1
3
\
>i
Is-
— ^
V
\,
\
se
Vs
\
\
N,
0.1 0.05 0.01 0.005 0.
SILT on CLAY
0
10
20
30 Ł
(S
t-4
LU
40 *
>-
m
a:
en UJ
bO (,,
cr
-------
evaluated based on Atterberg limits, classification under the Unified
Soil Classification System (USCS) and percent fines (determined
through both sieve and hydrometric gradation analysis). The six sources
were identified as samples BM-1, BM-2, BM-3, BM A BM-5 and BM-6.
According to D'Appolonia2, SB slurry wall hydraulic conductivity is
a function of the percent fines (percent of material that passes through
a No. 200 sieve) of the borrow material used in the formulation of the
SB backfill material. The greater the percentage of fines, the lower the
hydraulic conductivity of the SB backfill mixture.
Based on the USCS, sample BM-1 was a CH soil, samples BM-2,
BM-3 and BM-4 were CL soils, and samples BM-5 and BM-6 were
SC soils. The CH soil (sample BM-1) and CL soil (BM-2) were selected
for use in formulating the two SB backfill mixtures evaluated in perme-
ability testing. Soil sample BM-2 was chosen from the CL group of
sources because it had the second highest percentage of fines of the
CL class of soils. Therefore, source BM-2 was considered representa-
tive of the CL group of borrow material sources. Gradation curves for
the selected borrow materials are presented in Figures 2 and 3. Physi-
cal and chemical properties of the two selected borrow sources arc
presented in Table 4.
A 6.0% bentonite and COI tap water slurry with a Marsh Funnel
reading of 48 sec. was mixed with the two borrow materials to formu-
late two SB backfill mixtures. One SB backfill mixture was prepared
using the BM-1 material, while the second SB mixture was prepared
with the BM-2 material. The porosities (n) of the two SB backfill
mixtures were determined in order to calculate the pore volume of the
backfill mixture samples loaded into each permeameter. The slumps
of the BM-1 and BM-2 SB backfill mixtures were 4.0 and 4.5 in., respec-
tively. The water contents of the BM-1 and BM-2 SB backfill mixtures
TaW«4
Physical and Chemical Characterization of Ctay Samples
Formulation of SB Backfill Mixture*
Parameter
PH
CEC (meq/kg)*
Ca(mg/l)
Mg(mo/l)
K(mg/l)
Na(mg/l)
TOC(mo/f)
Liquid Umtt(%)
Plastic Umrt
Plasticity Index (%)
Water Content (%)
Specific Gravity
day Type-
Clay Samples
BM-1
5.37
2260
15200
9970
3690
246
4307
50
17
33
15.6
2.73
CH
BM-2
7.71
1960
1270
5470
4120
149
1081
39
17
22
8.0
2.73
CL
• Method 9061 USEPA SW-846 (Sodum Method)
•• Unified Sol Classification System
were 49.5 and 41.1%. respectively. The percentages of bentonite in the
SB backfill mixtures BM-1 and BM-2 were 2.30 and 2.33%. respec-
tively.
too
90
SO
^70
Ł
l»l
Ł60
>•
-«-.
• •<
1
^
^
1
^
1 '
^^e.
>0 tOO SO 10 3 109
GRAIN SIZE IN MILLIMETERS
COBBLES
a 39
GRAVEL
COAflSf I fltf
SAND
COAASC MID II* 1 FINI
PL l? |PI ga |GS 2 ?3 |WT*«8>0
CLASSIFICATION
SANDY CLAY (CL) GRAY
GRADATION CURVE
1
q
S
\
Ss
**"*"!
N.
\
\
\
S
^
V
^
^
Ss
\
\
0 1 0.05 0 01 0 005 0
SILT OP CLAY
0
10
20
30 J
(9
40 I
o
50 1
5
«;
tu
70 Ł
a
80
90
100
101
PROJECT 9TH. AVENUE
BORING NO. SAMPLE NO.
DEPTH/ELEV DATE 09 AUG 88
Figure 3
Sample No. BM-2
522 BARRIERS
-------
Permeameter Testing
Eight of the 16 permeameter cells were loaded with one of the two
SB backfill mixtures while the remaining eight cells were loaded with
the other SB backfill mixture. Initially, all test cells were permeated
with CGI tap water in order to determine the hydraulic conductivity
(K) of the SB backfill mixture sample in each cell. Although eight repli-
cate cells contained the same SB mixture, slight differences in packing
each of the cells could produce differences in the observed hydraulic
conductivity for each cell. For this reason, CGI tap water was permeated
in all cells so that a baseline hydraulic conductivity could be deter-
mined for each cell.
After at least one pore volume of tap water was permeated through
each cell, six of the eight cells for each SB backfill mixture were per-
meated with contaminated groundwater collected from the three site
observation wells. Samples from each of the three site observation wells
were permeated through two replicate cells for each SB backfill mixture.
Two of the eight cells for each SB backfill mixture continued to be
permeated with CGI tap water throughout the course of permeability
testing. These four cells (two cells for each SB backfill mixture) served
as test control cells. The control cells were used to help determine if
any changes in hydraulic conductivity were due to physical changes in
the SB backfill mixture samples caused by operational adjustments made
during testing (i.e., increased hydraulic gradient) and not due to chemical
interaction between the backfill mixtures and ground-water con-
taminants.
The test cells were downflow rigid wall permeameters constructed
as illustrated in Figure 4. SB backfill samples used for permeability
testing were 2.25 in. long with 4-in. diameters. The inside walls of
the permeameters were roughed up with a stainless steel brush and
coated with a 0.06-in. layer of bentonite paste to reduce sidewall leakage.
Porous stones, saturated with CGI tap water, were used to support the
samples inside the permeameters. The permeameters were setup as il-
lustrated in Figure 5. The average hydraulic gradients (i) used in
permeability-testing of the SB backfill mixtures BM-1 and BM-2 were
45 and 27 ft/ft, respectively.
Pressurized N
FLANGE BOLTS
•1/4" COPPER TUBING
1/4" THICK.
PLEXIGLAS
NEOPRENE
GASKET
4" ID
- PERMEAMETER
COLUMN
t
1/2" PLEXIGLAS
FLANGE PLATES
-8" FLANGE
-1/4" COPPER TUBING
Figure 4
Rigid Wall Permeameter
Permeameter
Permeant
Collection
Vessel
1/4" Copper 1/4" Ball Valve
Tubing
Figure 5
Rigid Wall Permeameter System
Bottled nitrogen was used as the pressure source for the permeant
reservoirs. One pressure reservoir was used to pressurize and deliver
the permeants to two separate permeameters. This arrangement served
as duplicate permeameter sets for each SB backfill mixture sample and
respective permeant. Copper tubing (0.25 in. OD) was used to connect
the reservoirs to the permeameters. Permeant volumes were collected
and measured daily in 100-mL graduated cylinders.
Table 5 summarizes the results of the permeability testing of the two
SB backfill mixtures. The period of permeability testing in which CGI
tap water was used as the permeant in all cells was identified as Phase I.
Phase II of permeability testing was the period of permeability testing
when contaminated groundwater samples were used as permeants in
some of the cells. Table 5 also lists the K ratio for each cell which is
simply the ratio of the average Phase II K over the average Phase I K.
A K ratio of unity indicates a test cell with no K deviation over the
course of permeability testing.
A K ratio of less than unity indicates a cell with decreasing K's. This
situation can occur when the sample within the cell slowly consoli-
dates over time. An example of this is best illustrated in Figure 6 which
is a plot of the number of pore volumes of permeant that flowed through
cell No. 5 versus the respective K. The K values in Figure 6 gradually
decrease over the course of permeability testing as the sample consoli-
dates over time thus closing off pore channels. Another example of less
than unity K ratios observed is illustrated in Figure 7 which presents
the K data for one of the two BM-2 control cells. In this case, during
the initial stages of Phase I permeability testing, the cell was very
Pore Volumes
Figure 6
Cell Number 5 Hydraulic Conductivity Data
BARRIERS 523
-------
ItableS
Summary of Preliminary Bermeameter Result*
Numbtr of
Pora Voluaaa
Number Panuant
KN-I BacUlll
I
2
}
*
5
6
7
e
BM-2 Backfill
9
10
11
12
1]
It
15
16
Tapvatar
Tapwatar
Tapwattr
Tapwatar
Tapwatar
Tapwatar
Tapwatar
Tapwatar
Tapwacar
Tapwatar
Tapwatar
Tapvatar
Tapwatar
Tapvatar
Tapwacar
Tapvatar
Phaae I
1.}}
1.73
1.56
1.27
1.27
3.21
2.72
1.87
10.81
5.9*
4.94
4.43
5.02
4.88
S.07
5.66
Avaraga K
Nuvbar o(
During fhaaa I It Fora VoluMt
Phaaa 1 Standard Dav. Phaaa 110 Par^atad Durlna
(c»/a«c)
3.07E-OS
3.90E-08
3.94E-08
3.09E-08
3.13E-08
B.67E-08
6.34E-08
4.67E-08
5.478-07
2.391-07
2.13E-07
I.90E-07
2.6IE-07
2.I5E-07
1.87E-07
2.03E-07
(cm/aac)
9.73E-09
7.92E-09
9.62E-09
1.021-09
1.06E-08
8.98E-08
3.3*E-08
3.29E-08
.26E-07
.56S-07
.52E-08
.8IE-08
.02E-07
.23E-07
.02E-07
7.70E-08
Famaant
Tapwatar
Tapwacar
X-l
X-l
X-l*
X-l*
1-25
X-25
Tapvatar
Tapvacar
X-l
X-l
X-l*
X-l*
X-2J
X-2>
Phaaa II
2.60
3.48
3.71
2.16
1.68
6.48
2.73
.33
.2)
1 .27
.09
.23
.64
.09
.73
Avarafa X
Ratio
During Phaaa II K Fhaaa 11
Fhaaa 11 Standard Dav. to Ph««« T
(emlue) (ca/aae)
2.48E-08
3.02E-08
4.08E-08
2.451-08
1.81E-08
1.S4E-07
5.12E-08
2.22E-08
1.27E-07
2.55E-07
I.J5E-07
1.58E-07
2.26E-07
1 . 38E-07
.93E-09
.32E-09
.24E-08
.80E-09
.MB-09
~
.15E-07
.UE-Ot
.16E-D8
.64E-08
.64E-08
.23C-08
.17E-07
.78E-07
.08E-08
1.49E-07 3.45E-08
11
0.10
0.77
1.03
0.79
O.S8
—
2.90
1. 10
0.04
0.53
1.20
0.82
0.61
1.05
0.74
0.73
• Phast I la whan all parvaanta wara tapwatar
t Phaat II la whan contaalnacad pan»anta wara run In non-control taat calla
I Calculated ualna avaraga K'a
Calculated ualna avaraga
fcr*v.i*i.'tcr fjllurv
dynamic with inconsistent Ks. After a period of variable Ks. the cell
stabilized with very consistent Ks (hence the extremely low K ratios
for both BM-2 control cells). Most of the cells experienced variable
Ks at the initiation of permeability testing (the early stages of Phase
I testing).
u
i = ,.
ts«J
U
I
a 10 ii
Pora Volumvt
Higurc 7
Cell Number 9 Hydraulic Conductivity Dam
Greater than unity K ratios were also observed. Greater than unity
K ratios observed for this study are indicative of those permeameters
that were experiencing increased variability in K over time. Greater
than unity K ratios can be an indication of adverse chemical interaction
between the contaminants in the groundwatcr and the SB backfill
materials. None of the greater than unity K ratios observed during this
study were believed to be caused by chemical interaction. Permeametcr
operations, such as permeant change out and reservoir refilling, were
believed to be the cause of the greater than unity K ratios.
None of the BM-2 SB backfill permeameters had K ratios signifi-
cantly higher than unity. Cell No. II, with a K ratio of 12, had the
highest K ratio of all the BM-2 SB backfill permeameters. After
approximately seven pore volumes of contaminated groundwater from
well X-l had permeated through Cell II, an increase in K was observed
as shown in Figure 8. This increase in K was only temporary, because
the Ks began to gradually return to within the range of Ks previous
to the elevated values. The cause of the increase is believed to be due
to the sample shifting during the removal of the pressure head while
refilling the permeant reservoir.
Figure 8
Cell Number II Hydraulic Conductivity Data
Pora Volumaa
Figure 9
Cell Number 7 Hydraulic Conductivity Data
524 BARRIERS
-------
The BM-1 SB backfill mixture cells had two cells with K ratios sig-
nificantly higher than unity. Cell No. 6 data were not used to evaluate
groundwater compatibility with the SB backfill mixture because sig-
nificant sidewall leakage was observed at the end of Phase I testing.
Cell No. 7, with a K ratio of 2.9, experienced variable Ks at the end
of Phase I, as shown in Figure 9. The variable Ks continued through-
out all of Phase II, however, no trend toward an increase in K was
observed. The high Phase n standard deviation listed in Table 5 for
cell No. 7 is indicative of the variability of cell No. 7 K data.
CONCLUSIONS
Acetone increased the free swell volume of all the bentonites tested
except for bentonite sample B-4. Bentonite samples B-l and B-2 had
increased free swell volumes when exposed to MEK during the free
swell tests, while samples B-3 and B-4 both exhibited reduced free
volumes when exposed to MEK. Sodium chloride reduced the free swell
volumes of all the bentonite samples. Toluene did not have a detrimental
impact on the free swell volumes of any bentonite samples.
SB backfill mixtures BM-1 and BM-2 were compatible with ground-
water samples from site well X-l, X-14 and X-25. Permeameter testing
demonstrated that hydraulic conductivities for the SB slurry mixtures
selected for this study were not affected by contaminants in ground-
water from the Ninth Avenue site. Elimination, or at least reductions,
in the on/off cycling of the pressure head that occurs during refilling
the permeant reservoir and permeant change out should reduce the
amount of K variability observed during permeability testing. The overall
average hydraulic conductivities for SB backfill mixtures BM-1 and
BM-2 throughout both phases of permeability testing were 4.7 x 10(-8)
cm/sec and 2.1 x 10(-7) cm/sec, respectively.
ACKNOWLEDGEMENTS
This work was funded by the U.S. Army Corps of Engineers, Omaha
District, in conjunction with the U.S. EPA, Region V. The authors wish
to thank Mr. Steve Rowe, USAGE, Omaha District, and Ms. Allison
Hiltner, U.S. EPA, Region V, for their assistance and support for this
study. The authors would also like to thank Warzyn Engineering, Inc.,
of Madison, Wisconsin, for their assistance during the development
of this study. Permission was granted by the Chief of Engineers to publish
this information.
REFERENCES
1. Anderson, D.C. and Jones, S.G., "Clay Barrier-Leachate Interaction," National
Conference on Management of Uncontrolled Hazardous Waste Sites, HMCRI,
Silver Spring, MD, p 154, 1983.
2. D'Appolonia, D.J., "Soil-Bentonite Slurry Trench Cutoffs," J. of the Geo-
technical Engineering Division, 106, No. GT4, 1980.
3. U.S. EPA, Slurry Trench Construction for Pollution Control, EPA
540/2-84-001, 1984.
4. Evans, J.C., Fang, H.Y. and Kugelman, I.J., "Organic Fluid Effects on the
Permeability of Soil-Bentonite Slurry Walls," Proc. National Conference on
Hazardous Wbstes and Environmental Emergencies, Cincinnati, OH, HMCRI,
Silver Spring, MD, p 267, 1985.
5. Hettiarachi, J.P. and Hrudley, S.E., "Influence of Contaminant Organic-
Mixtures on Shrinkage of Impermeable Clay Soils with Regard to Hazardous
Waste Landfill Liners," Haz. Wastes/Haz. Materials, 4,(4), pp 377-388, 1987.
BARRIERS 525
-------
Solute Migration Control in Soil-Bentonite Containment Barriers
Henry V. Mott, Ph.D., RE.
South Dakota School of Mines and Technology
Walter J. Weber, Jr., Ph.D., P.E.
The University of Michigan
Ann Arbor, Michigan
ABSTRACT
In many cases containment by a soil-bentonilc barrier is the most
attractive strategy for mitigation of subsurface contamination. Because
the hydraulic conductivity of these barriers is generally less than K) '
cm/sec, molecular diffusion can be the dominant transport process.
"Effective" diffusion coefficients for low-molecular-weighl organic
solutes in soil-bentonite barriers are reduced only several factors from
those in free aqueous solution. Molecular diffusion can then result in
solute breakthrough within a relatively short time as well as in significant
solute transport through the barrier when the diffusion process obtains
a near-steady-state condition.
Of the several potential means for improving the performance of
soil-bentonite barriers, this work examined the addition of class "F"
fly ash, which contains a significant fraction of unburned carbon, to
soil-bentonite mixtures for enhancement of sorpiion capacity. The
sorption capacity of the unburned carbon fraction of the fly ashes tested
was found to be roughly equivalent to that of natural soils and sediments.
Simulations of the performance of a typical barrier indicated that the
addition of a sorptive phase such as fly ash can significantly retard solute
breakthrough.
INTRODUCTION
Several authors" have recently claimed that the migration of water
soluble contaminants through cut-off barriers can be effectively curtailed
by restricting the hydraulic conductivity of such barriers to less than
10' cm/sec. Given that the magnitude of natural hydraulic gradients
seldom exceeds unity, under these conditions the bulk flow of water
will not exceed 3.0 cm/yr. The aqueous solution residing within the
pores of a soil-bentonite barrier is, then, effectively a stagnant fluid
and convection (advection) is insignificant. Under these conditions.
molecular diffusion is the dominant transport process. Two studies**
of dissolved solute transport into natural clay barriers of low hydraulic
conductivity concluded that molecular diffusion was the process
responsible for the net migration of solutes into the barriers. Another
theoretical study7 suggested that solutes could migrate through
soil-bentonite barriers by molecular diffusion even against on inward
directed hydraulic gradient of magnitude equal to O.SO.
Diffusive flux (Fu°) and convective flux (Ju°) may be represented
in mathematical form as:
An empirical relationship describing D in terms of the free aqueous
diffusion coefficient (D,) and the porosity («) of a porous medium or
the hindrance factor (H) is given as **:
D = e^D/H
(2)
Equation 2 has been verified for diffusion in packed beds of lead
shot, sand and calcite": and diffusion in unsaturated soils"0.
Additionally, diffusion coefficients for chloride in compacted
sand-bentonite mixtures" and in natural clay6 have been found to
conform with Equation 2".
Sorption within a porous medium can delay or retard transport through
reduction of aqueous phase concentration levels and, hence, the
macroscopic concentration gradient (dC/dz). The aqueous
concentration of solutes within soU-bentomte containments can approach
the solubility limit of the solutes in question: thus, the most applicable
relationship for describing sorption equilibrium is the non-linear
Freundlich sorption model:
(3)
where p^ is the adsorbed concentration at equilibrium, K,, is the
Freundlich capacity parameter. C, is the aqueous concentration at
equilibrium and n is an exponential fining parameter. In certain cases,
it is assumed that the conditions within a given medium very closely
approach equilibrium and that equation 3 is applicable throughout (the
local equilibrium assumption. LEA). Most systems, however, exhibit
rale-limited sorption rendering the LEA quite inapplicable. The
combination of non-linearity and non-equilibrium introduces additional
degrees of complexity into the mathematical structure of models
describing the transport process.
Diffusion, convection and sorption are combined with a pseudo
first-order transformation term (k^) through application of the mass
conservation law in one dimension at the differential scale to yield the
mathematical model describing solute transport within a porous
medium:
ac,
- 0,
ac
3z
at
(4)
ta° = D:i (dC/dz) and J,,0 = v(C
(D
where C is the aqueous concentration, v( is the superficial fluid
velocity, z is the coordinate in the gradient direction and DM is the
effective diffusion coefficient for solute i within the porous medium.
where '% is the density of the sorbcnt phase and t is time. If the LEA
is applicable. Equation 4 may be more simply stated as:
ac.
ac
a2c
ac
dz
(5)
526 BARRIERS
-------
where R = e + *s (1 -«) nKFC" 1 for a non-linear sorption
relationship. Note that for a linear sorption relationship, n equals unity
and K,, is replaced by the partition coefficient (Kp. If the LEA is not
applicable, the sorption term typeset equation here may take one of
many formulations. The simplest of these, considered here, is:
(6)
where k,, is a mass transfer coefficient derived from boundary layer
theory, as is the aqueous/solid interfacial surface area per bulk unit
volume of porous medium and Cs is the aqueous solute concentration
at the aqueous/solid interface. The assumption inherent with this
formulation is that intraparticle transport resistance is negligible. The
coefficient kf may be approximated for the case of insignificant
convection as15-16'
k/i/D =2.0 (7)
where d is the characteristic dimension (usually the diameter) of a
spherical particle. Substitution of Equations 6 and 7 into Equation 4
and de-dimensionalization of the interfacial transfer term leads to the
definition of a Damkhler number (ND):
0.1 0.01
Grain Size In Millimeters
Figure 1
Gradation Curves for Experimental Background Soil Mixture
ND. = kfocs"2/De.
(8)
Table 1
Pertinent Fly Ash Properties
Alternative formulations for ND have been developed for convective
systems, and it was found that when the magnitude of ND was 100 or
greater, simulations of solute transport using Equation 6 agreed quite
well with those employing the LEA 17-18 It would be reasonable to
assume that the mathematics of diffusive systems would behave in a
similar manner with respect to the appropriateness of the LEA.
The objectives of this work were: (1) to measure diffusion coefficients
for low-molecular-weight solutes in soil-bentonite media; (2) to define
the applicable relationships among Ds, D, and definable properties of
soil-bentonite barriers; and (3) to investigate the use of high-carbon
fly ash as an additive to soil-bentonite mixtures in order to enhance
retardation capacity.
EXPERIMENTAL PROGRAM
Materials
Background soil for soil-bentonite mixtures consisted of a mixture
containing 77% silica sand, 10.5% silica flour and 12.5% kaolinite.
These materials were chosen for their properties of low organic carbon
content, high purity and inert mineral surfaces. Grain size distributions
for each constituent and for the composite background soil are shown
in Figure 1. Sodium bentonite (Slurry Ben 90, American Colloid Co.)
was added to the background soil in various quantities to obtain the
experimental soil-bentonite mixtures. Relationships between porosity
and confining stress were developed for each experimental mixture using
a slurry consolidometer and, subsequently, were employed in the
analysis of the diffusion data.
Samples of fly ash were obtained from the B.C. Cobb, Karn and
Trenton electrical power generating plants owned and operated by the
Consumers Power Co. of Michigan. The fly ashes were tested for loss
on ignition, carbon content and various other properties. The results
of this characterization are shown in Table 1.
Target solutes were chosen to represent the several classes of
low-molecular-weight priority pollutants. A listing of these solutes and
pertinent properties is given in Table 2. All chemicals used were of
reagent grade or better.
Diffusion Experiments
Both quasi-steady-state (QSS) and transient diffusion experiments
were conducted to determine the magnitude of effective diffusion
coefficients for target solutes in soil-bentonite mixtures. The device
Property
Density2 (g/cm )
Loss on ignition W
Percent carbon1
Iodine number
Phenol value
Tannin value
Specific surface (m /g)
Raw
Fired
Karn
2.25
6.47
4.69
525
nd
3086
1.14
0.78
Fly Ash
Trenton
2.29
9.14
6.14
562
nd
951
2.65
1.08
Cobb
2.24
10.23
6.52
595
39.0
3501
3.52
1.18
Measured as C02 recovered during wet combustion
2Bergstrom and Gray (19)
formalized to the carbon fraction of the fly ash
^Primary surface area by B.E.T. nitrogen adsorption
Table 2
Properties of Target Organic Solutes at 25 °C
Solute Henry's log i Aqueous
Constant KW Solubility
(atm) (mg/L)
CTET 1659* 2.64 7851
TCE 521; 2.29 1100
TTCE 965* 2.88 1261
1,4-DCB 1385 3.39 79
1,2,4-TCB 190^ 3.98 301
POP 0.055 2.39 27100*
lindane 0.000016 3.72 7.21
^Values given at 20° C.
Hayduk and Laudie Correlation (20)
3Based on LeBas Volume (20)
4Gosset (21)
Calculated from vapor pressure and
6Mackay and Leinonen (22)
n •!
Aqueous Molecular
Diffusivity Radius
(x!05cm2/sec) (A)
0.983
1.068
0.961
0.920
0.847
0.930
0.635
solubility data
3.55
3.49
3.70
3.79
3.98
3.67
4.68
employed in the QSS experiments is shown schematically in Figure 2.
Experiments were conducted by allowing dissolved solutes to diffuse
from the upper reservoir through the composite barrier into the lower
reservoir. Traces of lower reservoir concentration verses time formed
BARRIERS 527
-------
the basis for the calculation of the magnitude of effective diffusion
coefficients. Exact details of the experimental procedures and analyses
are given elsewhere1" The device employed in the transient
experiments and the arrangement of the apparatus are shown
schematically in Figures 3 and 4, respectively. Experiments were
conducted by allowing dissolved solutes to diffuse into the packed
columns from the small reservoir located directly below the boundary
between the aqueous solution and the column packing. The solute
regeneration and recirculation system provided for a known
concentration condition at the boundary. Diffusion coefficients were
evaluated by matching experimental concentration profiles with those
generated by simulation of the process using a time-linearized, implicit
finite difference numerical approximation of Equation 4. Additional
details of the experimental procedures and analyses arc given
elsewhere h
22 ma
-------
o
TWELVE TRANSIENT
DIFFUSION COLUMNS
Solute
Figure 4
Transient Diffusion Column Experimental Arrangement
liable 3
Results of Quasi-Steady-State Diffusion Experiments
m Hindrance Factor (ff)
e - 0.40 e - 0.43 e - 0.48
DCB
PCP
1.31
1.46
3.32
3.81
3.02
3.43
2.62
2.92
the diffusion process at the three experimental times. The effective
diffusion coefficients corresponding to the respective layers were
calculated using Equation 2 and used in simulations of the diffusion
process.
Lindane is known to undergo solvolysis in aqueous solution; thus,
a first-order transformation coefficient (kT) was determined
experimentally and employed in the simulations. The sorption capacity
of the background soil for lindane was determined and used in the
simulations.
The experimental concentration profiles and two selected simulations
are shown in Figures 5, 6 and 7. Note that only the portion within the
soil-bentonite is shown. Extrusion of the column contents disturbed
the confinement layers and, perhaps, the segments of the soil-bentonite
that were situated near the confmement/soil-bentonite interface. Total
solute penetration was used to ascertain the correctness of the diffusion
coefficients. Effective diffusion coefficients determined from Equation
2 appear to successfully describe the diffusion process which occurred
in these experiments.
Sorption Experiments
Karn, Trenton and Cobb fly ashes were tested in raw form for sorption
of CTET, TCE, TTCE, DCB, TCB, PCP and lindane. Additionally,
O)
<
fJC
UJ
O
o
o
800
600-^
•P
400 H
2004
TRANSPORT TIME = 194.4 hr
Conf Hfac - 3.50
Freundlich "n" = 0.854
Vs = 0.0 cm/hr
0123456789 10
DISTANCE FROM INTERFACE (cm)
Soil-Bentonite column at 194.4 hr
(1) S-BHfac=2.63;Kd=0.0/hr;Kf=3.84E-4
(2) S-BHfac=2.63;Kd=9.02E-4/hr;Kf=6.98E-4
Figure 5
Transient Column Data and Simulations for
Transport Parameter Evaluation
3000
O)
TRANSPORT TIME = 384.3 hr
ConfHfac = 3.50
Freundlich "n" = 0.854
Vs = 0.0 cm/hr
2000
UJ
o
z
o
o
<
o
012345678910
DISTANCE FROM INTERFACE (cm)
a Soil-bentonite column at 384.3 hr
• (1) S-BHfac=2.63;Kd=0.0/hr;Kf=3.58E-4
• (2) S-BHfac=2.63;Kd=9.02E-4/hr;Kf=6.98E-4
Figure 6
Transient Column Data and Simulations for Transport Parameter Evaluation
the fly ashes were tested for sorption of DCB after firing at 600 °C.
From these experiments it was concluded that the fraction lost upon
ignition was responsible for the sorption capacity of the fly ash.
Therefore, the K^. values are based on the carbon fraction of the fly
ashes rather than total mass. The resulting parameters based on least
squares fitting of sorption data to Equation 3 are listed in Table 4.
The K^ data of Table 4 were plotted against octanol/water partition
coefficient (Kw) in Figure 8 and against aqueous solubility (S) in
Figure 9. Because PCP sorption is not solvophobically driven, data for
this solute are not included in Figures 8 and 9. The resulting regression
equations and statistics are listed in Table 5. The correlations are highly
significant, suggesting that physical relationships exist. Of particular
BARRIERS 529
-------
Figure 7
Transient Column Data and Simulations for
Transport Parameter Evaluation
4000
Z 3000 •
O
z
111
o
o
o
<
o
2000
1000
TRANSPORT TIME = 774.6 hr
ConfHfac = 3.50
Freundllch "n" = 0.854
V« = 0.0 cm/hr
0123456769
DISTANCE FROM INTERFACE (cm)
• Soil-bentonlte column •( 774.6 hr
• (1) S-BHfac=2.63;Kd=0.0/hr;Kf=3.58E-4
• (2) S-BHfac=2.63;Kd=9.02E-4/hr;Kf=€.98E-4
Table 4
Summary of Fly Ash SorjXton Capacity (Kr)
and Intensity (n) Parameters
10
Salute Fly Aeh
95« C.I.
95» C.I.
CV
CTET
TCE
TTCE
DCS
TCB
rcr
Llndene
Kern
Trenton
Cobb
Kern
Trenton
Cobb
Kern
Trenton
Cobb
Kern
Trenton
Cobb
Kern
Trenton
Cobb
Kern
Trenton
Cobb
Kern
Trenton
Cobb
0.
0.
1.
0.
0.
1.
1.
2.
4
4.
4.
6.
8.
5.
14
2.
3.
8.
8.
7.
19
348 0
387 0
53 0
658 0
920 0
71
37 0
39
88
02
71
20
66
85
.0
95
10
09 (
05
28 I
.5 1
160-0.
156-0.
777-3.
400-1.
581-1.
1.04-2.
808-2.
.70-3.
.09-7
82-5.
.82-5.
.39-11
.67-11
.35-7.
.07-21
.01-4.
433.
13-10
.02-12
• .35-12
.55-44
756
961
00
08
46
82
33
36
69
72
80
.4
2
87
.6
33
96
.7
.9
.2
.3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
o
0
0
o
o
o
609
745
581
477
498
428
466
320
428
322
267
386
256
360
305
267
211
149
348
330
430
0
0
0
0
0
0
0
0
0
0
0
0
0
o
0
0
0
0
o
0
0
356 0
477-1
258-0
309-0
343-0
264-0
204-0
142-0
170-0
181-0
174-0
172-0
134-0
211-0
180-0
131-0
121-0
049-0
202-0
192-0
252-0
863
01
903
644
653
593
729
499
685
463
361
600
378
508
430
404
302
250
494
467
607
161
141
48
145
111
55
27
12
18
20
14
40
18
32
42
14
8
3
65
81
103
The veluei of lf correspond co unite of «fc/L end »j/«, for equeoui
end eolld pheee concencreclone. reepecclvely.
Coefficient of variation
interest is the fact that the carbon associated with the respective fly
ashes behaved differently with respect to its sorption capacity. The
surface area associated with the fraction of the fly ash lost on ignition
was determined from measurements of the surface area of both raw
and fired fly ashes. The values were 63, 18 and 24 nr/gr for Karn,
Trenton and Cobb ashes, respectively. The pre-exponential coefficients
shown in Table 5 are a measure of the base sorption capacity of the
fly ash. This coefficient was divided by the surface area lost on ignition
to obtain a measure of the specific surface sorptivity. Respective values
based on the IC-K()W relationships for Karn, Trenton and Cobb ashes
are 7.6x10 ,, 1.0x10 and 2.0x10,. The value of the specific sorptivity
increased dramatically from Karn to Cobb ash which suggests that the
surface of the Cobb ash has a greater density of sorption sites, due to
one or more of many potential factors which cannot presently be
identified.
Tables
RegreMkm Data for K.-K-, and
K,-S Relationships
Fly eih
Regretllon Equation
ec«lc Significance
Kern
Trenton
Cobb
Kern * Trenton
Cobb
A> - 0.0048*0,°""
*> - 0.0187V «7
Kf - 0. 0461*3,° 651
JCp-0.048SS-°-"1
*>-o.U20s-0*74
0.94
0.87
0.94
0.90
0.98
5.51
3. S3
5.51
4.13
9.80
It
5*
U
11
0.05%
100-s
KF
100
• KARN ASH
• TRENTON ASH
• COBB ASH
1000
Kow
KF = 0.0048 ' Kow*0.837 ; R = 0.94
KF = 0.0187 • Kow*0.667 ; R = 0.87
KF = 0.0468 • Kow*0.651 ; R = 0.94
10000
notes: 1) Parachtorophenol data is omitted from regressions
2) KF has units of (mo/OV{mg/L)*n
Figure 8
Fly Ash Freundlich K, vs. Octanol-Watcr Partition Coefficient
The sorption data discussed above were obtained for non-linear
systems. Unfortunately, most other data taken for systems involving
organic carbon associated with natural soils and sediments apply to the
very low concentration range where partitioning relationships are
functionally linear. To compare the sorptivity of fly ash carbon and
naturally occurring organic carbon, a partition coefficient (Kr) was
defined as:
Kt = K/V/C.
(9)
The value of C, chosen was 100 pg/L and the resulting estimated
values of Kc. were regressed on a log-log basis against both K_, and
aqueous solubility in millimoles per liter (X). The results oithese
regressions based on Kc are compared in Table 6 with representative
regressions from the literature that are based on K^.. Upon
comparison of the correlations presented in Table 6, the conclusion may
be drawn that the sorption capacity of the carbon associated with the
fly ashes examined by this investigation is at least as great as that of
naturally occurring organic carbon.
Dlflusion-Sorption Experiments
Three transient diffusion columns were packed with a soil-bentonite
mixture containing 0.31% Cobb fly ash. The experimental concentration
profiles were simulated using transport and sorption parameters derived
above. It was assumed for the simulations that the LEA applied within
530 BARRIERS
-------
100
KF
10-:
1 •:
PGP
OMITTED
0 .0001 .001 .01 .1
MOLAR SOLUBILITY, X
D KARN & TRENTON ASHES
KF = 0.0485 * XM1.521 ; RA2 = 0.81
• COBB ASH
KF = 0.162 * XMJ.474 ; R*2 = 0.96
Figure 9
Between
and Solute Solubility for Sorption by Fly Asl
The Relationship Between the Freundlich K
\sh
Tabled
Regression of KC on KQW
and Kr on x
Sorbent
Regression Equation
Significance
Karn and Trenton ash
Cobb ash
Natural Sediments (23)
Karn and Trenton ash
Cobb ash
Natural Sediments (24)
logXc
loglCc
lo§*0<
logJCc
logXc
1°&*0<
-logxw
- logXj^
; ~ l°lFa
- 5.65
- 5.81 -
- - 4.28
+ 1.17
+ 2.15
0.451og*
0.441ogy
0.561ogx
0.
0.
0.
0.
.84
.90
.87
.94
1%
1%
1%
1%
the columns. The concentration profiles from the experiment and
simulations for one experimental time are shown in Figure 10. It may
easily be noted that the simulation fails miserably in describing the
experimental behavior. Based on the effective diameter of the spherical
fly ash particles, ND was calculated for the system and found to be
0.04, well removed from the value of 100 suggested as appropriate for
application of the LEA. The value of ND for a soil-bentonite mixture
containing 40% fly ash was calculated to be approximately 4.0, which
also is well removed from the acceptable value of 100. The conclusion
may be drawn that successful modeling of solute transport in
soil-bentonite barriers that are modified by the addition of high-carbon
fly ash must employ non-equilibrium as well as non-linearity in the
sorption term.
Simulations of the Performance of
Soil-Bentonite Barriers
Two extreme conditions envelope the realm of potential conditions
within a soil-bentonite barrier: (1) no sorption capacity; and (2) sorption
that may be described by the LEA. Simulations were performed for
these two conditions and were based on a hypothetical barrier of
thickness equal to 3.3 ft, containing 10 ac and extending 50 ft below
an unconfined water table. The target solute considered was CTET and
the effective diffusion coefficient was calculated from Equation 2. The
barrier was considered planar in geometry and infinite in areal extent.
The interior concentration (C^ was considered constant at 1 mg/L and
TRANSPORT TIME = 774.6 hr
Conf Hfac = 3.50
Soil Freundlich "n" = 0.854
Carbon Freundlich "n" = 0.430
Vs=0.0 cm/hr; Kd = 9.02E-4/hr
P |B [ B i | B "P i | i
123456789 10
DISTANCE FROM INTERFACE (cm)
0 0.309% Cobb Fly Ash Column at 774.6 hr
• (3) S-BHfac=2.63;Kfs=3.58E-4;Kfc=19.47
• (4) S-BHfac=2.63;Kfs=3.58E-4;Kfc=9.84
Figure 10
Transient Column Data and Simulations for 0.309%
Cobb Fly Ash Experiments
the exterior concentration was considered to be zero. The analytic
solution for the solute flux at the exterior face of the barrier which was
adapted for the applicable form of Equation 4 and stated boundary
conditions is25:
FLz°
De,iC0,i
1+21 (-l)nexp|-
n-1
-2 2,
(10)
where TB is the barrier thickness (3.3 ft) and all other terms are as
previously defined. The total solute migration (Q() through the barrier
is simply the integral over time of the flux:
(11)
In the case that the barrier was non-sorbing, De was obtained
directly from equation 2. For the condition of a barrier with sorption
capacity, the LEA was invoked, the partitioning relationship was
assumed linear, a retardation factor (R) was defined and the quotient
De/R was substituted for De in Equation 10. The partition coefficient
chosen corresponded to that for a mixture containing 40% Cobb fly
ash and was calculated using Equation 9.
Plots of Fiz" versus time for the two limiting cases considered are
shown in Figure 11 and plots of Q, versus time for the two limiting
cases are shown in Figure 12. For the case of no sorption capacity,
breakthrough is predicted to occur within 2 yr and near-steady-state
flux rates are predicted to obtain within 12 yr. Conversely, for the case
of barrier modification by fly ash, solute breakthrough is predicted to
occur in approximately 30 yr and near-steady-state flux rates are
predicted to obtain within approximately 220 yr.
The solute migration rates shown in Figure 12 are based on an aqueous
concentration of 1 mg/L within the barrier. To obtain an estimate of
total solute escape, these values must be multiplied by the true interior
concentration and the exterior area of the barrier (10s cm2). Given that
the solubility of CTET is 785 mg/L, the ultimate steady-state solute
migration rate through the hypothetical barrier could approach 85 kg/yr.
For the case of no sorption capacity, the total solute escape during the
BARRIERS 531
-------
first 12 yr could approach 600 kg. From these simulations it may be
concluded that diffusive transport can be quite significant and that the
addition of fly ash to soil-bentoniie barriers can significantly improve
performance.
The true behavior of a barrier will fell between the two extreme cases
considered here; thus, for proper evaluation of mitigative strategies,
it is imperative that methods be developed to more accurately predict
solute transport in these barriers. Moreover, if soil-bentonite barriers
are to be viable alternatives for mitigation of subsurface contamination,
means of reducing the magnitude of effective diffusion coefficients
should be sought.
EC
O
2 ^ 0.5
UJ n
x LU
UJ „•
C0.4
o>
UJ «
oc E
oc »:
< 0,3
o g
X «
R O"
O
w
0.2-
0.1 -
0.0
1 10 100
ELAPSED TIME (yr)
• D(aq) = 1.07E-5; porosity = 0.48; w/o fly ash
• D(aq) = 1.07E-5; porosity = 0.48; w/fly ash
Figure II
Model Simulations of Solute Flux from a Hypothetical
Soil-Benionite Containment Barrier
1000
fc»
fcw
CO
•o ir-
CO
.cŁ
<*•• ^™
O)
3 CO
0 *
H
.- E
»- o
o
o. tr
(A (0
C" D)
CO ^->
(1)
3
o
0.6-
05-
0.4-
0.3-
0.2-
0.1 -
o
•
O A
. ' X*
o ^*
_^y
sjpfi ^^^
id^.^****
0 100 200 300 400 50
TIME (YR)
H Barrier without fly ash
• Barrier with tly ash
Figure 12
Simulations of Solute Escape from a Hypothetical
Soil-Bcntonite Barrier
CONCLUSIONS
The results of both quasi-steady-suite and transient diffusion
experiments suggest that effective diffusion coefficients for
low-molecular-weight solutes in soil-bentonite barriers are reduced only
several factors from those in free aqueous solution. Effective and free
aqueous diffusion coefficients appear to be related through a power
function of porosity.
The capacity of carbon associated with fly ash for sorption of
representative low-molecular-weight solutes was found to be at least
equivalent to that of naturally occurring organic carbon. The Freundlkh
sorption capacity factor, Kp, was found to correlate well on a log-log
basis with both the octanol/water partition coefficient and aqueous
solubility.
Simulations based on a hypothetical barrier and performed for two
limiting cases suggest that solute migration through soil-bentonite
barriers by molecular diffusion can be significant, and that the addition
of a sorbcnt phase such as fly ash to soil-bentonite mixtures can
markedly improve the performance of such barriers
ACKNOWLEDGEMENTS
This work was supported in pan by Research Grant No. R8H570-OI-0
from the US EPA, initially, and subsequently, by Research Grant No.
1-P42-ES049I1-OI from the National Institute of Environmental Health
Sciences.
REFERENCES
I DAppolonia. DJ . -Soil-Benlonnc Slurry Trench Cutoffc," ASCE Jour
Geo-iechnical Engr.. 106(014). 1980.
2 Schul/c. D . Barvemk. M and A) res J . "Design of Soil-BeWomur Backfill
Mix for the First Environmental Protection Agency Supcrfund Cutoff Will,"
from The Proceedings of die Fourth National Symposium and Exposition
on Aquifer Restoration and Ground Waier Monitoring. Columbus. OH. May.
1984
3. Jepson. C.P.. "Sodium Bentmde: Still a Viable Solution for Hazardous Wfatt
Containment." Mlui. Eng.. Apr., 1984
4 White. LA.. Dasgupta, A andCota. M F. "Containment of Uncontrolled
Hazardous Waste Sites." Jour Haz, Mai.. 14. 1987.
5. Goodall. D.C and Quigley. R.M . "Pollutant Migration from Two Sanitary
Landfill Sites Near Sarnia. Ontario." Can. Geourk. J.. 14, 1977.
6 Johnson, R L . Cherry. J.A and Pankow, J.F.. "Diffusive Contaminant
Transport in Natural Clay: A Field Example and Implications for day-Lined
Waste Disposal Sites," Eimron. Sri. Technol . 23, 1989.
7. Gray, D H and Wtber. WJ.. Ir . "Diflusional Transpon of Hazardous W«t
Lcachate Across Clay Barriers." Presented ai the Seventh Annual Madison
Vvfeste Conference, University of Wisconsin-Extension. Madison. WI. Sept,
1984.
8. Millington. R J . "Gas Diffusion in Porous Media," Sciencr. 130, July. 1959
4 Millington. R.J. and Quirk. J.P.. "Permeability of Porous Solids." Kuaday
Tram. . J., Farmer. WJ. and Klute, A., "Lindane Diffusion
in Soil." Soil Sri ix Am. Proc.. 37, 1973.
12. Farmer. W.J.. \ang. MS. Letty, J. and Spencer. W.F.. "Hcjnchtorobenzene:
Its Vapor Pressure and Vapor Phase Diffusion in Soils," Soil Sri. Soc Am.
J . 44. 1980,
13. Gillham. R.W.. Robin. M.J.L. and Dytynyshyn. D.J., "Diffusion of
Nonrcactive and Reactive Solutes Through Ftne-Grained Barrier Materials."
Can. Geoifch. /, 21, 1984. p. 541-550.
14 Mott, H.V.. "Diffusive Transpon of Low-Molecular-Weight Organic Solutes
through Soil-Bcnlonitc Containment Barriers." Ph.D. Thesis, The University
of Michigan, Ann Arbor, Ml, 1989.
15. Bennett. CO. and Myers, J.E,, Momentum. Heal and Mass Transfer, 3rd
ed . McGraw-Hill, 1982.
16. Cussler. E.L., Diffusion: Mass Transfer in Fluid Sntems. Cambridge
University Press. 1984.
17. Jennings. A.A. and Kirkncr. D.J., "Instantaneous Equilibrium Approximation
Analysis," ASCE J. Hydraulic Eng., #0(12), 1984.
18. Bahr, J.M. and Rubin, J.. "Direct Comparison of Kinetic and Local
Equilibrium Formulations for Solute Transport Affected by Surface
Reactions." Witer Resources Kes.. 2.*(3), 1987.
19. Bergstrom, W.R. and Gray. D.H.. "Fly Ash Utilization in Cut-Off Wills:
A Progress Report to Consumers Power Company," The University of
532 BARRIERS
-------
Michigan, Department of Civil Engineering, Ann Arbor, MI, August, 1987. 9(13), Dec., 1975.
20. Lymann, W.J., Reehl, W.F. and Rosenblatt, D.H., Handbook of Chemical 23. Karickhoff, S.W., Brown, D.S. and Scott, T.A., "Sorption of Hydrophobic
Property Estimation Methods, McGraw-Hill, New York, NY, 1982. Pollutants on Natural Sediments," Water Res., 13, 1979.
21. Gossett, J.M., "Measurement of Henry's Law Constants for C, and C2 24. Chiou, C.T., Peters, L.J. and Freed, V.H., "A Physical Concept of Soil-Water
Chlorinated Hydrocarbons," Environ. Sci. Tech., 21(2), 1987. Equilibria for Non-ionic Organic Compounds," Science, 206, Nov., 1979.
22. MacKay, D. and Leinonen, P.J., "Rate of Evaporation of Low-Solubility 25. Crank,, J., The Mathematics of Diffusion, Second Edition, Clarendon Press,
Contaminants from Water Bodies to Atmosphere," Env. Sci. Tech., note, Oxford, 1975.
BARRIERS 533
-------
Use of Synthetic Liners in Recent Superfund Cleanup Projects
Mark W. Cadwallader
Gundle Lining Systems Inc.
Houston, Texas
ABSTRACT
Many NPL sites for which the Superfund is responsible are finally
being cleaned up after several years of engineering studies. Contaminants
at a large number of these sites are being contained with the help of
flexible membrane liners functioning as barriers to waste migration.
A number of different liner systems are used in the containment
process. The current concept in liner system design can be summa-
rized as incorporating redundant layers of liner and using composite
liners which takes advantage of the synergism between materials with
different mechanisms of barrier activity. Provisions also are made for
fluid drainage both above and between liners.
INTRODUCTION
The Superfund program has taken a first step toward addressing one
of the most perplexing environmental challenges of all time, hazardous
waste and its uncontrolled storage/burial. Future waste disposal will
be very expensive and billions of dollars currently are being set aside
to rectify past problems. Since we did not pay to properly dispose of
the waste in prior years.
But what are proper waste disposal techniques? Many people talk
about waste recycling or incineration as though they are the panacea
to the problem. Yet neither process is a complete answer to the hazardous
waste problem. There still remains non-reusable waste and ash that are
often more hazardous than before treatment. Deep well injection, another
viable disposal technique, also has limited application. The simple truth
is that disposal and containment of waste products on and below the
earth's surface must continue for lack of better alternatives.
Since surface containment of waste is necessary, the wisest approach
is to provide the best possible barrier for waste containment. If money
is spent on appropriate barriers to prevcni waste migration, savings will
result because future cleanup operations will not become necessary.
Also, if costs are increased for traditional surface containment because
of better barrier construction, desirable alternatives such as recycling
become more cost competitive. An incentive is therefore provided to
recycle waste and/or limit waste production. Liners are often used as
barriers to containment transport. Examples of liner systems arc shown
in Figure 1.
LINER TECHNOLOGY
With the advent of copolymer, pipe-grade HDPE technology, synthetic
liners in current landfill technology have achieved strength, toughness,
durability, chemical resistance and environmental stress crack resistance.
The desirable qualities of HDPE as a barrier material can be seen from
its increasing applications in the container market. Much growth is ex-
pected for HDPE containers of agricultural chemicals, insecticides,
herbicides, paint thinners and household chemicals as well as other
Txv^S^SHJff?"*'*
— W'''''''''-**
Figure I
Liner Systems Available
chemical products1. HDPE is expected to replace more and more
traditional metal and glass containers in the market place. High molecu-
lar weight HDPE is also used to contain low level radioactive waste.
A composite liner, formed by a synthetic liner such as HDPE in con-
tact with a clay liner, offers the greatest degree of impermeability. This
good performance occurs because the permeability of the synthetic liner
is very low and containment transport occurs only by diffusion. Unlike
convection, diffusion is driven by a concentration gradient instead of
a pressure gradient. The absorptivity and porosity of a day liner directly
underneath a synthetic liner means that the chemical concentration
gradient is reduced across the synthetic liner since diffused chemical
species accumulate in the clay pores at the clay/synthetic interface. With
chemical concentrations approximately the same on both sides of the
synthetic liner, the driving force for diffusion is eliminated.
A syntheiic/clay liner works well because the synthetic liner is a
barrier to pressure-driven mass transfer, while the underlying clay liner
forms a barrier to concentration-driven mass transfer. The combina-
tion liner comprised of a synthetic liner, drainage layer, synthetic liner
and clay, is the liner system required for hazardous waste containment
under Subtitle C of the Hazardous and Solid Waste Amendments to
the RCRA.
The drainage layer in the landfill can be either conventional sand
or gravel, or drainage netting. Drainage netting is a net-shaped material
made by overlapping high density polyethylene strands to conduct fluids
in the plane of the net. Drainage netting generally has a hundred times
more flow capacity than sand or gravel drainage layers.
534 BARRIERS
-------
SUPERFUND CLEANUP STRATEGY
Every Superftmd site requires its own special considerations before
a cleanup strategy can be mapped out. The goal, however, in every
case is to prevent the toxic waste at the site from causing an ecological
or human health problems.
Current barrier technology can offer a number of practical approaches
to site cleanup. These procedures generally can be classified into the
following groups:
• Removal of the contaminated material off-site for containment in a
RCRA-approved hazardous waste facility or treatment/cleansing of
polluted soils.
• Construction of a RCRA landfill at or adjacent to the site for transfer
and proper containment of the polluted soil
• On-site containment of the waste by construction of an impermeable
cap and barrier wall to prevent infiltration of surface water/precipi-
tation and the spreading of contamination to the surrounding
groundwater
Removal of contaminated material to a hazardous waste facility means
the construction of increased capacity at RCRA-approved disposal sites.
This option requires the transportation of hazardous waste.
Proper construction of a landfill at or adjacent to the Superfund site
would also demand the double liner technology required under RCRA.
All the considerations appropriate to hazardous waste facility construc-
tion centered around the installation of two layers of synthetic liner would
apply.
The construction of caps and barrier walls for on-site containment
of Superfund waste will likely be a frequent strategy in cleanup work.
The use of synthetic liners for cap construction has been proven effec-
tive. Barrier walls are primarily constructed with bentonite slurry.
With SARA setting timetables for cleanup activity at NPL sites, money
will be spent for actual cleanup rather than for just paper studies. In
the past, the major share of the Superfund money has gone to study
the problems rather than clean them up. One exception and an early,
now classic Superfund project, is the work at Nashua, New Hampshire.
PUMP-AND-TREAT FLUSHING WITH SYNTHETIC LINERS
Superfund cleanup work at Nashua, New Hampshire, utilized the
construction of a cap and barrier wall to meet a fast-moving plume of
groundwater contaminated with illegally dumped organic solvents
including chlorinated hydrocarbons. The contaminant plume was moving
at a rate of approximately 2 ft/day when work began in 1982. The 20-ac
synthetic cap was constructed with an HDPE geomembrane liner. The
barrier cut-off wall was made from a bentonite slurry; it extended down
to bedrock and ringed the site in an oval shape (Fig. 2).
Because of fractures in the bedrock and because of evidence that the
organic chemicals in the aquifer would tend to degrade the bentonite
by altering the mineral composition of the clay, groundwater interception
and treatment was implemented through the use of pumps. Contami-
nated water is thus being pumped out of the containment area, treated
and re-injected so as to flush out remaining contaminants. This inno-
vative and economical Superfund project at Nashua, New Hampshire,
likely is indicative of the approach to be used at many sites in the future,
i.e., cap and barrier wall construction with pumping of contaminated
water to lower the water table within the containment and remove
pollutants.
Construction of caps for other Superfund cleanup projects could utilize
other geosynthetic materials such a drainage netting for drainage of
surface precipitation above the impermeable geomembrane layer as well
as geotextile for separation of cover soil from the fluid flow zones.
Similar pump-and-treat techniques are now available using air as the
mobile phase to extract up organic contaminants from soils. Synthetic
liner caps are important in such cases to intercept rain water and prevent
further movement of the contaminants into and through the ground-
water. With a synthetic liner impermeable to the contaminated air being
flushed out, funnelling, collection and treatment of the air is facilitated.
CAP APPLICATIONS AT SUPERFUND SITES
Four very recent Superfund projects have utilized synthetic liners for
HnPE Cover
Bentonite Slurry Wall
Figure 2
Schematic of Cap and Barrier Wall On-Site Containment System
For Superfund Project
cleanup purposes. They include Pine Bluff Arsenal, Pine Bluff,
Arkansas (2.6 million ft2 of 40-mil and 80-mil HDPE); Lackawanna
Landfill, Old Forge, Pennsylvania (1.3 million ft2 of 60-mil HDPE);
Charlie George Landfill Cap, Tynsborough, Massachusetts (2.8 mil-
lion ft2 of 60-mil HDPE); and Helleva Landfill Cap, LeHigh County,
Pennsylvania (1.8 million ft2 of 60-mil HDPE).
Special situations at Pine Bluff Arsenal included the presence of
dangerous chemical weapons and the containment of 13 separate sites
- two hazardous waste landfills, nine landfill caps and two surface im-
poundments. Several of the sites used some rather recent geosynthetic
product developments such as specially textured sheets for extra slope
stability and synthetic drainage netting for fluid flows in the liner system.
OSHA levels B, C and D protection were required for the crews at
the various project sites. Installation crews were, in some cases, required
to have OSHA training before working around the hazardous waste.
Cooperation with unions was necessary at both Charlie George and
Helleva Landfills.
LINER INSTALLATION
In all of these Superfund projects, when liner construction began the
cell earthworks were already prepared and graded. A front-end loader
was used to deploy the HDPE, which was manufactured by Gundle
Lining Systems, Houston, Texas, in 22.5-ft. wide seamless rolls, each
weighing about 2,800 Ib. All field seams were welded by Gundle
employees using the company's patented extrusion welding machine
with mixing tips to improve heat transfer, or'the Gundle automatic dual
hot wedge welder. All of the seamed footage was either vacuum tested
for voids or air pressure tested in the case of the hot wedge welds.
Vacuum testing uses a plexiglass faced, rectangular box placed over
a section of the seam. A 5 psi vacuum is pulled on the box. Any voids
in the seam will form bubbles in the soap solution sprayed onto the
seam before the box is set in place (Fig. 3).
HO GUNDLINE LINER
(2) VISE CRIPS WELDED
TO STEEL BAR
Figure 3
Seam Air Pressure Test
BARRIERS 535
-------
In the case of the hot wedge welding, several advantages of (he seaming
method should be noted. The Gundle hot wedge welder automatically
feeds the sheet across the hot wedge and through the pressure rollers
The welder also automatically positions the wedge accurately at the
edge of the lop sheet and automatically adjusts the roller gap 10
accommodate different sheet thicknesses These features of the hoi
wedge enable it to achieve welding speeds of up to 15 ft in in
With the dual hot wedge welder, non-destructive testing is made more
efficient because of air pressure testing of the "split" or "dual" wedge
of the system. The dual wedge system leaves a gap between two separate
wedge weld tracks. In the air pressure test, the gap is pressuri/ed to
about 30 psi and possible leaks arc noted by the reduction in pressure
over 5 min.
Buffing of the sheet is not necessary with the hoi wedge, unlike
extrusion techniques This, along with the increased welding rales.
makes the hot wedge welder very cost effective for high quality
construction of liner systems in waste coniainmcm.
In addition to the non-destructive testing, samples for destructive scam
testing were cut from the field seams at regular intervals The samples
were tested for both shear and peel on a icnsiomcicr. In shear testing.
one applies a tensile stress across the weld from the lop through the
bottom sheet. Peel testing peels the lop sheet back against the over-
lapped section of the bottom sheet to observe how the weld is coming
apart. If the weld (ears (called a "film tear"), then the weld has formed
a homogeneous connection through the seam. This is the desired result.
If not, the lest signifies a defective scam and it must be repaired.
CONCLUSION
The suitability of high quality geomembranes as barriers in the con-
tainment of hazardous waste has been, and is continuing to be, demon-
strated on a very wide scale. The adaptability of the products and
construction techniques to marry different situations is continuing to
prove their usefulness and is extending their application lo the highly
important work of the Superfund.
REFERENCES
I. US. KTA, Minimum Technology Guidance of Double Liner Syoems/or Lond-
filh and Surface Impoundment]—Design. Construction and Operation. U.S.
I-PA. VWtthington. DC. 1985
2. Lcovenuch. R . "HOPE Resin of Choice for Barrier Functions." Modem
Plastics, pp 6J-7I, May 1986
536 BARRIERS
-------
Some Observations of the Influence of Deformation
On a Clay Liner
KJ.L. Stone, Ph.D.
U. Guttler, Ph.D.
H.L. Jessberger, Ph.D.
Ruhr-University of Bochum, F.R.G.
ABSTRACT
Centrifuge model tests have been performed to study the response
of clay barriers subjected to differential deformations. The modes of
deformation mat have been observed are relevant to those that might
occur as the result of differential settlements of waste material—leading
to deformation of cover liners—or from non-uniform soil strength
profiles or deep sited subsidence—leading to deformations of base liners.
Plane model liners were constructed both from pure kaolin and from
a mixture of sand, silica flour and bentonite. The integrity and perfor-
mance of these model liners was evaluated on the centrifuge at a force
of 50 gravities. Physical degradation of the model liners was monitored
photographically and their performance as effective hydraulic barriers
was assessed throughout the deformation process.
For all the model liners where no overburden was present, tension
cracking of the liner surfaces was observed. These tension cracks were
very significant in the kaolin models and led to a drastic reduction in
liner performance. However, the presence of an overburden suppressed
the formation of tension cracks, and no significant reduction in the kaolin
liner efficiency was observed.
The sand/silica flour/bentonite liner material proved to be highly
resistant to deformation with little evidence of tension cracking and no
significant reduction in performance.
INTRODUCTION
Solid waste disposal in shallow landfill depositories has been exten-
sively employed for the permanent disposal of both municipal and
industrial wastes. However, this practice has generated much concern
over the possiblility of environmental contamination. The principal con-
cern is the potential for groundwater contamination resulting from the
permeation of leachate out of the landfill. Such leachates are likely to
contain concentrations of chemical and biological pollutants produced
by the decomposition of the contained wastes. A typical engineered
landfill depository is shown schematically in Figure la.
The key element to the successful operation of such a landfill is the
presence of a hydraulic barrier surrounding the waste material. A further
reduction in the possibility of groundwater contamination can be
achieved by minimizing the potential for a buildup of leachate within
the deposit. This added protection requires the temporary covering of
the landfill during filling followed by placement of a permanent cover
once filling is complete and an acceptable degree of waste stabilization
has occurred. Both the base and cover liner systems commonly are fabri-
cated from compacted clay—usually a few per cent wet of optimum
proctor compaction.
However, such compacted clay liners may fail to perform satisfac-
tory for several reasons. For example, cover liners are susceptible to
climatic effects such as dessication cracking1 and frost action2, as well
as deformations of the liner itself3, (Fig. Ib), caused by differential
settlements of the contained wastes.
/ -c
r\{_^~ -S™M;:> ' Ł ( f I,
Vv^^C^l'J
v
/ /
SOIL COVER
COMPACTED
CLAY LINER
Figure la
Schematic Representation Of A Landfill Facility
.ORIGINAL LINER
'POSITION
CRITICAL REGION
OF DEFORMATION
DEFORMED LINER
Figure Ib
Illustration Of Liner Deformation Mode.
Similarly, base liners are prone to chemical attack from the contami-
nated leachate, as well as to differential liner settlements resulting from
non-uniform soil strengths below the landfill4 or from near surface
ground movements associated with deep sited substances.
The tests reported in this paper are concerned with the effect of
differential settlements on the performance and integrity of clay liners.
Previous studies5 where part of an underlying basement has been dis-
played to introduce a discontinuity of slope, but not of displacement,
have shown that such continuous boundary deformations can lead to
the formation of discontinuities or ruptures in the overlying soil. It is
of interest, therefore, to study the response of clay liners subjected to
such boundary deformations and to investigate the parameters which
influence the stress dependent liner response. In particular, the evalua-
tion and comparison of pure clay and fine/coarse mixture sand/clay liner
materials has been investigated, and their performance as effective
hydraulic barriers throughout the deformation process has been assessed.
CENTRIFUGE MODEL TESTS
Introduction To Centrifuge Model Testing
It is well known that the behavior of most soils is very dependent
on stress level. In conventional small scale model tests performed in
the earth's gravitional field, it is not always possible to maintain similarity
BARRIERS 537
-------
with prototype situations and to ensure that stress levels in areas of
interest reach prototype values. A geotechnical centrifuge can subject
small models to centripetal accelerations many times the earth's gravita-
tional acceleration. By selecting a suitable acceleration level, the unit
weight of the soil being tested can be increased by the same proportion
by which the model dimensions have been reduced, and thus stresses
at corresponding points in the model and prototype will be the same.
The centrifuge model tests reported here were performed on the
Bochum 10 m balanced beam centrifuge at an enhanced acceleration
level of 50 gravities. Details of the Bochum Geotechnical Centrifuge
can be found elsewhere'.
Scaling Relationships
Centrifuge scaling relationships have been extensively described
elsewhere7. However, if we consider a model where the prototype
dimensions have been reduced "n" limes such that d /d = n. where
p PI»
dp and dm are prototype and model dimensions respectively, and if 'n'
isP chosen as the gravity scaling factor, then the Table 1 illustrates the
basic scaling relationships associated with centrifuge modeling.
Table 1
Centrifuge Scaling Relationships
. AIR COEHUBI
(Parameter |
1 1
| Gravity |
| Length |
| Stress |
| Strain |
| Force |
Units
m/s*
m
PA
z
N
| Scaling Relationship |
1 1
1 n 1
1 Vn I
I 1 1
1 1 1
1 1/n' I
I TLne*
sec
1/nJ
Note: These relationships apply to laminar flow processes such as
consolidation.
If the same material is used in both the model and prototype, then
the similarity of stress levels ai corresponding points in the model and
prototype will result in a model response directly analogous to that of
the prototype. Furthermore, the prototype stress gradient present in
the model will ensure similarity of the primary permeability
distribution7,
Centrifuge Model Package
The centrifuge model tests were performed in a rectangular strong
box of internal dimensions 395 mm wide x 658 mm long x 395 mm
high. The front of the strong box is formed by a 70-mm thick Perspex
window through which deformations of the model can be photographi-
cally observed while the model is "in-flight" on the centrifuge.
The model test package is shown schematically in Figure 2. In order
to generate a displacement profile at the base of the model liner, a rec-
tangular piston is centrally located in the floor of the strong box. This
piston extends the full width of the strong box and has a maximum
throw of 25 mm. A flase base containing a pair of 95-mm hinged flaps
is located across the strong box and so arranged that when the piston
is lowered, the flaps rotate and induce a discontinuity of slope at the
base of the overlying soil—as represented by the dashed line in Figure 2.
Linear variable displacement transducers (LVDTs) arc used to monitor
water levels and liner deformation.
To minimize the possibility of leakage between the liner and the strong
box sides, the overlying water is contained within a shallow trench—
• TO IIP OHO
'MRftCt WATER SUPPW
IY0I
couECTKw tistBi
tOO UACHATE
Figure 2
Schcnutic Mliutnuon Of Centrifuge Model Ted Package
model landfill—as illustrated in Figure 4a. The depth of surface water
present is monitored by a LVDT and float and can be increased by rekas-
ing water from vessel A. The water table below the liner is maintain^
at the pre-set level of outlet B. The overflow from this outlet B is
collected into the cylindrical vessel D. Consequently, by monitoring
the rise of the water level in D, the rale of water flow through the liner
can be deduced and hence the permeability can be estimated.
MODEL PREPARATION PROCEDURE
Choice Of Uncr Materials
Two different liner materials were chosen. The first was a commer-
cially available kaolin clay (2096c kaolin) supplied by Erbsloh & Co.,
W. Germany. This clay has a liquid limit of 44.4% and a plastic limit
of 28.1% There is much debate as to the optimum design water content
that should be used for compacted clay liners, but for the purposes of
this experimental study, a moisture content corresponding to 95% satu-
rated Proctor density was adopted.
The second liner material was a sand/silica flour/bentonite mixture
(hereafter referred to as the fine/coarse liner material) of the following
proportions (Table 2).
Table 2
Composite Model Uncr Mixture
I Material I Particle site (approx.)l Percentage by «ci|ht|
I
Coartc sand I 1 - 1 . 5 »•
Silica flour I 30 pa
Bcntanitc | 80 I < 2 )•
1
6* J
22 1
14 1
1
This model liner material was chosen to represent the prototype mix-
ture shown in Figure 3. As can be seen in this figure, the prototype
mixture contains a large gravel fraction. This gravel fraction has been
scaled down and replaced by the quartz sand fraction in the model liner
mixture.
Model Preparation
After all the internal components have been fitted into the strong box,
a 30-mm layer of coarse sand overlain by a further 45 mm of fine sand
was poured into the strong box. A layer of Filter paper was placed just
below the final sand surface to prevent fine panicles of the liner material
from being washed out. A row of discrete markers was placed against
538 BARRIERS
-------
PARTICLE SIZE DISTRIBUTION CURVE
100
90
1"
K70
§60
r
U. 60
o
I30
Z 20
|10
* 0
CLAY
—
FINE M
I
— - •
SILT
EDIUM
'
1 1 1 1
COARSE
1
s^™»
I
V • H
D001 0,002 0.006 0,01 0,02 0,06
DIAMETER d in mm ^
FINE
1 1 M
•-
MM
SAND
MEDIUM
T
^
1
X*
^
S
COARSE
x-
1111
:— =
1 1 1 1
FINE
I
,— —
I
.. —
GRAVEL
MEDIUM
- -
I 1 1 I
/
/
x^
1 1 1 1
COARSE
I
/
/
/
I
1
STON2
0.1 0,2 0,6 1 2 6 10 20 60 100
Figure 3
Grading Curve For Prototype Fine/Coarse Mixture Liner Material.
the Persplex window on the sand surface for subsequent digitisation
from "in-flight" photographs. The sand was then saturated by the up-
ward percolation of water introduced via a network of drainage holes
at the base of the strong box. After greasing the internal sides of the
strong box, the model was ready for liner fabrication.
Kaolin Liner Preparation
Kaolin slurry was placed by hand—to avoid air entrapment—to a pre-
determined depth over the saturated sand. A consolidation unit then
was attached to the strong box and the slurry was one-dimensionally
consolidated to a final vertical effective stress of 630 kPa. This final
effective stress level is consistent with the moisture content associated
with a 95% (saturated) proctor compaction density.
After removing the consolidation unit, a shallow landfill was formed
in the consolidated liner (Fig. 4a).
Composite Liner Preparation
The preparation of the fine/coarse mixture liner was somewhat
arbitrarily arrived at for the test reported here (test KDB1). The material
was mixed to a moisture content of 35% and placed by hand to the
required depth. The consolidation unit was then fitted and a pressure
of 100 kN/m2 was applied. This pressure was maintained for 3 days.
After removal of the consolidation unit, a shallow landfill was formed
again in the consolidated liner. This preparation procedure was simply
intended to produce an initially saturated model liner in a reasonably
short time. However, for future tests the material will be placed dry
and vibro-compacted before saturating by upward percolation. This
process is more representative of the prototype placement method, but
it has the disadvantage of requiring a very long time for saturation.
4 CENTRIFUGE MODEL TEST RESULTS
The corresponding model and prototype boundary conditions for the
tests reported here are given in Table 3.
The principal objective of the model tests was to investigate the phys-
ical response of model liners, subjected to various degrees of deforma-
tion, as illustrated through the development of cracks and ruptures. The
Table3
Model and Prototype Boundary Conditions
(Test No.| Liner material
I I
I Liner thickness JDepth of overburden I
|Model(mn) (Prototype(m) |Model(ran) |Prototype(m) |
.1 I I I I
| IDS
| TD9
| TD10
| KBD1
I
I 2096c Kaolin
I 2096c Kaolin
I 2096c Kaolin
I Sand/Silica flour/
I Bentonite mixture
35 |
35 |
40 |
I
40 |
1.75
1.75
2.0
2.0
I 0 |
I 50 |
I 0 |
I I
I 0 |
0 I
2.5 |
0 I
I
0 I
effect of overburden and choice of liner material on the model response
was observed. In addition, the performance of the liners as effective
hydraulic barriers was monitored throughout the deformation process.
This test process enabled the effects of losses in liner integrity—such
as cracking—to be quantified.
As mentioned earlier, the liner deformations are induced by the ver-
tical translations and rotations of the piston and flap arrangement lo-
cated at the base of the sand layer. However, it is difficult to relate these
movements to the actual degree of deformation suffered by the liner.
Consequently, the degree of liner deformation is defined as the degree
of rotation, 0, (Fig. Ib) that has occurred at the base of the liner. This
angle is deduced from digitized recordings of the discrete markers placed
at the sand/liner interface.
Tension Cracking And Rupture
Figure 4b shows a post-test photograph of the model liner surface
of test TD10 with a liner deformation of 8°. Severe tension cracks are
clearly evident in the regions of maximum liner deformation. The de-
velopment of such tension cracks was a typical feature of the pure kao-
lin'clay liner tests where no overburden was present. The degree of
liner deformation at the onset of tension cracking will be a function
of liner thickness for similarly prepared models8.
BARRIERS 539
-------
TENSION CRACKING
Figure 4a
Prc lest Photographs For Kaolin Model l.incr
Figure 4b
Post-test Photographs For Kaolin Model Liner. Tni TDK)
For the tests reported, here the onset of tension cracking was observed
at 3 to 3.5° for the pure kaolin tests TD8 and TDK) After washing
away the sand overburden in test TD9. no tension cracking was evident
with a liner deformation of 11° For (he benumite/sand liner (test KBD1).
slight surface cracking was observed at a liner deformation of 7.5°. but
no significant cracking developed even after the maximum liner defor-
mation of 16° had been introduced.
After each test, the Persplex front face of the strong box was removed
and the model liner was sectioned to examine the depth of tension cracks
and (he presence of any other internal damage For tests TD8 and TD
10, the tension cracks extended vertically to (he base of the liner
(Fig. 5a). Careful sectioning of (csi TD9 revealed no further evidence
of tension cracking; however, a scries of multiple shear ruptures in the
regions of greatest liner deformation was clearly observed (Fig. 5b).
These ruptures curved out over the break in slope, indicated an arching
type mechanism of material response Finally, the post-test examina-
tion of the deformed bcnlonite/sand liner did not reveal any significant
material degradation, and only shallow surface cracking in (he regions
of greatest liner deformation was visible.
Assessment of Liner Performance
The performance of the model liners as effective hydraulic barriers
is besl illustrated through the rate of Icachate (water) flow through the
liner This flow rate is directly observed by the rise of the water level
in collection vessel D (Fig. 2). However, conversion of this flow rate
to an average value of liner permeability is complicated by the non-
uniform hydraulic gradient present across the model liner.
This non-uniform hydraulic gradient arises from (wo conditions. The
RUPTURE
SURFACE
Figure Sa
Trace Of Crack and Rupture Pattern* Obterved Without Overburden, Tot TM
SAND OVERBURDEN
MULTPIE
SHEAR RUPTURE
KAOLIN LINER
TEST TO 9
Figure 5h
With Overburden. Ten TD9
Tint condition is unique to centrifuge modeling and is the tendency of
water levels in centrifuge models to align along lines of equal radius
from the axis of centrifuge rotation. Thus, the surface water level within
the model landfill will, at all limes, maintain a concave curvature
resulting in higher hydraulic gradients away from the center line of the
model. Second, the deformations introduced during the test will result
in increased hydraulic gradients over areas of liner depression. Thus,
in light of these and other complications the values of average permea-
bility stated herein should be treated with caution.
Figure 6 shows the settlement record of the center of the model liner
superimposed on the LVDT trace monitoring the water level in the col-
lection vessel D for test TDK) These data are typical lest results far
a kaolin clay liner with zero overburden, and they illustrate the following
characteristic behavior From A to B the centrifuge is accelerated op
to speed and (he collection vessel D is rapidly filled and discharged
as water in the underlying saturated sand is centrifuge down to the led
of connection B (Fig. 2) From B to C there is still quite a rapid flow
into vessel D as water continues to be expelled from the sand and also
from the self-weight consolidation of the clay liner. From C to C it
can be assumed that no more water is being forced out of the sand and
that the amount of water deriving from the liner itself is minimal.
From this flow rate, an initial value of average liner permeability was
found (o be 1.3 x 10* m sec From C' to D to flow rale is seen to in-
crease, while a deformation to 3° is introduced at the base of the liner.
This increase in flow rate again will be partly due to water being
squeezed out of (he liner itself, so a realistic calculation for permea-
bility cannot be made during the actual deformation process. However,
for the region D to E. the value of average permeability was found to
be 1.18 x 10*. This value suggests that within experimental error no
delectable change in permeability has occurred with a liner deforma-
tion of 3°
Further deformations are introduced (E 10 F). but again there ait
no significant increases of flow rate through the liner until a deforma-
tion o! 6.4° is obtained at F. at which point a dramatic increase in flow
rale in(o vessel D is observed. The subsequent reduction in flow rale
(region F to G) before further deformation indicates a self-healing poten-
tial of the clay. The behavior of all the kaolin clay liners without over-
burden exhibited a behavior similar to thai illustrated by Figure 6.
The onset of a sudden increased flow rate through the liners can be
considered to be a serious failure of the liner and corresponds to the
development of deep tension cracks and ruptures forming a preferen-
tial flow path through the liner in the regions of maximum deforma-
tion. For the lest performed with an overburden pressure (test TD9),
no such liner "failure" was observed, and it can be concluded that the
540 BARRIKRS
-------
TIME IN HOURSIITEST TD1
LINER SURFACE
SETTLEMENT
IN
Figure 6
Selected Transducer Records For Test TD10
(kaolin liner, zero overburden).
presence of shear ruptures did not significantly affect the liner's
performance.
Comparison of Kaolin and Fine/Coarse
Mixture Liner Performance
The fine/coarse mixture liner model was made to the same initial
boundary conditions as test TD10 and subjected to a similar time/defor-
mation test history. Figure 7 shows a plot of liner settlement and leachate
collection level for the full duration of the test. Comparison of this plot
with the corresponding record for test TD10 (Fig. 6) illustrates some
fundamental differences.
First, the initial flow rate reduces to virtually zero (region A to B)
indicating extremely low permeability of the model liner. As observed
in the kaolin tests, the flow rate increases during deformation but
approaches zero again soon after stopping the deformation (region B
to C). This behavior is repeated until 9.5° of liner deformation is
achieved (at D). At this point a permanent increase of flow rate is
observed from which an average permeabiliity of 1.89 x 10'1U m/sec has
been calculated. On further deformation to 11.5 °, the flow rate into vessel
D increases slightly but then remains constant for the remainder of the
deformation process. The final average permeability for the liner at the
end of the test with a liner deformation of 15° was estimated to be 2.915
-60 u
Figure 7
Selected Transducer Records For Test KBD1
(fine/coarse mixture material liner, zero overburden).
x 10'10 m/sec.
These increases in flow rate through the liner at high degrees of liner
deformation are likely to be the result of local changes in permeability
in regions of severe deformation; thus, the values of average permea-
bility are somewhat misleading. It is not possible at this stage to make
any statements about localized permeability changes, but clearly the
severe liner failure observed for the kaoline test TD10 is not evident
with this material. In fact, it could be argued that at such large defor-
mations the assumption of a smooth profile of differential settlements
at the base of the liner is no longer applicable, since deformations within
the underlying sand will have localized into thin bands of intensely
shearing material. The interaction of these shear planes with the base
of the liner will generate local discontinuities of both slope and
displacement.
DISCUSSION
The tests reported in this paper were performed to investigate the
effect of overburden and choice of liner material on the response of
a model liner to imposed deformations.
The liner response was significantly different in the presence of an
overburden where no tension cracking was evident and the formation
of multiple shear surfaces was observed. The suppression of tension
cracking is explained by the increased lateral stresses generated within
the liner. Greater deformation (straining) of the liner therefore is pos-
sible before tensile stresses necessary for cracking to occur are
generated. However, before such stress levels are reached, localization
of deformation occurs with the formation of multiple shear ruptures
in regions of greatest liner deformation. Consequently, tensile stresses
do not arise and no tension cracking is observed once rupturing has
occurred. It is not possible at this stage to make any statements as to
when shear rupture occurs and what combination of overburden and
liner thickness is necessary to prevent tension cracking. The presence
of shear ruptures did not affect the performance of the kaoline model
liner as an effective hydraulic barrier. However, there is some evidence
that in the presence of large hydraulic gradients this is not necessarily
the case, and such ruptures could provide preferential flow paths8'9
reducing liner effectiveness.
Where no overburden was present, the growth of tension cracks in
regions of large deformation resulted in failure of the pure kaolin model
liners to function as effective hydraulic barriers. However, in liners of
greater thickness, the larger lateral stresses present in the lower depths
of the liner may also result in the onset of shear rupture rather than
tension cracking—as argued above for the case of an overburden. In
such instances, a liner 'failure' would arise from the creation of preferen-
tial flow paths consisting of a combination of shear rupture and tension
cracking. This was thought to be the case in test TD8 (see Fig. 5a for
interpretation).
Comparison of the fine/coarse mixture and pure kaolin model liners
illustrates a much greater capacity of the fine/coarse mixture liner
material to function effectively when subjected to even large deforma-
tions. The reasons for this response are not entirely clear, but the
following interpretation is suggested. First, the absence of significant
tension cracking suggests that the material possesses a very small
cohesive strength and, hence, large unsupported tension cracks cannot
appear, i.e., the response to deformation of the material is as might
be expected for a sand. Second, the very low permeability of the
material, which is derived from the nature and size distribution of the
fine fraction filling the voids of the sand, is maintained under imposed
deformation by the ability of this fine fraction—which would behave
like a slurry of zero effective strength—to flow within the sand matrix.
Hence, the material would exhibit an extremely quick and efficient self-
healing property.
CONCLUSIONS
The ability of the centrifuge to induce prototype stress levels within
a small scale model allows the stress dependent response of the model
to be directly interpreted to the corresponding prototype situation. Thus,
from the model tests presented in this paper, where model liners have
BARRIERS 541
-------
been subjected to deformations of a similar nature to those that might
occur in the field, the following conclusions can be drawn:
• Tension cracking and rupture are likely responses for pure compacted
clay liners. The dominant mode of liner response will be dependent
on the lateral stress level. For example, small lateral stresses will
favor the development of tension cracking, whereas larger lateral
stresses will promote localization of deformation into shear ruptures.
• The development of severe tension cracks can lead to failure of the
liner to function as an effective hydraulic barrier.
• The presence of ruptures alone is unlikely to affect the satisfactory
performance of the liner in a prototype situation.
• Liners manufactured from fine/coarse (sand/clay) mixtures may pro-
vide hydraulic barriers virtually unaffected by likely prototype defor-
mations. This design is thought to be due to a highly efficient
self-healing system.
REFERENCES
1. Kleppc, J.H and Olson, R.E.. "Desiication Cracking of Soil Barncm." m
Hydraulic Barriers in Soil and Rock, ed. A.I. Johnson. R K Frobel. M J
Cavalli. C.B. Peaenon. ASTM STP 874. Philadelphia, P*, pp. 263-275,1985.
2. Anderstend, QB. aiidAJ-MouMawi, H.M., "Cim* Formation to So« Undffl]
Coven due to Thermal Contraction," Wait Mono. Ra., 5, pp. 445-452,1987.
3. Sterling. M J and Ronayne, M.C., "Centrifugal Modelling of Subsidence
of Landfill Covers," Proc. Symp. on Keceni Advancei in Gtauchnlcal Cm-
rrifiige Modelling, University of California, Davis, CA. pp. 71-81, 1982.
4. Jessberger, H.L. and Thiel. G , Abtchaetzung dc* Seizungiverhaltem von
Deponie-BatiMbdichfiingeo—Benchnung und ModeOvenuch. In preparation.
5 Stone. KJ.L ani Wx>d. D.M., "Some Observation of FatibugmSoAQayC
Centrifuge '88. ed. IF. Cone, Balicema Publ , Rotterdam, 1988
6 Jessbcrger. H.L. and Guttler. U., Bochum Geotcchnical Centrifuge, Cen-
trifuge '88, ed. J.F Cone, Balkema Publ.. Rotterdam, 1988.
7, Arulanandon, K., Thompson, P.Y, Kuiter. B.L., Meegoda, NJ.,
Munucetharam, K K and Ybgadundran. C. "Centrifuge Modelling of Ite*
pon Proceues for Pollutants." / Geoiech Eng.. ASCE, 114 (2). pp. 185-205
1988
8 Jcifbergcr. H.L . Guttler. U. and Stone. KJL ."Centrifuge Modelling of
Subsidence Effects on Clay Barriers," paper submitted lo Sardinia '89. Second
International Landfill Symposium. 1989
9. Gronow. J.R., "Migration Pathways in Seabed Sediments." in Dispoialaf
radioactive Hbstet in Seabed Sediment!. ed. T J Freeman, puW. Grahao
and Trotman Ltd.. kmdon. pp. 179-199. 1988
542 BARRIERS
-------
Composite Liner System to Retain Inorganic
and Organic Contaminants
George Alther
Bentec, Inc.
Ferndale, Michigan
Jeffrey C. Evans, Ph.D., P.E.
Stephen Zarlinski
Bucknell University
Lewisburg, Pennsylvania
ABSTRACT
Hazardous waste disposal facilities presently rely on RCRA-required
liner systems, composed of two geomembrane liners overlying natural
or compacted clay, to impede the migration of contaminants into the
environment. Limitations to the present liner systems include the poten-
tial for construction defects, long-term changes in the properties of the
membrane and diffusion of organic contaminants through the other-
wise intact liners. The present design philosophy, simply stated, is to
design a liner system that does not leak.
Since liner systems cannot be perfectly constructed and diffusion
causes contaminant migration through liners, a composite liner system
is proposed which would further reduce the rate of contaminant migra-
tion into the environment. The design philosophy of this composite liner
approach is markedly different than the present design philosophy. While
present designers strive to minimize the hydraulic conductivity of the
system, it may be more prudent to acknowledge the potential for
contaminant transport across the barrier layers and design the system
to adsorb these contaminants.
The proposed composite liner starts with the same features as a con-
ventional RCRA liner system including two geomembrane liners, two
leachate collection systems and the appropriate drainage and filter layers.
However, the proposed composite also includes components to sorb
the contaminants in the leachate. A sequence in the liner system is also
proposed but this can be left to the individual designer's preference.
We propose that the liner include a layer of calcium bentonite and natural
zeolite attached to the bottom of the uppermost geomembrane. Cal-
cium bentonite and zeolite will preferentially adsorb and filter inor-
ganic species which may penetrate this liner. We also propose that the
liner include a layer of organically-modified clay attached to the lower-
most geomembrane. This liner composite will adsorb organics which
would otherwise migrate into the environment. Performance data are
presented to support these concepts.
INTRODUCTION
The present philosophy for the design of liner systems is to reduce
the hydraulic transport of contaminants through the system. The design
strategy recognizes the potential for imperfections and accommodates
these through redundancy and through quality control measures. The
present design philosophy does not, however, explicitly acknowledge
diffusion as a contaminant transport mechanism. Research has shown
that contaminant transport in response to diffusion gradients may be
significant. We propose that liner systems be designed as composites
which include sorption layers as well as barriers to hydraulic transport.
In this way, a mechanism is provided through which the liner system
can minimize the rate of inorganic and organic contaminant transport
into the environment in response to both hydraulic and diffusion
gradients.
Described in the following paper are several proposed materials that
have demonstrated the capability to adsorb contaminants. Calcium
bentonite and natural zeolite are suitable for inorganic constituents, and
organically-modified clays are suitable for organic constituents. In
addition, high carbon fly ash may be considered for both organics and
inorganics.
PROPOSED COMPOSITE LINER
Liner systems presently employ two geomembrane barrier layers
overlying natural or compacted clay as shown in Figure 1. This system
is subject to constraints including:
• A clay source of adequate quality and quantity within an economic
distance
• Compaction to remove natural defects (preferential contaminant
migration pathways) such as root or desiccation cracks
• Quality control to prevent punctures in the synthetic liner not be
during construction or during waste disposal
• No diffusion of contaminants occurs through the synthetic liner
;RBC__-iSS=ijiiiiii_••mmffl *'
Figure 1
RCRA Liner System
Each of these constraints can be addressed with the present design
approach to some extent. Clays of adequate quality and quantity are
available, although costs may be quite high. Preferential contaminant
migration pathways can be removed from the underlying clay layer, with
difficulty, by mechanical means. Geomembrane liner's defects have been
a problem commonly associated with synthetic liners. The final item
presents the greatest difficulty for the present liner systems as diffu-
sion through geomembranes does occur'. It also has been shown that,
as chemicals at the molecular level penetrate geomembranes, synthet-
ic liners probably deteriorate2. When geomembranes are exposed to
chemicals over a long period of time, thickness, crystallinity and molecu-
lar structure are affected by chemical and thermal changes in the im-
BARRIERS 543
-------
mediate environment. A landfill is a chemical and biological
environment in which chemical and biological degradation of the con-
tents takes place over time. A composite liner system must be designed
to accommodate these processes to achieve a maximum life span com-
mensurate with the predicted time rate of chemical degradation.
A number of modifications to the conventional liner system are pro-
posed in this paper to improve the performance of landfill liner systems
for both inorganic and organic constituents (Fig. 2). The liner system
performance can be enhanced using a series of natural and chemically-
altered natural materials to create an improved composite liner system
Figure 2
Proposed Composite Liner System
As shown in Figure 2. a layer of calcium bentonite and natural zeolite
beneath the primary geomembrane liner is proposed. These two mineral!.
effectively adsorbed any heavy metals which have not been removed
by the leachate collection system and have passed through the primary
geomembrane barrier. The thickness of this natural material layer can
be based on the required attenuation capacity. The layer may be rela-
tively thin because the ion exchange capacity of both minerals is high
(i.e.. over 60 meq/100 g).
These materials underlie the primary geomembranc barrier and may
be attached to the geomembrane consistent with presently available tech-
nology (the Paraseal System from Paramount Co.. Spcarfish, South
Dakota). Alternatively, the added materials may be prepared as separate
layers and serve as bedding for the primary geomembrane. In either
instance, the sorptive layer will attenuate the rate of inorganic mass
transport through the liner system.
The proposed composite liner is also equipped to attenuate the rate
of mass transport of organic contaminants. This attenuation process is
accomplished with the inclusion of a layer of organically-modified clay.
The organically-modified clay is designed to adsorb organic compounds
that may have diffused through the liner. The organically modified clay
may be attached on the bottom side of the secondary geomembrane.
This liner may rest on a soft soil or clay to assure no damage to it.
In this fashion, organics which may diffuse through the liner will be
sorbed onto the organically-modified clays. Alternatively, the
organically-modified clay may be attached to the upper side of the secon-
dary geomembrane. In this fashion, the organically modified clay will
protect the geomembranc from degradation as well a«> reduce con-
taminant transport across the layer
The inclusion of a layer of carbon-rich fly ash also has been consi-
dered. Alkaline fly ash could increase the pH of the Icachale, which
would in turn cause precipitation of heavy metals. The layer also could
contain an inexpensive reducing agent, such as reduced iron powder
or graphite, to reduce hcxavalent chromium to mvalcni chromium to
enhance precipitation. The fly ash should contain at least 10% carbon
and be free of phenolics. This high carbon fly ash would then efficiently
absorb low molecular weight organics'.
This proposed system maximizes retention of contaminants through
sorption processes. This new liner system contrasts to designs which
focus on hydraulic transport mechanisms. The clays and zeolites will
sorb metals and the environment is protected by the organically-modified
clay against diffusing organic chemical transport. The remainder of this
paper describes the individual components in detail and discusses some
test results which support the composite liner concepts.
MATERIALS DESCRIPTION
Calcium Bentonite
Calcium bentonites arc non-gelling bentonites containing montmoril-
lonite as their major constituent. This montmorillonite has, as its
primary exchangeable cation, calcium (typically from 35 to 99% of
the exchangeable cations). Calcium bentonites, such as those from Mis-
sissippi and Alabama, have ion exchange capacities of more than 70
meq/100 g.
Calcium bentonites are in abundance worldwide*. These clays ex-
pand little upon wetting with water a desirable characteristic since their
function is as an ion exchange barrier rather than as a water barrier.
Calcium bentonite transmits water, where the secondary leachate col-
lection system can collect it. The calcium ions on the surface of this
clay exchange readily with heavy metals such as lead, copper, nickel
and others, with strong preference for lead'.
Zeolite
A natural zeolite (of a 200 mesh grain size) can be blended with the
calcium bemonitc. or it can be placed as a separate layer to enhance
the sorption of inorganic species. Placement of the zeolites as a separate
layer would prevent possible exchange reactions of the calcium from
the bentonite with the sodium from the zeolite. Natural zeolites are also
hydrous aluminum silicates, like monlmorillonite. but with an entirely
different crystalline lattice arrangement. Zeolites are crystalline-
hydrated, three-dimensional aluminosilicates with a cage structure.
wherein the exchangeable ions are sorbed. Its exchange capacity is up
to 250 meq/g. The major exchangeable ions are sodium and calcium.
The zeolite acts like a sieve when permeated by water, trapping metals
that pass through by ion exchange. The zeolite would complement the
calcium bentonite to assure maximum adsorption of metals.
There arc five major zeolite classes including: Analcime. Chabaziie.
Clinoptilolite. Enonite and Mordenite. Zeolites presently are used to
purify water contaminated with radioactive cesium and other ions (e.g.
Three Mile Island), remove ammonium from wastewater treatment plant
effluent and fish tanks, control odor cat litter and adsorb metals from
industrial wastewaiers in columns.
Arsenic, lead and cadmium are preferentially adsorbed by zeolites.
A study by the Montana College of Mineral Science and Technology*
showed that a chabazite from Bowie County. Arizona was an excellent
adsorption media for metals in groundwater. The structure and behavior
of zeolites have been studied extensively7 but further details are
beyond the scope of this paper
Organically-Modified Clays
Organically-modified clays were developed in the 1940s. Jordan1*
published the first original vvork on the subject. Organoclays are further
described by Alther. et al.," and Evans, et al." In summary, the in-
organic cation on the surface of a clay such as a montmorillonite, is
exchanged with a suitable organic cation, preferentially with an anune
of a composed chain length of at least 12 carbons*. The important
reactions that take place in this process are adsorption, ion exchange
and intercalation.
According to Mortland". an organically-modified clay adsorbs or-
ganic molecules due to two controlling factors: (1) adsorbent-adsorbate
interactions and (2) adsorbate-solvent interactions. The adsorption
capacity of an organically-modified clay is thus dependent on the amine
used to convert the clay and the properties of the medium, such as its
temperature, pH and type of solvent. Additional insight into the behavior
of organically-modified clays is provided by Boyd, et al.°, Mortland.
et al.';. and Wolf, et ul.H. These authors and others have designed
organically-modified clays by exchanging onto them dioctadecyl
dimethyl, hcxadecyltrimethyl ammonium chloride (very hydrophobic
in nature), the less hydrophobic tetramethyl ammonium chloride and
hexadecyl pyridinium compounds. Boyd, ct al.". and Mortland, et
al.'1, have shown those clays (the ones modified with strongly
hydrophobic organic molecules) to be excellent adsorbents of chloro-
phenols, while not suitable for straight phenols.
544 BARRIERS
-------
Wolfe, et al.'4, tested the efficiency of three organically-modified
clays as to their efficiency (versus activated carbon) in the removal of
11 organic compounds from water, with encouraging results. Boyd, et
al.13, have shown that, as hydrophobicity of the sorbate increases,
sorption by the organically-modified clay increases while the reaction
of primary amines with other molecules is pH-dependent. This is not
the case for quaternary amines, making them more suitable sorbents
for water decontamination.
The entire process of sorption of organics onto organically-modified
clays is described as a partitioning process15'16. Boyd, et al.13, describe
the process such that the quaternary ammonium ion attached to the clay
surface acts as a solubilizing (partitioning) medium to remove organic
molecules from water, being functionally and conceptually similar to
a bulk organic solvent such as hexane or octanol13 They used the
hexadecyl trimethyl ammonium chloride ion (16 carbons long) to re-
move non-ionic organic contaminants such as benzene, dichloroben-
zene and perchloroethane from water, with good success. The evidence
points to the effectiveness of organically-modified clays for the removal
of such toxic compounds.
Fly Ash
A description of the type of fly ash properties best suited for landfill
liners desired for this application is found in Mott and Weber13. The
properties of the fly ash from Trenton, Michigan are shown on Table 1.
These authors showed that molecular diffusion of organics through
soil/bentonite backfills in slurry walls result in solute breakthrough
within a relatively short time. The addition of high-carbon fly ash within
the barrier caused a substantial delay in breakthrough.
Table 1
Fly Ash Properties
Density (g/cm3)
Loss on Ignition (%)
Carbon1 (%)
Phenol2
Tannin2
Specific Surface (m2/g)3
2.29
9.14
6.14
nd
951
2.65
1 Measured as C(>2 recovered during wet combustion
2 Normalized to the carbon fraction of fly ash
3 Primary surface Area by B.E.T. nitrogen adsorption
LABORATORY TESTING
Laboratory studies were undertaken to further demonstrate the via-
bility of the proposed composite liner concepts. Samples of sorptive
media were compacted into cylindrical specimens for permeation in
a triaxial cell permeameter. Samples were permeated with distilled water
spiked with several inorganic and organic species including copper, lead
and nickel at concentrations of 10 mg/L. Samples of the influent and
effluent were taken throughout the permeation period and analyzed for
the spiked constituents.
The results are shown in Figures 3 through 5. As shown, the calcium
bentonite, chabazite (a natural zeolite) and a naturally occurring silty
clay all attenuate the contaminant concentrations. In fact, breakthrough
does not occur for the lead and nickel even after a pore volume
displacement of 12.
This study also used a commercially available organoclay (PT-1,
Bentec, Inc., Ferndale, Michigan) which is modified with dimethyl di-
hydrogenated tallow ammonium chloride, an 18-carbon alkyl ammonium
molecule. As expected, the organically-modified clay had little effect
upon the sorption of inorganic species. Analytical data on the concen-
1.0
0.9 -
OJ -
0.7 -
0.6
0.5
0.4 •
0.3 -
0.2 -
0.1 -
0.0
Copper
Pore Volume Displacement
+ Load o Nickel
Figure 3
Permeation of Silty Clay
Pore Volume Displacement
a Copper + Lead » Nickel
Figure 4
Permiation of Calcium Bentonite
D Copper
Pore Volume Displacement
+ Lead » Nickel
Figure 5
Permeation of Chabazite
trations of the organic species were not available at the time of publi-
cation but are expected to demonstrate the adsorption capacity of the
organically-modified clay for organic species.
BARRIERS 545
-------
Results for the fly ash were less conclusive. Although some inor-
ganic ions were adsorbed from the influent, copper concentrations sig-
nificantly increased in the effluent. Further studies investigating
alternative fly ash sources and types are needed to justify the use of
a fly ash layer within the composite liner system
CONCLUSIONS
An alternative philosophy for the design of secure landfill liner systems
is proposed. Whereas designers presently focus on the need for low
hydraulic conductivity, we propose that designers recognize that liners
may leak and contaminants will migrate across the barrier layers due
to the hydraulic imperfections and molecular diffusion. With this recog-
nition of contaminant transport across the barriers, comes the need to
include adsorptive layers in the liner system. Thus, a composite liner
would use technologies presently employed to reduce the hydraulic con-
ductivity (such as geomembranes) coupled with media to sorb con-
taminants. Adsorption media include calcium bentonites and natural
zeolites for inorganic species and organically-modified clays for organic
species.
REFERENCES
I. Haxo, H. E . Haxo, R S . Nelson. N. A . Max... P. D. While. R M ,
Dakessian, S. and Fong. M. A . Liner Materials for Hazardous and Toxic
Wastt and Municipal Solid Hbsir Leachair, Noyes Publications, Park Ridge.
NJ, 1985.
2. Ycrschoor, K.. Brinon. L. and Thomas. R . "Chemical Compatibility Testing
of Geosynthetics to be Used as Containment Barriers in Hazardous Waste
Management." Ha; Htisie/Ha: Mai, S, pp. 205-209, 1988.
3. Moct. H. V. and Weber, W. J , "Diffusive Transport and Attenuation of
Organic Leachaies in Cut-Off Wall Backfill Mixes," Presented at the Twelfth
Annual Madison Waste Conference, Department of Engineering Professional
Development. University of Wisconsin. Madison. WI. Sept.. 1989.
4. Grim, R. E. and Gucven, N., Bentoniiei, Development* in Sedimentology,
Elsevicr Scientific. New Vbrk. NY. 978. 256 pp., 1978.
5 International Minerals and Chemical Corporation, Laboratory Report, 1986.
6. Montana College of Mineral Science and Technology, Open File Report. 1989.
7. Mumpton, F. A.. Mineralogy and Geology of Zeolites, Mineralogical So-
ciety of America, Short Course Notes, Southern Printing Co., Blacluburg,
W. 4. 233 pp., 1977.
8. Jordan. J. W., "Alteration of the Properties of Bentonile by Reaction with
Amines." Mineralogical Magazine and J. afihe Mtneralogual Soc, 28,205,
pp. 598-605 June 1949
9 Jordan. J. W., "Organophiuc Bentonjics Swelling inOrganic Liquids." Phys.
& Colloid Chem.. S3, (2). pp. 294-306. 1949
K). Althcr. G.R . Evans. 1C and Pancofki, S.E., "Organically-modified Clays
for Stabilization of Organic Hazardous Waste*," Superfund '88. Proceeding!
of the Ninth National Conference. Hazardous Materials Control Research
Institute. Silver Spring. MD, pp. 440-445. November 1988.
II Evans, J.C.. Pancotki. S.E. and Allher, G.R., "Organic Wane Treatment
with Organically-modified Clays," Third Imemaltonal Conference on New
Fronlien for Hazardous Waste Management. (Submitted May, 1989).
12. Mortland. M.M . Shaooai. S and Boyd, S A . "Clay-OrgankComplexet
as Adsorbents for Phenol and Chlorophenols." Clays and day Minerati,
34. (5), pp. 581-585, 1986.
13. Boyd. S. A . Lee. J-F and Mortland. M. M 'Attenuating Organic Con
uminam Mobility by Soil Modification." Naiurt. 333, pp. 345-347. May.
1988
14. Wolf. T. A.. Dcmirel.T and Bauman, R.E.. "Adsorption of Organic Pollu-
tants on Mommonlloniie Treated with Amines." JWPCF. 58 (1). pp. 68-%
1986.
15. Chiou, C T. Porter. P E. and Schmedding. D. W., "Partition equilibria
of Nonionic Organic compounds between Soil Organic Matter and Water."
Envir. Sri. Trchnol. 17(5}. pp 227-231, 1983
16. Chiou. C. T. Peten. L. J. and Freed. V. H.. "A Physical Concept of Soil-
Water Equilibria for Nonionic Organic Compounds." Science. 213, pp.
684-685
546 BARRIERS
-------
Recent Advances in Asbestos Assessment at
Superfund Sites
Paul C Chrostowski, Ph.D.
Sarah A. Foster
Clement Associates, Inc.
Fairfax, Virginia
INTRODUCTION
As of August, 1989, asbestos has been identified as a chemical of
concern at 22 Superfund sites. At many of these sites, it is the sole
chemical of concern. The sources of the asbestos are quite variable.
On one hand, the source may be mining and/or milling activities or
emissions from natural deposits (Atlas/Coalinga, California, Globe,
Arizona). Other sites consist primarily of material disposed improperly
from secondary processing (Ambler, Pennsylvania). At still other sites,
the source of the asbestos is tailings which have been put to a supposedly
beneficial use (South Bay/Alviso, California). In addition to the sites
where asbestos is the sole chemical of concern, there are numerous
other sites where it occurs with other toxic chemicals. These sites con-
taining mixed wastes range from abandoned hazardous waste disposal
sites (Hardage, Oaklahoma) to sites were asbestos was used as a struc-
tural or insulation material (Sharon Steel, Utah).
Due to its unique nature, asbestos presents some special problems
in its analysis, remediation and risk assessment. Specialized analytical
concerns regarding asbestos at Superfund sites have been dealt with
in the literature for asbestos in air1 and in soil2. The U.S. EPA has
produced a Health Effects Assessment3 and a review of the health
effects of asbestos4 which incorporates risk-based information.
Additionally, ATSDR is in the process of producing a toxicological
profile for asbestos. However, a methodology has not been developed
for asbestos risk assessments at hazardous waste sites which parallels
the methods developed by the U.S. EPA for chemicals5. The develop-
ment of this type of methodology has been hampered by many questions
of scientific controversy which surround asbestos. These unanswered
questions relate to the definition of biologically active fibers, the shape
of the dose response curve, the relevance of analytical measurements
to risk assessment techniques, the problem of route-specific carcinoge-
nicity and the presence of ubiquitous background concentrations at levels
associated with relatively high cancer risk.
The U.S. EPA has conducted considerable research under regulato-
ry programs other than Superfund such as AHERA6 and the Phase-
down rule7. Other regulatory agencies both within the United States8
and abroad' also have developed programs for asbestos management.
The purpose of this paper is to present a methodology for asbestos risk
assessment which we have synthesized from our activities at three Su-
perfund sites in addition to risk and exposure assessments performed
under AHERA.
A RISK BASED DEFINITION OF ASBESTOS
Asbestos is a generic term referring to a family of naturally occurring
silicates having a fibrous crystalline structure. There are six fibrous
silicates defined as asbestos types: chrysotile, actinolite, amosite,
anthophyllite, crocidolite and tremolite. Of these six silicates, only
chrysotile is typically detected at Superfund sites.
Deposition and absorption of asbestos fibers can be influenced by
fiber characteristics such as fiber length, fiber diameter, aspect ratio
(ratio of length to diameter), fiber number, stability of fibers in the
body, surface chemistry of the fiber, interactions between fibers and
other surfaces, fiber translocation and migration, overall fiber dose and
fiber type10. Specific data relating individual asbestos type and
physical characteristics of the fiber with biological activity via ingestion
are lacking.
Following inhalation, there is some evidence to suggest a relation-
ship between asbestos fiber dimension and carcinogenic potential. This
relationship is known as the Stanton Hypothesis and is based on corre-
lations between pleural sarcomas in rats and dimensions of fibers in
addition to human epidemiologic data". Long, thin fibers (> 5 ^ in
length, aspect ratio >3) appear to elicit the greatest biological response.
However, a critical fiber length below which there would be no car-
cinogenic activity has not been demonstrated. Fibers less than 5 fi in
length appear to be capable of producing mesothelioma4, and the
results of one analysis show that carcinogenicity appears to be a con-
tinuously increasing function of the aspect ratio12.
A re-analysis of Stanton's original data13 concludes that factors other
than size and shape may play a role in asbestos carcinogenicity. There-
fore, for purposes of risk assessment, all asbestos fibers will be consi-
dered to be carcinogenic, although direct preparation total TEM fiber
counts (where available) must be converted to PCM equivalents for pur-
poses of using human health data derived from epidemiologic studies
in which exposure was measured by PCM.
HAZARD IDENTIFICATION/
DOSE-RESPONSE QUANTIFICATION
The primary non-carcinogenic health effect of asbestos is asbestosis,
a chronic lung disease associated with function disabilities and early
mortality; however, development of asbestosis usually is associated only
with high-level occupational exposure4. For low-level environmental
exposure, cancer is considered a more appropriate endpoint for criteria
development than asbestosis.
The carcinogenicity of asbestos following ingestion has not been con-
clusively demonstrated by direct studies. In a National Toxicology
Program (NTP)'4 bioassay in male rats, a significant increase in
benign epithelial neoplasms in the large intestine was interpreted as
limited evidence that orally ingested chrysotile fibers may be carcino-
genic. Available data from occupational studies also suggest a link
between inhalation and subsequent ingestion of asbestos and gastro-
intestinal cancer4.
The U.S. EPA15 developed an oral unit risk factor of 1.4xlO~'3
(fiber/liter)-' based on the NTP bioassay in which benign neoplasms
BARRIERS 547
-------
were observed in male rats exposed to asbestos (> 10 n in length) in
drinking water; this cancer potency factor was used by the U.S. EPA
as the basis for the drinking water maximum contaminant level goal.
There are a number of uncertainties associated with this approach
including the absence of adequate dose-response data from human popu-
lations exposed via ingest ion, the induction of benign tumor, only and
the fact that the criterion is limited to fibers greater than 10 n in length.
Inhalation exposure in humans and experimental animals can result
in both lung cancer and mesothelioma. The calculation of risk for
inhalation exposure is based primarily on the methodology set forth
in the U.S. EPA's Airborne Asbestos Heallh Effects Update1. This cal-
culation relies on the use of risk tables which give maximum likelihood
estimates for mesothelioma and lung cancer as a function of sex, age
at onset of exposure, years of exposure and ambient atmospheric con-
centrations. Situations not exactly described by the risk tables arc
evaluated by linear interpolation among values on the tables.
Since the tabulated function is non-linear at high concentration.-!, for
situations when concentrations are substantially higher than those shown
on the tables, the resulting risks presented should not exceed the maxi-
mum risk shown in the health risk table (3x10 '). but should rather
be listed as ">3xlO\ For situations below the range of the tables
(e.g.. exposure periods of less than 1 yr, concentrations substantially
lower than K) ' to 10" fiber/cm'), the linearized unit risk of
2.3186x10 ' (fibers/cm1) ' developed by the U.S. EPA" for a lifetime
exposure may be used. The risk calculated by this method must be
adjusted to compensate for less than lifetime exposure.
APPLICABLE OR RELEVANT AND
APPROPRIATE REQUIREMENTS
The U.S. EPA's interim guidance" defines ARARs as follows.
Applicable requirements means those cleanup standards, standards
of control or other environmental protection requirements, criteria or
limitations which are promulgated under Federal or State law and
specifically address a hazardous substance or other circumstance at a
CERCLA site. On the other hand relevant and appropriate requirements
address situations which are sufficiently similar to a CERCLA site that
their use is well suited to the particular site or situation.
Federal regulatory action on asbestos has taken on a variety of forms.
Regulations have been promulgated by numerous agencies including
the US EPA. the Consumer Product Safety Commission (CPSC), the
Department of Transportation (DOT), the Food and Drug Administra-
tion (FDA), the Mine Safety and Health Administration (MSHA) and
the Occupational Satiety and Health Administration (OSHA). Although
none of the promulgated regulations may be applicable to a specific
asbestos site, they may be relevant and appropriate.
In 1980, the National Institute for Occupational Safety and Heallh
(NIOSH) recommended to OSHA a maximum level for asbestos in the
workplace in addition to several measures which would act to minimize
exposure. Recently, OSHA issued a rule implementing many of these
regulations and lowering the old workplace standard of 2 f/cm' to 0.2
f/cm' of air as an 8-hr time weighted average (51 FR 22612, 1986). As-
bestos was, first designated by the U.S. EPA as a hazardous air pollu-
tant under the Clean Air Act in 1971. Since their initial promulgation
in 1973, the National Emissions Standards for Hazardous Air Pollu-
tants (NESHAP) for asbestos have been revised several times. As they
currently read, the regulations call for no visible emissions from mill-
ing, manufacturing and asbestos waste disposal activities (43 FR 26372.
1977) and require that asbestos containing waste be kept thoroughly
wet with water during handling. The standard of no visible emissions
may be relevant and appropriate to the asbestos soils and wastes at
hazardous waste sites.
Asbestos was first determined to be a hazardous water pollutant in
1973. Effluent limitation guidelines for asbestos manufacturing have been
promulgated (40 FR 1874, 1975). Also under the Clean Water Act, the
U.S. EPA published an Ambient Water Quality Criteria document for
asbestos1*. The document noted that data were inadequate to issue
criteria for the protection of aquatic life. For protection of human health,
the estimated levels of asbestos in water which would result in increased
lifetime cancer risks of 10 \ 10 ' and 10 ' are 300,000 total fibers/L,
30JOOO total fibere/L, and 3000 total fibers/liter, respectively. The U.S.
EPA also has proposed an MCLG for asbestos in drinking water under
the Safe Drinking Water Act which equaled an excess cancer risk of
10 "• to an asbestos concentration of 7.1xH) * fibers greater than K) p
in length/L"
Another significant regulation is the Asbestos Hazard Emergency
Response Act (AHERA) which enacted Title II of TCSA. Under this
act, the U.S. EPA has published regulations related to inspection and
management of friable asbestos in schools (52 FR 42826, 1987). The
monitoring procedures specified in this regulation may be relevant to
hazardous waste sites. In the rule, local agencies must consider an area
to contain asbestos if asbestos fibers are found in any sample at greater
than 1% as analyzed by PLM. Since 1% generally is accepted as the
detection limit for asbestos in soil, it could be argued that this represents
a standard for non-detectable asbestos, i.e., thai any delectable asbestos
in soil warrants action.
EXPOSURE ASSESSMENT
There are two general routes through which individuals may be
exposed to asbestos at a Superfund site: inhalation and ingestion.
Experimental and epidemiologicaJ studies indicate that inhalation
exposures to asbestos are of greatest potential concern to human health.
Exposure to asbestos via ingestion. although not considered to be at
important lexicologically as inhalation, may also be associated with
an increased risk of neoplasms. These ingestion exposures include direct
ingestion of contaminated soil, direct ingcsuon of contaminated surface
water and indirect ingestion of asbestos which has been inhaled. Dermal
contact and subsequent absorption is not an exposure route of concern
since asbestos is not likely to be absorbed through intact skin.
Although a thin crust usually forms on the asbestos material in mining
tailings piles, mill tailings piles, soils, etc. after rainstorms, this crust
is easily disturbed with the net result that the asbestos materials become
friable and may be eroded by high winds. Persons in downwind off-
site residential areas may be exposed to ambient airborne asbestos. In
addition, individuals who trespass or engage in recreational activities
in the vicinity of a site may inhale ambient air concentrations of asbestos
on-siie. Ambient concentrations of airborne asbestos are usually detected
in various on-site and off-site areas by air monitoring conducted during
an RI; therefore, this exposure pathway is considered complete.
Problems with sampling and analysis of airborne asbestos, however.
often make airborne data difficult to use for risk assessment. At one
site, for example, air samples were collected both in summer and winter
and both during the day and at night in order to account for the complex.
changing nature of meteorological conditions affecting air concentra-
tions at different times of the year. A threshold wind velocity is neces-
sary to cause soil entrainment of asbestos; however, air monitoring
samples were collected during periods when this threshold velocity was
exceeded. Therefore, using the available air monitoring data and the
site-specific meteorological monitoring data collected during an RI.
modeling was conducted to determine annual average air concentra-
tions of asbestos at various locations. Annual average air concentra-
tions of asbestos were determined for each sampling location based on
the air monitoring concentrations and modeling which accounted for
periods when the threshold velocity was exceeded.
A further limitation of air monitoring concerns the presence of more
than one potential source of airborne contamination. If more than one
site is contributing to airborne contaminant levels, ambient air
monitoring data often do not allow determination of the fraction of
measured concentrations associated with each source. The concen-
trations used to estimate exposures in a risk assessment should in theory
reflect site-specific conditions, but a wide variety of potential off-site
sources of asbestos can contribute to background asbestos concentra-
tions in air. Air monitoring for asbestos often does not allow dif-
ferentiation between these background ambient asbestos levels, whether
naturally occurring or anthropogenic. It is also difficult to distinguish
between primary sources of asbestos (e.g., the mining and mill tailings
piles on the sites) and secondary sources of asbestos (e.g., sedimentary
deposits of asbestos which have been transported downstream during
rain storms and subsequently become dry and subject to wind erosion).
548 BARRIERS
-------
For many sites, it is necessary to conduct modeling which estimates
the fractional contribution of the different sources to calculate annual
average air concentrations of asbestos in various locations.
Although of short duration compared to inhalation of contaminated
ambient air, asbestos air concentrations resulting from activities which
disturb contaminated source materials may be elevated above ambient
levels by several orders of magnitude"- Thus, in addition to chronic
exposures to lower ambient asbestos levels, activity-related exposures
of the individuals participating in the activity to airborne asbestos could
potentially result in cumulative asbestos exposures of concern to human
health if they occur with sufficient frequency.
For a typical non-residential mining or milling site, on-site activities
which could potentially generate increased levels of airborne asbestos
(i.e., above ambient) include off-road vehicle traffic (e.g., motorcycles,
cars and trucks) on asbestos piles and on jeep trails, horseback riding,
hiking, camping and hunting. Vehicle traffic and horseback riding are
expected to stir up greater quantities of dust than are hiking, camping
and hunting. In addition, hikers, campers and hunters are not expected
to spend extensive periods of time on the asbestos piles present at the
sites. Therefore, only inhalation exposures related to vehicle traffic and
horseback riding activities usually need to be assessed for the activity-
related air exposures.
For a typical site involving secondary contamination, activities which
could potentially generate airborne asbestos include vehicle traffic on
dirt roads and agricultural tilling of a contaminated area.
Air monitoring is not usually designed to measure exposure to point
concentrations of airborne asbestos for on-site and off-site activities.
These exposures are, instead, estimated using a combination of results
from empirical experiments in addition to emission and air dispersion
models. An emission model is required to predict the rate of release
of the contaminant from the site into the air. The dispersion model uses
the estimated emission rate to predict concentrations of a contaminant
in air around the source.
Emissions of this type may be evaluated by using the results of the
U.S. EPA's Environmental Asbestos Roads Study21. The results of
some experiments performed by California Department of Health
Services19 at the South Bay/Alviso Superfund Site may be used to
evaluate individual exposures during some routine activities which
involve soil disturbance on a small scale. Due to the lack of quality
assurance for these activity studies, the experiments can be used only
to yield an order-of-magnitude (or qualitative) estimate.
Efforts to correlate paniculate matter emissions with asbestos emis-
sions have been undertaken in hopes that fugitive dust emission models
could be adapted for estimating ambient concentrations of asbestos.
Addison, et al.,22 have shown, for instance, that trace amounts of
asbestos in soils (e.g., 0.001%) may yield airborne asbestos concentra-
tions greater than 0.001 fibers/mL, and that a 1% concentration in soil
may yield up to 20 fibers/mL in air. In general, however, current ef-
forts have shown that airborne concentrations of particulate matter do
not correlate in a consistent manner with asbestos concentrations, and
the U.S. EPA has not developed emission factors specifically for the
release of asbestos from soil. The emission models mentioned above
have, however, been developed to characterize releases of fugitive dust
(soil) from exposed sites due to mechanical disturbances (e.g., vehicle
traffic and agricultural tilling)20. The U.S. EPA considers their air pol-
lution manual, AP-42, emission factors for total suspended particulate
matter to be a reasonable approach for risk assessments of activity-
related exposures providing that the asbestos fibers are relatively short.
The fact that TEM fiber counts are much greater than PLM counts
at most chrysotile asbestos sites is an indication of short fiber lengths.
The U.S. EPA also has provided precedent in using a box model for
asbestos decision-making purposes.
It should be noted that for usual activity modeling, the units for the
soil asbestos concentrations are expressed in PLM area percent, which
were assumed to be equal to PLM weight percent. These soil concen-
trations are applied to the air emissions modeling, and air concentra-
tions of asbestos in units of PLM ug/m3 were calculated. In order to
assess the potential human health impact of inhaling airborne asbestos
fibers, air concentration units must be PCM fibers/cm3. Therefore, the
mass of fibers reported as ug by PLM may be assumed to be equal
to mass in units of ug by TEM analysis and a conversion factor of 30
TEM ug/m3 = 1 PCM fiber/cm3 may be applied to obtain air concen-
trations in PCM fiber/cm3.
As mentioned above, a few research efforts have been undertaken
in an attempt to determine the impact of some routine activities (e.g.,
playing and gardening) which involve soil disturbance on a small scale.
These studies have been conducted by the California Department of
Health Services (DHS)19 at the South Bay/Alviso Superfund Site. Four
general types of activities have been examined: (1) a worst-case scenario
in which asbestos contaminated soil was thrown in front of a fan and
air concentrations were measured 10 ft downwind at 30 in. above the
ground19; (2) vehicle scenarios in which a truck or car was driven
along an asbestos contaminated dirt road and samples were collected
upwind and downwind'"'; (3) a playing scenario in which a toy dump
truck was filled and emptied for 15 min. and personal air samples were
collected at 1 and 4 ft above ground"; and (4) a gardening scenario
in which loose dirt containing asbestos was turned over with a shovel
for 15 min. and personal air samples were collected again at 1 and 4 ft
above surface level'9.
The results of these activity-related experiments can provide an
indication of the potential air concentrations and resulting exposures
and risks that may be associated with similar activities conducted at
the residential areas located in the vicinity of Superfund sites. The
playing scenario was conducted in soil containing approximately 5%
asbestos (approximately 13% by TEM); breathing zone air concen-
trations were estimated to be roughly 1.7 NIOSH fibers/cm3 (equiva-
lent to PCM fibers/cm3). If activities similar to the simulated playing
scenario were to occur repeatedly (e.g., every other weekend for several
months of the year for a period of several years) among children playing
in the residential areas, cumulative asbestos exposures could result in
increased lifetime cancer risks exceeding one in one million.
The worst-case and vehicle scenarios for the California DHS experi-
ments resulted in greater impacts on air concentrations than did the
playing scenario. The worst-case scenario was conducted in soil con-
taining less than 1% chrysotile based on PLM and roughly 30% asbestos
based on TEM and resulted in air concentrations of approximately
200 f/cm3 by PCM. The most experimentally rigorous activity-related
experiment was a vehicle scenario conducted by the U.S. EPA21. In
this experiment, a car was driven back and forth along a 100-ft test
section of dirt road in California containing approximately 0 to 4%
asbestos, during which 1-hr and 8-hr air samples were taken. Two
upwind and four downwind air samples were collected. One-hour
median and maximum upwind air asbestos results were 0.01 and 0.09
PCM structure/cm3, respectively. One hour median downwind air as-
bestos sample concentrations varied from 0.08 to 0.21 PCM struc-
tures/cm3, depending on the distance of the sampling station from the
experiment location. One-hour maximum downwind air asbestos con-
centrations varied from 0.23 to 0.9 PCM structures/cm3. Since the
majority of soil sample results from the dirt road were less than 1%
(by wt), ratios of air concentrations resulting from the vehicle move-
ment to the soil concentrations cannot be calculated. From a qualita-
tive standpoint, however, these results indicate that vehicle use on
unpaved surfaces containing asbestos at elevated concentrations could
result in elevated air concentrations of potential concern to nearby resi-
dents or on-site workers.
Individuals may directly contact and subsequently inadvertently ingest
chemicals present in contaminated soil which may adhere to hands,
toys, tools, etc. Inadvertent ingestion of chemicals present in soil is most
likely to occur in young children, although exposures could possibly
occur among adults who engage in activities involving soil contact such
as farming or gardening. In asbestos risk assessment, two types of
exposures via soil ingestion typically may be evaluated; one involving
the lifetime exposure of local off-site residents and one involving the
intermittent exposure of adult hikers, campers and hunters to on-site
contaminated materials.
Estimating cancer risks for incidental ingestion of asbestos present
in soil is complicated because the U.S. EPA15 has developed a unit risk
factor for exposure to asbestos in surface water [1.4 x 10~13
BARRIERS 549
-------
(fibers/L) '] only and noi for exposure to asbestos from other environ-
mental media where concentrations may be reported on a mass (noi
fiber) basis. This unit risk factor was used as the basis of the proposed
Maximum Contaminant Level Goal (MCLG) for ingestion of water.
There are several important uncertainties associated with even this unit
risk factor, as discussed in the Hazard Identification section. In order
to quantify risks associated with incidental ingestion of asbestos in soil,
the U.S. ER\ unit risk factor was convened into a mass-based potency
factor (mg/kg/day) '
This conversion was done only for the purposes of providing a rough
indication of the potential excess lifetime cancer risks associated with
direct contact with asbestos in soil and subsequent incidental ingestion
at a site. The conversion factor from fibers to mass of asbestos was
taken to be 0.129x10' fiber per mg asbestos based on TEM drinking
water measurements performed at the Illinois Institute of Technology
Research" in conjunction with development of the proposed MCLG.
The convened asbestos potency factor can be multiplied by the CDI
to derive an approximate estimate of the excess lifetime cancer risks
associated with the specific exposure scenario.
RISK CHARACTERIZATION
Summaries of risks calculated by these methods arc presented for
two Superfund sites (Tables 1 and 2); background air (Table 3) and
public buildings (Table 4). For site A, upperbound cancer risks asso-
ciated with inhalation of ambient air on-site and downwind are less than
an order of magnitude different than the upwind station. These risks
are also similar to risks associated with inhalation of ambient air out-
doors and in public buildings. Risks associated with specific activities
are much higher, however and would be high enough to initiate
remediation based on EPAs risk range of 10 ' to 10~4. For site B, on
the other hand, risks associated with both inhalation by near site resi-
dents and casual recreational users exceed the U.S. EPA's risk range.
Activity related risks are also high. The highest ingestion risk at cither
of the sites only slightly exceeds the U.S. EPAs risk range. It should
be kept in mind, however, that this calculated value is derived from
data on benign neoplasms rather than malignancies.
Table l
Summary of Excess Individual Lifetime Cancer Risks for
Exposure to Asbestos Superfund Site A
Exposure
Average Case
Haiinun Case
Inhalation Ambient Air
Station I (off-site/up»1nd)
- wsotheliana 1E-06 - 2E-05
- lung cancer 3E-07 IE-OS
Stations 2-4 (on-site)
•MSOtheHona 6E-06 - IE-04
lung cancer JE-06 9E-OS
Station S (off-site/domiilnd)
- •esothellona 1E-06 3E-OS
lung cancer 4E-07 2E-05
6E-06
IE-OS
SC-OS
IE-OS
7E-06
2E-06
IE-OS
7E-05
1E-04
7E-04
2E-04
9E-05
Inhalation Activity Generated Airborne Asbestos (a)
Child Playing In:
Residential laUlngi Portion:
Non-Residential Tailings Portion!
Vacant Lots
Trailer Park Yards
2E-06 3E-OS
IE-OS - 2E-05
NC
NC
5[-06 1E-03
5E-06 - BE-05
1E-06 - 2E-05
2E-06 4C-04
Inhalation Activity Generated Alrbrone Asbestos (aj
Adult Gardening In:
City Yards NC
Trailer Park Yards NC
Inhalation - Activity Generated Airborne Asbestos (a)
Truck Traffic on Unpaved Surfaces
Residents Inhaling:
Oust from Truckyards NC
Dust frost Unpaved Roadways NC
Street Dust from Paved Roadways NC
Workers Inhaling:
Dust from Truckyards NC
Ingestion of Sol!
Children Playing on:
Residential Portions of Tailings 7[-06
Non-Residential Tailings Portions 2E-06
All Other Areas NC
8E-06
IE-OS
1E-03
2E-03
>3E-03
>3E-03
>3E-03
>3E-03
3E-04
1E-04
3E-05
Table 2
Estimated Excem Lifetime Cancer Risk* for
Exposure to Asbestos Superfund Site B
Resident
Average
Kaxfcu
Inhalation Residential Air (a)
Inhalation Recreational Users (a)
Inhalation - Off -Road Vehicles
- nesotheltotna
lung cancer
Inhalation - Agricultural Tilling
- atsothelioma
- lung cancer
Ingestion Recreational
Ingestion Residential
7E-04
ZE-04
5E-06
4E-07
SC-04
9E-04
6E-07
2E-06
1E-02
1E-03
ie-oz
6E-04
3E-02
4E-02
K-05
9E-05
(a) Hesotheliona and lung cancer combined
Table 3
Summary al Ambient Air Asbestos Concentrations and
AsMdattd Excess Lifetime Cancer Risks
loul
(««• K«f-
tMo Cmv
luirurlr COIpMltcs 0* S IO J
14-hour US Mopbjf
MM. ll«l>ohw «•«
MMI
Ouoriorlr envoi n« e' S <• '
M-hour u S >«*>*•
ill S CM ItX)
It-tour me 1*1 l' ICoxift
ot «l ISSJ)
ISJ4-7S
iMo-ai
M
III
M
J 4
It
0 M
s.s
X «i
I(-M
n-n
11-hour UK In fro. Ior
OnUrlo (bwjtfUld I9U)
U.I iirMn Md run! Wckorowl
(Lotuoo M •! ISS3)
Urtn SulUorlMd
(llttllorf tt «l ISU)
Surol WIUorlMd
(lltlltorf ot il ISS1)
ISSO-S1 It
isee-ai M
itM-ai a
an 10
IMI-4J 10
e u
0 rt
58
0)4
s.n
*•*
«-«
«-«s
«•*
at-«
(«) SOWCM: llcliolw {••' N); OuolMK « il (lof M).
(a) Mesothelloma and lung cancer combined.
REFERENCES
I Decker. J A. andSuder. Q. "Asbestos: r>oWeim and O>nskJeiwioos Related
lo Airborne Asbestos Sampling in an Outdoor Environment." Proc, Sia*
Ann Con/ Hazardous Hbsits and Hazardous Materials. HMCRI, Silver
Spring. MD. pp. SOS-SOS. 1989.
2 Decker. J.. Wao, N. and McDonald, A.. "Environmental Asbestos: Problems
Associated with PLM Soil Analysis." Proc. ftft Nai. Conf. Supeifiatd '*&
HMCRI. Silver Spring. MD, pp. M5-I51. 1988.
3. U.S. EPA, Health effects assessment for Asbestos. ECAO, Cincinnad, OH,
1984.
4. U.S. EPA. Airborne Asbestos Health Assessment Update. Office of Research
and Developmcnl. Research Triangle Park, North Carolina. EPA 600/8-84
003f. June 1986.
S. U.S. EPA. Superfund Public Health Evaluation Manual. U.S. EPA Office
of Emergency and Remedial Response. EPA 540/1-86-060). 1986.
6. "Asbestos: Proposed Mining and Import Restrictions," fed. Keg. . 5/. p. 3738.
Jan. 29. 1986.
7. "Asbestos-Containing Materials in Schools," frd, Reg.. 52, p. 41826, Oct.
30. 1987.
8. Occupational Safety and Health Administration (OSHA), 20 CFR Parts 19K)
550 BARRIERS
-------
Table 4
Indoor Asbestos Aire Levels in Schools and
Associated Excess Lifetime Cancer Risks
Sample Set
Total Excess Total Excess
Air (a) Lifetime Lifetime
Concentration Cancer Risk* Cancer Risk*
(ng/m3) for Teachers for Students
Air in U.S. schoolrooms
without asbestos
Air in Paris buildings
with asbestos surfaces
Air in U.S. buildings
with cement itious asbestos
Air in U.S. buildings
with friable asbestos
Air in U.S. schoolrooms
with asbestos surfaces
Air in U.S. schools
with damaged asbests
Air in U.S. schools
with asbestos surfacing
Ontario buildings
U.K. buildings
63
35
15
48
183
217
61
2.1
1.5
5E-05
3E-05
IE-OS
4E-05
1E-04
EE-04
5E-05
3E-06 (b)
1E-06 (b)
IE-OS
8E-06
3E-06
1E-05
3E-05
5E-05
IE-OS
NA
NA
* « Calculated using teacher exposure assumptions, converting ng/m3 - PCMe f/cm3,
and summing lung cancer and mesothelioma.
(a) DATA SOURCE: Ouelette et al. (Ref 24): Nicholsen (Ref 23).
(b) Calculated for an office worker.
and 1926. Occupational Exposure to Asbestos, Tremolite, Anthophyllite,
and Actinolite. Fed. Reg., 51, pp.22612-22790, June 20, 1986.
9. "Report of the Royal Commission on Matters of Health and Safety Arising
from the Use of Asbestos in Ontario," J. Stefan Dupre, Chairman, Toronto,
Ontario, 1984.
10. Schneiderman, M., Nisbet, I.C.T. and Brett, S.M., "Assessment of Risks
Pbsed by Exposure to Low Levels of Asbestos in the General Environment,"
Bga-Berichte, 4, pp. 31-37, 1981.
11. Stanton, M.F., Layard, M., Tegris, et al., "Relation of particle dimension
to carcinogenicity in amphibole asbestos and other fibrous minerals," J. Natl.
Cancer Inst., 67, pp. 965-975, 1981.
12. Bertrand, R. and Pezerat, H., "Fibrous Glass: Carcinogenicity and dimen-
sional characteristics," In Wagner, J.S. and Davis, W. eds. Effets Biologiques
des Fibres Minerales, W.H.O., Lyons, France, pp.901-011, 1980.
13. Wylie, A.G., Virta, R.L. and Segreti, J.M., "Characterization of Mineral
Population by Index Particle: Implication for the Stanton Hypothesis,"
Environ. Res., 43, pp.427-439, 1988.
14. National Toxicology Program (NTP), "Broad draft. NTP Technical Report
on the Toxicology and Carcinogenesis Studies of Chrysotile Asbestos in
F344/N Rats," DHHS, Research Triangle Park, NC, NTP TR 295, 1984.
15. U.S. EPA, "Drinking Water Criteria Document for Asbestos," Environmental
Criteria and Assessment Office, Cincinnati, OH, Mar. 1985.
16. U.S. EPA, "Memorandum from Brenda Riddle," Office of Air Quality
Planning and Standards, Research Triangle Park, NC, Feb. 19, 1987.
17. U.S. EPA, "Interim Guidance on Compliance with Applicable or Relevant
and Appropriate Requirements," Fed. Reg., 52, pp. 32496-32499, Aug. 27,
1987.
18. U.S. EPA, "Ambient Water Quality Criteria for Asbestos," NTIS PB81-117335,
1980.
19. Aqua Terra Technologies, Interim Report South Bay Asbestos Area, 1,
Pleasant Hill, CA, 1986.
20. U.S. EPA, "A method for estimating Fugitive Paniculate emissions from
hazardous waste sites," EPA 600/2-87-066, 1987.
21. U.S. EPA, "Environmental Asbestos Roads Study: Field Work Report,"
Emergency Response Section, San Francisco, CA, Jan. 1988.
22. Addison, J. Davies, L.S.T., Robertson, A. and Willey, R.J., "The Release
of Dispersed Asbestos Fibres from Soil," Institute ofOcc. Med., Edinburgh,
UK, 1988.
23. Nicholson, W.J., "Airborne levels of mineral fibers in the non-occupational
environment," Mount Sinai School of Medicine of the City University of
New York, NY, pp. 1-39, 1987.
24. Ouelette, R.P., Dilks, C.F., Thompson, W.C., Jr. and Cheremisnoff, P.N.,
Asbestos hazard management, Technomic Publishing Company, Lancaster,
PA, 1987.
BARRIERS 551
-------
Alternatives to the Remedial Investigation/Feasibility
Study Process
Paul C Chrostowski, Ph.D.
Lorraine J. Pearsall
Clement Associates, Inc.
Fairfax, Virginia
INTRODUCTION
Since CERCLA was passed in 1980, the Super-fund process ai inactive
hazardous waste sites has become more institutionalized and cumber-
some as time progresses. Although sites are being cleaned up, the U.S.
EPA has been severely criticized by environmentalists, industry and
Congress alike for lack of progress in the program. A few statistics
will enlighten this situation.
As of Mar. 31, 1989. the U.S. EPA' estimated that there was a total
of 30.844 potentially hazardous sites in the United States. Removal
activities had been completed at 1,121 sites: 642 sites were completed
through removal action; and work at 41 NPL sites had been completed.
Thus, it is apparent that a considerable amount of work still needs to
be done. This work proceeds slowly.
For the program as a whole, only 680 Rl/FS have been started Some
RJ/FS can take many yean to accomplish. For example, activity started
at the Lowry Landfill in 1984, however the last feasibility study for
an operable unit is not scheduled to be completed until 1994. fully K) yr
from initiation of work at the site.
The purpose of this paper is to suggest some alternatives to the tradi-
tional RI/FS process which could help to streamline the program with
the net result that more sites would be cleaned up in a shorter time
period and at less cost. We begin with a discussion of the Superfund
process as it currently operates. This background material is followed
by a discussion of the current status of the program, from the stand-
point of numbers of sites at various stages of the process, types of sites
and costs for performing various activities. Last, we make concrete
recommendations in several areas which can be readily implemented
to streamline the process. The reader should keep in mind that many
of the comments in this paper also apply to corrective action under
RCRA. In fact, one suggested goal is to merging of these two programs.
CURRENT SUPERFUND PROCESS
This discussion is intended to furnish a baseline which describes the
Superfund process as it currently operates. It should be recognized that
a Superfund cleanup is an extremely complex process and no two sites
are treated in exactly the same way. Additionally, as this is being writ-
ten (August, 1989), we are between two versions of the National Con-
tingency Plan. Different elements of each version are being applied at
different sites to varying degrees Thus, the information in this section
is intended as a paradigm and does not necessarily to represent any
particular site.
The first step of the Superfund process is the identification of poten-
tially hazardous sites which may require remedial action and their entry
in a data base known as CERCLIS. At this point, or at any time there-
after, a removal action may be conducted at a site due to emergency
conditions which may require rapid response or because the situation
at the site may worsen considerably before a full-scale remedial action
can be implemented.
In the pre-remedial process, sites undergo a preliminary assessment
(PA) and a site inspection (SI) which usually culminates in a scoring
by the hazard ranking system (HRS). Currently, if a site scores over
28.5 on the HRS. it is placed on the NPL where it is eligible for inves-
tigative and possible remedial action. Approximately 10% of all sites
which are initially identified are finally listed on the NPL. Concomi-
tant with this, the Agency for Toxic Substances Disease Registry
(ATSDR) conducts a health assessment to determine if an imminent
health threat exists or if runner community public health studies (e.g.,
epidemiology and biological monitoring) are necessary.
Once a site is listed on the NPL, it undergoes an RI/FS to determine
the nature and extent of contamination and to evaluate alternatives for
remedial action. The RI and FS usually overlap in time; for example,
there can be initial scoping of alternatives while field data are being
collected. The RI starts off with the preparation of a work plan. This
process is an evaluation of all data previously collected (e.g.. during
the SI/PA or by other parties) and an in-depth cost and time proposal
for the conduct of the RI FS. A preliminary risk assessment, identifi-
cation of applicable or relevant and appropriate requirements (ARARs),
determination of data quality objectives (DQOs) arid an initial screening
of remedial alternatives often accompany the work plan is approved,
actual investigative work commences at the site.
The majority of this work involves the collection of samples for chemi-
cal analysis with the results being used to determine the nature and
extent of contamination. Samples are collected by site personnel and
analyzed by the U.S. EPA's Contract Laboratory Program Regular
Analytical Services (CLP-RAS). In cases where the RAS cannot meet
the requirements of the RI/FS work plan. Special Analytical Services
(SAS) are employed. During the time when field work is occurring,
the FS engineers are screening initial alternatives.
After the data leave the CLP laboratory, they go through a process
of validation which ensures that the data meet the U.S. EPA's QA/QC
requirements. The data are then used in the RI report to describe the
nature and extent of contamination.
Another use of analytical data from site samples is in the human health
risk assessment or public health evaluation which is performed as part
of the RI/FS. The objective of the risk assessment is to assist the U.S.
EPA in decision) making at the site, especially regarding remedial
decisions which have a public health basis.
Additionally, during this time, the FS progresses through its final
evaluation of alternatives, with the result that one alternative is recom-
mended to the U.S. EPA. Two additional risk assessment activities
accompany the FS. The first assessment is a determination of prelimi-
nary remediation goals (cleanup levels) for contaminants in various
552 BARRIERS
-------
media at the site; this determination takes health effects and ARARs
into account. The second assessment is a health based screening of
remedial alternatives which accompanies evaluations of long) and short-
term effectiveness and reduction of toxicity as required by SARA.
Following the completion of the RI/FS, the U.S. EPA issues a ROD
which states the chosen remedy, justifies its choice and responds to
comments received from the public on the RI/FS. The ROD may decide
on a no-action alternative. Additionally, a ROD may be issued for a
portion or single operable unit at a site. In essence, the ROD is similar
to a combination of a final environmental impact statement as used under
the National Environmental Policy Act and a contract in which the
government commits to ensuring that the site will not present a current
or potential threat to public health, welfare or the environment.
After the issuance of a ROD, the site proceeds to the remedial design
(RD) stage in which the details of construction for remediation are
worked out. This step may be preceded by a conceptual design and
also may require additional sampling and analysis over what was
performed for the RI/FS. Once the RD is approved, the remedy is
implemented as a remedial action (RA). When an effective cleanup
has been accomplished, the site is removed from the NPL.
Most sites have ongoing operations and maintenance activities which
typically last for 30 yr to cover post-closure monitoring requirements
of RCRA. Additionally, if hazardous materials are left on-site in a form
where they are still toxic and potentially mobile, the site may be revisited
every 5 yr to ensure that the cleanup is still effective.
STATUS OF THE SUPERFUND PROGRAM
As mentioned previously, there are 30,844 potentially hazardous waste
sites which have been entered into the CERCLIS inventory1. There are
numerous estimates of the total potential number of hazardous sites in
the United States. The General Accounting Office2 has analyzed this
aspect of the program and determined that there are between 130,000
and 425,000 potentially eligible sites. Preliminary assessments have been
completed at 28,101 sites and 9,902 sites have had site inspections
performed'. As time progresses and the worst and most obvious sites
are remediated, it becomes more difficult to perform these pre-remedial
activities. More sophisticated sampling techniques such as groundwater
wells and air monitoring programs are required at sites where wastes
are present, but not obvious. It is estimated that current costs for SI/PA
activities may exceed $100,000 and may be as high as $200,000 per site.
Thus fer, 2,053 sites have been scored by the HRS to date1. In Dec.,
1988, the U.S. EPA proposed a revision to the existing HRS which is
more sophisticated and uses more principles from risk assessment than
the current version. No further action is planned at 12,416 sites, while
the U.S. EPA proposed to list 1,163 sites on the NPL. If we use the
rule of thumb that 10% of all identified sites are finally added to the
NPL, the number of NPL sites could ultimately exceed 40,000.
RI/FS activities have been started at a total of 845 sites, including
Federal Facilities'. Since an RI/FS is usually carried out for all sites
listed on the NPL, up to 40,000 RI/FS studies ultimately will be per-
formed. The U.S. EPA currently3 estimates that the cost of an RI/FS
is $1,100,000. Note that this cost has escalated rapidly. As recently as
1985, the comparable figure was $800,000 per site. Remedial design
activities have been performed for 300 sites'. The costs associated
with a remedial design are approximately $750,000 per site3.
Remedial activities have been implemented or are in progress at a
total of 204 sites'. The U.S. EPA estimates the cost of an average
remedial activity to be $13,500,0003.
Taken as a whole, the costs for an average site, excluding pre-remedial
activities, but including the RI/FS, remedial design, remedial action
and O & M, are approximately $19,000,000. If this cost factor were
to be applied to the upper bound potential of 40,000 sites, the ultimate
cost of the Superfund program would be over $700 billion.
In addition to cost considerations, there are time considerations. A
typical RI/FS requires approximately 1 yr to perform. The U.S. EPA
would like to see this time reduced if possible. At some sites, however,
this time is substantially prolonged. At the Lowry Landfill site, there
are five operable units. One of these units involves an expedited removal
action which will be completed by the middle of 1991. The remainder
of the activities leading up to the FS for the last operable unit, however,
will not be completed until 1994. The RI/FS studies for the individual
operable units will each require about 4 yr to complete.
ALTERNATIVES TO THE RI/FS PROCESS
The remainder of this paper is devoted to recommendations for
streamlining the RI/FS process. It is recognized that some of these
recommendations may be controversial and that all of them could not
be implemented at once. The goal of making these recommendations
is to achieve greater flexibility and less institutionalization in the Super-
fund process. Some of these recommendations could be implemented
readily. However, some would require regulatory action such as a
revision to the NCP.
Maintain Consistency in Contractors
The U.S. EPA uses numerous mechanisms to procure professional
services at hazardous waste sites. These contracts include field inves-
tigation teams, technical assistance teams, laboratory management and
remedial management.
Throughout the history of the program, various types of contracts
have been in place. For example, the REM n contract covered the whole
United States and involved a small group of contractors. The REM III
and REM IV contracts divided the country in half and used a greater
number of contractors. The ARCS contracts are regionalized and will
use almost 50 contractors. PRPs and State governments also use a series
of contractors.
Often no contractor continuity is maintained and inexperienced con-
tractors are used to satisfy procurement rather than technical require-
ments. Additionally, artificially strict conflict of interest rules often make
it difficult for the most experienced and qualified contractors to work
on government contracts and they are relegated to the private sector.
The bottom line is that one contracting team should be hired for all
activities at a site from SI/PA through remedial design. Unless the con-
tractor fails to perform, the firm should be kept at the site until all these
activities have been completed.
Minimize Reliance on Contract Laboratory Program
One of the largest cost elements of an RI is chemical analysis. The
stringent quality assurance and record-keeping requirements of the con-
tract laboratory program can add as much as 50% to the cost of per-
forming an analysis. Additionally, there seems to be a common
perception that DQO Level IV analysis is required for risk assessment,
evaluation of alternatives and engineering design 4. In reality, the
largest source of uncertainty in risk assessment lies in the quantitative
toxicological parameters used to characterize risk and the mathematical
fate and transport modeling used to calculate exposure point concen-
trations.
Since the risk assessment will only be as accurate and precise as its
component parts, it does not seem reasonable to make the chemical
analytical data more accurate than toxicological or modeling results.
Evaluation of alternatives also is often carried out in a qualitative fashion
with order of magnitude estimates of cost. As with risk assessment,
this activity does not require rigorous QA/QC.
In addition to the problem associated with perceived data quality
needs, there are some problems with the CLP itself. First, a large
proportion of data received from the CLP are often "qualified" (this
is the CLP term for data which do not meet the contract requirements).
The net result is that decisions for risk assessment and engineering
design often are made on the basis of estimated data. In one recent
site, for example, fully 85 % of the soil analytical data were estimated.
Last, there often is a time delay associated with obtaining CLP ana-
lyses due to the large number of samples in the program and the limited
number of qualified laboratories. It is difficult to rationalize the addi-
tional incremental costs of CLP QA/Q.C with the large amount of quali-
fied data and long time delays endemic to the program.
In lieu of sending large numbers of samples through the CLP with
its attendant cost and tune constraints, we propose maximizing reliance
on field measurement techniques and on lower degrees of quality
assurance (e.g., DQO Level III) for measurements actually performed
BARRIERS 553
-------
in the laboratory, the U.S. EPA1 has developed an automated system
for specifying a wide variety of field analytical methods ranging from
atomic absorption and x-ray fluorescence for metals to mobile gas chro-
matographs for volatile organic compounds to on-sitc GC analysis for
PCBs. Although not mentioned by the U.S. EPA. there are no techni-
cal or cost barriers to using on-silc, mobile GC/MS.
Data collected by these field measurements could be used for risk
assessments, engineering design and selection of alternatives. Only in
cases where litigation was anticipated would the more rigorous CLP
QA/QC program be used.
Perform the RI/FS Critical Path
At most sites, a complete RI/FS/RD may not be necessary. Evalua-
tion of numerous completed RJ/FS studies reveals that there are many
redundant elements and that there often is a critical path through a RI/FS
which would eliminate redundancy. Since SI/PA activities have expanded
recently, for a number of sites, data from an SI/PA may be all that is
required to go directly to remediation. An example of this type of site
would be one where groundwater contamination from a known source
was a problem. If sufficient samples are taken during pre-remedial
activities, the identity and a rough idea of the concentration of the chemi-
cals of concern will be known. The remedy is obvious: source control
and pump-and-treat.
Groundwater cleanup technology has been studied extensively and
the U.S. EPA has developed guidance on remedial actions for contami-
nated groundwater at Superfund sites6. Most typical groundwater con-
taminants such as volatile organics and heavy metals have ARARs in
the form of maximum contaminant levels; thus, the cleanup objective
is already known. A remedial response would involve the following:
(1) remove the source; (2) install one or more extraction wells in the
zone of known contamination; (3) initiate pumping and treatment of
contaminated water; (4) use well points (possibly in conjunction with
geophysical techniques) to explore the extent of the problem, taking
measurements with a portable GC as wells are drilled; and (5) install
new extraction wells as necessary.
Similar scenarios could be envisioned for contaminated soils. Again,
the U.S. EPA has produced guidance for treatment of CERCLA
soils7). When contaminated soils are discovered as pan of pre-
remedial activities, field screening techniques may be rapidly deployed
to determine the extent of the problem. Preliminary remedial objec-
tives (numerical cleanup goals) may be calculated using the Prelimi-
nary Pollutant Limiting Value' (PPLV) approach.
Excavation or in situ treatment could be preceded by field analysis.
For example, at an inactive secondary lead refinery site, a portable x-ray
fluorescence unit could be used to determine on a real-time basis those
areas where lead concentrations exceed the cleanup goal calculated by
the PPLV approach. The analytical unit could be followed by the con-
struction unit which would excavate the lead-contaminated soils for sub-
sequent treatment.
Information from other U.S. EPA programs could be used to identify
treatment technologies. For example, if a listed RCRA waste was iden-
tified, then the best demonstrable available technology as identified in
the land disposal restrictions could be used.
Combine all Public Health Related Activities
Currently three public health evaluation activities take place at all
Superfund sites. These three separate evaluations include the HRS
scoring, the risk assessment performed by the U.S. EPA and the health
assessment performed by ATSDR. Again, with increased SI/PA activities
and HRS scoring taking on more of the attributes of risk assessment,
it may be possible to eliminate the HRS entirely and go directly to the
baseline risk assessment. Preferably, risk assessments would be per-
formed in an iterative manner. The first risk assessment would be based
on data from the SI/PA. If these data were adequate to demonstrate
an actionable level of risk, no further work would be required. If the
data were not adequate, exposure-based sampling plans could be
designed to fill the data gaps. In no case would data which were
extraneous to the risk assessment process be collected
To eliminate the redundancy caused by the ATSDR Health Assess-
ment, we suggest that ATSDR propose guidelines for conducting these
activities. These guidelines could be provided to risk assessment con-
tractors and integrated into the baseline risk assessment. For example,
statistical analysts of cancer incidence in a given area to determine if
the local observed incidence is greater than the norm observed for the
stale is an activity which is delegated to ATSDR. often is performed
by state or local health departments and fails to be integrated into the
Supertund process. This activity could be performed by qualified risk
assessment contractors and become part of the baseline risk i
Privatize UK Superfund Process
Recently, a great deal of attention has been paid to the role which
contractors are playing in Supertund. The U.S. EM has been criticized,
with allegations being made that contractors are writing Agency policy,
for example. One solution to this problem would be to remove all U.S.
EPA personnel from the Superfund management process and transfer
them to policy formulation rotes. Supertund management could be taken
over by contractors. There are, indeed, many contractors with the
requisite experience to manage large engineering projects which would
be directly transferrable to Superfund management. Under this option,
the Superfund staff would be located at U.S. EPA headquarters and
be responsible for policy, enforcement and ensuring consistency among
contractors.
REFERENCES
1. Memorandum from T. Juszczak. Director Resource Management Staff,
OSWER. 10 J. Cannon and R. Duprey, dated May 16, 1989. entitled "Super-
fund Progress Report as of March 31. 1989."
2. General Accounting Office (GAD) 1987. Supertund: Extent of Nation's Poten-
tial Hazardous Vfeie Problem Still Unknown. GAO/RECD-88-44.
3. US. EPA National Priorities List for Uncontrolled Hazardous Waac Saes-
Fraal Update No 5. federal Kegiaer. 54(61). pp. I3296-U3I7. Mar. 31,1989.
4. U.S. EM . Data Quality Objecrivrs for Remedial Response Activities,
ER\/S4(VG-87/003. Office of Emergency and Remedial Response, Washington,
D.C.. 1987
5 U.S. EM. Field Screening Methods Catalog, EPA/MOtt-88/005. Office of
Emergency and Remedial Response. Washington, D.C.. 1988.
& U.S. EPA. Guidance on Remedial Actions for Contaminated Groundwtaer
at Superfund Sites, EPA/54(VG-88AX)3, Office of Emergency and Remedial
Response, Washington, D.C.. 1988.
7. U.S. EPA, Technology Screening Guide for Treatment of CERCLA Soils and
Sludges. EPA/54(V2-88/004. Office of Solid VAstt and Emergency Response.
Washington. DC. 1988.
8. Dacre, J.C.. Rosenblatt, DiH. andCogley, UR, "Preliminary Pollutant Limit
Values for Human Health Effects," Environ. Set. TMinot, M(7). pp. 778-784,
1980.
554 BARRIERS
-------
Optimizing the Use of Soil Gas Surveys
David S. Naleid
Franco E. Godoy
The Earth Technology Corporation
Alexandria, Virginia
ABSTRACT
Recently there has been a significant increase in the use of soil gas
surveys as an assessment tool to investigate hazardous waste sites. The
conventional soil gas survey technique was developed to investigate sub-
surface contamination from volatile organic compounds by measuring
the concentration of their vapors in shallow soils. With benefits that
include cost-effectiveness, thoroughness, safety and speed, soil gas
investigations have become commonplace staple in the remedial inves-
tigation process.
Widespread acceptance and use of this technology has spawned
numerous techniques and vendors with a wide array of sampling,
analysis and interpretation services. Soil gas surveys range from simple,
do-it-yourself techniques to in situ testing providing detailed, real-time
analytical data.
This paper presents an overview of available techniques and methods
to determine the level of survey sophistication required to meet specific
remedial investigation objectives. Recent studies conducted by Earth
Technology are discussed to illustrate the implementation of site-specific
soil gas surveys.
INTRODUCTION
Volatile organic compounds establish a vapor phase plume in the intra-
granular space in the vadose zone above contaminated areas. Soil gas
survey techniques allow one to exploit this transport process specific
to volatiles in order to build a two-dimensional picture of soil and/or
groundwater contamination. This information can be utilized in both
contaminant source detection and plume delineation. The key charac-
teristics which make the soil gas survey an effective component of field
investigations are:
• Rapid acquisition and turn-around time of data
• Ability to use soil gas data to focus or redirect project resources
• Capability to thoroughly screen large areas at a relatively low cost
The foundation for successful application of soil gas techniques lies
in clearly identifying remedial investigation objectives and designing
a survey to support these goals. Figure 1 presents the life-cycle of a
typical phased remedial investigation along with the stages where soil
gas surveys may be appropriate.
Available Survey Techniques
A broad assortment of methods exists for soil gas sample acquisi-
tion, analysis and interpretation which provide a range of qualita-
tive/quantitative results. Each method has its appropriate
application—from site screening to plume boundary delineation for
recovery well placement, etc. Selecting the optimal approach requires
examining the site and contaminant characteristics, determining
analytical needs and determining the level of interpretation required.
PHASES OF REMEDIAL MVESnOATON WHERE
SOB. OA9 SURVEYS MAY BE Al
Figure 1
Appropriate Phases of Remedial Investigation
For Soil Gas Survey
Economic and scheduling constraints must also be incorporated into
this analysis. Figure 2 presents a step-by-step approach designed to iden-
tify the degree of survey sophistication necessary for project success.
The following sections address these key elements and present perti-
nent technical data essential to proper survey scoping.
Sampling
Sampling for soil gas can be done at any depth above the ground-
water table due to the existence of a vapor concentration gradient from
the contaminant source/plume to the surface. Standard practices involve
sample collection in the vadose zone at depths where surface/atmosphere
effects are minimal. Soil gas collection depths typically range in the
3- to 5-ft depth level. Compounds which are amenable to soil gas
detection at this depth can be characterized as those which possess high
vapor pressures under ambient conditions, low aqueous solubilities and
degrade slowly in the environment. For compounds which degrade
quickly, a depth of 10 ft is preferable to reduce the effects of oxida-
tion/microbial breakdown. Other factors which influence sampling depth
include the subsurface conditions such as perched aquifers or ground-
water at great depth.
Soil gas samples are obtained through headspace volatilization of a
containerized soil sample or through in situ extraction of gas using
shallow probes. Soil samples can be obtained using a hand auger, while
in situ gas samples require more sophisticated equipment. In situ
sampling can be performed with specialized equipment such as small-
volume, slide hammer driven probes or hydraulically driven cone
penetrometer units mounted on a vehicle. Once a representative sam-
ple has been obtained, on-site or remote analytical qualitative or quan-
titative work can be completed.
VAPOR CONTROL 555
-------
OTOMO I
aoKD I
WTW I
LMM.-IM
HTAANALYM
>
•040U1
—
KTWf«tT»T10N WO I
Tram n or OAT* r^~
Figure 2
Decision Tree for Optimization of Soil Gas
Analysis
At this point in the survey, there is little variation in the costs (sample
acquisition), it is the analytical techniques which generally drive project
costs and the quality/detail of the results. Techniques employed in the
characterization of soil gas samples are shown in Table I and range
from qualitative measures of total volatile organic compounds to a quan-
titative compound by compound analysis. Qualitative results require
ionization and detection of volatile sample components, while quan-
titative data require chromatographic separation of the volatile compo-
nents followed by ionization and detection using electron capture
methods.
Table 1
Soil GUI AnahitLs Techniques
Analytical
Technique
Quality of
Result*
Application
(aenoltIvlty)
Organic Vapor
Analyzer (OVA)
Photoionization
Detector
GC/PID
CC/ECD
Qualitative Roal-tiBe analyilc. Flame
lonizttlon Detection for
Total Organ Kn (ppn)
Qualitative Real-tin* analysis,
lonization of volatile
organic coBpounda lie!ted
by laap photoionization
energy (pp«)
Quantitative Pull range of EPA volatile
priority pollutants
All these analytical methods can be used on-sitc us samples are
acquired or remotely at a full-service laboratory. Key factors which
should be considered when selecting the optimal analytical strategy
include:
• Required turn-around time
• Value of qualitative vs. quantitative results
• Number of compounds to be analyzed
• QA/QC requirements
Costs and quality/detail need to be balanced at this point for the most
useful results. Typical unit costs for the acquisition, analysis and
interpretation of soil gas samples are provided in Table 2.
Soil Gm Amlytif toil Cods
Analtyical
Technique
Analysis Location/
Sample Collection Rat*
Unit CocU
Qualitative: Total Volatile Organic*
Organic Vapor Analyzer (OVA) Real-tiiw $ 17.50
Photoionization Detector (3 eanplee par hour)
Quantitative 8-10 compounds
CC/P1D
CC/ECD
CC/PID
OC/ECO
Real-tiM $ 180
(2 samples per hour)
Re»ote Lab $ 195
(3 samples per hour)
INTERPRETATION
Data presentation and subsequent interpretation is best done by
developing isoconccntration maps of the soil gas survey mulls. Soft-
ware is available which can quickly contour results for real-time use.
Qualitative/quantitative data can be superimposed on a site map which
contains all essential features such as (he survey grid, existing boring
and well locations and related data, groundwater flow direction, etc.
Once all pertinent site-specific data have been collected, the soil gas
data can yield:
• Relative location of subsurface contaminant sources
• Boundary of contaminant plumes
• The existence of preferential groundwater flow patterns
High quality data can provide information about the physical slate
of contaminants (i.e., dissolved species or free product), and the con-
dition of the contaminant (weathered or degraded).
APPLICATION OF SOIL GAS SURVEY TECHNIQUES
Recent soil gas studies conducted by Earth Technology have ranged
from qualitative measures of total volatile organic compounds to quan-
titative, compound-specific surveys, based on project needs. The
following discussions present information about two site-specific surveys
along with relevant details to illustrate the range of soil gas uses.
Abandoned Fuel Line Investigation
A preliminary screening soil gas survey was recently completed on
a I.S-mi long abandoned jet fuel line. Remedial investigation tasks
included locating (he 6- in. fuel line using geophysical techniques fol-
lowed by a soil gas survey designed to delineate segments of the line
where subsurface contamination had occurred.
Sampling involved probe penetrations at approximately K)0-ft inter-
vals, 2-ft off the center of the fuel line. A total of 55 samples was col-
lected over the entire length of the line. Head-space analyses were
conducted using a hand-held OVA. Sampling points, along with survey
results, are presented in Figure 3.
This survey very clearly delineated segments of the line where vola-
tile organic contamination existed due to jet propulsion fuel (JP-4) leaks.
Based on these results, several soil borings were completed along the
line to further delineate vertical and lateral contaminant migration.
556 VAPOR CONTROL
-------
Cont»rnlnit»d S*etlon»
Figure 3
Abandoned Fuel Line Qualitative Soil Gas Survey
Plume Delineation to Support Remedial Action
A quantitative soil gas survey is currently underway to support place-
ment of groundwater monitoring wells and a subsurface groundwater
collection system. Contaminants have been thoroughly characterized
through previous work at the site; however, the leading edge of the con-
taminant plume has not been defined. Figure 4 shows the soil gas survey
area located downgradient of existing groundwater monitoring wells.
Groundwater contaminants at this site include trichloroethene (TCE)
and its degradation products. The survey incorporates GC/FID/ECD
analysis of the samples to build a detailed picture of contaminant
migration.
CONCLUSIONS
Soil gas survey techniques are powerful, cost-effective tools to survey
and delineate volatile organic contaminant sources and migration.
Several techniques exist which provide a wide range of qualitative and
quantitative information. Careful review of investigation objectives to
identify data quality needs results in the selection of the appropriate
method to support future remedial efforts.
u ? n ^ a * n ,
Potential Ground-
water well locations
Soil Cos Survey
Figure 4
Plume Delineation to Support Remedial Action Development
VAPOR CONTROL 557
-------
Comparison of Air Stripping Versus Steam
Stripping for Treatment of Volatile Organic
Compounds in Contaminated Groundwater
Gerald E. Fair, RE.
Forrest E. Dry den, P.E.
Walk, Haydel & Associates, Inc.
New Orleans, Louisiana
ABSTRACT
In many cases, steam stripping can be a viable alternative to air
stripping for removal of VOC contaminants from ground water Each
method is applicable to a broad range of VOC compounds normally
found in contaminated groundwaier. but each situation should be
thoroughly researched or pilot-tested prior to design. When compared
to air stripping systems requiring vapor recovery, steam stripping offers
many advantages including greater environmental protection, better
operating performance, less operating attention and lower capital cost
Pre-assembied air strippers and vapor recovery units, including
instrumentation and internals, are readily available from vendors.
Although packaged steam stripping units are not as readily available,
they can be designed using basic process engineering principles and
a relatively simple process flow train. While design of air strippers
requires mass transfer data or pilot testing, steam strippers often can
be designed from boiling point data readily available in the literature
thereby eliminating the need for expensive pilot testing.
INTRODUCTION
While a great deal of information has been published regarding air
stripping of groundwaters contaminated with volatile organic compounds
(VOCs), relatively little has been published regarding steam stripping.
Although steam stripping has not been as widely used for groundwater
treatment as air stripping, it has been used extensively in the chemical
process industry for many years for solvent recovery and is a
well-demonstrated technology. Air stripping probably is used more
frequently because most cases involve low groundwater solvent
concentrations (in the ug/L range) and many are in remote locations.
These conditions favor air stripping since VOC-laden exhaust air often
can be released uncontrolled to the atmosphere without significant
impacts on ambient air quality.
However, for facilities treating relatively high concentrations of VOCs
and/or large flow rates, air emissions can be significant and therefore
subject to control prior to release. Increasingly more facilities are being
required to control VOC emissions from air stripping facilities.
Typically, air purification is accomplished with vapor recovery systems
using activated carbon as a VOC-adsorption medium. Some special
circumstances (high toxicity) may justify exhaust air incineration as
a method of VOC control. However, these systems add significantly
to the project cost and operating complexity.
For cases in which treatment of the exhaust air for VOC removal
is required, steam stripping should be considered as an alternative for
treatment of contaminated groundwater. In appropriate situations, it may
offer a number of advantages over air stripping with vapor recovery.
such as:
• Greater environmental protection
• Simplicity of operational control
• Better operational efficiency
• Lower capital costs
• Smaller equipment space requirements
The purpose of this paper is to provide information regarding the
applicability of air and steam stripping, and factors to be considered
in the evaluation of each method, for treatment of VOC-contaminated
groundwaters. Information provided in this report is based on a study
and a preliminary design evaluating air stripping versus steam stripping
for an industrial facility required to treat 25 gpm of groundwater
containing an average trichloroeihylene (TCE) concentration of
300 mg/L and a maximum concentration of 1200 mg/L.
AIR STRIPPING
Air stripping is a well-documented method for the removal of small
amounts of volatile organic chemicals from water. The potential
effectiveness of air stripping can be evaluated using Henry's law, which
provides a measure of the relative volatility of the VOC According to
Henry's law, the relative concentrations of VOC in the water and air
will be functions of the VOC vapor pressure. Values for Henry's law
constants can be calculated, determined from pilot studies or obtained
from the published literature. Table I presents information published
for five chlorinated solvents commonly found in contaminated
groundwater'.
Compounds with relatively large Henry's law constants are more
easily stripped than compounds with lower values. As can be seen in
Table 1. the value of Henry's law constant is very much temperature
dependent. While many VOCs can be air stripped at ambient
temperatures, those with low volatiles at ambient temperatures, those
with low volatilities at ambient temperatures may require preheating
of the groundwater entering the stripper1 Preheating of the stripper
feed also might be required during winter operations in extremely cold
climates.
A schematic diagram of a typical packed tower air stripper with vapor
recovery is shown in Figure 1. Contaminated groundwater is sprayed
on the top of the tower and allowed to flow by gravity down through
the packing. Air is blown upward through the packing and then passed
through a vapor recovery unit. In the tower. VOCs are transferred from
the water phase to the air phase. The water exiting the tower normally
is acceptable for discharge.
To size a stripping lower, mass transfer coefficients are required to
determine how much VOC contaminant is transported from the water
into the air per unit volume of packing per unit time. Appropriate mass
transfer coefficients are difficult to obtain because their values are
dependent upon air loading, water loading, VOC concentrations,
operating temperatures, packing type, packing size and properties of
558 VAPOR CONTROL
-------
Table 2
Most Frequently Reported Organic Solvents3
TREATED EXHAUST
TO ATMOSPHERE
Figure 1
the fluids and solutes to be removed. However, because of the large
number of interdependent factors, the overall efficacy of air stripping
is sensitive to temperature, influent VOC concentrations and the air/water
ratio, even for a well-designed tower. To assure consistent operational
results in an air stripping system, the design must make provisions to
accommodate possible variations in such conditions as air and water
temperature and VOC concentrations in the groundwater.
Vapor recovery systems used with air stripping usually consist of two
or more carbon adsorption vessels connected in parallel. The use of
multiple vessels allows for on-site regeneration of one vessel while the
other is on-stream. While various regeneration systems are available
using hot inert gas, steam is used most often since it is economical
and normally is readily available. Flow between the vessels can be
controlled automatically by a timer or via an in-stack monitor. Following
the desorption step, the steam is condensed and the VOCs are separated
in a decanter. The condensed water layer is returned to the stripping
tower inlet system.
STEAM STRIPPING
Table 2 lists the organic solvents most frequently found in remedial
action projects at NPL sites3. Most of the solvents on the list (13 out
of 14) are readily amenable to steam stripping.
Steam stripping is technically feasible when the following conditions
are present:
• The VOC will form an azeotrope with water which has a boiling
point less than that of water
• The condensed overhead azeotropic product separates into an organic
layer and a water layer
Table 1
Henry's Constant for Five
Chlorinated Solvents1
Compound
1,1, 1-trichloroethane
4262/T)
tetrachloroethylene
5119/T)
trichloroethylene
4929/T)
chloroform
4180/T)
methylene chloride
Henry's Constant
(m3 -atm/fflol
@ 20°C)
0.0132
0.0130
0.00764
0.00333
0.00225
Temperature
dependence
equation (°K)
He = exp (10. 21-
Hc = exp(13.12-
Hc = exp (11. 94-
Hc = exp(8.553-
Hc exp(8.200-
4191/T)
Except for phenol, all of the VOCs listed in Table 2 form azeotropes
with water, separate into distinct layers upon condensation and thus
are amenable to steam stripping. Data regarding azeotropic compositions
and boiling points for most of these VOC compounds are presented
in Table 3. All of these materials form low boiling azeotropes, i.e.,
Rank
Substance
Percent of Sites
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Trichloroethylene
Toluene
Benzene
Chloroform
Tetrachloroethyleneie
Phenol
1,1, 1-Trichloroethane
Ethylbenzene
Xylene
Methylene chloride
Trans-1 , 2-Dichloroethylene
1 , 2-Dichloroethane
Chlorobenzene
1 , 1-Dichloroethane
Carbon tetraohloride
33
28
26
20
15
14
13
13
12
11
8
8
8
7
the boiling point of the solution is less than the boiling point of either
pure constituent. Also, all of the azeotropic compositions shown in Table
3 are substantially higher in VOC content than in water content. Both
of these conditions are favorable factors for steam stripping. Additionally,
all of the VOCs have limited solubility in water, thereby yielding a
two-phase system upon condensation.
Table 3
VOCs Commonly Found in Contaminated Groundwater
Which can be Steam Stripped3.
Substance Azeotropic
Boiling Point
(°C Q760 mm)
Trichloroethylene
Toluene
Benzene
Chloroform
Tetrachloroethylene
1,1, 1-Trichloroethane
Ethylbenzene
Xylene
Trans-1 , 2-Dichloroethylene
Chlorobenzene
Carbon tetrachloride
73.
85.
69.
56.
88.
86.
92.
94,
71.
90.
66,
1
0
,4
,3
.5
.0
,0
.5
.0
.2
.8
Azeotropic
Composition
(Weight %
water / solvent)
6.3 /
20.2 /
8.9 /
3 /
17.2 /
16.4 /
33.3 /
40 ,
8.2 ,
28.4 ,
4.1 ,
' 93.7
' 79.8
' 91.1
' 97
' 82.8
' 83.6
' 66.7
1 60
/ 91.8
1 71.6
I 95.9
It should be noted that for cases in which several different VOCs
are present, ternary or complex azeotropes may be formed; however,
these also may boil below the boiling point of water.
A block flow diagram of a steam stripping system that can be used
for' removal of VOC from groundwater is shown in Figure 2.
Contaminated groundwater is recirculated through a steam reboiler,
VAPOR CONTROL 559
-------
which provides heat for vaporization of the azeotrope. Instead of the
indirect heating method using a rcboiler, live steam can be added directly
into the column. Direct overhead vapor, consisting of the VOC azcotrope
and some excess water, is condensed and collected in a decanter vessel.
In the decanter, the VOC and water separate; the water layer flows by
gravity or is pumped via a reflux line back to the stripper column. The
VOC layer can be controlled by an interface controller which can scn.sc
eitherspecific gravity or conductivity. Steam-stripped groundwater from
the column is passed through a feed preheater to cool the effluent and
to heat the feed entering the column The water is then discharged
CONDIMMft
Figure 2
Unlike air stripping which is sensitive to many variables, steam
stripping is dependent only upon the operator maintaining the
temperature ai the column bottom above the boiling point of the
azeotrope. This temperature level ensures that the VOC cannot exist
in liquid form within the column. In practice, the temperature in the
column can be maintained at or near the boiling point of water to ensure
proper VOC removal.
A temperature control loop regulates the steam supply so that the
column temperature remains above the azeotrope boiling point or the
boiling point of water. Because the system is totally enclosed, there
is no contaminated off-gas requiring treatment. Sizing of the steam
control system is dependent only upon the flow rate of the feed and
is insensitive to the VOC concentration as long as (here is an excess
of water present and the column bottom temperature is maintained above
the boiling point of the azeotrope.
CONSIDERATIONS IN THE SELECTION OF
AIR OR STEAM STRIPPING
Items to be considered in selecting an air or steam stripping system
include design and operating simplicity, capital and operating costs.
environmental impacts and requirements for prctreatmem
Design and Operating Simplicity
Design and operating simplicity are important factors to be considered
in the evaluation since they affect capital and operating costs, operating
performance and operating and maintenance labor requirements
While both air and steam stripping systems can be fully automated.
a steam stripper requires less operator attention and less instrumentation
than an air stripper since the only critical operating variable is (he
temperature in the stripper column bottom. On the other hand, air
stripping systems are sensitive to a number of variables including:
• Air and water temperatures
• Air and water flows
• Air humidity
• Contaminant concentration variability
• Contaminant composition
• Air/water ratio
• Tower pressure drop
Additionally, air stripping systems may require VOC recovery units
which add significantly to operating complexity and cost. Such recovery
units are not usually required with steam stripping where efficient
condensers are installed. These vapor recovery units often are required
to have a high capture efficiency (95 to 98%), which approaches
state-of-the-art performance for such equipment. For these applications,
there are a number of design and operating factors which can affect
system performance.
Vapor recovery for air stripping is comprised of three closely related
process operations: adsorption, desorption and cooling/drying. After
carbon regeneration with steam, a hot, wet carbon bed will not remove
organics from air effectively since high temperature and humidity do
not favor complete adsorption. Therefore, it is important to allow an
adequate cycle lime to completely dry and cool the beds with air*. It
maybe necessary to add an air prehcaicr to dehumidify such air entering
the adsorbers. In hot humid climates it may be necessary to precede
the air preheater with an air cooler to condense moisture, since only
preheating may increase the temperature of the beds to such an extern
that adsorption efficiency will be reduced.
Proper construction materials are critical to the design life of the
adsorber vessels due to the moisture present during regeneration cycles.
Stainless steel may not be acceptable for chlorinated solvents because
of the possibility of chloride stress attack. Therefore, exotic material
such as Hastelloy may be necessary.
In order to achieve consistently high removal efficiencies of 95 to
98% in the vapor recovery unit, careful operating and maintenance
attention is required. Minor failures of the many vapor recovery system
components (such as small teaks in valves, piping or adsorber vessels)
can affect operating performance Valve selection is especially important
because valves isolate the air outlet from the air inlet stream. Thus,
a small valve leak may result in unacceptable outlet VOC concentrations
even when adsorption is accomplished in the operating adsorber vessel.
Also, while sequencing between the adsorber vessels may be
automatically controlled, periodic operator surveillance is required due
to the potential for release of VOC air pollutants. In summary, air
strippers with vapor recovery units are more complex systems to design
and operate than steam stripping systems.
Capital and Operating Costs
The choice of materials of construction has a great influence on capital
costs. While fiberglass reinforced plastic is usually a cost-effective
choice for air strippers, steam stripping towers and vapor recovery
adsorber vessels may require more exotic materials. However, since
equipment used with steam stripping systems is relatively small-scale.
the use of expensive materials such as glass-lined vessels and titanium
heat exchangers may not be cost-prohibitive. In a 1986 industrial cast
study of the treatment of 25 gpm. we found that a glass-lined vessel
could be purchased for $22,000 and a titanium heat exchanger could
be purchased for $10.000.
Typically, the purchased costs of vapor recovery units are
significantly higher than the air stripping towers. In the study described
in this paper, the bare equipment cost for the vapor recovery unit was
$83,000 (for lined, steel adsorber vessels) compared to the equipment
cost of $30,000 for the air stripper. Using Hastelloy adsorber vessels
would have increased the equipment cost of the vapor recovery unit
to $140,000. Also, associated costs such as foundations, piping,
installation costs and instrumentation can significantly add to the cost
of vapor recovery units.
Table 4 shows an annual operating cost analysis for a packed tower
air stripper versus steam stripper for the treatment of 25 gpm of
TCE-contaminated groundwater. As shown in Table 4, the annual costs
are slightly lower for the steam system. The comparison is based on
installation cost of the stripping equipment and associated ductwork,
piping and instrumentation. The costs do not reflect water handling
equipment to and from the strippers such as pumps, storage tanks and
other ancillary equipment that would be common to both systems. Also,
the capital cost analysis does not include the cost of foundations, which
can be expected to be higher for the air stripping/vapor recovery system
due to its greater space requirements.
560 VAPOR CONTROL
-------
cost
Item
Table 4
Annual Cost Comparison of Air versus Steam Stripping
Air Stripper/ Steam
Vapor Recovery Stripper
Capital Charges* 60,000
Utilities
Electricity (S0.05/KWH) 3,500
Steam ($6/1000 Ib) 3,600
Carbon Replacement 2,000
Total S69.100
26,000
38,000
$64.000
(1) Annual capital charges @24% of installed cost including:
depreciation and interest @15%; taxes, insurance, and
administrative charges §4%, maintenance and materials 65%.
The installed cost of the air stripper/vapor recovery system is
$250,000 versus $108,000 for the steam stripping system. The steam
stripper cost includes a glass-lined column and two titanium heat
exchangers required by the high chloride content and low pH of the
water entering the column. Use of such exotic materials is probably
not necessary in most cases. Use of carbon steel equipment would lower
the installed equipment cost by about 25 %. The use of carbon steel
may be acceptable if sufficient corrosion allowance is provided to
achieve the equipment design life expectancy.
Since operating labor requirements are very site-specific, estimation
of labor costs associated with each system is difficult. For simplification,
the cost analysis assumes labor and maintenance costs are a fixed
percentage of the overall capital cost. However, the air stripping system
is likely to require more operating attention since it is more complex.
For this example, treating 25 gpm of groundwater contaminated with
300 mg/L of TCE, the capital cost of an air stripper/vapor recovery
system is greater than the steam system but utility costs are lower.
While this cost picture will generally be the case in comparing these
two types of systems, the more concentrated the groundwater stream,
the more the economics will favor steam stripping. Since the
temperature of the entire waste stream must be raised to approximately
the boiling point of water in steam stripping applications, the utility
cost associated with steam stripping is proportional to the volume of
water treated and independent of the concentration of the contaminant.
For air stripping with vapor recovery, the amount of steam required
for regeneration of the carbon bed is proportional to the amount of
carbon present (which is proportional to the amount of VOC in the
groundwater), since the steam is used to increase the temperature of
the entire bed for desorption. Therefore, using our example of a more
concentrated stream (> 300 mg/L), steam stripping would offer greater
annual savings per pound of VOC removed, while at some concentration
below 300 mg/L, the air stripper would begin to offer more cost savings.
Of course, if waste stream is available, the utility cost of the steam would
be insignificant and steam stripping most likely would have lower
operating and capital costs.
System Performance and Environmental Protection
Emission control requirements for stripping units often are in the
range of 98% removal of VOC emissions, which approaches
state-of-the-art performance for such systems. Because of the difficulties
that can be encountered with the vapor recovery units associated with
air strippers, steam stripping can offer greater environmental protection.
Since steam stripping is a totally enclosed system, there are no air
contaminants released to the atmosphere. Also, steam stripping is not
prone to an air emission release as can occur in the event of an upset
in the vapor recovery system used for air stripping. Furthermore, the
additional emission source presented by an air Stripping/vapor recovery
system will increase the environmental regulatory burden of the operator
since it may involve stack emission testing, permit modification and
associated paperwork burdens. In addition to the cost savings involved,
the elimination of an air emission source with potential to emit
hazardous pollutants offers significant non-economic rewards.
Other Considerations
Some groundwaters have a high degree of hardness which can cause
scaling problems in both the air and steam stripping equipment. While
the hardness can be a problem for both systems, steam stripping systems
usually are more susceptible to scaling since they operate at higher
temperatures. Also, heat exchangers in steam stripping systems are
added potential problem areas for scaling. To strip groundwater with
high hardness, the water may need some type of pretreatment such
as pH adjustment or water softening.
Packed tower air strippers often plug due to biological growth
accumulation on the packing, a phenomenon that does not occur in
steam stripping. Biological fouling requires periodic shutdown of the
air stripping system to wash the internals with a hypochlorite solution
or some other chemical. This maintenance process adds to the operating
cost and requires periodic system shutdown for cleaning.
CONCLUSIONS
The preceding paper has demonstrated that, in many cases (especially
those cases of highly contaminated groundwater), steam stripping offers
many advantages over air stripping for cleanup. These advantages
include: cost savings, better controllability, easier recovery of the
contaminants and fewer operating problems. Moreover, data for design
of steam strippers are readily available from the literature, while design
of air stripping systems often requires pilot testing.
REFERENCES
1. J.M. Gosset, et al., Mass Transfer Coefficients and Henry's Law Constants
for Packed-Tower Air Stripping of Volatile Organics: Measurements and
Correlations, prepared by Air Force Engineering and Services Laboratory
at Tyndall Air Force Base, FL, June, 1985.
2. K. Sullivan, et al., Pilot Testing and Design of a Modular High-Temperature
Air Stripping System for Waste Cleanup, presented at Annual Technical
Meeting of the Water Pollution Control Federation, New Orleans, LA, Sept.,
1984.
3. U.S. EPA, CERCLA 301 "C" Study, Dec., 1984.
4. C.S. Parmele, O'Connell, W.L. and Basdekis, H.S., "Vapor-phase
adsorption cuts pollution, recovers solvents," Chem. Eng. 86(58), 1979.
VAPOR CONTROL 561
-------
In Situ Vapor Stripping: Preliminary Results
Of A Field-Scale U.S. EPA/Industry Funded Research Project
Robert D. Mutch, Jr., P.Hg., P.E.
Ann N. Clarke, Ph.D.
James H. Clarke, Ph.D.
Eckenfelder Inc.
Nashville, Tennessee
David J. Wilson, Ph.D.
Vanderbilt University
Nashville, Tennessee
INTRODUCTION
A 2-yr. long in situ vapor stripping research program is being con-
ducted at the CIBA-GEIGY Plant in Tbms River, New Jersey. The
research is being conducted by ECKENFELDER INC. and is co-
founded by the U.S. EPA Small Business Innovative Research Program
and CIBA-GEIGY Corporation. The research project, which began in
August of 1988. involves the closely monitored installation of in situ
vapor stripping technology. The research program calls for the vapor
stripping facilities to operate for a period of 1 yr. This paper reports
on findings through the first 10 mo. of operation.
The general objectives of the research program are: (1) to improve
the scientific foundation for this remedial technology; (2) to better de-
fine its technical limitations; and (3) to further refine the mathematical
model of the stripping process developed in an earlier phase of the
research program. Further objectives of the research are: (4) to study
this technology at the lower concentrations of volatile organic consti-
tuents as experienced near the errf of remedial actions, and (5) to evaluate
the performance of granulated activated carbon as a treatment agent
for the extracted vapors at these low levels.
Site Geology
The CIBA-GEIGY Toms River Plant lies in the Atlantic Coastal Plain
Physiographic Province in Toms River, New Jersey. The site is under-
lain by the Cohansey Sand, a geologic formation consisting
predominantly of moderate to high prcmeability sand, intcrbedded with
finer-grained, often lenticular, strata of silt and clay. The site chosen
for the research program lies within the central production area of the
1200 ac. plant she at the location of several recently demolished chemical
process buildings. Soil contamination was detected in the razing of the
buildings, presumably resulting from underground storage tank leaks
and process pipeline leaks.
INITIAL CHARACTERIZATION OF THE RESEARCH SITE
A drilling program was undertaken in order to characterize initial
levels of soil contamination within the study area. The program con-
sisted of 26 exploratory borings and collection and analysis of 40 soil
samples representing horizontal and vertical locations. The complete
menu of organic priority pollutant analyses was run. Table 1 lists the
specific chemicals detected and their respective concentration ranges
(including limits of detection). Not all chemicals were identified as
present at all probe locations or at all depths at a given location. No
acid extractable compounds were detected at a limit of detection of 2.0
ppm (mg/kg) with the exception of one sample which exhibited a phenol
value of 3.0 ppm. The most prevalent soil contaminants in the study
area are 1,1-dichlorobenzenc, 1,3-dichlorobenzene. 1,4-dichlorobenzene
and 1,2,4,-trichlorobenzcne.
of Chemkab Ifldentlfled to the Sofi
At The Study Site
r..Hj
*!<•! [Halt or ««<>etlon)
«••»• ef
Ceanatmlm
VoUtlll (ppbl
Tr cl»loro«Cliyl*A« (10)
1. -dlloro«lten« (10)
0> orMlluix (40)
H* hyifMoti*. (JO)
1... H.,,1,.1 ^M|4,n. ,,',""
Dt-n-tetylplKtuUt* (I)
floor UCIWIM (1)
PK.n.ntnr.n. (1)
fyr.i.. (1)
1 .2 F4-(r lcblerofr«nt«n« (1)
»l»(2-«lh,lh»»rl) pkttoUi. (1)
rir«w (1)
!««•«<> (l.>.3-<4) pyr.n. (1)
Anlhrftc*n« (1)
Mntoll) tkr«c«M (1)
t«n«o>(*) pyr»n« (t)
B«neo((hl) p^rflttn* (1)
»«njo(k) (Uor.nih.n. (I)
Chry.tn. (i)
l.)-<]|chlarob«r.c«n< (1)
1 . dichlorab«fi>«n« (1)
HjphtMUnc (1)
""""""""* U>
(Mk,
•HI
M-1.WD.On
m-ii
•O-Jt.S
UK-IS;
•D-M
m-S.l
»-J.4
TO- 13
TO-IS
HO- 11
Htt-W
HO- 10
Mb-1 4
m-i i
m-j i
w-i J
MB-* J
m-i.i
TO-4..1
m-j.o
H6-IOO
A-161
HO-J-*
m-n
Characterization of soils in the areas including and adjacent to the
proposed study area was performed prior to the initial site characteri-
zation. An area of high organic (semi-volatile) constituent content MB
identified approximately 35 ft from the location where the extraction
well was unplaced. This area of high organic concentration was desig-
nated "Test Pit 9". The chemical constituents present in the soil taken
from approximately S ft below the surface at Test Pit 9 are listed in
Table 2. Approximately 1,600 ppm additional semi-volatile constituenc
were estimated in a non-target library search.
RESEARCH PROGRAM FACILITIES
AND EQUIPMENT
One of the initial tasks was the installation of an extraction well in
the approximate center of the area of contamination. The well consisted
of a 4-in. diameter, 5-ft long, factory-slotted PVC screen which vns
set slightly above the water table on a 4-in. diameter PVC casing. In
the area of the project, the water table is at a depth of approximately
20 ft.
562 VAPOR CONTROL
-------
Table!
Chemical Constituents Identified In Soils From Test Pit 9
1,4 dichlorobenzene
1,2 dichlorobenzene
nitro benzene
1,2,4 trIchlorobenzene
naphthalene
2-chloro naphthalene
Concentration
(ppm)
31
300
21
200
55
27
A series of 38 soil gas probes was installed at radial distances of
approximately 20,40, 60 and 80 ft from the extraction well. A number
of the probes were constructed as clusters with individual probes at
depths of 5, 10 and 15 ft below ground surface. A sketch of a typical
soil gas probe is shown in Figure 1.
The probes were constructed of Teflon tubing and were installed by
a truck-mounted hollow steam auger rig. The screened section of the
probe was sand-packed and the remaining annular space was sealed
by bentonite pellets and grout. The probes allow for measurement of
in situ soil vacuum and also permit sampling of soil gas quality. Twelve
of the probes were fitted with thermisters to permit measurement of
the soil temperature.
TUBIHd VALVC
I/V TEFLON -
TUBING
PERFOflATED-
l/«" TEFLON
TUBING
,PRESSURE-VACUUM
MANOMETER
100 MESH SILKSCREEH
Figure 1
Sketch Of Typical Soil Gas Probe
Extraction of the soil gas vapors and much of the monitoring is per-
formed by ECKENFELDER INC.'s In Situ Vapor Stripping Pilot Unit.
The 8-ft by 12-ft long pilot unit trailer houses two New York Blower
Model 2606-A pressure blowers. Each blower utilizes a 26-in. aluminum
compression fan blade encased in a steel-frame housing and is powered
by a 7-12 hp, 460 v, 3-phase motor. At the rated 3500 rpm fan speed,
the two blowers produce 50.5 in. water column pressure on the outlet
at a flow rate of 400 scfm. The blowers and associated ducts are con-
figured for individual, series or parallel operation, depending upon flow
rate and pressure requirements.
The pilot unit trailer also contains a baffled demister to remove water
droplets from the air stream and instrumentation and controls for oper-
ation of the system and the monitoring of system performance. A lay-
out of the pilot unit trailer is illustrated in Figure 2.
There are five sampling ports in the duct work to allow sampling
of extracted gas quality at various points in the system. Measurements
of temperature and pressure can be taken remotely at each sampling port.
Treatment of the extracted gas is accomplished by use of granular
activated carbon. A carbon canister is set up outside the trailer as indi-
cated in Figure 2.
An HNU Model PI-201 photoionization monitor with an Esterline
Augus Model 410 chart recorder is utilized to continuously record gas
quality. An electronic control panel, in conjunction with a Masterflex
pump, automatically samples each of the five gas monitoring probes
and a calibration gas cylinder once every hour. The automatic sequencing
can be overridden if manual readings are desired. The HNU
photoionization-detected output data are stored on the chart recorded
for manual interpretation. The flow rate of the system is monitored by
means of a Dwyer pilot tube and micromanometer.
An air permit was obtained from the New Jersey Department of
Environmental Protecton in order to operate the vapor stirring system.
The permit established a maximum discharge concentration of 50 ppm
total volatile organic compounds.
PRELIMINARY FINDINGS
The preliminary findings of the research project center upon the
measured zone of influence of the extraction well, the change in the
quality of the extracted soil gas with time, the treatability of the extracted
gas by means of the granulated activated carbon system, temperature
variations occurring in the system and the observed rate of the ground-
water table induced by the vacuum extraction.
The preliminary findings in each of these areas are briefly discussed
below.
Zone of Influence
Mathematical modeling of the in situ vapor stripping process indi-
cates that in an isotropic soil the zone of influence of an extraction well
screened near the base of the unsaturated zone should produce a zone
of influence with a radius approximately equal to the depth of the well
(i.e., unsaturated zone depth). In a soil with vertical anisotropy, the
radius of the zone of influence is proportional to the degree of anisotropy.
Because the Cohansey Sand was expected to have a vertical anisotropy
of two to three, the in situ soil gas monitoring probes were set out at
radial distances of ID, 2D, 3D and 4D, where "D" equals the depth
to the water table (or well depth).
Soil gas extraction was commenced on Sept. 6, 1988, at a rate of
180 cfm. In situ soil gas vacuum levels were observed almost imme-
diately throughout the study area and reached a steady-state condition
in less than 15 min. The in situ vacuum levels have remained essentially
constant throughout the course of the research program. Contours of
in situ vacuum levels are depicted in plan view and in cross-section
in Figures 3 and 4, respectively. As indicated in these figures, a wider
zone of influence was established than anticipated, even considering
the vertical anisotropy of the Cohansey Sand. It can be extrapolated
from the measured in situ vacuum levels to be approximately 150 ft.
This distance is more than twice the anticipated radius of influence.
In order to determine the causative factor behind this observed
phenomenon, an attempt was made to calibrate Wilson's two-
VAPOR CONTROL 563
-------
dimensional, axial-symmetric, ISVS numerical model 10 ihe measured
in situ vacuum levels'. A finite difference grid which was 165 ft. wide
along the r axis and 23 ft. deep along the / anix was sci up. A uniform
grid spacing of 3.3 ft. was used along both axes, producing u total of
350 nodes. Because the model currently simulates an extraction well
as a sphere, the 5-ft long, 4-in. wide well screen and surrounding gravel
pack were represented as a sphere with a radius of 1.2 ft. The following
parameters were also held constant during the modeling:
Extraction rate = 180 scfm
Depth of extraction well intake = 22 ft.
Soil porosity = 0.25
Temperature = 63 °F
Extraction well vacuum = 0 94 atm
It was quickly discovered in the modeling process (hat no combina-
tion of soil permeability and vertical amsotmrn, would permit the model
to adequately reproduce the observed in situ vacuum levels The
modeling results and Held data suggested (hat observed in situ vacuum
levels were attributable to more complex hydrogcologic conditions than
simple vertical anisotropy.
The numerical model was therefore revised to permit modeling of
dual layer stratigraphy with varying anisoiropics in each layer The
calibration was then continued. An excellent match to the ft
was obtained when two ml layers were employed in the mode
upper, 3.3-ft. thick layer with an air permeability of 6.5 x 10' i
and a lower, 20-ft. thick layer with an air permeability o
m1/aim-sec A comparison of predicted and measured data is |
ed in Figure 5.
Because the upper layer is 191.5 times less permeable than th
layer, vertical anisotropics within the individual layers are not
cant to the results of the modeling. Air flow is nearly vertica
upper layer, thereby negating (he importance of lateral permeal
this stratum. Similarly, air flow is nearly horizontal in the lowc
correspondingly diminishing the importance of vertical perm
in this layer.
The apparent reasons for this phenomenon can be tied to th
cial conditions As mentioned earlier, (he research project is
the location of previous production buildings. Approximately'.
(he surface <>f the study area is comprised of I0-fl.:, 5-ft. thic
crctc footings. Moreover, the interveming soil around the foo
(o .1 large extent Till material of a finer-grained character U
underlying native soils of (he Cohansey Sand. Consequently, th
few feet of soil impede the influx of atmospheric air from the s
causing the zone of influence to spread laterally beneath this s
layer.
CONTROL BOXES
FOR BLOWERS
EXHAUST
V.O.C. COLLECTION
SYSTEM
PITOT TUBE
INSULATED EXHAS
EXTRACTION WELL
-FLEXIBLE DUCT
NOT TO SCALE
A VALVES
*i BLOWERS
- FLEXIBLE DUCT CONNECTION
CD SAMPLE PORTS
'LOCATED BELOW TRAILER FLOOR
DIRECTLY BELOW DEMISTER
' OPTIONAL SYSTEM SPECIFIC
TO EACH APPLICATION
Figure 2
Layout Of In Situ Vapors Stripping Pilot Scale Research Trailer
564 VAPOR CONTROL
-------
Extracted Soil Gas Quality
The quality of soil gas extracted during the first 10 mo. of the research
program is presented in Figure 6. The soil gas concentrations are
reported as a function of the days of system operation. Days of system
shut-down for maintenance and installation of additional granulated
activated carbon canisters are omitted from the graph.
Extracted soil gas concentrations initially were in the range of 100
to 140 ppm and have fairly steadily declined to current levels of ap-
proximately 60 to 70 ppm. Chemcial analysis of the extracted soil gas
reveals that the principal gas contaminants are: 1,1-dichloroethane,
1,1,2,2,-tetrachloroethane, 1,1,1-trichloroethane, trichloroethylene, 1,2,-
dichlorobenzene and 1,3-dichlorobenzene. Upon completion of the
testing, only 1,1,1-trichloroethane (TCA) and toluene were present in
measurable quantity in the vapor from the extraction port. Below method
detection limit quantities of TCA were noted in probe 11D.
Figure 7 presents a graph of discharge gas quality after granulated
activated carbon treatment. The five peaks in the graph represent
progressive exhaustion of the granulated activated carbon canisters.
These peaks represent exhaustion of the typical 1,200 Ib of granulated
activated carbon used in the project. The peaks represent a more rapid
exhaustion of a standby granulated activated carbon system consisting
of two parallel 55-gal drum carbon canisters.
The treatment efficiency of the granular activated carbon has been
significantly diminished by sorption of water vapor in the carbon. The
demister has removed relatively little water since the water occurs in
the form of water vapor rather than as a mist.
Temperature Variations
The temperature of the extracted gas, ambient air and the soil have
been measured throughout the course of the study. Figure 8 depicts
the variations in extracted soil gas temperature and ambient tempera-
ture. The temperature of the extracted soil gas was initially approxi-
mately 64 °F (18 °C) and has steadily declined during the fall and
beginning of winter to temperatures of between 52 °F (11 °C) and 54 °F
(12 °C). December through March exhibited the lowest gas and ambient
air temperatures recorded. There was a steady increase exhibited
throughout the spring as anticipated. June, 1989 temperatures were about
9° below last August, 1988 readings.
Figure 9 is a graph of in situ soil temperature variations occurring
within the study area. The graph illustrates that, initially, soil gas
temperatures were in the range 64°F (18 °C) to 72 °F (22 °C). Also, the
deeper soil probe (ID) exhibited a consistently lower temperature than
the intermediate (II) and shallow (IS) probes. This result is not sur-
prising considering the time of year. With the onset of fall and winter,
in situ temperatures declined and reversed their relative positions. The
deeper probe, probe ID, exhibited the highest temperature and the shal-
low probe, probe IS, the coolest temperature. Upon the arrival of spring,
the relative relationships were again reversed.
it — LOCATION cr ton. «AI r*OKi
\t.<)—— UCASUHEO IN SITU SOt. VACUUM
ecwTxira or IXSITU JCHL VACUUM
Figure 3
Contours Of In Situ Soil Vacuum
VAPOR CONTROL 565
-------
65r
: so
- 55
SO
45 -
- 4O
EXTRACTION
WELL
©
0
_— — to — -
_- '
f
1
rzi^H
— —
U
LEGEND:
SOIL GAS PROBE
MEASURED IN SITU SOIL
, VACUUM
^^ SCREEN SETTINO
-, 0 CONTOUR OF IN SITU
SOIL VACUUM.
Figure 4
Cross-SecUonal In Soil Silu Vacuum Contour
85
40
38
30
tXTHACTIOM
•ILL
- - " - 0,^ ,
Mil1'
L_l_._sJi. lj.±i
"-to-
SOIL 8A3
MCAtUMO IN SITU SOIL
VACUUM IN IMCMC* OF WATCK.
9CNCEN KTTINO
• MCASUNCO CONTOUR Of
IN SITU SOIL VACUUM.
10 1 0
CONTOURS Of
IN SITU SOIL VACUUM.
Figure 5
Comparison Of Measured And Predicted In Silu Soil Vacuum Levels
566 VAPOR CONTROL
-------
a
a
z
o
p
UJ
o
z
o
o
i
80
120 160
DAYS OF OPERATION
200
240
Figure 6
Extracted Gas Quality
Groundwater Levels
A rise in groundwater levels beneath in situ vapor stripping facilities
has been both predicted and observed. The phenomenon results from
the feet that the groundwater table represents the point in the subsur-
face where the voids in the soil or rock are not only fully saturated,
but also at equilibrium with atmospheric pressure. Consequently, if soil
gas pressures are reduced to below atmospheric pressure, a corre-
sponding rise in the groundwater table should result. The magnitude
of groundwater table rise (in inches) should coincide with magnitude
of the pressure drop below atmospheric occurrring at any point in the
system (in inches). Monitoring of groundwater levels during the course
of the research study confirms that the water table does indeed rise
a level commensurate with the soil vacuum levels produced by the
extraction well. The maximum rise in the water table of nearly 2.5 ft
occurred immediately beneath the extraction well.
MATHEMATICAL MODELING
A mathematical model has been developed for predicting various
aspects of a full-scale in situ varpor stripping system. ' This model has
been calibrated to the conditions of the Toms River field site. The model
was originally calibrted using laboratory data generated from specially
designed equipment which simulated actual field parameters and
operating conditions 2 A model parameter critical to the estimation
of chemical mobilities is the determination of a lumped partitioning
coefficient for each chemical constituent of interest. This coefficient
addresses the constituent's interacton with water, soil and other chemi-
cals present and dictates the constituent's strippability.
The model can be used to generate important design criteria and
optimize operating parameters form pilot-scale studies for use in full-
scale remediations. The model can be run on a PC. A list of the model
capabilities is provided in Table 3.
VAPOR CONTROL 567
-------
a
a
ui
O
z
o
o
70
60 -
30
40 -
30 -
20 -
10 -
NEW CARBON CANISTERS INSTALLED
PERMITTED DISCHARGE CONCENTRATION-
40
BO
DAYS OP OPERATION
Figure 7
Discharge Cms Quality
240
o^
UJ
§
UJ
UJ
SYSTEM SHUTDOWN
TOR MAINTENANCE
25-Aug 04-Oct 13-Nov 23-D«c 01-Feb 13-Mor :2-
-10
DAYS OF OPERATION
01-Jun
+ EXTRACTED '.'•:,
Figure 8
Exlracicd Gas and Ambient Temperatures
568 VAPOR CONtROL
-------
o
•wx
UJ
e
UJ
0.
SYSTEM SHUTDOWN
FOR MAINTENANCE
25-Aug 04-Oct 13-Nov 23-Doc 01 -Feb 13-Mar 22-Apr
D PROBE 1<
DAYS OF OPERATION
+ PROBE II
o PROBE 1D
Figure 9
Soil and Ambient Temperatures
Table 3
List Of Model Capabilities For The Prediction Of Design And
Operating Parameters For Full-Scale In Situ Vapor Stripping1*3"6
. Predict clean-up time to reach a target level of residual
contamination.
, Predict residual contamination levels after a given period of
operation.
, Predict location of hot spots through diagrams of contaminant
distribution.
Develop system design:
horizontal well placement
vertical well placement
screen placement
Predict impact of implemeable cap placement
Predict impacts of passive wells
Predict vapor stripping from fractured bed rock.
Predict clean-up levels around buried debris from various system
designs.
Predict impact of ambient air temperature on removal.
Calculate the anisotropy of the soil or rock.
Predict recontamination time of the remediated vadose zone from slow
moving contaminated groundwater.
12. Predict the rate of remediation of floating pools of LNAPLs.
FURTHER RESEARCH OBJECTIVES
While our findings to date have answered a number of questions con-
cerning the behavior of in situ vapor stripping systems, several of the
research objectives remain to be accomplished. These unfulfilled
objectives include the following:
• Description of the temporal variations in overall gas quality, as well
as the relative proportions of individual constituents within the gas
stream
• Determination residual levels of various contaminants in die soil at
the conclusion of the project
• Description of the relationship between extracted gas flow and the
resultant zone of influence and impact upon cleanup times
REFERENCES
1. Wilson, D.J., Clarke, A.N. and Clarke, J.H., "Soil Clean Up by In Situ
Aeration. I. Mathematical Modeling", Sep, Sci. Technol., 23, pp 991-1,037,
1988.
2. AWARE Incorporated, Phase I, Zone I Soil Decontamination through In Situ
Vapor Stripping Processes!, U.S. EPA, SBIR, Contract No. 68-024446, Apr.
1987.
3. Wilson, D.J., Gannon, K., Clarke, A.N., Mutch, Jr., R.D. and Clarke, J.H.,
"Soil Clean-up by In Situ Aeration. II. Effects of Impermeable Caps, Soil
Permeability, and Evaporative Cooling". (Sep. Sci. Technol.)
4. Wilson, D.J., Clarke, A.N. and Mutch, Jr., R.D., "Soil Cleanup by In Situ
Aeration, in Passive Vent Wells, Recontamination, and Removal of Under-
lying NAPL", (Sep. Sci. Technol.)
5. Mutch, Jr., R.D. and Wilson, D.J., "Soil Clean-up by In Situ Aeration. IV.
Anisotrophic Permeabilities", (Sep. Sci. Technol.)
6. Wilson, D.J., "Soil Clean-Up by In Situ Aeration, V. Vapor Stripping from
Fractured Bedrock", (Sep. Sci. Technol.)
VAPOR CONTROL 569
-------
Mathematical Evaluation of Volatile Organic Compound Transport
Via Pore-Space Dispersion Versus Advection
Mark J. Lupo, Ph.D.
K. W. Brown & Associates, Inc.
College Station, Texas
ABSTRACT
The environmental behavior and fate of hazardous organic waste con-
stituents in an unlined landfill was modeled in order to determine the
speed and effectiveness of dispersion in the unsaturaied pore space of
soil as a contaminant transport mechanism In addition to the well-known
effects of downward advection (leaching) of contaminants in the water
phase and upward air dispersion into the atmosphere, the results show
the potential of downward dispersion of hazardous chemicals in
unsaiurated pore space as an important contamination pathway. This
is due to the speed at which these constituents are transported by this
mechanism.
The Vadose Zone Interactive Processes (VIP) computer model was
used to simulate the migration and decay of several hazardous waste
constituents in a variety of soil types based on physical and chemical
properties of the constituents and local conditions (climate, soil proper-
lies, etc.). The model used partition coefficients to distribute the
hazardous constituents into four "phases;" the soil, the waste, the water
and the air phase.
The constituents were shown to move rapidly through the air phase.
even if the leaching of the hazardous constituent was retarded by
adsorption onto the soil. The constituents repartition into the water phase
at detectable concentrations several meters below the lowest extent of
the contamination by leachaic advection. This effect is more pronounced
in dryer soil, because the dispersion of organic constituents in the soil
is a strong function of air porosity.
If this transport mechanism is neglected in a system where it is
important, numerous misinterpretations are possible. For example, the
high rate of transport could cause the misidemification of fractures.
Groundwater monitoring data can underestimate the eventual concen-
tration of constituents by orders of magnitude or give misleading clues
as to the size of the mother lode because the transport was caused by
a rapid but less efficient mechanism than leaching. The rate of trans-
port can be underestimated by orders of magnitude if unsaiurated flow
equations are considered alone. Remedies could be chosen that limit
liquid phase flow when the air phase is the key transport route.
This study shows that it is not sufficient to consider the exient or
rate of transport of liquids in the soil in order to characterize soil or
groundwatcr pollution at a Superfund site. Motion in the unsaiurated
pore space in the soil also must be taken into account.
INTRODUCTION
There are several contaminant transport mechanisms for volatile
organic compounds in unsaiurated zone soils: advection in the pore
water, non-aqueous phase advection, dispersion in the pore water and
dispersion in the unsaturatcd pore space. Nevertheless, it is common
practice to consider migration only in terms of advection in the soil
pore water or as a distinct organic phase.
In this study, a comprehensive computer model was used to simulate
the fate and transport of two common volatile constituents of hazardous
wastes in a variety of soil types and conditions. The objective was to
quantitatively evaluate the transport mechanisms in terms of speed and
effectiveness.
570 VAPOR CONTROL
HYPOTHETICAL CASE
A hypothetical case was devised for (he purpose of this investigation.
Figure I illustrates an unlined landfill with barrels of hazardous waste.
not atypical of many sites that are on or may be added to the NPL.
The barrels are buried at a depth of approximately 5 m below the sur-
face, and the water table is located at a depth of K) m below the surface.
only 5 m from (he waste. In this hypothetical case, most of the barrels
have leaked, and the soil is heavily contaminated from a depth of 4.5 m
to 55 m. Soil cores from this depth contain 100 mg/kg of benzene and
chlornben/cne in addition R> other hazardous organic constituents
OO-i
so
GMOUNOMIDt
UtCOtt
OMTAMMAIBI IOC I
GMMNDnlE*
Figure I
Unlined, Unregulated Landfill. S m Deep
The spread of chJorobenzene and benzene from the contaminated soil
layer (Fig. 2) was computed for four soil types and three different water
budgets. The soil types chosen were sand, sandy loam, loam and clay
Table I lists typical properties for these soil types'. The three water
budgets, representing low, high and excessive recharge rates, yielded
deep percolations of K)-1, K)1, and K)J in,day. respectively (1.43, W.3
and 143 in./yr).
MATHEMATICAL METHOD
Organic molecules in the unsaturated soil can adsorb onto the organic
matrix on soil grains. They also can reside in soil pore water, the
unsaturated pore space or in any free organic or oil phase that often
can be found in contaminated soils at CERCLA sites. For this reason,
in evaluating the fate of organic contaminants (especially the more vola-
tile compounds), it is necessary to quantify the behavior of the hazardous
constituents for each of these four phases. In this study, the phases will
be referred to as the soil, water, air and oil phases, respectively.
-------
Ł4.5
Ł 5.0
& 5.5
10.0
SURFACE-
QO-I
0.3- -2i""i^M
.•.v AEROBIC ZOflT?
CROUNDWATCR
AEROBIC ZONE
CONTAMINATED ZONE
GROUNDWATER
Figure 2
Initial Condition
Figure 3
Partitioning Between Phases in Unsaturated Soil
Table 1
Properties of the Hypothetical Soils Used in this Study
Soil Type
Sand
Sandy Loam
Loam
Clay
Empirical
Constant
4.05
4.90
5.39
11.40
Porosity
(percent)
39.5
43.5
45.1
48.2
Dry Bulk
Density
(gr./cc.)
1.61
1.50
1.45
1.37
Sat. Hydraulic
Conductivity
(cm. /sec. )
1.76 x 10~2
3.95 x 10"3
6.95 x 10~4
1.28 x 10~4
To quantify the fate and transport of volatile organic constituents,
it is necessary to solve a suite of four partial differential equations. These
equations would express the change in the concentration of a contaminant
in each phase over time as the sum of advection, dispersion, adsorp-
tion/desorption and degradation. The solutions of these equations will
be four functions of time and space, each representing the concentration
of the contaminant in one of the phases.
Clearly, it would be very difficult to solve such a suite of partial
differential equations analytically. Therefore, it was necessary to find
a computer model that solved these equations numerically. The model
would have to account for advection, dispersion, degradation, the par-
titioning between the phases and allow the input of initial conditions
as well. The Vadose Zone Interactive Processes (VIP) model has all
of these characteristics.
The VIP model has been described in detail in the literature2-3 It
has been tested mathematically4, experimentally5 and with field data6.
VIP treats the soil system as a four-phase system. The four equations
are coupled by the adsorption/desorption terms, which contain the
partition coefficients. Model inputs include the soil-water partition
coefficient K^ (often referred to as Kd in the literature), the air-water
partition coefficient Km and the oil-water partition coefficient K^. Be-
cause partition coefficients in the literature almost always include water
as a partition phase, the authors of the VIP model chose to assume
that molecules migrating from one phase to another must pass through
the water phase (Fig. 3).
In addition to adsorption and desorption, the model allows for degra-
dation in each of the four phases. Migration is carried out mainly by
advection in the water phase and dispersion in the air phase. Disper-
sion in the water equation was set equal to zero because all interphase
constituent motion in the model must pass through the water phase,
and the adsorption/desorption process provides dispersive phenomena
sufficient to simulate the data observed by the modelers2. Advection
in the air phase was set equal to zero by the author since it is respon-
sive to large barometric pressure changes that are not expected in the
general case.
MODEL INPUTS
The inputs into the VIP model include soil properties (Table 1), climatic
data (Table 2) and chemical properties (Table 3). The soil-water partition
coefficients presented in Table 3 were based on literature values for organic
carbon-water partition coefficients9'10 and a fraction organic carbon of
1%" The soil-water partition coefficients for soil below a depth of 0.3
m are a factor of 10 lower, because the fraction organic carbon at this
depth is assumed to decrease to 0.1%.
Month
January
February
March
April
May
June
July
August
September
October
November
December
Temperature
0.0-0.3 meters
(Deg. Fl*
53.6
55. B
61.3
68.5
76.0
81.6
83.0
83.2
79.2
71.4
60.6
55.7
Table 2
Soil Temperature
Temperature
0.0-0.3 meters
(Deg. C)
12.0
13.2
16.3
20.3
24.4
27.6
28.3
28.4
26.2
21.9
16.0
13.2
Temperature
0.3-10.0 meters
(Deg. C) '
23.5
22.5
21.5
20.5
21.5
22.5
23.5
24.5
25.5
26.5
25.5
24.5
Average monthly temperatures for Houston, Texas (7).
Based on the Fluker model for a depth of ten feet (B).
Table 3
Chemical Properties
Property
Decay Rate (I/day)
Octanol-water Partition Coefficient
Air-water Partition Coefficient
Soil-water Partition Coefficient
Chlorobenzene
2 x 10"4
512.9
.146
3.89
Benzene
1 x 10"2
129
.224
0.89
The oil-water partition coefficients were octanol-water partition
coefficients9"". Air-water partition coefficients were computed using
Henry's law. Degradation rates obtained from the literature'2-13 were
reduced by an order of magnitude in Table 3 because biodegradation
is primarily anaerobic at depths greater than 0.3 m14
Air dispersion coefficients were obtained from the literature7 for
chlorobenzene and benzene and were corrected for site temperature.
The soil correction term for the air dispersion coefficient15 differed for
VAPOR CONTROL 571
-------
each of the 12 cases modeled (four soil types and three water budgets).
Therefore, it had to be computed for each case based on the varying
total porosity and air porosity in each soil (Table 4).
Table 4
The Size of the Phases in Each Soil Computed by the Model
TW»le6
VIP Results for Chlorobenzene Without Air Dispersion
Ca»
Soil
(percent)
Mater
(percent)
on'
(percent)
Mr
(percent)
Excessive recharge:
Sand
Sandy loam
Loam
Clay
High recharge
Sand
Sandy loam
Loam
Clay
Low recharge:
Sand
Sandy loan
Loam
Clay
60. 5
56.5
54 9
51 .6
60.5
56. 5
14.9
51. 6
60.5
56.5
54.9
si. e
20.4
27 6
33.5
41.9
16.6
:j.O
?8.3
40.2
13. S
19.2
24 0
36,1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
19
15
11
4
22
20
It
1
25
24.
21,
11.
.0
.a
.5
.2
.8
.4
.7
.9
.9
.2
.0
.4
The oil phase MAS included to 3tab!HE* the
•1 at the bound*ri«a.
MODEL RESULTS
A total of 48 simulations was conducted, half for benzene and half
for chlorobenzene. For each constituent, four different soil types were
used in the model and three different water budgets. The 12 simula-
tions were run with and without air dispersion to determine the conse-
quences of neglecting this transport mechanism.
The simulations were for a period of 1 yr. The chlorobenzene results
are presented in Tables 5 and 6. The benzene results are presented in
Tables 7 and 8. By comparing the results of the VIP simulations, with
and without air dispersion, it is immediately clear that the air disper-
sion transport mechanism is far more rapid than advection in moving
volatile organic constituents in the water phase.
Tables
VIP Results for Chlorobenzene with Air Dispersion
Simulation
Sand
Sandy loan
Loan
Clay
High recharge:
Sand
Sandy loan
Loam
Clay
Low recharge:
Sand
Sandy loan
Loan
Clay
Percent
Degraded
5.12
5 20
5,31
5.49
4.99
5 06
5.16
5.39
4.89
4.74
5.06
5.31
Depth
Reached
(metera)
Releaae
Releaae
Releaae
7.*
ftaleaao
Release
Releaae
8.«
Release
Releaae
Release
Release
Releaae
to Mr
(percent)
4.54x10°
4.87xlO"5
5.62X10"7
0.0
1.47x10"'
2.32x10"*
S.lOxlO"4
0.0
4.82X10"1
1.67x10"*
2.78x10"*
2.16«10"6
Releaae
to Groundvater
(percent)
1.28x10"*
3.01x10"'
4.02x10"*
0.0
1.52x10"*
2.44x10"'
I.8
-------
6-
5-
H
^
3
Ul
CO
^f
a,
Ld J-
o:
o
o
2-
1-
0-
1 i
_L1 Pi
WAD WAD WAD WAD
CLAY LOAM SANDY SAND
LOAM
SURFACE
LEGEND
' —
CLAY
LOAM
SANDY LOAM
SAND
Figure 4
Release vs. Soil Type
concentration of the contaminant increases by orders of magnitude as
one approaches the source from the water table. At every level, the
contaminant in the air phases partitions with the other phases, con-
taminating the soil and the soil pore water (Fig. 5). Thus, the phases
besides the air are also contaminated outside of the soil that had been
in contact with primary leachate. Contaminants delivered to soil and
soil pore water by the air phase are free to advect and disperse. Thus,
the air dispersion mechanism not only moves contaminants, but also
augments other transport mechanisms.
A clear illustration of this augmentation can be seen with benzene
under excessive recharge. For the sand, the sandy loam and the loam,
release would have occurred without air dispersion because of water
advection. In Table 8, the size of the release is small in each case,
because the concentrations represent the fringe of the downward-moving
contaminant plume (given a few more months, the release would have
been much greater as the bulk of the plume crossed the 10 m datum).
\et in Table 7, the predicted releases for these three cases are large,
with concentrations in the groundwater in the mg/L. This is due to the
advection of benzene in the water phase that had been carried to lower
strata by dispersion in the air phase.
Air dispersion also is active in the upward direction. In drier soils,
in which dispersion is more rapid, there is a symmetry in the concen-
tration of contaminants in each phase, above and below the contaminant
source (Fig. 6). This symmetry results from simultaneous upward and
AIR DISPERSION
TO OIL
TO SOIL
TO WATER
CONTAMINATION
TO WATER
TO OIL
TO SOIL
AIR DISPERSION
GROUNDWATER
Figure 5
The Contamination of All Fourphases of the Soil by Air Dispersion
downward dispersion. Rapid advection (as in the case of higher recharge
rates) and faster degradation (as in the case of benzene) decrease the
amount of symmetry.
Hazardous constituents can be released to the atmosphere by upward
dispersion in the air phase. But this mechanism also can contaminate
soil and soil pore water above the contaminant source. Thus, if the
barrels were covered with clean soil during the operation of this land-
fill, the soil above the barrels could be quite contaminated. For example,
one year into a simulation involving loam with a recharge rate of 1.43
in./yr, soil can be contaminated in the mg/L range with chlorobenzene
as much as 2 m above the barrels (Fig. 6).
DISCUSSION
One may underestimate the importance of the effect of dispersion
of volatile organic contaminants in the air phase by reasoning that the
mechanism, although much more rapid than advection, is much less
effective than advection because it transports smaller masses. It could
be argued that the hazardous waste constituents released by dispersion
eventually would have been released by advection in the water phase
anyway. Although such a view might be justifiable in the sense of long-
term damage to the aquifer, it fails to address several key issues from
the standpoint of Superfund.
First, in the negotiations between PRPs and the Agencies, it is in
the interest of both sides to have an accurate understanding of the site
soil system in order to implement an effective remedy. An overly
optimistic view of the situation can lead to an ineffective cleanup.
Remedies could be chosen that limit liquid phase migration when the
air phase is an important transport route.
For example, neglecting air dispersion could cause an engineer to
conclude that benzene would degrade before it could be released.
Neglecting this mechanism can lead to overestimates in the time avail-
VAPOR CONTROL 573
-------
AIR DISPERSION ACCOUNTED FOR
-3
-2-10 1
LOG CONCENTRATION (ppm)
Figure 6
Chlorobenzcne in Dry Loam After One Year
able to consider options before a release to groundwaier. One could
conclude from testimony that a site rich in volatiles could be capped
with clean soil, neglecting the potential for contamination of that soil
from below. Testimony by site personnel that they used clean soil to
cover the barrels could lead to lax application of safety rules on-site
in dealing with shallow soil.
Errors on the side of optimism can lead to errors that impact the
environment as well as humans. Highly dispersive constituents with
lower octanol-water partition coefficients such as xylene may be assumed
to decay in place. Yet relatively low concentrations initially reaching
an aquifer directly below a site could become substantial over time,
especially if the landfill is large in area. It also should be remembered
that VIP is only one-dimensional. It may be accurate in the middle of
such a hypothetical landfill, but at the edges there will be lateral dis-
persive transport of volatile organic compounds as well as vertical trans-
port. If advection is the only transport mechanism considered in a site
rich in volatiles in a densely populated area, it would be easy to consider
only downward transport, completely missing the peril to homeowners
from horizontal motion of the constituents.
Another serious error of interpretation could occur if detections in
monitoring wells were assumed to be caused solely by advection. The
magnitude of the detections could be orders of magnitude lower than
they will be when the advective plume finally reaches the monitoring
well network. After a few well sampling events, the size of the mobile
contaminant mass or of the source load could be grossly underestimated.
A risk assessment based on such monitoring well data clearly would
be misleading.
A pessimistic view can cause errors in selecting a remedy. It can
discourage the use of an effective remedy in favor of a lesser one more
suited to the pessimistic view of site conditions. For example, a site
characterization could yield parameters for an unsaturated flow model
that would erroneously predict the time that organic contaminants would
reach an aquifer by advection. A suite of water samples taken from
the aquifer could then show contaminants in places where they should
not be. This finding could lead the investigators to assume fracture flow
where it is not taking place, or conclude the presence of a highly trans-
missive zone. The presence of such fractures could have an adverse
influence on the Agency or the PRPs in selecting a remedy.
It should be pointed out that some of the assumptions in this study
were conservative. For example, the concentration of the contaminants
in the source area in this simulation was only 100 mg/L. Data for clays
in this study did not take into account the effect that organic constituents
have on clays"11 Organic constituents can desiccate clays and create
fracture systems in the clary. Thus, results for the clays presented in
the tables and figures may be optimistic.
Furthermore, if the VIP model is used to conduct site-specific simu-
lations with actual soil parameters, it should be pointed out that degra-
dation is not the end of concern for halogenated hydrocarbons. The
daughter products of halogenated hydrocarbons are often also halo-
genated hydrocarbons" Unlike the hazardous constituents found in
petroleum refinery wastes, the daughter products of chlorinated
hydrocarbons may be more hazardous than their parent compounds.
Finally, Figure 7 shows the concentration of chlorobenzene released
in one year versus the dispersion coefficient of chlorobenzene in the
soil for Rather, the size of the release differs from soil to soil. There-
fore, no formula can be derived to present release strictly as a function
of dispersion coefficient. The time and magnitude of die release cm
only be obtained by actually conducting a model run. It should be
remembered that all predicted releases are functions of time. In Tables
5 through 8, values are presented for a time 1 yr after the start of the
simulation. When the center of the advection plume reaches the water
table, the bulk of the release will take place. For chlorinated hydrocar-
bons such as chlorobenzene, a release is an eventuality in spite of bio-
degradation, because the daughter products of chlorobenzene decay are
also chlorinated hydrocarbons.
0 -
Ul
SAND
•s. SANDY LOAM
A«. LOAM
DC CLAY
Figure 7
Chlorobenzene, Log Dispersion vs. Log Release
CONCLUSIONS
The release of two selected volatile organic compounds from a
hypothetical unlined hazardous waste landfill was modeled using a
574 VAPOR CONTROt?
-------
detailed unsaturated zone model. It was found that dispersion of the
constituents in the air phase was rapid, and the size of the release was
significant in some cases. Coarse soils with lower recharge rates (i.e.,
greater unsaturated pore space) allowed the greatest transport by this
mechanism. In clays or wetter soils, advection in the air phase was less
effective.
The model results show that at a Superfund site, it is not sufficient
to consider contaminant transport mechanisms of liquids alone. Failure
to take motion (both vertical and horizontal) in the unsaturated pore
space into account can lead to errors, both in site assessment and in
remedy selection.
ACKNOWLEDGMENT
The author would like to thank Gordon Evans for helpful comments
in the preparation of this manuscript.
REFERENCES
1. Clapp, R. B. and Hornberger, G. M., "Empirical equations for some soil
hydraulic properties," Water Resources Research, 14, pp. 601-604, 1978.
2. Caupp, C. L., Grenney, W. J. and Ludvigsen, P. J., A Model for the Evalua-
tion of Hazardous Substances in the Soil, Version 2, Civil & Environmental
Engineering, Utah State University, Logan, UT, 1987.
3. Grenney, W. J., Caupp, C. L., Sims, R. C. and Short, T. E., "A mathe-
matical model for the fate of hazardous substances in soil: Model descrip-
tion and experimental results," Hazardous Hbste and Hazardous Materials,
in press, 1987.
4. Yan, Z., Evaluation of the Vadose Zone Interactive Processes (VIP) Model
Using Nonequilibrium Adsorption Kinetics and Modification of VIP Model,
Thesis, Department of Civil and Environmental Engineering, UT State
University, Logan, UT, 1988.
5. Symons, B. D., Sims, R. C. and Grenney, W. J., "Fate and transport of
organics in soil: model predictions and experimental results," JWPCF, 60(9),
pp. 1684-1693, 1988.
6. Stevens, D. K., Grenney, W. J., Yan, Z. and Sims, R. C., Sensitive Parameter
Evaluation for a Vadose Zone Fate and Transport Model, U.S. EPA Rept.
No. EPA 600/2-89/039, US. EPA, Ada, OK, 1989.
7. Thibodeaux, L. J., Chemodynamics: Environmental Movement of Chemi-
cals in Air, Water and Soil, John Wiley and Sons, New York, NY, 1979.
8. Fluker, B. J., "Soil temperatures," Soil Science, 86, pp. 35-46, 1958.
9. Karickhoff, S. W., Brown, D. S. and Scott, T. A., "Sorptionof hydrophobic
pollutants on natural sediments," Water Resources, 13, pp. 241-248, 1979.
10. Roberts, P. V., McCarty, P. L., Reinhard, M. and Schreiner, J., "Organic
contaminant behaviour during groundwater recharge," JWPCF, 52, pp.
161-172, 1980.
11. Schwarzenbach, R. P. and Giger, W., "Behavior and Fate of Halogenated
Hydrocarbons in Ground Water," in Ground Water Quality, ed. C. H. Ward,
W. Giger, and P. L. McCarty, pp. 446^-71, John Wiley and Sons, New York,
NY, 1985.
12. Environmental Research and Technology, Inc., The Land Treatability of
Appendix VIII Constituents Present in Petroleum Industry Wastes, Document-
B-974-220, 1984.
13. Office of Air Quality Planning and Standards, Air Release Screening
Assessment Methodology, U.S. EPA Draft, 1987.
14. Brown, K. W., and Donnelly, K. C., "Influence of the soil environment
on biodegradation of a refinery and a petrochemical sludge," Environ. Poll.
(Series B), 6(2), pp. 119-132, 1983.
15. Farmer, W. J., Igue, K. and Spencer, W. F, "Effects of bulk density on
the diffusion and volatilization of dieldrin form soil," Journal of Environ-
mental Quality, 2, pp. 107-109, 1973.
16. Anderson, D. C., Brown, K. W. and Thomas, J. C., "Conductivity of com-
pacted clay soils and organic liquids," Waste Management and Research,
3, pp. 339-349, 1985.
17. Brown, K. W., "Use of Soils to Retain Waste in Landfills and Surface
Impoundments," in Utilization, Treatment, and Disposal of Waste on Land,
ed. K. W. Brown, B. L. Carlile, R. H. Miller, E. M. Rutledge, and E. C.
A. Runge, pp. 279-300, Soil Science Society of America, Inc., Madison,
WI, 1986.
18. Wood, P. R., Lang, R. F. and Payan, I. L., "Anaerobic Transformation,
Transport, and Removal of Volatile Chlorinated Organics in Ground Water,"
in Ground Water Quality, ed. C. H. Ward, W. Giger, and P. L. McCarty,
pp. 493-511, John Wiley and Sons, New York, NY, 1985.
VAPOR CONTROL 575
-------
Radiological Monitoring of Select Faunal Species
Indigenous to Environs of the Maxey Flats
Shallow Land Burial Facility
Robert B. Burns
Kentucky Department for Environmental Protection
Morehead, Kentucky
ABSTRACT
This stud> examines the radionuclide concentrations of four fauna!
species indigenous to the environs of the Maxey Flats Shallow Land
Burial Facility located in northeastern Kentucky. The sample species
were limited to the smokey shrew (Sore.\ fume us), short-tailed shrew
(Blarina brevicauda), white-looted mouse (/Vnwivu m leucopusl and
the eastern box turtle (Termpene c. camlina). Body fluids were analyzed
for tritium by liquid scintillation. Whole body samples were ashed and
analyzed for gamma-emitting radionuclidcs
The most abundant gamma-emitting radionuclide encountered was
potassium-40. The smokey shrew exhibited the greatest concent rai ion
of gamma-emitting radionuclides. The smokey shrew also exhibited the
greatest tritium concentration.
INHALATION
With the exception ni cesium-137, the gamma-emitting radionuclides
identified in this study were endogenous. The cesium-137 concentra-
tion may be attributed to global fallout from nuclear bomb testing. The
disposal l.iuht) did not seem to have influenced the concentrations of
gamma-emitting radionuclides assimilated h> the fauna! species
sampled. The concentrations of intium measured indicate that tritium
Irom the disposal facility has entered the biological systems of fauna!
species in the immediate area of the waste disposal facility
INTRODUCTION
Information concerning the concentration of radionuclides assimi-
lated by wildlife is essential in determining the env ironmental impact
ot waste disposal facilities. The purpose ol this study was to determine
UU'viUMPTION OF
AQUATIC ANIMALS
I'igiirc I
Potential Environmental Exposure Pathways at .1
I ow-Level Waste Burial Ground''
CONSUMPTION
OF ANIMAL
PRODUCTS
576 RAD WAS 11.
-------
the concentrations of radionuclides assimilated by four faunal species
indigenous to environments surrounding the Maxey Flats Shallow Land
Burial Facility located in northeastern Kentucky. This research was the
first wildlife study conducted at the facility and was compared with
other Maxey Flats studies to determine the environmental impact of
the facility.
The sample species used in this study were the smokey shrew (Sorex
jumeus); the short-tailed shrew (Blarina brevicauda); the white-footed
mouse (Peromyscus leucopus); and the eastern box turtle (Terrapene
c. Carolina). These species were chosen for the following reasons:
(1) the availability of indigenous species; (2) the opportunity to sample
large numbers without causing a great impact on population numbers;
(3) the burrowing habits of these animals; (4) food sources include
ground dwelling animal matter.
It has been suggested that the species most exposed to radionuclides
are the detritovores, the arthropods and earthworms living in the forest
floor litter. It also has been suggested that the species that consume
arthropods and earthworms are the most probable radionuclide
vectors'. A diagram of the potential environmental exposure pathways
is shown in Figure 1.
Tritium was chosen as an indicator in this study for the following
reasons: (1) its abundance in the burial facility3; (2) its abundance in
the off-site environment4; (3) the ready uptake by the animals; and
(4) its mobility in the environment.
The gamma-emitters were chosen for: (1) their abundance in the burial
site5; (2) their presence in the off-site environment6 and (3) their ex-
cellent data base in the environment. The majority of nuclides in wastes
from power plants, with physical half-lives of greater than 5 yr, are
gamma-emitters7.
SITE DESCRIPTION
In January, 1963, the Nuclear Engineering Company was issued a
license to operate the disposal facility and in May, 1963, the first radio-
active material was buried at Maxey Flats. The major users of the facility
included hospitals, power plants and various industries. Solid wastes
were buried in large rectangular trenches that ranged from 5 to 20 m
in length, 3 to 22 m in width and 3 to 10 m in depth. Liquid wastes
were solidified, on-site, by mixing them with cement and paper. This
mixture was then poured into polyethylene-lined trenches8.
From 1963 through site closure in 1977, 135,000 m3 of wastes were
buried at Maxey Flats. The volume of wastes buried at Maxey Flats
has been estimated to contain over 2.4 million curies of by-product
materials and 64 kg of plutonium9.
The burial trenches geologically lie within the Nancy Member of
the Borden Formations (Fig. 2). The Nancy Member consists of shale
and sandstone interbeds. Water is discharged from the facility by three
routes: (1) surface run-off, (2) interflow through the shallow soil zones
and (3) subsurface bedrock flow (Fig. 2). These routes form potential
water pathways for the migration of radionuclides from the facility. Due
to the extensive grading and earthmoving and a natural dip in the stra-
ta, the land surface slopes southeasterly, channeling most surface run-
off into a main east drainage channel".
In December, 1974, the Kentucky Department for Human Resources
released a report entitled "A History and Preliminary Inventory Report
on the Kentucky Radioactive Waste Disposal Site." The report stated
that radioactivity had been detected in the unrestricted environment of
the disposal facility. The conclusions and recommendations of this study
were:
2000
1060
1020
980
w
w 940
w" 900
Q
Farmers Member and lower pin of the
Ohio Shale arc saturated Other units
above the the Ohio Shale may be saairatcd
(as shown by flowlincs), or may be unsatunicd
Upper pan of the Ohio Shale k unsatuntcd
_ WEST
860
820
780
740 —
700
Evaporation and
transpiration
Rainfall
EAST —I
Co 11 u v iu m
and soil
Weathered Nancy Member
Sandstone marker bed
Unwcathcrcd Nancy Member
Upper part of the Farmers
Member
Lower pan ol the Farmers
Member
Henley Bed _
.>, .Sunbury Shale
\V Bedford Shale
Ł —
t.
o
u-
\\v\\\ \^
Upper part of Crab Orchard Formation
Figure 2
Hydrogeological Position of the Maxey Flats
Shallow Land Burial Facility10
RAD WASTE 577
-------
• The facility was contributing radioactivity to the area environment,
but the activity was not creating a public health hazard.
• Some samples collected showed the presence of man-made radio-
nuclides in the unrestricted environment.
• Further geological, hydrogeological, and climatological studies should
be conducted to determine a profile for the site11
A followup report providing additional data on the status of the dis-
posal facility was released by the Department in December, 1976. The
followup report concluded that the disposal facility was still contributing
small amounts of radioactivity into the immediate environment, but it
did not post a public health hazard. The report also concluded that the
major mode of environmental contamination was surface run-off. It was
suggested that other modes, such as subsurface movement also con-
tribute to environmental contamination. The report recommended that
routine analysis of water and sediment from specific sampling stations
be performed by the Radiation Control Branch of the Department for
Human Resources and by the licensee".
METHODS AND MATERIALS
Trapping sites were established in the vicinity of the three major chan-
nels that drain the disposal facility (Fig. 3). These areas are covered
by deciduous forest consisting mainly of oak, hickory and maple trees.
Mouse-sized snap traps were randomly placed in areas that appeared
to offer suitable habitat for the species being sought. The traps were
checked and rebaited with peanut butter daily. Traps were discharged
over the weekends due to the dehydration of the small mammals after
death. When a sample was recovered, it was weighed, placed in a zip-
top bag and labeled with the species name, collection locality and date
of capture. The sample was then frozen until prepared for analysis.
The capture of Terrapene c. Carolina was performed by searching
the forest surrounding the facility. These samples were documented and
stored in the same manner as the mammals.
Mammalian body fluids were obtained by making an incision from
the xiphistcmum to the anus and folding the skin laterally. Internal organs
were removed and fluids extracted with a vacuum runnel. Tritium it
not selective for any one organ, and the quantity of body fluids was
small; therefore no effort was made to distinguish between fluids from
different tissue components. Any fluids remaining in the body cavity
were collected with a syringe. The fluid was centrifuged at 1500 rpm
until separation of liquid from solid constituents was evident. Only the
liquid portion was transferred into vials and used for tritium analysis".
Body fluids from Terrapene c. Carolina were obtained by separating
the carapace from the plastron and removing the internal organs. Fluids
were removed with a vacuum funnel. The fluids that remained in the
carapace were collected with a syringe. Attempts to separate the solid
constituents from the liquid by centrifugation were ineffective. There-
fore, to obtain a homogenous sample necessary for scintillation analy-
sis, aezoetrophJc distillation was performed using benzene. The sample
and solvent were placed in a flask and allowed to reflux. Benzene was
distilled and collected in a Barrett water/oil collecting tube. As the body
fluids condensed and collected in the tube, they separated from the
benzene which formed the top layer. The benzene layer was pipetted
and discarded. The body fluid layer was pipetted into scintillation vi-
als. A new pipette tip was used for each sample to prevent the possibil-
ity of cross-contamination.
"Insta-gel," used as the scintillator, was added to each sample.
Samples were analyzed for tritium in a Hewlett-Packard Tri-Carb Liquid
Scintillation Counter for 300 min.
To prepare collected fauna! species for gamma ray spectroscopy, whole
body samples were placed in evaporating dishes and ashed in a Fonn-8
oven at 300 °C for 12 to 16 hr. The charred samples were ground with
a mortar and pestle and transferred to petri dishes for analysis. Evapo-
rating dishes and the mortar and pestle were thoroughly cleansed with
nitric acid before each preparation to prevent the possibility of cross-
10
9
S -
3 -
2 -
1 -
.9
.8
.7.
.6-
.5-
.3-
.2-
5.20
1.66
1.56
0.55
Sraokey
Shrew
White-footed Short-tailed
Mouse Shrew
Eastern Box
Turtle
Figure 3
Approximate Trapping Locations
Figure 4
Total Gamma Radionuclide Concentrations by Species Concentrations in pCi/g
578 RAD WASTE
-------
contamination. The samples were analyzed for gamma-emitting radi-
onuclides using a Nuclear Data 680 system.
RESULTS
The smokey shrew showed a gamma radionuclide concentration
greater than any of the other species studied. The mean gamma radio-
nuclide concentration for the smokey shrew was 5.19 pCi/g. The mean
concentration of gamma-emitting radionuclides for the white-footed
mouse was 1.66 pCi/g. The mean gamma-emitting radionuclide con-
centrations for the short-tailed shrew and the eastern box turtle were
1.56 pCi/g and 0.55 pCi/g, respectively (Fig. 4).
The most abundant gamma-emitting radionuclide encountered was
potassium-40, with a mean concentration of 5.93 pCi/g. Radium-226
and cesium-137 showed mean concentrations of 0.68 pCi/g and 0.28
pCi/g, respectively (Fig. 5).
10-
9 —
8 —
7 —
6 —
5-
4 —
3 —
2 —
1 —
.9 —
.6 —
7
.6 —
.5 —
.4 —
.3 —
.2 —
.1
5.93
0.68
0.28
40K 226Ra 137Cs
Figure 5
Variable Gamma Radionuclide Abundance in pCi/g
The smokey shrew exhibited the greatest concentrations of
potassium-40, with a mean concentrations of 13.03 pCi/g. The mean
concentration of potassium-40 for the white-footed mouse and the short-
tailed shrew were 4.49 pCi/g and 4.13 pCi/g, respectively. The mean
concentration of potassium-40 assimilated by the eastern box turtle was
1.28 pCi/g (Fig. 6).
The smokey shrew exhibited the greatest concentrations of
radium-226, with a mean concentration of 1.87 pCi/g. The mean con-
centration of radium-226 assimilated by the short-tailed shrew was 0.38
pCi/g. The mean concentrations of radium-226 for the white-footed
mouse and the eastern box turtle were 0.34 pCi/g and 0.16 pCi/g, respec-
tively (Fig. 7).
The smokey shrew exhibited the greatest concentrations of cesium-137,
with a mean concentration of 0.67 pCi/g. The white-footed mouse
exhibited the lowest concentrations of cesium-137, with a mean con-
centration of 0.14 pCi/g. The mean concentrations of cesium-137
15-i
10
9.
8
7
6-
5-
4.
3-
2-
.3-
.2-
.01
13.02
4.49
4.13
1.28
Smokey White-footed Short-tailed Eastern Box
Shrew Mouse Shrew Turtle
Figure 6
Variable"°K Concentrations by Species Concentrations in pCi/g
1.87
2 —I
.01
5-
9-
8-
6-
5—
4-
3—
2-
1-
9-
3-
7-
i-
0.38
0.34
0.16
Smokey
Shrew
Short-tailed White-footed Eastern Box
Shrew House Turtle
Figure 7
Variable 226Ra Concentrations by Species Concentrations in pCi/g
RAD WASTE 579
-------
1 -
.9 -
.8 -
.7 -
.6 -
.5 -
.4 -
.3 -
.2 -
.1 -
.09—
.08-
.07-
.06-
.04-
.03-
.02—
.01
0.67
0.20
0.17
0.14
Sraokey Eastern Box Short-tailed White-footed
Shrew Turtle Shrew Mouse
Figure 8
Variable irCs Concentrations by Species Concentrations in pCi/g
400 —
300 —
200 —
100 —
90 —
80 —
70 —
60 —
50 —
40 —
30 —
20 —
312.58
2
37.26
1
52.64
41.05
Smokey Short-tailed White-footed Eastern Box
Shrew Shrew Mouse Turtle
Figure 9
Variable JH Concentrations by Species Concentrations in pCi/mL
assimilated by the eastern box turtle and the short-tailed shrew were
0.20 and 0.17 pCi/g, respectively (Fig. 9).
The smokey shrew exhibited a concentration of tritium greater than
any of the species studied. The mean concentration of tritium assimi-
lated by the smokey shrew was 312.58 pCi/mL. The eastern box turtle
exhibited a concentration of tritium less than any of the species studied.
The mean tritium concentrations assimilated by the eastern box turtle
was 41.05 pCi/mL. The mean concentrations of tritium assimilated by
the short-tailed shrew and white-footed mouse were 237.26 pCi/mL anl
152.64 pCi/mL, respectively (Fig. 9).
DISCUSSION AND CONCLUSIONS
Extensive radiological monitoring studies of botanical species haw
been performed by Battelle Pacific Northwest Research Laboratory at
the Maxcy Flats Disposal Facility. In the fall of 1981, potassium-40 was
found in newly fallen leaf samples collected from the perimeter of die
site. These concentrations ranged from I to 3 pCi/g dry weight.
Cesium-137 was measured in the newly fallen leaf samples, with con-
centrations ranging from 0.02 to 0.20 pCi/g dry weight1*.
In 1981, tritium concentrations present in leaf water extracted from
oak trees growing around the perimeter of the disposal facility indi-
cated that tritium from the facility had entered the transpiration processes
of these trees. The tritium concentrations ranged from 48 to 560 pCi/ml.
Maple tree sap was also analyzed for tritium concentrations. These con-
centrations ranged from 30 (o 151 pCi/ml for the 1981-82 sampling
period*
Soil samples from the immediate area surrounding the facility have
been analyzed for gamma-emitting radionuclides. These analyses have
shown the presence of potassium-40, with concentrations ranging from
17 to 26 pCi/g. Radium-226 was present, with concentrations ranging
from 0.95 to 2.2 pCi/g. Exogenous cesium-137 concentrations ranged
from 0.47 to 0.51 pCi/gr
Poiassmm-40 was measured in all fauna) samples collected These
concentrations ranged from 0.83 to 21 pCi/g and had a mean concen-
tration of 5.93 pCi/g. Analysis of fauna showed that the amount of
radium-226 assimilated ranged from 0.01 to 4.40 pCi/g, with a mean
concentration of 0.68 pCi/g. Cesium-137 measured in fauna showed con-
centrations ranging from 0.03 to 1.40 pCi/g. The mean fauna! concen-
tratjon for cesium-137 was 0.28 pCi/g. Tritium was present in the species
analyzed. These concentrations ranged from 2.3 to 562DpCi/mL, with
a mean concentration of 153.92 pCi/mL.
With the exception of cesium-137. the gamma-emitting radionuclides
encountered in this study were endogenous. The concentrations of
cesium-137 assimilated by fauna indigenous to the environs of the dis-
posal facility may be attributed to fallout from nuclear bomb testing;
they do not appear to be influenced by the disposal facility.
The concentration of tritium assimilated by fauna appears to be com-
parable to the concentrations in the sample trees surrounding the dis-
posal facility which were reported by Kirby, in 1983, to be in excess
of background concentrations. A control leaf sample from the vicinity
of Cave Run Lake, Kentucky, showed a tritium concentration of 2.4
pCi/mL. The tritium concentrations found in leaves near the Maxey
Flats Shallow Land Burial Facility ranged from 48 to 560 pCi/mL. Con-
centrations of tritium assimilated by faunal species ranged from 23
to 562 pCi/mL. This indicates that tritium from the disposal facility
has entered the biological systems of fauna in the immediate area.
REFERENCES
1. Kirby, L. J., Research Program ai Maxey Flats and Consideration of Other
Shallow Land Burial Site*. United States Nuclear Regulators Commission
Rcpt. No. NUREG/CR 1982.
2. Murphy, E S. and Holler, G. M.. Technology, Safety, and Costs of Decom-
missioning a Reference Lo*--Level Hbste Burial Ground, United States
Nuclear Regulatory Commission Repl. No. NUREG/CR-0570, 1980,
3 Operational Records. 1981.
4. Operational Records. 1981
5. Operational Records, 1981
6. Operational Records. 1981.
7. Razor, J. E.. Personal Communication. 1982.
8. Clark. D. T.. Radioactivity Canceniraiiitns at the Maxey Flats Area of Fleming
Ci'uniy, Kentucky. Commonwealth of Kentucky. Department for Human
Resources. Frankfort, KY, 1976.
9. Operational Records, 1981.
10, Zenher. H. H., Preliminary Hydrolagic Investigation of the Maxey Flats Kadi-
oactiw Hbsle Burial Site, Fleming County, Kentucky. USGS Rept. No.
79-1329, USGS, Louisville, KY, 1976.
580 RAD WASTE
-------
11. Zenher, H. H., Preliminary Hydrologic Investigation of the Maxey Flats
Radioactive Waste Burial Site, Fleming County, Kentucky, USGS Rept. No.
79-1329, USGS, Louisville, KY, 1976.
12. Clark, D. T., A History and Preliminary Inventory Report on the Kentucky
Radioactive Waste Disposal Site, Commonwealth of Kentucky, Department
for Human Resources, Frankfort, KY, 1973.
13. Clark, D. T., Radioactivity Concentrations at the Maxey Flats Area of Fleming
County, Kentucky, Commonwealth of Kentucky, Department for Human
Resources, Frankfort, KY, 1976.
14. Horrocks, D. L., Measuring Tritium with Liquid Scintillation Systems, in
Tritium, ed. A. A. Moghissi and M. W. Carter, pg. 807, Messenger Graphics,
Phoenix, AZ and Los Vegas, NV, 1973.
15. Kirby, L. J., A. P. Toste, W. H. Richard, D. E. Robertson, Radionuclide
Characterization, Migration and Monitoring at a Commercial Law Level
Waste Disposal Site, Battelle Pacific Northwest Research Laboratory Rept.
No. IAEA-CN-43/470, Richland, WA, 1983.
16. Kirby, L. J., A. P. Toste, W. H. Richard, D. E. Robertson, Radionuclide
Characterization, Migration and Monitoring at a Commercial Low Level
Waste Disposal Site, Battelle Pacific Northwest Research Laboratory Rept.
No. IAEA-CN-43/470, Richland, WA, 1983.
17. Operational Records, 1981.
RAD WASTE 581
-------
Framework for Assessing Baseline Human Health Risks at
Superfund Sites Contaminated with Chemical
and Radiological Wastes
Lynn M. Sims
Judith M. Liedle, Ph.D.
William C Borden
Bechtel Environmental, Inc.
Oak Ridge, Tennessee
ABSTRACT
Historically, U.S. Department of Energy waste disposal practices have
resulted in the contamination of numerous sites with chemical and
radiological wastes. In many instances, a single site will contain both
types of waste, presenting unique risk assessment challenges. Currently,
protocols for uniformly assessing risk from both types of contaminants
at such sites do not exist. Consequently, the approach presented in this
paper was developed to provide a uniform assessment of the risk to
human health from sites contaminated with both hazardous chemical
substances and radiological materials.
INTRODUCTION
This paper presents a suggested framework for assessing risk at U.S.
Department of Energy (DOE) federal facility Superfund sites contami-
nated with both chemical and radiological wastes. Historically, the prin-
cipal contaminants of concern have been the radioactive materials, with
less or no emphasis placed on hazardous chemical contaminants.
However, with the passage of the SARA in 1986, and with greater con-
cern for the characterization and cleanup of federal facilities, both chemi-
IDENTIFY EXPOSURE
SCENARIOS
1
SELECT CONTAMINANTS
OF POTENTIAL CONCERN
i
1
EXPOSURE ASSESSMENT
TOXICITY ASSESSMENT
i
RISK CHARACTERIZATION
Figure I
Baseline Human Health Assessment Process
cal and radiological contamination are now being assessed at these DOE
sites.
The baseline human health assessment is an analysis of site condi-
tions in the absence of remedial action (i.e.. "no action" alternative).
This evaluation requires an understanding of the nature of contaminant
releases from the site, the pathways of human exposure and a measure
of the potential risk to human health as a result of the releases'. The
process includes identification of potential exposure scenarios for media-
specific pathways, selection of indicator contaminants, exposure assess-
ment, toxicity assessment and risk characterization.
As pan of the baseline human health assessment, several types of
effects can be evaluated relative to radioactive and chemical con-
taminants. This paper will address radionuclide carcinogenic (e.g.,
stochastic) effects and chemical carcinogenic and non-carcinogenic
effects. The radionuclide non-carcinogenic (e.g., non-stochastic) effects
of radioactive contaminants will not be discussed since they are the
result of relatively higher doses, hence are only of interest in special
and limited risk assessment scenarios. Further, effects from radionuclide
and chemical contaminants may be combined to obtain mixed waste
carcinogenic and non-carcinogenic effects. However, consideration of
mixed waste effects are beyond the scope of this discussion and will
not be addressed.
Each of the five major components of the baseline human health
assessment presented in Figure 1 is discussed in the following sections
for chemical and radiological contaminants.
IDENTIFICATION OF EXPOSURE SCENARIOS
The exposure scenarios describe the components for potential human
exposure pathways The pathways describe the mechanisms by which
a receptor may be exposed to contaminants originating from a site. An
exposure pathway is comprised of the following components: source;
mechanism of contaminant release; an environmental transport medium;
likely route of human intake or exposure; and potential human recep-
tor or exposure point. Figure 2 presents an example of the components
used in the development of an exposure scenario. To be considered as
a potential exposure scenario, all of these components of the source-
pathway-receplor scenario must be present.'
Under the authority of CERCLA, the U.S. EPA requires that a site-
specific risk assessment be conducted to characterize current and poten-
tial threats to human health3. This assessment involves the develop-
ment of a current exposure scenario for each site (e.g., resident nearest
to site, nearest population magnet and sensitive individuals), as well
as a reasonable maximum exposure scenario (e.g., resident/population
at point of highest contaminant concentration). It should be noted,
however, that the reasonable maximum exposure scenario can be
determined in a number of ways besides maximizing concentration.
582 RAD WASTE
-------
CONTAMINATED
SITE
SOURCE
RELEASE MECHANISM
(e.g., WASTE PIT LEACHATE
MIGRATES INTO SOIL)
TRANSPORT
MEDIUM
(e.g.. GROUND-
WATER)
EXPOSURE ROUTE
(e.g., INGESTION OF WATER FROM WELL)
> PATHWAY
HUMAN
RECEPTOR
Figure 2
Exposure Scenario Components
For example, exposure can, in some cases, be more sensitive to
individual activities (e.g., ingestion rates) than to contaminant
concentration3 In many exposure assessments, adjusting all
parameters to their limiting values would certainly maximize exposure
but may not have a realistic chance of happening in the real world. For
this reason, the concept of "reasonable maximum" scenario is used.
In such cases, the exposures may be high, but the combination of ex-
posure parameters for an individual are those that are more likely to
occur in the actual population 3
The U.S. EPA also requires consideration of all potential pathways
of radiation exposure from a disposal system to a receptor. This may
include inadvertent and intermittent human intrusion into a radioactive
waste disposal system by exploratory drilling for resources4. For this
scenario, the U.S. EPA assumes that passive institutional controls (e.g.,
markers, intrusion barriers and fences) or the intruders' own explora-
tory procedures would be adequate for the intruders to detect, or be
warned of, the incompatibility of the area with their continued activities.
Other agencies also have established regulatory guidance pertaining
to potential receptor scenarios at radioactive waste disposal sites. The
NRC assumes that after loss of institutional controls (i.e., the respon-
sible agency no longer has control of the site), an intruder may uninten-
tionally access a closed waste disposal site and subsequently modify
it for a specific purpose5. Oztunali and Roles have identified several
intruder scenarios relative to the NRC regulations, including a housing
construction scenario and an agriculture scenario6. As a result of
intrusion into the waste, short- and long-term radiation exposures to
the individual could occur.
In addition, DOE has identified two general human receptor scenarios
for assessment of dose resulting from DOE on-site operational releases
from nuclear facilities: the population and maximally exposed individual
in the vicinity of DOE-controlled facilities7, and individuals who in-
advertently may intrude into a disposal facility after the loss of institu-
tional controls8. Occasional exposure of inadvertent intruders has also
been considered at DOE sites if restricted public use of the land is per-
mitted during the controlled period9. For example, deer hunters,
hikers, campers, wildlife enthusiasts or joggers in or near contaminated
areas may receive external and inhalation exposures if they intrude into
or near contaminated areas.
When developing exposure scenarios at sites contaminated with radio-
active and chemical contaminants, consideration of institutional control
time periods may be necessary. Institutional control time periods refer
to the time during which controls (e.g., physical, deed, regulatory res-
trictions, etc.) are placed upon a site by an institution (e.g., govern-
ment agency, private party, etc.). Institutional control time periods have
been established by the U.S. EPA'4'*10"1", NRC5, and DOE3.
CONTAMINANTS OF POTENTIAL CONCERN
Contaminants of potential concern are those contaminants that are
site related and for which data are of sufficient quality for use in a quan-
titative risk assessment. For the human health assessment at sites con-
taminated with both chemical and radiological substances, chemical
and radionuclide contaminant data must be evaluated and validated prior
to selection of the contaminants of potential concern.
EXPOSURE ASSESSMENT
Exposure assessment is the determination or estimation (qualitative
or quantitative) of the magnitude, frequency, duration and route of
exposure. Numerous variables are used to quantify exposure. These
include estimation of exposure point concentrations, estimation of con-
taminant intakes/exposures and quantification of pathway specific
exposures.
Exposure point concentration is the average concentration contacted
over an exposure period. Methods for estimating exposure point con-
centrations include direct use of environmental media monitoring data
and use of environmental fate and transport models. These models help
to predict contaminant release and migration when monitoring or charac-
terization data are unavailable at specified points of exposure. They also
help assess future risks to receptors from hazardous substances present
in the environmental media.
Chemical intake is the amount of contaminant at the exchange
boundaries of an organism that is available for absorption. These data
are normalized for time and body weight and expressed as mg chemi-
cal/kg body weight-day. The generic equation for calculating chemical
intake is presented in Table 1.
Table 1
Generic Equations for Calculation of Intakes/Exposures
CHEMICAL INTAKES
(C v CR X EF X ED) / BW X AT
I Intake (mg/kg body weight-day)
C Chemical concentration at the exposure point
CR Contact rate
EF Exposure frequency (days/year)
ED Exposure duration (years)
BW Body weight of exposed individual (kg)
AT Averaging time (days; period over which exposure
is averaged)
RADIONUCLIDE INTAKE/EXPOSURE
I X C X O X exp ( -D X T )
ETF Environmental transport factor
I - Annual intake of contaminated environmental medium
C = Average concentration of contaminant in
environmental medium
O = Other pathway specific factors (e.g., occupancy
factors, transfer factors, depth factors, etc.)
D = Radiological decay constant for contaminant
T = Time for decay
Radionuclide intake/exposure is determined by an environmental trans-
port factor (ETF). This transport factor consists of pathway factors that
affect the migration of a radionuclide or transmission of ionizing radia-
tion along a pathway from the source to the point of human exposure.
The generic equation for estimating the ETF is presented in Table 1.
RAD WASTE 583
-------
Specific chemical and radionuclide intakes can be calculated for many
different exposure scenarios, A partial compilation for possible resi-
dential exposures is presented in Table 2
Table 2
Possible Residential Exposure Scenarios
1. Ingestion of contaminants in drinking water
2. Ingestion of contaminants in surface water while
swimming
3. Dermal contact with contaminants in water
4. Ingestion of chemicals in soil
5. Dermal contact with contaminants in soil
6. Inhalation of airborne (vapor phase) contaminants
7. Ingestion of contaminated fish and shellfish
8. Ingestion of contaminated fruits and vegetables
9. Ingestion of contaminated meat and dairy products
TOXICITY ASSESSMENT
Toxicity assessment is the determination of the potential for adverse
effects resulting from human exposure to the contaminants. If possi-
ble, an estimate of the relationship between the extent of exposure to
a contaminant and the incidence of disease also is provided. Compo-
nents of the toxicity assessment include estimation of effects from
exposure to both chemicals and radionuclides.
Effects from Exposure to Chemicals
A reference dose. RfD. is the toxicity value used most often to evaluate
non-carcinogenic chemical effects resulting from exposures at Super-
fund sites. Various types of RfDs are available, depending on the
exposure route (oral or inhalation), the critical effect and the length
of exposure being evaluated (chronic, sub-chronic or single event). The
length of exposure may include exposures lasting 7 vr to a lifetime (i.e.,
chronic RfD), exposures lasting 2 wk to 7 yr (i.e.. sub-chronic RfD),
exposures lasting less than 2 wk (i.e., 1- or 10-day health advisories)
and exposures from a single event (generally 1 day) (i.e.. developmental
RfD).
A slope factor and the accompanying weight-of-evidence determina-
tion are the toxicity data most commonly used to evaluate potential
human chemical carcinogenic risks. The weight-of-evidence determi-
nation is used to determine the likelihood that the agent is a human
carcinogen. The slope factor represents a toxicity value that quantita-
tively defines the relationship between dose and response. The slope
factor is used to estimate an upper bound probability of an individual
developing cancer as a result of a lifetime of exposure to a particular
level of a potential carcinogen.
Effects from Exposure to Radionuclides
There are two broad classes of effects resulting from radiation
exposures: (l)stochastic and (2)non-stochastic. Non-stochastic effects
are those that have observable thresholds and that increase in severity
with increasing dose. Examples of non-stochastic effects include cell
death, lens opacification, cosmetically-unacceptable changes in the skin
and amenorrhoea. These effects, observable only al relatively high
doses, would only be of concern in certain inadvertent intruder
scenarios. Stochastic effects are those effects that arc random (e.g..
probabilistic) in nature, for which linearity in dose respoasc is assumed.
and for which the degree of severity is independent of dose. Stochastic
effects can be divided into three broad classifications: (l)genelic effects.
(2)teratological effects and (3)carcinogenic effects. For the purpose of
this discussion only, the risk from carcinogenic stochastic effects will
be considered. Teratological effects data are insufficient to predict a
linear dose-response relationship at low doses". The International
Commission on Radiological Protection (1CRP) places the risk of genetic
eflects at approximately a factor of two lower than the risk of carcino-
genic effects (e.g., 4 X 10! rem' versus 10J rem') °. Further, the
ICRP slates that risk assessment of the detriment due to hereditary
damage should be made over the total population. Most risk assess-
ment scenarios, however, will not include populations sufficiently large
to represent a potential threat to the gene pool.
The ICRP and the National Council on Radiation Protection and
Measurements (NCRP) make the assumption that the frequency of
occurrence of health effects per unit dose at low-doses is the same as
at high doses. This linear, non-threshold hypothesis assumes that the
risk of radiation-induced effects (i.e., cancer) is linearly proportional
in dose, nn matter how small the dose might be. Since no threshold
associated with exposure to ionizing radiation is assumed, any dose,
no matter how low, might give rise to cancer.
Ideally, human cpidemiological data and animal data regarding
radiation-induced cancer are used in the calculation of numerical risk
estimates. However, since the epidemiologies! data are incomplete in
many respects, mathematical models are used to estimate the risk. The
result from the model is an effective dose equivalent for external radi-
ation pathways or a committed effective dose equivalent for internal
radiation pathways.
RISK CHARACTERIZATION
Risk characterization uses information from the exposure assessment
and toxicity assessment to assess risks to human health from con-
taminants at a site. Components of the risk characterization include
reviews of toxicity and exposure assessments, quantification of risks
from individual contaminants, quantification of risks from multiple con-
taminants, combining risks across exposure pathways and assessment
and presentation of uncertainties associated with the estimation of risks.
Quantification of Chemical Risks
Chemical risks from carcinogenic substances are estimated as the
incremental probability of an individual developing cancer over a life-
time as a result of exposure to the potential carcinogen. For low risk
levels (i.e., for estimated risks below 0.01), the linear low-dose cancer
risk equation presented in Table 3 can be used to estimate risk. Where
risk levels are greater than 0.01, the one-hit equation for high carcino-
genic risk levels can be used (Table 3). The total cancer risk from mul-
tiple carcinogenic substances can be obtained by summing the individual
substance risk estimates.
Table 3
Estimation of Carcinogen* Risks
CER RISK EQUATION
Risk - CDI X SF
whore: Risk - unities* probability of fatal canc«rs
CDI - Chronic daily intake averaged over 70 years,
expressed as mq/kq-day
SF - slope factor, expressed as (»g/kg-day)"1
ONE-HIT EQUATION
Risk - 1 exp<-CDI x SF)
where: Risk - unitless probability of fatal cancers
exp - exponential
CDI - chronic daily intake averaged over 70 years,
expressed as mq/kq-day
SF - slope factor, expressed as (»q/kq-day)"l
584 RAD WASTE
-------
The potential for non-cancer health effects is evaluated by comparing
an exposure level over a specified time period with a reference dose
derived for a similar exposure period. This ratio of exposure to toxi-
city is called a hazard quotient. The generic equation for calculation
of hazard quotient is presented in Table 4. The total non-carcinogenic
hazard index from multiple substances can be obtained by summing
individual substance hazard quotients.
Table 4
Generic Hazard Quotient Equation
Noncancer Hazard Quotient = E/RfD
where: E = exposure level (or intake)
RfD = reference dose
E and RfD must:
(1) be expressed in the same units
(2) represent the same exposure period (i.e.,
chronic, sub-chronic, or shorter-term)
If E/RfD < 1, it is unlikely that even sensitive populations
would experience adverse health effects from the contaminant.
If E/RfD > 1, there may be concern for potential non-cancer
effects.
Quantification of Dose from Radionuclides
The annual dose for each radionuclide can be estimated as illustrated
in Table 5. By summing the external and internal radiation doses
individually for all contaminants at an exposure point, the annual
external and internal dose to individuals is obtained. Further, the annual
external and internal doses are summed resulting in a total individual
annual effective dose equivalent at an exposure point.
The annual effective dose equivalent may be converted to a health
risk by using the risk coefficient of 2 x 10^ risk of fatal cancer per
person-rem of radiation dose as calculated using the linear non-threshold
model14. The lifetime risk is based upon the further assumption that
the exposure level is the same for each year of a 70-yr lifetime.
Combining Risks Across Pathways
Whether risks or hazard indices for two or more pathways should
be combined for a single total exposure point can be determined by
considering: (1) the identification of reasonable exposure pathway com-
binations and (2) the likelihood that the same individuals would con-
sistently face the reasonable maximum exposure by more than one
pathway. If it is reasonable to combine risks across pathways, the cancer
risks and the non-cancer risks must be combined separately.
UNCERTAINTIES
There are many uncertainties that are inherent in the baseline risk
assessment process. Each component of the process, identified in
Figure 1, has uncertainties associated with the input parameters, the
evaluation methodology and the results. Specifically, some sources of
uncertainty in the baseline risk assessment include input variable un-
certainties modeling uncertainties, scenario uncertainties, and risk
estimate uncertainties.
The process of analyzing the uncertainty can be either quantitative
or qualitative depending on the time, resources and parameters or
processes being analyzed. Selecting the appropriate way to characterize
an uncertainty depends upon the type of decision the analysis supports,
confidence level required, model type, quantity type, extent and quality
of information and understanding available and the method used to
propagate uncertainty.
Tables
Annual Effective Dose Equivalent for Radionuclides
DCF X ETF
where: D = Annual Dose to an individual from external or
internal exposure from a radionuclide
DCF = Dose Conversion Factor
ETF = Environmental Transport Factor
Dose Conversion Factor: The dose conversion factor is the
committed effective dose equivalent per quantity of a radionuclide
inhaled or ingested (for internal exposure) or the effective dose
equivalent rate per concentration of a radionuclide in the air,
water, or ground (for external exposure).
Dose Equivalent: Dose equivalent is the product of absorbed dose
in tissue, a quality factor, and other modifying factors.
Effective Dose Equivalent: Effective dose equivalent is the sum
of the products of dose equivalent and weighting factor for each
tissue.
Weighting Factor: The weighting factor is the decimal fraction of
the risk arising from irradiation of a selected tissue to the
total risk when the whole body is irradiated uniformly to the same
dose equivalent.
Committed Effective Dose Equivalent: Committed effective dose
equivalent is the sum of the committed dose equivalents to various
tissues in the body each multiplied by the appropriate weighting
factor.
Committed Dose Equivalent: The committed dose equivalent is the
predicted total dose equivalent to a tissue or organ over a
specified time period after an intake of a radionuclide into the
body. It does not include contributions from external dose.
REFERENCES
1. U.S. EPA, Superfund Public Health Evaluation Manual, EPA/540/1-86/060
(OSWER Directive 9285.4-1), Office of Emergency and Remedial Response,
Office of Solid Waste and Emergency Response, Washington, D.C.
2. U.S. EPA, National Oil and Hazardous Substance Pollution Contingency
Plan, 40 CFR 300, Proposed Rule, 50 FR 38066, Dec. 21, 1988.
3. U.S. EPA, Proposed Guidelines for Exposure-Related Measurements, 53 FR
48830, Dec. 2, 1988.
4. U.S. EPA, Environmental Radiation Protection Standards for Management
and Disposal of Spent Nuclear Fuel, High-Level and Transuranic Radio-
active Wastes. 40 CFR 191, Final Rule, 50 FR 38866, Sept. 19, 1985.
5. U.S. Nuclear Regulatory Commission, Licensing Requirements for Land Dis-
posal of Radioactive Waste, 10 CFR 61, Final Rule, 47 FR 57482, 10 CFR
Parts 2, 19, 20, 21, 30, 40, 51, 61, 70, 73, and 170, Preamble, 47 FR 57446,
Dec. 27, 1982.
6. Oztunali, O. I., and G. W. Roles, Update of Pan 61 Impacts Analysis Metho-
dology, NUREG/CR-4370, Vol. 1, Envirosphere Company, Prepared for the
U.S. Nuclear Regulatory Commission, Jan. 1986.
7. U.S. Department of Energy, Radiological Effluent Monitoring and Environ-
mental Surveillance, DOE Order 5400.xy, Draft, Sept. 14, 1988.
8. U.S. Department of Energy, Radioactive Waste Management, DOE Order
5820.2A, Effective Nov. 28, 1988.
9. Bechtel National, Inc., Remedial Investigation Plan for ORNL Waste Area
Grouping 7, ORNL/RAP/Sub-87/99053/16&Rl, Prepared for Oak Ridge
national Laboratory, Oak Ridge, TN, Nov. 1988.
10. U.S. EPA, EPA Interim Status Standards for Owners and Operators of
Hazardous Waste Facilities, 40 CFR 265, Effective Nov. 19, 1980, 45 FR
33232.
11. U.S. EPA, Environmental Protection Standards for Uranium Mill Tailings,
40 CFR 192, Final Rule, Effective Apr. 22, 1980, 45 FR 27366.
12. United Nations, Scientific Committee on the Effects of Atomic Radiation,
Source and Effects of Ionizing Radiation: 1977 Report to the General
Assembly, New York.
13. International Commission on Radiological Protection, Recommendations of
the International Commission on Radiological Protection, ICRP Publication
26, Pergamon Press, Oxford, England, 1977.
14. U.S. Nuclear Regulatory Commission, Policy Statement on Exemptions from
Regulatory Control, 53 FR No. 238, pp. 49886-49891, Dec. 12, 1988.
RAD WASTE 585
-------
Decontamination and Decommissioning of a
Plutonium Fabrication Facility
Robert A. Hunt
Michael L. West
Donald Paine, PhD
EcoTek, Inc.
Erwin, Tennessee
ABSTRACT
EcoTck, Inc. was contracted in July, 1987 to manage an active project
involving the decontamination and decommissioning (D&D) of a plu-
tonium fabrication facility at Nuclear Fuel Services, Inc. (NFS) in
Erwin, Tennessee. Approximately 10,500 ft1 of currently unused plu-
tonium fabrication facilities are located in two separate buildings on
the NFS-Erwin site.
Waste processing strategy centers around decontamination and
sectioning with an ultra-high pressure water jetting system incorporating
a recirculated medium; volume reduction in a high capacity shear/baler;
and material control accountability utilizing a five station active-passive
neutron non-destructive assay (NDA) system. A stainless steel contain-
ment structure has been constructed to house the sectioning and decon-
tamination station. This containment structure attaches directly to the
shear/baler, which has been modified to encapsulate all surfaces subject
to contamination. The NDA system consists of five stations: (1) pre-
decontamination inventory station, (2) decontamination assay station,
(3) nuclear safety and accountability monitoring system. (4) bale and
drum counter and (5) bulk mixed uranium-plutonium oxide assay system.
The majority of waste consists of 136 gloveboxes containing process
equipment. Additional sources are ventilation ductwork, piping, conduit,
scabbled concrete and soil. This paper will present a brief synopsis
of the overall decommissioning approach which received United States
Nuclear Regulatory Commission (USNRQ approval on June 20, 1989.
INTRODUCTION
The primary objective of the decontamination and decommissioning
effort is to remove all transuranic (TRU) waste by Apr. IS, 1992. By
contract, all TRU waste must be received by the Department of
Energy—Idaho National Engineering Laboratories (DOE-INEL) no later
than this date. Specific plan objectives are:
• To restore the existing facilities and site to levels of contamination
which will permit "unrestricted" use, including possible use for future
NFS requirements
• To accomplish the work in a safe and environmentally acceptable
manner in accordance with all applicable federal and state regulations
• To minimize the volume of waste shipments
• To keep the TRU waste volume below 5,500 ft'
• To complete all shipments to DOE no later than Apr. 15, 1992
• To meet the above objectives while performing the work in the most
cost-effective manner
• To maintain exposures As Low As Reasonably Achievable (ALARA)
BACKGROUND
Site Description
The NFS site encompassing approximately 58 ac, is located within
the Erwin city limits. The City of Erwin has a population of approxi-
mately 5,600 people and is the seat of Unicoi County (population
approximately 16jOOO). The area is within the mountainous region of
east Tennessee. The site occupies a relatively level area 25 to 50 ft above
the Nolichucky River. To the north, east and south, the mountains rise
to elevations of 3,500 to 5JOOO ft within a few miles of the site.
Plutonium Facilities History
The plutonium facilities at NFS-Erwin were constructed in 1964 and
1965. Figure 1 shows the plutonium facilities in relation to the Erwin
plant site. Table 1 provides a description of the plutonium facilities.
Between 1965 and 1972, NFS processed 812 kg of plutonium for four
primary customers. The largest order covered the manufacture of
approximately 2 XXX) PuO2-UOr mixed oxide (MOX) fuel rods for the
Southwest Experimental Fast Oxide Reactor (SEFOR). This project
was a joint undertaking of General Electric, the AEC and several utility
companies. The GE-SEFOR order (746 kg Pu) and the DuPont-SROO
order (16 kg Pu) comprised 94% of the Erwin job orders which utilized
plutonium as shown in Table 2.
In the years following completion of the final order (1973 to 1985),
NFS was unsuccessful in finding a disposal site for TRU wastes (hat
would be generated from decommissioning activities. NFS was finally
successful in negotiations with the DOE-INEL office in 1985. These
efforts culminated on Apr. 15. 1986, with the signing of the contract
which allows NFS to ship its TRU wastes to DOE-INEL.
Process and Equipment Description
Capabilities of the NFS-Erwin plutonium facilities included: dissolu-
tion of plutonium metal and oxide; co-precipitation of uranium-
plutonium; blending of MOX powders; pellet production and inspec-
tion; rod loading, welding and inspection; scrap dissolution; and full
laboratory services.
Equipment in the facilities is located primarily in gloveboxes or in
a single limited-entry cell adjacent to the conversion area. In addition
to gloveboxes, the plutonium facilities contain equipment such as: metal
tanks (some containing Raschig rings); glass columns; pumps; mixing
vessels; blenders; drying, conversion and sintering furnaces; pellet press;
cut-off machine and centerless grinder; outgasing equipment; inspec-
tion jigs; welders; leak test equipment; liquid and high efficiency
paniculate air (HEPA) filters; miscellaneous laboratory equipment;
ventilation fans; wet scrubbers; and piping. Figures 2 and 3 show
detailed layouts of equipment in Building 234 and Building 110, respec-
tively. Equipment listings by type and volume for each building are
shown in Tables 3 and 4.
Radiological Status
Initial radiological surveys were made in each building to provide
586 RAD WASTE
-------
Figure 1
Plutonium Facilities NFS-Erwin Plant Site
N
PLANT NORTH
Figure 2
Building 234: Detailed Equipment Layout
input to planning efforts. Continuous radiological surveillance is
performed at each building to maintain exposures ALARA, to prevent
spread of contamination, to determine extent of decontamination
required and to segregate radioactive waste. A final radiological survey
will be made at each facility at the conclusion of decontamination and
decommissioning (D&D) operations in compliance with the USNRC
"Guidelines for Decontamination of Nuclear Facilities and Equipment
Prior to Release for Unrestricted Use or Termination of License for
Byproduct, Source, or Special Nuclear Material," Division of Fuel Cycle
and Material Safety, July, 1982.
Facility Disposition
As previously discussed, the primary objective of the NFS D&D
project is to reach an "unrestricted" use status for all remaining equip-
ment/facilities after completion of the Final Survey. The equipment that
is not contaminated or has been successfully decontaminated will either
be retained for use by NFS or processed as excess equipment. There
is no intent during D&D to remove the building structures since any
remaining structures should meet the unrestricted release criteria.
Decommissioning Project
The purpose of the project is to decommission the NFS facilities and
dispose of all contaminated waste generated at off-site burial and/or
storage locations. The waste material is weighed and assayed to deter-
mine the concentration of radioactivity. The three waste classifications
are Class A (< 10 nCi/gm), Class C (10-100 nCi/gm) and Greater Than
Class C (>100 nCi/gm) of TRU. Following dismantling, decontami-
nation, shearing, baling, packaging and classifying, the waste is sent
RAD WASTE 587
-------
PLANT
NORTH
sgŁA,
A
Figure 3
Building 110: Detailed Equipment Layout
to one of the approved disposal or storage sites. INEL's facility will
be used for Greater Than Class C materials. Class A and Class C
material will be disposed of at an approved commercial burial facility
As decommissioning progresses, bulk quantities of MOX will be
encountered. As that occurs, these materials will be collected under
rigid criticality, security and material control conditions. Processing
of these materials, such as screening, cleaning and drying, will be re-
quired. Bulk MOX material will be packaged, stored and shipped to
the Department of Energy (DOE).
Modifications to Building 234. Area D have been performed in order
to construct a decontamination and volume reduction area (Fig. 2). The
existing plant ventilation systems and criticality monitoring systems have
been supplemented to support this area.
DECONTAMINATION AND VOLUME
REDUCTION FACILITY
The Decontamination and Volume Reduction Facility (DVRF) located
in Area D of Building 234 is utilized to decontaminate and volume
reduce gloveboxes and other components and equipment. The major
pieces of equipment associated with the DVRF arc the General
Dynamics Corporation PERMA-CON™ enclosure (decontamination
cell), the ADMAC* JETPAC™ ultra-high pressure water jetting sys-
tem, the MAC Corporation Shear Power Baler™ Model 5200, Pajarito
Scientific Corporation Five Station Active-Passive Neutron NDA Sys-
tem, General Electronics digital scales, the data acquisition system
(DAS) electronics and the bale packaging and storage areas. Figure 4
shows the layout of the DVRF.
Decontamination Cell
The decontamination cell is a modified version of General Dynam-
ics PERMA-CON1" enclosure. The enclosure consists of interchange-
able, modular panels constructed of a carbon steel frame sheathed with
stainless steel.
The overall dimensions of the containment structure are 30 ft long
and 10 ft wide. It is divided equally into two 10- x 15-ft rooms. The
first room is a material receipt airlock. The airlock is 12 ft high. The
second room is the main decontamination area which is 16 feet high.
Bi-fold doors lead into the airlock and provide access between the airlock
and decontamination room.
Table 3
Equipment Lifting and Volume
Building 234
STATION I STATION DESCRIPTION VOLUME (CU.
I Pu Nitrate Load-in Box 199
2 Pu Scrap Prep. Bo* 89
3 Pu Scrap Prep. Bo« 48
4 Scrap Dissolution 2H
S Nitrite Storage Coluans 60
6 Nitrate Puap 1
7 Pw U Mailer HI* Tank 294
8 Sur^c Tanks 85
9 Surge Tank! 85
10 Evaporator 26
II Vacuw Cunp 28
13 Evaporator Tank 183
14 Condmier Receiver 40
S-l Scrubber Coluans 8$
IS Precipitation Bo> 334
li filtrate Coluon 8o> 11
17 Precipitation Box 212
21 Pu AOU Furnace Load Boi 102
22 Drying Furnace 90
23 Pu AOU Unloading Box 17S
24 Storage Box SB
2* Tray toad 38
26 Oxide Conversion Furnace 141
27 Tray Unload/Return Box 134
28 One* Halter Station SO
29 Tvln Shell Blender 232
30* Oilde Loading Station 141
308 Storage Rockets 139
31 Oxide Slugging Press 100
32 Oxide Transfer Box 3S9
33 Pellet Press 2«3
34* Inspection I Tray Return II
348 Inspection 131
34C Pellet Press lie
35 Pellet Press 438
36137 Pellet Inspect/Tray Load S3
38 Sintering Furnace 1S3
39 Unload Box t Tray Return 83
40 Pellet Inspection 83
4| Pellet Cut-off Station 144
42 Density Check Station 29
43 Pellet Mash Station 116
44 Pellet Storage (2
45 Pellet Storage 104
46 fuel Loading Station 188
4? Rod Cleaning Station 16
48 oxide Storage Rockets IK
62 Pellet Grinder 177
63164 Pellet Transfer/Storage 212
65 Pellet Out-gas Station 115
67 Condensate Station 209
68 Condensate Station 27
201MB Oxide Prep. Station 206
202 Drue, Unload Station 214
203 Scrap Prep. Station 177
204 Leaching Station 193
20S Oxide Load Station 128
206 Oxide Storage Rockets 62
207 Oxide Dissolution Station 370
208 Pu Nitrate Storage 202
209 Pu Nitrate Load-Out 212
103 Rod Cleaning Station 46
109 Rod Cleaning Station 80
Rod Welder 240
TOTAL CUBIC FEET FOR BUILDING 234 8,883
Both rooms are serviced by 1 ton capacity bridge cranes. Lighting
is provided from above through polypropylene panels in the ceiling of
the containment structure. Access ports for ventilation, supplied
breathing air lines, pneumatic tool air lines and ultra-high pressure water
jetting lines are provided through the modular panels with scalable bulk-
head penetrations.
The foundation of the floor is a built up concrete pad, sloped to a
sump in the middle and covered with stainless steel. A metal grating
cover provides traction on a level surface and prevents water used in
the decontamination process from pooling around the operators' feet.
Air is supplied to the containment from conditioned room air in Area
D. The air is pulled in at 500 cftn with an exhaust tan. The exhaust
588 RAD WASTE
-------
is discharged to the atmosphere through a HEPA-filtered exhaust stack.
Routine stack samples are taken to monitor airborne contaminants dis-
charged to the atmosphere. The air intakes to the containment are also
HEPA filtered to prevent spread of contamination into Area D in the
event of positive pressure in the cell.
All personnel operating in the cell wear supplied air encapsulated
suits which provide protection with a safety factor of 2,000. Careful
personnel surveying by radiation monitors and decontamination tech-
niques during transition out of the containment prevent the spread of
contaminants outside the cell.
Table 4
Station Descriptions and Estimated Volumes
Table 2
Summary of Ervvin Plutonium Processes
AREA
110-C
110-C
110-C
110-C
110-C
110-C
110-C
110-C
110-C
110-C
110-C
110-C
110-C
110-0
110-D
110-D
110-D
110-0
110-0
110-D
110-D
BUILDING 110
riON » STATION DESCRIPTION VOLUME (CU.FT.l
1 Screening & Sizer 27
2 Balance Box 178
3 Autoradiography 318
4 Cut-Off Machine 144
5 Mount Prep. Box 397
6 Fume Hood 273
7 Ion-Exchange 232
8 Sample Prep/Wet Chemistry 389
10 Auto Titration 351
11 Balance 174
12 Density & Porosity Test 207
13 Polarograph Box 331
14 Kjeldahl Apparatus 350
1 Small Hood Box 59
2 Prep. Box for Pulse 149
Height Analyzer
3 Fume Hood Box 225
4 Prep. Box for Emission 239
Spectography
5 Arc Stand Box 72
6 Auto Radiography 23
7 Standard Prep. Box 148
8 Vacuum Fusion Box 62
TOTAL CUBIC FEET FOR BUILDING 110 4,346
DATE
1965
1966
1967
1971
CUSTOMER
DuPont/SROO
SEFOR/GE/AEC
MOX
Fuel rods
HOX
Fuel rods
& scrap
di ssolution
Sub-Total AEC Programs:
1972
1972-
1973
Halden/NFS/RFD
Big Rock Point/
Consumers/NFS/RFD
MOX
Fuel rods
MOX fuel
assemblies
Sub-Total NFS Programs:
TOTAL ALL PROGRAMS:
KGS PLUTONIUM
16
746
762
3
47
50
812
234 BLDG.
Area C
Area A
Area B
Area D
Area E
Area F
Area G
Area H
Area I
Area M
Area 67
Area 68
Wet Cell
110 BLDS.
Area C
Area D
Table 1
Description of Plutonium Facilities
FUNCTION
Pelleting
Batch Weigh
Former U-233 Process
Fabrication
Lab
Office
Clean Change
Process Change
Material Unloading
Air Lock
Condensate Station
Condensate Station
Cell
FUNCTION
Wet Chemistry Lab
Spectrographic Lab
SQUARE FEET
3,000
108
994
1,550
228
135
246
249
288
360
168
66
630
TOTAL 8,022
SQUARE FEET
1,800
617
TOTAL
2,417
Figure 4
Decontamination and Volume Reduction Facility
RAD WASTE 589
-------
ADMAC® JETPAC™ Ultra-High Pressure
Water Jetting System
The ADMAC* JETPAC™ provides the primary means of decontami-
nation for gloveboxes, piping and equipment. The JETPAC™ produces
up to 40,000 psi of water pressure at a 2 gpm flow rate. This system
has effectively decontaminated hard surfaces at nuclear facilities across
the country. Additionally, the JETPAC™ can be fitted with options that
provide sectioning and concrete scaling capabilities.
During operation, the residual water from the decontamination process
is picked up from the floor sump through the metal grating by a slurp
pump. The water is processed through a series of filters and ion
exchanges and stored for re-use by the JETPAC™
MAC Corporation Shear Power Baler™ Model 5200 (Modified)
A shear/baler, with modifications specified to ensure containment
of contaminants, has been purchased from the MAC (Mobile Auto
Crusher) Corporation. The shear/baler provides 377 tons of shear
capacity and 180 tons of compaction force. The shear/baler can accept
gloveboxes, piping, conduit and other equipment into its loading hopper,
perform shearing and compacting operations with an extremely effi-
cient hydraulics package and deliver a 16-in. square bale of variable
thickness. The thickness is determined by the amount of material sheared
before the compaction stroke.
Major modifications have been specified to the basic shear/baler unit
to ensure that contaminants are kept within the shear and compaction
chambers. All internal wear plates have been seal welded to form a
continuous chamber from the loading hopper to the outlet chute. A sheet
metal enclosure surrounds the sliding plate and area behind the shear
ram bead. All hydraulic ram pistons have been fitted with bellows as-
semblies to prevent adhesion of airborne contaminants to the piston
walls which would lead to contamination of the hydraulic system.
The top of the hopper has been fitted with a sheet metal enclosure
over the top and three sides. The fourth side is flanged, fined with a
gasket and mated to the decontamination cell. The bridge crane installed
in the decontamination cell extends over the top of the loading hopper
to assist the operator in loading material into the shear/baler. Addi-
tionally, the side of the loading hopper adjacent to the decontamina-
tion cell has been hinged with electrically driven worm gears installed
to lower the side, allowing easier access for the operator to the charg-
ing box of the shear/baler. A vertical extension, which rides in a track
and folds down with this side, has been installed to provide protection
to the operators in the decontamination cell from small pieces of material
which might break loose and become projectiles during the shearing
operation.
The controls for the shear/baler are located at a central panel out-
side of the containment. A Lexan™ window has been installed in the
containment to allow the shear/baler operator a clear view of the loading
area. Emergency shut-off switches have been installed at the control
panel, inside the decontamination cell and at the bale outlet chute.
A glovebox is attached to the outlet chute to provide an enclosure
to seal the bales in plastic as they emerge from the shear/baler. The
bale is subsequently bagged out of the glovebox to provide a double
seal. Gravity roller conveyors have been installed in the glovebox to
assist the operator in handling the bales.
Pajarito Scientific Corporation's Five Station
Non-Destructive Assay System
Pajarito Scientific Corporation designed and fabricated a non-
destructive assay (NDA) system that consists of five stations. The stations
utilize combinations of passive neutron detectors, an active neutron
generator and a Canberra* Big MAC** (multi-attitude cryostat) hyper-
pure germanium detector.
Station 1 is a series of passive neutron detectors used to provide an
initial inventory of holdup in equipment prior to entering the decon-
tamination cell airlock. This inventory is used to verify holdup meas-
urements taken during the in situ characterization effort and to localize
holdup to optimize decontamination efforts. The equipment also is
weighed at this station with electronic scales to determine an overall
concentration. This station can detect 200 mg of plutonium.
Station 2 is an active-passive neutron differential die-away station
utilizing a neutron generator to induce reactions in the contaminated
material which can be detected by 'He neutron detector tubes. Station
2 is located in a chamber immediately adjacent to and accessible from
the decontamination cell. This station is utilized to determine decon-
tamination effectiveness and can detect K) mg of plutonium.
Station 3 consist of two passive neutron detector packages used for
real-time nuclear safety monitoring of holdup and material accounta-
bility. These detectors are placed on the water recirculation system and
the shear/baler. In addition to providing continuous monitoring, the
detectors trigger an alarm at a threshold of neutron activity well below
nuclear safety concern. Station can detect 200 mg of pluionium.
Station 4 is an active-passive neutron differential die-away station
used to assay bales and drums. The bales are accurately weighed,
assayed, measured and labeled with a bar code. The labeled bales are
stored in identifiable locations in a shelving area. As sufficient bates
are produced, packaging is accomplished in a batch mode. A computer
program optimizes the packaging of the drums based on bale heigh
and concentration, selecting bales that, when packaged together in a
55-gal drum, minimize void space and maximize specific activity within
the waste acceptance criteria of the designated burial or storage site.
Station 4 can delect 2 mg of plutonium.
Station 5 is a passive neutron detection chamber for assaying bulk
mixed plutomum-uranium oxide (MOX) retrieved from the process
equipment prior to decontamination. A hyper-pure germanium detec-
tor is used for isotopic analysis of the bulk MOX which is processed
and packaged for shipment to DOE.
Data Acquisition System
In order to provide an accurate history of decommissioning activity,
every opportunity has been taken to utilize electronic monitoring,
recording, retrieval and reporting. Equipment is tagged and tracked by
bar code from the moment it is removed from the process line to the
time it is placed in a drum for burial. This audit trail provides a valida-
tion of facility characterization, real-time material accountability con-
trol, and assists in management of the decommissioning effort. Records
required for shipment, storage and disposal are generated by the Data
Acquisition System (DAS) from the data base.
DECOMMISSIONING MATERIAL FLOW
Before gloveboxes and equipment are removed from the process areas,
they are surveyed for contamination. All contamination is either removed
or fixed in place to eliminate regeneration of airborne paniculate. Glove-
boxes or equipment that require dismantling or sectioning before removal
are completely contained inside temporary containments (e.g., tents).
All work associated with equipment removal or sectioning is conduct-
ed with respiratory protection and layered anti-contamination clothing.
All equipment removed from Building 110 is transported approxi-
mately 0.25 mi to Building 234 and the DVRF. Material transport is
conducted in a 16-ft trailer pulled by a tow motor. The trailer has been
lined with formica sheathing and linoleum with all cracks and crevices
sealed. The equipment is transferred into and out of the trailer through
dock seals attached to the buildings. The trailer is equipped with a roll
up door. The equipment removed from Building IK) enters the DVRF
through the Building 234, Area M airlock.
The equipment is then moved into the DVRF to NDA Station 1, The
initial assay and weight are recorded in the data acquisition system.
A determination is made to ensure that safe mass limits will not be
exceeded by the introduction of the material being assayed into the
decontamination cell. The initial assay also is used to validate the facility
characterization and to localize material holdup for the decontamina-
tion effort.
The equipment is then introduced into the decontamination cell
through an integral airlock. The material handling in the airlock and
the decontamination cell is assisted by 1 ton bridge cranes. Once the
equipment is in the decontamination cell, it can be sectioned as neces-
sary, opened and cleaned. After initial decontamination is complete,
NDA Station 2 is utilized to determine if decontamination has been
effective and if further cleaning is required. The goal is to decontaminate
590 RAD WASTE
-------
all equipment to below 10 nanocuries of TRU per gram of waste.
When the equipment is decontaminated to as low a level as achieva-
ble, it is hoisted into the loading hopper of the shear/baler. After each
shearing cycle is completed, a baling cycle is performed. If additional
shearing cycles are made before producing a bale, the bale becomes
too thick and unmanageable. A single cycle bale is nominally 4 in. thick
and weighs 50 Ib. The 16-in. height and width are determined by the
inner dimensions of the baling chamber.
The bale exits the shear/baler into a glovebox where it is sealed in
flame retardant plastic. It is then bagged out of the glovebox and heat
sealed in a second layer of flame retardant plastic. The sealed bale is
weighed, assayed in Station 4 and the thickness is measured. A bar
code label identifying the bale is attached and the associated informa-
tion is recorded in the DAS. The bale is then placed in an identified
cell in a temporary storage area. When sufficient bales are produced
to generate drums for shipping, the DAS is accessed and the computer
selects bales for optimum packaging of 55-gal drums.
CONCLUSIONS
During the planning phase of the NFS Plutonium facilities decom-
missioning project, several alternative approaches to the ultra-high pres-
sure jetting decontamination and shear/baler operation were evaluated.
However, the D&D action presented in this paper provided the best
means of disposing of the contaminated materials in an environmentally
sound manner in association with the implementation of a program that
minimized the amount of TRU and low level waste (LLW) materials
requiring disposal. Costs were substantially reduced due to the decreased
volume of the waste to be buried, the reduced time schedule and the
fewer number of personnel required to accomplish the tasks.
RAD WASTE 59 L
-------
Mixed Funding as an Enforcement Tool in
Superfund Settlements
Deborah Swichkow
William O. Ross
U.S. EPA
Washington, D.C.
INTRODUCTION
The goal of the U.S. EPA in implementing CERCLA. as amended
by SARA, is to achieve effective and expedited cleanup of as many
uncontrolled hazardous waste sites as possible. To meet this goal, the
U.S. EPA may enter into settlement agreements with PRPs to carry out
cleanups, or, where such agreements are not reached, use enforcement
actions to order such cleanups. Where enforcement actions are not avail-
able for a variety of reasons, however, the Agency may conduct Fund-
financed cleanups and subsequently inmate litigation against PRPs to
recover the costs of cleanups. The U.S. EPA's objective is to have the
party responsible for the release undertake the actions necessary to
mitigate such releases.
For this reason, response actions undertaken by PRPs arc essential
to the Government's goal of achieving effective and expedited cleanup
of hazardous waste sites. Prior to the enactment of 1986, SARA, the
Agency issued the "Interim CERCLA Settlement Policy (50 FR 5034
etseg., Feb. 5, 1985). The Interim CERCLA Settlement Polio set out
the conditions under which the government may settle for less than 100%
of the costs of a cleanup authorized by CERCLA through its strict, joint
and several liability. This guidance states that the U.S. EPA will negotiate
only if the initial offer from the PRPs constitutes a substantial propor-
tion of the response action costs or a substantial portion of the needed
remedial action.
To provide additional incentives for voluntary cleanups. Congress
through SARA provided the Agency with the authority to allow, private
panies (including PRPs) to carry out a response action and then reim-
burse the panies for the costs incurred. CERCLA Section 122 provides
the U.S. EPA with the authority to enter into negotiated settlements
with PRPs that provide for: (1) conduct of the response action by the
settlors in return for the reimbursement of a portion of the costs of the
response action from the Fund (i.e., "prcauthorization"); (2) the con-
duct of discrete portions of the work by the settlors in satisfaction of
their liability (i.e., "mixed-work"); and (3) cash payments by the settlors
in satisfaction of liability for the release at issue (i.e., "cash-outs").
In the broadest sense, these three different types of settlements are
referred to as "mixed funding" agreements. The Agency, at its discre-
tion, may enter into these agreements with PRPs to conduct and/or pay
for a portion of a response action.
In summary, the three types of mixed funding agreements to be
discussed in this paper are:
• Preauthorizution is an arrangement in which the PRP agrees to
conduct the response action, and the U.S. EPA agrees to allow the
PRP to assert a claim against the Fund for a portion of his response
costs
• Cash-out is an arrangement in which a PRP pays the U.S. EPA for
a portion of the response action and the U.S. EPA agrees to conduct
or arrange for the conduct of the response action
• Mixed-Ubrk is an arrangement in which the PRP and the U.S. EPA
agree to conduct discrete portions or segments of the response action.
While Section 122 of CERCLA authorizes mixed funding settlements.
Section I22(b)(l) also requires that the U.S. EPA make all reasonable
efforts to recover Fund reimbursements from other parties through the
authority of Section 107 of CERCLA.
MIXED FUNDING IN SETTLEMENT NEGOTIATIONS
After mixed funding settlements were expressly established by statute
as settlement tools for the U.S. EPA in the 1986 amendments, the US.
EPA published guidance on "Evaluating Mixed-Funding Settlements
Under CERCLA (53 FR 8279 et seq.. Mar. 14. 1988). This settlement
guidance provides that the criteria of particular importance include the
strength of the liability case against settlors and any non-settlors, die
size of the portion for which the Fund will be responsible, and other
mitigating and equitable factors.
Shortly after it issued the settlement guidance, the U.S. EPA elected
to implement the authority of Section 122(b)(l) to reimburse parties
to settlement agreements in the same manner as claims for response
costs authorized pursuant to Section lll(a)(2) of CERCLA. Specifically,
Section lll(a)C) of CERCLA authorizes the U.S. EPA to any claim
for response costs incurred by "any other person" as a result of carrying
out the National Contingency Plan. The NCP requires, among other
things, U.S. EPA's prior approval, referred to as "preauthorization."
THE PURPOSE OF A MIXED FUNDING SETTLEMENT
The purpose in pursuing a mixed funding settlement is as follows:
• Provides the government with an additional means to initiate response
activity
* Provides a means to expedite cleanup, thereby avoiding protracted
litigation
• Provides up-front financing of a cleanup by the PRPs
CRITERIA FOR ELIGIBILITY
The following criteria are specified in the guidance document Evalua-
ting Mixed Funding Settlement.'! Under CERCLA (Oct. 20, 1987) and
the ten-point settlement criteria contained in the Interim CERCLA Set-
tlement Policy. These criteria should be used in the evaluation of any
mixed funding settlement proposal.
• The strength of the liability case against seniors. This criterion
includes any litigative risks in proceeding to trial against settlors as
well as the nature of the case remaining against nonsettlors after the
settlement.
• The size of the portion or operable unit for which the Fund will be
responsible or the amount of the PRPs' offer. A substantial cost of
592 STATE PROGRAMS
-------
cleanup should be offered by the PRPs usually over 50%. The higher
the PRPs' portion, the greater the incentive for them to keep their
costs down.
• Good-faith negotiations and cooperation of settlors and other
mitigating and equitable factors.
• The government's options in the event settlement negotiations fail.
There should be some assurance, for example, that if negotiations
break down, a State cost share will be available for a Fund-lead action.
PREAUTHORIZATION IN SETTLEMENT NEGOTIATIONS
The preauthorization process in settlement negotiations consists of
three distinct but interrelated steps:
• Mixed funding offer, negotiation and U.S. EPA/PRP agreement in
principle
• Submittal of the application for preauthorization and review by the
U.S. EPA
• Final negotiation of the consent decree and development of the FDD
by the U.S. EPA.
Before the U.S. EPA will consider preauthorization of a response
action by a PRP, the PRP and the U.S. EPA must be involved in settle-
ment negotiations. The first step that a PRP may take to initiate the
preauthorization process is to propose a mixed funding settlement offer
through the appropriate U.S. EPA Regional Office. This proposal typi-
cally takes the form of a "good faith" offer in response to a special
notice letter from the U.S. EPA, pursuant to Section 106(b) of CERCLA,
through which the U.S. EPA advises the PRP that he is liable for
response costs as a result of a release or threat of release. Such an offer
should be substantial and provide a basis for evaluation by the U.S. EPA.
The Interim CERCLA Settlement Policy contains the "Ten-Point
Criteria" the U.S. EPA will use in determining whether it is appro-
priate to settle for less than 100% of response action costs. The U.S.
EPA will evaluate the PRP's offer against these criteria. If the offer
appears acceptable, the Regional Office will determine if preauthori-
zation, or some other mixed funding approach, is appropriate. Once
the Regional Office determines that preauthorization is appropriate, an
agreement in principle is generally sent to the PRP. At this point, the
Regional Office may also provide the PRP with guidance on the next
step—the Application for Preauthorization.
The Application for Preauthorization, consistent with Section
300.25(d) of the NCP [300.700(d) of the proposed revision to the NCP],
formally notifies the Agency of the PRP's intent to submit a claim
against the Fund, demonstrates the PRP's knowledge of the NCP and
the site-specific remedy and demonstrates the PRP's technical, finan-
cial and other capabilities to carry out the response action in a safe
and effective manner. Once the PRP submits an acceptable application,
the U.S. EPA will formulate the Preauthorization Decision Document
(FDD).
The development of the Consent Decree and the FDD is the final
step in the preauthorization process. The Consent Decree sets out the
requirements of the parties to the agreement and is enforceable by the
court. The non-negotiable, site-specific FDD, an attachment to the Con-
sent Decree, sets forth the terms and conditions for reimbursements
from the Fund, including the maximum amount of such reimbursements
and the schedule for reimbursements.
To date, the U.S. EPA has authorized six design and construction
mixed funding agreements through court approved consent decrees: the
Re-Solve site in Region I; the McAdoo, Harvey & Knotts and Tybouts
Corner sites in Region III; the Motco site in Region VI; and the Col-
bert site in Region X. In addition, a Consent Decree was lodged in
September 1989 for the Bailey Waste Disposal site in Region VI. The
U.S. EPA has authorized one mixed funding agreement for an Area-
wide RI/FS through an Administrative Order on Consent (Peak
Oil/Reeves/Bay Drums in Region IV). The costs of design and con-
struction range from $7 to $45 million. The maximum reimbursement
from the Superfund ranges from $1.4 to $9.3 million.
SUBMITTAL OF A CLAIM UNDER PREAUTHORIZATION
Claims will be awarded from the Fund in accordance with the terms
and schedule contained in the Consent Decree. An important compo-
nent of the claims award process is the presentation of the claim to non-
settling PRPs (i.e., other parties liable for the release that are not parties
to the settlement). Section 112(a) of CERCLA states that a claim may
not be submitted against the Fund unless it is first presented to the owner,
operator or guarantor of the vessel or facility from which the hazardous
substance has been released (if known) and to any other person who
may be liable under Section 107 of CERCLA. If additional PRPs are
unknown, the potential claimant must make a good-faith effort to iden-
tify any other parties believed responsible for the release. If the claimant
cannot locate other PRPs or the claim remains unsatisfied for 60 days
after presentation, the potential claimant may then submit the claim
to the U.S. EPA.
Once the U.S. EPA has received the claim and has determined that
the claim contains the information and documentation necessary for
evaluation (i.e., it has been "perfected"), the Agency will review and
analyze the claim according to the criteria set forth in the proposed
CERCLA Response Claims Procedures (54 FR 37892 et seq., Sept.
13, 1989). The Agency may use the services of a claims adjusting firm
in reviewing the claim to determine that the costs are "necessary."
The Agency's proposed Response Claims Procedures defines "neces-
sary" response costs as: (1) required (based on site-specific circum-
stances), (2) reasonable (nature and amount do not exceed that estimated
or which would be incurred by a prudent person), (3) allowable (in-
curred specifically for the site at issue) and (4) otherwise allowable
(consistent with the Federal cost principles). If the claim fulfills the
established criteria, the Agency will award the claim and reimburse
the claimant the amount of the approved response costs.
If the U.S. EPA determines that the claim has not fulfilled the re-
quirements contained in the Consent Decree, the claim will be denied
in whole or in part. As provided by Section 112(b)(2) of CERCLA,
if the claimant is dissatisfied with the award from the U.S. EPA, that
claimant may request a hearing before an Administrative Law Judge.
All decisions by such an ALJ shall be final, but either party may appeal
a decision within 30 days of notification of the award or decision to
the Federal district court. Pursuant to Section 112(b) of CERCLA,
decisions by an ALJ shall not be overturned except for arbitrary and
capricious abuse of discretion.
To date, settling defendants have not filed any preauthorized claims
against the Superfund.
MIXED WORK SETTLEMENTS
As stated above, a mixed work settlement involves an agreement which
addresses the entire response action, but the PRPs and the Agency agree
to conduct and pay for discrete portions of the remedial action.
Evaluation of a Mixed Work Proposal
Mixed work may be appropriate for cases in which the U.S. EPA
can identify discrete phases or operable units of the remedial action
and when PRP cooperation can be assured. Frequently, a removal action
or an RI/FS (where it involves an area-wide contamination) is the most
plausible situation to consider mixed work since specific tasks are broken
out easily. The U.S. EPA may agree to conduct soil removal actions
at one specific portion of the site (such as an impoundment area), while
PRPs would concentrate their activities removal at other areas of the
site (such as a building complex). Mixed work also can be a settlement
option in a RD/RA scenario. In the Love Canal, New York mixed work
agreement, the PRPs agreed to implement portions of the sewers and
creeks remedial program associated with implementation and the State,
through a Cooperative Agreement, implemented remaining portions
of the remedy (construction of a dewatering containment facility and
excavation of the creeks).
One approach used in carrying out a mixed work settlement is the
use of a "carve out." A "carve out" is a form of mixed work where
a particular task or tasks will be carved out by settling PRPs or by the
U.S. EPA from those tasks the settlors objected to carrying out. The
"carve out" tasks can be imposed exclusively on the non-settlors through
a Unilateral Administrative Order (UAO). Settlors would have tasks(s)
STATE PROGRAMS 593
-------
stipulated in the consent decree. If the recalcitrant fails to comply with
the Administration Order (AO), the U.S. EPA can either sue for
injunctive relief seeking compliance with the Order or judicial referral
with daily penalties imposed, or carry out the work itself using the
Fund. The U.S. EPA would follow by suing for cost recovery and treble
damages under Section 107.
Consent Decree Language Requirements
Since the implementability and overall managemeni of a mixed work
settlement is based largely on how well each activity is defined. Consent
Decree language should be developed with the same objectives in mind.
Any covenants not to sue should be clearly limited to the operable unit(s)
addressed in the Consent Decree; areas of responsibility and timeliness
should be delineated in the agreement. The covenant not to sue in the
Love Canal, New York decree is extremely narrow and only covers
claims under CERCLA. RCRA and State common law for work actually
performed or costs actually paid by the PRPs under the Decree. The
Decree also contains the standard "reopeners" provision for informa-
tion or conditions that emerge after entry of the Decree or completion
of the work. In the Ottati and Goss, New Hampshire settlement, the
reopeners also limit the PRPs' liability consistent with the Decree of
any additional work that must be done at the site.
State Cost Share Requirements
When the Federal government uses its response authority to conduct
a remedial action. Section 104 (C)(3) of CERCLA requires the State
to share a percentage of the cost of the remedial action. Since response
actions through mixed work may be carried out under Section 104 (C),
State cost share is required, including all future maintenance.
There are a variety of ways the State can "pay" or "assure payment"
of its cost share. For instance:
• the State and the PRPs may enter into an agreement under Slate law
and CERCLA where the PRPs pay 10% to the State and the Stale
obligates funds for use at the site; or,
• the State may use its own funds to pay for any portion of its share
that cannot be paid for by the PRPs.
State involvement in mixed work settlements are best illustrated in
the Ottati and Goss, New Hampshire, the PRPs undertook the soils
portions of the remedy, e.g. the aeration, incineration, soil replace-
ment and regrading, and establishment of the site cover. The State agreed
to conduct the necessary post-closure maintenance. The PRPs also
agreed to pay a portion of past and future oversight costs and a percen-
tage to the State for long-term maintenance of the site.
At Love Canal, New York, the State entered into a cooperative agree-
ment to conduct specific portions of the remedy (excavation of sewer
and creek sediments) while the PRPs processed, bagged and transported
materials form the Love Canal site to the PRP's main plant site for
temporary storage and incineration.
In either case, the U.S. EPA and the State should enter into a State
Superfund Contract (SSC) to assure cost share and O&M responsibility.
The State cost share docs not have to be incorporated into the Consent
Decree between U.S. EPA and the PRPs. In general, mixed work settle-
ments should only be considered when the State cost share is reasona-
bly certain.
CASH-OUT SETTLEMENTS
The third type of mixed funding arrangement described in Section
122 (b)(3) is a cash-out. A cash-out occurs when the U.S. EPA conducts
the response action and the PRPs pay U.S. EPA for a portion of the
costs. A cash-out settlement can be prepared at any time in the remedial
process. One common use of cash out settlements involves PRPs who
have contributed a low percentage of waste to a site, and who are not
technically or financially capable of conducting the entire response
action. Once the PRPs pay their allocated share of costs, they are no
longer liable for any further participation in site remediation.
Evaluating A Cash-Out Proposal
Since the U.S. EPA conducts the response action in a cash-out set-
tlement, factors such as the proposed remedy and public interest are
not decisive in evaluating a cash-out by all or some of the PRP». The
Interim Settlement Policy (Section III) does, however, identify the fol-
lowing key concerns that should be considered when evaluating such
a settlement:
• The U.S. EPA should have a high level of confidence in the infor-
mation concerning liability at the site and expected cost of the remedy
to determine an appropriate cash out settlement.
• The U.S. EPA should have sufficient information related to both
settlors and nonsettlors to determine a settlement amount for the
settlors requesting a cash-out (should be based on the settlement
policy, including waste contribution).
• The U.S. EPA should ensure that the percentage of total costs to be
paid by settlors is "substantial."
While cash-out settlements do not have to involve de minimis parties
as defined in Section 122(g), they do share many of the same analyti-
cal factors. The U.S. EPA guidance entitled, "Interim Guidelines on
settlements with De Minimis Waste Contributors under Section 122(g)
SARA" should be consulted when reviewing cash-out proposals.
Consent Decree Language Requirements
As mentioned previously, a key the U.S. EPA concern related to cash-
out settlements is the strength of the information related to PRP lia-
bility and the cost and development of the remedy. These issues have
particular bearing on the scope of any covenant not to sue in such an
argument—particularly, early in the remedial process when informa-
tion is limited.
The Regions should ensure that the covenant not to sue, if any, is
carefully drafted to cover only the specific response action covered by
the mixed funding settlement and is otherwise consistent with Section
122(0 and the "Covenants Not to Sue Under SARA," (52 FR 28038
July 27, 1987). This provision would include the standard reopenen
for unknown conditions and new information indicating that the remedial
action is not protective and which reopeners would be triggered in the
event of remedy failure. It also would include a provision ensuring that
the PRPs are responsible for contributing toward any cost overruns
arising during completion of the mixed-funded response action fin at
least the same percentage as the initial agreement), unless a risk
premium payment is received. Further, if information relating to the
settling PRP's waste contribution to the site is limited, then the Region
should also consider including a reopener which would allow the Agency
to seek additional funds from the settling PRP if new information relating
to its waste contribution to the site is discovered.
Stale Cost Share Requirements
The State cost share requirements apply as in mixed work settlement
(See Section on mixed vtorfc). Prices Landfill, New Jersey is an example
of a cash-out where a 10% State cost share was provided. The settle-
ment included PRP payment for a portion of the remedy, (the U.S. EM
is implementing it) plus accrued back interest to the United Stales, the
Stale of New Jersey and the Atlantic City Municipal Utilities Authority
(ACMUA) in exchange for a covenant not to sue for past and future
liability, subject to limited reopeners. In addition, a future Order of
the U.S. District Court provides that some money be set aside for future
remedial expenses, a percentage returned to the Superfund, the State
and the ACMUA. It should be noted that in the Prices case, the PRPs
are paying a sum of money which is not high enough to constitute a
"premium" and, yet, for which they will be relieved of future liability
at the site unless materially different information about the landfill or
new findings on imminent and substantial endangerment emerges. Pro-
visions in the Consent Decree are also included to maximize govern-
ment access to evidence linking additional generators to waste at the site.
CONCLUSION
Settlement agreements incorporating mixed funding provisions offer
an alternative to either up front Fund financing of the total costs of
a response action from the Superfund, or possible delays in initiating
a response action as a result of litigation to compel a action by a PRP-
594 STATE PROGRAMS
-------
Mixed funding is only one of the settlement tools available to the U.S.
EPA. Before one can begin an assessment of mixed funding for a
particular site, he must first determine whether under the Ten-Point
Settlement Criteria it is appropriate to settle for less than 100% of
response action costs.
Mixed funding is not appropriate for all circumstances and requires
a site/case-specific determination by the parties. Some of the consider-
ations to be made by parties to any settlement include:
• Whether the PRP desires to conduct the response action. If the PRP
is willing to conduct the cleanup, both mixed-work and preauthori-
zation may be considered.
• The knowledge and capabilities of the PRP to carry out the remedy
as designed. The more technical the remedy, the more important the
capabilities of the PPR.
• The amount of the work to be conducted by the PRPs. This may enable
the U.S. EPA to authorize a mixed-work settlement and conduct the
balance of the work, or "carve-out" the work to be conducted by
a subsequent settling PRP.
• The portion of the costs to be assumed by the PRPs. A cash-out may
be the appropriate result if the PRP is liable only for a small propor-
tion of the cost. Liability for a significant portion of the costs may
enable the use of mixed funding. The incentive for the PRP to manage
the response action in a manner to control costs is diminished when
the Fund pays a higher proportion of the costs.
• The most critical consideration is the PRP's "good faith offer'7 to
implement the response action, a significant portion of the work
and/or pay a significant portion of the costs.
Mixed funding settlements will be encouraged and adopted by the
Agency when such settlements are in the best interest of the Govern-
ment, the general public and the environment.
DISCLAIMER
The contents and conclusions of this paper are those of the authors
and do not necessarily reflect the views and policies of the U.S. EPA.
STATE PROGRAMS 595
-------
The U.S. Army's Installation Restoration Program:
Achievements and Initiatives
LTC Christopher P. Werle, P.E.
Office of the Assistant Chief of Engineers
Washington, D.C.
Kathleen A. Hutson
U.S. Environmental Protection Agency
Washington, D.C.
ABSTRACT
The Congress, the US EPA, the Stales and the public have clearly
stated thai Federal agency facilities should be the model of environ-
mental compliance for (he regulated community. In an effort to meet
this challenge in the hazardous waste program area, the Defense En-
vironmental Restoration Program (DERP) was established in 1984 to
expand the Department of Defense's (DOD) existing efforts to clean
up contamination from hazardous waste sites. SARA of 1986 provided
continuing authority for the Secretary of Defense to carry out this
program in consultation with the U.S. EPA. Furthermore, Executive
Order 12580 on Superfund Implementation delegated authority to the
Secretary of Defense to carry out the DERP within the overall frame-
work of SARA and CERCLA.
A component of the DERP is the Installation Restoration Program
(IRP) which is designed to identify and remediate contamination from
hazardous substances and wastes on DOD installations and at formerly
used properties. The objective of this paper is to outline the Army's
Installation Restoration Program which has evolved since its inception
in 1974 to be comprised of projects at over 401 installations involving
3,208 sites. The paper highlight!, some of the success the Army has
achieved in its quest to clean up its hazardous waste sites. Efforts under-
way at Anniston Army Depot are used to illustrate remedial actions
taken at an Army NPL site and the degree to which the Army is com-
mitted to restoring the environment.
INTRODUCTION
Pick up any newspaper or magazine and you will almost certainly
find an article addressing some sort of major environmental compliance
issue: the conviction and sentencing of three senior civilian managers
at Aberdeen Proving Ground, Maryland on multiple felony counts for
violating the Resource Conservation and Recovery Act; the FBI's
conduct of "Operation Desert Glow", where more than 70 agents
searched the Rocky Flats nuclear weapons plant for evidence of illegal
disposal of radioactive and hazardous materials; and the Hxxon Valdez
striking a reef off Prince William sound to create the largest oil spill
in United States history.
The magnitude of the hazardous waste problem is so great that it
almost defies comprehension. In addition to what lies buried in the thou-
sands of disposal sites that have been identified nationwide, the U.S.
EPA estimates that hazardous waste is generated in this country at the
rate of 700,000 tons/day; more than 250 million tons/yr. Recognizing
that Federal laws were needed to address the potential dangers of
abandoned hazardous waste sites, lawmakers passed CERCLA. As the
first major piece of legislation to address the problem on a national
level, CERCLA had the following key objectives:
• To establish priorities for cleaning up the worst hazardous waste sites
• To hold responsible parties liable for payment for those cleanups
(where possible)
• To establish a $1.6 billion Hazardous Waste Trust Fund (Superfund)
to perform cleanups when responsible parties could not be held
accountable, and to respond to emergencies involving hazardous
materials
• To improve scientific and technological capabilities in all aspects of
hazardous waste management, treatment and disposal.1
SARA, passed in 1986. outlined the framework for CERCLAs Super-
fund hazardous waste cleanup program during the next five years.
A major feature of the re-authorization was the clarification that "each
department, agency and instrumentality of the United States.. " was
required to comply procedurally and substantively with the statute
to the same extent as private entities (Section 120(a)(l) )•
The RCRA, an amendment to the Solid Waste Disposal Act, was
passed in 1976 to regulate the transportation, storage and disposal
of hazardous wastes (hat are being generated now. Passage of these
and a number of other statutes covering virtually all forms of pollu-
tion serve as positive testimony to the serious regard with which the
issue of environmental protection is being taken.
Within the regulated community, the Congress, the U.S. ER\ and
the public have clearly stated that Federal agency facilities should
serve as models of environmental compliance. In an effort to meet
this challenge in the hazardous waste area, the Department of Defense
(DOD) has continued to expand its effort through a wide array of
initiatives implemented as part of the Defense Environmental Resto-
ration Program (DERP). A component of the DERP is the Installa-
tion Restoration Program (IRP), which is designed to identify and
remediate contamination from hazardous substances and wastes on
DOD installations as well as at formerly used defense sites (FLTDS).
THE ARMY INSTALLATION RESTORATION PROGRAM
As one of the largest real estate holders in DOD (12.000,000 ac of
land on 1391 installations), the Army is keenly aware of its respon-
sibilities to protect and enhance the environment. In consonance with
its defense mission, the Army has established environmental quality
goals that will ensure the long-term protection of the land and
resources entrusted to its care. These environmental quality goals are:
• Demonstrate leadership in environmental protection and improvement
• Minimize adverse environmental and health impacts while maxi-
mizing readiness and strategic preparedness
• Assure that consideration of the environment is an integral part of
Army decision-making
• Initiate aggressive action to comply with all Federal. State and local
environmental quality laws
• Restore lands and waters damaged through our past waste disposal
596 STATE PROGRAMS
-------
activities
• Support Army programs for the recycle and reuse of materials to
conserve natural resources, prevent pollution and minimize the
generation of waste
• Pursue an active role in addressing environmental quality issues in
our relations with neighboring communities.
This philosophy has been carried forward from the initiation of the
Army IRP in 1974, a full 6 yr before the Superfund statute was even
on the books. From the first actions taken to clean up the monumental
problems at Rocky Mountain Arsenal to current research activities in
waste minimization, the Army has led the way in DODs efforts con-
cerning hazardous waste site remediation2. Figure 1 recaps Army IRP
performance as of the end of Fiscal Year 1988. It is obvious that sub-
stantial resources have been allocated to the program and the positive
results that have been obtained serve as the best testimony to its
success3.
3000
2000
1987
1988
Figure 1
Army IRP Progress
Under the direction of the U.S. Army Toxic and Hazardous Materials
Agency (USATHAMA), Preliminary Assessments/Site Inspections
(PA/SI) have been completed on almost all potential contamination
sources. The purpose of the PA/SI is to determine which sites may pose
a threat and require further cleanup action. Where remedial action is
anticipated, an RI/FS has been initiated to determine the extent of con-
tamination and develop a range of options to remediate the site. Working
with State authorities and U.S. EPA Regional and Headquarters offices,
a remedy will then be selected and documented in a Record of Deci-
sion. As shown in Figure 1, over 28% of those sources scheduled for
remedial action have been cleaned up. All remedial actions are
programmed for initiation by 1994.
In some cases, the magnitude of the hazardous waste problem has
warranted placement of Army IRP sites on the NPL. The NPL identi-
fies abandoned or uncontrolled hazardous waste sites that warrant further
investigation to determine if long-term "remedial action" is necessary.
Sites on the NPL are eligible for such action under CERCLA. However,
CERCLA Section lll(e)(3) generally prohibits use of Superfund dol-
lars for remedial action at Federally owned facilities. As of the July
1989 update, there are 18 Army facilities promulgated on the NPL and
an additional 17 Army facilities proposed for listing. The installations
proposed/final on the NPL are listed in Figure 2a and Figure 2b.
In conjunction with the cleanup of NPL sites, the Army has worked
closely with U.S. EPA Regional and Headquarters offices and the States
to execute several comprehensive cleanup agreements referred to as Inter
Agency Agreements (IAG). These agreements were first authorized
under Section 120 of the 1986 amendments and are binding, enforce-
able documents that cover the entire cleanup process from investiga-
tion through construction and operation of the remedy. lAGs are
generally designed to meet all of the facility's cleanup obligations under
1. ROCKY MOUNTAIN ARSENAL, CO
2. MILAN ARMY AMMUNITION PLANT, TN
3. ANNISTON ARMY DEPOT (SE IND. AREA), AL
4. CORNHUSKER ARMY AMMUNITION PLANT, NE
5. SACRAMENTO ARMY DEPOT, CA
6. SHARPE ARMY DEPOT, CA
7. SAVANNA ARMY DEPOT ACTIVITY, IL
8. LETTERKENNY ARMY DEPOT (PDO AREA), PA
9. FORT DIX (LANDFILL SITE), NJ
10. ALABAMA ARMY AMMUNITION PLANT, AL
11. JOLIET ARMY AMMUNITION PLANT (LAP AREA), IL
12. LETTERKENNY ARMY DEPOT (SE AREA), PA
13. FORT LEWIS (LANDFILL NO. 5), WA
14. LAKE CITY ARMY PLANT (NW LAGOON), MO
15. JOLIET ARMY AMMUNITION PLANT (MFG. AREA), IL
16. LONE STAR ARMY AMMUNITION PLANT, TX
17. UMATILLA ARMY DEPOT (LAGOONS), OR
18. LOUISIANA ARMY AMMUNITION PLANT, LA
Figure 2a
Army National Priority List Sites (Final)
1. FORTWAINWRIGHT.AK
2. FORTORD, CA
3. RIVERBANK ARMY AMMUNITION PLANT, CA
4 SCHOFIELD BARRACKS, HI
5. IOWA ARMY AMMUNITION PLANT, IA
6. FORTRILEY.KS
7. FORTDEVENS.MA
8. FORT DEVENS - SUDBURY TRAINING ANNEX, MA
9. ABERDEEN PROVING GROUND - EDGEWOOD AREA, MD
10. ABERDEEN PROVING GROUND - MICHAELSVILLE LANDFILL, MD
11. WELDON SPRING FORM. ARMY ORDDNANCE WORKS, MO
12. PICATINNY ARSENAL, NJ
13. SENECA ARMY DEPOT, NY
14. TOBYHANNA ARMY DEPOT, PA
15. LONGHORN ARMY AMMUNITION PLANT, TX
16. TOOELE ARMY DEPOT (NORTH AREA), UT
17. FORT LEWIS LOGISTICS CENTER, WA
Figure 2b
Army National Priority List Sites (Proposed)
CERCLA, RCRA and applicable State laws, and serve as "regulatory
blueprints" for the cleanup of the facility. In addition, they provide for
State and U.S. EPA oversight of the cleanup process and generally pro-
vide for reimbursement of State costs associated with the agreement.
In June, 1988, the U.S. EPA and DOD agreed to model language for
lAGs which resolved a number of national policy issues that were ham-
pering facility-specific cleanup negotiations. These model provisions,
to be included in each agreement, provide specific language on juris-
diction, funding, enforceability, dispute resolution, stipulated penal-
ties and RCRA/CERCLA integration. The agreements establish U.S.
EPA and State jurisdiction at the facility, provide for State and citizen
enforcement, as well as the assessment of penalties for failure to com-
ply with the schedule or terms and conditions of the cleanup, and for
the U.S. EPA Administrator to make the final decisions on cleanup or
any dispute arising under the agreement. By mid Fiscal Year 1989, seven
of the 16 lAGs in place with the U.S. EPA were for Army facilities.
In addition, the Army and the U.S. EPA currently are negotiating eight
additional agreements.
In August 1987, the Army, the U.S. EPA and the State of Minnesota
made history when all three parties signed the Twin Cities Army Ammu-
nition Plant (TCAAP) Federal facility InterAgency Agreement. The
TCAAP agreement was the first of its kind under Section 120 of SARA.
This agreement was the culmination of negotiations between the Army,
the U.S. EPA and the State and established provisions for:
• Coordination of overlapping requirements of RCRA and CERCLA
• Policies and procedures consistent with those for non- Federal
STATE PROGRAMS 597
-------
facilities
• U.S. EPA approval of selected remedies
• U.S. EPA and State oversight of the Army's activities at the site.
In addition to TCAAP, the Army has six other tri-pany IADs that
have been signed to-date, including:
• Joliet Army Ammunition Plan, Illinois (June 9, 1989)
• Letterkenny Army Depot, Pennsylvania (Feb. 3, 1989)
• Louisiana Army Ammunition Plant. Louisiana (Jan. 31, 1989)
• Milan Army Ammunition Plant, Tennessee (July 25, 1989)
• Sacramento Army Ammunition Depot, California (Oct. 27, 1988)
• Sharpe Army Depot, California (Mar. 16, 1989)
Besides evaluating its own installations, the Army, in particular the
U.S. Army Corps of Engineers (COE), is the DOD Executive Agent
for the implementation of Environmental Restoration Program opera-
tions at formerly used properties. As Executive Agent, the COE
(working through its Huntsville Division) is responsible for hazardous
waste cleanup activities, building demolition and debris removal, and
unexploded ordnance removals on lands formerly owned or used by
any of the DOD components. The investigation and cleanup procedures
at formerly used sites are similar to those at currently owned installa-
tions. Determinations must be made as to the origin of the contamina-
tion, land transfer and current ownership before a site is considered
eligible for restoration by the DOD.
ANNISTON ARMY DEPOT-A TYPICAL SUCCESS STORY
Situated on more than 15.000 ac in Calhoun County, Alabama, Annis-
ton Army Depot is one of 13 depots in the Army's Depot System Com-
mand (DESCOM). Anniston's primary mission includes combat vehicle
rebuild and conversion programs, small arms and artillery rebuild, main-
tenance of numerous missile systems and the storage of large quanti-
ties of ammunition. It is principally involved in the rebuilding of main
battle tanks for the U.S. Army and allied nations.
The industrial processes inherent with daily operations at the plant
result in the generation of several hazardous waste streams. Primarily
these wastes are degreasing solvents and metals processing sludges. For
many years, these wastes were disposed of either through: (I) being
placed in lagoons, trenches or pits which were later capped with earth,
or (2) being sealed in metal drums which were then buried. Ground-
water monitoring wells were emplaced around the chemical sludge dis-
posal trenches and old lagoon sludge piles (Fig. 3 and 4) in 1979 and
1980 to determine whether any contamination had migrated from the
disposal sites. It was subsequently determined that both sites had a high
potential for migration.
In February, 1981, DARCOM (now AMC) asked USATHAMA to
conduct an assessment at Anniston to determine the extent of contami-
nation and to develop plans for appropriate remedial action. USATHA-
MA worked in close coordination with U.S. EPA Region IV in Atlanta
to establish a program that would ultimately accomplish four main tasks:
• Conduct a geotechnical evaluation
• Prepare a contamination survey and assessment
• Conduct an alternatives analysis
• Effect closure operations
The first three tasks were completed by August, 1981. with
USATHAMA concluding that the most feasible technical/economic
remedial action would call for physical removal of contaminated soil
and other wastes from both sites. The wastes would then be transported
to a permitted hazardous waste disposal/treatment facility located in
Emelle, Alabama.
Closure operations began in March, 1982 with completion of
hazardous waste excavations by May, 1983. In all, more than 62,000
tons of waste were transported to Emelle without a single incident or
spill during the process. After completion of the excavations, exten-
sive soil sampling was conducted to insure that complete removal was
attained. The remediation process was completed by backfilling the
entire area, regrading to match the natural contours surrounding the
site and seeding to restore vegetative cover. According to Ron Grant,
Chief of the Environmental Management Division, Directorate of
Figure 3
Location of Survey Area: Calhoun County. Alabama
Figure 4
Site Layout Plan: Hazardous Waste Trenches
Engineering and Logistics for Anniston, work is now underway to
remediate contaminated groundwater by air stripping and returning it
to the environment4. The bulk of this work was completed 6 yr prior
to the Anniston site being listed by the U.S. EPA on the NPL.
NEW TOOLS FOR TRACKING
ENVIRONMENTAL COMPLIANCE
The preponderance of environmental legislation that Congress has
generated over the past decade has been instrumental to insuring that
598 STATE PROGRAMS
-------
our past mistakes will ultimately be rectified and that the likelihood
of future contamination will be greatly reduced or eliminated. Next
to passage of the landmark Superfund law, no other single statute has
had a more pronounced effect on the regulated community than RCRA.
Its broad application to the multitude of Army TSD (transport, storage
and disposal) facilities and the potentially severe consequences of non-
compliance have dictated the need for managers at all levels to have
immediate access to timely, accurate compliance data.
To establish such a data base from scratch would have been cumber-
some and resource intensive. A better solution would be to tap the
existing compliance data bases maintained by EPA. Housed on a large
mainframe system at Research Triangle Park, N.C., RCRA compliance
information could be accessed through modem by both DA staff ele-
ments and field commanders as well. Known as the Hazardous Waste
Data Management System (HWDMS), it tracks the complete compliance
history of all RCRA regulated facilities to include inspection perfor-
mance, enforcement actions taken and progress made to return facili-
ties to full compliance.
In July, 1989, an Interagency Agreement was signed between EPA
and DOD establishing a special account on the system for use by DOD
agencies. Use of this information will enable Army managers to main-
tain up to date status of all facilities, identify problem installations and
conduct long range planning to correct minor compliance problems
before they result in major enforcement actions.
In addition to HWDMS, EPA has also made available to both DA
and DOD use of a new PC-based multimedia data base called the Federal
Facility Tracking System (FFTS). Recently fielded in each of the 10
EPA Regional offices, FFTS enables the agencies to track compliance
in all media program areas. The systems' flexibility also allows the user
to access permitting data, enforcement histories, A-106 pollution preven-
tion project records and progress reports relating to the cleanup of
hazardous waste sites under the IRP. The capability to generate reports
in virtually any format, both on and off line rounds out the systems
attributes.
Use of both systems will enable the Army to closely monitor progress
made in all areas and to ensure better compliance in accordance with
the full spectrum of environmental regulations.
THE U.S. ARMY/U.S. EPA ENVIRONMENTAL
EXCHANGE PROGRAM
In early 1987, recognizing the need to expand the technical compe-
tence of his environmental staff, John Shannon, Assistant Secretary of
the Army (Installations and Logistics), and Lee Thomas, U.S. EPA Ad-
ministrator, sought to initiate an exchange program. Established under
the purview of Training With Industry (TWI), the program would afford
selected Army officers the opportunity to study environmental policy
and regulatory requirements while working within the major program
offices of the U.S. EPA.
The program is co-sponsored and administered through the Army
Environmental Office/Assistant Chief of Engineers on the Army staff,
and by the Office of Federal Activities at the U. S. EPA. Having just
completed its second year, with a total of five officers participating (two
at U.S. EPA HQ in Washington, D.C. and three at Regional offices
around the country), the program has been a resounding success. Under
the agreement established between the two agencies, these officers are
free to develop their own programs of study based on personal preference
and the job-specific requirements of their follow-on assignments. After
completing a full year with the U.S. EPA, the officers return for a mini-
mum 1 yr utilization tour to an environmental position within the Army.
Assignment of U.S. EPA personnel has thus far been limited to DA
staff in the Pentagon, but steps are being taken to place personnel on
several MACOM staffs as well.
The Army has just begun to realize the benefits of the program, not
only through the education of its environmental staff, but also through
the improvement in communication between the two agencies. The
program has done much to promote a willingness to work together in
tackling the seemingly overwhelming task of cleaning up the environ-
ment. With the continuation of programs such as this, we hopefully
will reach that end more quickly and efficiently.
CONCLUSION
Federal facility compliance with environmental laws and regulations
is one of U.S. EPA's highest priorities. Overall, Federal facilities have
made significant progress in improving their environmental compliance
records and in establishing/expanding their environmental programs.
The Army exemplifies the efforts being undertaken by the Federal sector
to "be the model of compliance."
DOD has established the Defense Environmental Restoration Account
(DERA), as outlined in SARA section 211, for the cleanup of its in-
active hazardous waste sites. FY 90 DERA funding for Army is expected
to increase from the $204.5 million allocated for FY 89. These funds
are exclusively earmarked for CERCLA activities.
As part of the effort to ensure that Federal agencies meet their en-
vironmental obligations and complete required CERCLA actions, the
U.S. EPA is dedicating significant resources to the Federal facilities
program as part of the effort to ensure high levels of environmental
compliance.
REFERENCES
1. Superfund: Looking Back, Looking Ahead, EPA 'Journal, April 1987.
2. Environmental Quality J987-1988, Annual Report of the Council on Environ-
mental Quality, Washington, DC
3. Defense Environmental Restoration Program, Annual Report to Congress for
Fiscal Year 1988, Office of the Deputy Assistant Secretary of Defense (Envi-
ronment), Washington, DC, Mar. 1989.
4. Remedial Action of Hazardous Waste Sites, Anniston Army Depot, Anniston,
AL, brochure published by the U.S. Army Toxic and Hazardous Materials
Agency, Aberdeen Proving Ground, MD, June 1983.
STATE PROGRAMS 599
-------
Assessing a Potentially Responsible Party's Ability to
Pay Superfund Cleanup Costs
Anthony M. Diecidue
U.S. EPA
Washington, D.C.
Mark F. Johnson, MBA, ARM
Laurie A. Manderino
Kenneth E. Anderson
PRC Environmental Management Inc.
McLean, Virginia
ABSTRACT
This paper presents a methodology for assessing a PRP's ability to
pay past and future costs under CERCLA as amended. The objective
of the Superfund program is to impose the ultimate liability for the cost
of cleanup of hazardous waste sites on the responsible panics who
generated the waste and on those who owned or operated the sites. The
U.S. EPA's enforcement program can accomplish this goal in two ways.
No matter how the U.S. EPA chooses to proceed with enforcemem,
an accurate financial assessment of the PRP's ability to pay developed
early in the enforcement process will assist the U.S. EPA in determining
the most effective enforcement strategy and in establishing the most
realistic settlement or cost recovery amounts.
This paper presents an approach to assessing a PRP's ability to pay
based on the PRP's operating and Financial statements. The paper goes
on to describe four types of standard financial indicators, measured
according to eight standard Financial ratios, that can provide an overall
picture of a PRP's past and present financial condition and ability to
pay. These four financial indicators or ratios are: (1) liquidity, (2)
leverage, (3) solvency and (4) profitability. After discussing and inter-
preting these financial indicators, the paper illustrates the methodology
through a hypothetical assessment of a PRP's ability to pay. This case
study: (1) analyzes a hypothetical PRP's ability to pay CERCLA cleanup
costs by calculating and interpreting the financial ratios. (2) presents
the results of sensitivity analyses that measure the impact of various
cost recovery schedules on the hypothetical PRP's current financial con-
dition and (3) discusses the sources of information on a PRP's financial
status.
INTRODUCTION
The Superfund program attempts to place the ultimate responsibility
for the cost of cleaning up hazardous waste sites on responsible parties
who generated the wastes and on those who owned or operated the sites
The U.S. EPA can accomplish this goal in two ways, but in either case-
will seek to directly impose the cost of cleanup on responsible parties
The Agency can conduct the cleanup using money from the Superfund
and later seek to recover the cleanup costs through cost recovery actions;
or the U.S. EPA can use a variety of CERCLA enforcement authori-
ties to directly compel responsible parties to finance or conduct cleanups.
The U.S. EPA's recent 90-day management review of the Superfund
program further indicates that the U.S. EPA's increasing emphasis on
enforcement will further induce PRPs to rapidly achieve enforceable
agreements to finance or carry out more cleanups under the U.S. EPA
direction.
The U.S. EPA will rely on the various enforcement authorities to
pursue responsible parties to recover costs and replenish the Super-
fund when the Agency conducts a Fund-financed cleanup without a
negotiated settlement. Congress authorized the U.S. EPA in Section
104 of CERCLA to spend Superfund monies to clean up a site and in
Section 107 to pursue responsible parties for recovery of cleanup cons.
Although the U.S. EPA does not hesitate to pursue cost recovery actions
or litigation in the interest of public health and the environment, the
Agency still prefers to negotiate settlements with responsible parties.
Section 122 of CERCLA provides the U.S. EPA with the discretion
to enter into settlement agreements with PRPs. establishes procedures
for negotiating settlements with PRPs for financing or conducting Super-
fund site cleanups and codifies, with some additions, the settlement
process established under the U.S. EPA's interim settlement policy.
The US EPA's Interim CERCLA Settlement Policy (50 FR 5034)
sets as the U.S. EPA's objective in Superfund negotiations the collec-
tion of 100% of cleanup costs or complete conduct of cleanup from
PRPs. The Agency recognizes, however, that in certain circumstances,
exceptions to this goal may be appropriate. The interim CERCLA set-
tlement policy sets forth K) criteria for determining when such excep-
tions are allowed. Based on a full evaluation of the facts and a
comprehensive analysis of the K) criteria, the U.S. EPA may consider
accepting offers of less than 100% of the total amount. One of the tea
criteria the U.S. EPA may consider in evaluating a settlement proposal
is the ability of the settling parties to pay.
Determining a PRP's ability to pay cleanup costs is important whether
the U.S. EPA conducts the cleanup and seeks cost recovery or obtains
a negotiated settlement with PRPs to finance or conduct the cleanup.
Assessing the PRP's ability to pay early in the settlement negotiation
process is an important step toward achieving rapid and effective settle-
ments An accurate financial assessment of PRPs early in the enforce-
ment process will assist the U.S. EPA to determine the most
cost-effective enforcement strategy, establish realistic settlement or cost
recovery amounts, decide whether a PRP is financially sound enough
to conduct future remedial work, obtain corporate information that can
facilitate a rapid and effective negotiation process and verify whether
a PRP has a genuine ability-to-pay problem.
DISCUSSION OF AN APPROACH TO
PRP FINANCIAL ASSESSMENT
The proposed methodology for evaluating a PRP's financial condi-
tion and ability to pay is based on a standard method used by the finan-
cial community. The methodology relies on four financial indicators
commonly used by the financial community (for example, bank loan
officers, investment bankers, security analysts, a firm's management
and stockholders) to measure corporate performance, generally to assess
a firm's viability where possible bankruptcy is not an issue, (In bank-
ruptcy, the same liquidity, leverage, solvency and profitability ratios
arc used to check a reorganization plan and to monitor progress.)
600 STATE PROGRAMS
-------
Managers generally monitor liquidity to determine the timing of pay-
ments, while lenders generally focus on leverage and solvency (or
coverage) ratios as decision-making criteria for establishing new lines
of credit or loans. Conversely, stockholders generally look at profita-
bility to make stock buy/sell decisions7. Financial analysts measure a
firm's financial indicators against industry norms to assess the firm's
current financial condition and to identity positive or negative trends
in a firm's liquidity, leverage, solvency and profitability positions.
Limitations of Financial Ratio Analysis
Financial analysts indicate that there are four key limitations to
assessing a firm's financial condition using the financial ratio analysis
methodology. First, values for financial ratios vary substantially across
industrial sectors. Industry-specific norms or averages should there-
fore be used as benchmarks for comparison when evaluating individual
firms. To overcome this limitations, data base reports can provide in-
formation on industry norms for liquidity, leverage, solvency and profita-
bility indicators. These norms are fairly stable from year to year in
the absence of extreme economic conditions.
Second, no single ratio or indicator is a perfect test of ability to pay,
particularly when a firm is in distress2. However, a set of indicators
that considers liquidity, leverage, solvency and profitability will generally
be informative1'7. Third, analysts have established minimum accepta-
ble or threshold levels for financial ratios that reflect values the finan-
cial community considers "red lights" signaling possible financial
distress. Fourth, historical ratios over the past 3 to 5 yr should be
checked for trends and unique occurrences.
One approach to assessing a PRP's current financial condition and
ability to pay is based on the firm's operating and financial statements
and four types of financial indicators designed to measure a PRP's finan-
cial condition and performance. In the proposed approach, a set of
standard financial indicators is calculated to measure a PRP's past and
present financial performance and to determine its ability to pay past
and future response costs. Generally, four types of financial indicators
are-calculated: (1) liquidity, (2) leverage, (3) solvency, and (4) profita-
bility.
To conduct the financial assessment, eight standard financial ratios
are examined, including two each that measure liquidity, leverage, sol-
vency and profitability. First, a financial spreadsheet is developed from
the PRP's audited financial statements to calculate historical and cur-
rent financial ratios to obtain an overall picture of a PRP's past and
present financial condition and ability to pay. Each of the PRP's ratios
is then compared with industry ratio averages. An assessment of a PRP's
financial condition and ability to pay is made in part by comparing a
PRP's ratios relative to the average ratio values for other firms in its
industry. Finally, the ratios are subjected to a sensitivity analysis to
determine the effect of various CERCLA cost recovery levels on the
PRP's financial viability. The eight financial ratios used in the proposed
approach are discussed below.
1. Liquidity Ratios
Liquidity ratios measure a firm's ability to meet its short-term ex-
penses and other financial obligations in a timely manner. The current
ratio (CR) and quick ratio (QR) described below are two standard
liquidity ratios that measure a firm's resources or available cash7.
Current Ratio: CR = (Current Assets)/(Current Liabilities)
The CR is the sum of cash and cash equivalents (principally, accounts
receivable and inventories) divided by current liabilities (principally,
accounts payable, taxes and short-term bank loans); that is, the CR test
is calculated as the ratio of current assets to current liabilities. The CR
measures liquid assets available to pay expected invoices and monthly
and periodic bills; it is equivalent to the multiple of a firm's current assets
to its liabilities.
A CR greater than 3.0 indicates more than adequate cash and cash
equivalents to meet short-term requirements. A CR in the range of 2.0
to 3.0 generally indicates sufficient resources. A CR value of less than
2.0 generally signifies potential future liquidity problems'.
Quick Ratio: QR = (Current Assets -
Inventories)/(Current Liabilities)
The QR is the current ratio adjusted for the value of inventories by
excluding inventories in process because finished and unfinished projects
generally cannot be converted into cash immediately to pay current
liabilities. The QR is also known as the "acid test" ratio. A QR greater
than 1.0 indicates sufficient liquidity for expected short-term business
expenses.
2. Leverage Ratios
Leverage ratios provide information on the extent of debt in the com-
pany's capital structure, the long-run ability of the firm to repay borrowed
funds and, indirectly, management's degree of risk aversion and its
business philosophy. These ratios reflect the firm's financing or capitali-
zation by comparing debt to equity and debt to assets; large debt balances
indicate a higher probability of credit risk and default plus substantial
debt servicing costs. Two standard indicators of a firm's degree of
leverage are the debt to equity ratio (DER) and debt to assets ratio
(DAR).
Debt to Equity Ratio: DER = (Long Term Debt
+ Capitalized Leases)/(Stockholders' Equity)
The DER is defined as long-term debt plus capitalized lease obliga-
tions divided by stockholders' equity or net worth. Capitalized lease
obligations are included because leases are, in many respects, equiva-
lent to secured loans4. For blue-chip Fortune 500 companies, the DER
is substantially less than 1.0 and is generally in the 0.3 to 0.4 range7.
A DER greater than 1.5 signals possible debt servicing problems14.
Debt to Assets Ratio: DAR = (Current
+ Long Term Liabilities)/(Current + Long Term Assets)
The DAR is defined as total debt or total liabilities (the sum of current
liabilities and long-term debt) divided by total assets (the sum of cur-
rent assets plus long-term assets). A DAR greater than 0.65 and in-
creasing in subsequent years is evidence that the firm has doubtful ability
to service its debt2. A value of 1.0 (or greater) demonstrates that the
firm has zero equity (or negative equity or net worth) since liabilities
equal (or exceed) assets.
3. Solvency Ratios
Solvency or coverage ratios measure a firm's ability to remain in
business without substantial infusions of new equity, major liquidation
of corporate assets or other significant changes in operations or cor-
porate behavior. Solvency ratios are designed to evaluate the firm's ability
to cover its financing charges and debt exposure. Two principal indi-
cators of solvency are the fixed charge coverage (FCC) and cash flow
coverage (CFC) ratios.
Fixed Charge Coverage Ratio: FCC =
(Earnings Before Interest & Income Tax)/
Fixed Payments plus Current Debt Due)
The FCC ratio is calculated as the ratio of earnings before interest
and income tax to fixed charges that must be "covered." These fixed
expenses include lease payments, insurance, interest charges on debt
and current-period principal payments due on long-term debt. An FCC
ratio greater than 2.0 signifies acceptable coverage or solvency. Values
less than 1.5 generally point to questionable viability14. Negative
values indicate inability to pay fixed expenses and potential imminent
financial insolvency.
Cash Flow Coverage Ratio: CFC =
(Cash Flow/Total Liabilities)
The CFC ratio is calculated as cash flow (net income after tax plus
depreciation and amortization) divided by total liabilities. This indi-
cator represents internally generated sources of funds that are availa-
ble to meet the company's long-term debt obligations and current
liabilities2'9'10. A CFC ratio greater than 0.4 indicates more than suffi-
cient cash flow to service liabilities with internal resources. A value
in the 0.2 to 0.4 range reflects adequate cash flow, and a CFC ratio
STATE PROGRAMS 601
-------
less than 0.15 indicates serious financial uncertainty.
4. Profitability Ratios
Profitability indicators reflect operating performance as measured by
net earnings or after-tax income in relation to capital supplied by the
firm's owners or in relation to assets deployed. These ratios compare
profitability to funds invested in the firm. Two accepted measures of
profitability are return on equity (ROE) and return on assets (ROA)'.
Return on Equity: ROE = (Net Income
After lax)/(Stockholders' Equity)
The ROE is calculated as net earnings or income after tax divided
by stockholders' equity. The ROE measures the percentage return to
the firm's owners or stockholders and is an index of the productivity
or efficiency or shareholders investment. ROE values must be higher
than the firm's cost of capital for long-term continued viability.
Return on Assets: ROA = (Net Income
After Tax)/flbtal Assets)
The ROA is calculated as net earnings or income after tax, adjusted
for the tax savings associated with debt financing divided by total assets
(current and long-term). The ROA measures the productivity or effi-
ciency of asset deployment and corresponds to the net operating profit
margin. A value less than 0.06 may be a sign of doubtful viability1
SENSITIVITY ANALYSIS
Once an assessment of the firm's overall financial condition has been
completed by calculating the eight financial ratios, the analyst can
perform a sensitivity analysis to determine the impact of different levels
of cost recovery or cleanup payments on the financial condition of the
firm. The current-year financial data, and the ratios derived from them.
serve as a "baseline" for completion of the sensitivity analysis. Gener-
ally, the impact of a one-time, lump-sum payment on the current-year
condition is analyzed, but it also is possible to examine the impact of
a "structured settlement," periodic payments over a specified number
of years.
There are two primary ways in which a company could fund a cost
recovery or cleanup payment; by drawing on current assets, (assuming
that these assets can be liquidated) or by securing a loan. The analyst
can calculate the effect of both scenarios on a company's baseline finan-
cial condition. Generally, four different levels of cost recovery pay-
ments are analyzed to provide an indication of the maximum amount
that the company could afford to pay without imposing a severe finan-
cial burden on the firm's operations.
To analyze a payment funded from a firm's current assets, the analyst
will first revise the firm's financial spreadsheet to reflect the reduction
in current assets that would result from each level of payment examined.
Then, the financial spreadsheet is used to calculate revised liquidity
ratios for the current year that reflect the impact of each of the four
payment levels. Finally, the revised liquidity ratios are compared with
the firm's baseline ratios and industry averages.
A payment funded by long-term debt affects the firm's leverage and
solvency ratios. To analyze this situation, the financial spreadsheet is
revised again, this time to reflect the addition to the firm's liabilities.
The financial spreadsheet is then used to recalculate the ratios, and these
new ratios are compared with the baseline ratio and the industry aver-
ages to provide an indication of how the financial status of the firm
would be affected by different payment levels.
CASE STOW: XYZ COMPANY
A hypothetical case was created to illustrate how the financial
assessment methodology presented in this paper may assist the U.S.
EPA to assess a PRP's financial position and determine its ability to
pay past and future CERCLA cleanup costs. This hypothetical case was
patterned after several real CERCLA cases involving settlement negotia-
tions for future Rl/FS work and cost recovery for past costs.
Background on the Hypothetical XYZ Company Case
The hypothetical XYZ Company (hereafter XYZ) has owned and
operated a small manufacturing plant in the eastern United States for
more than 30 yr. XYZ has approximately $25 million in annual revenue
and has experienced net income losses for 2 yr, primarily because of
unplanned plant shutdowns to repair aging manufacturing equipment;
however, sales continue to be strong and XYZ managers are confident
they can modernize the plant and restore profits in the near future. XYZ
is the major employer in a small rural town, employing approximately
TOO people.
Over the years, XYZ has stored or disposed of a variety of hazardous
contaminants on-sitc in surface impoundments, waste piles and drums.
Hazardous waste is known to be migrating off-site, contaminating
groundwatcr below the site and at least one nearby off-site well. The
U.S. EPA believes that XYZ is the sole PRP.
To date, the U.S. EPA has conducted a preliminary assessment and
site investigation (PA/SI) at the site and has proposed the site for inclu-
sion on the NPL. The U.S. EPA currently is developing a work plan
with details for conducting future Rl/FS work at the site. The U.S. Eft
has incurred approximately SI million in past costs at the site and has
placed a $1 million lien on XYZ's real property. The U.S. EM ant
XYZ are negotiating to allow XYZ to assume lead responsibility for
future Rl/FS work at the site.
Before allowing XYZ to take on the Rl/FS work, the U.S. EM in-
tends to recover its past costs. In negotiations with the U.S. EM, XYZ
has been cooperative and willing to fund future Rl/FS work out of none
earnings or bank loans, but professes an ability-to-pay problem to
currently repay past costs. Furthermore, the U.S. EPA's lien prevents
XYZ from borrowing from commercial banks. During settlement negoti-
ations, XYZ requested that the U.S. EPA allow the company to reply
past costs over time and fund Rl/FS work out of future earnings starting
in 1989. To successfully resolve the settlement negotiations, the U.S.
EPA must quickly assess whether XYZ has a genuine ability-to-psy
problem, establish a realistic amount that XYZ can pay without
bankrupting the company and determine whether XYZ is financially
sound enough to complete the future Rl/FS work the company hopes
to undertake.
Conducting the Hypothetical Financial A
To prepare the hypothetical ability-to-pay analysis for XYZ. a spread-
sheet is constructed to calculate the eight financial ratios using the finan-
cial ratio formulas previously discussed. The spreadsheet (Lotus 123)
allows the analyst to quickly calculate the ratios and conduct sensiti-
vity analyses. Data from XYZ's audited 1984-1988 financial statements
are entered into the financial spreadsheet. Using the spreadsheet, the
analyst first calculates XYZ's historical (1984-1987) and current 0988)
financial ratios to observe the trends in XYZ's financial position. These
ratios are then compared with XYZ's industry norms for the eight ratios
(compiled yearly and published in standard financial reference boob
and by Dun & Bradsirect*, and potential problems are identified. The
analyst then conducts sensitivity analyses on XYZ's 1988 financial ratios
to test the impact of various cost recovery payments on XYZ's current
financial position.
Table 1 summarizes the results of the spreadsheet analysis. It presents
XYZ's eight historical financial ratios (1984-1987) and XYZ's current
financial ratios (1988), XYZ's industry average for each of the eight
financial ratios, and generally accepted danger zone levels for the finan-
cial ratios. Table 1 also shows the impact of four different cost recovery
payments ($250,000, $500,000, $750jOOO and $1 million) on XYZ's 1988
financial ratios. The cost recovery payments used in the sensitivity
analysis were selected by the analyst based on the facts of the case and
the range of costs that the analyst considered possible for use in settle-
ment negotiations. The 1984-1988 financial ratios provide an overall
picture of XYZ's historical and present financial condition and ability
to pay. XYZ's 1988 financial ratios are compared with industry ratio
averages; Table 1 shows XYZ's industry average for each of the eight
financial ratios. An assessment of XYZ's financial condition and ability
to pay is made using professional judgment by comparing XYZ's finan-
cial ratios to the average ratio values for other firms in its industry and
assessing the impact of various cost recovery payments on XYZ's finan-
cial condition through sensitivity analysis.
602 STATE PROGRAMS
-------
Table 1
XYZ Company Financial Ratios (1984-1988) and Impact of Various Cost
Recovery Payments
On the 1988 Financial Ratios
,. Cuncnl Rub) 0.90 113 3J» 160 1.90
i Oiick Rilio 030 1.80 147 112 0.90
1.75 1.67 1J9
0.82 0.74 0.66
IJ1
OJ4
}. Debt/Equity 0.00 0.00 0.00 0,00 0.00
<. Dda/AlKU 039 016 033 0.24 031
1. F«td diirjtd
Ccwl(e 1.45
2.10 0.88 -3.03 -153
OSS 0.80 032 -0.12 -0.10
7. Return m
Equity
0.10 030 0.09 0.01 4.10
0.12 0.14 -OJ)1 -OJM 4.15
0.00 0.00 0.00 0.00
0.32 033 034 035
•103 -1.73 -U3 -1.40
-0.16 -0.22 -0.28 -034
•O.U -0.17 421 -036
•0.17 -0.19 -0.22 -OiS
1.90
1.00
0.40
0.50
>L5
>.65
Interpreting the Hypothetical Financial Assessment
The financial ratio calculations presented in Table 1 assist the analyst
to understand historical trends in XYZ's financial position over time
and enable the analyst to make professional judgments about XYZ's
current financial position and ability to pay. The sensitivity analyses
assist the analyst to determine the impact of various cost recovery pay-
ments on XYZ's current financial ratios and present a range of possi-
ble cost recovery amounts to be suggested in settlement negotiations.
By interpreting the financial ratios, the analyst is able to decide the
U.S. EPA's strategy in negotiations with XYZ. Although it is beyond
the scope of this paper to discuss all the results shown in Table 1, an
example related to a specific financial ratio (the cash flow coverage
ratio) will illustrate the concept.
Table 1 summarizes XYZ's cash flow coverage (CFC) ratio for the
years 1984-1988 and at the four different cost recovery payment levels
for 1988. The industry norm for the CFC is approximately 0.156. A
CFC ratio greater than 0.4 generally indicates that a company has more
than sufficient cash flow to service liabilities with internal resources.
A value in the 0.2 to 0.4 range reflects adequate cash flow, and a CFC
ratio less than 0.15 indicates serious financial uncertainty. By reviewing
Table 1, the analyst determines that the CFC ratio for XYZ varies from
year to year depending on fluctuations in XYZ's net income. The CFC
ratio was excellent in 1985 and more than adequate, compared to the
industry norm, in 1984 and 1986. The negative CFC ratios for 1987
and 1988 reflect XYZ's net losses as a result of reduced revenues from
plant shutdowns and increased expenses from plant maintenance. By
reviewing the CFC ratio in Table 1, the analyst is able to determine
that if the U.S. EPA imposes a cost recovery payment in 1988, XYZ's
CFC ratio would drop into the "danger zone." This indicates that XYZ
would be unable to fund a payment from current net income generated
by operations, without seriously affecting existing operations. A nega-
tive CFC ratio indicates that XYZ may be headed toward bankruptcy,
since the ratio must be positive for financial viability. As a result of
the analysis, the analyst concludes that even a cost recovery payment
in the range of $250,000 to $500,000 may seriously impede XYZ's ability
to maintain current operations and could force XYZ into bankruptcy.
To complete the analysis, the analyst must review all eight ratios and
assess their meaning in relation to XYZ's historical ratios, industry
norms and danger zone levels. In summary, XYZ's liquidity position
is good and should not be severely affected by a cost recovery payment
in the range of $250,000 to $500,000. Although liquidity ratios would
fell below industry averages, they would not drop into the "danger zone"
for the financial ratios discussed previously. A cost recovery payment
above $500,000 may, however, seriously impair XYZ's current opera-
tions and may lead to bankruptcy. Table 1 indicates that XYZ's leverage
indicators are strong, but that the company's ability to borrow addi-
tional funding for a cost recovery payment is doubtful because of the
U.S. EPA lien on XYZ's real property. The solvency ratios indicate
XYZ's cash flow is insufficient, so that current obligations may be
difficult to meet. Finally, XYZ's profitability ratios are negative, in-
dicating an inability to raise new capital through stock issues and
reflecting current, and possibly future, financial problems.
It is apparent from the financial ratio analysis presented in Table 1
that XYZ may be able to fund an immediate cost recovery payment
only from current assets. Both the solvency ratios and the profitability
ratios are negative, indicating an inability to fund a cost recovery pay-
ment by using earnings from XYZ's operations or by raising new capi-
tal. In addition, the lien on XYZ's real property discourages prudent
lenders from providing XYZ with additional funds. As a result, XYZ
cannot borrow money to make a cost recovery payment or fund future
RI/FS work. If XYZ makes a cost recovery payment of $250,000 to
$500,000, it is possible the company may be forced to raise additional
funds from external sources to continue existing plant operations. This
may be unlikely for XYZ in the short term. The analysis indicates that
if the U.S. EPA wishes to recover a portion of past costs immediately,
a lump-sum cost recovery payment from XYZ appears possible in the
range of $250,000 to $500,000. However, given the company's current
tenuous solvency, the U.S. EPA may risk hastening XYZ's insolvency
by imposing even this range of cost recovery payment on XYZ at this
time.
AVAILABLE FINANCIAL INFORMATION SOURCES
It is necessary to gather financial information on the PRPs to
accurately assess their ability to pay. Depending on the level of analy-
sis required, many sources of financial information are available to as-
sist the U.S. EPA in establishing the financial viability of PRPs. To
assess a PRP's ability to pay CERCLA cleanup costs, a minimum of
three financial information sources should be examined: (1) National
Enforcement Information Center (NEIC) Superfund Financial Assess-
ment System (SFFAS) financial information, (2) recent audited finan-
cial statements of the PRP and (3) commercial financial data bases and
financial references. This section discusses these three primary sources
of financial information on PRPs and briefly reviews other common
sources of financial information.
NEIC SFFAS Financial Information
NEIC provides financial information to the U.S. EPA personnel to
assist them in determining a PRP's financial status. NEIC operates
SFFAS on publicly held companies as a tool to assist enforcement
personnel in negotiating with PRPs. SFFAS was designed to: (1) calcu-
late the amount of response action costs a PRP can afford to pay and
(2) provide a concise financial evaluation of the PRP.
The model consists of two components. First, it calculates a PRP's
ability to pay by measuring the cash flows from the company's opera-
tions and their variability to determine the company's ability to main-
tain its current business and pay response costs. Second, it applies three
standard financial ratios to assess whether additional debt may be feasi-
ble for the firm. SFFAS requires a, minimum of 3 yr of annual data
on net income and depreciation, and data on the company's current
liabilities, long-term debt, net worth, interest expense and income tax
rate for the most recent year.
The NEIC SFFAS does not include privately held companies, where
financial information is not readily available. In these cases, NEIC can
usually provide Dun and Bradstreet reports for privately held compa-
nies not listed in the SFFAS data base. The Dun and Bradstreet
reports' can be used to initially determine various aspects of a PRP's
financial status; however, in-depth financial analysis based on the PRP's
actual financial statements is recommended to determine ability to pay.
Recent Operating and Audited Financial Statements
of the PRP
If the PRP's capital stock is publicly traded, it is required under the
Security Acts of 1933 and 1934 to file various reporting forms with
the Securities and Exchange Commission (SEC) in Washington, D.C.
These forms contain valuable information about the company's opera-
tion, financial condition and ownership that can be used to assess a
PRP's current and future financial condition. Three primary SEC forms
STATE PROGRAMS 603
-------
are discussed below:
Form S-l: This is the most comprehensive form that companies
may be required to file with the SEC under the
1933 Act. The S-l provides specific information
on a company's use of the proceeds when it issues
capital stock. Generally, this information helps in-
vestors to make an informed decision. The S-l
contains financial statements and descriptions of
the filing company's property, directors and execu-
tive officers, and security ownership of
management.
Form 10: This form is similar to the S-l. Companies that
issue capital stock, and are not required to file
under the 1933 Securities Act, file under the 1934
act using Form 10. Form K) contains similar.
though less detailed, information as the S-l.
Form 10-K: This is the annual report that most reporting com-
panies file with the SEC. It provides a compre-
hensive overview of the registrant's business,
including a description of the business, select
financial data, financial statements, management's
discussion and analysis of the financial condition
and results of operations, and security ownership
of management.
These documents may be obtained by contacting the company directly,
or from the SEC in Washington, D.C. The SEC is located at 450 5th
Street NW, Washington, D.C., (202) 272-7450; documents must be
obtained in person at the Public Reference Room (PRR) in room 1024.
An information data base in the PRR provides key word searches for
companies included and describes the forms these companies have filed.
The most important source of financial information needed to con-
duct a financial ability to pay assessment are the PRP's audited finan-
cial statements. Generally, the past 5 yr of a PRP's audited financial
statements (including balance sheets, income statements, sources and
uses of funds) should be obtained to assess the PRP's financial condi-
tion and ability to pay. The financial statements should be audited by
a certified public accountant to ensure their validity. PRPs may volun-
tarily submit financial statements, or they may be demanded in CERCLA
)04(e) notice letters.
The annual report is another valuable source of information about
a PRP's financial position. The financial information in an annual report
is, however, usually unaudited. Annual reports can be obtained from
state corporation commission offices in the state where the company
is incorporated or located. The corporation commission may be the
only alternative for information about PRP companies whose stocks
are not publicly traded and, therefore, do not report financial informa-
tion to the SEC.
Commercial Financial Data Bases
and Financial References
Commercial financial data base services can expedite the search for
financial information when PRPs are privately held and their financial
statements are not readily available, or when the analyst must screen
a large number of PRPs. Standard financial references also can provide
valuable information on historical and future trends on the PRP's finan-
cial position. Several commercial financial data base services and
standard financial references are available to assist in determining a
PRP's ability to pay. The primary financial data base services and
references are briefly discussed below.
Dun & Bradstrect Financial Data Base
The Dun & Bradstreet (D&B) data base provides financial statements
for up to 3 yr on more than 850,000 United States companies, both
public and private. Along with financial statements, D&B provides key
business ratios for more than 800 industries. These ratios compare the
company's performance to others in its industry, providing insights into
its financial condition. The ratios show a company's solvency, busi-
ness efficiency and profitability, providing a basis to evaluate a
financial condition quickly according to objective, quantitative measures
of performance. It should be noted, however, that D&B reports are not
always available for many privately held companies and may lag as much
as 6 to 12 mo behind the company's most current financial data.
D&B also can provide a financial profile report. This report shows
financial trends, profit performance and a company's relative position
within its industry. This report also contains industry norms that provide
the analyst the ability to compare quickly each item on the financial
statement with the industry's average.
Information America Financial Data Base
Information America provides on-line Secretary of State Corporate
and UCC filings, local county records, U.S. bankruptcy court filings,
as well as on-line document ordering services. Duns Business Recoidi
Plus reports can be obtained through Information America. The Duns
report provides in-depth historical information on the operation and
finances of private and public United States businesses. An analyst can
quickly obtain information on businesses of any size, and financial data
on more than 750,000 private and public companies. Information
America's main advantage over a strictly financial data base a that it
offers multiple information services from one source.
Standard & Poors Data Bases
Standard and foors (S&P) maintains a number of data bases on United
States businesses. These data bases are available through the Dialog
Information Retrieval Service and include financial information on
36,000 corporations and 340,000 "key executives," with 74jOOO profile
biographies. S&P also can provide substantial information on finan-
cial institutions.
Standard Financial References and Directories
Standard financial references and directories may provide another
excellent source of historical financial information. These references
generally are available through business libraries. While too numerous
to discuss in this paper, some of the more common standard financial
references include America's Corporate Families, America's Corporate
Families and International Affiliates, Directory of Companies Filing
Annual Reports With the Securities and Exchange Commission, Direc-
tory of Corporate Affiliations—Who Owns Whom. Dun and Bradsntet's
Middle Market Directory and Million Dollar Directory, Funk and Scon's
Index of Corporations and Industries, Moody's Industrial Manual and
other publications. Standard & Poor's Register of Corporations, Direc-
tors and Executives, Value Line, Almanac of Business and Industrial
Financial Ratios, Handbook of Business and Financial Ratios, the Direc-
tory of Public High Technology and Medical Corporations, the Inter-
national Directory of Company Histories, and the Corporate Technology'
Directory.
Other Potential Sources of Financial Information
When conducting a financial assessment, it may be necessary in some
cases to research a variety of other information sources to gain a dear
picture of the PRP's financial position and ability to pay. Additional
financial information often can be obtained by contacting various
agencies at the county, city, state and federal levels. Although these
sources are too numerous to discuss in this paper, a U.S. General
Accounting Office publication, "Investigator's Guide to Sources of
Information" (GAO/OSI-88-1), provides a wealth of information on the
types of financial information available from governments, governmental
agencies and commercial sources.
CONCLUSION
Assessing a PRP's ability to pay early in the enforcement process
is an important first step toward achieving rapid and effective cleanups
and settlements. An accurate financial assessment of a PRP's ability
to pay early in the enforcement process will assist the U.S. ERA to de-
termine the most cost-effective enforcement strategy, determine realis-
tic settlement or cost recovery amounts, obtain corporate information
that can help to facilitate a rapid and effective negotiation process.
604 STATE PROGRAMS
-------
determine whether a PRP is financially sound enough to conduct future
remedial work and verify whether a PRP has an ability-to-pay problem.
This paper has discussed one approach to assessing a PRP's ability
to pay. The approach is based on the PRP's operating and financial
statements and four types of standard financial indicators designed to
measure financial condition and performance.
The objective of the financial assessment is to assist the U.S. EPA
personnel to accurately determine a PRP's ability to pay past or future
response costs. The approach can be adapted to a variety of CERCLA
settlement or cost recovery cases involving individuals and single or
multiple companies who are PRPs. The approach also can be used to
determine appropriate RCRA penalty assessment levels to impose on
companies facing penalties under RCRA.
DISCLAIMER
Any views or opinions expressed in this paper are the views of the
authors and do not necessarily represent the views of the U.S. the U.S.
EPA.
REFERENCES
1. Beaver, W. H., "Financial Ratios as Predictors of Failure. Empirical Research
in Accounting: Selected Studies," Supplement to J. of Accounting Res., pp.
77-111, 1966.
2. Beaver, W. H., "Alternative Accounting Measures as Predictors of Failure,"
Accounting Rev., pp. 113-122, Jan. 1968.
3. Beaver, W. H., "Market Prices, Financial Ratios, and the Prediction of
Failure," /. of Accounting Res., pp. 179-192, Autumn, 1968.
4. Brealey, Richard, et al., Principles of Corporate Finance, McGraw-Hill Book
Company, New York, NY, 2nd Edition, pp. 195-226, 1984.
5. Davidson, Sidney, Financial Accounting: An Introduction to Concepts,
Methods, and Uses, Third Edition, Dryden Press, Chicago, IL, pp. 202-229,
1982.
6. Dun & Bradstreet, Industry Profile, Oct. 17, 1988.
7. Foster, G., Financial Statement Analysis, Prentice-Hall, Englewood Cliffs,
NJ, Chapters 2, 14, 1978.
8. Hartley, W.C.F. and Meltzer, We L., Cash Management, Planning, Fore-
casting, and Control, Prentice Hall, Englewood Cliffs, NJ, pp. 90-123.
9. Industrial Economics, Inc. (IEC), Superfund Financial Assessment System,
Technical Support Document, U.S. EPA/Office of Policy and Resource
Management, Washington, DC, Chapter 4, 1982.
10. International Research & Technology (IR & T), Financial Tests as an Op-
tion for Demonstrating Financial Responsibility, U.S. EPA/Office of Solid
Waste, Washington, DC, 2, pp. 4-18, 1980.
11. Johnson, M., Mason, R., "Structured Settlements: A New Settlement In-
centive," Proceed, of the 9th National Conference on Hazardous Waste and
Hazardous Materials, Washington D.C., HMCRI, Silver Spring, MD, pp.
23-29, 1988.
12. Mays, R.H., "Revised Hazardous Waste Bankruptcy Guidance," U.S.
EPA/Office of Enforcement and Compliance Monitoring, Washington, DC,
May 23, 1986.
13. Morris, R., 1988 Annual Statement Studies, Robert Morris Associates,
Philadelphia, PA, p. 135, 1988.
14. Putnam, Hayes and Bartlett (PHB), Financial Screening Criteria for Privately
Held Firms, U.S. EPA/Office of Policy Analysis, Washington, DC, pp. 19-21,
1984.
15. Schnepper, J.A., The New Bankruptcy Law, A Professional's Handbook,
Addison-Wesley Publishing Company, Philippines, 2nd Ed, pp. 67-100, 1982.
16. Siegel, J.G., How to Analyze Businesses, Financial Statements and the Quality
of Earnings, Prentice Hall, Englewood Cliffs, NJ, pp. 93-168, 1987.
17. Tyran, M.R., Handbook of Business and Financial Ratios, Prentice Hall,
Englewood Cliffs, NJ, pp. 145-162, 1986.
18. Troy, L., Almanac of Business and Industrial Financial Ratios, Prentice Hall,
Englewood Cliffs, NJ, pp. 122-123, 1989.
19. The U.S. EPA, Office of Waste Programs Enforcement, Technical Enforce-
ment Support Potentially Responsible Party Search Orientation Manual,
Washington, DC, pp. 20-24 and C-l to C-2, July, 1989.
20. Vichas, R.P., Handbook of Financial Mathematics, Formulas, and Tables,
13th Edition, Prentice-Hall, Inc., Englewood Cliffs, NJ, pp. 341-372, 1979.
21. Vickman, T.M., Handbook of Model Accounting Reports and formats, Pren-
tice Hall, Englewood Cliffs, NJ, pp. 4-16 and 241-245, 1987.
STATE PROGRAMS 605
-------
Development of a Projection Model for
State Hazardous Waste Disposal Capacity Assurance
Diane Sharrow
U.S. EPA
Region 5
Mark Johnson, MBA, ARM
Nancy Willis, MAS
Butch Fries, MA
PRC Environmental Management, Inc.
McLean, Virginia
ABSTRACT
Section HM(c)(9) of CERCLA as amended requires states to demon-
strate adequate capacity for the destruction, treatment or secure dis-
posal of all hazardous wastes that are reasonably expected to be generated
within the state during the next 20 yr. States that fail to provide an ade-
quate capacity assurance plan (CAP) face loss of Superfund monies
for remedial actions. To demonstrate capacity, states first calculated
hazardous waste generation and disposal capacity for a "base year"
and then projected generation over a 20-yr period. The projection con-
sidered the impact of economic growth or contraction, waste minimi-
zation and new regulations on the state's waste generation rates and
disposal capacity. Few states, however, had comprehensive data avail-
able or the technical resources necessary to complete the CAP.
Through its Alternative Remedial Contracting Strategy contract.
Region S of the U.S. EPA commissioned Planning Research Corpora-
tion (PRC) Environmental Management, Inc., to assist states in the CAP
projection. This paper discusses the approach PRC conceived to devise
reliable methods for calculating the impact of economic change, waste
minimization and new regulations despite the absence of comprehensive
data and discusses future research needs and problems encountered in
completing the CAP projection.
INTRODUCTION
CERCLA Section 104(c)(9) requires each state to assure by Oct. 17,
1989, that adequate RCRA Subtitle C hazardous waste management
capacity will be available for waste generated within the state during
the next 20 yr. The legislative history for the 1986 CERCLA amend-
ments indicates that Congress enacted Section I04(c)(9) because politi-
cal pressures or public opposition prevent some states from siting or
issuing permits for adequate hazardous waste management capacity
within their boundaries, perhaps leading to the creation of future Super-
fund sites.1 The perceived inaction over hazardous waste siting and
permitting prompted Congress to require the state capacity assurance
plans (CAPs) and to make future Superfund assistance for remedial
actions contingent on the capacity assurance.
In December, 1988, guidance,2 U.S. EPA specified that states not
able to demonstrate capacity could provide the required assurance by:
• Demonstrating the intent to site new facilities
• Describing or implementing, and demonstrating the effectiveness of,
a waste minimization program
• Assuring access to facilities in other states through interstate or region-
al agreements
To complete the CAP, states first calculated capacity and generation
for a "base year"—usually 1987 — and then projected hazardous waste
generation for 1989, 1995 and 2009, finally comparing projected waste
generation to current capacity to determine where surpluses or short-
falls exist. For the base year and projection calculations, the more than
700 RCRA waste categories and the full array of available waste manage-
ment techniques were compressed into 17 SARA waste types and 15
SARA management categories. The U.S. EPA's guidance asked stales
to modify the projections by considering expected economic growth
or contraction, waste minimization activities and the impact of new
federal and state regulations.
U.S. EPA Region 5 provided guidance and coordination among the
Region 5 states and between the stales and EPA Headquarters through-
out the CAP process. Region 5 also commissioned PRC to support die
states in methodology development and data manipulation and to pro-
vide a consistent approach to data manipulation methodologies.
To assist the states, PRC first developed an economic projection matrix
that, in sum, calculated current waste volume, updated the figures to
the projection year based on the factors in the U.S. EPA's guidance and
measured projected demand for management capacity against availa-
ble management capacity. Second, although several U.S. EB\ Region 5
states intended to rely on waste minimization to assure capacity, none
maintained waste minimization records that allowed measurement of
the reduction in waste generation achieved. Therefore, to estimate waste
minimization potential, PRC developed a strategy for modifying U.S.
EPA research according to factors specific to each state and derived
probable reduction coefficients Third. PRC's research targeted pending
regulations—in particular the RCRA "land ban" and newly listed RCRA
wastes—likely to affect future hazardous waste management. The
research was translated into coefficients that adjusted the capacity projec-
tion to account for the impact of future regulations. (Although the projec-
tion assessed future rates of waste generation rather than expected
capacity, the undertaking was known commonly as the "capacity projec-
tion.") Finally, PRC produced computer software and an accompanying
users manual and provided them to the states to implement the overall
projection methodology. Subsequent sections discuss each component
of the methodology.
ESTIMATING DEMAND FOR WASTE
MANAGEMENT CAPACITY
As an aid in state CAP development, the U.S. ER\ Office of Solid
Waste developed software that calculated hazardous waste management
capacity information for each state for the 1987 base year. The SARA
Analytic Software (SARA-ASW) accepts data only in the format deter-
mined by the U.S. EPA Biennial Report Data System (BIRDS) derived
from the U.S. EPA's RCRA biennial report forms. Depending on the
amount of information available—whether comprehensive or limited—
the SARA-ASW program applies varying levels of processing. Few stales
have adopted the biennial report form that is the basis for BIRDS, so
PRC was first required in most cases to convert state data to the BIRDS
format.
606 STATE PROGRAMS
-------
Predicting a state's future hazardous waste management capacity
requirements involves comparing probable future generation to the
current demand for, or utilization of, existing capacity. The general
approach to projecting hazardous waste generation within the state con-
sists of four features and assesses recurring and nonrecurring wastes
(for example, wastes generated from spills or CERCLA remedial
actions) separately:
• An initial estimate of future waste generation is developed based on
the projected growth or decline in industrial activity for recurring
wastes, on planned remedial actions for nonrecurring wastes and on
wastes affected by new regulations.
• The estimates are then adjusted on the basis of the projected changes
in waste generation per unit of activity (any constant measure of
industrial output). This approach estimates change in the generation
of specific wastes by specific industries or in remedial actions on
the basis of information on waste minimization programs and new
regulations.
• Projected waste generation is aggregated across all industry groups
and remedial activities and divided into the 17 SARA waste types
specified in the U.S. EPA guidance. (Examples of the SARA waste
types include halogenated solvents, inorganic liquids with metals and
non-halogenated organic liquids.)
• The total projected waste volumes are classified into demand for the
15 SARA waste management categories specified in the U.S. EPA
guidance. (The SARA management categories include metals and
solvents recovery, incineration, aqueous treatments and stabilization.)
The output of the projection is expressed as demand for the different
types of waste management techniques. The demand estimates can be
compared to existing waste management capacity to determine the state's
future ability to handle instate waste production. Information on capacity
deficits and surpluses also enabled judgments about the manageable
level of interstate shipments of hazardous waste in the future for states
that intended to rely on interstate agreements as a component of capac-
ity assurance.
ECONOMIC PERFORMANCE
Economic performance must be incorporated when projecting a state's
future recurring waste generation because, as a state's economy expands
or contracts, waste generation is likely to react accordingly. For the
CAP analysis, the most reliable measures of industrial activity are those
that correlate most directly with waste generation, such as production
in terms of physical unit or value of production. Value of shipments
is a reliable indicator, but may include shipments out of inventories
as well as current output and so is less effective as a measure of indus-
trial activity. Other variables such as value added and employment can
be used, but their relationships to production, and therefore to waste
generation, may change over time on the basis of other factors such
as labor productivity. If employment is used to measure industrial ac-
tivity, production employment is preferable to total employment.
Many state agencies forecast industrial economic activity using
methods such as regional input-output models that translate projected
state demand for such elements as consumption, investment and govern-
ment spending into industry production and employment estimates. In
some cases, however, no state industry projection data were available
beyond 5 yr. In those cases, 20-yr projections specific to each state,
by industry group, were drawn from sources published by the U.S.
Department of Commerce. PRC's flexible projection methodology al-
lowed the states to rely on default values or to input economic data on
a facility-specific basis.
For the default values, PRC selected employment projections by state
at the two-digit Standard Industrial Classification (SIC) code level,
prepared by the Regional Economic Analysis Division of the Depart-
ment of Commerce Bureau of Economic Analysis. The projections,
named OBERS for the offices formerly in charge of their preparation,
use a "step-down" approach. The step-down approach is based on the
premise that data for larger aggregates are generally more accurate than
the same type of information for more detailed classifications. Thus,
the OBERS program first develops national projections and then dis-
tributes them among states to arrive at state-level projections.
WASTE MINIMIZATION
The U.S. EPA guidance asked states to consider the potential for waste
minimization in forecasting recurring waste generation. For purposes
of the CAP projection, waste minimization was defined as the reduc-
tion of hazardous waste that is generated or subsequently requires treat-
ment or disposal.
Most information produced so far on waste minimization potential
is anecdotal only. The evaluation of the potential for waste minimiza-
tion in die U.S. EPA Region 5 states, therefore, was based on review
and analysis of the information the U.S. EPA developed for the 1986
Report to Congress on the Minimization of Hazardous Hbstes. No origi-
nal research was conducted, although the general minimization factors
developed as default values for the projection methodology were modi-
fied when information was available specific to the state. The generic
information has limitations when applied to a specific facility, but the
analysis represented a systematic approach to considering the potential
for waste minimization in various industrial categories examined as a
whole.
The analysis first deleted the data on the considerable quantities of
contaminated rinse water. Then, for those industries identified as major
generators of hazardous waste, an industry-specific waste reduction
potential was extrapolated from the U.S. EPA Report to Congress and
included in the software as a default value. This default value may be
applied to all wastes the industry generates if no information on the
nature of the waste by waste code is available, or it may be used for
waste codes without waste-code-specific reduction factors.
Potential reductions were classified and adjustment factors derived
according to two possible scenarios. Under the "most likely scenario,"
a particular type of industry would adopt a broad range of minimiza-
tion techniques, resulting in a moderate reduction in the quantity of
waste produced. Potential reduction factors in the most likely scenario
generally ranged from 5% to approximately 20%. Under the "most
optimistic scenario," the industry as a whole would adopt the most
effective minimization techniques available, resulting in substantial
reductions in waste generation. Potential reduction factors in the most
optimistic scenario ranged from 18% to 50%; the most optimistic reduc-
tions would, however, probably not be achieved without a tough and
aggressively enforced state program.
Default reduction factors were derived by two-digit SIC code where
information was available in that form and were obtained by analyzing
the industrial processes examined for the U.S. EPA Report to Congress.
If the process generated a listed hazardous waste,.the waste reduction
potential for that process was averaged with the potential from other
processes within the industrial category generating that waste. For the
RCRA wastes not generated by a specific process (the "F" wastes),
the methodology resulted in an aggregation of reduction potentials
drawing on one to seven processes. No aggregation was required for
the RCRA "K" wastes because the category refers to hazardous wastes
generated during specific industrial processes.
IMPACT OF REGULATORY CHANGES
Regulatory changes may alter future waste generation and manage-
ment in a variety of ways. New hazardous waste regulations may ex-
pand the universe of regulated hazardous waste or force a shift to new
waste management techniques—for instance, moving wastes from land-
filling to treatment by incineration. PRC used a two-step process to
account for the effects of regulatory changes on capacity assurance,
leading to development of a coefficients matrix to modify the CAP
projection. First, the impact of regulatory changes on the types and
quantities of hazardous waste generated was assessed. Second, the
impact of regulatory changes on hazardous waste management options
was evaluated. The assessment of the impact of new regulations for
the current CAP analysis relied on the following assumptions:
• Newly listed wastes will be the primary regulatory change affecting
the types and quantities of wastes generated.
• The RCRA land disposal restrictions, commonly known as the "land
ban," are the primary regulations affecting hazardous waste manage-
ment options.
STATE PROGRAMS 607
-------
• The analysis need examine only federal regulations because no state-
specific new or proposed laws or regulations could be identified in
EPA Region 5 that would affect hazardous waste management
capacity.
• Given the uncertainties associated with proposed or anticipated
rulemakings. only federal regulations considered final as of July I,
1989. were considered. The statutorily required land ban, whose
effects are both substantial and predictable, is an exception to the
general approach.
Effects of Newly Listed Wastes
A process that includes use of a computer program was developed
to project the impact of newly listed wastes on hazardous waste dis-
posal capacity. First, the newly listed wastes that must be considered
were determined: currently, the only wastes that are the subject of a
final rulemaking are the six mining wastes listed Sept. 13. 1988 '
Each state in U.S. EPA Region 5 was asked to compare a list of industry
types by SIC code that generate the newly listed wastes against a list
of the industries active in each state. The comparison determined
whether the mining wastes were likely to be generated within (he state.
The state determined responsible industries (or. when possible, specific
facilities) and estimated the quantity of each newly listed waste cur-
rently generated. The information was incorporated into the base year
data, and the program completed a projection of future generation iden-
tical to that for other wastes. The best management technique was iden-
tified for each newly listed waste, and the waste volume was apportioned
among the suitable SARA waste management categories.
Effects of Land Ban
Future management of many wastes will be significantly affected
by the land ban because it prohibits the land disposal of many untreated
hazardous wastes and requires their management by alternative treat-
ment technologies. The land ban is being implemented in three phases,
each described as a "Third" of the total, that began in August, 1988
and will end in May, 1990. Because of certain provisions of RCRA,
the phases do not always occur in discrete segments.
The best demonstrated available technology (BOAT) was identified
for each waste affected by the land ban. In cases where the waste was
generated in different forms — for example, wastewater versus non-
wastewater—more than one BOAT was identified for each waste and
a portion was assigned to each BOAT. Because BDATs have yet to
be proposed for most Third wastes, the BOAT for the most closely
analogous First Third and Second Third wastes was assigned.
Some management techniques produce residual wastes that require
additional waste management. BDATs were identified for these residual
wastes, and factors were developed to account for the increased waste
management capacity they require. For example, incineration produces
a residual waste, ash, that requires stabilization before landfilling so
that total capacity would equal the original volume of waste, plus the
residual volume, plus the volume of the treated residual. These fac-
tors were developed based on the regulatory development documenta-
tion for the land ban and on best engineering judgment.
Next, the SARA waste management category or categories corres-
ponding to the BOAT for each affected waste were identified. The to-
taJ volume of waste requiring alternative treatment (including residual
volumes) was assigned to waste management categories, and total
demand for each management technique in each projection year was
calculated. As a final step, the waste generation data, identified by
RCRA waste code, were aggregated into the 17 SARA waste types.
PROBLEMS, FUTURE DATA NEEDS, AND RESEARCH
Several problems and areas for future research became apparent
during completion of the CAP. The lack of data on waste minimization
already has been mentioned. In fact, the overall questionable quality
of the data and inconsistencies among states presented a major problem
throughout completion of the CAP. In addition, the interstate agree-
ment required by the U.S. EPA's guidance raised legal questions about
the prohibition under the commerce clause of the U.S. Constitution
of state control of the interstate movement of ha/.ardous wastes. Most
states attempted to reach some form of regional agreement, although
the process was at best imperfect. Other research areas are briefly dis-
cussed below.
The U.S. EPA's guidance required states to include in the capacity
analysis wastes considered "exempt" under RCRA, such as recycling
of hazardous wastes during continuous industrial processes, discharges
to facilities permitted under the Clean Water Act national pollutant
discharge elimination system (NPDES) program and discharges to pub-
licly owned treatment works. The exempt wastes are by definition ex-
cluded from the reporting requirements of RCRA, complicating (he
states' attempt to produce an acceptable CAP. Even where limited
information was available on the exempt wastes, it was frequently
incomplete or difficult to adapt to the CAP. For instance, records on
NPDES flows show only pollutant levels in total gallons discharged
and do not readily enable regulatory agencies to extrapolate influent
levels of hazardous wastes. Furthermore, no clear statutory authority
enables the stales to collect the data on exempt wastes, and the effect
of the exempt wastes on RCRA Subtitle C hazardous waste manage-
ment facilities has not been demonstrated.
In addition, few state reporting systems could directly relate waste
generation to industry types except in general terms. Enhanced stale
data collection thai would accumulate data on a more detailed level,
for example, at the four) rather than two-digit SIC code level, would
aid the accuracy of future capacity plans.
CONCLUSION
The capacity assurance plan required development of a method to
project future hazardous waste generation according to factors such
as economic performance, waste minimization potential and impact
of future regulations. The states desired consistent projection metho-
dology and data manipulation. This paper reviewed one approach to
preparing the CAP.
DISCLAIMER
The views expressed in this paper are those of the authors, and do
not necessarily reflect the views of the U.S. Environmental Protection
Agency.
REFERENCES
I. Senate Report No. II. 1985. 99th Congress. 1st Session, p.22.23.
2. U.S. EPA/Office of Solid Waste and Emergency Response, Assurance of
Hazardous Waste Capacity: Guidance to Slate Officials. Assistance in Ful-
filling the Requirements o/CEKCLA 104
-------
Implementation of Environmental Evaluation Policy
In the Superfund Program
Michael J. Dover, Ph.D.
The Cadmus Group, Inc.
Waltham, Massachusetts
Patricia A. Mundy
Office of Emergency and Remedial Response
U.S. EPA
Washington, D.C.
John Bascietto
U.S. Department of Energy
Washington, D.C.
ABSTRACT
In late 1988, the Directors of the U.S. EPA's Office of Emergency
and Remedial Response and Office of Waste Programs Enforcement
expressed concern "that thorough and consistent environmental evalua-
tions are not always being performed at Superfund sites in both the
removal and remedial programs." Pointing out that "the law calls for
protection of human health and the environment," they asked the U.S.
EPA Regional Offices to take steps to address this issue in all current
and future Remedial Investigations and Feasibility Studies.
As part of the implementation of this policy, the U.S. EPA has issued
Risk Assessment Guidance for Superfund—Environmental Evaluation
Manual. The primary audience for this manual is U.S. EPA Regional
Office Staff—the On-Scene Coordinator (OSC) for removal actions and
the Remedial Project Manager (RPM) for remedial investigations and
cleanup actions. The purpose of the manual is to provide a scientific
and conceptual framework for overseeing environmental evaluations at
Superfund sites. A key ingredient in the implementation of environ-
mental evaluation at all relevant sites is the establishment of Biological
Technical Assistance Groups (BTAGs) in the U.S. EPA Regional Offices.
The manual is designed to facilitate communication between the RPM
or OSC and the BTAG by describing:
• The statutory and regulatory basis of environmental evaluation in
the Superfund program
• Basic scientific concepts relating to environmental evaluation of
hazardous waste sites
• The role of the BTAGs and the information needed by specialists
serving on the BTAGs
• Steps and information needed in planning an environmental evaluation
• An outline to guide the organization and presentation of the environ-
mental evaluation in Superfund reports
The paper describes the respective roles of contractors, U.S. EPA
staff and BTAGs in the planning, implementation and review of environ-
mental evaluations. Progress in implementing the policy in the Super-
fund program also is discussed.
INTRODUCTION
From its inception, the Superfund program has relied on risk assess-
ment to determine both the need for remediation and the level of
remediation at uncontrolled hazardous waste sites. Until recently, formal
Superfund risk assessments have focused almost exclusively on threats
to public health. Beginning in 1988, the Superfund program has been
increasing its concern for assessing and controlling environmental
hazards as an integral part of the remedial process.
Environmental evaluation (ecological assessment), as .it is applied
in the Superfund program, is defined as the qualitative and/or quanti-
tative appraisal of the actual or potential effects of a hazardous waste
site on plants and animals other than people and domesticated species.
The program recognizes, however, that the health of people and domesti-
cated species is inextricably linked to the quality of the environment
shared with other species. Information from ecological studies may point
to new or unexpected exposure pathways for human populations, and
health assessments may help to identify environmental threats.
Ecological assessment of hazardous waste sites is an essential element
in determining overall risk and providing protection of public health,
welfare and the environment. The U.S. EPA considers ecological factors
in hazard assessment and in reviewing alternative remedial actions
because:
• Through the authority found in the Superfund legislation and other
statutes, the U.S. EPA seeks to protect wildlife, fisheries, endangered
and threatened species and valued habitats.
• From a scientific viewpoint, the U.S. EPA needs to examine ecological
effects and routes of exposure so that: (1) important impacts and trans-
port pathways are not overlooked, and (2) reasonable estimates are
made of health and environmental effects.
CERCLA, as amended by SARA in 1986, requires the U.S. EPA to
ensure the protection of the environment in: (1) selection of remedial
alternatives and (2) assessment of the degree of cleanup necessary.
Several sections of CERCLA make reference to protection of health
and the environment as parts of a whole: Section 105(a)(2) calls for
methods to evaluate and remedy "any releases or threats of
releases.. .which pose substantial danger to the public health or the
environment;" Section 121(b)(l) requires selection of remedial actions
that are "protective of human health and the environment;" Section
121(c) calls for "assurance that human health and the environment con-
tinue to be protected;" and Section 121(d) directs the U.S. EPA to at-
tain a degree of cleanup "which assures protection of human health
and the environment."
CERCLA Section 104(b)(2) calls upon the U.S. EPA to promptly
notify the appropriate Federal and State natural resource trustees about
potential dangers to protected resources. Table 1 provides a partial listing
of natural resource trustees. It is important for Federal trustees to identify
and notify all cognizant trustees because co-trustees, such as States,
Indian tribes or other Federal agencies, may have overlapping or primary
jurisdiction over natural resources potentially affected by releases from
any Federal facility.
Section 122(j) of the amended CERCLA requires the Agency to notify
the Federal natural resource trustees of any negotiations regarding the
release of hazardous substances that may have resulted in natural
resource damage. Section 122(j)(l) also calls upon the U.S. EPA to
encourage Federal natural resource trustees to participate in negotia-
tions with potentially responsible parties (PRPs). If the U.S. EPA seeks
to settle with a PRP by signing a covenant not to sue, the Federal natural
STATE PROGRAMS 609
-------
Table 1
Natural Resource Trusteeships
Departnent of the interior
Fish and Mildllfe Service
Migratory birds
Anadroaous fish
Endangered/threatened species
Critical habitati
National wildlife refuges
National fish hatcheries
National Park Service (Including)
National parka
National seashores
National recreation areaa
National historic sites
National battlefields
National scenic and recreational clv«rs
Department of Coaaerce
Coastal environments and habitats
Habitats (rivers and tributaries) of anadroaous and
catadroBoua fish
Endangered/threatened species
Tidal wetlanda
Marine sanctuaries
COMerclal and recreational euirine fishery resouroet
Department of Agriculture
National forests
Pepartawit of Defense
Department of Defense installations
Department of Energy
Department of Energy inatallatlons
Statea
The responsibility of the states are atate specific to be
decided by each state Governor. Similarly, the Oovernor also
designates the appropriate atate agency to act as the. Tnutee.
Contact should be ude with the respective state envlronwntal
department or attorney general's office for the intonation
regarding trustee designations and responsibilities.
Indian Tribes
Tribal lands
resource trustee must agree to this covenant in writing. Section 122(j)(2)
states that:
"The Federal natural resource trustee may agree to such a
covenant if the potentially responsible party agrees to under-
take appropriate actions necessary to protect and restore the
natural resources damaged by such release or threatened
release of hazardous substances."
In December, 1988, the Directors of the U.S. EPA's Office of
Emergency and Remedial Response (OERR) and Office of Waste
Programs Enforcement (OWPE) expressed concern "that thorough and
consistent environmental evaluations are not always being performed
at Superfund sites in both the removal and remedial programs." Pointing
out that "the law calls for protection of human health and the environ-
ment," they asked the U.S. EPA Regional Offices to take steps to address
this issue in all current and future Remedial Investigations and Feasi-
bility Studies.
Implementation of U.S. EPA policy regarding environmental evalua-
tion is proceeding through five initiatives: publication of a manual for
U.S. EPA site managers; formation of technical assistance groups in
the Regional Offices; development of additional information resources;
production of training materials for U.S. EPA Regional staff; and
communication of U.S. EPA policy to remedial contractors.
ENVIRONMENTAL EVALUATION MANUAL
In March, 1989, the U.S. EPA completed its manual on environmental
evaluation1. The primary audience for this manual is U.S. EPA
Regional Office Staff—the On-Scene Coordinator (OSC) for removal
actions and the Remedial Project Manager (RPM) for remedial inves-
tigations and cleanup actions. The purpose of the manual is to provide
a scientific and conceptual framework for overseeing environmental
evaluations at Superfund sites.
The approach taken in the Environmental Evaluation Manual differs
significantly from its companion volume on human health evaluation1
Whereas the Human Health Evaluation Manual contains considerable
technical detail on how to conduct a risk assessment with respect to
health threats, the Environmental Evaluation Manual focuses on de-
veloping a general understanding of the concepts and strategy of eco-
logical assessment. This difference stems from the key task of defining
the scope of ecological investigations, which is inherently more com-
plex than scoping human health evaluations. Due to the wide array of
possible habitats, species and effects that may be involved, responsi-
bility for planning and interpreting environmental evaluations needs to
be snared with technical specialists who understand both the questions
that need to be asked and the most efficient means of answering those
questions. Hence, the Environmental Evaluation Manual is designed
primarily as a means for facilitating communication between the OSC
or RPM and these specialists.
The Environmental Evaluation Manual contains six chapters. Chapter
I. the introduction, defines environmental evaluation and its role in the
Superfund program. The chapter also briefly discusses the relation-
ship between environmental and human health evaluation.
Chapter 2 discusses the statutory and regulatory basis for environ-
mental evaluation, including citations of the amended CERCLA and
the proposed revisions to the NCP. This chapter also describes relevant
sections of the U.S. EPA's guidances for removal actions and RJ/FSs.
Finally, the chapter lists numerous Federal laws that may contain
applicable or relevant and appropriate requirements (ARARs). In
addition to such commonly applied statutes as RCRA and the Clean
Water Act, this section of the manual discusses such laws as the Fish
and Wildlife Coordination Act. the Endangered Species Act and the
Marine Mammal Protection Act.
Chapter 3 of the manual describes the basic scientific concepts
underlying ecological assessment. It is intended to assist the RPM or
OSC in working with the ecologisis who will provide technical advice
or perform the studies, by describing the conceptual framework within
which these specialists make their judgments. This chapter defines
numerous terms that are used later in the manual.
Chapter 4 details the role of technical specialists in ecological
assessment. It discusses the kinds of information that the RPM or OSC
should make available to these specialists so that a suitable characteri-
zation of the site and its contaminants can be made. This information
is likely to include data on the site's location, the site's history,
contaminants of concern and the site's environmental setting. The chapter
goes on to discuss the assistance that technical specialists can provide
in site screening, identification of information gaps, advice on Work
Plans, data review and interpretation, advice on remedial alternatives
and enforcement support.
Chapter 5 discusses the process of developing an appropriate study
design for assessment of a site. It discusses the principal components
of defining the scope and design of an environmental evaluation:
• Determination of the objectives and level of effort appropriate to the
site and its contaminants
• Evaluation of site characteristics
• Evaluation of the contaminants of concern
• Identification of exposure pathways
• Selection of assessment endpoints
The outcome of the planning process is the Sampling and Analysis
Plan, which specifies the methods for data collection and analysis and
the procedures for QA/QC. If new data are to be collected for the en-
vironmental evaluation, it is essential that data quality objectives reflect
specific programmatic goals and management objectives to ensure that
time and funds spent to gather and analyze data are used efficiently
and effectively.
Chapter 6 describes a basic outline for an assessment. Although each
site's assessment will differ according to the details of the contaminants,
exposure routes, potentially affected habitats and species, this chapter
provides a checklist of items for the RPM or OSC to expect when over-
seeing the preparation of an assessment, including:
• Specifying the objectives of the assessment
• Defining the scope of the investigation
• Describing the site and study area
610 STATE PROGRAMS
-------
• Describing the contaminants of concern
• Characterizing exposure
• Characterizing risk or threat
• Describing the derivation of remediation criteria or other uses of quan-
titative risk information
• Describing the study's conclusions and any limitations of the analysis
The Environmental Evaluation Manual intentionally avoids detailed
discussion of field and laboratory methods, concentrating instead on
basic concepts, design strategies and information resources. Its intent
is to foster communication between RPMs/OSCs and the Biological
Technical Assistance Groups that have been established to provide the
ecological expertise necessary for effective design, execution and in-
terpretation of environmental evaluations.
BIOLOGICAL TECHNICAL ASSISTANCE GROUP
In the December, 1988 memorandum, the Directors of OERR and
OWPE pointed to the successful use of biological technical assistance
groups (BTAGs) in U.S. EPA Regions 2, 3, 4 and 10. In March, 1989,
a workshop for Regional Superfund managers was held in Cherry Hill,
New Jersey, sponsored by Region 3, which focused on the activities
of BTAGs. As a result of these efforts, all U.S. EPA Regional Offices
have now established BTAGs. Membership in the BTAG varies from
Region to Region, but may include staff from:
• U.S. EPA Regional Environmental Services Divisions
• The U.S. EPA Environmental Response Team
• U.S. EPA Regional NU.S. EPA coordinators
• Ecosystem-specific U.S. EPA programs, such as the Great Lakes
National Program Office in Chicago, Illinois or the Chesapeake Bay
Program Office in Annapolis, Maryland
• Laboratories of U.S. EPA's Office of Research and Development
• Regional and field offices of the U.S. Fish and Wildlife Service and
the National Oceanic and Atmospheric Administration (especially
NOAA's Coastal Resource Coordinators)
• Other Federal and State environmental and resource-management
agencies (such as State fish and game departments)
Generally, specialists on the BTAG serve an advisory role. Their
function is to assist the RPM or OSC with information collection and
evaluation and to help ensure that ecological effects are properly
considered in investigations and decisions. In specific cases, arrange-
ments may be made for individual BTAG members to be involved directly
in conducting the work.
BTAGs are expected to be consulted at all appropriate stages of the
remedial process, from the Preliminary Assessment and Site Investi-
gation to the review of Remedial Designs and Remedial Actions. Perhaps
the most frequent and most important use of the BTAGs occurs during
the RI/FS process, including site screening, review of Work Plans and
review of data.
Following collection of existing data, the BTAG members should be
in a position to determine the nature and extent of ecological assess-
ment that will be necessary for the site. If no ecological exposure path-
ways have been revealed in this initial review, little or no additional
work may be needed. Alternatively, certain exposure pathways might
be eliminated from further study while others might require more data.
For instance, if there is no surface water on the site and no opportunity
for contaminants to reach surface waters off the site, further data on
aquatic effects would very likely be pointless, even though concern about
exposure to terrestrial organisms might warrant extensive sampling and
testing.
Effective ecological assessment will require a design that is tailored
to each site's specific characteristics and the specific concerns to be
addressed. Choosing which of the many possible variables to inves-
tigate in the study will depend on the nature of the site, the types of
habitats present and the objectives of the study. The BTAG is expected
to assist the RPM in specifying technical objectives for the investiga-
tion. Such objectives might include:
• Determination of the extent or likelihood of impact
• Interim mitigation strategies and tactics
• Development of remedies
• Remediation criteria
The BTAG can then help the RPM develop data quality objectives
to support these technical objectives.
Although each assessment is in some way unique, it is possible to
outline the general types of data that may be required. For terrestrial
habitats, the BTAG specialists may specify such data needs as:
• Survey information on soil types, vegetation cover, and resident and
migratory wildlife
• Chemical analyses to be conducted in addition to any previous work
done as part of a Preliminary Assessment or Site Investigation
• Site-specific toxicity tests to be conducted
For fresh-water and marine habitats, the information needed will most
likely include:
• Survey data on kinds, distribution and abundance of populations of
plants (phytoplankton, algae and higher plant forms) and animals
(fish, macro- and micro-invertebrates) living in the water column
and in or on the bottom
• Chemical analyses of samples of water, sediments, leachates and bio-
logical tissue
• Sediment composition and quality, grain sizes and total organic carbon
• Toxicity tests designed to detect and measure the effects of contami-
nated environmental media on indicator species, or on a representa-
tive sample of species
BTAG members will also provide guidance on such QA/QC issues as:
The area to be covered in biotic and chemical sampling programs
The number and distribution of samples and replicates to be drawn
from each habitat
The preferred biological analysis techniques to be used
Adherence to the assumptions of predictive models used in the analysis
The physical and chemical measurements (e.g., dissolved oxygen in
a water sample, pH of water or soil, ambient temperature) to be taken
at the time of the survey
• Any special handling, preservation methods or other precautions to
be applied to the samples
The BTAG also may be called upon to review data and provide com-
ments on the interpretation of data. In most situations, extensive and
long-term ecological studies are unlikely to be undertaken, and informed
professional judgment will be required to determine if the weight of
evidence supports a particular decision regarding the site.
OTHER ACTIVITIES
Publication of the Environmental Evaluation Manual and establish-
ment of BTAGs constitute the core of the U.S. EPA's implementation
of its environmental evaluation policy. The Superfund program is also
providing additional information to support environmental evaluation,
revising its training program for RPMs and Regional risk assessors,
and communicating its policy to remedial contractors.
While OERR and OWPE were developing the Environmental Evalua-
tion Manual, the U.S. EPA's Corvallis Environmental Research Labora-
tory sponsored the preparation and publication of a companion
volume3 containing detailed discussion of field and laboratory methods
for ecological assessment of hazardous waste sites. This reference
document covers such topics as:
• Types of ecological endpoints, criteria for selecting endpoints and
defining assessment goals
• Assessment strategies and designs, and selection of appropriate
assessment methods
Field sampling design
Quality assurance and data quality objectives
Aquatic, terrestrial and microbial toxicity tests
Use of biomarkers
Field assessments of aquatic ecosystems, terrestrial vegetation, terres-
trial vertebrates and terrestrial invertebrates
Data analysis and interpretation
STATE PROGRAMS 611
-------
The U.S. EPA's Office of Information Resource Management is
developing a data base for use by Regional and Headquarters staff which
will list ecological expertise found throughout the Agency. These experts
will include ecologists in other U.S. EPA program offices, scientists
in the Office of Research and Development at Headquarters and in
research laboratories, and specialists in the Regional Environmental
Services Divisions and ecosystem-specific programs. This data base
will be made available in electronic form to Regional Offices, allowing
for rapid updates and ready recovery of the information contained in
the directory.
The Superfund program currently is revising its training program for
RPMs and Regional risk assessors to include material on environmental
evaluation along with the updated approach to human health evalua-
tion. Among the topics to be covered in the training arc the following:
• Statutory and regulatory basis for health and environmental evalu-
ations
• Ecological principles and concepts relevant 10 environmental
evaluation
• The role of the BTAG
• Sampling and analysis for health and environmental evaluations
• Planning and evaluation of site assessments
The courses will include detailed examination of real and hypothetical
case studies. For environmental evaluation, emphasis will be placed
on directing and reviewing contractor products, rather than on con-
ducting the studies themselves.
In April, 1989, Superfund ARCS contractors were invited to the US.
EPA for a day-long workshop on Superfund policy and procedures. At
that time, the Chief of the Toxics Integration Branch for Superfund
presented information on the revised Human Health Evaluation Manual
and the Environmental Evaluation Manual. Contractors also were in-
formed of the establishment of BTAGs in the Regional Offices and were
encouraged to work with these specialists, through the site-specific
RPM, to develop environmental evaluations that are both scientifically
sound and capable of being conducted within program-mandated time
and budget constraints.
Ecological assessment is, and will continue to be, a process combining
careful observation, data collection, testing and professional judgment.
Through close coordination with the RPM and BTAG, and by following
U.S. EPA's guidance manuals and other reference materials, Superfund
contractors should be able to conduct site assessments that will result
in effective and efficient protection of environmental as well as human
receptors.
REFERENCES
I U.S. F.PA, Risk Assessment Guidance for Superfund. Volume II: Emimn-
menial Evaluation Manual. Interim Final. US EPA/540/l-89/001. US. B*.
Cincinnati. OH. March 1989
2 U.S. EPA. Risk Assessment Guidance for Superfund. Volume I: Human HeaUi
Evaluation Manual Pan A, (Review Drift) OSWER Directive 928S.70IA.
July, 1989
3. US. EPA. Ecological Assessments of Hazardous Htisie Sites: A Field and
Laboratory Referenrr. US EPA/600/3-89/013. U.S. EPA. Cincinnati, OH.
Mar.. 1989.
612 STATE PROGRAMS
-------
Review and Comparison of State Statutes for
Natural Resource Injury
Lloyd Landreth, J.D.
PRC Environmental Management, Inc.
Denver, Colorado
ABSTRACT
When a natural resource is injured due to release of a hazardous sub-
stance, restoration of that natural resource to pre-release levels is
generally very expensive. CERCLA provided for use of the Hazardous
Substance Response Fund to redress injury to natural resources, but
SARA has taken away the use of the fund for this purpose. While the
United States government is undoubtedly interested in the preservation
of natural resources, it is the state wherein injury to a natural resource
occurs that stands the greatest loss from unavailability of fund monies
for restoration.
Because the burden of payment for restoration of a natural resource
may rest ultimately with the state, it is important for all states to have
statutory mechanisms in place that will provide the legal basis for recov-
ery of damages for injury to a natural resource. This paper provides
an overview of the current issues in the law relating to recovery for
injury to natural resources. With this overview as background, the dis-
cussion then summarizes each state's statutes pertaining to a hazardous
substance release injuring natural resources. Finally, this paper reviews
certain state statutes that are especially sensitive to natural resource
injury and provides recommendations for improvement by the states
in this area.
INTRODUCTION
CERCLA as amended by SARA is a monumental piece of legisla-
tion that attempts to provide a framework for response to the release
of a hazardous substance. The enormity of the task to draft such a com-
plex public law had the understandable result of falling short when
certain provisions were acted upon.
Much of CERCLA/SARA gave only the statutory basis for response
to an event, and left for the President the task of drafting regulations
that describe the methodology to be followed in implementing each
section. One such example is the law and regulations related to release
of a hazardous substance causing injury to natural resources.
Under CERCLA/SARA, the provision for liability in responding to
natural resource injury can be found in §9607(a)(4)(C). This section
is straightforward in its meaning, and complete in that it provides for
recovery of costs in response to natural resource injury.' To implement
recovery for injury to natural resources, the President was required by
CERCLA to promulgate regulations for the assessment of injury to
natural resources.2 The regulations were finally published on Aug. 1,
1986.3 SARA was passed on Oct. 17, 1986, and Congress gave the
President 6 mo from that date to revise the Aug. 1, 1986 regulations
to conform with SARA.4 These revised regulations were published in
February, 1988.5
The Type A and B regulations for assessment of damages to injury,
destruction or loss of natural resources, as published by the U.S. Depart-
ment of the Interior, were immediately challenged in federal court. The
U.S. Court of Appeals, D.C. Circuit, ultimately reached a decision in
these cases on July 14, 1989.6 The opinions are extremely well
reasoned and provide an excellent historical perspective of the
CERCLA/SARA natural resource damage provisions. Generally, the
holding in these opinions is that the U.S. Department of Interior must
revise the assessment regulations and provide a broader approach to
ensure injured, lost or destroyed natural resources are made whole.
The D.C. Circuit Court of Appeals realized that natural resources
are not easily replaceable. As a result, when injury, destruction or loss
occurs to a natural resource, any damages recovered to address the harm
must, if possible, be sufficient to reestablish said natural resource in
the environment. The message sent by the D.C. Circuit Court of Appeals
is that natural resources are finite, and every legislative mechanism
should be fully utilized to maintain those natural resources remaining
in the environment.
A majority of the plaintiffs challenging the U.S. Department of the
Interior's assessment provisions were state governments.7 It is the state
that suffers the most immediate loss when harm to a natural resource
occurs through release of a hazardous substance. A state is more per-
sonally involved in the natural resources existing within its boundaries.
Often, it is the state that is uniquely knowledgeable about particular
natural resources and their value to the public. As a result, it is the
state that stands in the best position to promote the preservation and
growth of its natural resources. A recent opinion confirms the state's
role in enforcement of laws for the protection of the environment.8
States can avail themselves of the natural resource damage provisions
within CERCLA/SARA to establish a claim for recovery of costs for
injury to, destruction of or loss of natural resources. States can utilize
those natural resource damage assessment provisions published as regu-
lations. What the state cannot do is take money from the Hazardous
Substance Response Fund to pay for assessment and resultant harm to
a natural resource.9 With the forthcoming changes to promulgated
natural resource damage assessment regulations, state reliance on this
area of federal laws and regulations is problematic.
A state is not required by federal law to rely solely on the
CERCLA/SARA natural resource damage laws and associated regula-
tions. States are sovereigns, and as such can pass their own laws per-
taining to natural resource injury and damage assessments. Generally,
said laws cannot be less strict than federal laws, or conflict with the
intent and purpose of the federal law. With the absence of Superfund
monies to pay for injury to natural resources, a state must rely on other
laws (i.e., state common law) that provide a basis for recovery. Or, as
a number of states have done, they can pass their own laws pertaining
to natural resource injury and recovery for damages.
The following sections provide a general discussion of state laws that
STATE PROGRAMS 613
-------
Table 1
State Statutes Pertaining to Release of a
Hazardous Substance and Natural Resource Injury
anoo Mwift •
mtma
mm
• m
ssr—
i vn n K M* « I
M. MTU. •*»,-»*.-»•
* AMMUdMf TM* H
it-m •*•*?> tm
614 STATE PROGR>tM$
-------
pertain to release of a hazardous substance causing injury, loss or des-
truction of a natural resource. From this general survey of state laws,
the discussion focuses on specific states whose laws in this area are
of particular note. Finally, recommendations are made for states to
improve their laws and regulations in this area.
STATE ENVIRONMENTAL LAWS
All states have some form of law concerning the natural environment.
Since the advent of federal laws on hazardous waste management and
release of hazardous substances, the majority of states have passed laws
similar to those passed by the U.S. Congress. Some states have simply
passed the federal law as a state law, with little or no change in the
substantive language. Other states have passed environmental laws that
are, for the most part, original pieces of legislation. Still others have
passed very little in the way of environmental legislation, perhaps relying
on the federal laws and common law theories of recovery.
Table 1, column 3, provides a list of each state's environmental laws.
To develop this list of state statute citations, certain criteria were utilized.
Initially, a state's statutes were surveyed for laws pertaining to the release
of hazardous substances in the environment, a'la CERCLA/SARA. The
next statutes of interest were those concerning management of hazardous
wastes, a'la RCRA. Then the state statutes were surveyed for other laws
pertaining to environmental pollution, such as air and water statutes.
Those statutes concerning radioactive material and waste were not
included in the list of state environmental statutes in Table 1.
Column 4 of Table 1 is a list of those state statutes related to release
of a hazardous substance causing injury, loss or destruction to natural
resources. To develop this list of state statutes, certain criteria were
utilized. Initially, the environmental law statutes of a state were reviewed
to determine if there were any specific statutes concerning injury to
natural resources. If no statutes could be found directly on point, then
the state environmental laws were reviewed to determine if related
statutes could be used as authority to bring a state claim for injury,
loss or destruction of natural resources. An absence of either type of
statute led to a review of general water pollution or similar type statutes
whereby a claim for natural resource injury, loss or destruction could
be made.
There are two caveats to reliance on the statutes listed in Column 4.
The first is that these statutes are in many cases current only through
the 1988 legislative session. Every attempt was made to secure laws
as up-to-date as possible using the Advance Legislative Service, etc.
At best, this list is current through July, 1989. The other caveat con-
cerns legal interpretation of a state's statutes by the state attorney general
or other legal representative. A state may rely on other types of damage,
nuisance, fish and wildlife, agricultural, etc. statutes as a basis for mak-
ing the same claim as the federal government would under the language
of CERCLA/SARA §9607(a)(4)(C).'° With these caveats in mind, the
next section discusses particular state statutes that have specifically ad-
dressed a hazardous substance release causing injury, destruction or
loss to a natural resource.
SELECTED STATE NATURAL RESOURCE INJURY STATUTES
As discussed previously, a state can have laws similar to the federal
laws, as long as the state law does not conflict with the purpose and
intent of the like federal law. One distinct advantage for a state to have
its own law on a certain subject, and not rely on federal law, has to
do with the forum in which a case is litigated. In a number of situa-
tions, a state may find it more advantageous to present its case in a
state court instead of a federal court. A state in this situation would
have the opportunity to be heard by a judge who is very familiar with
the intent of the state's laws." Forum shopping is a practice actively
engaged in by many litigants, and a state action to enforce a state law
has a better chance of remaining in the state court.
It is to the state's advantage to draft its own laws when the subject
matter holds a special interest for the state, is already highly regulated
by other state laws and/or is of a unique and complex nature in which
the state has previously invested time and money to gain a better
understanding. State statutes that are carefully and specifically drafted
to treat such technical subject matter usually will be given great
deference by a court of law.
State statutes that apply to release of a hazardous substance causing
injury, destruction or loss of a natural resource fall into the definition
of unique subject matter as discussed in the previous paragraph. Natural
resources, as defined by federal and state laws, include almost every-
thing living in the natural environment.12 Most states have statutes con-
cerning hazardous substances and/or pollution in general. All states
have statutes concerning quality and regulation of fish, wildlife, air,
water, soil, biota and other natural resources. Natural resources, as they
exist, are of great interest to a state. Why then have more than half
of our states not considered it worthwhile to pass laws that address in-
jury, destruction or loss to a natural resource?
The following is a selection of states that utilize their statutes to address
harm to a natural resource from release of a hazardous substance. For
each state, a brief comment regarding the state's statutory scheme is
included. There are states not discussed here that have statutory
mechanisms for addressing injury to a natural resource. However, the
states that were selected for inclusion have unique or comprehensive
approaches to natural resource injury, destruction or loss.
California
In the California code, there are a number of sections concerning
injury to and damages for natural resources. California has a trust
fund/account set up to provide for the assessment and replacement of
injured natural resources. ° California provides for punitive damages
when the injury, loss or destruction to a natural resource occurred after
Sept. 25, 1981.M Due to the state's interest in marine natural resources,
there is a provision specifically for release of a hazardous substance
causing injury in this type of environment.15
Colorado
Colorado relies on CERCLA/SARA and other federal environmental
laws, but also provides specific statutes where natural resource injury
results in recovery of damages." Monies recovered for the CERCLA
fund in Colorado, unlike the federal Hazardous Substance Response
Fund, can be used to restore, replace, etc. natural resources.
Connecticut
Connecticut not only provides for liability to a person causing injury,
destruction or loss of a natural resource, they also provide for imposi-
tion of civil penalties towards such action.17
Iowa
A statute directly on point that includes cost of damage assessment
and punitive damages for willful release.18
Louisiana
Louisiana, like several states, has placed within its constitution a
section concerning the public policy interest of protecting natural
resources and the environment.19 Thus, natural resources could
arguably be considered a constitutionally protected interest. Monies
collected from a responsible party can be used for restoration of the
natural resource.
Maine
Maine has an extensive liability statute on recovery of damages for
injury, destruction or loss of a natural resource.20 This statute also in-
cludes punitive damages.
Maryland
As with a number of states, Maryland provides a reference to natural
resource injury when discussing removal or remedial actions.21 The
state also provides for a State Hazardous Substance Control Fund and
that expenditures from this fund can include natural resource injury,
loss or destruction.22
Massachusetts
Massachusetts has an extensive statute on liability for injury, des-
truction or loss of natural resources that paraphrases
STATE PROGRAMS 615
-------
CERCLA/SARA.-'1 Massachusetts also provides a definition of
"damage to the environment.""
Minnesota
Minnesota has one of the more extensive statutes on release of a
hazardous substance causing injury, destruction or loss of natural
resources.-'5 The statute is modeled after CERCLA/SARA. The state
also describes who is the trustee for natural resources.J* There is an
Environmental Response. Compensation, and Compliance Fund from
which monies can be spent on natural resource injury, loss or
destruction."
Montana
Montana's liability statute is to the point, but does not specifically
refer to recovery of assessment costs.31 The statute has a provision for
punitive damages as well. Montana is also included as an example of
many state statutes. Based on the language under the section providing
for remedial action, Montana could bring an action for liability to injury.
loss or destruction of the environment. The term environment is arguably
analogous to natural resources."
North Carolina
North Carolina has an excellent, extensive statute on liability for
injury, destruction or loss of natural resources. *' The statue also dis-
cusses assessment of damages more extensively than most states.
Punitive damages are available for intentional or negligent release."
Oregon
In addition to the language of federal law placed in their environ-
mental statutes. Oregon also provides a price list for natural resource
replacement.^
Pennsylvania
Pennsylvania has an excellent series of provisions for protection of
natural resources. The Pennsylvania constitution affirms that preser-
vation of natural resources is an important public consideration." The
state basically has modeled its laws concerning liability for injury, des-
truction or loss of natural resources after CERCLA/SARA.") The
state has a Hazardous Sites Cleanup Fund which provides for expendi-
tures to restore, rehabilitate or acquire natural resources." Pennsylva-
nia also has a statute focusing on economic evaluation methods to arrive
at damages for loss of fish or wildlife.1* The language in this section,
while not referring directly to hazardous substances, could arguably
include a release of such in causation.
South Dakota
South Dakota laws are somewhat general as they relate to damages
for injury, destruction or loss to natural resources, with a need for
inference. Several sections of the South Dakota law provide very specific
procedures for actions by the state to recover damages for injury, des-
truction or loss of natural resources." The reference is not to release
of a hazardous substance, but to pollution. However, in this context.
these terms arc analogous.
Washington
The Washington statute concerning liability fur injury, loss or des-
truction to natural resources should be considered a model in this area
of legislation.u The statute makes it very clear that restoration of the
environment is the goal to be achieved and damages will be sought for
the full amount. This statute shows unusual foresight in light of the
recent natural resource damage assessment opinions previously dis-
cussed." The Washington Department of Fisheries and Game is given
the responsibility of determining the pre-harm condition of the natural
resource.
CONCLUSIONS AND RECOMMENDATIONS
When the release of a hazardous substance occurs, concerns related
to impacts on the public health and welfare are immediately addressed.
Simultaneous with this release event, there may have been injury, loss
or destruction of natural resources. The public is a vocal and persis-
tent advocate for remedy of potential health impacts resulting from the
release. The natural resources that are harmed, however, cannot com-
municate the degree of present and potential injury that has occurred.
Often, a significant period of time elapses before the effects of a release
on natural resources are addressed.
The federal government has an interest in the protection of natural
resources. But this interest is dedicated to a nationwide responsibility.
Federal management of natural resources, and representation as trustee
in matters affecting said resources, cannot begin to adequately protect
localized resource related interests. It is, therefore, the responsibility
of individual states to ensure their natural resources are being given
the proper degree of protection.
For a stale to provide adequate protection of natural resources when
the release of a hazardous substance occurs does not require large capital
expenditures and additional bureaucracy. Every state has the basic tools
required to ensure protection of their natural resource interests. For
example, each state has some form of a Department of Natural Resources
or equivalent agency. This department, along with assistance from
universities within the state, can provide the technical resources required
to determine the impacts of a hazardous substance release on the natural
environment. Stales also have an office of the attorney general. This
office can provide the legal resources to hold those persons responsi-
ble for resultant harm from the release of a hazardous substance.
What many stales do not have in place, however, is the comprehen-
sive statutory language needed to adequately recover the monies for
restoration of injured, lost or destroyed natural resources. While the
specific language and approach of individual state's statutes will vary,
said statutes should, at a minimum, address the following: who is the
trustee and what is the level of response authority; how is the injury,
destruction or loss to be assessed; liability for injury, loss or destruc-
tion of a natural resource to pre-release levels; punitive damages and/or
criminal penalties for recalcitrant responsible panics; if the state has
a hazardous substance response fund, provide access to the fund monies
for restoration, replacement, etc. of natural resources.
Of the 15 states discussed in this paper for their statutory approach
to protect natural resource interests, the following five should be con-
sidered noteworthy: Minnesota. North Carolina, Pennsylvania, South
Dakota and Washington. These state statutes, along with the language
of CERCLA/SARA. should be utilized to derive a series of statutes
that will adequately protect the natural resources located within an in-
dividual state.
REFERENCES
I 42 USC §9607(aX4MO holds a person liable for "damages for injury W,
destruction of, or loss of natural resources, including the reasonable oasis
of assessing such injury, destruction, or loss resulting from such a release,
and. . "
2. 42 USC §30I|c)(l>,(2) The President delegated the authority to promul-
gate natural resource damage assessment regulations to the U.S. Depart-
ment of the Interior
3. 43 CFR $11.10 through 11.93 (1987).
4 42 USC §9651(c)OU2)
5 53 F.R 5166 (1988)
6 Type B Regulations - Ohio et. al. v. U.S. Department of the Interior, No.
86-1529 (D.C. Cir 1989). Type A regulations - Colorado v. U.S. Dqwt-
mcni of the Interior, No. 87-1265 (D.C Cir. 1989).
7. Id. note 6
8 Colorado v. Idarado Mining Co., 707 F. Supp. 1227 (D.C. Colorado 1989).
9. Landrcih, Lloyd W., Recovery for Natural Resource Damages on Super-
fund Sites, p. 605 n. 8, Hazardous Maif rials Control Research Institute,
9th National Cortfcrence Proceedings, 605-607 (November 1988).
10. The third caveat is that a statute may have been so obscured within a general
area of law thai it was never uncovered.
II. This is not to say that a federal district court residing in the state would
not be any less competent in matters of state law.
12 For the CERCLA/SARA definition of natural resources see 42 USC
§9601(16).
13. West's California Code. Health and Safety Code §25352 (1989).
14. Id at 25359, 25359.1.
15. Id al Harbors and Navigation Code §293 (1989).
16. Colorado Revised Statutes §25-16-104.7, 201 (1988).
17. Connecticut General Slat. Ann. §22a-6a, 6b (1989).
616 STATE PROGRAMS
-------
18. Iowa Code Annotated §455 B.392 (1989).
19. Louisiana Constit. Article IX, §1.
20. Maine Revised Stat. Ann. Title 38, §1367 (1988).
21. Ann. Code of Maryland. Environment §7-222 (1988).
22. Id at §7-220-221 (1988).
23. Massachusetts General Laws Ann. Chapter 21 E §5 (1989).
24. Id at Chapter 214 §7A (1989).
25. Minnesota Stat. Ann. §115B.04 (1989).
26. Id at 115B.17.
27. Id at 115B.20.
28. Montana Stat. Ann. §75-10-715 (1988).
29. Id at §75-10-711 (1988).
30. General Stat. of N. Carolina, Art. 21A §143-215.90 (1989).
31. Id at §143-215.91 (1989).
32. Oregon Revised Stat. Ann., Hazardous Waste and Materials, Chapter 466.890
(1988). While this statute is roughly on point, the prices for these natural
resources are likely low compared to restoration value.
33. Purdon's Pennsylvania Stat. Ann. Constitution Art. 1 §27.
34. 34 Purdon's Pennsylvania Stat. Ann. §6020.702 (1989).
35. Id at 35 P.S. §6020.902 (1989).
36. Id at 34 P.S. §2161 (1989).
37. South Dakota Codified Laws, Chapter 34A-10, §34A-10-1 through 10-15 (1989).
38. Revised Code of Washington Annotated, Title 90, §90.48.142 (1989).
39. See note 6.
STATE PROGRAMS 617
-------
Institutional Controls of Waste Sites:
The Groundwater Management Zone
Michael A. Apgar, P.G.
John T. Barndt
State of Delaware
Department of Natural Resources and Environmental Control
Division of Air and Waste Management
Dover, Delaware
ABSTRACT
The management of waste sites which are contaminating groundwater
typically involve source control or removal, and aquifer restoration.
Generally this process relies on federal authority and involves the
expenditure of massive amounts of money. All too often, these expen-
ditures far exceed any damages or even potential damages they are
intended to rectify or prevent. Further, they do not preclude inappro-
priate future occupation of the site and potential future exposure to
contaminants.
Usually, the protection of human health and the environment can be
accomplished just as adequately—and far less expensively—by using
alternate unthreatened water supplies in combination with State and local
authorities controlling well construction, water withdrawals and land
use. In fact, such controls should be considered for application to any
waste*related case of groundwater contamination regardless of other
remedial activities.
The employment of state water controls and local land use restric-
tions to create groundwater management zones has been undertaken
in Delaware as a means of protecting existing and future local residents
from exposure to contaminated groundwater near waste sites. These
same state and local authorities have been used to augment federal
authority at Superfund sites as well as to manage groundwater con-
tamination situations where federal authority was not thought to bear.
INTRODUCTION
Groundwater contamination has been documented at many sites of
past waste disposal. It can be argued that groundwater contamination
is the inevitable result of past waste disposal practices where little effort
was employed to minimize water entrance to the waste and prevent
leachate percolation to the subsurface. However, this groundwater con-
tamination is virtually always a local problem for which local tailored
solutions are the most efficient.
Sometimes the contamination from old waste sites is of such a toxic
and persistent nature and of such considerable area extent that valua-
ble aquifers and, in a relatively few cases, high capacity water supplies,
are affected by the contaminants. Sometimes, but even less commonly,
the groundwater contaminants discharge into streams at rates sufficient
to degrade stream quality to the point that uses are impaired. However,
the loss of irreplaceable groundwater supplies to contamination is rare
and adverse impacts on major streams or significant segments of minor
streams is even less frequent.
Nonetheless, the typical public and governmental response to the
discovery of groundwater contamination at abandoned waste sites is
to insist on cleanup. The cleanup process often involves the invocation
of the federal Superfund program or its state equivalents. These
processes entail a comprehensive investigation of the nature and extent
and existing and potential impacts of contamination, an assessment of
alternatives, selection of a remedy (usually designed to achieve strict
cleanup levels on-site). implementation of remedial action(s) and long-
term operation, maintenance and monitoring. The cost for a single site
in this process averages approximately $20 million per site*.
Fortunately, the majority of cases of groundwater contamination from
old waste sites have limited the relatively minor adverse impact on water
supplies and the environment. Ninety-five percent of Delaware is in
the Atlantic Coastal Plain, much of which is blanketed by permeable
unconsolidated sands and annual recharge averages 12 to 16 in./yr. Wtfer
supplies in the coastal plain are derived from water wells, so ground-
water is both vital for the health and economy and is often very vulner-
able to contamination. Even so, few of the Superfund sites have mpwfrA
major wells or even threatened aquifers capable of supporting large
supply wells.
The typical effect of old waste sites is a contaminant plume of rela-
tively limited area which extends downgradient to or towards nearby
streams. Generally, however, the concentrations of many contaminants
are effectively reduced below levels of detection and/or concern during
transport through porous earth materials. In fact, natural attenuation
of contaminants is sufficient in many areas to adequately treat ground-
water before it reaches a water supply well or discharges to a stream.
In several cases, nearby wells have been contaminated or are threa-
tened by contamination to a degree which would render the water un-
suitable for use without treatment.
Although volatile and semi-volatile organic compounds generally are
present in the leachate of these waste sites in concentrations objection-
able for public water supply, these compounds rarely are present in
appreciable concentrations in ground water downgradient of the sites
and have little or no significant impact on local streams. The most com-
mon objectionable off-site contaminants in the groundwater are dissolved
inorganics, iron and/or manganese. Although relatively mundane, these
contaminants still render local groundwater unfit for use without treat-
ment. Treatment costs for iron removal typically are greater than for
removal of dissolved organics. Also, individual well water treatment
systems usually exchange sodium for the dissolved iron, increasing the
intake of a substance linked to high blood pressure.
In most cases, locally contaminated or threatened water wells can
be replaced by deeper wells or by the extension of a public water system
whose source is not threatened by the contaminants. The cost of
providing alternative safe water supplies by use of treatment or replace-
ment is typically only a fraction of the costs to remove or control the
contaminant source and to decontaminate the aquifer to a quality suitable
for use at the waste site boundary.
The only practical dilemma with a cost-effective remedy of providing
safe replacement water supplies is the potential for future occupation
618 STATE PROOR/TMS
-------
of the waste site or drilling of water wells in the contaminant plume.
However, the incorporation of state authority over water development
and local controls over land-use make it possible to prevent future
exposure to the contaminants.
The use of both state and local authorities to prevent the exposure
to contaminants is prudent regardless of whether an outright no-action
alternative is followed or a very complicated and lengthy remedial
alternative is selected. In either case, some form of controls to prevent
water well development or inappropriate land-use are warranted. Res-
triction of land and water uses will prevent further exposure to con-
taminants both on-or off-site. Typically, remedial actions require many
years to complete and even longer to achieve desired results.
APPLICABLE STATE WATER MANAGEMENT POLICIES
State water management policies vary considerably. However, good
policy should support a practical, common sense approach to problem
solving. Delaware's official Groundwater Management Policy states that
"The geologic materials which comprise an aquifer possess a limited
capacity to attenuate certain contaminants. This capacity should be con-
sidered in remedial actions for existing incidents of groundwater con-
tamination. . .if (the abilities of earth materials to attenuate
contaminants) are judged effective, then dependence on these proper-
ties may be a workable remedial strategy." It also states that "ground-
water quality management should be integrated with the management
of water supplies such that one activity does not contravene another."3
The state groundwater management plan takes the position that "Dela-
ware's official water management policy allows consideration of the use
of earth materials to attenuate contaminants in groundwater if (1) no
environmental harm results and (2) water supply development is
integrated with waste management so that wells do not draw water with
objectionable concentrations of contaminants."... or further, "Based on
the costs and limited success of groundwater recovery programs, an
alternative to pumping and treating contaminated groundwater should
be developed."4
Adequate protection - not cost - should be the primary consideration
in any contamination management program. We should remember that
the key role of government in managing water and wastes is to prevent
the exposure of unacceptable levels of contaminants to people or the
environment. In Delaware, this state goal is expressed as "to ensure
sufficient groundwater quality for the protection of pubic health and
for such beneficial uses as may be desired, including the preservation
of significant ecological systems, now and in the future."3
This goal can be accomplished by requiring safe sources of water
for users in areas of contamination and preventing the development of
wells or activities near or on waste sources to prevent exposure to the
contaminants. The basic concept for groundwater contamination
management zones is not new7. In fact, the avoidance of contaminant
plumes by provision of alternate water has been a traditional though
unofficial response to cases of groundwater contamination. Besides ef-
fectively preventing exposure to contaminants, this approach usually
is at least an order of magnitude less expensive than major waste source
controls, groundwater decontamination and continuous extensive
monitoring of groundwater quality.
The authorities to regulate water supply development and to restrict
land use traditionally have been vested with state and local governments,
respectively. In Delaware, the authority to regulate water supplies rests
with DNREC and includes:
• Licensing of water well contractors
• Requiring permits to construct all wells
• Issuance of permits for well construction only to licensed water well
contractors
• Requiring a separate water withdrawal permit for withdrawal rates
greater than 50,000 gpd
The authority to regulate land-use exists at the county and municipal
levels. These local authorities include land-use planning, zoning and
building and occupancy permits. Additionally, local authorities can insert
use restrictions into property deeds.
Delaware's groundwater management plan concludes by stating that
"regulatory controls, existing policy and economic practicality would
allow natural attenuation of contaminants from existing waste sites in
groundwater where 1) such attenuation would have no significant adverse
impact on public water supply sources, potential water supply sources,
the ecosystem, or aquatic life; and 2) uncontaminated water would be
available to meet existing and future water supply needs." These con-
ditions can be met in a number of instances4.
To bring existing state and local authorities are brought to bear on
managing groundwater contamination by attenuation without con-
taminating any water supplies, a joint effort of several agencies is
required. These include:
• Delineation of a contaminant attenuation/well restriction zone by the
state's Department of Natural Resources and Environmental Control
in cases where groundwater contaminants would have no significant
impact on the environment at the point of discharge to the surface.
Use of land-use controls by local governments to ensure that homes
are not build and- occupied in the restriction zones unless a safe,
uncontaminated source of water is available to the occupants.
• A commitment by the party(ies) responsible for the contamination
to provide a safe, uncontaminated water supply to those water users
in the well restriction zone.
"Where these conditions are met and the controls previously out-
lined can be jointly arranged by the parties, groundwater attenuation/well
restriction zones should be formally delineated to manage incidents of
groundwater contamination."4
Groundwater management zones have been officially designated at
several abandoned waste disposal sites in Delaware1 including, recent-
ly, a Federal Superfund site. These designations have included the deline-
ation of areas in which existing threatened wells must be replaced and
no future threatened wells can be constructed (these include negotia-
tions to include deed notices and deed restrictions) and the provision
of alternate safe water supplies at the expense of the responsible party.
These designations have been made at sites addressed entirely by the
state's environmental regulatory authority and at a Superfund site by
incorporation into the Record of Decision.
CASE 1: SUSSEX COUNTY LANDFILLS
(BRIDGEVILLE LANDFILL)
During the 1970s the Sussex County government operated six land-
fills for the disposal of solid waste generated in the county. These land-
fills were unlined and were constructed in permeable sandy soils with
shallow water tables. Monitoring wells installed at the landfills detected
contamination of groundwater beneath and adjacent to each site. This
contamination included low concentrations of hazardous substances and
highly elevated dissolved solids and iron concentrations. These latter
contaminants render the water immediately downgradient of the land-
fills unfit for water supply purposes. However, little adverse impact
is observed or anticipated in streams to which the landfill contaminated
groundwater is or will be discharging.
The contaminants from landfills which abut streams are discharged
with groundwater directly to the streams baseflow. There is little poten-
tial for these landfills (Omar, Stockley and Anderson's Crossroads) to
contaminate groundwater supplies. The impact on surface water—though
detectable—is not significantly adverse to biota or other possible uses
of the streams. The other landfills (Bridgeville, Laurel and Angola)
have groundwater flow paths extending up to several thousand feet to
streams into which they discharge. The landfill-contaminated ground-
water from these sites will have no detectable impact on the surface
waters to which they eventually will discharge, but do threaten exist-
ing and possible future groundwater supplies.
Generally little groundwater development occurs downgradient of
these landfills (which are in rural areas). However, a few wells do exist
in the unconfined aquifer in areas threatened by contamination and
additional future development is possible. A report documenting the
groundwater conditions at these sites was required by the state. The
county's report8 documented these conditions and recommended the
establishment of Groundwater Management Zones (GMZs).
Subsequently, the State of Delaware and Sussex County executed
STATE PROGRAMS 619
-------
a Memorandum of Understanding9 establishing the GMZs as a prac-
tical, cost-effective and sure way of resolving the threat of water con-
tamination posed by the landfills. The GMZs consist of three concentric
areas, drawn (on maps) for each of the landfill sites, wherein ground-
water withdrawal rates and well designs are regulated. The GMZs are
designated as follows:
• "No well zone:" all wells prohibited from the unconfined aquifer;
• "GMZ A:" wells screened in the unconfined aquifer to be pumped
at a rate of 10 gpm are prohibited;
• "GMZ B:" wells screened in the unconfined aquifer to be pumped
at a rate of more than 100 gpm are prohibited.
The principal points of this agreement provide that:
• Permit applications for wells to be located in the "no well zone"
may be issued only after a joint review has been conducted by the
DNREC and Sussex County. Where (1) a central public water supply
is available well permits will be denied; or (2) public water will be
available within five years of the request for water service, a well
permit may be issued by DNREC and continued that it be abandoned
once public water is available; or (3) public water is not provided.
a method of supplying water safe from landfill contamination, such
as a double-cased well screened in a confined aquifer, may be
permitted by DNREC.
• Sussex County must replace all existing wells (with the exception
of irrigation wells) located in the "no well zone" and screened in
the unconfined aquifer with either a public water supply of accepta-
ble quality or an alternative source such as double-cased well in a
confined aquifer. A schedule for accomplishing this work prioritized
according to threatened wells was required within 6 mo of the date
of execution of the MOU This well replacement schedule had to
be such that existing wells will be free from contamination at all
times. Existing irrigation wells in GMZs had to be investigated to
determine the effects of such pumping, and the county and DNREC
were to determine the best course of action to be taken.
Figure I
Calibrated Hydraulic Head Distribution for Bridgeville Landfill
For new wells that would, but for the landfill contamination or threat
of contamination, be constructed within a "no well zone," Sussex County
must pay the difference in cost between a water supply like the cost
of a typical domestic well completed in the unconfined aquiferand one
safe from landfill contamination, such as a double-cased well in a con-
fined aquifer or an extension of an acceptable public water supply. This
requirement did not include properties that are leased, transferred or
subdivided after the notification or property restrictions are provided
by Sussex County.
Sussex County had to create, within 6 mo of the execution of this
MOU, a procedure that would indicate that a particular parcel is located
within a GMZ when a proper title search is performed. Additionally,
the county had to develop, within the same six month period, other
mechanisms (deed restrictions, zoning requirements, etc.) that will alert
potential buyers to the intent and content of (his MOU.
The Bridgeville Landfill received mixed municipal, residential and
industrial solid waste between 1968 and 1984, Wastes were buried in
unlined trenches excavated in sandy loam soil to the top of the zone
of saturation (approximately) 10 ft below ground surface). The landfill
covers approximately 135 ac.
The unconfined (Columbia) aquifer is an unconsolidated, medium
textured, predominantly quartz sand which was deposited by Quater-
nary streams. The aquifer has a saturated thickness in the vicinity of
the Bridgeville landfill of approximately 120 ft and a iransmissiviiy of
about 83,000 gpd/ft.
No deeper aquifers have been explored locally because of the high
productivity and generally excellent water quality of the unconfined
aquifer. Regionally, deeper aquifers often contain water with objection-
able concentrations of (naturally occurring) iron.
The water table beneath the Bridgeville landfill slopes toward the
Nanticoke River as shown in Figure 1. The Nanticoke functions as the
regional groundwater drain.
Groundwater beneath and immediately downgradient of the Bridge-
ville landfill contains objectionably high concentrations of iron and has
a putrescible odor. The maximum concentrations of priority pollutani
volatile and semivolalile organics total less than 100 p/L. This water
is unsuitable for supply to the growing trailer park and individual resi-
dences downgradient of the landfill. However, it will not have any signifi-
cant adverse (or likely, even delectable) impact on the Nanticoke River
because of attenuation prior to reaching the river and dilution with the
river water.
The GMZs for the Bridgeville Landfill are shown in Figure 2. The
proposed source of safe alternate water supply for existing and future
water users within the GMZ is a public well field located north of an
outside the GMZ. A water transmission main will extend from this new
wellfield to the threatened water users. Because of the number of
threatened water users, the Bridgeville landfill is the number one priority
of the Sussex County government. The new replacement public voter
system is scheduled for completion during 1990.
CASE 2: WILDCAT LANDFILL
The Wildcat Landfill was a privately owned and operated landfill
which accepted both municipal and industrial wastes from 1962 until
1973. Following its inclusion on the NPL in 1982, an RI/FS3 was con-
ducted followed by two RODs in June and November, 1988. The
remedial action selected for the landfill and the adjacent areas was
detailed in the June, 1988 ROD. Besides the general requirements for
alleviating problems at the landfill, the ROD specified the need for
administrative and institutional controls both upon the landfill and in
areas adjacent to the landfill which were susceptible to groundwwer
contamination originating from the landfill. Figure 3 shows the general
features of the landfill flow.
620 STATE PROGRAMS
-------
Legend
—60— Concentration Contours
" ,s All New Wells Restricted;
' ' ' Required Monitoring of
Existing Weils
SHU Wells Less Than
Ł2m 10GPM Allowed
JgSSSSj Wells Less Than
SSSSS 100 GPM Allowed
Figure 2
Groundwater Restriction Zone
for Bridgeville Landfill
(Weston, 1987)
N
Scale in Feet
1.000
475000
APPROXIMATE LANDFILL BOUNDARY
N\\\\l AREA OF GMZ "A"
AREA OF GMZ "B"
Figure 3
Groundwater Management Zone for the Wildcat Landfill
Superfund Site (After CH2M Hill, 1988)
Since the source of contamination, the landfill, was not to be removed,
surficial aquifer which was affected by the off-site movement of
contaminants was of very limited useable value and since the landfill
discharges into nearly surface water bodies, administrative and institu-
tional controls were deemed to be an appropriate and cost-effective
approach to preventing human exposure from contaminated ground-
water. Contaminated groundwater discharged directly into the St. Jones
River and Tidbury Creek where contaminant levels would not likely
exceed federal limits.
The administrative and institutional controls associated with the
selected remedy included the following:
• Water well installation (except monitor wells) would not be permitted
within the landfill boundary;
• Water wells within the shallow, unconfined aquifer in areas down-
gradient and other nearby areas would not be permitted. Deep wells
constructed in to the local confined aquifers may be permitted
provided special conditions established the DNREC are met (e.g.,
double cased;;
• Existing shallow wells within the GMZs are to be replaced by the
responsible party with deep wells screened in a confined aquifer;
• Commercial and residential building on the landfill would not be
permitted;
• Governmental agencies would work toward obtaining agreements from
property owners to have restrictive language placed into deeds to
prevent future building or other activities which could expose humans
to landfill wastes or contaminants in the water.
Both DNREC and DPA recognized that some form of institutional
control was necessary to prevent future exposure to the population in
the future. Although the general guidelines for some form of control
were mentioned in the ROD, it was left to DNREC to develop the specif-
ic mechanism for developing reliable controls. DNREC had recently
concluded the agreement (Case 1) with Sussex County. In the case of
the Wildcat Landfill, however, the landfill had been privately owned
and operated.
To meet the requirements of the ROD, DNREC developed an inter-
nal mechanism for ensuring that water wells were not constructed within
the restricted areas. A Memorandum of Agreement (MOA) was for-
malized between state agencies responsible for management of the
Superfund program and for groundwater management programs. This
agreement6 defined the groundwater management zones (GMZs)
associated with the landfill and adjacent areas. Figure 3 illustrates the
GMZs established in the agreement:
• GMZ "A": areas where no water wells are permitted, except for
monitor wells
• GMZ "B": areas where no shallow water wells are permitted but
where deep wells may be permitted following joint review of the
permits by the DAWM and the DWR
A copy of this agreement was subsequently provided to U.S. EPA
and included in the administrative record for the site.
Concurrent with the purely administrative nature of the MOA,
DNREC and U.S. EPA negotiated with a PRP group to implement the
STATE PROGRAMS 621
-------
remedies selected in the RODs for the landfill and an adjacent pond.
The PAP group for the Wildcat landfill included the properly owner.
Consequently, DNREC and U.S. EPA requested that the owner volun-
tarily include restrictive language into the property deed which, among
other things, served notice to future property owners of the presence
of landfill areas.
The MOA complimented the voluntary cooperation of the property
owner in providing the restrictive language in placing "permanent" con-
trols on preventing future exposure to contaminants both on the site
and adjacent to the site.
CONCLUSIONS
The formal designation of groundwatcr management zones has been
made at several abandoned waste disposal sites and one Superfund site
in Delaware. These zones include portions of an aquifer in which con-
taminants will be allowed to attenuate, existing threatened water wells
must be replaced new threatened water wells are prohibited and supply
of water to existing and future occupants of the zones is provided from
an unthreatened source by the responsible party.
These GMZs are a creative, practical, adequate, cost-effective alter-
native or supplement to a remedial alternative. However, as Delaware's
Groundwater Management Plan cautions:
"Obviously not all contamination instances will allow for a ground-
water attenuation/well restriction zone remedial management alterna-
tive. If. after transportation and attenuation in the subsurface.
groundwater used for supply purposes or in the protection and propa-
gation of aquatic organisms fails to meet the criteria for its designated
use and no alternative water source is available, corrective actions may
be required.
The choice between corrective action and a groundwater attenua-
tion/well restriction zone option must be made on a case-by-case basis
after careful consideration of the technical and administrative merits
of each case. Clearly, the intent in both management options is the
protection of human health and the environment."
DISCLAIMER
Any opinions expressed are (hose of the authors' and are not neces-
sarily those of the State of Delaware Department of Natural Resource!
and Environmental Control.
REFERENCES
I Apgar, MA. and Cherry PJ.. Jordan J.H and S. N. William, "The Grouod-
waicr Management Zone - An Alternative to Cosily Remediation," Ground-
witcr Monitoring Review, 1988.
2. CHEM Hill Southern*. Inc. Wildcat Landfill Remedial Imtstiganvt Report
Vol. I. 1988
3. Delaware Department of Natural Resources and Environmental Control,
"Groundwater Quality Management." From The Management of Wider
Resources in Delaware (officially adoped as DNREC policy). Prepared under
the Direction of the Comprehensive Water Resources Management Commioce,
1983. 100 pp.
4. Delaware Department of Natural Resources and Environmental Control. Sue
of Delaware Groundwaier Management Plan. 1987. 42 pp.
5. Delaware Department of Natural Resources and Environmental Control,
"Memorandum of Undemanding Between Sussex County and the Delaware
Department of NatunI Resources and Environmental Control,'' *«*fmpd Aug.
9. 1988. 3 pp.
6. Delaware Department of Natural Resource* and Environmental Control,
1' Memorandum of Agreement Between the Division of Air and Waste Mamg-
ment and the Divison of Water Resources Wildcat Landfill GrouodwMer
Management Zones," 1989, 2 pp.
7 Landon. R A.. "Waste Disposal Zoning." Presented at the NWWA Annul
Convention. Boston. MA. 1977.
& Wesson. Roy F.Inc.. "CtDundwuer Management Investigatiom brSixSuaa
County LandilU, Final Report prepared for Sussex County Delaware."
December. 1987. 250 pp.
9. US EPA. 40 CO? Pan 300 "National Priorities List for Uncontrolled
Hazardous Waste Sites Final Update No. 5," Federal Register 54. (69,
p. D304, 1989.
622 STATE PROGRAMS
-------
Implementation of Permanent Remedies in
New York State
Chittibabu Vasudevan, Ph.D., P.E.
New York State Department of Environmental Conservation
Division of Hazardous Waste Remediation
Albany, New York
ABSTRACT
The use of destruction/treatment technologies at inactive haz-
ardous waste sites has been underutilized primarily as a result of
the cost of such technologies. SARA and RCRA, which restrict
land burial, provide incentives to use treatment technologies in
remedial programs. SARA clearly gives preference to treatment
technologies "that, in whole or in part, will result in a permanent
and significant decrease in the toxicity, mobility, or volume of
hazardous substances, pollutants or contaminants," to the maxi-
mum extent practicable. The State of New York strongly sup-
ports this position. A New York State guidance document ex-
pected to be adopted in September of 1989 uses the same criteria
to evaluate and analyze remedial alternatives, proposed in the re-
vised NCP, dated Dec. 12, 1988; however, there are significant
major differences between the proposed NCP and New York
State's guidance document.
INTRODUCTION
The use of treatment technologies at Inactive Hazardous Waste
Sites has been underutilized primarily as a result of the cost of
such technologies. SARA and RCRA, which restrict land burial,
provide incentives to use treatment technologies in remedial pro-
grams. SARA requires that preference be given to remedies that
permanently reduce the toxicity, volume or mobility of the haz-
ardous substances, pollutants or contaminants, and to remedies
using alternative treatment technologies (SARA Section 121). In
addition, the 1984 amendments to RCRA restricted land disposal
of all listed hazardous wastes by 1991.
A New York State Department of Environmental Conserva-
tion (NYSDEC) guidance document which is expected to be
adopted in September 1989 uses the same criteria to evaluate and
analyze remedial alternatives proposed in the revised NCP, dated
Dec. 21,1988; however, there are significant differences between
the proposed NCP and the guidance document. This document
presents detailed guidelines for evaluation and selection of remed-
ial alternatives for some on-going and all new RI/FS projects at
Federal Superfund, State Superfund and PRP sites. NYSDEC
would consider exempting an inactive hazardous waste site from
this document if deemed appropriate. For example, if a remedial
action for a site is readily apparent, it would not be beneficial
in selecting remedies in accordance with this guidance document.
IMPLEMENTATION OF REMEDIAL ACTIONS
In order to eliminate the significant threat to public health and
the environment, NYSDEC prefers to implement permanent
remedies in accordance with SARA's preference for treatment
technologies, wherever practicable. When remedies such as con-
ventional isolation and/or control technologies are selected, the
ROD shall discuss why a remedial action resulting in a permanent
and significant reduction of the toxicity, volume or mobility of
hazardous wastes was not selected.
If a remedial action that leaves any hazardous wastes at the
site is selected, such remedial action shall be reviewed no less
than once each 5 yr after completion of the remedial action; this
review will take place in addition to the regularly scheduled mon-
itoring and operation and maintenance, even if the monitoring
data indicate that the implemented remedy does not contravene
any "cleanup criteria or standards." The objective of the review
will be to evaluate if the implemented remedy protects human
health and the environment and to identify any "permanent"
remedy available for the site. In addition, if upon such review, it
is determined that action is appropriate at such site, New York
State shall take or require such action. Before taking or requir-
ing any action, all interested parties including the responsible
parties and the public shall be provided an opportunity to com-
ment on New York State's decision.
Hierarchy of Remedial Technologies
The following provides the hierarchy of remedial technologies
for hazardous waste disposal sites, from most desirable to least
desirable. The Department shall consider only destruction or sep-
aration/treatment or solidification/chemical fixation of inorganic
wastes as permanent remedies.; However, solidification/chemical
fixation of wastes containing "low" level organic constituents
may be considered as a permanent remedy if justified.
Destruction
This type of remedy will irreversibly destroy or detoxify all or
most of the hazardous wastes to "acceptable cleanup levels." The
treated materials will have no residue containing unacceptable
levels of hazardous wastes. This type of remedy will result in
permanent reduction in the toxicity of all or most of the haz-
ardous wastes to "acceptable cleanup level(s)."
Separation/Treatment
This type of remedial action will separate or concentrate the
hazardous wastes from the wastes; this remedy would leave a
treated waste stream with acceptable levels of hazardous wastes
and a concentrated waste stream with high levels of contaminants
—e.g., treatment of contaminated leachate by granulated acti-
vated carbon. This type of remedy will result in permanent and
STATE PROGRAMS 623
-------
significant reduction in volume of hazardous wastes. In these in-
stances where the concentrated waste stream can be destroyed or
detoxified, preference shall be given to this additional treatment.
Solidification/Chemical Fixation
This type of remedy will, for a site containing predominantly
inorganic hazardous wastes, significantly reduce the mobility of
inorganic hazardous wastes. This type of remedy may not signif-
icantly reduce the toxicity or volume of the inorganic hazardous
wastes, but will significantly and permanently reduce the mobility
and hence the availability of the inorganic hazardous wastes to
environmental transport and uptake.
Control and Isolation Technologies
This type of remedial action will significantly reduce the mobil-
ity of the hazardous wastes, but will not significantly reduce the
volume or toxicity of the hazardous wastes. It also includes con-
struction of a physical barrier to control migration of leachate,
contaminated groundwater and surface runoff, solidification/fix-
ation of organic hazardous wastes and pumping and treatment of
contaminated leachate/groundwater.
In evaluating treatment technologies, NYSDEC shall give or
require that preference be given to technologies which have: (1)
been successfully demonstrated on a full-scale or a pilot-scale
under the Federal Superfund Innovative Technology Evaluation
(SITE) Program; (2) been successfully demonstrated on a full-
scale or pilot-scale at a Federal Superfund site, at a Federal facil-
ity, at a State Superfund site anywhere in the country or at a PRP
site overseen by a State environmental agency or U.S. EPA; (3)
a RCRA Part B permit; (4) a RCRA Research and Development
permit; or (5) a documented history of successful treatment, such
as a granulated activated carbon unit.
DEVELOPMENT OF REMEDIAL ALTERNATIVES
Alternatives typically are developed concurrently with the RI.
This process should consist of five general steps briefly presented
below:
• Develop remedial action objectives specifying the contaminants
and media of interest and exposure pathways. The objectives
developed are based on contaminant-specific cleanup criteria
and ARARs.
• Develop general response actions for each medium of interest
that may be taken to satisfy the remedial action objectives for
the site or specific operable unit.
• Identify volumes or areas of media to which general response
actions might be applied, taking into account the requirements
for protectiveness as identified in the remedial action objec-
tives and the chemical and geological characterization of the
site or a specific operable unit.
• Identify and screen the technologies applicable to each medium
of interest to eliminate those technologies that cannot be imple-
mented technically at the site for that medium.
• Assemble the selected representative technologies into appro-
priate remedial alternatives.
Initial set of alternatives developed shall include appropriate
remedial technologies that are representative of each of the four
categories of remedial technologies as described previously.
PRELIMINARY SCREENING OF REMEDIAL
ALTERNATIVES
The objective of remedial alternatives screening is to narrow
the list of potential alternatives that will be evaluated in detail.
Hence, alternatives will be evaluated more generally in this phase
than during the detailed analysis. In some situations, the number
of viable alternatives to address site problems may be limited
such that screening may be unnecessary or minimized. During the
screening, the extent of remedial action (e.g., quantities of media
to be affected), the sized and capacities of treatment units and
other details of each alternative should be further defined, u
necessary, to conduct screening evaluations.
Individual remedial technologies should be screened primarily
on their ability to meet medium-specific remedial action objec-
tives, their implementability and their short-term and long-term
effectiveness. At this time, cost will not be used to screen remed-
ial technologies or alternatives.
EffecthretMM Evaluation
Each alternative should be evaluated as to the extent to which it
will eliminate significant threats to public health and the environ-
ment through reductions in toxicity, mobility and volume of the
hazardous wastes at the lite. Both short-term and long-term effec-
tiveness should be evaluated; short-term referring to the construc-
tion and implementation period, and long-term referring to the
period after the remedial action is in place and effective.
The expected lifetime or duration of effectiveness should be
identified for each alternative. The control and isolation technol-
ogies may fail if any of the following is expected to take place:
(1) significant lots of the surface cover such u clay cap with i
potential for exposure of waste material underneath the cap; (2)
contamination of the groundwater by the leachate from the waste
material; (3) contamination of the adjoining surface water by the
leachate from the waste material or by the contaminated ground-
water; or (4) structural failure of the control or isolation tech-
nology.
Table 1 should be used in evaluating the effectiveness of each
alternative in protecting human health and the environment. If an
alternative is scored less than 10 out of a m««imum score of 25,
the project manager may consider rejecting that remedial altern-
ative from further consideration.
Tskfel
Skort-T«m/L««t-Tcni EffeeftreweM
(Mntau Seon - 25)
Anal/tit fader
bin <•' [valuation Dvrlei
»rel l*tft*rj Screeolnf
I. Protection of ca
oil,
SettoUl Inutoue « «)
2. tnvtrOflB»nl«l InpacM
hettotel (BulM • 4)
>. Tleo to loeleneM th«
I)
a Are there significant short-lore) risks Ten
to the coonutltr that OMt bo addressed? lo
(If ino»or It no), go to) factor I.)
e C*n the short-tore risk be easily Tes
controlled? lo .
o Oees the at! I leal I ve effort to control Tes __
short-late) risk tested tha cceomjnltj oa
life-stylo?
o Are than significant ihort-tero risks fe»
ta the envlroneent Out a»ut ba lo
addressed? (If answr Is M. go to
Factor J.)
a Ara Uie avillabla ottleaUn tMUirei To
reliable to •Inle.lie potential teowcU? lo
o Miat Is tha required tine to Inple
el < tjr-
> ».
o lecjulrod deration of UN •Kigali" < tir. __
effort U control short-tore rlsl > tJT- __
Subtotal
4. Peraenonce of the reeedlal o Kill the rened* be classified as tM.
alternative. permanent In eccoreenco wltn Section !• .
J.l(a). |b), or
-------
(1) the ability to construct, reliably operate and meet technical
specifications or criteria and (2) the availability of specific equip-
ment and technical specialist to operate necessary process units. It
also includes operation, maintenance, replacement and monitor-
ing of technical components of an alternative, if required, into
the future after the remedial action is complete. Administrative
feasibility refers to compliance with applicable rules, regulations
and statutes and the ability to obtain approvals from other offices
and agencies, the availability of treatment, storage and disposal
services and capacity.
Determination that an alternative is not technically feasible and
not available for implementation will preclude it from further
consideration unless steps can be taken to change the conditions
responsible for the determination. Often, this type of fatal flaw
would have been identified during technology development, and
an alternative which is not feasible would not have been
assembled. Remedial alternatives which will be difficult to imple-
ment administratively will not be eliminated from further con-
sideration for this reason alone.
Implementability of each remedial alternative should be eval-
uated using Table 2. If an alternative does not score a minimum
of eight out of a possible maximum score of 15, then the Project
Manager has the option of screening out this alternative from
further consideration.
Table 1 (continued)
Short-Tenn/Long-Term Effectiveness
(Maximum Score = 25)
Analysis Factor
Basis for Evaluation During
Preliminary Screening
5. Lifetime of remedial
actions.
Subtotal (maximum - 4)
6. Quantity and nature of
waste or residual left
at the site after
remediation.
o Expected lifetime or duration of
of effectiveness of the remedy.
1) Quantity of untreated hazardous
waste left at the site.
25-30yr.
20-25yr.
15-ZOyr.
< ISyr.
None
< 25%
25-501
> 50*
4
3
2
0
3
2
1
0
ii) Is there treated residual left at
the site? (If answer Is no. go to
Factor 7.)
iii) Is the treated residual toxic?
iv) Is the treated residual mobile?
Yes
Ho
Yes
No
Yes
No
Subtotal (maximum « 5)
7. Adequacy and reliability
of controls.
i) Operation and maintenance required < Syr. 1
for a period of: > 5yr. 0
ii) Are environmental controls required Yes 0
as a part of the remedy to handle No 2
potential problems? (If answer 1s
no. go to "1v")
iii) Degree of confidence that controls Moderate to very
can adequately handle potential confident 1
problems. Somewhat to not
confident 0
iv) Relative degree of long-term Minimum 2
monitoring required (compare with Moderate 1
othor remedial alternatives Extensive 0
evaluated 1n the Detailed Analysis).
Subtotal (maximum = 5)
TOTAL (maximum = 25)
IF THE TOTAL SCORE IS LESS THAN 10. PROJECT MAI1AGER MAY REJECT THE REMEDIAL ALTERNATIVE FROM
FURTHER CONSIDERATION.
DETAILED ANALYSIS OF ALTERNATIVES
The detailed analysis of alternatives follows the development
and preliminary screening of alternatives and precedes the actual
selection of a remedy. During this phase, remedial alternatives are
analyzed in detail and relevant information is presented to allow
decision-makers to select a remedy. The evaluations conducted
during the detailed analysis phase build on previous evaluations
conducting during the development and preliminary screening of
alternatives. This phase also incorporates any treatability study
data and additional site characterization information that may
have been collected during the RI.
Table 2
Implementabillty.
(Maximum Score = 15)
Analysis Factor
Basis for Evaluation During
Preliminary Screening
1. Technical Feasibility
a. Ability to construct
technology.
b. Reliability of
technology.
c. Schedule of delays
due to technical
problems.
d. Need of undertaking
additional remedial
action, If necessary.
Subtotal (B.UI.U = 10)
2. Administrative Feasibility
a. Coordination with
other agencies.
Subtotal (mud.**! = 2)
3. Availability of Services
and Materials
a. Availability of
prospective
technologies.
b. Availability of
necessary equipment
and specialists.
Subtotal (maximum = 3)
TOTAL (maximum = 15)
1) Not difficult to construct.
ii) Somewhat difficult to construct.
No uncertainties In construction.
ill) Very difficult to construct and/or
significant uncertainties in construction.
i) Very reliable in meeting the specified
process efficiencies or performance goals.
11) Somewhat reliable in meeting the specified
process efficiencies or performance goals.
1) Unlikely
ii) Somewhat likely
1) No future remedial actions nay be
anticipated.
ii) Some future remedial actions may be
necessary.
I
1) Minimal coordination is required.
ii) Required coordination is normal.
iii) Extensive coordination is required.
1) Are technologies under consideration Yes
generally commercially available No
for the site-specific application?
ii) Will more than one vendor be available Yes
to provide a competitive bid? No
i) Additional equipment and specialists Yes
may be available without significant No
delay.
3
2
1
3
2
2
1
2
1
Z
1
0
1
0
1
0
1
0
IF THE TOTAL IS LESS THAU 8. PROJECT MANAGER KAY REJECT THE REMEDIAL ALTERNATIVE FROM
FURTHER CONSIDERATION.
Detailed Analysis of Remedial Alternatives
During the detailed analysis, each alternative is assessed against
the seven evaluation criteria. The seven evaluation criteria listed
encompass technical, cost and institutional considerations and
compliance with specific statutory requirements. The seven cri-
teria and their relative weights are presented in Table 3. Each eval-
uation criterion has been further divided into specific factors to
allow a thorough analysis of the alternatives.
Table3
Criteria for Detailed Analysis of Remedial Alternatives.
Criteria Weight
1. Short-term effectiveness 10
2. Long-term effectiveness and performance 15
3. Reduction of toxicity, mobility and volume 15
4. Implementability 15
5. Compliance with ARARs 10
6. Protection of human health and the environment 20
7. Cost 15
TOTAL
100
STATE PROGRAMS 625
-------
Short-Term Effectiveness (Relative Weight: 10)
This evaluation criterion assesses the effects of the alternatives
on human health and the environment during implementation of
the remedial action. The following factors of this criterion should
be addressed for each alternative; (1) Protection of the commun-
ity during remedial actions—This aspect of short-term effective-
ness addresses any risk that results from implementation of the
proposed remedial action, such as dust from excavation or air-
quality impacts from the operation of an incinerator; (2) En-
vironmental impacts—This factor addresses the potential adverse
environmental impacts that may result from the implementation
of an alternative and evaluates how effectively available mitiga-
tion measures would prevent or reduce the impacts; (3) Time
until remedial response objectives are achieved—This factor in-
cludes an estimate of the time required to achieve protection for
either the entire site or individual elements associated with spe-
cific site areas or threats; and (4) Protection of workers during
remedial actions—This factor assesses threats that may be posed
to workers and the effectiveness and reliability of protective
measures that could be taken.
The score for this criterion should be assigned based on the
analysis of factors (1), (2), (3) presented in Table 4. Analysis of
the factor "protection of workers during remedial actions,"
should be used to design appropriate safety measures for on-site
workers.
T»bk4
Sbort-Ttnn Effect.vent**.
(Relative Wdibl - 10)
Ani1ri>i Factor Milt for [valuation During »•!»*
1. »r«Ur. 0
e teojulrod duration of tha olltgaU»o < tyr 1
effort to control ahort-ter. rlrt S Zjrr 0
S*total (eulei
IOTA1 (eaulemal
taking into account their toxicity, mobility and propensity to bio-
accumulate; (3) Adequacy of controls — This factor assesses the
adequacy and suitability of control*, if any, that are used to
manage treatment residuals or untreated wastes that remain at the
site. It may include an assessment of containment systems and
institutional controls to determine if they are sufficient to ensure
that any exposure to human and environmental receptors is with-
in protective levels; and (4) Reliability of controls—This factor
assesses the long-term reliability of management controls for pro-
viding continued protection from residuals. It includes theasseu-
ment of the potential need to replace components of the alterna-
tive, such as a cap, a slurry wall or a treatment system; the poten-
tial exposure pathway; and the risks posed should the remedial
action need replacement. This factor should also include systems
to warn of the failure of a remedial alternative, once in place.
Table 5 should be used during the analysis to assign a score for
this criterion.
TibkS
LMf-T«fo* EffccttrceWM ud
(R«tmU»t Wdtfct . 15)
Anetjrtlt factor
lotlt for (valuation Ourtnf
Detallee*
Per»en«« nature of
•ana or reiloatll left
el the file «ft«r
reeanllatlcpn
*lTT the reawjy be cloBtlfioo: u
porvaMont to accordance wttA lection
.' tin ft), rr |c) (If a«p»or li
•M. fa to Foctaw 1.)
llfetie. or torettoo of
of of fa** Ivomeit of the roaaoey
Br-ZSjT.
li-Hr/r
< Or.
I) Quantity of MfftreatH h
*ei't toft •( UM alu.
It) It there trooteel rntoWal lift It
the lite? (I' e«e»»r le »e. re U
Ftctaw «.)
Ill; it UM tre«le« reatexttl t*a*cT
'. It lh« IrMtoe reileiKll labile'
luttetal (BHiteaao • 1)
Adequacy end reliability I) rv^rit'oo) one! OialoteAenco ri
of contrela for e »orlo4 of
at I port of tho r-OBoely t* I
xtaotlel arooleeo? (If UB
HI) 0* «4»a»*totj r^ndlo polofttlol
I.] lilitln «o|i M of 1o«(-tor>
•xilurinf r««>lro4 (capon «1t»
oUior rvMU) ilUnHtlm
..•luUd lo tto O.U11«1 A«*lnl«)
ion
< Bl
Z5-MI
To
lo
> J»r. I
T« I
b J
IS)
(amilaa • t)
TOTJU (Kiln. . D)
Long- Term Effectiveness and
Permanence (Relative Weight: 15)
This evaluation criterion addresses the results of a remedial
action in terms of its permanence and the quantity/nature of
waste or residual remaining at the site after response objectives
have been met. The primary focus of this evaluation is the extent
and effectiveness of the controls that may be required to manage
the waste or residual remaining at the site and operating system
necessary for the remedy to remain effective. The following com-
ponents of the criterion should be addressed for each alternative:
(1) Permanence of the remedial alternative; (2) Magnitude of re-
maining risk—The potential remaining risk may be expressed
quantitatively such as by cancer risk levels, margins of safety
over NOELs for non-carcinogenic effects or the volume or con-
centration of contaminants in waste, media or treatment residuals
remaining at the site. The characteristics of the residuals that
should be considered to the degree that they remain hazardous,
Reduction of Toxicity, Mobility and Volume
(Relative Weight: 15)
This evaluation would focus on the following specific factors
for a particular remedial alternative: (1) The amount of haz-
ardous materials that will be destroyed or treated, including how
the principal threat(s) will be addressed; (2) The degree of
expected reduction in toxicity, mobility or volume measured as >
percentage of reduction (or order of magnitude); (3) The degree
to which the treatment will be irreversible; and (4) The type and
quantity of treatment residuals that will remain following treat-
ment.
Table 6 lists factors to be addressed during the analysis of toxic-
ity, mobility or volume reduction.
Implementability (Relative Weight: IS)
Of the total weight of IS, the technical feasibility shall receive
626 SI ATI. PROGRAMS
-------
a maximum score of 10 while administrative feasibility and avail-
ability of services and materials shall be assigned a combined
maximum score of 5.
Table 6
Redaction of Toriclty, Mobility or Volume.
(Relative Weight = IS)
Analysis Factor
Basis for Evaluation During
Detailed Analysis
Height
1. VoluM of hazardous
waste reduced (Reduction
In volune or toxlclty).
1) Quantity of hazardous waste destroyed
or treated.
11) Are there concentrated hazardous waste
produced as a result of (1)7
(If answer 1> no. go to Factor 2.)
HI) Now 1s the concentrated hazardous
waste streaai disposed?
BO-9M
60-80%
40-60%
20-401
< 2M
Itt
No
Subtotal (maximum = 12)
(If subtotal = 12, go to :
2. Reduction In nobility of
hazardous waste.
Subtotal (maximum = 5)
3. Irreverslbtllty of the
destruction or treatment
of hazardous waste.
Subtotal (maximum = 3)
TOTAL (maximum = IS)
1) Method of Reduction
- Reduced mobility by containment
- Reduced mobility by alternative
treatment technologies.
11) Quantity of Hastes Immobilized
Completely Irreversible
Irreversible for most of tile hazardous
waste constituents.
Irreversible for only SOM of tht
hazardous waste constituents
Reversible for M>st of the hazardous
waste constituents.
On-lite land
disposal 0
Off-site
secure land
disposal 1
On-3lte> or off-
stte destruction
or treatment
2
< 100%.
5 60%
? 60*-
Technical feasibility: This criterion relates to the technical
difficulties and unknowns associated with a technology. This was
initially identified for specific technologies during the develop-
ment and preliminary screening of alternatives and is addressed
again in the detailed analysis for the alternative as a whole.
• Reliability of technology—This criterion focuses on the ability
of a technology to meet specified process efficiencies or per-
formance goals. The likelihood that technical problems will
lead to schedule delays should be considered as well.
• Ease of undertaking additional remedial action—This criterion
includes a discussion of what, if any, future remedial actions
may need to be undertaken and how difficult it would be to im-
plement such additional actions. This is particularly applicable
for a FS addressing an interim action at a site where additional
operable units may be analyzed at a later time.
• Monitoring considerations—This criterion addresses the ability
to monitor the effectiveness of the remedy and includes an eval-
uation of the risks of exposure should monitoring be insufficient
to detect a system failure.
Administrative feasibility: Activities needed to coordinate with
other offices and agencies (e.g., obtaining permits for off-site
activities or rights-of-way for construction) should be evaluated
for the alternative.
Availability of services and materials: The following should be
considered during the analysis: (1) Availability of adequate off-
site treatment, storage capacity and disposal services; (2) Avail-
ability of necessary equipment, specialists and skilled operators
and provisions to ensure any necessary additional resources; and
(3) Availability of services and materials, plus the potential for
obtaining competitive bids, which may be particularly important
for alternative remedial technologies.
Table 7 lists typical factors to be addressed during the analysis
of the implementability criterion.
Table 7
Implementabllity.
(Relative Weight = 15)
Analysis Factor
Basis for Evaluation During
Detailed Analysis
Weight
1. Technical Feasibility
• .
b.
t.
d.
Ability to construct
technology.
Reliability of
technology.
Schedule of delays
due to technical
problems.
Need of undertaking
additional remedial
action, If necessary.
1) Not difficult to construct.
No uncertainties in construction.
11) Somewhat difficult to construct.
No uncertainties 1n construction.
111) Very difficult to construct and/or
significant uncertainties In construction.
1) Very reliable 1n meeting the specified
process efficiencies or performance goals.
11) Somewhat reliable in meeting the specified
process efficiencies or performance goals.
1) Unlikely
11) Somewhat likely
1) No future remedial actions may be
anticipated.
3
2
1
3
2
2
1
2
11) SO.M future remedial actions nay be
necessary.
Subtotal (Bui...* - 10) NlnlM Required Score = 7
2. Administrative Feasibility
1) MiniMl coordination Is required.
11) Required coordination Is normal.
Ill) Extensive coordination Is required.
. Coordination with
other agencies.
Subtotal (maximum = 2)
3. Availability of Services
Availability I
and Materials
1) Are technologies under consideration
generally commercially available
for the site-specific application?
11) Kill more than one vendor be available
to provide a competitive bid?
lit) Additional equipment and specialists
may be available without significant
delay.
Yes
No
Yes
No
Yes
Ho
Subtotal (maximum = 3)
TOTAL (maximum = 15)
Compliance with ARARs (Relative Weight: 10)
This evaluation criterion is used to determine how each alterna-
tive complies with applicable or relevant and appropriate Fed-
eral and State requirements as defined in CERCLA Section 121.
There are three general categories of ARARs: chemical-, loca-
tion- and action-specific. ARARs for each category are iden-
tified in previous stages of the RI/FS process (e.g., chemical-
specific ARARs should be preliminarily identified during scoping
of the project). The detailed analysis should summarize which
requirements are applicable or relevant and appropriate to an
alternative and describe how the alternative meets these require-
ments. When an ARAR is not met, justification for use of one of
the six waivers allowed under CERCLA and SARA should be dis-
cussed.
Other information in the form of advisories, criteria and guid-
ance that are not ARARs may be available. Compliance with
such guidance may be necessary to ensure protectiveness and may
be appropriate for use in the evaluation of a specific alternative.
If an alternative complies with all ARARs, it should be assigned
a full score of 10. If an alternative complies with none of the
above-mentioned four specific aspects of the ARARs, it should
receive a score of 0. Each component of the four specific aspects
of the ARARs shall receive a maximum score of 2.5. If an alterna-
tive does not meet the ARARs and a waiver to the ARARs is not
appropriate or justifiable, such an alternative will not be further
STATE PROGRAMS 627
-------
considered. Table 8 should be used to evaluate remedial alterna-
tives.
Table!
Compliance With ARARj.
(Relative Weight - 10)
Analvilt factor
»li for (valuation During
Dalallod Anatvitl
I Ccaapllanca Kith chaailtal- Noati cha»l» ailnli
ipoclflc ARAltl. tochnolooj llandardt
3. Coavllanca wHh location- Roan l 4
0
I J
0
Protection of Human Health and the Environment
(Relative Weight: 20)
This evaluation criterion provides a final check to assess
whether each alternative meets the requirement that it is protec-
tive of human health and the environment. The overall assess-
ment of protection is based on a composite of factors assessed
under other evaluation criteria, especially long-term effectiveness
and performance, short-term effectiveness and compliance with
ARARs.
Evaluation of ihe overall protectiveness of an alternative dur-
ing the RI/FS should focus on how a specific alternative achieves
protection over time and how site risks are reduced. The analysis
should indicate how each source of contamination is to be elim-
inated, reduced or controlled for each alternative.
Table 9 outlines pertinent questions to be answered to help the
evaluator assign relative weighing scores to remedial alternatives.
Table 9
Protection of Human Health and the Environment.
(Relative Wdjbl - 20)
t'l for [valuation During
Dvtallod Analvtlt
Vvlotit
1. UM of UM til* afur
raawdtallon.
TOTAL (Itulaw > 20)
2. Huu« hoalUi and UM
tnvtronawnt Mpotura
aftar tha raa»»dlatton
Sufctottl (amilau > 10)
1 HagnlUxla of roalouil
public haalth rlika
aftar tho rdaaodUt Ion.
Subtotal (a»jlam. • 5)
4. MainlUido of r.ildual
onvlrorvontal rlika
aftor tha raaBtcHatton.
Subtotal (amitaia. • i)
TOTAL (akulau • 20)
UnraitrUlod |>U of tha land and
•ator (If anaaaar la jroa, 90 to
UM and of UM Tibia )
I) \\ tha axpoiura to contaailnantl
via air rowto accaptabla)
II) l> Uia ..oo>ura U tontaa>tnant>
via oroundMator/iurraca Katcr
actoptabI«T
lit) If tM cipotura to contaalnanlt
via l.oi«anli/»olli .naptabl.'
I) HoalUi rut
II) Haalth rMk
i I In
< 1 In
1,000.000 .
100,000
I) 1.ii than accapubla
U) Slightly graalar than accaplabla
III) Significant Mil HIM ailata
20
0
Cost (Relative Weight: 15)
The application of cost estimates to evaluation of alternatives
should include capital costs, operation and maintenance costs,
future capital costs and costs of future land use. The U.S. EPA'j
guidance for conducting RI/FS under CERCLA4 may be referred
to for detailed descriptions of cost elements such as capital costs,
operation and maintenance costs and cost-sensitivity analysis.
Capital Costs. Capital costs consist of direct (construction)
and indirect (non-construction and overhead) costs. Direct costs
include expenditures for the equipment, labor and material!
necessary to install remedial actions. Indirect costs include ex-
penditures for engineering and other services that are not pan of
actual installation activities but are required to complete the
installation of remedial alternatives.
Operation and Maintenance Costs. Annual costs are post-con-
struction costs (such as operating labor costs, maintenance
materials and labor costs) necessary to ensure the continued
effectiveness of a remedial action.
Future Capital Costs: The costs of potential future remedial
actions should be addressed and, if appropriate, should be in-
cluded when there is a reasonable expectation that a major com-
ponent of the remedial alternative will fail and require replace-
ment to prevent significant exposure to contaminants. It is not
expected that a detailed statistical analysis will be required to
identify probable future costs. Rather, qualitative engineering
judgment should be used, and the rationale should be well docu-
mented in the FS report.
Cost of Future Land Use. Any remedial action that leaves haz-
ardous wastes at a site may affect future land use and perhaps
groundwater use. Access or use of such sites will be restricted,
resulting in loss of business activities, residential development and
taxes to the local, state and federal governments. During the
feasibility study, potential future land use of the site should be
considered. Based on this potential land use, economic loss attrib-
utable to such use should be calculated and included as a cost of
the remedial alternative. In addition, the continuing presence of
an inactive hazardous waste site, even though remediated, may
have a negative effect on surrounding property values. This loss
in value should be considered as a cost of the remedial program
developed for the site. Economic loss due to the future land use
should be derived based on comparison with a neighboring com-
munity not affected by any hazardous waste sites.
Cost of future land use should be determined for sites only
when such cost is deemed appropriate and significant. When cost
of land surrounding an inactive hazardous waste site located in
the urban/suburban area is determined to be significant in rela-
tion to the cost of a remedial alternative, then cost of future land
use as described above should be determined for inclusion in the
present worth analysis of the remedial alternative.
Accuracy of Cost Estimates. Site characterization and treat-
ability investigation information should permit the user to refine
cost estimates for remedial action alternatives. It is important to
consider the accuracy of costs developed for alternatives in the
FS. Typically, these study estimate costs made during the FS are
expected to provide an accuracy of +50 to -30^ should be
identified as such in the FS.
Present Worth Analysis. A present worth analysis is used to
evaluate expenditures that occur over different time periods by
discounting all future costs to a common base year, usually the
current year. This figure allows the cost of remedial action
alternatives to be compared on the basis of a single figure repre-
senting the amount of money that, if invested in the base year and
disbursed as needed, would be sufficient to cover all costs asso-
ciated with the remedial action over its planned life.
In conducting the present worth analysis, assumptions must be
made regarding the discount rate and the period of performance.
NYSDEC recommends that a discount rate equivalent to the 30-
yr U.S. Treasury bond rate before taxes be used; this discount rate
should take inflation into account. In general, the period of per-
formance for costing purposes should not exceed 30 yr.
An alternative with the lowest present worth shall be assigned
628 STA'II. PROGRAMS
-------
the highest score of 15. Other alternatives shall be assigned the
cost score inversely proportional to their present worth.
Presentation of Comparative Analysis of Alternatives
Once the alternatives have been individually assessed against
the seven criteria, a comparative analysis should be conducted to
evaluate the relative performance of each alternative in relation to
each specific evaluation criterion. The purpose of this compara-
tive analysis is to identify the advantages and disadvantages of
each alternative relative to the others so that the key trade-offs to
be evaluated by the decision-maker can be identified.
The first five criteria (short-term effectiveness; long-term effec-
tiveness and permanence; reduction of toxicity, mobility and
volume; implementability; and cost) will generally require more
discussion than the remaining criteria because the key trade-offs
or concerns among alternatives will most frequently relate to one
or more of these five. The overall protectiveness and compliance
with ARARs criteria will generally serve as threshold determina-
tions in that they either will or will not be met.
The comparative analysis should include a narrative discussion
describing the strengths and weaknesses of the alternatives rela-
tive to one another with respect to each criterion, and how rea-
sonable variations of key uncertainties could change the expecta-
tions of their relative performance. If destruction and treatment
technologies are being considered, their potential advantages in
cost or performance and the degree of uncertainty in their ex-
pected performance (as compared with conventional/isolation
technologies) should also be discussed.
The presentation of differences between alternatives can be
measured either qualitatively or quantitatively, as appropriate,
and should identify substantive differences (e.g., greater short-
term effectiveness concerns, greater cost, etc.) between alterna-
tives, differences in total scores, etc. Quantitative information
that was used to assess the alternatives (e.g., specific cost esti-
mates, time until response objectives would be obtained and levels
of residual contamination) should be included in these discus-
sions.
The Final Draft RI/FS or the Proposed Remedial Action Plan
(PRAP) should present the remedial alternative recommended for
the site and a clear rationale for the recommendation.
COMMUNITY ASSESSMENT
The community assessment incorporates public comment into
the selection of a remedy. There are several points in the RI/FS
process at which the public may have previously provided com-
ments (e.g., first phase of the RI/FS). The Department will solicit
public comments on the remedial alternatives and the recom-
mended alternative in accordance with the New York State Inac-
tive Hazardous Waste Site Citizen Participation Plan and statu-
tory and regulatory requirements. A document titled, "New
York State Inactive Hazardous Waste Site Citizen Participation
Plan," dated Aug. 30, 1988, should be used as a guide to solicit
the public comments on the remedial alternatives and the recom-
mended alternative at New York State inactive hazardous waste
sites. The public comments shall be considered. The remedy for
the site will be selected and documented in accordance with the
Organization and Delegation Memorandum #89-05 Policy-
Records of Decision for Remediation of Class 2 Inactive Haz-
ardous Waste Disposal Sites.
SARA's preference for the selection of remedies that permanent-
ly reduce the toxicity, volume or mobility of the hazardous
wastes.
The New York State guidance document identifies a hierarchy
of preferred remedies toward meeting the State's goal of imple-
menting remedial actions which would result in a permanent and
significant decrease in the toxicity, mobility or volume of haz-
ardous wastes. This hierarchy is consistent with the New York
State policy in hazardous waste management, SARA and RCRA
land disposal restrictions.
The guidance document does not consider cost of remedial
alternatives in initial screening of the remedial alternatives. Effec-
tiveness and implementability are the only criteria used to screen
the remedial technologies.
The New York State guidance document assigns numerical
weighting factors for criteria to ensure objectivity, uniformity and
consistency in the initial screening and detailed analysis of remed-
ial alternatives; this document also outlines guidelines to be used
in assigning weighting factors for evaluation criteria.
In addition to capital cost and operation and maintenance cost,
the present worth analysis of a remedial alternative includes: (1)
cost of future capital cost when there is a reasonable expectation
that a major component of a remedial alternative will fail, and
(2) cost of future land use when such cost is deemed appropriate
and significant.
The U.S. EPA has proposed to use the Office of Management
and Budget's circular A-94 discount rate of 10% when deter-
mining the present value of remedial alternatives. If the discount
rate is high, the cost of operation and maintenance in the future
will appear low when evaluating costs, thus favoring remedies
which have low initial capital costs. Permanent treatments and
remedies often will have high initial capital costs but lower oper-
ating and maintenance costs in the future than less-permanent
remedies which will need longer operation or continual oversight.
Therefore, the use of a high discount rate may result in an unfair
bias against permanent or treatment remedies.
The New York State guidance document recommends the use
of 30-yr U.S. Treasury bond rates before taxes as the discount
rate; this discount rate should also take inflation into account.
CONCLUSION
The New York State guidance document assigns weights to
evaluation criteria to ensure objectivity, uniformity and consis-
tency in initial screening and detailed analysis of remedial alterna-
tives; this process would facilitate implementing permanent rem-
edies. Deletion of the cost criterion in initial screening of remed-
ial alternatives would help to carry permanent remedies over to
the next phase of detailed analysis. Inclusion of the cost of future
land use and a discount rate of a 30-yr U.S. Treasury bond rate
(instead of 10%) would eliminate unfair bias against permanent
remedies.
The draft guidance document was distributed to the public and
other interested state and federal agencies for review and com-
ment. The guidance document was revised to reflect public review
and comments.
ACKNOWLEDGEMENTS
DIFFERENCES BETWEEN PROPOSED NCP AND
NEW YORK STATE GUIDANCE DOCUMENT
The following are major differences between the proposed
NCP and New York State guidance document in appb'cation of
The author would like to express his appreciation to Charles
N. Goddard, P.E., Assistant Director, Division of Hazardous
Waste Remediation, New York State Department of Environ-
mental Conservation, for his valuable guidance in the develop-
ment of the New York State guidance document.
STATE PROGRAMS 629
-------
REFERENCES
1. Comprehensive Environmental Response, Compensation, and Liabil-
ity Act of 1980. Public Law 96-510, Dec. 1980.
2. Superfund Amendments and Reauthorization Act of 1986. Public Law
99-499, Oct. 17,1986.
3. National Oil and Hazardous Substances Pollution Contingency Plan;
Proposed Rule. Federal Register. 53 (245), pp. 51394 51520, Dec. 21,
1988.
4. Environmental Protection Agency, Outdance for Conducting Re-
medial Investigations and Feasibility Studies Under CEKCLA, EPA/
540/0-89/004. Oct. 1988.
5. New York State Department of Environmental Conservation, Tech-
nical and Administrative Guidance Memorandum for the Selection
of Remedial Actions at Inactive Hazardous Wasit Sites, Sept. 1989.
6. New York State Department of Environmental Conservation, New
York Slate Inactive Hazardous Waste Site Citizen Participation
Plan. Aug. 30. 1988.
630 STATI-; PROGRAMS
-------
U.S. EPA'S Federal Facility Hazardous Waste Compliance Program
Gordon M. Davidson
Deborah K. Wood
U.S. EPA
Washington, D.C.
ABSTRACT
Ensuring compliance by Federal facilities with hazardous waste
requirements is a challenging task because of the number and size of
Federal facilities, the types and sources of contamination involved and
the potential for overlap of jurisdictional and/or statutory authority.
Although Federal facilities must comply with RCRA and CERCLA
requirements similar to private parties, there are some differences. For
example, Federal agencies are delegated certain CERCLA response
authorities by executive order. Also, under both RCRA and CERC-
LA, dispute resolution procedures differ.
RCRA-regulated activities occur at most Federal facilities that require
CERCLA action, and the U.S. EPA is developing a mechanism for
creating a coordinated response between statutory and jurisdictional
authorities. For Federal facilities, the integration solution may be a site-
specific three-party CERCLA interagency agreement with the state, the
U.S. EPA and the Federal facility as parties. This paper provides a statu-
tory overview of requirements that apply to Federal facilities. It out-
lines two issues that are unique to the executive branch and discusses
RCRA/CERCLA integration.
INTRODUCTION
At the U.S. EPA, the Office of Waste Programs Enforcement (OWPE),
within the Office of Solid Waste and Emergency Response, is respon-
sible for ensuring compliance by Federal facilities with RCRA and
CERCLA requirements. In 1987, OWPE established a Federal Facility
Task Force to focus dedicated resources on achievement of Federal
facility compliance. The task force now has a permanent role within
OWPE and has been renamed the Federal Facilities Hazardous Waste
Compliance Office (FFHWCO).
The primary goals of FFHWCO are to assist U.S. EPA regions to
reach CERCLA cleanup agreements at NPL sites and ensure compliance
with RCRA in a nationally consistent manner. FFHWCO develops
guidance and policy for Federal facility compliance, assists in resolu-
tion of issues that arise in negotiations with Federal facilities, tracks
ongoing negotiations and supports enforcement actions.
Over 1,100 Federal facilities that will require investigation and pos-
sible remediation under CERCLA have been identified. These facilities
range in size from hundreds of acres to tens of thousands of acres, and
many contain multiple contaminated areas.
Federal facilities that require investigation are those that manage
hazardous waste or may have potential hazardous waste problems. The
Departments of Defense (DOD), Interior (DOI) and Energy (DOE)
account for 84% of the Federal sites that require investigation.
Hazardous waste contamination at Federal facilities may result from
such activities as weapon manufacturing, testing, loading and packaging;
aircraft and vehicle maintenance and repair; metal plating; and
producing, processing and recovering nuclear materials. Types of
hazardous waste disposed of include explosives, solvents and cleaning
agents, paints, heavy metals, pesticides, waste oil and various organics.
At DOE facilities, disposal of high and low level radioactive and mixed
hazardous and radioactive waste is a common problem. Past disposal
practices at Federal facilities include disposal in unlined pits, drainage
ditches, holding ponds, drying beds and landfills; discharge on the
ground; and burning.
The number of Federal facilities to be investigated, their sizes, and
the types and sources of contamination involved combine to create the
challenge of ensuring compliance by Federal facilities with hazardous
waste laws. This challenge is heightened by the potential at each site
for overlapping jurisdiction both among federal programs and between
states that are authorized for the RCRA base or HSWA programs (i.e.,
the 1984 RCRA amendments, called the Hazardous and Solid Waste
Amendments) and the federal CERCLA or HSWA programs. There
is also a potential overlap with other Federal laws, such as the Atomic
Energy Act, and with other state and local hazardous waste-related
authorities.
STATUTORY OVERVIEW
Federal facilities must comply with the requirements of RCRA and
CERCLA. This section contains an overview of those requirements.
RCRA
Section 6001 of RCRA expressly subjects Federal facilities to RCRA
provisions and implementing regulations, including requirements for
permits, corrective action and reporting. Federal treatment, storage and
disposal facilities that handle hazardous waste must have RCRA permits
and must address hazardous waste releases.
There are approximately 336 Federal facilities that treat, store or
dispose of hazardous waste. Eighty of these facilities are land disposal
facilities and 256 are treatment and/or storage facilities. This total
represents less than 7% of the universe of RCRA treatment, storage
and disposal facilities in the country.
The U.S. EPA or an authorized state conducts an annual inspection
at all RCRA-regulated Federal facilities, as required by section 3007(c).
As of December, 1988, 46 Federal land disposal facilities were out of
compliance. Compliance mechanisms are discussed in the section
following this statutory overview.
CERCLA
CERCLA devotes a special section to Federal facilities, section 120,
enacted in the 1986 Superfund amendments. Section 120 (a) states that
Federal departments, agencies and instrumentalities are subject to
CERCLA like non-governmental entities, includiing CERCLA's liability
RCRA / SUPERFUND ACTIVITIES 631
-------
provisions. Pertinent guidelines, rules, regulations and criteria apply
in the same manner and to the same extent, with the exception of
requirements on bonding, insurance and financial responsibility.
Section 120 establishes special requirements and timetables regarding
Federal facilities. For example, section 120 (c) requires establishment
by the U.S. EPA of a Federal Agency Hazardous Waste Compliance
Docket that lists Federal facilities which manage hazardous waste or
may have potential hazardous waste problems. The docket identifies
the universe of Federal facilities to be evaluated for possible contami-
nation by compiling information submitted under RCRA and CERCLA.
The docket is updated every 6 mo. and includes;
• Information on releases of reportable quantities of hazardous sub-
stances under section 102 of CERCLA
• Information submitted to obtain a permit under section 3005 of RCRA
• Information submitted under section 3010 of RCRA from genera-
tors, transporters, owners and operators involved with waste desig-
nated as hazardous under RCRA
• Information submitted for the inventory of Federal agency hazardous
waste facilities that is compiled every 2 yr, under section 3016 of
RCRA
The docket is available for public inspection at U.S. EPA regional
offices. Each regional docket contains the documents submitted under
the reporting provisions described above, and any relevant corespon-
dence. for each facility in that region. A complete national index is
maintained at US. EPA headquarters. The docket was first published
on Feb. 12. 1988 in the Federal Register 153 p 4280 with 1,095 facili-
ties on the list. The first update was published on Nov. 16, 1988 (53
Federal Register 46364); this list contained 1,170 facilities.
Once a Federal facility is listed on the docket, the Federal facility
must conduct a preliminary assessment (PA) and, if necessary, a site
inspection (SI) within 18 mo. The statute requires the U.S. EPA to assure
that PAs are conducted, while the authority to conduct PAs is delegated
to Federal agencies by executive order 12580. As of August. 1988, 987
of the 1.095 facilities listed in the original docket submitted PA infor-
mation to the U.S. EPA. The U.S. EPA currently is reviewing this in-
formation for completeness and to determine whether further action
is required.
Following the PA, the U.S. EPA applies the hazard ranking system
(MRS) and includes sites that score 28.50 or above on the NPL. Inclusion
on the NPL does not mean Superfund monies are available for cleanup
as is the case with non-Federal sites; section lll(e) (3) specifies that
the Fund is not available for remedial actions at Federal facilities (except
for providing alternative water supplies where groundwater contami-
nation is outside the Federal facility boundaries and the Federal facility
is not the only potentially responsible party involved). Still, NPL listing
of Federal facilities serves the purpose of alerting the public and pro-
viding information concerning risks to public health or the environ-
ment from the site. In addition, NPL listing assists Federal agencies
to set cleanup priorities. There are currently 41 Federal facilities listed
on the NPL and 74 proposed for inclusion.
If a Federal facility is included on the NPL, section 120(e) of man-
dates that it begin an RI/FS. in consultation with the U.S. EPA and
the state, within 6 mo. of listing. The U.S. EPA and the state must pub-
lish an enforceable timetable and deadlines for Rl/FS completion, and
the U.S. EPA must review the RI/FS when completed.
Section 120(e) also requires the Federal facility to enter into an inter-
agency agreement (IAG) with the U.S. EPA for the remedial action
within 180 days of the U.S. EPA's review of the RI/FS. An IAG is the
vehicle for remedy selection. At a minimum, the IAG must include
a review of cleanup alternatives considered and the remedy selected,
a schedule for cleanup accomplishment and arrangements for opera-
tion and maintenance.
U.S. EPA policy, reflected in the model lAGs developed with DOD
and DOE. is to enter into an IAG at the RI/FS stage. This procedure
meets the requirements of an RI/FS start and a published timetable and
deadlines and provides for early input by the U.S. EPA and the state
to the RI/FS and remedy selection process. U.S. EPA policy is to have
three-party lAGs with the state joining the U.S. EPA and the Federal
facility as an active partner and signatory. lAGs are enforceable by the
parties to the agreement and by citizens and states using CERCLA Sec-
tion 310 authority.
Section 120(e) requires cleanup to begin at a Federal facility no later
than 15 mo. after RI/FS completion. In their annual budget submis-
sions, Federal agencies must include a review of alternative funding
that might be used to provide for cleanup costs. The annual budget sub-
mission also has to include a statement on the hazards posed to public
health, welfare and the environment, and the consequences of failure
to begin and complete remedial action. In addition, an annual report
to Congress must be submitted by each Federal agency participating
in the CERCLA program. This report must describe the Federal agency's
progress in such areas as reaching lAGs and conducting RI/FSs and
cleanups.
To facilitate the negotiation of site-specific LAGs, the U.S. EPA
developed model lAGs with DOD and DOE in 1988. The models cover
the following areas:
• Jurisdiction
• Purpose
• Statutory Compliance/RCRA-CERCLA Integration
• Consultation with U.S. EPA
• Dispute Resolution
• Enforccabilily
• Stipulated Penalties
• Extensions
• Force Majeure
• Funding
The models are captioncd as CERCLA section 120 agreements and
are designed to apply at NPL sites where CERCLA is the lead response
authority. Compliance with substantive RCRA requirements as appli-
cable, relevant or appropriate requirements (ARARs) is assured through
section 121 of CERCLA and the model's statutory compliance section.
For installations that include both NPL sites and RCRA units, language
in the jurisdiction section that cites RCRA authorities may be used.
Although these model agreements do not reflect stale involvement (be-
cause it was impossible to have SO state representatives at the negotiating
table), it is the U.S. EPA's policy to integrate state participation into
the IAG provisions at site-specific negotiations.
The consultation section establishes the procedures for U.S. EB\ and
state review of documents. Documents designated as primary, including
discrete portions of RI/FS and remedial design and remedial action
(RD/RA) activities, are subject to dispute resolution procedures. Docu-
ments designated as secondary are subject to review and comment.
Secondary documents are feeders to primary documents and are subject
to dispute resolution when incorporated into primary documents or when
the corresponding primary document is issued.
The dispute resolution section provides the parties to the agreement
with the ability to formally dispute issues associated with primary docu-
ments. This process assures that the work being conducted by the Federal
facility is in compliance with the requirements of CERCLA, the NCP
and applicable state law. The U.S. EPA administrator makes the final
decision in disputes should the parties not resolve these disputes at lower
levels. The U.S. EPA expects that in all but the most extraordinary situa-
tions, disputes will be resolved at the project manager or director level.
The dispute resolution section also includes a threshold for stopping
work affected by a dispute. The threshold is crossed in the event of
inadequate or defective work which is the U.S. EPA's or the state's
opinion is likely to yield an adverse effect on human health or the en-
vironment or to have a substantial adverse effect on the remedy selection
or implementation process.
The enforceability section preserves citizen suit rights under section
310 of CERCLA. States arc "persons" under CERCLA and therefore
can sue to enforce the IAG in Federal district court. The enforceability
section specifically establishes that deadlines related to the RI/FS and
terms and conditions of the IAG are enforceable, as is final dispute
resolution, by any person pursuant to section 310. Also, all parties have
the right to enforce IAG terms
The stipulated penalties section allows the U.S. EPA to assess stipu-
632 RCRA / SUPERPUND ACTIVITIES
-------
lated penalties to be payable to the Hazardous Substances Response
Trust Fund in the event of specified failures under the agreement. The
amount assessed and the reasons for the assessment must be reported
on a facility-specific basis by DOD or DOE in its annual report to
Congress on progress under section 120.
The funding section requires DOD and DOE to seek sufficient funds
for response and include estimates in their annual reports to Congress.
The U.S. EPA reserves its rights against any other party if funding is
not available.
Provisions describing the work to be accomplished at a specific site,
the schedule for its accomplishment, and any individual state concerns,
are negotiated on a site-specific basis. As of August, 1989, lAGs have
been concluded for 21 Federal facilities covering 26 NPL sites.
Federal Agency Authority Under CERCLA
Executive order 12580, which delegates authorities contained in
SARA, delegates Section 104 response authority to DOD and DOE for
releases on or originating from their facilities. It requires that such
response authority be exercised in accordance with seciton 120, which
provides the U.S. EPA administrator with the final decision on remedy
selection should the U.S. EPA and a Federal agency disagree. All Federal
agencies are delegated section 104 response authority for non-
emergencies at non-NPL sites where the release is on or originating
from their facilities.
Executive order 12580 may have singled out DOD and DOE for ad-
ditional response authority because both agencies established cleanup
programs prior to the passage of SARA in 1986. DOD established its
Installation Restoration Program (IRP) in 1975. Under IRP, each service
operates a program whose goals are to identify and evaluate past waste
disposal practices at DOD facilities. Studies and remediation are con-
ducted as necessary. Section 211 of CERCLA governs management of
the program. DOD funding for IRP in FY'89 is $500 million.
DOE initiated an informal program in 1984 designed to identify,
evaluate and remediate hazardous waste contamination at DOE facili-
ties. DOE has not yet established a formal response program analo-
gous to DOD's IRP.
ISSUES UNIQUE TO THE EXECUTIVE BRANCH
This section of our paper describes two issues that are unique to the
executive branch. These issues are funding and dispute resolution.
Funding
Unlike the private sector, Federal agencies cannot use earnings to
fund their hazardous waste cleanup responsibilities. Federal funding,
including funding for cleanups by Federal facilities, is requested by the
President and appropriated and overseen by Congress.
Clearly, Congress plays an essential role in Federal facility cleanups
by appropriating sufficient funding for those cleanups. Compliance by
Federal facilities with RCRA and CERCLA is subject to available appro-
priations, although the U.S. EPA's RCRA compliance agreements state
that failure to obtain funding does not release the Federal facility from
its obligations to comply with RCRA and the terms of the agreement.
Dispute Resolution
Federal agencies are created and supported by Congress and report
to the President, who ultimately is accountable for agency missions.
Federal agencies are immune form suit except to the extent that sovereign
immunity is specifically waived in legislation by Congress.
In the view of the Department of Justice (DOJ), executive branch
agencies may not sue each other; nor may one issue an administrative
order to another without providing a prior opportunity to contest the
order within the executive branch1. Like lawsuits, unilateral order
authority is viewed as inconsistent with the constitutional principles
of unity and unitary responsibility within the executive branch2.
Executive branch disputes of a legal nature are properly resolved by
the President or his delegate, in DOJ's opinion, because lawsuits and
unilateral administrative orders interfere with the President's ability to
manager the executive branch3.
RCRA
The DOJ has distinguished between section 3008(a) compliance
orders and section 3008(h) corrective action orders in regard to the
U.S. EPA's authority to issue RCRA orders to Federal facilities. In the
DOJ analysis, the U.S. EPA may issue section 3008(h) corrective action
orders buy may not issue section 3008(a) compliance orders.
Section 3008(a) Orders
According to the DOJ, the U.S. EPA may not issue a section 3008(a)
order to a Federal facility to address compliance violations because
an order is not a "requirement" under Section 60014. Section 6001 de-
fines the obligation of Federal facilities to comply with RCRA.
Section 6001 states in part that Federal agencies dealing with solid
waste:
.. .shall be suject to, and comply with, all Federal, State, inter-
state, and local requirements, both substantive and procedural
(including any requirement for permits or reporting or any
provisions for injunctive relief and such sanctions as may be
imposed by a court to enforce such relief)... in the same manner,
and to the same extent, as any person is subject to such
requirements. . .
The DOJ found that the issue turned on whether a section 3008(a)
order constitutes a substantive or procedural requirement, and cited
RCRA's legislative history and case law to determine that section 3008(a)
orders are not requirements; they are instead a means by which require-
ments are enforced5.
Instead of issuing section 3008(a) orders to address compliance at
Federal facilities, the U.S. EPA will issue a Notice of Noncompliance
(NON)6. A NON is similar to a section 3008(a) order in content and
format; it details the violation, remedy and remedy implementation
schedule.
After issuance of the NON, the U.S. EPA and the Federal facility
negotiate a Federal Facility Compliance Agreement (FFCA)7. The
FFCA is the document that resolves compliance violations by specifying
the agreed-upon remedy, compliance schedule, reporting requirements
and record-keeping requirements. Also included in a FFCA is dispute
resolution language, which emphasizes resolution at lower levels, and
an enforceability clause, which clarifies that the FFCA may be enforced
under RCRA's section 7002 citizen suit provision.
To ensure that negotiation of FFCAs is concluded in a timely manner,
the U.S. EPA has established an elevation process for resolution of
issues8. The U.S. EPA's goal is to conclude FFCA negotiations within
120 days. At day 90, U.S. EPA regions evaluate the negotiations and
determine whether agreement is likely within 120 days. In any case
where agreement does not appear likely in that time-frame, the case
is referred to U.S. EPA headquarters for resolution. Upon referral, the
assistant administrator for the Office of Solid Waste and Emergency
Response meets with an equivalent representative from the Federal
agency involved. If the dispute is not resolved within 30 days, it is ele-
vated for resolution to the U.S. EPA administrator and his Federal agency
counterpart.
The DOJ's opinion that the U.S. EPA may not issue section 3008(a)
orders to Federal facilities does not prohibit the U.S. EPA from issuing
such orders to the contractor at a government-owned contractor-operated
(GOCO) facility'. Contractors at GOCO facilities are subject to
RCRA to the same extent as any non-Federal entity, including orders
assessing penalties. Several courts have held that penalties may not be
assessed against Federal facilities because, under section 6001, enforce-
ment sanctions are distinct from requirements10. Although some Fed-
eral agencies indemnify their contractors, so that a fine assessed for
environmental violations against the contractor ultimately is paid by
the Federal agency, there is authority for the proposition that private
contractors may not be afforded the privileges of the Federal
government"
Section 3008(h) Orders
Section 3008(h) corrective action orders, as opposed to section
3008(a) compliance orders, are viewed by the DOJ as integral to the
RCRA / SUPERFUND ACTIVITIES 633
-------
permitting process, which Federal facilities are required to comply with
under section 6001';, Thus, the U.S. EPA may issue section 3008(h)
corrective action orders to Federal facilities. Federal facilities that receive
section 3008(h) orders may confer with the U.S. EPA on such orders
and raise any issue that cannot be resolved at the regional level to the
U.S. EPA administrator for final resolution" The U.S. EPA may also
issue section 3008(h) orders to the contractor at GOCO facilities.
CERCLA
The application of CERCLA authorities at Federal facilities is less
subject to interpretation than the application of RCRA authorities.
Section 120 requires Federal facilities to enter into lAGs for remedial
action at NPL sites. lAOs are enforceable through CERCLA's section
310 citizen suit provision" In addition, section 122(1) specifically
authorizes imposition of civil penalties for failure or refusal to comply
with a section 120 IAG.
The U.S. EPA may issue an order under section 104(e)(5)(A) to ob-
tain access to a Federal facility or to collect information. The U.S. EPA
may also issue a section 106 order to ensure compliance at a Federal
facility where there is an imminent or substantial endangerment to public
health, welfare or the environment due to an actual or threatened
hazardous substance release from the facility. In either case, however,
executive order 12580 requires that the U.S. EPA receive DOJ concur-
rence on the order. The U.S. EPA may use any of its administrative
and judicial authorities under CERCLA against a contractor at a GOCO
facility.
RCRA/CERCLA INTEGRATION ISSUES
RCRA commonly applies at a Federal facility that is subject to
CERCLA. For example, a Federal facility that is listed or proposed
on the NPL may also have interim status or a permitted unit under
RCRA
There are some unresolved issues about which statute should be used
as the primary vehicle to ensure cleanup when both RCRA and
CERCLA apply, and how the statutes may be used together. This is
particularly so where a RCRA-regulated release is the cause of NPL
listing, rather than a contributing factor. Neither statute is entirely clear
on these issues.
Most states are authorized to run the RCRA base program, and several
have gained authorization for HSWA authorities. Federal facilities have
their own delegated authorities under CERCLA; DOD and DOE have
response authority for all releases on or originating from their property.
The U.S. EPA, states and Federal facilities share the ultimate goal of
cleaning up Federal facilities, although there may be disagreement about
which statute should control in a particular case.
The statutory overlap may be broader than RCRA and CERCLA.
Another federal statute, such as the Atomic Energy Act, may apply
in a given case. Also, state or local hazardous waste-related authori-
ties, independent of RCRA or CERCLA, may apply to a particular
facility.
Inherent in the statutory overlap is a jurisdictional overlap. The juris-
dictional overlap may be between Federal programs, such as RCRA
and CERCLA, or between state and Federal programs, such as an
authorized state RCRA base program and the Federal CERCLA or
HSWA programs.
Such overlaps contain a potential for slowing cleanup while disagree-
ments over which statute to use are resolved and for inconsistent or
duplicalive cleanup activities if disagreements are not resolved. To speed
cleanup and avoid inconsistency or duplication, the U.S. EPA is working
to develop a mechanism to create a comprehensive, coordinated response
at Federal facilities with a RCRA/CERCLA overlap. This mechanism
is a three-party IAG with the slate, the U.S. EPA and the Federal facility
as signatories.
A three-party IAG can address site-specific state concerns and
maximize slate involvement in the cleanup process. Regulatory or over-
sight authority for work conducted under an IAG can be allocated in
a manner consistent with the concerns of the parties. Such an agree-
ment could satisfy a Federal facility's corrective action responsibilities
under RCRA as well as the public participation requirements of both
RCRA and CERCLA; a RCRA permit could later incorporate require-
ments of the IAG, if appropriate.
A three-party IAG may be developed for either NPL or proposed
NPL Federal facilities. A three-party IAG is also flexible enough to
include a non-listed RCRA-regulated portion of a Federal facility where
that makes sense from a technical standpoint, thus providing for a swifter
comprehensive cleanup. An IAG allows the parties to include in a
response action releases of CERCLA hazardous substances that are not
regulated under RCRA (e.g., radionuclides).
The US. EPA is focusing on RCRA/CERCLA integration at Federal
facilities through a variety of efforts including policy development and
a work group of stale and U.S. EPA regional representatives. The state
representatives have taken the lead on developing language for a three-
party IAG. While the U.S. EPA has expressed a strong preference for
three-party lAGs, it will enter into two-party lAGs or issue '106 orders
to the facility at Federal facility NPL sites where necessary to fulfill
its statutory mandate.
CONCLUSION
The Federal facility hazardous waste compliance program is on track.
The provisions of CERCLA section 120 are being implemented. With
DOD and DOE agreement on model IAG language, the number of site-
specific lAGs concluded is rising. Unique dispute resolution and en-
forcement procedures have been designed where deemed necessary by
DOJ to ensure Federal facility compliance. To further enhance progress,
the U.S. EPA is developing three-party lAGs as a mechanism to inte-
grate RCRA and CERCLA at Federal facility sites.
REFERENCES
1 Statement of F. Henry Habichi 0. Assistant Attorney General, Lands and
Natural Resources Division, Department of Justice, before the Subcommittee
on Oversight and Investigations, Committee on Energy and Commote,
House of Representatives, April 28, 1987. p.29.
2. Ibid.
3. Id.
4. Habicht statement. Appendix B.
5. Ibid.
6. Memorandum from J. Winston forte*. Assistant Administralorfcr Solid V*«
and Emergency Response, to EPA Regional Administrators, "Enforcement
Actions Under RCRA and CERCLA at Federal Facilities," January 25,1988,
p.3.
7. Porter memorandum, January 25. 1988, p.3.
8, Memorandum from J. Winston Porter, Assisttnt Administrator for Solid WK*
and Emergency Response, to EPA Regional Administrators, "Elevation
Process for Achieving Federal Facility Compliance Under RCRA," March
24, 1988.
9. Porter memorandum, January 25, 1988, p.4.
K). Habichi statement. Appendix B, pp. 6-9.
II, Inside EPA. "In Boon To Enforcement, Judge Ban DOE from Intervening
for Contractor," December 18, 1987, p.5.
12. Porter memorandum, January 25, 1988, p.5.
13. Ibid.
14. Conference Report on Superfund Amendments Reauthorization Ad of 1986,
99th Congress, 2d Session, Reprt 99-962, p.242.
634 RCRA / SUPERFUND ACTIVITIES
-------
Community Assessment: A Planned Approach to Addressing
Health and Environmental Concerns
Barbara Shyette Barnett
Barnett-Beese & Associates
Chicago, Illinois
Susan Pastor
U.S. EPA Region V
Chicago, Illinois
ABSTRACT
The U.S. EPA develops and implements plans for public participa-
tion at each site on the NPL. Public participation is required under
SARA during the investigation and cleanup of Superfund sites. These
provisions are included in SARA to ensure that residents in communi-
ties potentially affected by hazardous waste sites are informed of and
have the opportunity to participate in site-related activities.
The core of each site-specific public participation program is a
planning document called the Community Relations Plan (CRP). The
CRP evaluates the community within the context of the site investiga-
tion and cleanup, and outlines the goals and activities to be undertaken
by U.S. EPA to address the concerns and participation interests of the
community. The CRP is developed through a process that the authors
call community assessment.
Community assessment evaluates the potential economic, social and
political impact of the site on the community and, conversely, the poten-
tial impact of the community on site remediation. Each community is
distinct, frequently presenting several unique publics within the com-
munity structure. The publics which may include different levels of
public officials, environmental activists, community groups, people
residing in close proximity to the site or groups, such as the elderly,
with special needs. The various groups within the community often
have conflicting concerns and agendas.
Community assessment involves several activities, including iden-
tifying potentially affected or interested local residents and groups,
making field trips to the affected community, conducting numerous in-
person interviews and doing extensive file research. The result is an
analysis that assists the U.S. EPA in addressing concerns, initiating a
dialogue with the community and establishing credibility among com-
munity members.
The authors discuss a series of assessments they conducted in three
Dane County, Wisconsin communities. The sites are located within a
15-mi region of the county in three different, but adjacent, communi-
ties in the Madison, Wisconsin, metropolitan area. Madison is the state
capital and the home of a major midwestern university. The sites are
all landfills in which similar types of hazardous wastes were disposed.
The authors will examine the community assessment process, its value
as a tool in identifying potential community concerns and its use in
anticipating potential community controversies and opposition in other
areas of environmental planning.
INTRODUCTION
Several federal environmental programs address a growing public
demand for knowledge about environmental risks and participation in
decisions that will affect the health, welfare or the environment of their
community. The Superfund community relations program is one such
program. The program involves a process'of identifying potentially
affected publics, defining their concerns and developing and imple-
menting a.comprehensive plan to address them.
The first step in this process is called community assessment. Com-
munity assessment identifies the extent and nature of community interest
and concern about the site in the same way that a hazardous waste site
assessment identifies the nature and extent of a potential contamina-
tion problem. Site assessments often involve qualitative and quantita-
tive biological, physical and chemical analyses; community assessment
includes qualitative political, social and economic analyses of a com-
munity in relation to a potential or existing hazardous waste problem.
Each community assessment includes file research, library research,
field interviews, analysis and report writing.
The product of the community assessment is a document called the
Community Relations Plan (CRP). Among other things, the CRP
proposes activities and actions to address the concerns and issues iden-
tified during community assessment.
Implementing the activities suggested in the CRP then becomes the
ongoing basis of the site community relations program. Activities often
include public and small group meetings, newsletters, library exhibits
and public involvement programs.
One observation the authors have made over the course of working
with numerous communities is that each community is unique. Even
adjacent communities, sharing common schools and history, frequent-
ly react quite differently to Superfund, hazardous waste and environ-
mental issues.
Three of the more interesting communities the authors encountered
as a team were in Dane County, Wisconsin. Dane County is the home
of Madison, the state capitol, the University of Wisconsin and several
old hazardous waste sites. Three of these sites are included on the NPL:
(1) the City Disposal Corporation Landfill site in Dunn Township,
(2) the Hagen Farm site in Dunkirk Township (3) the Stoughton City
Landfill in the City of Stoughton. Two additional sites, the Every Farm
site in Dunkirk Township and the Madison Metropolitan Sewage System,
were proposed for inclusion on the NPL but are not yet on the final list.
WHY COMMUNITY ASSESSMENT?
Community assessment serves a dual purpose. This phase of the com-
munity relations program primarily is designed to identify community
concerns and understand their basis. However, the assessment also ena-
bles the U.S. EPA to become known by the community early in the
site investigation and cleanup process. We were aware that making our-
selves accessible to residents and officials would be crucial to the suc-
cess of future community relations activities and the ultimate cleanup
of the three sites. By letting people know at the beginning of the process
\ who-.was accountable for the technical and community relations aspects
RCRA / SUPERFUND ACTIVITIES 635
-------
of each project, we hoped to win at least the cautious trust of Dunn,
Dunkirk and Stoughton residents. Also, early intervention enabled the
U.S. EPA to be proactive in addressing concerns, thus avoiding sur-
prises and having to Tight Tires later in the process.
By the time the technical staff went into the field, both the commu-
nity relations staff and technical project manager were knowledgeable
about community concerns, potential controversies, information require-
ments and potentially sensitive issues. The U.S. EPA generally initiates
formal contact with the community approximately 4 mo before the
remedial investigation begins. Shortly thereafter, we start the commu-
nity assessment.
Sometimes the U.S. EPA community relations coordinator becomes
involved long before formal community assessment is started. In August,
1988, an Oneida Indian tribe contacted the U.S. EPA about some site
related concerns shortly after the nearby Fort Howard Paper Company
sludge lagoons site in Green Bay, Wisconsin was proposed for the NPL.
So, the U.S. EPA representatives traveled to Green Bay with offi-
cials from two other agencies to hold public meetings and meet with
the Oneida Business Committee to explain the Superfund process.
When community assessment is not done, the potentially affected
community may feel shut out of the cleanup process. This exclusion
of interested parties may lead to increased uncertainty about the site
and its associated risks. The credibility of the U.S. EPA may be severely
decreased in the community. As a result, the community may become
hostile toward the U.S. EPA. This was the case at one Superfund site
in Indiana. While a well-planned program is now being developed and
implemented, it may take a long time before a trusting relationship is
established between the community and the U.S. EPA.
THE COMMUNITY ASSESSMENT PROCESS
Each community assessment began with extensive file research to
understand as much about each site as possible. We reviewed technical
reports and other documents to understand how the site was brought
to the attention of the U.S. EPA, why it was placed on the NPL and
to identify evidence of past community interest.
Technical information about the site's history was used in our later
discussions with community officials and residents. In order to talk
intelligently and to ask questions that addressed the basis of their con-
cerns, we needed to have a good working knowledge of the nature and
extent of the potential contamination problem.
While this background research was being done, the first telephone
contacts with members of the communities were made. These com-
munity contracts were made while we established locations for the site
information repositories, where site-related documents and informa-
tion about the U.S. EPA and Superfund would be placed for public
review. Establishing an information repository is one of the first com-
munity relations activities undertaken. The repository is an ideal loca-
tion to place information that may be of interest to the community.
Because the likely locations for the repositories include the local library
and municipal hall, the repository establishment process provided an
opportunity to meet one or more public officials and the local librari-
an. Conversations with these local officials frequently provide insight
into the local political climate and general feelings about the U.S. EPA
and Superfund.
Sue Pastor (U.S. EPA) contacted the Stoughton city clerk's office to
set up the Hagen Farm repository because our information told us thai
Stoughton was the location of the NPL site. We learned that our infor-
mation was not entirely accurate. We also learned three additional
lessons from this one telephone call.
First, the city clerk refused to house the information repository,
making it very clear that the NPL site was not located in Stoughton.
The site was actually located in unincorporated Dunkirk Township which
surrounds Stoughton on all sides and shares its schools, post office and
Scandinavian ethnic culture. The clerk wanted us to be sure that there
was no misunderstanding of the local geographic boundaries.
Lesson two was that Stoughton officials were very sensitive about
the presence of a Superfund site within the city's corporate limits.
Stoughton already was host to one Superfund site, the Stoughton City
Landfill (which we had yet to start working on). The City of Stough-
ton was named as potentially responsible for the cleanup cost of the
Stoughton City Landfill. The city wanted nothing to do with the Hagen
Farm site.
The third lesson was that we needed to be very cautious to distin-
guish sites and communities and not to address them in our community
assessment as a unit.
Through these preliminary contacts, Ms. Pastor also identified the
local newspaper which covered all three communities and their respec-
tive governments. She learned that the Madison papers also covered
that pan of the county. She then contacted the local reporters. This con-
tact provided her an opportunity to introduce herself as the U.S. EPA
contact and learn more about the three communities as well.
During this early stage of community assessment, we tried to identify
residents, officials and groups that might be interested in each site. Iden-
tifying the appropriate people to interview was critical in correctly
characterizing communities' concerns and interests.
We identified potentially interested individuals through the background
research using several valuable resources. Some of the most useful were:
a plat map, a local telephone directory, a visitors' kit from the local
Chamber of Commerce and the local health department.
The plat map of the area surrounding each site helped us identify
specific residences or businesses located near the sites. We made a
special effort to contact those individuals living near the site because
they would be most directly affected by the site and its remediation.
The local telephone book, which served all three communities, helped
us locate telephone numbers of people we identified. In addition, by
thumbing through the yellow pages, we identified additional local com-
munity groups. The telephone book, as much as any other resource,
also gave us a sense of the community's flavor and culture. It gave us
an insight into the types of businesses, clubs, civic groups and services
present in the community.
The Chamber of Commerce packet provided us with specific geo-
graphic, social and demographic facts about the community, including
its major employer, the Uniroyal plant in Stoughton, population, com-
munity history and local services.
The local health department was an important contact at this stage.
The health department is often the first official agency to hear health
or environmental complaints from concerned residents. We were for-
tunate that these three sites were in contact with a county environmen-
tal health sanitarian knowledgeable about all three sites. He lived in
Stoughton (in fact near the Uniroyal plant) and was able to supply us
with a history of past community involvement and concerns and names
of concerned residents for us to contact and interview.
Our research identified numerous parties to interview. As we tele-
phoned the people on our list, we were often referred to neighbors and
other interested residents. People were generally forthcoming and
seemed to welcome our interest. The types of individuals we identi-
fied are shown in Table 1.
After the preliminary research and identification, we scheduled in-
person interviews. We scheduled about 15 interviews in each commu-
nity, trying to get a variety of perspectives. Some additional interviews
were conducted over the telephone to accommodate those who were
unavailable during the field trip. Good scheduling ensured that every-
one who wished to had an opportunity to be interviewed and that ap-
pointments were not missed. We allowed 1 to 2 hr for each interview.
We also left a half hour between meetings to find our way, over rural
and unknown terrain, to the next appointment.
We limited the number of people present at each interview. WE did
this for several reasons. People tend to speak more openly and honestly
about their concerns in a one-on-one situation. A local alderman, for
example, might be more willing to disagree with the mayor if the mayor
is not present. We also did not want to interview a large group of resi-
dents together at one meeting. The interview could easily turn into i
public meeting, and the goal of information gathering could have been
lost.
Based on the preliminary research, we developed an interview ques-
tionnaire to guide the interview and provide topics for discussion. The
questions addressed site history, past community involvement, percep-
tions about the U.S. EPA, community culture, specific past and present
636 RCRA / SUPERFtlND ACTIVITIES
-------
Table 1
Typical Local Contacts for Community Interviews
Local officials:
Town Board Chairmen (Dunn and Dunkirk)
Mayor (Stoughton)
Town/City Clerk
Town Board members
Aldermen
public Works Director
County Officials:
County Board Representative
County Environmental Health Sanitarian
State Officials:
State and district WDNR personnel
State Legislators from affected district
Wisconsin Division of Health personnel working on site health
assessment
Groups:
League of Women Voters (Dane County and Stoughton Chapters)
sierra club
Audubon Society
Environment Wisconsin
Residents:
Long-time establishment residents
Newer term residents mostly from the University and working
in Dane county
People living directly around the site
Table 2
Community Profiles
Town of Dunn
o Site closed for many years
o Population: 5,000
o Unincorporated Township
o Madison, WI bedroom community
o Environmental issues a strong concern
o Politically active community
o Town meeting well attended
o Strong latent interest
o Rely on private drinking water wells
Town of Dunkirk
o Site closed for many years
o Population: 1,800
o Unicorporated township
o Rural, fanning community
o Environmental issues not of great concern
o Current interest in site is low
o Rely on private drinking water wells
Stoughton
o Site closed for several years
o Incorporated municipality
o Public water supply
o City had suit pending against U.S. EPA to delete site
from Superfund List
o Site was being developed as a park
o New senior citizen housing constructed adjacent to site
o Very high level of interest in the site
environmental concerns and the way in which the group or individual
would like to be informed about, or involved with, the site remediation.
After returning home from the field trips, we reviewed and analyzed
the many additional documents and newspaper clippings we obtained
from local officials, residents, state files and the library. We also sum-
marized our interview notes, identified common concerns and began
to formulate how we might address them. The three communities were
quite distinct and resulted in three different community relations ap-
proaches. Profiles of the communities based on the assessments are
presented in Table 2.
Based on the field and other research, we compiled a comprehen-
sive community relations plan for each site. Each CRP contained a site
history, community profile, information about the local media, a his-
tory of past community involvement with the site and discussion of the
concerns raised by residents, officials and other interested parties. Each
CRP included a discussion of public participation goals for the com-
munity based on the concerns raised, and activities designed to achieve
the public participation goals. While many of the activities outlined
in the CRP are guided by CERCLA, others are designed to address
specific and general environmental and health concerns expressed during
the interviews.
DIFFERENT APPROACHES
In Stoughton, we addressed the city's interest in community relations
by calling on it for logistical support in planning meetings and reposi-
tories. At the same time, we addressed the concerns of some of the
residents by installing a groundwater and risk assessment exhibit in the
water department foyer. We also addressed the special needs of the senior
citizens we identified during our field work. We planned a special day-
time public meeting at the development's clubhouse to accommodate
the less mobile senior citizens. We developed written fact sheets to help
residents understand Superfund, risk assessment and the nature of the
work to be conducted at the site.
In Dunn, our research indicated that there was latent interest in the
site. Our research was right. Nearly 100 people attended our first public
meeting. People at the meeting wanted to be very involved and informed
about the cleanup program. After the meeting, the residents formed
a citizens' committee to monitor activities. The U.S. EPA technical
project manager has since attended several of those meetings.
In Dunkirk, as predicted, things have remained fairly quiet with only
minimal expressed interest. Thirty people attended the public meeting.
Another meeting will be held there this summer to keep interested resi-
dents up-to-date on site activities. But low interest has indicated no need
for special activities.
During our Dane County community assessments we examined three
communities, three sites and three worlds. Only by careful book and
field research could we have anticipated and planned for these vast
differences.
APPLICABILITY TO OTHER AREAS OF PLANNING
Community assessment is applicable whenever an issue may have
an impact on the public health or environment of a community. One
such issue is recycling, surely prominent in environmental planning.
Community assessment might be used to identify the factors that will
encourage individuals to recycle more materials. It also could identify
the best avenues and vehicles for disseminating information about the
recycling program. That information could then be used to increase
participation in the recycling program.
Public health officials might use community assessments to identify
and then act on local health concerns like AIDS, water pollution or
environmental cancer risk. Programs geared to the nature and level of
concern may then be more effectively developed.
Community assessment conducted before a chemical plant, incinera-
tor or other locally unacceptable land use is planned and "imposed"
on a community might help developers and planners choose host com-
munities more carefully. Understanding the nature and extent of the
community's concern may enable the developer to address legitimate
concerns and work with the host community in planning the project.
Sometimes, no matter how well a community and its concerns are
understood, conflict, controversy and opposition cannot be overcome.
However, community assessment is a good first step in establishing a
meaningful dialogue and a strong base for planning decisions.
RCRA / SUPERFUND ACTIVITIES 637
-------
Community Relations Programs:
Improved Planning through Better Understanding
of Communication Systems
Ifemara L. Reeme
Charlotte F. Young, Ph.D.
Argonne National Laboratory
Environmental Research Division
Argonne, Illinois
ABSTRACT
Thorough and situation-specific planning is often neglected during
development of community relations programs. However, planning is
critical to developing effective and ultimately successful community
relations programs. Such planning involves several steps: sctnng goals
and objectives, identifying community characteristics, understanding
communication systems, determining target audiences and developing
an evaluation scheme.
Although crucial to effective planning, (he step "understanding com-
munication systems" often is overlooked. This paper therefore not only
emphasizes the importance of understanding communication systems,
but also includes brief descriptions of each planning step.
Communication systems generally have seven components or aspects:
objects (e.g., receivers, senders, communication media and messages),
attributes of objects, relationships among objects, environments in which
the system functions, balance, hierarchical organization and goal orien-
tation. Each of these components or aspects is described as it pertains
to communit) relations programs, and the main methods for obtaining
information about each component or aspect are discussed.
INTRODUCTION
Implementation and operation of environmentally regulated projects
often require the inclusion of programs to address community concerns.
For example, NEPA, CERCLA/SARA and RCRA projects often require
community relations programs because these types of environmentally
sensitive activities often attract public attention and invoke (he interest,
curiosity, concern and at times outrage of local communities. Whether
community relations programs are required by law or not, they are
strongly recommended as a means to mitigate negative community
reaction as well as to foster community support.
The value of community relations programs often is underestimated
Consequently, thorough and situation-specific planning for such
programs is frequently is neglected during project development. A major
benefit of such planning is eventual accomplishment of program goals
and objectives. Also of benefit are improved corporate or agency images
and increased credibility within the community. On the other hand,
poorly planned and communicated community relations programs may
damage one's image and one's credibility. Because community percep-
tions often extend beyond (he bounds of a single projec(, they cun
influence the outcome of future, unrelated projects.
Future community relations programs will benefit from the experience
gained during current program planning and implementation. Although
planning effective communication strategies is initially energy-intensive
and time-consuming, the effort pays off in the long run. Planning and
implementing community relations programs becomes increasingly cost-
effective as execution becomes more efficient with practice. For example,
experience assists organizations to respond more quickly and effectively,
and with less preparation, in the event of unanticipated events or when
time is not available for thorough advance planning. Furthermore, the
costs associated with preparing community relations programs are small
relative to the overall costs of the associated projects. Because roost
environmentally sensitive projects require public support to be success-
ful, well-planned community relations programs can greatly influence
their outcome.
Communities and their communication systems are complex and need
to be characterized and understood before community relations activi-
ties are initiated. The planning so essential to success includes the
following steps: setting goals and objectives, identifying community
characteristics, understanding communication systems, determining
target audiences and developing an evaluation scheme. Although the
step "understanding communkation systems" is particularly important,
it frequent!) is overlooked by program planners and guidance docu-
ments. This paper therefore not only briefly discusses each planning
step, but also describes communication systems and their components
in more detail, thereby providing understanding of their functions and
interrelationships.
COMMUNITY RELATIONS PROGRAM PLANNING STEPS
Planning provides the basis for a coordinated effort by providing
program direction, reducing the effects of unexpected changes, minimiz-
ing waste and redundancy, and facilitating control over communication
and information exchange. Community relations program planning steps
include the following major activities: (1) setting goals and objectives,
(2) identifying community characteristics, (3) understanding commu-
nication systems, (4) determining target audiences and (5) developing
an evaluation scheme.
Setting Goals and Objectives
Community relations program goals and objectives underpin the
planning process by indicating the purpose of the program in terms
of expected accomplishments and desired outcomes. They should
provide direction, yet be flexible enough to accommodate new infor-
mation and necessary revision. Nevertheless, the desired outcomes of
the program should be specifically stated so that each step has clear
criteria for measuring program effectiveness. A program typically has
multiple goals and objectives. Typical goals are to disseminate infor-
mation to special interest groups, heighten the awareness of uninformed
publics, gain project support from local government officials or agencies,
resolve specific issues with local business groups or solicit the partici-
pation of affected community members.
638 RCRA / SUPERFUND ACTIVITIES
-------
Identifying Community Characteristics
Once the program goals and objectives have been set, planners should
then learn about the community. Identifying community characteris-
tics is essential to determining target audiences and to understanding
the communication systems so that community relations messages can
be meaningful. Numerous variables affect program development and
outcome. Measuring these variables and analyzing their respective sig-
nificance provides a community profile that is very valuable in deter-
mining community subgroups and target audiences, identifying
communication systems, choosing appropriate communication channels
and formulating messages. Table 1 lists the variables most important
to community relations programs.
Table 1
Variables to Consider When Identifying Community Characteristics
Variable
Comment
Media attention, coverage, and opinion
Risk perception
Actual hazard or risk
Sallency of Issue or proposed activity
Degree of polarization within the
comunlty
Past public Interest, Involvement,
or position
Demographics and social structure
Comunlty leaders with public
Influence
Geographical proximity
Camunlty concerns and environment
Project duration
Benefits and costs to community
Recipients of benefits and costs
Required regulatory oversight
Voluntary or Involuntary participation
Public participation
Alternatives
Urgency of Issue or activity
Image and credibility of players
Conmunications system
Public visibility (supportive or
adversarial)
Relative to actual risk
Relative to perception of risk
Importance of Issue relative to
other community Issues
Alignment of community opinion or
support
Expected level of activity
Race, age, marital status, and
Income
Special Interest groups, media
personalities, elected officials,
union representatives, educational
and religious leaders, and
neighborhood associations
"NIMBY"3 response
Competing or complimentary Issues
Long- or short-term effects
Municipal revenues and property
values
Equitable distribution thereof
Compliance enforcement
Amount of control over Involvement
and assumed risks
Timing and significance of Input;
token Involvement or real
decision-making opportunities
Identify and consider technological
and geographical options
Immediate (e.g., spill response) or
planned (e.g., treatment facility
construction)
Perceived sincerity and community
respect
Channels of communication and flow
of information
'Not In my backyard.
Understanding Communication Systems
Once acquainted with a community's characteristics, planners should
move on to understanding its communication systems. They should
assess how members of the public find out about environmental
problems and how they learn new community information. This step
usually is limited to identifying interested citizens, responsible govern-
ment officials affected organizations and community subgroups.
However, simply identifying these entities and compiling a list of con-
tacts is insufficient for effective community relations program planning.
Rather, the communication systems within which the community rela-
tions program must function should be identified. Not only identifying,
but also learning about these systems, constitutes a more comprehen-
sive approach to developing contacts.
A communication system—also called a communication network-
comprises the multiple communication links between people, agencies
and businesses. A communication system can be thought of as a net-
work of individuals linked by information exchange through mass media
or interpersonal communications. Any individual will likely partici-
pate in several communication systems. He or she will tend to partici-
pate in networks involving individuals who share interests and espouse
similar values.
Community relations programs are less effective when the commu-
nication links are incorrectly identified or inadequately understood. For
example, using inappropriate methods to inform community members
about an environmentally sensitive project or to solicit public involve-
ment will lead to costly errors and inefficiencies.
The following discussion provides a framework for understanding
how information flows through a community and techniques for iden-
tifying communication links. Understanding communication system
components promotes improved community relations program planning
and selection of the most appropriate communication channels for
conveying program information.
Communication System Components
Every communication system has several components. Although the
components are common to all systems, their characteristics will vary
according to specific circumstances (e.g., CERCLA/SARA remedial
response or NEPA EIS preparation). However, component types should
be identified first and then the major community-relations-specific
aspects of these components should be described. Communication
systems all have seven major components: objects, attributes of objects,
relationships among objects, environments in which the system func-
tions, balance, hierarchical organization and goal orientation.
Objects are the elements of communication systems. In the case of
community relations communication systems, the three main objects
are the senders, the receivers and the communications media and
messages. For example, receivers could be the community groups and
subgroups to which a message is directed.
Attributes are the qualities or properties of the system's objects. Table 2
lists selected attributes for the three main objects mentioned above.
Example attributes of a communicated message include how it is
designed and delivered. Attributes also include the characteristics of
the people or organizations that communicate, such as their beliefs and
value orientations and their previous exposures to similar environmental
issues (e.g., an agency's attitude toward planning and conducting a com-
munity relations program).
Table 2
Attributes of Objects within a Communication System
Object
Attribute
Communication sender
Communication medium and message
Communication receiver
Knowledge of community concerns
Attitude toward Involvement
Previous public Involvement
Attitude toward communication receivers
Environmental knowledge
Personal proximity to Issue
Risk perception
Position in community (e.g., opinion
leader or environmental group president)
Previous exposure to problem
Length
Channel (e.g., personal or interpersonal)
Type (e.g., brochure or newspaper)
Timing and frequency of information
dissemination
Amount of information
Amount of coverage
Credibility of Information
Suitability of language (e.g., technical
or regulatory Jargon)
Environmental knowledge or awareness
Diversity of demographic characteristics
Previous public Involvement
Previous exposure to problem and attitude
regarding problem
Personal proximity to Issue
Risk perception
Position In community
Community group membership
Effect of message (e.g., 1s the message
actually understood?)
Relationships among objects are how the objects interrelate and affect
each other. A change in one part of a system causes change in another
part. Moreover, relationships and their effects may be direct or indirect
RCRA / SUPERFUND ACTIVITIES 639
-------
and are usually multidirectional. In other words, information moves
from senders to receivers, but also vice versa. Objects are discrete units;
however, when they are combined with other objects, and as a result
of the relationships between those objects, they become a communica-
tion system.
Aspects of relationships to consider during community relations
program planning include frequency and credibility of communications,
past community involvement and number of communication sources
providing either similar or different information. Another aspect to con-
sider is the system's openness with respect to other systems (e.g.,
whether Native American or boat marina communication systems are
being accessed in addition to those of environmental groups).
Each system Junctions within an environment and is therefore affected
by this environment. Furthermore, the type of environment will likely
influence system relationships. For instance, open systems (i.e., those
that are strongly linked to other communication systems) can be effective
in reaching community subgroups.
A balanced communication system not only produces outputs, but
also receives inputs. For example, if an agency gives information to
community members, then the agency should be prepared to receive
information. Community members receiving information will react and
send other information to other parts of the system, as well as back
to the sender. The response information may be communicated in a
different form, such as anger, involvement or awareness. Therefore,
to maintain balance, each system must adapt and change. As a result,
the goals and objectives set early in the planning process may have to
be changed.
Hierarchical organization implies that objects combine to form sub-
systems within the larger system. A subsystem might include the
communications network of an agency or a firm with environmental
interests.
All communication systems are goal oriented. Planners should set
goals for the communication system, but recognize that there are mul-
tiple pathways to achieving those goals. The goals set should reflect
the goals and objectives set in the initial planning step, as described
previously. Like the overall program goals and objectives, the commu-
nication system goals may change as a result of system components
having different or conflicting goals. Likewise, the pathways to achieving
these goals also may change.
Techniques for Identifying and Describing Communication S\-stem
Components
Once the communication system components for a particular
community relations situation have been identified, three techniques
generally are used to describe the components more specifically: con-
duct surveys; search and review documents; and use available expertise
Each technique has certain strengths and weaknesses. The level and
type of information obtained by each, as well as the lime required to
implement each, also vary. In some cases, it may be appropriate to use
combinations of these techniques.
Conducting surveys entails formal questioning of community members
about their communication systems. This technique is (he most com-
prehensive, but also the most time-consuming. In addition, it requires
the most expertise. Three types of surveys are used: (1) mail, (2) tele-
phone and (3) face-to-face.
All three survey approaches should be carefully considered before
selecting the most appropriate one for a given situation. For example,
mail surveys are easier to send out, but may produce lower response
rates than telephone surveys. Telephone and face-to-face surveys re-
quire trained interviewers and time to conduct the interviews. In some
cases, using more than one type of survey may be appropriate. For ex-
ample, it may be most effective to interview influential community mem-
bers (e.g., opinion leaders) by telephone and mail questionnaires to
a random sample of community members.
Design and implementation of the survey also require careful
consideration. Questions should be carefully worded so that they solicit
the desired information. Furthermore, to ensure representative results
and absence of bias, survey participants should be selected through an
appropriate sampling scheme. A survey specialist can be very helpful
in this regard.
Searching and reviewing documents entails examining materials that
contain hints about the communication links within the community, for
example, telephone books and newspapers may help identify groups
interested in the issue (e.g., yacht club members for a water quality
issue or the Sons of the American Revolution for an excavation site
containing historical artifacts). Other useful documents might include
listings from the local Chamber of Commerce and government contact
lists. (Suggestions for compiling contact lists are found in Table 3.)
TaMc3
Contacts for Helping to Determine Communication System*
Contact type
Eiample
dieted federal. it«tc, and local
government officials
environment*) he*ltd and safety
agencies
environment*I organizations
Service groups and neighborhood
associations
Press and mtdla representatives
Special-Interest citizen groupt
Senators, congressional representatives,
governors, mayors, and council waters
U.S. EPA regional offices and branches;
state environment*! protection agencies
or department* of natural resources and
conservation; city, county, or UxmsMp
environmental commissions; local
•i)visor7 commissions and planning
boards; and health department sanitary
engineers
local Sierra Club chapter
Local bridge clubs, Kl.anlt Club, and
League of women Voters
Television, radio, and newspaper staff
Boating, hunting, and other recreation-
oriented groups
Searching and reviewing documents is less time-consuming than con-
ducting surveys and requires less specialized expertise. It facilitates
identification of communication system objects, but it does not facili-
tate understanding of other system components (e.g., attributes, rela-
tionships and hierarchies). Nevertheless, assembling contact lists and
identifying system objects provide the basis for preliminary diagramming
of community relations-specific communication systems.
Using available expertise entails relying on experience gained from
previous community relations program planning and implementation
efforts and informal contacts with potentially influential community
members such as government officials and service club leaders. Some
of the contacts may be asked some of the same questions asked in a
survey. However, one must be careful about the representativeness of
the answers when only a handful of community members have been
contacted.
Although this final technique may be the least time-consuming, it
also is the least systematic of the three. Furthermore, it is limited to
gathering information on system objects.
Determining the Target Audiences
After identifying community characteristics and understanding com-
munity communication systems, planners should proceed with deter-
mining the community relations program target audiences. The affected
community is composed of many subgroups. A subgroup may be an
individual, individuals with shared concerns or formal organizations.
Participating subgroups are those whose members have expressed
interest in being involved, have a record of participation or have special
interests (e.g., homeowners concerned with the effect of the project
on property values or local Sierra Club or League of Women Voters
chapters). Nonparticipating subgroups generally account for a greater
number of (but not necessarily the most influential) people and often
are referred to as the "general public."
The community relations program goals and objectives for each com-
munity subgroup may be different. For instance, an objective for non-
640 RCRA / SUPERF0ND ACTIVITIES
-------
participating community subgroups might be to increase their awareness
of the issues by distributing general fact sheets. An objective for par-
ticipating subgroups might be to educate them through project-specific
cost/benefit analyses.
Each community subgroup is a discrete audience with its unique set
of attributes and community characteristics (Table 1). To be effective,
community relations program planners must recognize the existence
of multiple target audiences and tailor the communications accordingly.
Developing an Evaluation Scheme
Mechanisms for community relations program evaluation should be
formalized during the planning process so that evaluations can be con-
ducted during program implementation as well as following comple-
tion. Moreover, program evaluation before full-scale implementation
provides intermediate feedback, which allows decisions made in the
previous planning steps to be appropriately modified to increase over-
all program effectiveness. For example, proposed messages and com-
munication methods or formats could be pilot-tested to determine their
appropriateness and effectiveness.
Early recognition and correction of ineffective messages and com-
munication methods greatly increase program efficiency by minimizing
production, administration and distribution activities. The effective-
ness of community relations programs can and should be evaluated
against measurable criteria developed during the setting of program goals
and objectives both during and after program implementation. Evalua-
tion results can then be used to modify inadequate program compo-
nents. Post-program evaluations lead to better understanding of
accomplishments and outcomes and provide the basis for the success
of future community relations programs.
CONCLUSION
There are several benefits to completing the planning phases of com-
munity relations programs in a timely and thorough manner. Planning
enables quicker and better responses to unanticipated changes. It also
increases the level of experience, which is useful for future program
development, and it can enhance the image and credibility of involved
agencies and corporations. Such factors can influence the success of
both current and future programs.
ACKNOWLEDGMENT
This paper is related to work performed for the U.S. Army Corps
of Engineers through an agreement with the U.S. Department of Energy,
under Interagency Agreement E8589D164.
REFERENCES
1. Babbie, E., Survey Research Methods, Wadsworth Publishing Co., Belmont,
CA, 1973.
2. E. Bruce Harrison Company, Inc., Environmental Communication and Public
Relations Handbook, Government Institutes, Inc., Rockville, MD, 1988.
3. Fazio, J. R. and Gilbert, D., Public Relations and Communication for Natural
Resource Managers, Kendall-Hunt, Dubuque, IA, 1983.
4. Hanchey, J. R., Public Involvement in the Corps of Engineers Planning
Process, IWR Research Report 75-R4, U.S. Army Engineer Institute for Water
Resources, Fort Belvoir, VA, 1975.
5. ICF, Inc., Community Relations in Superfund: A Handbook, U.S. EPA Report
No. EPA/540/G-88/002, U.S. EPA, Washington, DC, June 1988.
6. Littlejohn, S. W., Theories of Human Communication, Wadsworth Publish-
ing Co., Belmont, CA, 1983.
7. McDonough, M. H., "Audience Analysis Techniques," in Supplements to
a Guide to Cultural and Environmental Interpretation, prepared for Office,
Chief of Engineers, U.S. Army, Washington, DC, by Environmental Labora-
tory, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS,
1984.
8. Robbins, S. P., Management: Concepts and Practices, Prentice-Hall, En-
glewood Cliffs, NJ, 1984.
RCRA / SUPERFUND ACTIVITIES 641
-------
Treatment of Coal Tar Contaminated Soil and Lagoon Closure
Richard T. Cartwright
Daniel S. Schleck
Chemical Waste Management, Inc.
Research and Development Group
Geneva, Illinois
ABSTRACT
During 1984 and 1985, Chemical Waste Management's ENRAC
division removed 1.000,000 gal of coal tar from an open 1 acre lagoon.
Two thirds of the coal tar was treated on-site to make a supplemental
fuel blending component. The final one third of the coal tar was treated
using an in situ stabilization process. The stabilized material was then
disposed of in a hazardous waste landfill. After satisfying the cleanup
requirements of the slate regulatory agency, the coal tar lagoon under-
went final closure.
Over 1500 coal gassification "town gas" sites left over from plants
that operated in the United States between 1850 and 1950 have been
located. Coal tar waste from these plants was accumulated in open
lagoons. Many of these lagoons have since been filled in with debris
and covered with soil. The following treatment technologies are
evaluated for remediating coal tar contaminated soils at former town
gas sites:
1. Biorcmediation
2. Thermal Desorption
3. Organic Solvent Extraction
4. Surfactant Soil Washing
INTRODUCTION
For 100 yr, until the mid 1940s, "town gas" produced by coal gassi-
fication was a major source of energy for many cities throughout the
United States. Town gas was formed by carbonizing coal.
Coal was heated in a reactor to drive off volatile compounds that
became part of the town gas. At the same time, the healed coal was
reacted with steam to produce "water gas", which also was mixed with
the town gas.
In the town gas manufacturing process, 6 to 9 gal of waste by-product
coal tar were produced for each ton of coal fed. The yield of coal tar
from carbonization is approximately linearly related to the amount of
volatile components in the coal1 A typical plant using 50 tons of coal
a day, over a 100 yr lifetime, produced 10 to 16 million gallons of coal
tar7, which usually was placed in trenches or lagoons on-site. Many
of these trenches and lagoons have since been filled with debris and
covered over to meet pressing real estate demands.
It is estimated there are 1500 to 2000 of these town gas sites poten-
tially in need of remediation across the country'. Some of these sites
are currently part of the Federal Superfund program.
COAL TAR CHARACTERIZATION
Coal tar made from the carbonization of anthracite, lignite or
bituminous coal will have varying chemical compositions and physical
properties depending on the constituents of the original coal. The total
amount of volatile material in the coal has an impact on the amount
of coal tar produced when coal is carbonized. Composition also varies
depending on whether benzene/to I uenc/xylene (BTX), creosote or pitch
components were recovered from the tar. BTX often was recovered as
a solvent or as a liquid fuel component. Creosote was recovered and
used as a wood treating agent. Pitch was in demand as a sealing agent
for roofs.
If no auxiliary recovery of the BTX, creosote, or pitch components
took place, the typical composition of the resultant coal tar is illustrated
in Table 1.
Tabk 1
Composition at Coal Tkr4
Benzene
Toluene
Xylene
Light Oil
Naphthalene Oil
Heavy Creosote Oil
Anthracene Oil
Soft Pitch
Medium Pitch
Hard Pitch
15%
3%
1%
4%
8%
9%
13%
16%
14%
17%
PROJECT COAL TAR STRATIFICATION
Chemical Waste Management's (CWM) Environmental Remedial
Action division (ENRAC) gained "hands on" experience in remediating
a coal tar contaminated site between 1984 and 1985. During the project,
ENRAC removed 1,000,000 gal of coal tar from a 1-ac lagoon. ENRAC
found that the coal tar tended to stratify into several layers after settling
for over 40 yr. Each layer exhibited unique physical and chemical
characteristics'-'. At room temperature, separate liquid, semi-solid and
solid phases appeared. A description of each coal tar phase in the lagoon
from the project is found in Table 2.
COAL TAR REMEDIATION PROJECT
At room temperature, a composite blend of coal tar phases is a sticky,
tacky semi-solid material. However, significant changes in physical
properties occur with variations in temperature. At higher temperatures
the viscosity drops dramatically, causing the semi-solid material to
liquify. Table 3 illustrates this relationship between temperature and
viscosity for the coal tar found in the Illinois lagoon*.
642 RCRA / SUPERFUND ACTIVITIES
-------
EEASE
I
II
III
IV
•feblel
Phases Coal Tar Lagoon
DESCRIPTION
Light Oil (LO)
Viscous-Rubbery (VR)
Hard and crumbly (HC)
COMPONENTS
BTX, Light Oil
Naphthalene Oil,
Anthracene Oil,
Creosote Oil
All Pitches,
and any Coal
Solids (including
coke)
Contaminated Soils (CS) All Phases of Tar
Iable3
Viscosity of Coal Tar vs. Temperature for Coal Tar6
Temp fFl
72
120
160
200
Viscosity (Poise)
3400
435
85
16
Note: Coal tar becomes pumpable at 25 poise (roughly
185'F)
These changes in viscosity were exploited during the remediation
of the site. ENRAC found that coal tar can be excavated when cold
and pumped when heated. Because the coal tar in the project had sepa-
rated into three separate organic phases plus a contaminated soils phase,
individual materials handling strategies were developed for each phase.
The strategies are described in the next four sections of the paper.
Light Oil Phase Strategy
In 1984, the light oil phase of the lagoon was removed using a high
speed, open impeller submersible pump suspended from an overhead
crane. Because coal tar is black, the surface of the lagoon absorbed
significant heat energy from the sun during the summer to liquify the
light oil. Eventually, enough of this energy was absorbed to cause a
"sweet spot" to develop. This is an area where the sun had heated the
light oil sufficiently that it flowed and could be pumped. The pump
was continually moved to other new "sweet spots," until all of the
pumpable light oil phase was removed from the top of the lagoon.
The light oil was pumped directly from the lagoon into liquid tankers.
The tankers were heated using internal steam coils until shipment of
the contents to a liquid fuels blending facility. The fuels blender blended
waste solvents with the coal tar light oil prior to disposal as a liquid
supplemental fuel in a blast furnace, cement kiln or industrial furnace.
Viscous Rubbery Phase Strategy
After the coal tar light oil was pumped off, the submersible pump
begin to pick up the heavier viscous/rubbery coal tar phase. Pumping
operations slowed down. In addition to encountering a more viscous
coal tar phase, ambient temperatures had decreased (the fall months),
which made additional pumping almost impossible without auxiliary
heating. Coal tar remediation was shut down during the winter of
1984-1985.
The following spring, 1985, ENRAC began the remediation of the
viscous rubbery (VR) phase. A small pit was excavated adjacent to the
lagoon. Then a box pattern 20 ft x 20 ft x 10 ft heat exchanger made
of 2-in. pipe was placed into the pit. A hot glycol/water solution was
circulated inside the 2-in. pipe. The glycol/water solution was heated
by a steam package boiler system. The temperature of the coal tar in
the pit was kept below 190 °F, minimizing fugitive emissions of BTX
during heating.
The heat exchanger had a large heat transfer area, which was needed
to promote both convective and conductive heat transfer (coal tar has
a low thermal conductivity). Once the semi-solid material is liquified,
convective rather than just conductive heat transfer will occur.
The VR phase material was placed in the heat exchanger pit using
a clam shell and heated until it became fluid. The material was then
pumped, using a submersible pump, into steam-heated tankers for ship-
ment to the liquid fuels blenders as a liquid supplemental fuel
component.
Hard and Crumbly Phase Strategy
Hard and crumbly material was stabilized in situ using proprietary
blends of cement and lime kiln dust, then exhumed and transported
to a hazardous waste landfill. This stabilization was done because it
was the most practical option available at the time. No option for dis-
posal as a solid supplemental fuel was available in 1984 or 1985. After
May, 1990, untreated disposal of the hard and crumbly phase in a
hazardous waste landfill will not be permitted, due to the HSWA
land-ban hammers.
Contaminated Soils Phase Strategy
According to the 1985 closure plan, the contaminated soils below
the hard and crumbly phase were stabilized and mixed with clean soil
backfill and left on-site. Today, because of the HSWA land-ban hammers,
further treatment to remove coal tar residuals is required. Strategies
to remove coal tar residuals from contaminated soil are discussed in
the next section.
In the past, the methods described above for the 1984-1985 project
were sufficient to accomplish the remediation of an open coal tar lagoon.
Today, with stricter environmental regulation of fugitive air emissions
and land ban regulations, these remediation strategies will have to be
modified. CWM's Research and Development Group is actively
pursuing new and innovative technologies for dealing with each phase
of the coal tar lagoon remediation problem.
FUTURE COAL TAR REMEDIATION PROCEDURE
Supplemental Fuels Recovery
CWM is developing a devolatization/detackification process to convert
a VR-HC mix into a solid supplemental fuel, which can be utilized
as a coal substitute. The unit being developed will combine VR, HC
and a blend of proprietary inorganic/organic detackifying agents and
mix them in a totally enclosed, indirect, heated paddle mixer to produce
a coal-like solid supplemental fuel. The devolatization step in this
process removes BTX, thus rendering the fuel non-hazardous.
VR and HC components having a fuel value of 5000 Btu/lb or more
can be recovered and used as a solid supplemental fuel in industrial
processes. The 5000 Btu/lb criterion is set because 5,000 Btu/lb is
required to sustain combustion. Burning material of less than this value
for fuel is considered sham recycling. If a compound has less than this
fuel value and it is burned as fuel, technically it is being incinerated,
because additional fuel must be added to sustain combustion.
It is the opinion of the authors of this paper that any facility that
engages in this practice of sham recycling should be regulated as an
incinerator. Generators should be very selective from a liability stand-
point when choosing a fuels blender and energy recovery facility.
Solvent Extraction
Laboratory studies show that solvent extraction effectively removes
coal tar from contaminated soil and rocks. CWM has tested solvents
such as propane, pentane, Freon, methanol, trichloroethylene, and
triethylamine (TEA).
Solvent extraction processes are designed to remove contaminants
from contaminated soil by extraction into the solvent phase. The sol-
vent is then recovered and reused. Solvent recovery is necessary because
of its cost and fugitive emission control. The residual solvent in the
soil is driven off thermally and recovered.
Solvent extraction may be expensive because of the costs associated
with the hazard of using large volumes of solvent, the need to control-
fugitive VOC emissions, and material handling difficulties that may
occur.with the use of flammable organic solvents.
Since most of the town gas sites in need of remediation are in highly
populated urban areas, the use of large volumes of organic solvent may
be undesirable. The safety and toxicological risks of this technology
must be carefully considered before choosing this option.
RCRA / SUPERFUND ACTIVITIES 643
-------
Soils Washing
In soils washing systems, surfactant/water solutions are mixed with
the contaminated soil/rock in an agitated vessel. After sufficient agita-
tion, the contaminant/aqueous solution is removed from the soil by filtra-
tion. CWM has developed a proprietary soils washing process, in which
water and surfactant can be recovered without distillation and be re-
used in the process. Benefits of soil washing include: the biodegrada-
tion properties of the surfactants used; the absence of solvents; and the
non-thermal nature of the process.
Soils washing may be difficult, however, because of the need to emul-
sify the coal tar for good removal from soil. Hard pitch materials are
harder to emulsify than the light oil phase of the coal tar. Soils washing
will work for coal tar contaminated soil if the right situations exist.
Adequate testing of bench-scale systems and stringent economic analysis
must be completed before selecting this option.
Incineration
When a site has a significant amount of contaminated soil/debris that
occurs after a lagoon has been filled in or covered, incineration may
be the only practical remedial alternative. When using on-site
incineration, all phases of the material can be disposed of collectively.
Incineration, in some cases, may be required by the regulatory agencies,
CWM has developed PYROX, a transportable rotary kiln incinera-
tion system designed to handle these types of materials. The PYROX
unit has a horizontal, rotary kiln, primary combustion chamber into
which the contaminated soils are fed. The vertical, secondary com-
bustion chamber takes gases from the primary chamber and exposes
them to a temperature of 2200 T. Exit gases are scrubbed with a dry
scrubber to remove HC1 and particulates prior to discharge.
Incineration can treat all phases of coal tar effectively. However, it
is the most expensive treatment technology discussed in this paper.
CONCLUSION
In most cases, more than one technology should be applied during
town gas remediation projects. Certain technologies will work well for
certain phases of the contaminated material, but one technology will
not necessarily be the best solution for the whole cleanup. Careful testing
of these technologies in the laboratory must be performed before com-
mitting to a full-scale project.
Table 4
Viability of Contaminated Soil Remedies
Bloremediatlon
Thermal Desorptlon
Solvent Extraction
Soils Mashing
Incineration
Applicability/Effectiveness
Potentially Viable for low
Contaminant Concentration - Nay
not achieve needed levels of
clean-up.
Not Viable - Thermal Desorption
will not effectively remove high
boiling point PNAs from soil.
Viable Works well to remove
coal tar. Hatards of use of
large volumes of organic
solvents must be considered.
Potentially Viable Must do
treatabllity studies. Still
under development.
Very Viable Mill treat all
phases of coal tar and
contaminated soil.
Supplemental fuels are a cost-effective disposal outlet for all of the
coal tar phases having a high heating value. The viability of contami-
nated soil remediation must be closely examined. Table 4 summarizes
the viability of various soil remedies discussed in this paper.
When the final third HSWA land-ban treatment standards are enacted
in May of 1990, most coal tar will be listed as characteristically
hazardous because of CCW levels of BTX organics. This base will mean
that untreated land disposal of coal tar waste will be prohibited. Because
this waste is a final third waste, if no treatment standards have been
established by May of 1990, land disposal of this waste, treated or un-
treated, will be prohibited regardless of what treatment technology is
viable.
Until standards for the treatment and disposal of coal tar are set. Table
5 summarizes the recommended disposal options.
TibieS
Recommended Strategy for Owl Tar Lagoons Remediation
CQM! Tetr yrtitrt Ion
Light Oil:
Viscous/Rubbery:
Hard and Crumbly:
Contaminated Soil:
Recovery aa liquid supplemental foal
blending component. Mo auxiliary heating
is required. Control of VOC emissions 1*
necessary.
Development of solid supplemental fuel
utilising deteckificetion and
devolatillzation process. Control of VOC
emissions is necessary.
as above for viscous rubbery)
1. Incineration
3. Soils Hashing
1. Solvent Extraction
4. Bioremediation
As each coal tar segment is encountered, the appropriate remedial
remedy changes. This paper is intended to aid in the development of
a management strategy for the selection of feasible remedies at town
gas sites. This paper can be viewed as a road map for the remedy of
town gas sites. By dividing the remedy of town gas sites into different
phases, this coal tar problem can be conquered.
REFERENCES
1. Droegkamp. R. E , Schusskr. M.. Lambert. J. B and Taylor, a F, "Tar
and Pilch" in Encyclopedia of Chemical Technology, Kirk and Othmcr. \W.
22. 3rd Ed.. John Wiley and Sons. New tork, NY, pp. 564-600, 1981
2. Lafomara, J.P. et. a) . "Coallar: PblluuuMsof the Past Threaten the Future".
Proc. of lite I9S2 Hazardous Materials Spills Conference, Milwaukee, WI.
1982
3. Fbchtman. E. G. andCanwright R, T., "Closing a Coal Tar Lagoon," Amen
can Institute of Chemical Engineers Annual Meeting. New Tfork, NY, Nov.,
1987.
4. Kent. J. A . Kegfl's Handbook of Industrial Chemistry, 7th ed. Van Nosffland
Reinhold Publishing. New \brk, NY, 1974.
5. Austin, O. T . Shreeve t* Chemical Process Industries, 5th ed. McGraw-Hill
Book Company. New Ybrk. NY. 1984.
6. Fochtman. E. G.. and Cam/right R. T.. "Closing a Coal Tar Lagoon," Ameri-
can Institute of Chemical Engineers Annual Meeting, New M>rk, New Tfork,
Nov.. 1987.
7. Swanstrom. C. "X^Tl\X Low TenipcratureTraraportabteTltamw^Process
for Organic Contaminated Solids," Proc. of the 1989 HosMat Central Con-
ference, Rosemonl, IL, Mar. 15, 1989
644 RCRA / SUPERFUND ACTIVITIES
-------
Owner, Contractor, Government Relationships
John W. Buckley, P.E., J.D.
R. W. Beck and Associates
Seattle, Washington
INTRODUCTION
Superfund site cleanups are becoming more complex and involved,
particularly with regard to the relationship between the remediators
(Owner, Contractors) and the Regulators (Governments).
The relationship between the Owner and cleanup Contractor is
basically contractual and governed by the principles of construction
contract law. However, the relationship between the Owner/Contrac-
tor and the "Governments" can be relatively complex. Some of the
governmental entities that can be involved are Federal (U.S. EPA), State
(pollution control authority - usually both regional and headquarters
staffs), local County and City authorities [police department, health
department, fire department, building department, planning department,
legal department, community relations department, engineering depart-
'ment and utility departments (water, sewer, storm drainage, streets),
etc.] In addition, quasi-state authorities such as regional wastewater treat-
ment authorities and regional air quality authorities often are involved
in the cleanup.
Each of the governmental entity listed above has power to regulate
construction activity within the "scope of their authority." The local
authority's power usually is exercised by the granting of permits. Often
the "scopes of authority" of the various governmental units overlap or
are unclear. In this situation, the effect on the remediators can be either
conflicting permit requirements imposed by different authorities or in-
action on the part of the permitting authorities (each claiming another
agency is responsible for some action). In addition, the Governments
responsible for oversight of the cleanup (U.S. EPA or state pollution
control agency) often delegate review authority for portions of the
cleanup activity to other state and local agencies (i.e., regional
wastewater treatment authority, regional air quality authority, etc.).
Potentially superimposed over this governmental review activity is
"community action group" oversight. The formation and use of such
community action groups was encouraged by SARA; Federal funding
is available for their activity.
The purpose of this portion of the seminar is to apprise Owners/Con-
tractors of the general regulatory scheme imposed on hazardous waste
cleanup activities and some of the potential pitfalls they may face.
Experience with the Western Processing Superfund site cleanup located
in Kent, Washington, one of the largest site remediations (more than
$80 million) currently in progress, will be described relative to
Owner/Contractor requirements imposed by the local government.
REGULATORY SCHEME
In most site remediations currently in progress, either the Federal
government (U.S. EPA) or the State government (pollution control
authority) is the lead agency. The state would be a lead agency if the
remediation were conducted under a State "Superfund" program, or
site cleanup responsibility was delegated to the State by the Federal
government. Lead agency status usually is decided internally between
the Federal and State agencies. Thus, from the standpoint of the site
remediation Owner/Contractor, the lead regulatory agency is clearly
defined, and the set of rules/regulations governing the technical aspect
of the cleanup is known. What is less clear in a site remediation is the
status of the local governments (County or City). The balance of this
paper will deal with the legal powers of the local governments and how
the exercise of this power can be perceived by the local government.
Under the United States governmental system, all power emanates
from "the people." The people delegated certain specific powers to the
Federal government. The legislative powers, which are enumerated in
the U.S. Constitution (Article 1), include spending, foreign affairs, war
powers, immigration, taxation and commerce. The Constitution also
grants implied powers (Article 1, Section 8) which allows Congress
to regulate outside the enumerated areas so long as the result furthers
an enumerated power (necessary and proper clause). Powers not
delegated to the Federal government are specifically reserved to the
states or the people under the 10th Amendment.
State constitutions likewise define the relationship between "the
people" and state government for sharing of non-Federally delegated
powers. Each state is somewhat different in terms of specific power
allocation. However, certain non-Federal powers traditionally have been
reserved "to the people" (i.e., local governments). These areas of power,
as perceived by most local governments, include land use planning,
police protection, fire protection, building code enforcement, local trans-
portation, utility services, parks and recreation and community health
services. Thus, the local governments expect that if a Federal or State
site remediation activity impacts one of the above areas traditionally
regulated by the local governments, the local governments will have
control over that activity. Whether a local government does, in fact,
have control, and the extent of that control, can be the subject of pro-
tracted legal and political maneuvering. The Owner/Contractor should
be aware of the potential for an intergovernmental power struggle, par-
ticularly in areas traditionally regulated by local governments.
In addition to the power allocation issues, it makes good public rela-
tions sense for the Owner/Contractor to keep the local governments
involved in the cleanup process to the maximum practical extent. Local
governments tend to feel that they are the true grass roots representa-
tives of the people, and thus usually are concerned about the health
risks, timing and community impacts of the remediation effort. Over-
sight groups of private citizens, whose activities may be funded under
SARA, may be established for each remediation site. In general, these
groups either include members of local government or report to local
government. Also, their opinions often are considered newsworthy by
the press.
CONSTRUCTION MANAGEMENT 645
-------
WESTERN PROCESSING SUPERFUND SITE
The Western Processing Superfund site is located in Kent. Washing-
ton, about 20 mi south of Seattle, Washington. The area is heavily
industrialized, primarily with firms serving the aerospace industry.
Western Processing was operational during the 1960s and 70s,
providing chemical reclamation and recycling for materials generated
by over 400 public and private customers. Waste materials included
animal by-products, metal finishing solutions, oils, paints, solvents,
cyanide and battery acid. Due to spills and other releases, the materials
over time contaminated the soil and groundwater beneath the site and
Mill Creek which runs through the site. In 1983. the site was forced
by the EPA to close.
The site is underlain with an alluvial aquifer that extends to a depth
of approximately 150 ft. The aquifer is bisected by a 10) to 15-ft-lhick
semi-permeable stratum located at a depth of approximately 45 to 60 ft.
The upper portion of the aquifer is referred to as shallow groundwater,
and the lower portion is referred to as the regional aquifer. The shallow
groundwater and soil beneath the site are contaminated with over 90
contaminants including lead, zinc, cadmium, phenol, toluene, methy-
lene chloride, oxyzolidinone, trichloroethylene and other solvents. The
regional groundwater is largely uncontaminated.
The largest single contributor of waste to the site was the Boeing
Company. Boeing organized the negotiations between the Regulators
and approximately 200 PRPs. A consent decree was entered in U.S.
District Court in August, 1984, initialing Phase I of the remediation.
Phase I consisted of surface cleanup. Approximately 2,400 truck-
loads of various wastes were removed for off-site treatment and dis-
posal. Approximately 7,400 gal of dioxin-contaminated liquid were
treated using the potassium polyethylene slycol process. This process
destroys dioxin in a low-temperature, low-pressure reaction with no air
or water emissions.
Phase 0 of the remediation process began in April, 1987, and dealt
with subsurface cleanup. Approximately 22,000 yd' of specific waste
were hauled off-site for disposal at Arlington, Oregon, a U.S. EPA-
approved hazardous waste landfill. A slurry wall was installed around
the 16-ac site, extending down into the semi-permeable stratum (±50
ft). A vacuum groundwater extraction system using 206 wells was
installed along with an infiltration system. The groundwater extraction
system discharges to a treatment system consisting of a stripping tower
with a Calgon CADRE fume incinerator for volatile removal, and liquid
phase phenol oxidation/heavy metal precipitation. Precipitated heavy
metal sludge is dcwatered in a plate and frame press and hauled to the
Arlington, Oregon disposal site.
The general operational scheme is to pump, treat and re-infiltrate
the groundwater in order to remove the heavy metal contamination.
The pump- and treat-system has to operate for at least 5 to 7 yr and
may have to be operated for as long as 30 yr to achieve the desired
cleanup level.
GOVERNMENTAL ISSUES
The Federal and State Regulators (U.S. EPA and Washington Slate
Department of Ecology) have dealt primarily with enforcement of the
Consent Decree. The City of Kent, which was not a party to the Con-
sent Decree, desired to be involved in the remediation process since
it impacted their community and governmental services. Kent passed
an ordinance requiring remediation contractors to pay an annual permit
fee in order to fund the City's involvement in the remediation process.
In April, 1987, Kent retained R. W. Beck and Associates, a Seattle-
based consulting engineering firm, to provide technical oversight of
the remediation process.
In addition to reporting to Kent on remediation progress. Beck assisted
the Owner/Contractor and Regulators in dealing with a number of local
issues. The following is a summary of some key local issues and
suggestions for dealing with such issues on other sites.
Land-use planning
A key assumption of the risk assessment associated with establishing
the level of cleanup centered on the future use of the Western Processing
site and surrounding property. Cleanup levels were predicated on the
property retaining an industrial classification (i.e., residential type
exposure not anticipated). The local planning agency should thus be
involved in reviewing any assumptions regarding future land use rela-
tive to risk assessments.
Police protection
Site security is always a key issue in hazardous waste site remedia-
tion. If the Owner/Contractor expects the local police force to enforce
trespass law and the local city attorney to prosecute trespassers, then
the applicable law must be followed closely. It would be desirable to
get in writing from the city attorney's office the exact steps needed to
post the property for trespass enforcement.
Fire protection
Special training and equipment may be required for personnel
expected to combat a fire on a hazardous waste remediation site.
Frequent communication on potential problems and availability of
supplies such as foaming agents, etc., is a must. Evacuation plans and
air toxic control measures for the surrounding area also should be
coordinated with the local fire department.
Utility services
At Western Processing, plans are to use up to 60 gpm (86000 gpd)
of potable water for infiltration flushing. Added to this are potabte water
requirements for equipment seals and cooling water. The Western
Processing site remediation is one of the largest users of potable water
in the city.
Code enforcement
The city enforces the Uniform Building Code, Uniform Plumbing
Code and Uniform Fire Code. All construction should comply with
these codes in order to reduce Owner/Contractor liability for negli-
gent construction.
Local transportation
Since a hazardous waste site may be in limbo for 30 yr or longer,
it can have a major impact on planning for local streets and arterials.
Consideration should be given to future transportation plans when
designing a remediation system.
Parks/Recreation
A major recreational facility called the Interurban Trail extends along
the east side of the Western Processing site. This trail had to be closed
for over 2 yr because of remediation activity. Consideration had to be
given for alternative safe routing for people using the trail. Failure to
adequately consider safe rerouting could substantially increase the lia-
bility of the Owner/Contractor for injury suffered by trail users.
CONCLUSION
The relationships among the site owners), remediation contractors)
and Federal/State regulators are fairly well defined. The relationships
of the above parties to local governments are less well defined.
Activities normally regulated at the local level include land-use
planning, police protection, fire protection, utility services, code en-
forcement, local transportation and parks/recreation. Owners and Con-
tractors should comply with local regulatory agency requirements to
improve public relations and to minimize legal liability for asserted negli-
gent conduct.
646 CONSTRUCTION MANAGEMENT
-------
Evaluation of Procedures
For Claims Presentation and Resolution
John A. Cooney
Bureau of Reclamation
Denver, Colorado
ABSTRACT
The objective of this paper is to explore those procedures which can
be taken by a construction manager to minimize the number of claims
which result from contractual changes and to enhance their resolution.
This objective is achieved through a discussion of mitigative actions
that can be taken during both the pre-bid and the post-bit periods.
Pre-bid actions include: (1) developing clear and concise bid docu-
ments, (2) including contractual provisions that clearly define each
party's responsibilities relating to change, (3) establishing a contracting
strategy and (4) clarifying ambiguities.
Post-bid actions include: (1) assigning experienced contract adminis-
trators and inspectors to the construction management team, (2) estab-
lishing procedures for documentation of events, (3) educating team
members on instruction procedures and claims elements and (4) avoiding
arbitration and litigation whenever possible.
Adherence to these procedures will be material assistance in reducing
the number of claims and in achieving their resolution in an efficient,
equitable and cost-effective manner.
INTRODUCTION
Like death and taxes, two things that can be said with certainty about
a construction project are: (1) changes will be made during the course
of construction, and (2) the construction manager and the contractor
will seldom initially agree on the effect the changes have upon the
project. The objective of this paper is to provide a description of proce-
dures which can be implemented by a construction manager to minimize
the number of claims which result from contractual changes and to
enhance their resolution.
Failure to properly address all aspects of the claims procedure will
result in unnecessary performance and construction management costs.
This oversight can also lead to the development of an antagonistic rela-
tionship between the construction manager and the contractor which
increases the likelihood of disputes. Mitigative actions can be divided
into pre-bid categories.
Requisite pre-bid actions include: (1) developing bid documents that
precisely define the work to be performed and the expected site condi-
tions, (2) including contractual provisions which clearly define each
party's responsibilities relating to change, (3) establishing a contracting
strategy and (4) clarifying ambiguities during pre-bid meetings and in
formal responses to questions received during the bidding period.
Post-bid actions include: (1) assigning experienced contract adminis-
trators and inspectors to the construction management team, (2) estab-
lishing procedures to identify, document and track changes from
inception through the issuance of change orders, (3) educating all
members of the construction management team on pricing procedures,
the elements of contractor costs and the necessity for documenting all
instructions to the contractor and (4) avoiding arbitration and litigation
for claims settlement whenever possible.
DEFINITIONS
Prior to beginning this discussion of the procedures required for an
effective claims control program, the following definitions are offered
to avoid confusion.
Change:
A change is any modification to the guidance provided within the
contract documents. Therefore, changes encompass modifications to
specifications, drawings and other written or oral guidance. They may
be generated as a result of design modifications, field orders, excusa-
ble delays, actions of the construction manager or other contractors,
and differing site conditions.
Claim:
A claim is a written assertion by one of the contractual parties seeking
as a legal right the payment of money, an extention of performance time,
an adjustment of contract terms or other relief under the terms of the
contract.
Change Order:
A change order is the formal instrument for establishing agreement
between the parties to alter the contract price and/or performance time.
It is the contractor's responsibility to notify the construction manager
if an event occurs which he believes justified an increase in contract
price or an extension in contract time. Conversely, the construction
manager must initiate deductive claims.
The following discussion of claims procedures is directed specifi-
cally toward a construction management function for commercial con-
tracting. However, most of these procedures are equally applicable to
a governmental contract. Although changes will affect both cost-
reimbursable and fixed-price contracts, the impact is far more severe
on a fixed-price basis. Consequently, this discussion specifically ad-
dresses that type of contract.
PRE-BID ACTIONS
Completeness of Bid Documents.
Disagreement over the interpretation of contractual terms and con-
ditions and the scope of work is a major cause of contract claims, dis-
putes and litigation. A concerted effort should be made during the
preparation of the contract bid documents to ensure uniformity, com-
patibility and clarity of construction requirements and to accurately
define site conditions. The construction manager will be compensated
many times over for this effort by reductions in expenditures related
to claims and disputes and by reductions in performance costs. Achieving
CONSTRUCTION MANAGEMENT 647
-------
completeness and clarity rcquries the early freezing of design and the
performance of rigorous constructibility reviews during the design effort.
Many claims can be prevented by establishing a detailed definition
of the work scope and an accurate description of the payment bid
schedule items. Precise definition of contractor interface is particularly
important when multiple contractors will be involved. Including a
detailed description of payment items in the bid documents is a con-
venient and effective method of avoiding claims.
Contractual Provisions to Regulate Changes
Contractual provisions for change control procedures include both
general provisions and special conditions, both of which arc mandatory
for the orderly resolution of claims. A properly formulated contract
should, at a minimum, include provisions for the following:
• Changes in the work
• Changes in the contract price
• Changes in the contract time
• Decisions and disputes resolution
• Construction management responsibilities
• Differing site conditions
• Reference points
« Subcontractor Rejection
• Laws and regulations
• Related work at site
• Field Orders
* Defective work
• Suspension of work
• Contract termination
• Quantity variations and unit price adjustments
• Force account labor, equipment and mark-up rates
• Excusable delays
The most important change provisions are those which assure the
construction manager's right to order changes, prescribe equiiablc con-
tractor compensation and require the contractor to implement the change
without undue disruption to project progress. Establishing the joint
ownership of float is also a critical contractual requirement.
Contracting Strategy
There are two fundamental contract classifications: 0) fixed-price,
for which the contractor has primary cost responsibility and (2) cost-
reimbursable, for which the construction manager's client shares in the
cost responsibility. The selection of the appropriate contract type requires
an assessment of cost, time and quality priorities.
Fixed-price contracts provide the means for maximum cost control
but require the longest period of project performance time since
drawings and specifications must be completed prior to bid solicita-
tion. The use of a series of fixed-price contracts can reduce the total
project time by permitting construction to proceed on some initial work
packages concurrently with the design of subsequent work packages.
However, use of multiple fixed-price contracts requires a more defini-
tive work scope and closer coordination of contractor operations. This
method also creates the potential for increased claims resulting from
the simultaneous use of work areas by contractors and the interdcpen-
dency of various contgractors' work activities.
Clarification of Ambiguities
Invitations to bid should include a request for the submittal of written
questions from the contractors at the pre-bid meeting. In addition to
providing answers to these questions, the construction manager should
review his requirements for progress meetings, administrative submit-
tals, project control, quality control, health and safety, site security and
schedule compliance. Minutes of this meeting should be recorded and
issued as a solicitation addendum. Substantive questions received during
the bidding period also should be recorded and issued with their answers
as a contract addendum. Efforts to assure a mutual understanding of
contract language and intent will substantially reduce claims.
POST-BID ACTIONS
Experienced Staff
The experience of the administrative staff and inspection force is of
particular importance in minimizing claims and in achieving prompt
resolution and settlement. Staff personnel must be thoroughly familiar
with construction procedures and be aware of the manner in which a
change can impact the contractor's cost. Without this understanding,
administrators will not appreciate the full effect of their actions upon
the contractor's costs.
Change Documentation
The establishment of a procedure for documenting changes and poten-
tial claims is essential to the orderly settlement of the claims. Without
suitable documentation, the construction manager will be at a disad-
vantage in assessing the validity of the contractor's contentions.
Accurate project records should be maintained so that the day-to-day
work history can be recreated if necessary. Effective documentation
can be achieved by maintaining daily construction reports and a project
photograph log, performing time studies and establishing a correspon-
dence control system which prescribes procedures for logging,
serializing and filing correspondence and other written data.
Of particular importance to claims resolution is the maintenance of
a separate file for each incident that may result in a claim. A case file
should be established for each contractor with each incident assigned
a serial case file number. All pertinent information relevent to a specific
incident should be retained in the applicable case file. A confidential
case file log, similar to the example provided as Appendix A, should
be maintained to provide effective control of the required responses
to each incident.
Every effort should be made to promptly settle claims while the facts
and circumstances are current. Failure to quickly resolve claims will
generally result in a more cosily settlement. A construction manage-
ment policy of pricing and resolving claims prior to the commence-
ment of work, whenever possible, must be rigorously followed.
Retroactive pricing is undesirable since it promotes force-account
pricing and increases administrative control costs. However, where
situations make it unpractical to preprice the change or where opera-
tional conditions require the immediate execution of the change, a work
directive may be issued on a cost-reimbursable basis. This decision
should be based solely upon the nature of the uncertainties at the time
the change must be executed.
When work is performed on a force-account basis, the scope of work
must be clearly defined and daily records of the work effort maintained.
Appenix B provides an example of the type of form that should be
employed. These reports will provide invoice support for periodic pay-
ment by documenting daily agreement on labor, equipment and supply
costs expended while executing the change.
Agreement between the panics must be formalized in a change order
which includes the elements should in Appendix C. Forward-priced
change orders will be executed prior to the commencement of the work.
A formal change order would also be executed at the conclusion of force-
account work to document costs and to modify the contract price.
A final procedural requirement is the maintenance of a change order
log for each contract that contains the basic information shown in
Appendix D.
Educating the Construction Management Team
The members of the construction management team must be
thoroughly indoctrinated on the procedures necessary for suitable claims
documentation and control. The objective of forward pricing and the
necessity for only formal directives for field changes must be stressed.
Inadvertent directives to the contractor must be prevented, since such
instructions can result in costly and unnecessary claims.
All members of the construction management team should be
reminded of the various manners in which a change can affect a
contractor. A change may have both a direct and a consequential impact
upon cost and/or performance time. In addition, the timing of the change
can be critical. A change issued prior to the commencement of con-
648 CONSTRUCTION MANAGEMENT
-------
struction may be limited to only a possible schedule alteration plus an
increase in administrative processing cost. However, a change issued
during construction may require demolition, rework and redirection
of crews. This type of change also may have a significant impact on
schedule, crew costs and overhead. Furthermore, a single change can
be accomplished with less disruption than a package of changes or over-
lapping changes.
A change issued during construction may result in a cost increase
directly attributable to some combination of the following:
• Productivity degradation
• Delays
• Labor, equipment, materials and supplies incorporated in the original
work
• Labor, equipment and supplies in removing completed work
• Labor, equipment, materials and supplies associated with new work
• Non-productive periods during redirection of work
• Recovery scheduling
• Equipment standby costs
Consequential impacts of a change may include the cost increases
caused by the following:
• Production degradation of sequential activities
• Production degradation of concurrent activities
• Increased overhead costs
• Crash scheduling costs
• Changes to subcontracts
• Time value of money
• Change of work to unfavorable construction seasons
• Miscellaneous - bond and insurance
-small tools and consumables
-revisions to as-built drawings
-extra cleanup costs
-materials handling and disposal
-materials expediting
-increased warranty reserve
-escalation
The construction management team must be prepared to address con-
tractor allegations of cost increases and/or schedule delays resulting
from these situations as valid or invalid on the basis of factual substan-
tiating documentation. Perhaps the greatest obstacle to claims resolution
is the contractor's frequent demand for total cost recovery. The con-
tractor often fails to recognize his own contributing inefficiencies and
his claim may not supported by facts. In such instances, the construc-
tion manager must rely upon those individuals on his staff who are
most familiar with the details of the claim and on his own files and
information to arrive at an equitable offer for resolution.
Avoidance of Arbitration and Litigation
Arbitration and litigation are methods of resolving disputes which
customarily occur long after project completion. These remedies are
costly for both parties, require the expenditure of considerable resources
and necessitate the resurrection of past events, often by individuals un-
familiar with the circumstances of the claims. It is in the interests of
Appendix A
CONTRACTOR
CASE FILE SUMMARY DESCRIPTION:
UPDATE DATE-
CASE
NO.
DESCRIPTION
NOTIFICATION
Con-
tractor
Date /
/ Ref.
Const.
Mgr.
Date /
/Ref.
ESTIMATES
Con-
tractor
Orig. /
/ Rev.
Const.
Mgr.
Cone. /
/ Def.
Time
Contr. /
/ CM
DISPOSITION
1 Withdrawn or
Lapsed
1 Resolved -
No Change
1 Resolved with
Change
Disputed
Sht of
Change /
Order No. /
/ Amount
/ and Time
REMARKS
MOUNT 1
orecast /
/Exposure
CONSTRUCTION MANAGEMENT 649
-------
both parties to seek agreement as early as possible. Reaching early
agreement requires an attitude of compromise in lieu of combat and
the construction manager's continuous assessment of the contractor's
receptiveness to reach agreement from pre-claim submittal through the
trial and briefing stages.
CONCLUSIONS
To achieve a successful program for claims control and resolution,
a construction manager must endeavor to eliminate the potential causes
of claims which frequently are the result of differences in interpreta-
tion of contract documents and inadequate scope definition Success-
ful claims control, therefore, begins with the contract documents.
Clarity, uniformity and the inclusion of provisions to facilitate claims
resolution are mandatory.
Following contract execution, the construction manager must assemble
a project team which is skilled in contract administration and is familiar
with the requirements of the particular type of work to be performed.
The project team must be thoroughly trained in the manner in which
construction costs are influenced by change to avoid unnecessary con-
frontation. Accurate and complete project records are essential. This
documentation provides the construction manager with the means neces-
sary to respond promptly to a contractor's claim allegations and with
the ability to negotiate a settlement from a position of strength.
Applying the procedures that have been addressed will not eliminate
changes and resultant claims. However, these actions will be of material
assistance in reducing the number of claims and in achieving their
resolution in an efficient, equitable and cost-effective manner.
BIBLIOGRAPHY
I. Aroavu, D.P.. Ointbury, G. J., Simchik, M.S. and Pachter. J.S., Managing
Comma Changei. National Contract Management Association, McLeto
VA. 1987
2. The Business Roundiable. Contractual Arrangements. Construction Indus-
try Coil EfTectiveneu Project Report A-7, 1982.
3. Concepu and Methods of Schedule Compression. Publication 6-7. The Con-
struction Industry Institute. Austin, TX, 1989.
4 Contractor Planning for Fixed Price Construction, Publication 6-4. TheCon-
rtruction Industry Institute, Austin, TX, 1987.
5. The Impaa of Changes on Construction Cost and Schedule. The Construc-
tion Industry Institute. Austin, TX, 1987.
6, Impact of Hirioui Construction Contract Types and Clauses on Project ftr-
formance. PuMicauon 5-1. The Construction Industry Institute, Austin, TX
1986
7. National Contract Management Association. Claims. Disputes A Appeal*.
APPL-I. Active Procurement Program Library.
8. Project Control for Construction, Publication 6-5. The Construction industry
Institute, 1987.
9. Hfer* Packaging or Project Control. Publication 6-6. The Constructioa
Industry Institute, Austin. TX. 1989.
DAILY FORCE ACCOUNT REPORT
CHANCE ORDER
Scoot ol Q\*ra»
MMvxl ol P«ym«n(
A) otftM Mnru and condMant ol IM contract rwiun
Ortgkwl Contract Prtc*
Prtoi Clung* Ordcrad
Thrt OMngcd Ordw
Cwmn Condtcl Pne«
Conlrtckv
AinharMdRcp
TIM
0»lo
AMU
ONMT:
TIM
D*M
AB«»I
650 CONSTRUCTION MANAGEMENT
-------
Appendix D
CONTRACTOR:
CONTRACT CHANGE ORDER SUMMARY CONTRACT i>
UPDATE DAT
C.O.
NO.
DESCRIPTION
JO:
E: Sht of
SETTLEMENT AMOUNT
COST REIMBURSABLE
$
UNIT PRICE
$
LUMP SUM
$
TOTAL
$
TIME
DAYS
REMARKS
CONSTRUCTION MANAGEMENT 651
-------
Transportation and Disposal of Denver Radium
Superfund Site Waste
Elmer W. Haight, P.E.
Bureau of Reclamation
Denver, Colorado
ABSTRACT
The Bureau Of Reclamation (Reclamation) and Department of Energy
(DOE), both under contract to the U.S. EPA, have embarked on a
monumental task involving the excavation and disposal of an estimated
385.000 tons of radium-contaminated soil and debris in the metropolitan-
Denver. Colorado area. The efforts are divided into two separate con-
tract areas and are expected to continue well into 1992. DOE will handle
the excavation and site restoration, and Reclamation will provide (he
transportation and disposal of the waste.
The area of contamination has been designated as the Denver Radium
Superfund Site (DRSS). The contamination is believed to have come
from the residues from radium processing in Denver in the early 1900s.
Radium processing began in the United Stales about 1914. The
National Radium Institute (NRI) was located in Denver at about that
time. It extracted radium from Carnoiile, a radium-bearing material
available in Colorado.
The NRI refined the process for extracting radium and subsequently
closed about 1916 after successfully producing 8.5g of radium from
approximately 1.500 tons of ore. During this period, other radium
processing operators also were active in Denver.
The primary hazards of radium process residues known today are
that it produces radon gas as it degenerates, and if radioactive parti-
cles become airborne, enter the lungs or are ingested, they may cause
cancer.
The legacy of the NRI and the rest of the Denver radium industry
is present today in the form of tailings and unprocessed ore which, since
the 1920s, have been spread and used as Till under and around buildings,
as foundation material, parking lots, road base and otherwise
mishandled.
As called for in the Inicragcncy Agreement between the U.S. EPA
and the Bureau of Reclamation, Reclamation has provided the trans-
portation and disposal of the process residue (waste material). The con-
tractor performing the work for Reclamation is Chcm-Nuclear Systems,
Inc., of Columbia, South Carolina. The DOE contractors are excavating
the material and loading it into Chem-Nuclcar's containers for trans-
portation to the disposal facility.
Most of the material (approximately 85%) will be shipped in rail-
road gondola cars. The remaining 15% will be loaded into sealed 20-ton
containers which are trucked to the rail yard and placed onto railroad
flat cars for shipment. All material will be disposed of in Utah at a
facility operated by Envirocare of Utah, Inc., under a subcontract to
Chem-Nuclear.
The waste is considered Naturally Occurring Radioactive Material
(NORM) of Low Specific Activity. The primary radioactive con-
taminants arc Radium-226 (Ra'"1) and Thorium-230 (TH;").
Weights for payment and record purposes will be made on state-
certified scales. Scheduling and coordination, as well as recordkeeping,
are important aspects of the work and are essential in working with
the many agencies and entities involved in the DRSS effort.
INTRODUCTION
This paper discusses the Bureau of Reclamation's approach to
accomplishing the transportation and disposal aspects of the Denver
Radium Superfund Site (DRSS) work. Some background information
is presented to provide a better understanding of the overall project.
When Madam Curie discovered radium in 1898, she set in motion
a chain of events which left an unwanted legacy for following genera-
tions. By the early 1900s, radium was touted for its medicinal proper-
ties and ability to destroy or inhibit cell growth, and it became widely
used as a treatment for cancer. As a result, the demand for radium
skyrocketed, starting the radium boom of the early 1900s.
Prior to 1914, there was little or no domestic production of radium.
Rather, radium-bearing ore was shipped from the United States to
Europe where it was refined. About 1914, it became evident that
processing in the United Stales would be advantageous. The U.S. Bureau
of Mines entered into a cooperative agreement with a private corpora-
tion, the National Radium Institute (NRI). According to the agreement,
the Institute was to develop and operate a radium processing plant in
the United Stales. The demand for radium grew, and new sources for
radium were sought. Camotite. a radium-bearing material, was identi-
fied in Colorado about that time, and it seemed appropriate to locate
the NRI in Denver. Camotite provided the ore from which radium was
extracted by several processors in Denver from 1914 to about 1920.
The Denver radium industry remained strong until around 1920 when
very rich deposits of radium-bearing ore were discovered in the Belgian
Congo. The Denver producers could not compete, and the Denver
radium industry closed almost overnight.
The health-related implications of radium processing were not known
or considered a problem in those days. Although much of the radium
was recovered, process residues containing radioactive materials were
discarded.
In 1979, the U.S. EPA discovered a reference to the NRI in a 1916
United Slates Bureau of Mines report. Subsequent research revealed
the presence of many sites in the Denver-metropolitan area containing
material requiring remedial measures. One of these sites was the Robin-
son Brick Company, the location of the original NRI. This site contains
approximately 88,000 tons of contaminated material. Studies were sub-
sequently conducted to identify the potential hazards on all of the known
sites.
There are 44 properties that have low levels of radioactive contami-
nation that could potentially endanger public health or the environment.
The DRSS was placed on the NPL in 1983. Due to the enormity and
complexity of the DRSS, the U.S. EPA determined that response actions
652 CONSTRUCTION MANAGEMENT
-------
could be conducted in groups or operable units, and 11 operable units
were established. Nine of the 11 operable units are being serviced by
Reclamation's transportation and disposal contractor.
The work falls under the jurisdiction of the U.S. EPA Region VIII,
which is headquartered in Denver. The U.S. EPA's agreement with DOE
is to provide the final studies and site investigations and to develop
appropriate specifications for the excavation of the contaminated material
and restoration of each of the sites to as near the original condition
as possible. This is a difficult task, because each property where con-
taminated material is located is unique. The task involves open areas
in some cases; in others it involves removal of buildings and improve-
ments for later replacement after contaminated material is removed.
Strong efforts are made during all site work to keep existing active
businesses in operation. The logistics of this present a significant
challenge to DOE and UNC Geotech, the firm with which DOE has
contracted to provide the engineering and construction oversight for
the remedial action work.
The work involved for each operable unit is covered by its own con-
struction subcontracts. At least two operable units have undergone
excavation and stockpiling of material since 1988. Separate contracts
have been awarded for the loading of this stockpiled material, which
amounts to about 65,000 tons.
INTERAGENCY AGREEMENT
While the investigation and studies were underway and the U.S. EPA,
DOE and UNC Geotech were involved in determining the quantities
and full extent of the excavation/restoration portions of the work, the
U.S. EPA asked Reclamation to provide remedial action assistance in
the transportation and disposal phases of the work. The Interagency
Agreement (TAG), signed in September, 1988, provides for Reclama-
tion to finalize a solicitation including statement of work and to obtain
a contractor to perform all aspects of work involved in the transporta-
tion and disposal of material at an appropriate, properly licensed, per-
mitted disposal site. Reclamation is providing the Contract
Administration and Construction Management for the work, which is
expected to continue into mid-1992. Superfund money is made availa-
ble to Reclamation as needed during the performance of the work.
Most of the overall coordination with interested and affected parties
such as the owners and local, state and federal governments is handled
by U.S. EPA personnel. Matters involving cost recovery, obtaining State
of Colorado participation in funding and working with various entities
to assist in identifying and obtaining permits and licenses are handled
primarily by the U.S. EPA.
The matter involving cost sharing is important as it pertains to main-
taining a timely schedule of work, because remedial work could not
start on operable units until all agreements were finalized. Schedules
were directly tied to signing these agreements.
QUANTITIES AND LOCATIONS OF WASTE MATERIAL
Since Reclamation involvement started in 1988, the estimated total
amount of material to be transported has risen from 140,000 tons to
the present estimate of 385,000 tons. This increased amount of con-
taminated material is due to better information further defining limits
of contamination at each site. Determining the depths and lateral ex-
tent in some cases is quite difficult. Access to some sites is limited,
buildings remain in place and the sheer magnitude of the project all
make accurate computation of quantities difficult.
Of the nine operable units involved in Reclamation's transportation
and disposal work, the estimate of material from the smallest unit or
property within a unit is 160 tons. The largest operable unit contains
approximately 158,000 tons. Transportation and disposal service must
be provided to a wide variety of areas from a restaurant franchise to
a large scrap metal processing facility covering several city blocks.
CONTRACT INFORMATION
For the transportation and disposal work, Reclamation chose a
"requirements-type" contract. "Delivery Orders" will be made against
the contract as the work progresses.
A solicitation for bids was issued in November, 1988. The technical
qualifications of the firm receiving the award were of paramount
importance. Price was also of great importance. Interested firms were
asked to submit separate proposals, one for technical evaluation, and
one for price evaluation; the technical proposals carried 60% of the
total available points and the price 40%.
Technical proposals from the firms were evaluated by a committee
of professionals, performing each review without discussion among
themselves. Following the independent review and scoring, the com-
mittee met to discuss the proposals. Consensus scores were arrived at
for each item rated as it compared to the preestablished evaluation
standard.
After best and final proposals were submitted and evaluated in the
same manner as the initial proposals, a contract was awarded to Chem-
Nuclear Systems, Inc., of Columbia, South Carolina, a subsidiary of
Chemical Waste Management, Inc. Chem-Nuclear has been in busi-
ness since 1969 is highly qualified in the radiological waste disposal
field and has an excellent transportation safety record for this type of
material. The contract value is expected to be about $70 million if the
final quantity of material is near the 385,000 tons presently estimated.
Because it is a per ton price, the contract value will change depending
on the final quantities involved.
The major subcontracts involved under Chem-Nuclear's contract
include trucking and the disposal facility. The disposal facility is Enviro-
care of Utah, Inc., a facility located approximately 80 mi west of Salt
Lake City, Utah.
The base contract is set up to provide for transporting and disposing
of material from time of mobilization through Sept. 30, 1989. Option
years will include in sequence the fiscal years (Oct. 1 through Sept.
30) of each year until Sept 30, 1992. Chem-Nuclear's proposal con-
tained somewhat different prices to perform the work for each
succeeding year. The Government will place orders against the contract
based on the quantities to be hauled and the prices submitted by the
contractor for each calendar period of performance.
The quantities estimated by UNC Geotech are "in-place" volume.
Through experience, a conversion factor of 1.6 tons/yd3 was estab-
lished and applied to this project. The total estimated volume of material
is 258,000 yd3. Applying the conversion factor, this yields to the
385,000-ton estimate for total material. The contract was awarded
May 15, 1989, with the first Delivery Order issued June 13, 1989, for
$9.7 million, covering transporting and disposing of 57,500 tons of
material.
When the contract was awarded, it was anticipated that up to approxi-
mately 20% of the material might be hauled by Sept. 30, 1989. Due
to delays in the work for a variety of reasons, this figure will be sig-
nificantly less than originally estimated. The option years were expected
to include approximately 155,000 tons in 1990, 110,000 tons in 1991,
and 40,000 tons in 1992. These amounts,too, may change due to the
late start in 1989. Actual shipping of material started in mid-August 1989.
The 20-ton containers are top-loaded, containing top and end covers
with waterproof gaskets to prevent dust from escaping. These containers
have been impact-tested by Chem-Nuclear and the railroads to ensure
continued integrity during accident conditions.
The Remedial Action contractor (UNC Geotech and its excavation
subcontractors) will load all containers, mainly using front-end loaders.
In the case of loading gondola cars from existing stockpiles, large units
capable of loading large quantities of material in a short time will be
used. Some areas do not contain stockpiles and will be loaded at the
same time they are excavated. This is a slower process using smaller
loading equipment. Some material will be excavated from inside
buildings after floors are removed; this process is understandably slow
and better suited to the use of the smaller 20-ton containers.
Loaded rail cars are decontaminated by UNC Geotech as they leave
the operable unit. Gondolas are then switched and start their journey
to the disposal facility by Burlington-Northern tracks to Speer, Wyom-
ing, where they are switched to Union Pacific to continue to Enviro-
care's disposal facility. The disposal facility has direct rail service and
has easy truck access from U.S. Interstate Highway 80. The cars wait
on a siding until test results allowing disposal are received.
Material from operable units not served by rail is loaded into the
CONSTRUCTION MANAGEMENT 653
-------
20-ton containers. Chem-Nuclear has provided a transportation termi-
nal in Denver, located at 1960 A 31st Street, where empty containers
are stored and released as needed to operable units for loading. After
loading, the vehicle and container will be decontaminated by UNC Geo-
lech and will travel back to the transportation terminal for weighing.
The container then proceeds to the railroad's mtermodal yard for loading
on flatcars for the trip to Salt Lake City, Utah. It is then picked up
by truck and transported to a holding area at Envirocarc to wait for
test results allowing disposal.
Truckers must meet stringent qualification requirements. Vehicles are
inspected daily. City routes have been established to avoid residential
and school areas, and all routes meet the approval of local Transporta-
tion Engineering Departments.
Security is provided at the transportation terminal 24 hr/day. The
station is manned and is enclosed by a chain-link fence. No loaded con-
tainers will be held at the station. They will only pass through for
weighing and recordkeeping purposes.
All containers must be weighed using state-certified scales manned
by state-certified weighmasters. The weights will be used for payment
purposes; they also will provide a (actual record of how much material
originated at each operable unit. For gondola cars, rail scales capable
of weighing cars as they travel slowly over the scale area are used.
A sign is located on each container and gondola car. showing that
it is dedicated to the transportation of Denver Radium waste and must
not be loaded with any other materials. The sign given shows a long-
distance, toll free number to contact for information or notification in
case of problems.
The bid schedule contains only four pay items. The most significant
pay item is the per-ton, all-inclusive price for transporting and disposing
of waste. Other items include holding loaded containers while waiting
for waste certification test results, moving empty containers from one
unit to another to accommodate loading schedule changes and returning
loaded containers to the unit where loaded in the event the material
falls outside of the waste classification limits of the solicitation.
DESCRIPTION OF THE MATERIAL TO BE HANDLED
The waste is considered Naturally Occurring Radioactive Material
(NORM) of Low Specific Activity. It is not considered "radioactive"
under the Department of Transportation's (DOT) definition in 49 CFR
173, but the contract requires that certain portions of those regulations
be followed in transporting waste. Much of the material looks like or-
dinary soil, and the debris is mainly building materials, pavement
chunks, tree stumps and similar items.
The primary radioactive contaminants include Radium-226 (Ra"*)
approximately 100 picocuries per gram (pCi/g). with very limited
amounts, possibly 2 yd', containing up to 65,000 pCi/g. There is also
Thorium-230 (TH2") approximately HX) pCi/g with very limited
amounts, possibly 2 yd' containing up to 167,000 pCi/g.
Minimal amounts of asbestos-contaminated debris will be present
at times. Waste may also contain trace amounts of other non-radioactive
contaminants; however, it is not expected that the waste will be classi-
fied by characteristics or listed as hazardous waste under RCRA and
40 CFR Part 261. It is doubtful the waste will contain PCBs in concen-
trations of SO ppm or more.
SAMPLING AND TESTING
The sampling and testing program set up and conducted by the U.S.
EPA, DOE and UNC Geotech for waste certification provides needed
information concerning the character and composition of the waste.
Representative sampling will be done at the time of loading; thus, a
determination can be made concerning the average concentrations of
Ra226 and TH2!0 in the waste and to otherwise determine if the waste
is acceptable to the disposal facility.
TRANSPORTING THE WASTE
Chem-Nuclear plans to transport at least 85% of the material in
100-ton railroad gondola cars and the other 15 % in smaller containers
mainly of 20-ton capacity. The sampling and testing procedures will
accommodate these containers. Samples will be analyzed by the
Opposed Crystal System (OCS) gamma-ray spectrometer. The radium
concentration determined by the OCS will be used to confirm that the
average radium concentration does not exceed the maximum allowed
by the disposal facility, TH2n testing and numerous other tests are to
be performed as appropriate. Split samples will be provided to the dis-
posal facility for comparative testing upon their request. As test results
become available, containers will be released for disposal.
Since the first Delivery Order, Chem-Nuclear has been working
intensely, improving old railroad spur tracks and installing new ones
at two major operable units. This construction not only involves co-
ordination among the railroads, owners and others, but also involve*
coordination with UNC Geotech to ensure that the transportation phase
remains compatible with the loading operations. Railroads need to
provide the necessary switches, track and schedule availability of gondola
cars.
Operable units where rail service is not available, or where it is not
feasible to construct spur track into the areas, will be served by trucked
roll-on, roll-off, 20-ton containers.
All containers must meet DOT requirements for shipping radioactive
waste. They must be closed, tight containers set aside for exclusive use
for Denver Radium Supertund Site wastes. If the material is such that
it will stick to the gondola floor, the gondola car floor will be lined
with 6-mil polyethylene sheets. All can will be filled, and steel clad
foam covers will cover the entire car's top. The covers weigh approxi-
mately 1,200 Ib and are lifted on and off by a small forklift. Disposa-
ble Trak-Pak coven were used on some initial shipments until the
steel-clad foam covers were available.
DISPOSAL FACILITY
Envirocare of Utah, Inc., was chosen by Chem-Nuclear as the only
operating NORM waste disposal facility in the country that can receive
radium waste in bulk form. It has been used to receive material from
several sources including at least 2.5 million yd1 of mine tailings. It
became fully licensed in Feb. 1988. After years of comprehensive
studies, this disposal site was selected by DOE and the Stale of Utah
as the best of 29 potential sites in Utah. The facility is designed to handle
over 20 million tons of contaminated material.
The facility lies above a substantial clay layer which provides a good
bottom seal for the cells. The percolation rale through the layer is
extremely low. The facility is (ar from surface water or potable ground-
water. The DRSS cell will be excavated several feet down from (he
ground surface in an area approximately 600-ft wide by 800-ft long.
It will be filled layer by layer with waste until all waste under the con-
tract has been deposited in the cell.
Rail cars as they arrive will be held on Envirocare's railspur siding,
capable of holding more than 250 railcars at one time, until official
clearance to dispose of the material is received. The cars then proceed
to the area where the covers are removed, and onto a rollover machine
where each car is secured in the machine and turned over about 150°
to dump its contents onto a concrete pad beneath the machine. The poly-
ethylene liner, placed into the cars containing waste that may stick to
the gondola floor, facilitates dumping of these loads. Cycle time is about
6 min/car.
The waste is then loaded into dump trucks with a front loader for
the 4,000-ft trip to the cell. The dumped loads are spread into approxi-
mate 12-in. lifts, moistened if necessary to facilitate compaction and
control dust, and rolled with a standard roller to at least 90% of labora-
tory maximum dry density using the standard Proctor Method ASTM
D-698. When debris is present, it will be distributed so that adequate
space is provided for proper placing and compacting.
Dust suppression is an important safety consideration for this material.
If dust is present during the unloading, hauling and depositing process,
appropriate respiratory protection must be worn by the workers.
The 20-ton containers trucked to the facility will be held in storage
until the waste material is cleared for disposal. Material from the con-
tainers will then be dumped directly into the cell for spreading and com-
paction.
All containers are decontaminated using a high-pressure washer prior
to being released for return to Denver. Only the outsides need be decon-
654 CONSTRUCTION MANAGEMENT
-------
laminated, since the containers will be covered for the return trip and
reused for Denver waste. At the end of the job, the entire container,
inside and out, must be cleaned as necessary for the container to be
released for nonrestricted use.
The completed cell will be topped with a 7-ft layer of compacted clay
to provide a radon barrier. A 6-in. layer of gravel bedding topped with
18 in. of cobbles will provide the top and side slope erosion protec-
tion. A drainage ditch and Operation and Maintenance road will
surround the cell. The disposal area is designed to be relatively main-
tenance free for up to 1,000 yr.
The facility can accept waste 12 mo a year. The average precipita-
tion is only 5 in./yr so downtime due to heavy rains and snow is minimal.
Long-term assurances by trust agreement are provided for the con-
tinued maintenance of the facility. The facility is appropriately licensed
in accordance with the requirements of 40 CFR 192(a), is fully approved
by the State of Utah and is under its constant monitoring and inspec-
tion. Disposal activities at the site are in accordance with CERCLA,
Section 121(d)(3). Groundwater and air monitoring measures are
thorough.
PERSONNEL PROTECTION
The work is little different in many respects than other work involving
heavy equipment. The use of heavy equipment coupled with the special
hazards associated with radioactive materials and possible other con-
taminants, makes safety considerations of great importance. The con-
tract requires Chem-Nuclear to abide by all applicable regulations, the
most notable being OSHA, 29 CFR 1926/1910. In addition to these regu-
lations, Reclamation's Construction Safety Standards must be followed.
These standards closely parallel the OSHA regulations so do not sig-
nificantly impact the contractor. The contractor submitted an all-
inclusive safety program specific to the work before transportation and
disposal work began.
In addition to the typical personnel protective measures, any person
working on the Operable Units must have attended a 40-hr Personnel
Protection and Safety course as required by SARA and must have had
a recent (within the last 12 mo) physical examination meeting the SARA
and OSHA requirements including a baseline analysis for heavy metals.
The site workers must be certified to use respiratory protective
devices. When these devices are needed on the job, they are provided
by UNC Geotech. Prior to work on the operable units, all employees
also are required to attend an additional 4-hr training session conducted
by that firm. This training is specific to the operable units and covers
more in-depth information on handling radioactive materials of the type
expected to be encountered here.
External thermolyminescent dosimeters (TLDs) must be worn by all
site workers. UNC Geotech provides the TLD service, and the
dosimeters should never leave the site. They are picked up when a
worker enters the restricted area and are left at the guard shack when
he leaves.
The usual gear worn by workers (such as hardhats, foot gear, safety
glasses and hearing protection) is provided by Chem-Nuclear for its
employees.
All areas within the DRSS that contain radioactive contamination
or other identified potentially hazardous materials are considered to
be restricted for the purposes of access control. Only trained person-
nel are allowed in the area. No eating, drinking, smoking or chewing
of any substance is permitted. Even chewing on a toothpick or applying
lip balm are not allowed. Everyone must sign in and out on the access
log. A monitoring device (frisker) must be used each time anyone leaves
the area. The monitoring equipment is provided, maintained and
calibrated by UNC Geotech.
Once vehicles, tools or equipment enter the site, they may not be
removed until certified clean by UNC Geotech. Workers leaving res-
tricted areas must be monitored for contamination and must decon-
taminate their work clothes and/or wash their faces and hands, if
necessary.
PERMITS AND LICENSES
Chem-Nuclear obtained all local, state and federal permits and
licenses required from each governmental body having jurisdiction over
the transportation of the waste by virtue of the waste originating, passing
through or ending in their jurisdictional region.
PUBLIC RELATIONS
Public relations aspects of the work are highly important. When the
subject of radioactive waste comes up, the public perception is that it
is highly dangerous material. In the case of the Denver Radium Super-
fund Site material, the contamination averages approximately 10% of
the value to be considered radioactive by DOT guidelines.
Meetings with various groups helped dispel fears and were very im-
portant to the timely completion of the work. Contacts have been made
with local groups in the vicinity of the transfer station and also with
the cities and communities along the Colorado, Wyoming and Utah
material transportation routes. Fears subside to a great extent when
presented with the facts concerning the nature of the material and when
details of the Emergency Preparedness Plan are discussed.
SCHEDULING AND COORDINATION
The solicitation contained a master schedule for the work. This
schedule was intended to present only an indication of the sequence
and duration of the work expected for the operable units involved.
Weekly scheduling/coordination meetings are conducted involving the
U.S. EPA, UNC Geotech, Chem-Nuclear and Reclamation. These
meetings are very helpful in discussing progress in mobilization and
preparatory work and are a valuable tool as the work progresses in
providing a coordinated schedule whereby Chem-Nuclear will be aware
of where and how many containers will be needed for the next week.
Containers are to be supplied to best accommodate the schedule of the
loadout contractors.
Because the project is still in the early stages, DOE is still advertising
and awarding contracts for loading the waste. Chem-Nuclear is haul-
ing from only three operable units. As more of DOE's contracts get
underway, more operable units will be ready for waste transportation,
and scheduling of containers will become much more difficult.
Because waste may be hauled from as many as six operable units
at one time, so a long-range, 30- to 60-day forecast schedule is neces-
sary so there is some advance planning opportunity. In the early stages
of the job, funding for additional operable unit work was caught up
in the lack of agreement between the State of Colorado and the U.S.
EPA on the State's 10% contribution. This problem hampered the prepa-
ration of a meaningful long-range schedule.
With the many entities involved, the scheduling/coordination meetings
are essential. They provide an opportunity for the group to discuss
current problems and share ideas and information to help foresee future
problems.
The transportation and disposal work is expected to have peaks during
the better construction seasons and to slow significantly in winter
months. This is a natural tendency. Any effort to level out the hauling
schedule to eliminate the peaks' makes scheduling much easier. Chem-
Nuclear must provide an adequate number of containers to handle the
peak periods and still provide for transportation time and holding
periods. In slack times, the containers may be idle.
RECORDS AND REPORTS
Record-keeping in connection with Superfund work is very impor-
tant. Chem-Nuclear is required to report weekly on the tonnage handled
during the week from each operable unit. The report must contain the
shipment/container number, the date the container was sampled for
testing, the date it was released for shipment, the date it was actually
shipped and the date it was unloaded at the disposal site.
Records must include the containers that are returned to an operable
unit for any reason, including being overweight or because they con-
tain mixed waste. Chem-Nuclear must also supply copies of documen-
tation such as manifests. The location of each shipment/container at
the end of the week, whether in transit, in staging, or if disposal is
completed, must also be included. There is an annual report require-
ment that summarizes the activities for the year. A Health and Safety
Weekly report noting reportable occurrences for the period also is
CONSTRUCTION MANAGEMENT 655
-------
required. A final report is to be prepared when all work is completed.
CONCLUSION
The Bureau of Reclamation has utilized its knowledge of construc-
tion contracting to provide the support needed by the U.S. EPA in
accomplishing the transportation and disposal phases of the Denver
Radium Superfund Site work.
Reclamation's contractor. Chem-Nuclear, is successfully servicing
DOE's remedial action contractors by providing the types of containers,
in the required quantities for loading. The transportation and disposal
work is proceeding without significant problems.
ACKNOWLEDGMENT
Certain portions of background information for this paper were ob-
tained from various U.S. EPA documents and fact sheets. These sources
of information were of great help in developing this paper. The U.S.
EPA personnel whose work was used in some way include, John M
Brink, Holly Fliniau. Phillip C Nyberg, Sonya Pennock, Katharine
J. Teter and Timothy R. Redder. Information also was obtained from
Tim Hodgens of Chem-Nuclear Systems, Inc.
REFERENCES
I. U.S. EPA Final Draft. Remedial Investigation • Denver Radium Superfund
She SI-8LOIA Apr. 30, 1986.
2 U.S. EPA "ftct Sheett."
3 Solicitation No. 8 SP-81-15l3a Transportation and Disposal Service*—Denver
Radium Superfund Site. Denver. Colorado.
656 CONSTRUCTION MANAGEMENT
-------
1989 EXHIBITORS
3M Environmental Protection Products
3M Center. Bldg. 223-6S-04
St. Paul, MN 5 J144-1000 612/736-5335
3M Company—Environmental Protection Pro-
ducts—3M Foams. 3M Foams have proven their
suppression effectiveness during hazardous ma-
terial cleanup that involves release of volatile or-
ganic compounds (VOC), air toxics, odors and
dust. These water-based foams conform to the
terrain and last hours, days and even weeks, de-
pending on the site requirements.
All-Pik.Inc.
2260RosweUDr.
Pittsburgh, PA 15205
412/922-7525
SturdeeSeal Shipper—performance tested.
D.O.T. exempt packaging, VOA vials, bottles,
bags, overpack drums, pails, cans, sample con-
tainers, safety coated bottles.
ANDCO ENVIRONMENTAL
PROCESSES, INC.
595 Commerce Dr.
Amherst, NY 14150 716/691-2100
Wastewater treatment systems to remove heavy
metals, fluorides, phenol and other organics
from industrial wastewater, contaminated
groundwater and leachate. Also a portable
heavy metal pilot unit.
ARAMSCO
1635 Imperial Dr.
Thorofare, NJ no zip 609/848-5330
ARAMSCO specializes in safety products for
the hazardous environment. Introducing the
Blastrac—a portable shotblast cleaning system
for removing contaminants such as PCB, as-
bestos and radiation from concrete or metal
floors.
Acres International Corporation
140 John James Audubon Pkwy.
Amherst, NY 14228-1180 716/689-3737
Acres provides waste management expertise to a
wide variety of industrial firms, utilities and gov-
ernment agencies—federal, state and local. Site
investigations, permitting and regulatory compli-
ance evaluations, remedial investigations and
feasibih'ty studies, conceptual and detail design,
and construction supervision are among the com-
prehensive services offered. Acres offers a multi-
disciplined and experienced team of geologists,
hydrogeologists, chemists, biologists, geotechni-
cal, chemical, civil and hydraulic engineers, and
support staff to successfully complete a variety
of waste management projects.
AlrScp Corporation
84 Aero Dr.
Buffalo, NY 14225
1/800/426-0212
Adsorption Systems Inc., (ASI)
P.O. Box 387
Millsonn, NJ 07041
201/762-6304
Activated carbon adsorption systems and Re-
activation Services—Adsorption Systems Inc.,
(ASI), provides potential clients with complete
adsorption/reactivation programs that utilize ac-
tivated carbon for the removal of organic con-
taminants from process air and liquid streams.
Such services include the following:
• On-Site Adsorption Systems custom designed
to fit each clients specific needs.
• Carbon Transport Systems efficiently utilizing
trailers specially engineered for environmental
safety.
• Off-Site Segregated Reactivation Services
which receive the highest quality reactivated
products.
Advanced Environmental Technology Corp.
Gold Mine Rd.
Flanders, NJ 07836 201/347-7111
AETC is a full service company offering packag-
ing, transportation and disposal for virtually all
types of drum and bulk chemical wastes. Lab
chemical packaging, hazardous waste and site
cleanup, reactive and explosive disposal and
24-hour emergency responses are also part of
AETC services. Branch facilities are located in
New Jersey, New York, Pennsylvania, Massa-
chusetts and North Carolina.
Advanced Sciences Inc.
2620 San Mateo NE, Suite D
Albuquerque, NM 87110
505/883-0959
Professional and technical services firm special-
izing in environmental services, hazardous waste
management and advanced technologies.
AirSep Oxygen Generators utilize a unique Pres-
sure Swing Adsorption (PSA) air separation pro-
cess. AirSep manufactures generators with flow
rates of 0-40,000 ftVhr and discharge pressures
of 0-4,000 psi. The generators can completely
substitute for cylinder or liquid oxygen applica-
tions. They are safe, reliable, and carry a lifetime
service warranty.
All American Environmental Corp.
140 53rd St.
Brooklyn, NY 11232 718/492-7400
All American Environmental Corporation pro-
vides state-of-the-art hazardous waste incinera-
tion service. Two complete 43,000,000 BTU/hr.
transportable systems are available for inspec-
tion and project placement. All American En-
vironmental offers a wide variety of commercial
options including full service operation, joint
venture/teaming and system leasing.
Alliance Technologies Corporation
213 Burlington Rd.
Bedford, MA 01730 617/275-9000
Alliance develops detailed inventories of en-
vironmental contaminants, designs control and
treatment systems, evaluates environmental and
health damages from hazardous waste facilities,
and assesses new technologies. Alliance also con-
sults with government and private managers on a
range of environmental policy and management
issues.
American Colloid Company
1500 West Shure Dr.
Arlington Heights, IL 60004-1434 312/392-4600
American Colloid Company, the worlds largest
producer of bentonite clay, offers a wide variety
of environmental products and services. The
PureGold product line consists of bentonite pro-
ducts designed specifically for groundwater mon-
itoring construction and installation. Our En-
vironmental Division has bentonite based pro-
ducts for landfill and lagoon liners, slurry wall
construction, and waste stabilization.
EXHIBITOR PROFILES 657
-------
AmcrkM International Gitmp
2003 Market St.. Suite 2800
Philadelphia. PA 19103
215/981-7117
Meeting the insurance needs of industry by pro-
viding Environmental Impairment Liability and
associated coverages, for companies involved in
related fields, through experienced underwrit-
ing, comprehensive risk management, and dedi-
cated claims handling.
American Laboratories A
Research Service*, IK.
P.O. Box 15609
Hattiesburg. MS 39402 601/264-9320
American Laboratories A Research Services, Inc.
is a full service environmental analysis labora-
tory which provides complete characterization
and classification of soils, water, groundwater,
sludges, run-off, and air emissions. Instrumen-
tation includes OC, OC/MS, TCLP, HPLC,
AA, ICP. TOC, TOX, and complete wet labor-
atory capabilities.
Aatericaa NaKEM Corporate!
c/oAnalytiKEM, Inc.
28 Springdale Rd.
Cherry Hills, NJ 08003
609/751-1122
Chemical and radioactive waste process systems
engineering and management services. Cleanup,
removal and treatment of hazardous materials;
certified analytical laboratory and field services
are provided through subsidiary companies;
AnalytiKEM (609) 751-1122 and (803) 329-9690;
a full service environmental laboratory specializ-
ing in sample collection of environmental and
hazardous waste samples; ThermalKEM (803)
329-9690; Incineration of hazardous toxic ma-
terials employing state-of-the-art emission con-
trols; CyanoKEM (313)353-5880; Chemical
treatment of cyanides, suifides, corrosives, tox-
ics, and plating wastes and a wide variety of in-
organic wastes; An Emergency Response Treat-
ment Services Team (ERTS) (803) 329-9690; Co-
ordinate all of these integrated services on an
expedited basis. Repackaging and removal serv-
ices are also available.
Amoco Fabrics and Fibers Company
900 Circle 75 Pkwy., Suite 300
Atlanta, OA 30339 404/984-4480
Amoco Fabrics and Fibers Company manufac-
tures both woven and nonwoven geotextile fab-
rics for stabilization, filtration, heap leach, land-
fill and other containment applications. The
nonwoven products are commonly used in con-
junction with geonet and geosynthetic liner ma-
terials. A national distribution system proves to
serve immediate deliveries required in the con-
struction industry.
Aswan, Blckford A Flake
P.O. Box 239
West Springfield, MA 01090
413/733-O79I
We specialize in the recruitment and placement
of all types of environmental engineer!, mana-
gers, and executives nationwide.
Aqaa Tech Environmental Consultants, IDC.
181 S. Main St.. P.O. Box 436
Marion. Ohio 43302 614/382-5991
Aqua Tech Environmental Consultants, Inc.
proves accurate and precise analytical data, on a
timely basis, at competitive prices to industrial,
governmental and private clients. Aqua Tech's
services include complete capabilities for organic
and inorganic analysis, bioassay/biomonitoring,
sampling and mobile laboratory analysis.
Art's Manufacturing A Supply
103 Harrison
American Falls, ID 83211
1/800/635-7330
MAS will be displaying a full line of our soil
sampling equipment. We are happy to Introduce
our new patent pending liquid sampler and soil
gas vapor probe. Stop by booth 11012 for more
information.
Associated Dotga A Mfg. Co.
814 North Henry St.
Alexandria, VA 22314
703/549-5999
Associated Design provides suitable laboratory
equipment for TCLP, EP-Toxidty and Liquid
Release testing of solid waste. Featured products
include the Zero Headspace Extractor (ZHE) for
collection of volatile contaminants, two bench-
top filtration units, the new Liquid Release Test
device, and large-capacity rotary agitators which
hold bottles, separatory funnels or ZHEs.
BAY Watte Sdcact and Technology Corf.
4370 W. 109th St.
Overland Park. KS 66211 913/339-2900
A Black A Beatch Company, BVWST provide*
complete hazardous waste management services.
Including RJ/FS, design plans and specs, imple-
mentation oversight, RCRA services, regulatory
and permit support, and litigation assistance.
Other specialties include waste treatment, PCB
transformer replacement, public health evalua-
tions, facility closure services, environmental
audits, and community right-to-know planning.
BCMEagtsMcn
One Plymouth Meeting
Plymouth Meeting. PA 19462
215/823-3800
Quality engineering in hazardous waste manage-
ment and control; groundwater studies, geophys-
ical surveys, remedial design engineering, super-
fund site investigations, facility permitting,
closure plans, real estate contamination assess-
ments, asbestos surveys and analytical services.
BES Environmental Spcdalbl tec.
82-86 Boston Hill Road
Larksville, PA 18651 717/779-5316
Emergency and remedial construction services
for industry and government in site restoration;
characterization, excavation, transportation and
disposal of liquid, drummed and bulk wastes; se-
cure landfill and lagoon construction/closures;
facilities decontamination and demolition; de-
watering; storage tank testing, removal and re-
mediation; site assessment, and complete
sampling services.
BNA CoBtmnnkadou tec.
9439 Key West Ave.
Rockville. MD 20830
301/948-0540
BNA Communications Inc. will display bro-
chures in the literature center on our video train-
ing programs: Handling Hazardous Waste;
Splits Happen: A Training Program for Small
Spill Kesporat, and eight new safety training
videos for hazardous waste operations covered
under SARA 1910:120.
Baker/TSA, tec.
Airport Office Park, Bldg. 3.420 Rouser Rd.
Coraopolis. PA 15108 412/2694000
Perform remedial investigation/feuibility
studies, site assessments, hydrogeologk studies,
RCRA monitoring; implement remedial actioni
at hazardous waste management sites; provide
preliminary/final engineering design and con-
struction/cloture management at solid/hiz-
ardous waste sites, risk assessments, regulatory
reviews, economic analyses, market studies, tod
permitting; provide asbestos and tank manage.
men! services.
Bates Video Production
1033 O St., Suite 546
Lincoln, NE 68308
401/476-7951
Bates Video Production has experience develop-
ing projects dealing with the complex and polit-
ically sensitive issue of waste. We know how to
communicate the technical aspects of waste dis-
posal; we can help make the complictwd,
simple. If you're considering using video as i
communications, informational, educations)
tool, we'd like to help.
Bio-racoTcry Sysiinai, tee.
4200 South Research Dr.
Las Cruces, NM 88003
505/646-5188
Bio-recovery Systems, Inc. is a rapidly growini
company engaged In recovery of toxic, heavy
metals from industrial wastewaters, contami-
nated groundwaters, Superfund sites and mimni
process streams. The Las Cruces, NM, firm pro-
vides an economical, proprietary technology for
removing and recovering metallic hazardoui
wastes to meet pollution effluent limits.
Boot AJfea A Hamilton
4330 East-West Highway
Bethc*da,MD 20814
301/931-2690
Booz, Allen A Hamilton, Inc. is a leading tech-
nology and management consulting firm that hat
earned an outstanding reputation in environmen-
tal services through yean of direct involvement
developing and implementing key programs for
government and industry world-wide. The firm
has worked with the Superfund and RCRA pro-
grams since their inception and offers compre-
hensive mission and program-related expertise.
Technology and management services include:
risk management; audits and technical evalua-
tions; regulatory enforcement and policy sap-
port; records management; information sys-
tems development, and program planning, im-
plementation and evaluation.
300 Lafayette Blvd.
Fredericksburg. VA 2240 1 703/373-3482
Absorbents for emergency response, leak tod
spill control, and industrial maintenance: • Basic
Sponge: absorbs water, oil, inks, and non-
aggressive fluids • Oil Only Sponge: absorbs oil
and repels water both indoors or outdoors
• Chem-Sponge: absorbs any fluid— acids, caus-
tics, solvents, without guesswork. Available ss
rubes, pillows, booms, mats and rolls.
Browa aid CaMwefl
P.O. Box 8045
Walnut Creek, CA 94596-1220
415/937-9010
A full-service firm. Brown and Caldwell has ex-
perience in site assessments; remedial tavwugs-
658 EXHIBITOR PROFILES
-------
tions, feasibility and treatability studies; analyti-
cal programs; waste minimization; UST manage-
ment programs; design of remediations, includ-
ing soil and groundwater treatment; permitting;
construction supervision, and solid waste man-
agement. The firm has offices nationwide, in-
cluding three certified analytical laboratories.
The Bureau of National Affairs, Inc.
1231 25th St., N.W., 3-414
Washington, DC 20037
202/452-4229
BNA publishes regulatory, legal and working
guides providing the latest information concern-
ing the manufacture, transportation, safe han-
dling and disposal of hazardous materials.
C-E Environmental
261 Commercial St.
Portland, ME 04112
207/775-5400
Environmental consulting, monitoring and
chemical analysis; hazardous waste site investiga-
tions, remedial design, construction and clean-
up; thermal and non-thermal waste treatment
systems.
CECOS International
P.O. Box 3151
Houston, TX 77253
713/584-8850
CECOS International, Inc., a full-service com-
pany headquartered in Houston, Texas, and with
facilities throughout the country, is primarily in-
volved in the recycling, reclamation, remedia-
tion, treatment, transportation and disposal of
chemical and hazardous waste.
CHZM HILL, Inc.
P.O. Box 4400
Reston.VA 22090
703/471-1441
CH2M HILL provides waste management serv-
ices—including design, construction, investiga-
tion, and planning—to industry and govern-
ment. We are the largest environmental engi-
neering firm in the United States, with 4,000 em-
ployees in 57 offices worldwide. Over a third of
our business is managing hazardous, radioactive,
and solid waste.
TheCHEMTOX" System
Div. of Resource Consultants, Inc.
P.O. Box 1848
Brentwood.TN 37024-1848 615/373-5040
Software programs and services designed to save
time and control costs. Extensively tested over a
variety of industries, government agencies, in
more than a dozen countries. The CHEMTOX®
System can make right-to-know, safety jobs,
SARA reporting, and manifest tracking easier
and more productive.
CompnChem Laboratories, Inc.
3308 Chapel Hill/Nelson Highway
Research Triangle Park
NC 27709
1/800/833-5097
CompuChem Laboratories, Inc., a full service
organic and inorganic CLP laboratory, special-
izes in CERCLA, RCRA, Priority Pollutant,
Dioxin and Waste Characterization Analysis. In
1990, CompuChem will expand its analytical
services to include:
• Mixed Waste Analysis
• Air Analysis
• Low Level Radiological Analysis
CompuChem's Environmental Site Profile
(ESP), a proprietory data management system,
provides on-line access to test results which can
be downloaded to personal computers. For for-
ensic quality data and expedited turn-around
times, visit the staff of CompuChem Labora-
tories.
Calgon Carbon Corporation
P.O. Box 717
Pittsburgh, PA 15230
412/787-6700
Calgon Carbon Corporation supplies activated
carbon products, systems and services, and air
strippers to remove soluble and volatile organic
chemical compounds from contaminated
groundwater, surfacewater or wastewater.
Camp Dresser & McKee Inc. (COM)
One Cambridge Center
Cambridge, MA 02142
617/621-8100
Camp Dresser & McKee Inc. (CDM) provides
environmental engineering and consulting serv-
ices to government and industry for the manage-
ment of hazardous and solid wastes, water re-
sources, wastewater, and environmental facili-
ties. Hazardous waste services include remedial
design and construction; treatment facility de-
sign and operation; environmental assessments;
RCRA permitting; and groundwater modeling
and restoration.
Capsule Environmental Engineering, Inc.
1970 Oakcrest Ave., Suite 215
St. Paul, MN 55113 612/636-2644
An environmental management firm that special-
izes in the adaptation of waste reduction technol-
ogy as the primary method of meeting present
and future environmental regulations. While
working with clients on waste reduction. Cap-
sule provides engineering services in the areas of
environmental assessments, regulatory compli-
ance, management of site cleanups and educa-
tion programs.
CarbonAlr Services, Inc.
1624 5th St. South
Hopkins, MN 55343
612/935-1844
CarbonAir Services, Inc. is a major manufac-
turer of airstripping and carbon adsorption
equipment for use in the decontamination of
water and air.
Carnow, Conlbear & Associates
333W.WackerDr.,#1400
Chicago, IL 60606
312/782-4486
Carnow, Conibear & Associates is a full service
occupational and environmental health consul-
tant firm offering the following services:
• Asbestos Surveys
• AIHA Supervisor Laboratory
• Environmental Audits
• Medical Surveillance
• Hazard Material. Training
• Community Righl-to-Know Programs
• Health Risk Assessment
• Hazard Communication Programs
• Health Care Network for Occupational Health
Programs
Center for Hazardous Waste Management
10 West 35th St.
Chicago, IL 60616 312/567-4250
Sponsored by Illinois Institute of Technology
and IIT Research Institute, the Center assists
clients by performing research, developing im-
proved techniques, conducting seminars, and
providing quality assurance for waste manage-
ment programs. Recent activities include a tech-
nical, legal and policy study for the coalition on
Superfund, and licensing an in-situ soil decon-
tamination technique.
Chemical Waste Management, Inc.
3001 Butterfield Rd.
Oak Brook, IL 60521 312/218-1500
Chemical Waste Management, Inc. provides a
complete range of hazardous waste management
services. These include reduction, remediation,
treatment, recycling, transportation and dis-
posal.
Chen-Northern, Inc.
96 South Zuni St.
Denver, CO 80223
303/744-7105
Chen-Northern, Inc., is a full-service consulting
firm offering services in solid and hazardous
waste management, environmental engineering,
geotechnical engineering, asbestos consulting
services, tank testing services, geology, hydrol-
ogy, hydrogeology, materials engineering and
testing, construction quality control, construc-
tion contract administration, and management.
The firm has complete in-house subsurface ex-
ploration capabilities including drilling and soil-
gas investigation services. Headquartered in
Denver, Chen-Northern maintains 18 offices
throughout the Western United States.
Christensen Mining Products
4446 West 1730 South
Salt Lake City, UT 84130
801/974-5544
Christensen Mining Products booth features dia-
mond core and drill bits, core barrels and their
new casing advancer (which allows full-hole drill-
ing) and center section removal via wireline for
spot coring. We also display our environmental
and mud products.
Chromanetics Scientific Products
709 N. Black Horse Pike
Williamstown, N.J 08094 609/728-6316
Chromanetics, known as the "One Stop" Shop-
ping Service for environmental laboratory and
field products, offers its new 500-page environ-
mental products catalog to conference atten-
dees. There are seven unique sections to the cat-
alog, which describe specialty scientific glass-
ware, a complete line of chromatography pro-
ducts, chemicals/reagents, plasticware, general
labware, instruments (OC, IR, UV, AA,
Furnace, etc.), lab furniture/hoods/refrigera-
tors and field sampling and safety products, in-
cluding HNU PI 101 sales and rentals.
Claytro Environmental Consultants
22345 Roethel Dr.
Novi, MI 48050
313/344-1777
Services include: Environmental Engineering
Services • Environmental Risk Assessment and
Corrective Action Strategy • Point Source and
Ambient Air Quality Studies • Regulatory
Agency Liaison • Underground Storage Tank
Testing and Management • Hazardous Waste
Disposal, Storage, Handling, and Training Pro-
grams • Geological and Hydrogeological Eval-
uation • Chemical Emergency Response Pro-
grams •_ Comprehensive Surveys, Audits, and
Program Development • State and Federal Per-
mit Application Preparation and Negotiation
EXHIBITOR PROFILES 659
-------
Groundwater and Wutewater Studies • Fugitive
Emission Inventories and Odor Studies.
Columbia Scientific Industrie! Corporation
Box 203190
Austin. TX 78720 512/258-5191
Portable x-ray analyzers for in-situ monitoring
of inorganic contaminants. Directly measures
elements in soil or in sample* collected from the
site. Elements range from Al-u includes RCRA
elements such as Cr, Mn, Ft. Ni, Cu, Zr, Se,
As, Pb, Ag, Cd A Hg. Sensitivity from ppm to
100*.
Consolidated Rail Corporation
Room 919—One Liberty Place
Philadelphia, PA 19103-7399
215/851-7281
Conrail is one of the largest freight railroad sys-
tems in the Northeast-Midwest quarter of the
United States, operating over a network of
approximately 13,100 route miles. Conrail is a
licensed and registered transporter of hazardous
waste and sixty percent of all Superfund sites are
located within its territory. Conrail works close-
ly with connecting rail carriers, trucking, and
equipment companies to offer reliable service.
Corroon * Bteck Environmental
Inraruct Serrtces
6510 Grand Teton Plaza, Suite 102
Madison. WI 53719
608/833-2887
Environmental Insurance Specialists: Corroon
& Black is one of America's largest insurance
brokers. CAB has developed specialized insur-
ance and surety programs for many engineering
& construction groups. Among the specialized
programs is an insurance program for remedial
action contractors, engineers and asbestos abate-
ment contractors. It includes pollution legal lia-
bility, general liability, Contractors Pollution
liability, engineers professional liability, engi-
neers pollution liability, automobile and prop-
erty coverages.
Coca Instrument Corporation
55 Oak St.
Norwood, NJ 07648
201/767-6600
Cosa products include the TOX-10 Total Organ-
ic Halogen Analyzer and the TSX-IO Total Sul-
fur Chlorine Analyzer.
DartAmertca
61 Railroad St., P.O. Box 89
Canfield, OH 44406
216/533-9841
A group of companies dedicated to the transpor-
tation of hazardous waste and general commod-
ities in 48 states utilizing dumps, roll-offi, vans,
flatbeds, pneumatic and liquid tank equipment,
and LTL van service.
DataCbem Laboratories
960 W. LeVoy Dr.
Salt Lake City. UT 84123
801/266-7700
DataChem Laboratories, Inc. is one of the na-
tion's largest and most experienced providers of
comprehensive laboratory analytical services.
These services assist clients in objectively eval-
uating environmental/hazardous waste issues
and industrial hygiene/safety concerns. During
its 16 year history, the company has established
a national reputation for delivering accurate test
results on a cost-effective basis for both the pri-
vate and government sectors.
Davis PBgmill, Inc.
P.O. Box 60
Columbia, TN 38402-0060
615/388-0626
Twin-Shaft Twin-Drive pugmill and complete
plants. The feeding system can consist of belt
type feeder bins to specially designed screw type
bins or a combination of bins. Complete elec-
tronic weighing and ratio controlled packages
are available for precise measurement of added
material and chemicals. Modular and portable
units are available.
Donobnt A Associates, Inc.
4738 N. 40th St.
Sheboygan. W15308I
414/458-8711
Donohue is an ARCS contractor with a nation-
wide staff of over 1.000 and a 1989 ENR RANK
INO of 64. Our environmental scientists and en-
gineers are specialists in waste management, dis-
posal and cleanup. Donohue's hazardous waste
services include RCRA Investigations and com-
pliance monitoring, RI/FS studies, and engineer-
ing of remedial cleanup actions.
Da Pom
1007 Market St., BAD, NA-228
Wilmington. DE 19898 302/774-7248
Dunn Geo*cte*ct Corporation
12 Metro Park Rd.
Albany. NY 12205 518/458-1313
Full Service Environmental Consultants: Com-
plete staff of hydrogeologists, geologists, en-
vironmental scientists and engineers, toxicolo-
gists, and regulatory experts provides a range of
services including RI/FS and RCRA Corrective
Actions, Remedial Design and Construction
Management, Toxicology /Public Health Assess-
ments, Hazardous Waste Planning and Manage-
ment, Hydrogeologic Services and Property
Transfer Environmental Site Assessments.
Ebatco Environmental, A Division of
Ebstsco Services Incorporated
2 World Trade Center
New York. NY 10048-0752 212/839-2744
Ebasco Environmental, a division of Ebasco
Services Incorporated, is a leader in environ-
mental preservation, having remediated wastes
safely and economically for more than 30 years.
In addition to waste remediation services, we
offer complete environmental services including
permitting, siting studies, risk assessments, re-
medial investigations/feasibility studies and en-
vironmental impact reports.
ECOFLO, Inc.
8520-M Corridor Rd.
Savage, MD 20763
301/498-4550
ECOFLO provides answers to client-specific
waste management needs from our extensive
offering of services, including:
I. Waste characterization
2. Collection, transportation and treatment/dis-
posal of most wastes
3. Lab pack services
4. Remediation and clean-up services
5. Waste minimization advice
ECOFLO serves the mid-Atlantic region from
offices in Maryland and North Carolina.
ECOVA Corporation
3820I59thAve.. NE
Redmond, WA 98052
206/883-1900
Hazardous waste management technologies and
services for onstte remediation. ECOVA geolo-
gists, hydrogeologists, mkrobiologisu, chem-
ists, engineers, and field support staff perform
site assessments; sampling and analysis; under-
ground storage lank management; treatability
studies; pilot-scale demonstrations; and full-
scale remediation using biological, chemical,
physical, and mobile incineration lechnolofiej
In situ processes are available.
Elaco Proem Eqtlpnaat Coospuy
P.O. Box 300
Salt Lake City, UT 84110 801/526-2000
Eimco provides biological reactor for sou recla-
mation processes. Inclined plate clarifiers, belt-
press filters and other equipment for treating
wastes are available.
ELANCO Product Company
Lilly Corporate Center
Indianapolis, IN 46285
317/276-2299
Typar* Biobarrier™ is a long-lasting root pre-
vention system, which combines a fabric and a
time-released herbicide. Biobarrier will block
roots that attempt to penetrate closure caps over
waste burial sites without harming the ptanu,
tree* or other vegetation. Biobarrier is flexible
and easy to install.
ENR
1221 Avenue of the Americas
New York, NY 10020
212/512-313!
ENR is a weekly "»•§•**'"' reporting business
and technical news on many types of construc-
tion projects worldwide to over 400,000 readers.
The market includes design (architectural and
engineering) and contracting companies serving
industrial and commercial forms and govern-
ment agencies. Areas of coverage are buildings,
infrastructure, finance and environmental clean-
up.
EMRECO.tec.
P.O. Box 9838
Amarillo, TX 79105
806/379-6424
ENRECO is a technology company established
in 1982 to manage solid, liquid and gaseous
wastes. Our primary area of expertise, waste
stabilization, has been used successfully on over
200 projects. Recent developments include
specialized chemical separation techniques to
isolate a variety of organic contaminants from
the waste matrix.
ENSCO. IK.
333 Executive Court
Little Rock. AR 72205
501/223-4160
ENSCO is a premier hazardous waste manage-
ment company featuring transportation and in-
cineration of RCRA and PCB wastes. ENSCO
provides integrated hazardous waste manage-
ment services to private industry, public utilities
and governmental entities.
ERCE Environmental A Eaergy Services Co.
3211 Jermantown Rd. Oermantown Rd.
Fairfax, VA 22030 703/2464)539
ERCE is a professional and technical services
company that offers environmental, infrastruc-
660 I.XHIBITOR PROHI.I-IS
-------
ture and energy consulting and engineering serv-
ices to industrial and commercial companies,
electric utilities and governmental agencies. De-
sign services are supported by four accredited
analytical laboratories strategically located
throughout the U.S. ERCE maintains over 20
offices in the U.S.
The ERM Group
855SpringdaleDr.
Exton, PA 19341
800/544-5117
The ERM Group, a full-service environmental
consulting firm with more than 50 offices world-
wide, provides the following services: site remed-
iation; hydrogeology; hazardous/solid waste
management; management consulting; indus-
trial/municipal water and wastewater treatment;
underground tank management; environmental
science; air pollution control; computer services;
construction management; and health, safety
and toxicology.
Eagle-Picker Environmental Services
36B.J.TunnellEast
Miami, OK 74354 918/540-1507
Supplier of high quality glass and plastic ware
washed according to EPA Protocols. All con-
tainers available with sampling label and each
box custody sealed to insure sample container
cleanliness. Specify full QA or Wash Only. Ana-
lytical quality control is performed in our full-
service EPA Contract Laboratory Program. A
variety of styles and sizes of containers are avail-
able and are distributed on an exclusive basis by
Baxter Scientific Products.
Earth Resources Corporation (ERC)
P.O. Box 616961
Orlando, FL 32861 407/295-8848
Earth Resources Corporation (ERC) is a full-
service hazardous materials management firm
specializing in the containment, treatment, and
removal of all types of hazardous materials.
ERC has a highly trained professional and tech-
meal staff experienced in the design and imple-
mentation of innovative solutions to today's
waste problems. ERC's capabilities include but
are not limited to soil, groundwater, facilities,
containerized wastes and pressurized gas
cylinders.
The Earth Technology Corporation
100 W. Broadway, Suite 5000
Long Beach, CA 90802-5785 213/495-4449
As one of the nation's leading environmental,
earth sciences and geotechnical consulting firms,
The Earth Technology Corporation's primary
business activities include: Waste Management
and Environmental Services, Critical Facilities
Siting, and Related Advanced Technology and
Testing Services.
Visit Booth 1505 for information on superior
technical capabilities for government and private
industry.
EcoTek
219 Banner Hill Rd.
Erwin, TN 37650
404/843-3732
EcoTek, a subsidiary of Nuclear Fuel Services,
provides comprehensive services for the remed-
iation of sites contaminated by hazardous and
mixed waste materials. EcoTek's services include
hazardous/chemical (CLP) and radiological
laboratory services, process design, mixed waste
characterization, health and safety upgrades,
procedures and equipment, decontamination and
decommissioning, and resource recovery.
Ecology and Environment, Inc.
368 Pleasantview Dr.
Lancaster, NY 14086
716/684-8060
E & E is recognized worldwide as a leader in en-
vironmental science and engineering. The firm
performs remedial investigations, field studies,
mitigative engineering design and construction
management; hazards and risk analyses; spill
emergency response and asbestos removal man-
agement; regulatory compliance audits; environ-
mental impact assessments; air, water and
groundwater pollution control; industrial
hygiene; analytical laboratory services.
Empire Soils Investigations, Inc.
140 Telegraph Rd.
Middleport, NY 14105
716/735-3502
Empire Soils Investigations, Inc., along with its
laboratory division—Huntingdon Analytical
Services—and its wholly owned subsidiary,
Asteco, Inc., provides the following services:
Contract drilling and installation of ground-
water monitoring wells, geotechnical testing in-
cluding contaminated soils, geotechnical engi-
neering, chemical analytical testing, asbestos in-
spection and testing, and materials engineering
and testing.
Engineering-Science
75 North Fair Oaks Ave.
Pasadena, CA 91103
818/440-6101
Engineering-Science (ES) is a full-service, na-
tional and international environmental engineer-
ing firm providing complete services in haz-
ardous waste management. With offices in 23
domestic locations, ES is active in supporting in-
dustrial and military clients in all phases of site/
remedial investigations, feasibility studies, re-
medial action plan preparation, site cleanup/
closure and post-closure activities.
Entropy Environmentalists, Inc.
P.O. Box 12291
Research Triangle Park
NC 27709 919/781-3550
Entropy, the largest independent air emissions
testing firm in the country, provides air emis-
sions testing and consulting services to assist
firms in effectively complying with hazardous
waste regulations. Services include: RCRA/
TSCA incinerator testing and analyzing; VOC
emission inventories and testing; dioxin/furan,
PAH, PCB testing principal organic hazardous
constituents (POHC) selection, trial burn test
plan preparation and permitting assistance.
Envlrodyne Engineers, Inc.
1908 Innerbelt Business Center Dr.
St. Louis, MO 63114-5700 314/426-0880
Envirodyne Engineers, Inc. is a consulting engi-
neering firm and an analytical laboratory, with
offices in Chicago, New York and St. Louis. Our
certified laboratory offers full service capabilities
including: radioactive waste analyses, dioxins/
furans, explosives, Appendix VIII/IX, EP Tox-
icity, TCLP, Priority Pollutants, herbicides, and
all conventional inorganic parameters in waste-
water, potable water, soil, air, and biological
matrices. Our engineering services include site
assessments, UST, treatability studies, ground-
water monitoring, RI/FS, design and construc-
tion oversight.
Environmental Audit, Inc.
P.O. Box 322
Lionville, PA 19353
215/524-7002
EAI is an environmental information company
providing fast, cost-effective environmental
records from federal and state government agen-
cies. EAI's Federal Environmental Data (FED)
Report provides NPL, CERCLIS, RCRA,
FINDS and National Spill records for any area
of the country, guaranteed within 3 to 5 days.
Call (800) 542-8348 for service.
Environmental Business Journal
827 East Washington St.
San Diego, CA 92103
619/295-7685
EnviroQuest, Inc. acquires intelligence on the
business of environmental health and distributes
it through the Environmental Business Journal,
market research studies and strategic consulting
services.
Environmental Compliance Services, Inc.
721 East Lancaster Ave.
Downingtown, PA 19335 215/269-6731
ECS is an organization dedicated to assisting en-
vironmental companies with their insurance,
safety and compliance needs through the unique
combination of in-house expertise in environ-
mental regulation, risk management, and insur-
ance underwriting.
Environmental Instruments
2170 Commerce Ave., Unit S
Concord, CA 94520 415/686-4474
E.I.'s product line represents state-of-the-art en-
vironmental equipment. We specialize in ground-
water recovery and treatment systems, soil vapor
sampling and recovery systems. We also carry a
full line of monitors and samplers for water and
air. Come see the new Photoionization Detector,
the OVM 580B, and our new Cavitation-Oxida-
tion water treatment system at our booth.
Environmental Technology, Inc.
3705 Saunders Ave.
Richmond, VA 23227
804/358-5400
HazWaste Industries and its operating subsidies
(Environmental Technology, Bionomics, and
HazLabs) provide a full range of remediation
services:
• Site Investigations, Inspections, and Audits
• Feasibility and Treatability Studies
• Engineering Design and Construction Man-
agement
• Site Remediation, Including Emergency Re-
sponse and Removal
• Long-Term Monitoring
HazWaste provides complete, quality solutions
to environmental problems.
Environmental Medicine Resources, Inc.
4360 Chamblee-Dunwoody Rd., Suite 202
Atlanta, OA 30341 404/455-0818
Environmental Medicine Resources, Inc. (EMR)
specializes in the development and centralized
management of medical surveillance programs
for companies whose employees are potentially
exposed to OSHA regulated hazardous ma-
terials. EMR's regulatory and medical staff, in
conjunction with a national network of 300 clin-
ical facilities, ensures consistent, cost-effective,
compliance-assured programs.
EXHIBITOR PROFILES 661
-------
Envtroaafe Services, Inc.
900 E. Eighth Ave., Suite 200
King of Prussia. PA 19406
419/255-5100
Envirosafe Services, Inc. (ESI) is one of the lead-
ing providers of hazardous waste management
services in the United States, providing both off-
site disposal and on-site remediation services.
ESI's hazardous waste disposal services include
analysis, pretreatment to stabilize nonconform-
ing wastes and disposal, and are provided to in-
dustrial and governmental customers. In addi-
tion, a wholly owned subsidiary of ESI, Enviro-
safe Technology Group Inc. (ETC), furnishes
remediation services primarily at industrial sites
with hazardous substances. These services are
provided by the Field Services Oroup-A.C.E.S.,
vapor extraction treatment by Midwest Water
Resources, Inc.-
-------
Geo/Resource Consultants, Inc.
851 Harrison St.
San Francisco, CA 94107
415/777-3177
Geo/Resource Consultants, Inc. (GRC) is a full-
service environmental consulting firm dealing
with today's complex hazardous materials indus-
try. GRC's diverse professional staff have exper-
tise in numerous fields including: groundwater
monitoring programs; landfill characterizations;
UST/LUFT programs; RCRA training; environ-
mental assessments and audits; RI/FS programs;
and remedial action design.
Geochemlcal Engineering/
Rindol International
1658 Cole Blvd., #6-80
Golden, CO 80401 303/233-8357
Randal Mining Directory lists all U.S. mines.
Randal Buyer's Guide lists products and services
used by all mines and mineral processing indus-
tries. Special section on Environmental Protec-
tion and hazardous materials control are in-
cluded. Companies are invited to submit their
free listings.
Geonlcs Limited
8-1745 Meyerside Dr.
Mississauga, ON L5T 1C6 416/670-9580
Geonics Limited is the world's leading manufac-
turer of electromagnetic geophysical instrumen-
tation. Applications include: surface measure-
ments for delineation of contaminant plumes
and detection of buried metal; borehole conduc-
tivity measurements for mapping vertical plume
structure and well screen location.
Geoprobe Systems
607 Barney St.
Salina.KS 67401
913/825-1842
Manufacturers of the Geoprobe 8-A hydraul-
ically powered soil probing units, suitable for
soil gas, water, and descrete soil sampling. Small
diameter soil sampling kits, Geoprobe 1" O.D.
well points and stainless steel minibailers, stain-
less steel sampling implants, vacuum/volume gas
sampling systems, and complete mobile gas and
water laboratory vans.
Geosafe Corporation
303 Parkplace, Suite 126
Kirkland.WA 98033
206/822-4000
Geosafe Corporation offers in situ vitrification
(ISV) services for remediation of contaminated
soil and sludge sites. The ISV process destroys
hazardous organics through pyrolysis and simul-
taneously immobilizes hazardous inorganics in a
delistable, vitrified residual. This cost-effective
process offers significant advantages over con-
ventional soil treatment processes.
Geosclence Consultants, Ltd.
500 Copper NW, Suite 200
Albuquerque, NM 87102 505/842-0001
Geoscience Consultants, Ltd. (GCL) is a minor-
ity-owned small business (8(a) Certified) and is a
full-service environmental design and engineer-
ing consulting firm. GCL provides the following
services:
• UST Management
• Air Quality Assessment and Permitting
• Health and Safety Training
• Remedial Engineering
1 Waste Minimization and Compliance Reviews
• Groundwater Remediation
' Regulatory Negotiation
Geotechnlcal Fabrics Report
345 Cedar St., #800
St. Paul, MN 55101
612/222-2508
Geotechnical Fabrics Report (GFR) is an engi-
neer's guide to geomembrane liners, geotextiles
and related products. GFR regularly includes
case histories, papers, industry news and other
information on geosynthetic applications in haz-
ardous waste containment. GFR will serve
14,000 engineers, contractors, landfill owners
and operators and other industry members in
1990. GFR is published by the Industrial Fabrics
Association International.
Gtbbs & Hill, Inc.
11 Penn Plaza
New York, NY 10001
212/216-6000
Gibbs & Hill, Inc., a HILL Group Company, is
a full-service consulting engineering company
offering professional services in remedial investi-
gations, feasibility studies, chemical and envi-
ronmental engineering, geotechnical and hydro-
geological consulting, underground tank investi-
gation and remediation, litigation support, per-
mitting, landfill engineering design and closure,
wastewater treatment, right-to-know consulting
and program management, health and safety,
and construction management and claims. Addi-
tional available services from the HILL Group
include asbestos management, project manage-
ment oversight and PCB investigation/remed-
iation as well as program management.
Greenhorne & O'Mara, Inc.
9001 Edmonston Rd.
Greenbelt, MD 20770 301/982-2800, X442
Greenhorne & O'Mara, Inc. provides hazardous
waste management services to industry and gov-
ernment. Our experienced staff (most OSHA/
AHERA-certified) know the requirements of
RCRA, CERCLA, SARA, CWA, and CAA.
Services include site characterization, property
transfer assessments, asbestos management,
groundwater assessments, environmental audits,
RI/FSs, remedial design, waste minimization,
and surveying.
Griffin Remediation Services, Inc.
500 Winding Brook Dr.
Glastonbury, CT 06033 203/657-4277
Griffin Remediation Services, Inc. (GRS) is a
full-service remediation company with specialty
expertise in the design and implementation of
comprehensive solutions to groundwater-
oriented environmental problems. An affiliate of
Griffin Dewatering Corp., GRS utilizes over 50
years of groundwater control experience. From
their 18 locations throughout North America,
Griffin employs innovative technologies focused
on the containment, recovery, treatment and/or
disposal of hazardous and nonhazardous
groundwater pollutants. Services include: re-
medial dewatering, slurry trenches, landfill gas
vents, deepwells, wellpoints, monitoring wells,
soil vents, air stripping, and pump sale/rentals.
Groundwater Technology, Inc.
220 Norwood Park South
Norwood, MA 02062
616/769-7600
Groundwater Technology, Inc. is a full-service,
international environmental assessment and re-
mediation firm with 55 offices and 1,300 em-
ployees dedicated to finding innovative solutions
to today's environmental concerns. Considered
the world leader in integrated solutions to en-
vironmental problems, the firm has completed
more than 5,000 jobs since 1975.
Grundfos Pumps Corporation
2555 Clovis Ave.
Clovis, CA 93612
209/292-8000
Manufacturer of the Redi-Flo Environmental
Submersible Pump. The Redi-Flo is constructed
of stainless steel and teflon and is designed to
pump contaminated groundwater from a 4-inch
well or larger. The Redi-Flo can provide flow
rates up to 32 gallons per minute and to heads of
680 feet.
Gulf South Environmental Laboratory, Inc. and
Pacific Northwest Environmental Laboratory,
Inc.
6801 Press Dr., East Building
New Orleans, LA 70126 504/283-4223
Gulf South Environmental Laboratory, Inc. and
Pacific Northwest Environmental Laboratory,
Inc. are full-service analytical laboratories. Both
provide analytical support for all major regula-
tory programs including CERCLA, RCRA,
NPDES, drinking water and real estate transfer.
FSELI and PNELI are participants in the EPA
CLP program. PNELI is also California certified.
Gundle Lining Systems, Inc.
19103 Gundle Rd.
Houston, TX 77073 713/443-8564
Leaders in Synthetic Liners—Gundle Lining
Systems, Inc., Houston, Texas, is recognized as
the world leader in the manufacture and installa-
tion of High Density Polyethylene lining sys-
tems. Gundle manufactures HDPE (Gundline
HD) in 22.5 foot seamless widths from 20 to 140
mils thick. Gundle installs HDPE using their
patented extrusion welding machine and new
automatic hot wedge welder. Also from Gundle
is Gundnet, drainage net; Gundline HDT, a tex-
tured HDPE liner; and Hyperlastic, a very low
density polyethylene liner.
HAZCO Services, Inc.
2006 Springboro West
Dayton, OH 45439
513/293-2700
National Supph'er of Safety Equipment and Serv-
ices for the Hazardous Waste Industry including
personnel protective equipment and supplies, in-
strumentation rentals and repair services, hazmat
equipment, environmental sampling equipment
and supplies, decon trailers and information
systems.
HazMat Environmental Group, Inc.
P.O. Box 676
Buffalo, NY 14217
716/876-3957
HazMat Environmental Group, Inc. specializes
in full-service hazardous waste management and
transportation services. Our fleet of over 100 ve-
hicles perform safe and insured waste transport
with fully permitted 48 state authority. Waste
management consulting services include waste
minimization and management planning, site
audits, remedial project planning, and a full line
of personnel regulatory training programs.
HAZMAT Training, Information and
Services, Inc.
6480 Dobbin Rd.
Columbia, MD 21045 800/777-8474
The mission of HAZMAT Training, Informa-
EXHIBITOR PROFILES 663
-------
lion and Services, Inc. (Hazmat TISI) is to pro-
vide the highest quality training in the hazardous
materials industry. Our services are hallmarked
by a disciplined, systematic approach to training.
Our training programs, provided under direct
contract with our clients, can be presented either
at our Columbia, Maryland training facility or
on-site at a client's location, tailored to the
client's specific training needs. As a natural off-
shoot to our training program!, we also provide
a number of consulting and information serv-
ices. Additional information is available upon
request at 1-800-777-T1SI or 301/964-0940.
Siting and Design; Water Resource Manage-
ment; Mine Tailings and Water Management.
I, Inc.
8404 Indian Hills Dr.
Omaha, NE 68114
402/399-1000
HDR Engineering, Inc. specializes in hazardous
and industrial waste management, including re-
medial investigations, feasibility studies, remed-
ial design and construction management of haz-
ardous waste sites; design of treatment, storage
and disposal faculties; and closure/post-closure
planning; assessment and design of underground
storage tank facilities, environmental permitting,
real estate transfer audits, environmental risk
assessments, air pollution control and permit-
ting, environmental modeling (air, water,
groundwater, toxic contamination). Industrial
projects encompass the study, design and imple-
mentation of industrial waste treatment; ultra
pure water, gas and chemical systems; environ-
mental permitting; process facilities for high-
tech industries.
HMM Associates, Is*.
1% Baker Ave.
Concord, MA 01742
508/371-4305
HMM Associates is an environmental consult-
ing, engineering and planning firm. HMM
provides a wide range of hazardous materials/
waste services including: Superfund Rl/FS; re-
medial design and construction oversight; per-
sonnel protection and safety training; and Title
HI Emergency Preparedness Planning and Com-
munity Right-to-Know.
617/964-6690
S, IK.
160CharlemontSt.
Newton Highlands. MA 02161
HNU provides the following: Model HWlOl-
Hazardous Waste Analyzer, /S/07-lntrinsically
Safe Analyzer, /V/07-Photounization Analyzer
(portabk*)/Modr/ JOlDP-Dediatcd Program-
mable OC, Model j;;-Ponable Oas Chrom-
otograph, Model 301GC. Model J7/-Compact
Temperature Programmed OC, Model 331-Com-
pact Dedicated Capillary OC, SEFA -Portable
X-Ray Fluorescence Analyzer, 75 Meter-Pon-
able PHMV Temperature Meter, /5Ł-Ion Selec-
tive Electrodes.
Hydro-Search, IDC.
235 N. Executive Dr., *309
Brookfleld, WI 53005
414/784-4588
Services in hydrogeology, engineering, and pro-
ject management for: Remedial Investigations/
Feasibility Studies (RI/FS); Preparation of
Work Plans; Managing On-Slte Activities; De-
signing and Implementing Remedial Action Pro-
grams; Technical Ouidance for Responsible Par-
ties; Oversee EPA Contractors; Review Ground-
water Monitoring Plans and Reports; Under-
ground Storage Tank Management; Landfill
Haosoa Engineers incorporated
1525 S. 6th St.
Springfield. IL 62703
217/788-2450
Hanson Engineers Incorporated is a multi-disci-
pline engineering firm providing engineering
services in environmental/waste management,
geotechnical, structural, transportation, hydro-
logic/hydraulic, and material testing. HEI Is
registered to practice in all 48 contiguous states,
and has completed over 100,000 projects
throughout all 50 states and in 13 foreign coun-
tries.
Harmon Environmental Strrteaa, toe.
2066-A West Park Place
Stone Mountain, OA 30087 404/469-3077
Total site remediation, emergency response,
wastewater treatment and infectious hazardous
waste incineration services; dewatering and sta-
bilizing sludges; excavating contaminated soils;
closing and capping lagoons, landfills and waste-
pile*; incineration of liquids, sludges and con-
taminated soils.
Hart Eartroaaeatal Management Corporattoa
6981 North Park Dr.
Pennsauken, NJ 08109 609/663-0440
Engineering solutions to pollution control, haz-
ardous waste and environmental management
problems. Han of fen regulatory assistance, en-
vironmental audits, site investigation, risk assess-
ments, design engineering, construction man-
agement and remedial services with comprehen-
sive programs in underground storage tank man-
agement, asbestos hazard management, RCRA
and CERCLA strategic and technical support.
Hazardous Watte ActJoa CoaUtloa
1015 15th St., NW.«802
Washington, DC 20005 202/347-7474
The Hazardous Waste Action Coalition
(HWAQ responds to difficult questions facing
technical consulting firms active in hazardous
waste management. HWAC develops and pro-
motes approaches that are technically sound,
timely and cost-effective. HWAC also pursues
needed legislative and regulatory actions, pro-
motes sound business practices, and develops
effective technical practices.
Hewlett-Packard Company
P.O. Box 10301
Palo Alto, CA 94303-0890 415/857-1501
Hewlett-Packard will display Its instruments,
systems and capabilities for EPA approved en-
vironmental analysis and methods.
Howard Smith Scrtea Company
P.O. Box 666
Houston. TX 77001
713/869-5771
Major manufacturer and supplier of well screen
for application in environmental, water, oil and
gas wells.
Hoyl Corporation
251 Forge Rd.
Westport, MA 02790 508/636-8811
Manufacturer of Solvent Vapor Recovery/Air
Pollution Control Equipment, Distillation
Equipment, Odor Control Equipment, and
Liquid Purification Equipment.
Haater/ESEIac.
P.O. Box 1703
Gainesville, PL 32602-1703
904/332-331$
Hunter offers complete one-stop environmental
service*, with In-house capabilities that normally
require several firms. Areas of service include
RI/FS/RD; asbestos management; UST man-
agement; environmental audits; toxic and haz-
ardous materials control; water and wastewater
treatment technology; source and ambient sir
monitoring; regulatory analysis; permitting and
compliance; bioassay; and surface and ground-
water monitoring.
Hydro GfOMf The
97 Chimney Rock Rd.
Bridgewater, NJ 08807 201/563-1400
Contractors, manufacturers and consultants
specializing In contaminated groundwater. Will
be exhibiting air stripping towers, aerators, grav-
ity and pressure filters, clarified and GAC
systems.
HniasMtka,toe.
150 Causeway Si.
Boston. MA 02114
617/723-4664
Hygknetks. Inc. is an architectural-engiiwer-
ing/industrial hygiene firm specializing in asbes-
tos management consulting, indoor air quality
assessments, and hazardous *»^yyyfai* HianMn
mem consulting services. We are located in Bos-
ton. Hartford, New York, Washington, DC,
Chicago, Lot Angeles, San Francisco, Hooohuo,
and Frankfurt, West Germany.
415/782-3905
23717-FEkhlerSL
Hayward, CA 94545
A complete tine of glass and polyethylene
sample bonks, jars, and vials supplied with
Teflon-lined closures attached and available
chemically pre-ckaned and laboratory certified
to meet EPA specifications. Abo available are
custom cleaned sample containers, protective
shipping materials, convenient MmpBqg kits,
and preservatives in ampules.
ICF Kasstr Engineers
9300 Lee Highway
Fairfax. V A 22031 703/934-3300
ICF Kaiser Engineers, headquartered in Oak-
land, California, provides engineering, constrae-
tion, and construction management services to
public and private sector clients involved with
environmental, transportation, industrial, ad-
vanced technology, energy, and other infrastruc-
ture projects around the world.
ICM Laboratories
1152 Route 10
Randolph, NJ 07869 201/5844330
Full-service laboratory specializing in environ-
mental analysis. Laboratory services include
analysis for compliance with ECRA. RCRA,
NPDES, Hazardous Waste Classification, and
CERCLA. Monitoring well sampling also avail-
able.
IEA, Inc.
P.O. Box 12846
Research Triangle Park
NC 27709
919/467-9919
IEA, Inc.. a U.S. EPA CLP laboratory, pro-
vides complete environmental analytical services
664 EXHIBITOR PRON1.KS
-------
to the engineering/consulting, industrial and
governmental communities. IE A offers SW-846
3rd Edition, TCL/TAL, Priority Pollutants, and
Petroleum Hydrocarbon analyses utilizing mul-
tiple, dedicated GC/MSs, GCs, AA and ICP,
HPLC, IR OPC, SEM, TOC and TOX. Rapid
turnaround and a chemist-staffed Client Services
Group are just two examples of lEA's customer-
oriented commitment to meeting your analytical
requirements.
Inside EPA
P.O. Box 7167, Ben Franklin Station
Washington, DC 20044 703/892-8500
Inside EPA's environmental group of publica-
tions include Inside EPA, Superfund Report
and Environmental Policy Alert, the preeminent,
relied upon information sources for timely,
essential news on the environment today. To-
gether they cover the realm of environmental
issues and policies—the major legislative, regula-
tory and legal actions—facing the nation and
you.
In-Sltulnc.
210 South Third St., P.O. Box I
Laramie, WY 82070
307/742-8213
In-Situ Inc. designs and manufactures environ-
mental data loggers, water-level and water-qual-
ity probes, and UST Leak Detection Systems.
In-Situ develops hydrologic and geologic soft-
ware for both mainframe and personal com-
puters. In-Situ also has a professional staff for
hydrology consulting and water resource man-
agement services for industrial, municipal,
energy, and mining related projects and facil-
ities.
Institute of Gas Technology
3424 South State St.
Chicago, IL 60616
312/567-3794
IGT is a not-for-profit educational, energy and
environmental research and development organi-
zation established in Chicago, Illinois in 1941.
IFT's environmental capabilities include waste
incineration and detoxification, and catalytic
and biological decontamination of hazardous
and industrial wastes, soils and sludges, and
groundwater. IGT programs range from funda-
mental investigations through bench scale and
pilot plant process development to field testing.
Integrated Chemistries, Inc.
19700akcrestAve., Suite 215
St. Paul, MN 55113
612/636-2380
An environmental specialty chemical company
that develops chemical processes that create
more effective ways to reduce or destroy haz-
ardous waste. Our CAPSUR system exemplifies
this concept. CAPSUR is a PCB cleanup product
that is environmentally sound, easy to use and
has a high extraction efficiency while reducing
disposal costs.
International Technology Corporation
23456 Hawthorne Blvd.
Torrance, CA 90505 213/378-9933
International Technology Corporation (IT) is an
environmental management company with mul-
tiple technologies and human resources to solve
a wide variety of problems involving hazardous
chemical and nuclear materials. The company
Provides a comprehensive range of services to in-
dustry and governmental agencies in four busi-
ness areas: Analytical, Engineering, Remediation
Services and Environmental Products.
Jacobs Engineering Group, Inc.
529 14th St., NW., Suite 1234
Washington, DC 20045
202/783-1560
Jacobs Engineering Group, Inc. is an interna-
tional engineering and construction firm with ex-
tensive environmental experience in waste min-
imization; corrective and remedial action; and
planning, engineering, design and construction
management of hazardous, toxic, low-level and
mixed waste programs.
K-V Associates
281 Main St.
Falmouth, MA 02540
508/540-0561
Manufacturers of sub-surface tools and probes
for use with soil-gas sampling and water
sampling.
Kelchner Environmental Excavators, Inc.
6834 Loop Rd.
Centerville, OH 45459 513/434-1334
Kelchner Environmental Excavators, Inc. is a
service company providing support to environ-
mental consulting firms, industry, and govern-
ment. These projects include: Impoundment/la-
goon closures; soil/sludge solidification/stabili-
zation; slurry cutoff walls; landfill construction,
including HDPE; liner installation; leachate
collection systems; and underground storage
tank removal. For more information, call 8007
877-5352.
Keystone Environmental Resources, Inc.
3000 Tech Center Dr.
Monroeville, PA 15146 412/825-9600
Keystone Environmental Resources, Inc., a sub-
sidiary of The Chester Engineers, provides en-
vironmental services to industry and govern-
ment in the United States and Canada. Our areas
of expertise include analytical laboratory test-
ing; treatability studies; and conceptual, design
and project engineering for air quality, remed-
iation of contaminated soils and waters, and
hazardous wastes.
J.J. Keller & Associates, Inc.
8361 Hwy. 45, P.O. Box 368
Neenah, WI54957-0368
414/722-2848
J.J. Keller & Associates, Inc. currently re-
searches, writes, edits, and prints over 60 tech-
nical publications serving the CPI and transpor-
tation industry. Keller also offers software pro-
grams, videos, and regulatory compliance serv-
ices and products. Featured at Superfund '89 will
be Keller's Hazardous Waste Management
Guide; Hazardous Materials Guide; Right to
Know Compliance Manual; Chemical Crisis
Management Guide; Chemical Regulatory Cross-
reference; Small Quantity Generator Kit; Haz
Mat II Software; Rega-A-Dex Software; MSDS-
PC Software; Hazardous Waste Services Direc-
tory; and hazardous training booklets and
videos.
James T. Waning Sons, Inc.
4545 S St
Capitol Heights, MD 20743 301 /322-5400
All types and sizes of containers—new and re-
conditioned—fiber, steel, plastic. Our hazardous
waste containers are DOT approved and range
in size from 5 to 110 gallons. We accept orders
from one to truck loads and we ship anywhere.
You order a container—we don't have it—it's
special—we will get it for you. No order is too
small for James T. Warring Sons, Inc. Let us
help you contain your hazardous waste. Also
provided is empty drum removal with custom
shredding and crushing done on your site.
LaBounty Manufacturing, Inc.
P.O. Box B, State Rd.2
Two Harbors, MN 55616
218/834-2123
LaBounty RB 80 Barrel Handler Description—
LaBounty Manufacturing will exhibit the La-
Bounty RB 80 Barrel Handler which is specif-
ically designed to handle barrels and cylindrical
containers used for toxic waste or chemical stor-
age. LaBounty also manufactures TW Grapples
for handling solid waste and Mobile Shears for
cutting waste piping, tanks, and other contam-
inated materials.
Laboratory Resources, Inc.
363 Old Hook Road
Westwood, NJ 07675
201/666-6644
Laboratory Resources, Inc. provides high qual-
ity analytical testing services for commercial, in-
dustrial, and environmental clients with a labor-
atory network serving the northeast. Capabili-
ties include a wide variety of testing services
including organic, inorganic, asbestos, and in-
dustrial hygiene analysis along with being a CLP
laboratory. Quality is the cornerstone upon
which services are built.
Laborers—AGC Education & Training Fund
Route 97 & Murdock Rd., P.O. Box 37
Pomfret Center, CT 06259 203/974-0800
The Laborers-AGC Education and Training
Fund is a labor and management trusteed organ-
ization that develops and implements training
programs for over 70 training centers located
throughout the U.S. and Canada (32). Courses
offered include Line Foreman Safety Training,
Pipe Laying, Blasting, Laser Beam, Asbestos
Abatement and Hazardous Waste Worker Train-
ing.
Lancaster Laboratories, Inc.
2425 New Holland Pike
Lancaster, PA 17601
717/656-2301
Lancaster Laboratories, Inc. is an independent-
ly owned and operated testing laboratory lo-
cated in Lancaster, Pennsylvania. With a staff of
more than 350 scientists, technicians, and sup-
port personnel housed in a 78,000 sq. ft. facility,
Lancaster Labs provides a wide range of environ-
mental, industrial hygiene and health sciences
testing services.
Lancy International, Inc.
181 Thorn HillRd.
Warrendale, PA 15086-7527
412/772-0044
Innovative technologies for the recovery/treat-
ment of groundwater and the treatment of san-
itary and hazardous landfill leachates are on dis-
play by Lancy International, Inc., an Alcoa Sep-
arations Technology Company. Comprehensive
treatment/recovery equipment, services and
technologies are highlighted and information re-
garding the various treatment approaches are
available.
EXHIBITOR PROFILES 665
-------
Law Environmental, Inc.
112TownparkDr.
Kennesaw, OA 30144-5599
404/421-3400
Law Environmental, Inc.—A professional engi-
neering and earth science consulting firm. Serv-
ices include: remediation management and site
cleanup, environmental review for property
transfers and plant operations, environmental
siting and permitting, water resources and water
quality management, occupational health and
safety, tank management, hazardous and solid
waste management.
Layne-Western Company, Inc.
1900 Shawnee Mission Pkwy.
Mission Woods, KS 66205
913/362-0510
Layne-Western Company, Inc. brings technical
knowledge and practical experience to the
specialized fields of investigative drilling, remed-
ial action and environmental monitoring. From
offices located coast-to-coast, we provide clients
with a pool of talented professionals and a high
commitment to professionalism, safety and qual-
ity.
Lewfa Pnbttanera, ISK.
121 South Main St., P.O. Drawer 519
Chelsea, MI 48118
313/475-8619
Publisher of scientific and environmental books
that cover toxic and hazardous waste, ground-
water, wastewater, and other vital topics in the
environment*] Held.
Lopat Enterprise*. Inc.
1750 Btoomsbury Ave.
Wanamassa, NJ07712
201/922-6600
Lopat's K-20/LSC is used in the control and re-
mediation of all hazardous teachable toxic metals
mandated by the U.S. EPA, state and local
authorities in incinerator ash, soil, soil-like solids
or semi-solid wastes. K-20/TCC is used in the
control of PCBs and other chlorinated and
organic compounds in soil-like paniculate mat-
ter and on various cementitious surfaces.
Lo«Ma«« Snrety * Bonding. Inc.
P.O. Box 40371
Baton Rouge. LA 70835
504/272-7052
Louisiana Surety & Bonding, Inc. is a nation-
wide construction bonding agency specializing in
Bid, Performance, and Payment Bonding for
the hazardous waste remediation industry.
MAC Corporatfon/Sntnrn SB/widen
201 E. Shady Grove Rd.
Grand Prairie, TX 75050 214/790-7800
Manufacturers, designers and fabricators of re-
duction systems to address the needs of PCB, haz
waste, low rad waste, and steel-drummed chem
waste processors. If incineration or other treat-
ment requires preparing the infeed through
shredding, opening, separating, disengaging or
reducing the size of same, our expertise will pos-
itively contribute to your decision-making pro-
cess.
METCO Environmental, Inc.
P.O. Box 598
Addison, TX 75001
214/931-7127
Source emissions testing services including trial
burn testing, trial burn plan preparation, compli-
ance testing, ambient air monitoring, and contin-
uous emission certification.
MK-Eavlronmcntal Service*
P.O. Box 79
Boise, ID 83707
208/386-6172
Morrison Knudsen from start to finish...MK-
Environmenial Services offers full-service capa-
bility to manage environmental and hazardous
waste programs. Continuing Morrison Knud-
sen's 77-year history as a major International
engineer and constructor, MK-EnvironmentaJ
fully integrate* scientific, engineering, procure-
ment, and remediation activities to support pri-
vate- and public-sector clients.
MFC Environmental
8631 W. Jefferson
Detroit, MI 48209
313/849-2333
MPC Environmental is a full-service environ-
mental company. The show emphasis will be
with our high capacity, low viscosity pumping
system for movement of hazardous or petroleum
type products. We offer site cleanups, immed-
iate response to spills of all types, groundwater
remediation and marine services.
MSA
P.O. Box 426
Pittsburgh, PA 15230 412/967-3000
MSA will display a full line of personal protec-
tive equipment including products for respira-
tory protection and environmental monitoring.
MWR. Inc.
615 W. Shepherd St., P.O. Box 10
Charlotte, MI 48813 517/543-8155
Remedial services emphasizing patented soil
vapor extraction process.
Matarafc Industries, Inc.
1339 N.Milwaukee St.
Milwaukee. WI 53202
414/272-1965
Matarah Industries supplies a superior quality in-
dustrial sorbent for both in-plain and on-water
spill prevention, containment and/or clean-up.
Matarah Industries offers MATASORB. an oil
and chemical sorbent, which does not absorb
water and shows superior strength, sorbent
capacity, and cleanliness. Matarah also offers
SORBX, an advanced all-liquid sorbent ma-
terial. Both products incinerate cleanly for dis-
posal adding BTU value to the burning process.
MetcaifAEtfdy
P.O. Box 4043
Woburn, MA 01888-4043 617/246-5200
Metcalf A Eddy provides a full range of haz-
ardous waste services, from remedial investiga-
tions and feasibility studies, to remedial design
and construction, to long-term operations. Met-
calf A Eddy also provides emergency response
services using licensed equipment, as well as en-
vironmental audits, permitting assistance, and
underground storage tank management.
metaTRACE, Inc.
13715 Rider Trail North
Earth City, MO 63045
314/298-8566
metaTRACE is an analytical laboratory offering
full-service capabilities for organic, inorganic
and radio chemistry analyses of air, ground-
water, surface water, wastewater, potable water,
soil, hazardous wastes and biological samples.
Routine analyses include organic/inorganics, di-
oxins/furans, mixed waste, radiochemistry,
TCLP, Appendix VIII and IX, explosives, indus-
trial hygiene, air quality, and EPA priority pollu-
tant. RCRA and SARA analyses.
Minnesota Geophysical Associates
8616 Xylon Ave. North, Suite O
Brooklyn Park, MN 55445 612/493-3595
Minnesota Geophysical Associates offers coo-
suiting and contracting services in high-resolu-
tion surface and downholt geophysics. Surface
geophysical methods include seismic reflection
and refraction, electromagnetics (EM), and resis-
tivity. Downhote methods include natural
gamma, spontaneous potential (SP), single-point
resistance, caliper, and borehole video.
Naneo EnTtronmental Serried, Inc.
RDM Robinson Lane
Wappingers Falls. NY 12590 914/227-4100
Nanco Laboratories, a U.S. EPA, New Jersey
D.E.P. and Commonwealth of Virginia contract
laboratory, provide* complete environmental
analytical service* nationwide. Nanco is a Haz-
wrap approved laboratory. Nanco's services in-
clude: analysis for CLP, RCRA, ECRA. land
transfer, data interpretation, data validation,
electronic data delivery and quick turn-around
time analysis.
Ncri/Petm
309 Fannington Ave.. Suite A-100
Farmington. CT 06032 203/677-906
The Petrex Technique is an innovative geochem-
icaJ method for identifying and mapping volatile
organic compounds from soils and groundwater
contamination. The technique utilizes Petrex
monitors, which are placed in the ground in a
strategic pattern. After a representative subsur-
face sampling period, the monitors are removed
and analyzed by mass spectrometry.
NTH ConaaJtana Lid.
38955 Hills Tech Dr.
Fannington Hills. MI 48331-3432 313/553-6300
Geoenvironmental Services: hydrogedogk stud-
ies, groundwater modeling, environmental silt
assessments, geophysical surveys, underground
storage tank management, facility closure engi-
neering, compliance investigations, permitting
assistance, geoenvironmental monitoring, expert
witness testimony, health and safety plans for
environmental projects, construction monitor-
ing, as-built drawings and documentation, labor-
atory test services, leachate compatibility, stodge
stabilization and liner performance.
NUSCorp.
Wast* Managesnenl Service* Groin
Pittsburgh. PA 15275 412/788-1080
Oaithenburg. MD 20878 301/258-6055
For 27 yean NUS Corporation has provided the
environmental and engineering expertise to solve
industry and government waste management
problems with cost-effective solutions. Our staff
of over 2.000 multidisdplinary professionals
offers a full range of service* including environ-
mental assessment, environmental engineering,
remedial design and implementation, hydrogeo-
logic and geologic services, risk assessment, reg-
ulatory assistance, environmental health snd
safety and analytical service*.
666 EXHIBITOR PROFILES
-------
Nippi Tracking Corporation
P.O. Box 510
Matawan.NJ 07747
201/566-3000
Transportation and storage of hazardous and
non-hazardous waste.
National Draeger, Inc.
101 Technology
Pittsburgh, PA 15275
412/787-8383
National Draeger offers a wide range of pro-
ducts within the respiratory, instrumentation,
and detector tube lines. The Model 190 Data-
logger is the most advanced portable gas moni-
tor available for industrial hygiene and safety
professionals. It detects toxic gas and alarms in-
dependent of the microprocessor function. Na-
tional Draeger's air-purifying respirators include
cartridges for organic vapors, and gases and
ammonia, as well as high efficiency paniculate
filters for dust, fumes, mists, radionuclides and
asbestos.
National Environmental Testing, Inc.
Woodland Falls Corporate Park
220 Lake Dr. East, Suite 301
Cherry Hill, NJ 08002
609/779-3373
A growing nationwide network of environmental
testing laboratories dedicated to providing high
quality analytical data and superior customer
service. We offer a full range of environmental
analytical services backed by comprehensive field
services which include field sampling, stack test-
ing and industrial hygiene services.
National Lime Association
3601 N.Fairfax Dr.
Arlington, VA 22201
703/243-5463
Association of commercial lime producers who
supply quicklime and hydrated lime for treat-
ment of acidic and related wastes, many of which
are "hazardous." Lime is also used with fly ash
for the stabilization of hazardous waste. The
Association has movies and literature available
on lime including use in hazardous waste.
National Solid Waste Management
Association
1730 Rhode Island Ave., NW, #1000
Washington, DC 20036
202/659-4613
CWTI (Chemical Waste Transporters Institute)
and RCI (Remedial Contractors Institute) are
components of National Solid Wastes Manage-
ment Association to promote safe transport and
cleanup of hazardous waste sites. NSWMA is the
only association representing these interests for
Superfund and other state cleanups.
National Technical Information Service
5285 Port Royal Rd.
Springfield, VA 22161 703/487-4815
The National Technical Information Service
(NTIS) provides access to the results of both
U.S. and foreign government-sponsored R&D
and engineering activities. NTIS announces these
completed and ongoing results in its two data-
bases, the NTIS Bibliographic Database and
FEDRIP. The NTIS Bibliographic Database
contains over 1.4 million citations of completed
research, including technical reports, theses, bib-
liographies, software, datafiles, and inventions
available for licensing. It is available from BRS,
DATA-STAR, DIALOG, ORBIT, and STN in
the United States and several European vendors
including ESA-IRS and CISTI. The Federal Re-
search in Progress (FEDRIP) Database contains
citations of current and ongoing research from
nine Federal government agencies, including the
U.S. Public Health Service, NASA, and the De-
partment of Energy.
Northeastern Analytical Corp.
4 East Stow Rd.
Marlton, NJ 08053 609/985-8000
Environmental services include complete en-
vironmental field sampling, in-house gas chrom-
atography/mass spectometry (GC/MS) labora-
tory analysis, hazardous site training (40 hours),
asbestos inspection and management and abate-
ment monitoring services, asbestos analysis by
transmission electron and optical microscopy,
underground storage tank testing, excavation,
removal and installation, stack emission and am-
bient air testing.
OHM Corp.
16406 U.S. Route 224 East
Findlay, OH 45840
419/423-3526
OHM Corporation applies fully integrated en-
vironmental assessment, design, engineering,
implementation and treatment/disposal services
to hazardous waste and hazardous material con-
taminations nationwide and in Canada. Facilities
include 19 remediation service centers, 6 engi-
neering and technical centers, 6 laboratories, and
a fully permitted, fixed-base transfer, storage
and treatment facility.
OSCA Environmental Services, Inc.
1515 Poydras, Suite 2250
New Orleans, LA 70112
504/528-9184
OSCA Environmental Services, a subsidiary of
Great Lakes Chemical Co., provides a complete
line of Geo environmental engineering and field
remediation services. These services include pro-
ject management; site investigation and assess-
ments, UST remediation, installation of moni-
toring and recovery wells; excavation/decon-
tamination; dewatering; pit, pond, lagoon or
tank retrofit or closure.
OSCO, Inc.
P.O. Box 1203
Columbia, TN 38402
615/381-4999
OSCO is an environmental management com-
pany which operates a hazardous waste treat-
ment facility in Columbia, Tennessee. Materials
accepted are acids, bases, flammables, oily
waste, and many others in drums or bulk. Re-
medial services, including design, are provided
for clients. A fleet of transportation equipment
allows OSCO to handle the customers needs.
Occupational Hazards
1100 Superior Ave.
Cleveland, OH 44114
216/696-7000
Occupational Hazards Magazine is edited for
management officials who are responsible for
workplace safety, health and environment. Edi-
torial material includes coverage of major legis-
lative, regulatory, scientific and other develop-
ments affecting the field, as well as, practical
"how-to" articles.
Ogden Environmental Services
P.O. Box 85178
San Diego, CA 92138-5178
619/455-3045
Ogden Environmental Services has the solution
for Superfund cleanups. We specialize in provid-
ing a turnkey service for site remediation. This
service includes utilization of Ogden's trans-
portable circulating bed combustor (CBC). The
CBC is capable to safely and economically de-
stroying a wide variety of hazardous wastes to
levels over 99.99% without discharging harmful
emissions.
On-Slte Instruments
689 North James Rd.
Columbus, OH 43219
800/551-2783
On-Site Instruments/EnviroRENTAL sells,
rents, and services a complete line of industrial
hygiene, laboratory, and environmental moni-
toring instruments and equipment. Rent-to-own
and leasing options also available. Our service
department provides technical and applications
assistance, while our distribution center handles
all accessory orders. On-Site also offers training
classes at our Columbus, Ohio facility.
FACE Laboratories, Inc.
1710 Douglas Dr. North
Minneapolis, MN 55422
612/544-5543
PACE Laboratories, Inc. provides field services,
laboratory services, industrial hygiene services,
asbestos services and air sampling services.
PDC (Peorla Disposal Company)
4700 N. Sterling Ave.
Peoria, IL 61611 309/688-0760
PDC and its subsidiaries own and operate Part B
permitted hazardous waste disposal and treat-
ment facilities; and provide transportation serv-
ices, laboratory analysis, remedial response, and
consulting engineering.
Pollution Equipment News/
Rlmbach Publishing Inc.
8650 Babcock Blvd.
Pittsburgh, PA 15237 412/364-5366
Pollution Equipment News, published bi-
monthly, provides product information to the
person responsible for air, water, wastewater and
hazardous waste. An annual Catalog & Buyer's
Guide provides source information. Industrial
Hygiene News, published bi-monthly, provides
information on products and services for meas-
uring and controlling health hazards in the work
environment. An annual Catalog & Buyer's
Guide provides buying source information.
PRC Environmental Management, Inc.
303 East Wacker Dr., Suite 500
Chicago, IL 60601 312/856-8700
PRC EMI provides environmental services to
both government and industry. Headquartered
in Chicago, Illinois, PRC EMI maintains major
offices in McLean, Virginia and San Francisco,
California as well as 10 other offices throughout
the country. Specialties include remedial investi-
gations/feasibility studies, endangerment assess-
ments, remedial design and implementation,
compliance audits, permitting support, waste re-
duction audits, risk management support, en-
vironmental and systems engineering, policy and
regulatory analysis, economic analysis, and pro-
gram management support.
Pacific Analytical, Inc.
1989-B Palomar Oaks Way
Carlsbad, CA 92009 619/931-1766
Environmental analytical laboratory services
EXHIBITOR PROFILES 667
-------
with emphasis on CLP and SW-846 Method*:
Organic* by OC, HPLC and/or OC/MS. metals
by ICP-mass spectrometry. Special ties include
non-routine compounds, low level concentra-
tions, unusually complex matrices, sludge, bio-
logical tissue. CLP, CA DOHS certified.
Pennsylvania Drilling Company
500 Thompson Ave.
McKeesRocks.PAI5l36
412/771-2110
Pennsylvania Drilling Company has been install-
ing monitoring wells and drilling for hazardous
wastes for many years. The company began in
1900 to service the coal industry. To serve the
hazardous waste industry, we have OSHA trained
over 40 employees and invested in the latest
drilling equipment. In 1989 we moved into new-
ly renovated shops and offices to more efficient-
ly manage and coordinate projects. Our manu-
facturing facilities provide our customers with
unlimited options for trying out new and unique
ideas.
The PerklB-Efaater Corp.
761 Main Ave.
Norwalk. CT 06859-0012
203/762-1000
Perkin-Elmer offers products which perform a
wide range of analyses of hazardous materials
and materials of environmental concern. Spe-
cifically, gas chromatography, gas chromato-
graphy-ion trap detection; Zeeman furnace
atomic absorption, inductively coupled plasma-
mass spectrometry, and other techniques help
you solve both routine and complex problems.
Applications specialists will be on hand to dis-
cuss specific analytical situation.
PeroxMatte* Services. IMC.
4400 E. Broadway. Suite 602
Tucson, AZ 85711
517/456-4126
UV/peroridation water treatment equipment
and services. Destroys organics in contaminated
water with zero air emissions and no solid waste
residual.
Photon* lalcrutloaal lac.
741 Park Ave.
Huntington. NY11743
516/351-5809
Photovac will display portable instruments for
environmental toxic monitoring in groundwater,
soil, and ambient air: TIP™, a hand-held Total
Organics analyzer; the I OS Series Portable Oas
Chromatographs; and MicroTIP™. a hand-held
analyzer which incorporates advanced micropro-
cessor technology for real time digital or graphic
assessment of toxic gases and vapors.
Poly-America
2000 W.Marshall Dr.
Grand Prairie. TX 75051
214/647-4374. x355
Poly-America offers polyethylene geomembrane
for hazardous waste containment, landfill caps,
and waterways.
Polyfdt, Inc.
lOOOAbernathyRd.
Atlanta, OA 30328
404/668-2115
Producer of spunbond, continuous filament,
polypropylene nonwovens for soils engineering
applications. Specific uses include gas venting,
liner cushioning, and drainage infiltration.
Weight range from 2.7 to 22.0 oz/sq. yd.
Princeton Testing Laboratory
P.O. Box 3108
Princeton, NJ 08543
609/452-9050
Environmental analysis, industrial hygiene,
RCRA/ECRA, industrial wastewater NPDES.
groundwater, OSHA workplace surveys, asbes-
tos monitoring and evaluation, complete NIOSH
lab methodology, asbestos and HAZMAT train-
ing courses, Right-to-Know compliance, micro-
biology, bioassay, underground storage lank
testing, AIHA accredited. Certified for: NJ
DEP. NYDOH. PA DER, CT. RI. DE.
QED EaTlroameaUl Systems IK.
P.O. Box 3726
Ann Arbor, MI 48106 313/995-2547
The QED Environmental Systems family of pro-
ducts features Well Wizard* . the original ded-
icated bladder pump for groundwaier monitor-
ing; Sample Pro* portable samplers and sup-
plies; Pulse Pump™ pneumatic pumping sys-
tems for groundwater clean-up, and recovery,
and HydroPunch™ for groundwater sampling
without wells.
QU ALTEC. IK.
11300 U.S Hwy. One. Suite 500
Palm Beach Gardens, FL 33408
407/775-8396
QUALTEC, Inc. provides full-service on-site
environmental remediation including: consult-
ing, site assessments, remedial investigations/
feasibility studies, trea lability studies, bench/
full-scale pilot studies, landfill construction/clo-
sures, and the fixation/solidification of mod
types of contaminated soils, sludge* and ash.
QU ALTEC'S subsidiary, ENVECO. Inc.. pro-
vides leasing of specialized fixation equipment.
R.E. Wright AsMdates, lac. (ERS)
3240SchoolhouseRd.
Middletown, PA 17057 717/944-5501
Air stripping towers for removal of volatile or-
ganic compound* (VOCs) from groundwater or
process water; Auto-Skimmer for automatic re-
covery of floating hydrocarbons from water
wells; Wright modular recovery system pneu-
matic pumping system for recovery of floating
hydrocarbons or heavier-than-water contami-
nants; Depression pumps and controls for use in
subsurface oil spill recovery and aquifer restor-
ation projects; Liquid interface sampler for
quick sampling of liquids of differing densities
in the field.
R.E. Wright Aasociates, la*. (ERQ
3240 Schoolhouse Rd.
Middletown, PA 17057
717/944-5501
R.E. Wright Associates, Inc. (REWAI) offers
comprehensive environmental consulting and en-
gineering services. REWAI has extensive exper-
ience in the area of hazardous waste and hydro-
carbon pollution remediation. The firm's pro-
fessional involvement has spanned the full spec-
trum of necessary investigative functions.
REWAI routinely performs feasibility studies
utilizing contaminant transport modeling analy-
ses. In addition, REWAI regularly handles pro-
jects through final design and construction.
REWAI also provides technical services in the
areas of solid waste management, wastewater
treatment and disposal, groundwater resources
management, and environmental engineering
and planning.
RMCEavtrosusMsialSef
RD/Ci, Prick* Lock Rd.
Pottstown, PA 19464
215/3264(62
Analytical Laboratory Services; Geotechnkal
Services, including groundwater monitoring,
well siting and installation, hazardous site inves-
tigations; environmental consulting, induing
natural resource inventories, population studies,
biological sampling, tissue analysts, bioassayi,
wetlands studies.
RMT.IBC.
P.O. 80x8923
Madison. WI 53708-8923
608/832-4444
RMT is a consulting engineering firm specializ-
ing in solid and hazardous waste management,
groundwater monitoring and analysis, industrial
hygiene engineering, environmental control, and
lab services. Business includes planning, design,
and permitting of hazardous waste treatment,
storage, and disposal facilities, and the investi-
gation, design, implementation and monitoring
of remedial measures.
512/4544797
P.O. Box 201068
Austin, TX 78720-1088
Radian Corporation provides a full range of
process, solid and hazardous waste engineering
services...including rite •"•fftrrt to remedia-
tion design and construction, waste minimiza-
tion to the design of waste treatment or dis-
posal systems, and preparing permit applications
to responding to consent orders. In addition, the
company has three full-service laboratories pro-
viding complete characterization and classifica-
tion of soils, groundwater, run-off, kachatcs,
air emissions, soil vapors, and virtually any other
substance or material for which measurements
are required. Radian also has the unique ability
to perform remedial pilot studies on site. Thii is
accomplished through our transportable treat-
ment systems. The unit physical-chemical opera-
tions incorporated into these systems can be con-
figured to treat most contaminated waste
streams. These systems have sufficient capacity
to provide full-scale groundwater remediation.
Recn Eartroasawlal. lac.
10 Hazelwood Dr.. Suite 106
Amherst, NY 14150
716/691-2600
Recra Environmental, Inc.. with laboratory
facilities in Amherst. New York (Buffalo) and
Columbia. Maryland (Washington/Baltimore),
offers full-service environmental testing services
encompassing all matrices. Both of our faculties
are U.S. EPA CLP contractors and are certified
in numerous states. Services include data man-
agement and electronic data transfer.
Reaacor, la*.
701 Alpha Dr.
Pittsburgh, PA 15238
412/963-1106
Remcor provides the full spectrum of hazardous
waste consulting and remediation services. By
uniquely integrating expertise in engineering,
construction, and environmental field services,
Remcor performs projects ranging from investi-
gations through actual remediation. As a turnkey
contractor, Remcor has completed numerous
projects including building decontamination,
surface impoundment closure, landfill clean-
ups, storage tank management, asbestos removal
and groundwater remediation.
668 EXHIBITOR PR(5i;IU-S
-------
Remediation Technology, Inc. (REMTECH)
273 Franklin Rd.
Randolph, NJ 07869 201/361-8840
Remediation Technology, Inc. (REMTECH) pro-
vides services related to on-site remediation of
contaminated sites, engineering services aimed at
providing innovative solutions to environmental
problems, and technology transfer of unique and
novel cleanup technologies.
Resource Analysts, Inc.
Box 778
Hampton, NH 03842
603/926-7777
Resource Analysts, Inc., a subsidiary of Milli-
pore Corporation, provides environmental
chemistry, aquatic toxicology and field sampling
services to industrial and governmental clients.
Resource Analysts, Inc. participates in EPA's
Contract Laboratory Program and is Army and
Navy approved for DOD site restoration serv-
ices. Facilities include 24,000 sq. ft. dedicated to
chemistry laboratory and 10,000 sq. ft. dedi-
cated to fresh water and marine aquatic toxi-
cology laboratory. RAI maintains a professional
staff of 85 scientists, technicians and admin-
istrators. The company prides itself on its unique
customer satisfaction program.
Response Rentals
1460 Ridge Rd. East
Rochester, NY 14621
800/242-3910
Response Rentals provides rental instrumenta-
tion for remedial investigation studies, compli-
ance surveys and substance emergencies. The in-
strumentation is easy to operate, reliable and
represents the best names in the industry. Broad
product line meets virtually every application
need and includes OVAs, CGIs, PIDs, Iso-
thermal GCs and more.
HoyF. Weston, Inc.
Weston Way
West Chester, PA 19380
215/430-3025
Weston is a full-service environmental engineer-
ing firm specializing in analytical laboratory
services, consulting and engineering, remedia-
tion, facility construction and operations, tech-
nical information management and the manage-
ment of major programs. Weston employs more
than 2,600 people from various disciplines, wholly
owns 8 subsidiaries and now has 42 offices nation-
wide.
Roika Laboratories, Inc.
3601 Dunvale Rd.
Houston, TX 77063
713/975-0547
The Ruska PYRAN ThermaChromTM universal
spectrometer interface will be featured via photo-
graphs and data displays. This Thermal Chrom-
atograph1™ permits the direct analysis of soils,
sludges and other solid and semi-solid materials
for organic contamination without the need for
solvent extraction or other sample preparation
techniques. Both mobile and fixed lab configur-
ations will be depicted.
S-CUBED,
A Division of Maxwell Laboratories, Inc.
P.O. Box 1620
La Jolla, CA 92038 619/453-0060
Chemical Analysis Services: CLP organic analy-
ses, RCRA analyses, Methods 1618, 1624, 1625
analysis for OWRS samples, Appendix IX com-
pounds, inorganic Analytes; Quality Assurance
Support-BDAT, SITE, OPP Projects: QA pro-
ject plan reviews, final report reviews, field aud-
its, QA training; Analytical Methods Develop-
ment and Research; Environmental Engineering:
site investigation/field sampling and monitoring,
treatability studies, solidification/stabilization.
SCS Engineers
11260 Roger Bacpon Dr.
Reston, VA 22090
703/471-6150
Solid and hazardous waste consulting services
since 1970. Specialists in control and treatment
of subsurface gases; a subsidiary of the firm
offers construction services in this area. Remed-
ial investigations, feasibility studies, design, con-
struction management. Analytical laboratory,
underground tank testing, construction. Real es-
tate contamination assessments. Hazardous
waste facility design and permitting.
Sensidyne, Inc.
16333 Bay Vista Dr.
Clearwater, FL 34620 813/530-3602
Portable and fixed gas detection and air sampling
equipment. Equipment includes monitoring in-
struments such as Sensidyne's FID, Odor Moni-
tor, Hazardous Material Kits, Detector Tubes,
Personal Toxic Monitors, Personal Sampling
Pumps, Continuous Toxic and Combustible
Monitors.
Sentex Sensing Technology Inc.
553 Broad Ave.
Ridgefield, NJ 07657
201/945-3694
Manufacturer of: portable gas chromatographs;
used to monitor TLV levels of contaminants in
air, water or soil. Multi-point monitoring sys-
tems: for continuous PPM/PPB analysis of haz-
ardous gases. Portable purge and trap gas
chromatograph: for automatic/accurate analysis
ofVOCsinwater.
SLT North America, Inc.
16945 Northchase, Suite 1750
Houston, TX 77060
713/874-2150
The pioneer in HOPE lining systems, SLT man-
ufactures and installs 34-ft wide seamless mono-
lithic sheets, with engineered innovations such
as: Hyperflex™, PolyLock™, and Friction-
Flex™. Manufactured material thickness from
40 to 240 mils, SLT is your solution for ponds,
tanks, tunnels, landfills, and mining applica-
tions.
SMC Environmental Services Group
P.O. Box 859
Valley Forge, PA 19482
215/265-2700
SMC Environmental Services Group provides
consulting services in the fields of geology, hy-
drogeology, biology, civil and environmental en-
gineering, planning, surveying, environmental
science, and computer technology. Founded
more than 35 years ago, SMC is comprised of a
staff of nearly 60 engineers and scientists and
serves a broad range of clients from Fortune 500
industries to municipal authorities.
SSI Shredding Systems
28655 SW Boones Ferry Rd., P.O. Box 707
Wilsonville, OR 97070 503/682-3633
SSI Shredding Systems provides on-site volume
reduction and material processing of solid haz-
ardous waste prior to material treatment. Spe-
cific services include pre-processing, feedstock
preparation and volume reduction of solid haz-
ardous waste utilizing mobile low-speed, rotary
shear shredders. This low rpm equipment is easy
to trailer mount and once on-site is operational
within hours. OSHA certified operators are pro-
vided.
Systech Environmental Corp.
245 N. Valley Rd.
Xenia, OH 45385
513/372-8077
Supplemental fuel use of organic liquid wastes in
a rotary cement kiln.
Science Applications International Corp.
8400WestparkDr.
McLean, VA 22102 703/734-4302
SAIC has been providing environmental man-
agement services to government and commer-
cial clients for nearly 20 years from offices across
the country. Our services include RI/FS, design
engineering, construction management and com-
pliance programs for air, water and hazardous
waste regulations. Special capabilities include
laboratory services, health and safety training
and clean-up technology development and
demonstration.
Scientific Specialties Service, Inc.
P.O. Box 352
Randallstown, MD 21133
301/964-9666
Scientific Specialties Service, Inc. is showing its
line of environmental sampling supplies includ-
ing precleaned and regular vials, bottles, and jars
in both glass (which is also available Safety-
Coated if desired) and plastic. They are also
showing their Teflon® Capliners and Teflon0/
Silicone septa and their line of Teflon® sealing
tapes in an extensive range of sizes.
Serrot Corp.
5401 Argosy Dr.
Huntington Beach, CA 92647
714/895-3010
Serrot Corporation is a full-service company
specializing in the fabrication and installation of
geomembrane lining systems and floating covers.
Serrot is fully experienced in the full range of
materials available for a multitude of applica-
tions including: hazardous waste, sanitary land-
fills, potable water, wastewater, tanks, mining,
leach pads, process water, methane barriers,
floating covers and specialty applications. Serv-
ice is available on a national basis.
Sevenson Environmental Services, Inc.
2749 Lockport Rd.
Niagara Falls, NY 14302 716/284-0431
Sevenson Environmental Services, Inc. provides
remedial construction services to government
and industry in site remediation; excavation,
characterization, transportation, and disposal of
bulk and drummed wastes; secure landfill and
lagoon construction/closure; slurry wall con-
struction; sludge solidification and fixation; re-
covery and treatment systems installations for
groundwater soils; leachate collection and treat-
ment systems installations for groundwater
soils; leachate collection and treatment systems
construction; on-site incineration, biological re-
mediation; facilities decontamination and demo-
lition; dewatering; and storage tank removal/re-
mediation.
EXHIBITOR PROFILES 669
-------
Site Reclamation Syllcmi
P.O. Box 11
Howey-in-the-Hills, FL 34737
904/324-3651
Site Reclamation Systems. Inc. (SRS) designs,
constructs, and operates mobile soil volatilizeri
which clean soils contaminated with petroleum
hydrocarbons (e.g., gasoline, #2 fuel oil, jet fuel,
kerosene, etc.). The distinct advantages over
conventional disposal options are:
• Eliminates long-term liability associates with
conventional transportation and disposal
• The process is cost-effective and competitive
with conventional disposal methods
• Complies with new federal and state regula-
tions which require waste destruction ind
minimization whenever possible
Skolnlk Industries, Inc.
4900 S. Kilbourn Ave.
Chicago, IL 60632
312 735-0700
New steel containers (Overpack/Salvage. car-
bon, composite and stainless) from 8 to 110
gallons for shipping and storing hazardous and
non-hazardous waste materials, as well as con-
tainer liners, HazMat containment kits, com-
ponents, dollies, utility cans, tools and accessor-
ies. Custom design and fabrication is ako avail-
able.
SoliMl Canada Ltd.
The Williams Mill, 515 Main St.
Glen Williams, ON L7G 39S
416/873-2255
Solinst manufactures high quality groundwater
monitoring instrumentation. New on display at
Super fund '89 will be an oil/water "Interface
Probe." The probe has the rugged durability and
easy operation of Solinst Water Level Meters,
and accurately measures depth of product layers
and depth to water. The Waterloo Multilevel
System will also be on display.
Soil*wot Laboratory of Oklahoma
1700 West Albany. Suite C
Broken Arrow. OK 74012 918/251-2858
Quality and service oriented analytical labora-
tory offering comprehensive analysts for
CERCLA, SARA, RCRA, priority pollutants.
Dioxins/Furans, Appendix IX, explosives and
TCLP. SWL is a full participant in the U.S.
EPA CLP with multiple contracts in organics,
high hazards and inorganics. Also certified by
Corps of Engineers for DERA projects.
Southwell Research Institute
6220CulebraRd.
San Antonio, TX 78228-0510
512/522-2687
Southwest Research Institute provides commer-
cial leak location surveys of geomembrane liners
for landfills, impoundments, and lined tanks to
accurately locate leaks in the material and seams.
Analytical laboratory systems and techniques
will be presented for both the sampling and
analysis of environmental pollutants. Bio-degra-
dation techniques will also be discussed.
Specialized Environmental Equip., Inc.
Rt. 4, Box 216
Easley, SC 29540 803/859-8277
Specialized Environmental provides Mobile
Laboratories: chemical analysis units, water
pollution analysis units, and decontamination
units; Special Service Units; Bioassay Dilutor
Systems; and Water Baths, Dual Purpose Pumps
and Oxygen Demand Apparatus.
Sloul Environmental Inc.
2880 Bergey Rd.
Hmfield, PA 19440
215/822-2676
Stout Environmental, Inc. is a full-service en-
vironmental management company providing
treatment and disposal of hazardous, industrial,
and municipal wastes, along with a broad range
of specialized support services. Our 15 service
divisions enable us to offer a turnkey approach
to environmental problems providing timely and
cost-effective solutions.
Summit Interest*
1801 Sunset PI. .Suite D
Longmont, CO 80501
303/772-3073
SIP-1000 Portable Oas Analyzer—The SIP-
1000 is a small, portable, self-contained instru-
ment that is both a continuous monitor and gas
chromatograph and is capable of detecting gases
in the PPB range. "A New Twist on Detectors"
allows the operator to quickly change detector
types (PID, FID. TCD). Another feature is the
incorporation of a solid state carrier gas system
that utilizes metal hydrides.
Surety Specialist*. IK.
1501 2nd Ave. E., P.O. Box 5098
Tampa. FL 33675-5098
813/247-0118
Surety Specialists, Inc.—Our name says it all.
Surety provides Surety Bonds for all types of
contractors. We are experts with "tough" cases,
including asbestos, environmental and demo-
lition contractors. We specialize in creative
underwriting and represent Treasury-listed and
Best-rated bonding companies.
Sorely TeksuciasN, Inc.
6200 Courtney Campbell Cswy., Suite 685
Tampa, FL 3J607 813/872-1810
National bond agency specializing in the place-
ment of all types of hard to write contract bonds,
including environmental remediation, asbestos
abatement, and demolition contracts; represent-
ing over two dozen carriers including Treasury-
listed and Best-rated companies.
Sybron Chemical* Inc./Biochemical Division
Birmingham Rd., P.O. Box 66
Birmingham, NJ 08011 609/893-1100
Leaden in the application of augmented bio-
reclamation (ABR) for the treatment of contam-
inated soil and groundwater. Capabilities include
biosyslems engineering services and supply of
selectively adapted organisms for specific con-
taminants. Technology useful for cleanup of
chemicals from leaking storage tanks, pipeline
spills, train derailments, etc. Advantages are
ultimate disposal technology and low cost.
TAMS ConultanU, Inc.
655 Third Ave.
New York, NY 10017
212/867-1777
TAMS, a leading international engineering and
scientific firm, offers comprehensive services in
solid and hazardous waste management. Signif-
icant experience includes Rl/FS; health/safety;
risk assessment; community relations; remedial
design; construction oversight; site closure;
waste geotechnics; chemical/process design;
watershed management; hydrogeology/mathe-
matical modeling. TAMS provides services to
clients in government, military and private sec-
tors through offices in major cities.
TARGET Soil Gas Surveys
8940-A Rte. 108 Oakland Center
Columbia, MD 21045
301/992-6622
TARGET specializes solely in providing soil gas
services nationwide for fast and accurate screen-
ing of VOCs in the subsurface. TARGET'i
advanced soil gas surveys have been used for de-
tecting and assessing suspected VOC problems,
delineating the extent of a spill, and/or moni-
toring the progress and success of a remediation
effort.
TlggCorp.
P.O. Box 11661
Pittsburgh, PA 15228
412/563-4300
Manufacturers of modular adsorbers designed
for the remediation of vapor and water pollu-
tion. The combination of over 30 yean of exper-
ience with adsorbents and systems provides
unique capabilities of technical expertise and
product availability to address specific remedial
problems with the most appropriate technology.
TPS Technotofie* Inc.
2070 S. Orange Blossom Trail
Apopka, FL 32703
407/886-2000
TPS Technologies Inc. provides on-site thermal
treatment for petroleum-contaminated soils.
TPS Technologies' thermal treatment units have
statewide permits and can be operated anywhere
in the State of Florida, as well as in many other
nates. TPS Technologies offers the optimum
package: reduced liability exposure, competitive
pricing, minimum site disruption and immediate
response.
TecfcLaw Inc.
14500 Avion Pkwy., Suite 300
Chantilly. VA22021 1101
703/818-1000
TechLa*. an environmental consulting firm ex-
perienced in the application of legal and techni-
cal principles to tasks in support of RCRA and
CERCLA litigation activities, provides enforce-
ment services including: PRP search, imaging
services, tracking systems, evidence audits, docu-
ment control systems, legal research, full text
database, transactional database, data validation
and compliance audits.
Tekmar Company
P.O. Box 371856
Cincinnati, OH 45222-1856
513/761-0633
LSC 2000 Series Purge and Trap Concentrators
for analysis of volatile organic compounds in
environmental samples of water, soils, and
sludges. Also dynamic headspace analysis on
food samples and polymers. M5010GT Auto-
matic Thermal Desorber for analysis of volatile
organic compound* in air samples (both ambient
and industrial hygiene).
TenaxCorp.
8291 Patuxem Range Rd.
Jessup, MD 20794 301/725-5910
Tenax Corporation manufactures a full line of
Geosynthetics for waste management applica-
tions. The Tenax line of products include Drain-
age nets and Oeocomposites for leachite collec-
tion, Oeogrids for side slope reinforcement and
haul road stabilization, fencing for safety delim-
itation and litter control. Technical assistance
and design are available from Tenax Engineer-
ing Department.
670 EXHIBITOR PROFIU.S
-------
Tetra Tech, Inc.
630 N.Rosemead Blvd.
Pasadena, CA 91107
818/449-6400
Tetra Tech specializes in environmental science
and engineering, hazardous waste management,
and assessment of risks to human health and the
environment. Tetra Tech has more than 20 years
experience managing and conducting major en-
vironmental and engineering services contracts
for industry and government, including exten-
sive nationwide experience in hazardous waste
and underground storage tank management.
Thermo Analytical, Inc.
5635 Jefferson Blvd. NE
Albuquerque, NM 87109
505/345-9931
TMA's network of laboratories and service facil-
ities provides a full range of analyses of organic
and inorganic chemicals and radioactive ma-
terials in soil, water, air, industrial wastes, and
biological materials. In addition to these analyti-
cal services, TMA offers health physics, indus-
trial hygiene, and dosimetry consulting services.
Thermo Environmental Instruments
SW.ForgePkwy.
Franklin, MA 02038
508/520-0430
Thermo Environmental Instruments, Inc. is the
world's leading manufacturer of U.S. EPA ap-
proved Ambient Air Pollution Analyzers, Ex-
tractive and In-Situ Stack Emission Monitoring
Systems and Toxic Chemical Analyzers. Analy-
zers included in our product line are NOX,
C02, S02, CO. O3, OVM, HC, and OCs.
Toney Drilling SuppUes, Inc.
14060 NW 19th Ave.
Miami, PL 33054
305/685-2453
Complete line of drilling equipment: New/used
drill rigs, drill rods, subs and bits. Diamond
bits, core barrels, mud and additives; augers,
casing and plugs; stainless steel screens, PVC
screens, points and pcaps; monitoring and
sampling devices; safety clothing, masks, gloves
and boots. Consultation and instruction are also
available.
Tracer Research Corporation
3855 N. Business Center Dr.
Tucson, AZ 85705
602/888-9400
Tracer Research Corporation specializes in Leak
Detection for underground storage tanks, bulk
storage, above ground tanks and pipelines;
Tracer Technology for groundwater monitor-
ing and landfill liner tightness testing; and on-
site detection of subsurface volatile organic con-
taminants (soil gas analysis). Tracer offers full-
service organic analysis laboratory services.
TracorXray, Inc.
345E.MiddlefleldRd.
Mountain View, CA 94043
415/967-0350
Field model XRF system for on-site screening
for priority metals in contaminated soil without
site specific standards. System uses single source,
X-ray tube excibation, electrically cooled solid
state detector and an IBM PC.
TreaTek* , Inc.
2801 Long Rd.
Grand Island, NY 14072
716/773-8660
TreaTek* is an environmental service subsid-
iary of Occidental Chemical Corporation, and
has as its commercial objective the application
of advanced microbial and chemical treatment
technologies to the remediation of waste streams
and contaminated soil. TreaTek® can provide
remedial consultation, laboratory treatability
study (biological, chemical and physical), ana-
lytical support, system design and specification
and turnkey project management.
Triangle Laboratories, Inc.
801-10 Capitola Dr., P.O. Box 13485
Research Triangle Park
NC 27709
919/544-5729
Triangle Laboratories, Inc. is an EPA-approved
contract laboratory for Organics. The company
was founded in 1984 and is privately owned.
Its 20,000 sq. ft. facilities are located in Alston
Technical Park in Research Triangle Park, North
Carolina. Services include high resolution GC/
MS analysis for Dioxin and Furans.
U.S. Analytical Instruments
1511 Industrial Rd.
San Carlos, CA 94070
415/595-8200
Available for rent and immediate delivery—
HNU Model 101s, FOXBOROL OVA 128GCs,
and MIRAN IBs from U.S. Analytical Instru-
ments. In addition, USAI offers for rent or lease
GC, HPLC, Fluorescence, UV/VIS, AA and
ICP, IR and FTIR instrumentation from major
manufacturers such as Hewlett-Packard, Perkin
Elmer, Varian, Foxboro, Hitachi and Waters.
We offer flexible rental and purchase option
plans designed to meet your financial and instru-
mentation needs.
U.S. Army Corps of Engineers
P.O. Box 103, Downtown Station
Omaha, NE 68101-0103 402/69M533
The U.S. Army Corps of Engineers and the
U.S. EPA have joined forces to clean up Fed-
eral lead hazardous waste sites under the Super-
fund program. The booth will be manned by
Corps personnel to assist architect-engineer firms
and construction contractors to take advantage
of work available to them through the Corps of
Engineers.
U.S. Army Toxic & Hazardous
Materials Agency
Bldg. E4460, ATTN: CETHA-PA
Aberdeen Proving Ground
MD 21010-5401
301/671-2556
The U.S. Army Toxic & Hazardous Materials
Agency, located at Aberdeen Proving Ground,
Maryland, is a field operating agency of the
U.S. Army Corps of Engineers that offers a wide
spectrum of environmental support services to
Army installations nationwide.
U.S. Bureau of Mines
2401 E St., NW, MS 6201
Washington, DC 20241
202/634-1224
The Bureau of Mines is a Federal Government
agency under the U.S. Department of the Inter-
ior. The Bureau's mission involves conducting
research and gathering minerals-related data that
will help strengthen our domestic minerals indus-
try. Among its many research programs is a pro-
gram dedicated to alleviating or solving environ-
mental problems plaguing the minerals industry.
Promising technologies have evolved from this
program and are readily available to companies
desiring to use them through technology transfer.
U.S. Bureau of Reclamation
Mail Code D-3210, P.O. Box 25007
Denver, CO 80225-0007 303/236-8646
The U.S. Bureau of Reclamation provides Total
Project Management in hazardous waste site
cleanup—PA/SI, RI/FS, RD, RA, and O&M.
Work may be completed for other government
agencies in planning, design, construction, con-
struction oversight, reviews, or research. Work
has been completed under RCRA, Superfund,
and Federal Facilities section of CERCLA.
U.S. Envlrosearch, Inc.
445 Union Blvd., Suite 225
Lakewood, CO 80228
303/980-6600
A nationwide recruiting firm based in Denver,
Colorado, specializing in the recruitment of haz-
ardous waste, environmental and incineration
personnel. U.S. Envirosearch represents client
companies in the areas of hazardous waste dis-
posal, site remediation, environmental engineer-
ing, air quality, analytical labs, solvent recycling,
PCB disposal, industrial cleaning and generators.
U.S. Geological Survey
790 National Center
Reston,VA 22092
703/648-4377
Panels depicting research and products of the
U.S. Geological Survey dealing with earth
sciences will be displayed.
ULTROX International
2435 S.Anne St.
Santa Ana, CA 92704
714/545-5557
The innovative ULTROX® process utilizes
ultraviolet light with ozone and/or hydrogen
peroxide to destroy toxic organic contaminants
in groundwater, surface waters, wastewaters and
leachate, on site. No sludges or wastes are gen-
erated requiring regeneration, disposal or incin-
eration. ULTROX® is used as a stand-alone
treatment system and with other technologies.
USPCI, Inc.
515 West Greens Rd., Suite 500
Houston, TX 77067
713/775-7800
USPCI, Inc., headquartered in Houston, Texas,
is a professional hazardous waste management
company offering a complete range of services
involving the treatment, disposal, analysis and
transportation of hazardous industrial waste. It
has become one of the leaders in the industry,
providing services to a wide range of industrial
and government customers throughout the
United States.
Utensco/P&D
P.O. Box 710
Port Washington, NY 11050 516/883-7300
Utensco/P&D manufactures steel secondary
containment systems for the safe storage, dis-
pensing and transportation of hazardous waste
materials. All units feature an internal catch
basin designed to hold spills or leaks. The
volume of the catch basin meets or exceeds cur-
rent government environmental and safety regu-
lations.
Union Carbide Industrial Gases Inc./
Llnde Division
39OldRidgeburyRd.
Danbury, CT 06817 203/794-5601
America's leading producer of industrial gases,
including oxygen and nitrogen, Linde® Com-
EXHIBITOR PROFILES 671
-------
bustion System can safely double the capacity of
your incinerator, reducing CO excursions and
auxiliary fuel consumption. Sec us to learn about
our installation at the BROS Superfund site.
United Stale* Testing Co., Inc.
1415ParkAve.
Hoboken, NJ 07030 201 /792-2400 x325
Laboratory analytical services (EPA CLP) chem-
istry, environmental chemistry, radio chemistry,
asbestos, biology, geo-technical-specialists in pri-
ority turn-around with 10 OC and 17 OC/MS.
Tke University of Flndlay
1000 N. Main
Findlay, OH 4)840
419/424-4540
Training and education provided in the areas of
hazardous materials/waste, emergency response,
spill response, confined space entry, asbestos re-
moval, 40 hour OSHA, 8 hour OSHA and
OSHA site supervisor training. Hands-on train-
ing facility. On-tite training available upon
arrangement.
VIC Manufactoring
1620Central Ave., ME
Minneapolis, MN 35413
612/781-6601
VIC Manufacturing produces carbon adsorption
systems for solvent recovery and emissions con-
trol for a wide range of manufacturing applica-
tions. In addition to the sale and service of such
systems, VIC offers contract engineering services
including:
• Exhaust stack emission analysis
• Preliminary engineering designs and drawings
• Complete installation designs for adsorption
and related equipment
• On-site engineering supervision of contractors
• Start-up services and maintenance seminars
VTL Technology Corporation
42 Lloyd Ave.
Malvern. PA 193)5
215/296-2233
VTL Technology Corporation is a civil/gco-
technical construction firm specializing in the de-
sign and implementation of solutions to a variety
of waste management problems. Services include
soil/sludge solidification and stabilization,
lagoon/landfill closures, hazardous site remedia-
tion, groundwater recovery and treatment, on-
site treatment systems, excavation treatment and
disposal of contaminated materials on-lite or
off-site.
VSI Environmental Servket, li
P.O. Box 2878
Baltimore, MD 21225-0878
301/636-1490
When it comes to industrial cleaning, site decon-
tamination, toxic waste removal and paving,
count on VSI Environmental Services, Inc. to
handle every aspect of the job with competence
and care from initial consult to follow-through.
VSI provides equipment, labor, supervision,
and, most important, the expertise to get the job
done.
Venar, IDC.
68)0 Versar Center, P.O. Box 1)49
Springfield, VA 221) 1 703/7)0-3000
On a nationwide basis, Versar provides a wide
range of scientific and technical services to com-
mercial and industrial clients. We assist our
clients in hazardous waste (RCRA/CERCLA/
SARA) and toxic substance control, industrial
hygiene, air and water quality management,
regulatory compliance assistance, ecological and
human lexicological risk assessment, and mulU
media sampling and analysis.
Vlar and Company, Inc.
209 Madison St.
Alexandria. VA 22314
703/684-)678
Viar and Company delivers to the government
and its prime contractors practical solutions to
complex environmental information manage-
ment problems utilizing state-of-the-art telecom-
munications, microcomputer, local area net-
work, and expert systems technology. Our areas
of focus include: environmental data collection,
quality assurance, enforcement and compliance
monitoring, financial management and program
management and decision support.
WAPORA
7926 Jones Branch Dr.. Suite 1100
McLean, VA 22102 703/893-3904
KEMRON Environmental Servkes/WAPORA
are two nationally recognized laboratory and
environmental consulting companies. They pro-
vide full turnkey services from Site Investiga-
tion/Engineering Services, Remediation Serv-
ices, Chemical and Analytical Services, Asbestos
Management/Industrial Hygiene and Environ-
mental Management. Stop by booth M>919 to
see what they can do for you. MBE certified.
WATEC/ATEC Associates, Inc.
1300 Williams Dr.. Suite B
Marietta, OA 30066-6299 404/427-1947
Waste Abatement Technology. Inc. (WATEQ
is a full-service remedial construction contractor
based in Marietta, Georgia. Site restoration serv-
ices include: lagoon closures, drum characteriza-
tion and overpacking, soil excavation, ground-
water treatment, building decontamination, and
asbestos abatement. WATEC's parent company.
ATEC Associates, provides geotechnical and
environmental engineering support as well as en-
vironmental drilling services from 44 offices
across the nation.
Wa4iworth/Alert Laboratories
4101 Shuffel Dr.. NW
North Canton, OH 44720
216/497-9396
Complete environmental analytical services. Par-
ticipant in U.S. EPA Contract Laboratory Pro-
gram, sampling, mobile laboratories, industrial
hygiene services AIM A approved.
Watenaver Conspany, Inc.
P.O. Box 1646)
Denver. CO 80216
303/289-1818
Watersaver provides the world's most reliable
membrane lining systems. Meet all state and fed-
eral regulations with Watersaver. Liners and
closure caps for a wide variety of applications.
Custom fabrication and installation of CSPE,
CPER, PVC, XR-5, and others. Continuous
service for over 30 years.
Wayne Associates, Inc.
2628 Barrett Si.
Virginia Beach, VA 23452
804/340-0)))
Wayne Associates, Inc. Is responsible for the
staffing of many of the nation's environmental
firms. Our services include contract and con-
tingency search and our expertise involves na-
tionwide opportunities for chemists, engineers,
hygienists, sales/marketing and groundwater
specialists. Stop by Booth #1801 to discuss your
needs or investigate career alternatives.
Wekraa EnrlroTeck be.
666 But Main St., P.O. Box 2006
Middietown, NY 10940 914/343-0660
An environmental consulting firm of engineer!
and scientists thai provides a full range of haz-
ardous waste management services including:
site investigations, Rl/FS, hydrogeologk assess-
ments, real estate site assessments, aquifer pro-
tection and restoration, leachate/groundwater
treatment, and remedial engineering.
Wc*ib*y Isvtniaeni* Inc.
507 E. Third St.
North Vancouver, BC
Canada V7LIOA
604/984-4215
Wesibay manufactures the MP System, a multi-
level groundwater monitoring system capable of
providing access to any number of intervals from
a single casing. Westbay provides assistance in
selecting the monitoring well design appropriate
for each project as well as extensive field training
and support.
Weatfaaffcoue Environmental A
Geological Services, lac.
5240 Panola Industrial Blvd.
Decatur, OA 3003)
404/593-3464
Full-service supplier for all environmental and
hazardous waste management needs. Major serv-
ice areas include: risk and environmental assess-
ment, remedial engineering, site remediation,
emergency response, analysis, waste minimira-
tion and off-site treatment and disposal.
WUttasB Afford Associates,
122 East 42nd St.
New York, NY 10168
212/557-4742
We are leading executive recruiters in the solid
and hazardous waste industries. Assignments in-
clude collection and disposal; remediation; en-
vironmental, civil, mechanical and chemical en-
gineers; laboratory personnel; marketing and
sales of services and equipment.
WUson Laboratories
525 North 8th St.
Salina. KS 67401
913/825-7186
Wilson Laboratories services include: ground-
water, drinking water, and wastewater analysis;
toxic and hazardous waste analysis, including
standard analyses for inorganics, organics and
PCBs in various matrices.
YWC, Inc.
6490 Premier Ave., NW
North Canton, OH 44720
216/4994181
YWC, Inc. is an environmental services and pro-
ducts organization offering environmental en-
gineering capabilities, contract laboratory facil-
ities, wasiewater treatment design, build and
operate services, complete municipal and indus-
trial sludge pumping and dewatering, software
management of environmentally-mandated datt
(SARA. OSHA. etc.), OSHA/RCRA training,
transportation and disposal of RCRA and TSCA
wastes and distribution of polypropylene sor-
benis and bags, vapor suppressants, solidifica-
tion agents and bioculture agents for waste-
waiers and contaminated soils.
672 EXHIBITOR PHOI-II I.S
-------
Zlmpro/Passavant Inc.
301W. Military Rd.
Rothschild, WI54474 715/359-7211
Zimpro/Passavant is the developer of the
PACT® wastewater treatment system, which
uses powdered activated carbon and wet air oxi-
dation. Both technologies are part of EPA SITE
program. It is effective on contaminated ground-
water, leachates, process discharges, other haz-
ardous wastewaters. ZP has complete pilot plant
facilities for treatability studies.
EXHIBITOR PROFILES 673
-------
674 EXHIBITOR PROFILES
-------
AUTHOR INDEX
1980-1989
Abbott, C. K, 89-23
Abraham, John E., 88-524
Absalon, J. R, 80-53
Accardi, J., 85-48
Aceto, F., 89-273
Adamowski, S. J., 83-346
Adams, R. B., 84-326
Adams, W. M ., 83-108
Adams, W. R, Jr., 82-377, 83-352
Adaska, W. S., 84-126
Adkins, L. C, 80-233
Adrian, D. D., 89-519
Aguwa, A. A., 86-220
Ahlert, R. C, 82-203; 83-217; 84-393
Ahnell, C. P., Jr., 80-233
Ainsworth, J. B., 83-185
Alam, Abu M.Z., 87-111
Albrecht, 0. W., 81-248, 393
Aldis, H., 8343
Aldous, K., 80-212
Alexander, W. J., 82-107
Allcott, G. A., 81-263
Allen, Douglas C., 88-329
Allen, E. E., 89-485
AUen, Harry L., 81-110; 88-424
Allred, P. M., 88-528
Aim, Roger R., 87-480
Alther, George R., 88-440; 89-543
Alvi, M. S., 84-489
Amdurer, Michael, 87-72
Ammann, P., 84-330
Ammon, D., 84-62, 498
Amos, C. K., Jr., 84-525
Amster, M. B., 83-98
Anastos, G. J., 86-93, 322
Anderson, A. W., 84-511
Anderson, B. M., 89-396
Anderson, D. C, 81-223; 83-154; 84-131,
185; 85-80; 89-4, 503
Anderson, E. L., 86-193
Anderson, J. K., 84-363
Anderson, K. E., 89-600
Anderson, M. C., 89-4
Anderson, T., 89-27
Andrews, J. S., Jr., 86-78
Angelo, J. R, 89-374
Antizzo, James, 87-515
Apgar, M. A, 84-176; 89-618
Applegate, Joseph, 87-273
Appier, D. A., 82-363
Arland, F. J., 83-175
Arlotta, S. V., Jr., 83-191
Arnold, D. F., 84-45
Arthur, J., 84-59
Asheom, David W., 87-315
Asoian, M. J., 86-152
Assink, J. W., 82-442; 84-576
Astle, A. D., 82-326
Atimtay, A., 85-464
Atwell, J. S., 83-352
Aulenbach, S. M., 89-146
Aurelius, Marcus W., 88-495
Averett, Daniel E., 88-338, 347
Ayi-es, J. E., 81-359
Ayubcha, A., 84-1
Babcock, Kevin B., 87-97
Badalamenti, S., 83-202, 358; 84-489; 87-
111
Baer, W. L., 84-6
Bagby, J. R., Jr., 86-78
Bailey, P. E., 82-464
Bailey, T. E., 82-428
Bailey, W. A., 83-449
Baker, Katherine H., 88490
Baker, Sara B., 87-264
Balfour, W. D., 82-334; 84-77
Ballif, J. D., 82414
Banerjee, Pinaki, 87-126
Barbara, M. A., 83-237; 83-310
Barber, J. A, 89443
Barboza, M. J., 86-152
Bareis, D. L., 83-280
Barich, J. T., 89-264
Barich, John J., 87-172, 87-198
Barkdoll, Michael P., 88-164
Barker, L. J., 82-183
Barkley, Naomi P., 82-146; 85-164; 88-
419
Barndt, J. T., 89-194, 618
Barnes, D. L., 89-91
Barnett, B. S., 89-635
Barone, J., 84-176
Barrett, K. W., 81-14
Barsotti, Deborah A., 88-537
Bartel, Thomas J., 88-287
Bartel, Tom, 88-125
Barth, D. S., 84-94
Earth, Edwin F., 86-224, 87-172
Bartley, R. W., 84-35
Bartolomeo, A. S., 82-156
Bascietto, J., 89-609
Bashor, M. M., 89-72
Bath, R. J., 8941
Baughman, K. J., 82-58
Baughman, W. A., 86-126
Baumwoll, D., 86-22
Bausano, James, 89-306
Baxter, T. A., 84-341
Bayer, Hans, 88-219
Bayse, D. D., 84-253
Beam, P., 86-84
Beam, P. M., 81-84; 83-71
Beck, W. W., Jr., 80-135; '82-94; 83-13
Becker, D. Scott, 88-323
Becker, J. C., 83442
Beckert, W. F., 8245
Beckett, M. J., 82431
Beekley, P., 86-97
Beers, R. H., 81-158
Begor, K. F., 89468
Beflke; P. J., 82424
Beling, Christine, 87-296
Bell, R. M., 82-183, 448; 84-588
Ben-Hur, D., 84-53
Bennett, Doug, 88-208
Benson, B. E., 80-91
Benson, J., 86-386
Benson, R C, 80-59; 81-84; 82-17;
83-71; 85-112; 86465
Berdine, Scott P., 88-582
Berg, Marlene G., 87-337
Berger, I. S., 82-23
Berk, E., 83-386
Berkowitz, J., 83-301
Berkowitz, Joan B., 87471
Bernard, H., 80-220; 86463
Bernert, J. T., 84-253
Berning, W., 86-386
Berzins, Nick, 88-158
Beukema, P., 89497
Bhalla, S., 85-189
Bhinge, Deepak, 88440
Biggs, Richard K., 87-376
Bigham, Gary, 87444
Bilello, L. J., 83-248
Billets, S., 8445
Bilyard, Gordon R., 88-323
Binder, S., 85409
Bird, K. J., 86-126
Bird, Kenneth J., 88-594
Bissett, F., 89-190
Bissex, Donald A., 86-208; 88429
Bisson, D. L, 89413
Bixler, Brint, 88-1
Bixler, D. B., 82-141; 84493
Blackmail, W. C., Jr., 80-91; 84-39;
86407
Blais, L., 86441
Blasko, Marcello J., 87-367
Blasland, W. V., Jr., 81-215; 83-123
Blayney, E. K. H., 85476
Blowers, Mark A., 88-287
Boa, J. A., Jr., 82-220
Bode, B. D., 89463
Bogue, R. W, 80-111
Bonazountas, M., 84-97
Bond, F. W., 82-118
Bond, Linda D., 88-125, 287
Bond, Rick, 87-197
Bonneau, W. F., 84-509
Boornazian, L. Y., 86-398
Bopp, F., Ill, 84-176
Borden, W. C., 89-582
AUTHOR INDEX 675
-------
Borgianini. Stephen A . 88-7; 89-562
Clay. P I . 81-45. 82-40; 83-100: 86-120
Clcary. Joseph G., 88-474
Clem, Arthur G.. 87-512
Clemens, B., 84-49, 335. 85-419, 86-445
Clemens. R, 84-213
demons. G P.. 84-104
Clme. Patricia V., 84-217; 88-108
Clmc. S. P., 89-277
Clinton, R. .1 86-4
Coatcs, A L., 86-365
Cobb. William [-.. 87-436
Cochran, S R., Jr., 80-233: 85-275
Cochran, S. R., 82-131; 84-498
Cockcroft, Beth I . 87-367, 87-4%
Cogliano, V J , 86-182
Cohen, S A.. 81-405
Coia, Michael V, 86-322; 88-363
Coldeway, W. G.. 84-584
Cole. C R., 81-306; 82-118
Cole. I liirold, 87-280
Collins G., 89-41
Collins.. J 1'., 81-2, 83-326
Collins, I 0., 83-398
Colonnn. R., 80-30
Conibcar, Shirley A.. 86-455. 87-532
Conncll. A., 89-267
Connolly, John P., 88-359
Connor. John A., 88-234
Connor, Michael S., 87-426
Cook, D. K., 81-63
Cook. I,. R., 83-280
Cooncy. J. A.. 89-647
Cooper, C 81-185
Cooper, D 85-419; 86-457
Cooper, E. W., 83-338
Cooper, J. W, 82-244
Cooper, L. M., 86-415
Cooper, lanrc R. 87-231
Coover, M. P., 89-331
Copeland, L. G.. 86-287
Corbctt, C. R-. 80-6; 81-5
Corbin, Michael H., 86-322, 87-380
Corbo, P.. 82-203
Cord-Dulhmh, Emily, 88-429
Com, M. R., 81-70
Cornaby, B. W.. 82-380
Cotu, S.. 89-130
Cothron, T K., 84-452
Cotton, Thomas A. 88-39
Coutrc, P. E.. 84-511
Cox. G, V.. 81-1
Cox. R. D., 82-58. 334
Crawford, R. B.. 86-272
Crawley. W. W., 84-131. 185; 8S4»
Cromwell. John E, 87-53
Crotbie, J. R.. 89-338
Crosby, T. W.. 86-258
Cudaby. J. J., 85-460
Cullinane, M. John, Jr., 84-465; 88-435;
89-222
Cunningham. John M., 87-337, 87-515
Cuny, JT-lr., 84-103
Curry, M F. R., 86-297
Curtis, M . 89-181
Czapor, J V.. 84-457
Dahl. T. 0.. 81-329
Daigler, J. 83-296
Dairy, P L., 85-383
Dalton, D. S., 85-21
Dalton, T. P.. 81-371
Danko. J P.. 89-479
Dapore. J U, 89-493
DarUek, G. T, 89-56
Davey, J. R., 80-257
Davidson, G M., 89-631
Davis, A. 0.. 86-115
Davis, Andy, 89-145
Davis, B. D, 84-213
Davis. L. R., 86-303
Davis, N O.. Jr., 89-15
Davis, S. L. 84-449
Davol, Phebe, 87-66
Dawson, G. W., 81-79; 82-386; 8W53:
86^173
Day. A- R., 83-140
Day. P. T.. 89-417
Day, S. R-. 86-264
Day. Stephen R., 88-462
De Pcrcui, Paul R., 88-508
de WalJe, F. P., 88-479
DeCaito, V. J., 85-29
Deck, N., 86-38
Decker, Jennifer A., 88-145
DeGrood. T. J . 85-231
Dehn, W. T.. 83-313
Deigan, G. J.,86-287
Del Re, S., 86-110
Dclfino, Joseph J.. 88-108
DeLuca, R. /, 86-148
Demarest. H F-, 86-143
Demeny, D D., 86-247
Demmy. R. H., 81-42
Dempsey, J G., 85-26
Denbo, R. T.. 86-56
Denfeld. D Colt, 88-202
Dent. Marc J., 87-223: 89-313
DeRosa, C, 85-412
Derrington, D., 84-382
Dcsmarais, A. M. C, 84-226
DeSmidt, Pamela D., 88-55
Desvousges, William H., 87-517
Dev. Harsh, 88498
Devary. J. l_. 83-117
Devinny, J. S.. 89-345
Dhamotharan, D. S.. 86-56
Dickens. Ward. 87-280
Dickinson. R. P.. 84-306
Dickinson, R. Wayne, 86-258, 87-371
Dickinson, W., 86-258
Dickinson, Wade, 87-371
DiDomenico, D., 82-295
Diecidue, A. M., 82-354; 83-386; 86-22;
R9-600
AUTHOR INDEX 676.
-------
Dienemann, E. A., 84-393
Diesl, W. F., 80-78
DiGiulio, Dominic C., 88-132
Dikinis, J. A., 84-170
DiLoreto, John, 88484
Dime, R. A., 83-301
DiNapoli, J. J., 82-150
Ding, Maynard G., 88-575
DiNitto, R. G., 82-111; 83-130
Dinkel, Mary E., 87459
DiPuccio, A., 82-311
Dirgo, J. A., 86-213
Diugosz, E. S., 85429
Dodge, Elizabeth E., 88-1
Dodge, L., 85-255
Dole, L. R., 89476
Donaloio, Brenda, 88-234
Donate, Michael J., 88-353
Donnelly, Kirby C., 87-66
Dorau, David, 87-251
Dorrler, R. C., 84-107
Dosani, Majid A., 88419
Dover, M. J., 89-609
Dowiak, M. J., 80-131; 82-187; 84-356
Downey, Douglas, 88498
Downie, Andrew R., 88-103
Doyle, D. F, 85-281
Doyle, R. C., 82-209
Doyle, T. J, 80-152
Dragun, J., 86453
Drake, B., 82-350
Drever, J. I., 84-162
Driscoll, K. H., 81-103
Droppo, James G., Jr., 87409, 87465,
88-539
Dryden, F. E., 89-558
Du Pont, A., 86-306
Duba, G., 89-190
Duff, B. M., 82-31
Duffala, D. S., 82-289; 88-65; 89-13
Duffee, R. A., 82-326
Duke, K. M., 82-380
Duncan, D., 81-21
Dunckel, J. R., 85468; 86-361
Dunford, Richard W., 87-517
DuPont, Andre, 88-398
Durst, C. M., 85-234
Duvel, W. A., 82-86
Dwight, D. M., 89-241
Dybevick, M. H., 83-248
Earp, R. F., 82-58
Eastman, K. W., 83-291
Eastwood, D., 86-370
Eberhardt, L. L., 84-85
Eckel, W. P., 8449; 85-130, 88-282; 89-
86
Ecker, Richard M., 87465
Edwards, D. K., 89-286
Edwards, J. S., 85-393
Edwards, R. C., 89-309
Edwards, Sally, 87-254
Ehrlich, A. M., 86-167
Ehrman, J., 84-374
Eicher, A. R., 85460
Eimutis, E. C., 81-123
Einerson, Julie H., 88-157
Eisenbeis, John J., 88-177
Eissler, A. W., 84-81
Eklund, B. M., 84-77
Eley, W. D., 84-341
Elkus, B., 82-366
Ellis, H. V., Ill, 86-213
Ellis, R. A., 82-340
Eltgroth, M. W., 83-293
Ely, John, 87-5
Emerson, L. R., 83-209
Emig, D. K., 82-128
Emmett, C. H., 86467
Emrich, G. H., 80-135; 86412
Enfield, C., 89-501
Eng, J., 84457
Engelbert, Bruce, 88-32
Engels, J. L., 8445
Engler, D. R., 85-378
English, C. J., 83-453; 84-283; 86-173
Englund, E. J., 86-217
Enneking, Patricia A., 88-521
Epperson, Charles R., 88-72
Erbaugh, M., 85-452
Erdogan, H., 85-189
Esmaili, Houshang, 88-245
Esposito, M. P., 84486; 85-387
Ess, T. H., 81-230, 82-390,408
Evangelista, Robert A., 88-424
Evans, G. B., Jr., 89-503
Evans, G. M., 89425
Evans, J. C, 82-175; 85-249,357,369; 8
440, 403; 89-292, 543
Evans, M. L., 84407
Evans, R. B., 82-17; 83-28
Evans, R. G., 86-78
Evans, T. T., 84-213
Everett, L. G., 82-100
Exner, P. J., 84-226
Fagliano, J. A., 84-213
Fahrenthold, P. D., 89-259
Fair, G. E., 89-558
Falcone, J. C., Jr., 82-237
Falk, C. D, 86-303
Fang, H-Y, 82-175; 85-369
Farrell, R. S., 83-140
Farro, A., 83413
Fassbender, Alex G., 87-183
Fast, D. M., 84-243
Faulds, C. R., 84-544
Feild, Robert W., 88-255
Feld, R. H., 83-68
Feldt, Lisa G., 87-1, 87-28
Fell, G. M., 83-383
Fellman, Robert T., 87492
Fellows, C. R., 83-37
Fenn, A. H., 85476
Fenstermacher, T. Edward, 87476
Ferenbaugh, R. W., 86-1
Fergus, R. Benson, 87-376
Ferguson, J., 84-248
Ferguson, T., 80-255
Fields, S., 84404
Figueroa, E. A., 81-313
Filardi, R. E., 89-137
Fine, R. J., 84-277
Finkbeiner, M. A., 85-116
Finkel, A. M., 81-341
Fischer, K. E., 80-91
Fisher, W. R., 86-124
Fisk, J. F., 85-130; 89-86
Fitzgerald, William M., 88-55
Fitzpatrick, V. F., 84-191; 86-325
Flathman, Paul E., 88446
Flatman, G. T., 85442; 86-132, 217
Fleming, E., 89-222
Fogg, Andrea, 88-292
Fontenot, Martin M., 87-348
Ford, K. L., 84-210, 230
Forney, D., 85409
Forrester, R., 81-326
Fortin, R. L., 82-280
Foss, Alan, 88455
Foster, Allan R., 87-78
Foster, R., 89407
Foster, Sarah A., 88-292; 89407, 547
Foth, D. J., 86-176
Fournier, L., 89-273
Fowler, Alan S., 88-335
Francingues, N. R., 82-220
Francingues, Norman R. Jr., 88-338
Franconeri, P., 81-89
Frank, J., 84-532;
Frank, J. F., 89-377
Frank, James F., 87459
Frank, U., 80-165; 81-96, 110
Fredericks, S., 86-36, 120
Fredericks, Scott C., 87-14
Freed, J. R-, 80-233
Freestone, F. J., 80-160, 208; 81-285
Freudenthal, H. G., 82-346
Friedman, P. H., 84-29, 49
Friedrich, W., 83-169
Fries, B., 89-606
Frost, John D., 87-72
Fuller, P. R., 86-313
Fullerton, Susan, 88-598
Fullerton, Tod H., 88409
Funderburk, R., 84-195
Furman, C, 82-131
Furst, G. A., 85-93
Gabanski, Gilbert, 87-89
Gabry,- Jon C., 87-104
Galbraith, R. M., 86-339
Galer, Linda D., 88-521
Gallagher, G. A., 80-85
Gallagher, John, 88-199
Galuzzi, P., 82-81
Gangadharan, A. C., 88-592
Garczynski, L., 84-521; 8640
Garlauskas, A. B., 83-63
Garnas, R. L., 84-39
Garrahan, K. G., 84478; 86-167
Gaskill, Bart, 87439
Gay, F. T., Ill, 82414
Gee, J. R., 89-207
Geil, M., 85-345
Geiselman, J. N., 83-266
Gemmill, D., 83-386; 84-371
Gensheimer, G. J., 84-306
Gentry, John K., 87-273
George, J. A., 86-186
George, L.C., 88413
Geraghty, J. J., 8049
Gerst, Donna Lee, 87-5, 87-21
Gervasio, R., 89-15
Geuder, D., 84-29
Ghassemi, M., 80-160
Gherini, Steven A, 87444
Ghuman, 0. S., 84-90
Gianti, S. J., 84-200; 89459
Gibbs, L. M., 83-392
Gibson, S. C., 81-269
Giggy, Christopher L., 87-174
Gigliello, K., 84457
Gilbert, J. M., 82-274
Gilbertson, M. A., 82-228
Gill, A., 84-131
Gillen, B. D., 82-27; 83-237
Gillespie, D. P., 80-125; 81-248
Gillis, Thomas, 8741
Gilrein, S. A., 86-158
Ginn, Thomas C., 88-323
Gish, B. D., 84-122
Givens, R. C., 86-31
Glaccum, R. A., 80-59; 81-84
Glass, J., 89-246, 501
Gleason, Patrick J., 88-125, 88-287
Glynn, W. K., 86-345
Godoy, F. E., 89-555
Goggin, B., 81411
Gold, J., 84416
Gold, Jeffrey W., 88-183
Gold, M. E., 81-387
Goldman, L. M., 84-277
Goldman, Norma J., 88-273
Goldman, R. K., 81-215
Goldstein, P., 83-313
Golian, S. C., 86-8
Golian, Steven C., 88-1
Goliber, P., 80-71
Golob, R. S., 81-341
Golojuch, S. T., 85423
Goltz, Mark N., 87-129
Goltz, R. D., 82-262; 83-202; 84489;
85-299
Goode, D. J., 83-161
Goodman, J., 85419
Goodwin, B. E., 85-7
Gorton, J. C., Jr., 81-10; 84435
Goss, L. B., 82-380
Gossett, N. W., 89-306
Granger, Thomas, 88474
Gratton, P. F., 89-13
Gray, E. K., 85406
Graybill, L., 83-275
Grayson, Linda, 88-79
Greber, J. S. 84486; 85-387
Greber, Jack S., 88419
Green, Ermon L., 88440
677
-------
Green, J, 81-223
Greenburg, John, 87-502
Greene, Joseph, 87-198
Greenlaw, P. D., 8941
Greenihal. John L., 88-60
Greiling, R. W , 84-535
Gridley, G. M,, 88-467
Griffen, C N , 85-53
Griswold. 1:. D., 89463
Grube, W. 1- . Jr 82-191, 249; 89-413
Gruenfeld, M , 80 165; 81-96; 82-36
Grumnger. R. M., 89455
Grupp, D. J., 89-41
Guerrero, P., 83-453
Gupta, Ciopal D., 88-592
Gurba, P., 84-210. 230
Gurka, D f , 82-45
Gushuc. J J., 81-V-), 85-261
Gushue, John J . 87-138
Gustafson. M I .. 86-448
Guthne, J . 86-386
Guttler. L . 89-537
Gutzmcr, Michael P . 88-72
Hadzi-Antich. 1 . 86-18
Hacbercr. A. I . 82-45
Haffcrty, Andrew J., 87-107
Hagarty, E P.. 89-455
Hagel. W. A.. 86-186
Hager. Donald G , 82-259. 87-174
Hagger, C . 8145; 84-321; 85-7
Hahn. S. J.. 86-448
Haiges, Usa, 87-311
Haighl, E. W , 89-652
Haiali, Pans, 87-238
Haji-Diafan, S.. 83-231
Hale. David W., 87-223
Hale, F D., 83-195
Halepaska. J C. 84-162
Haley, Jennifer L, 88-19; 89-246, 501
Hi!!; P C* *f3SV27
Hallahan. F M., 85-14
Haller, P. H., 86-469
Hammond, J. W., 80-250; 81-294
Hamper, M J., 89-122
Hana. S L. 89-*
Hanauska. Chris P., 87-480
Hanford. Rjchard W . 88-462
Hangcland. Erik B, 87-380
Hanlcy. G . 89-452
Hartley. M M., 82-111
Hannmk, G.. 88-479
Hansel. M. J, 83-253
Hanson, B., 82-141; 83-4; 86-224. 462
Hanson, Bill, 88-5; 89-501
Hanson, C. R.. 84-189, 85-349
Hanson, J. B, 81-198; 84-493
Hardy, Mark J . 87-179
Hardy, U Z., 80-91
Harman. H. D., Jr.. 82-97
Harmon, G. R.. 89-387
Harrington. W II., 80-107
Hams, D. J , 81-322
Harris, M R., 83-253
Hartsfield, B., 82-295
Hartz, Kenneth \ . 88-295
Hass, H., 83-169
Hatayama, H. K., 81-149; 84-363
Hatch, Norm N.. Jr., 85-285. 87-300
Hatheway, A. W, 85-331
Hatton, J, W., 89-298
Hawkins, C., 83-395
Hawkins, Elizabeth T, 87-166
Hawlcy, K. A., 85-432
Hay, G. H, 89-392
Hayes, Douglas, 87-439
Hayes, H., 85-285
Hazaga, D., 84-404
Hazelwood, Douglas, 88-484
Head, H. N., 86-258
Hcarc, S., 83-395
Hcbert, Richard L., 88-113
Hcdigcr, I, M., 86-164
Hecb, M., 81-7
Hcffcrnan, A. '/.., 86-8
Heffcrmm, Amelia, 87-515
Heglund. William. 87-5
Hem, James C. 88-174
Heinle, D., 89-130
Hclsing, Lysc D.. 87-171
llctmlcy. W I . 80-141
Henderson, D R., 86-380
Henderson, R B, 84-135
llcndry, C D., 85-314
Hcnnclly. Alyson A., 87-53
Hennington, J. C, 83-21; 85-374
Herrinton, Lisa, 88-19
Herson. Diane S . 8849Q
Hess, J W , 83-108
Heysc, I 85-234
Hijazi. N , 83-98
Milker, D, 80-212
Hill, II David, 87-7
Hill. J A.. 86-292; 89-122
Hill. R., 82-233
Hill. Ronald D , 80-173; 86-356. 87-25.
88-S16
Hillenbrand. I , 82-357. 461
Hma, C 1 83-63
Hines, J. M , 81-70;85-349
Hinrichs, R. 80-71
Hiiucl, I J , 86-313
Hirschhorn. Joel S., 85-311. 87-251
Hitchcock, S. 82-97; 86-318
Hjerstcd. N. B . 80-255
Ho. Mm-Da. 88-575
Hoag. R. B. Jr., 85-202
Hodge. V . 84-62. 498
Hoffman. R. 1 86-78
Hoffmasier. Gary. 87-326
Hokanson. Sarah. 87-502, 88-184
Holberger. R L. 82-451
Holland, J. Kent, Jr.. 87-520
Holm, L. A., 89-436
Holmes. R, I-. 84-592
Holmes, T 89-222
Holsiem. Ii C. 84-251
Homer. David H.. 86-213. 87-126
Hoogendoom, D., 84-569
Hooper, M. W., 83-266
Hopkins. 1 , 80-255
Home. A., 81-393
Horton, K. A., 81-158
Hosfeld, R, K., 86415
Hostage, Barbara, 88-37
Housman, J. J., Jr.. 81-398
Housman. J., 80-25
Houston. R. C. 80-224
Howar. Michael. 87-439
Howe, R. W., 82-340
Hoylman, E. W., 82-100
Hubbard. A E., 86-186
Hubbard, Robert J , 86-186, 87-326
Hubner, R. P., 89-41
Hudson, Charles M.. 87-158
Hudson, T. B., 89-198
Huffman, G. L. 84-207
Huggins, Andrew, 88-277
Hughey, R. E,, 85-58
Huizenga, II, 85412
Hullingcr, J P.. 85-136; 86-158
Hunt. G. I:.. 80-202
Hunl. R A.. 89-586
Hunter, J. II. 85-326
Hupp. W. II, 81-30
Hushon, J M., 89-99
Hutchison, C., 89-282
Hutson, K A., 86-8; 87-515; 88-565: 89-
5%
Mutton, Daniel I , 88-557
Hwang. J C, 81-317; 84-1
Hwang, Seong I 84-346, 87-149, 87-485
Hyman, Jennifer A., 88-193
laccarino, T., 84-66
lanniello, Michael L., 88-251
Icrardi, Mario, 87-204
Ikalainen, Allen J , 88-329
Ing, R,, 84-239
Ingersoll, T. G.. 81405
Ingham, A. T., 85-429
Isaacson, L., 81-158
Isaacson. P. J., 85-130
Ubister, J. D.. 82-209
Ukandar. I. JC, 84-386
Islander. R L., 89-345
Jackson, D. R., 89-413
Jacob, T A.. 89-86
Jacobs. J R, 82-165
Jacobson, P. R., 86-233
Jacot. H J., 83-76
James. S. C., 80-184; 81-171. 288- 82-70
131; 84-265; 85-234
Jams. J. R,, 81-405; 82-354
Jams/. A. J., 82-52
Jankauskas. J. A. 85-209
Janosik, Vic. 88-363
Jansen. David J.. 88-335
Janssen, James A., 87-453
Jarvis, C E. 84-469
Jensen, Stephen U, 87-101
Jerger. Douglas &, 88-446
Jerrick. N J., S3-389; 84-368
Jessberger. H. L, 85-345; 89-537
Jewell, III, J. J., 88-67; 89-1
Jhaven, V.. 83-242; 85-239
Job, Charles A., 87-89
Johnson-Ballard, J., 81-30
Johnson, D.. 84-544
Johnson, D. W., 86-227
Johnson, E. 89-41
Johnson. G. M., 86-93. 105
Johnson, K., 89-267
Johnson. Leonard C, 87-326
Johnson. M.. 89-186
Johnson, M. G.. 81-154
Johnson. Mark F., 86-52, 87-34, 88-23;
89-600. 606
Johnson, Thomas L., 88-226
Johnson, W. J., 86-227
Johnston. R, I-L, 83-145
Jones, A. K_, 82-183, 448
Jones, R. 84-300. 85-412, 419
Jones. K. M., 82-63
Jones. Philip L.. 87-18
Jones, R. D., 83-123, 346
Jones. S. G , 83-154
Jordan. B. H.. 82-354
Jowett. James R-, 84-339; 86-40; 87-14
Jurbach. R.. 84-66
Kabnck, R M., 89-331
Kaczmar. S. W., 84-221
Kadish. J., 82-458
Kaelin. J J., 85-362
Kaltreider, R, 86-14, 398
Kanehiro, B. Y , 89-259
Kaplan, M., 82-131
Karalby, Louis S., 86-436; 87-97
Karas, Paul, 87-355
Karmazinski, Paul L, 87-213
Karon, J. M.. 84-243
Kaschak, W. M., 82-124; 84-440; 85-281;
86-393
Kastury, S.. 85-189
Katz, 1, 85-419
Kavanaugh, Michael C, 88-287, 88-125
Kay, R L, Jr., 84-232
Kay, W., 85-409
Keane, J.. 89-318
Keffcr, W., 84-273
Keirn, M. A., 85-314
Keitz, E L., 82-214
Kelleher, Timothy E, 87-7
Kemerer, J. A., 84-427
Kcmplin, Manin G., 87-18
Kennedy, S. M., 81-248
Kenney, Patricia J., 88-429
Kerfoot, H. B., 84-45; 87^23
Kerfoot, W. B., 81-351
Kesari, Jaisimha, 87-380
Kester, Paul E, 87-457
Keulen, RW., 8M79
Khan, A. Q., 80-226
Khan, B. H., 86-220
Kiefer, Michael I,, 88-188
AUTHOR INDEX 678
-------
Kilpatrick, M. A., 80-30; 84478
Kim, C. S, 80 = 212
Kimball, C. S., 83-68
Kincare, K. A., 89-146
Kinesella, J. V., 89-325
King, J., 84-273; 85-243
King, Wendell C., 88-152
Kirkpatrick, G. L., 89-277
Kirner, Nancy P., 87-403
Kissel, John C., 88-142; 89-67
Klein, George, 87-111
Kling, Timothy L., 88-419
Klinger, G. S., 85-128
Knapp, Joan O'Neill, 88-429
Knowles, G. D., 83-346
Knox, J. N., 86-233; 89-186
Knox, R. C., 83-179
Knox, Robert, 87-311
Koch, D., 89-152
Koerner, Robert M., 80-119; 81-165
317; 82-12; 83-175; 84-158; 86-272; 87-
390
Koesters, E. W., 84-72
Kohn, Douglas W., 87-34
Kolsky, K., 84-300
Konz, James J., 87-143
Kopsick, D. A., 82-7
Kosin, Z., 85-221
Kosson, D. S., 83-217; 84-393, 88451
Roster, W. C., 80-141
Koutsandreas, J. D., 83449
Kovell, S. P., 86-46
Kraus, D. L., 85-314
Krauss, E. V., 86-138
Krohn, Russell B., 87-306
Kuersteiner, J. D. Boone, 88-287
Kufs, Charles T., 80-30; 82-146; 86-110;
87-120
Kugelman, I. J., 85-369
Kumar, Ashok, 87-525
Kunce, E. P., 86-345
Kunze, M. E., 89-207
Kuykendall, R. G., 83-459
LaBar, D., 85449
LaBrecque, D., 83-28
Lacy, Gregory D., 88429
Lacy, W. J., 84-592
LaFaire, M. A. C, 89447
LaFornara, J. P., 81-110, 294; 85-128
LaGrega, Michael D., 88403, 88-277,
8142
Laine, D. L., 89-35, 56
LaMarre, B. L., 82-291
Lamb, Robert H., 88-67
Lambert, W. P., 84412
Lamont, A., 84-16
LaMori, Philip N., 87-396
Lampkins, M. J., 86-318
Landreth, Lloyd W., 88-605; 89-613
Lang, David J., 88-19
Lange, J. H., 89-78
Lange, R. M., 89-377
Langley, William D., 88-282
Langner, G., 82-141
Lamer, John H., 88-587
Lappala, E. G., 84-20
Larimore, D. R., 89-91
Larson, R. J., 80-180
Laskowski, Stanley L., 88-317
Laswell, B. H., 85-136
Lataille, M., 82-57
Laudon, Leslie S., 88-261
Lavinder, S. R, 85-291
Lawrence, L. T., 84481
Lawson, Frank D., 88-103
Lawson, J. T., 82474
Leap, D.R., 87405
LeClare, P. C., 83-398
Lederman, P. B., 80-250; 81-294
Lee, C. C., 82-214; 84-207
Lee, Charles R,, 88435
Lee, G. W., Jr., 83-123, 346
Lee, R. D., 85-157
LeGros, Susan P., 88-277
Leighty, D. A., 83-79
Leis, W. M., 80-116
Lemmon, A. W., 89-380
Lennon, G. P., 85-357
Leo, J., 82-268
Lepic, Kenneth A., 87-78
Leu, D. J., 86-303
Lewis, D. S., 84-382
Lewis, N., 89407
Lewis, Ronald A., 88-113
Lewis, W. E., 84427
Lia, Paula M., 87-72
Librizzi, William, 88-77
Lichtveld, Maureen, 88-524
Lidberg, R., 86-370
Liddle, J. A., 84-243
Lieber, Marc P., 87-72
Liedle, J. M., 89-582
Lincoln, D. R., 85449
Lincoln, David R., 88-259
Lindsey, W. B., 89-137
Linkenheil, R., 85-323
Linkenheil, Ronald J., 87-193, 87-533
Lippe, J. C., 83423
Lippitt, J. M., 82-311; 83-376
Lipsky, D., 82-81
Livolski, J. A., Jr., 84-213
Lo, T. Y. Richard, 83-135; 87-228
Locke, P. W., 89-95
Lockerd, M. Joseph, 88-93
Loehr, Raymond, 533
Lombard, R. A., 85-50
Lominac, J. K., 89-309
Longo, Thomas P., 88-39
Longstreth, J., 85412
Lord, Arthur E., Jr., 80-119; 81-165-
82-12; 83-175; 84-158; 86-272; 87-390
Losche, R., 81-96
Lough, C. J., 82-228
Lounsbury, J., 84498; 86457
Loven, Carl G., 82-259; 87-174
Lovett, John T., 88-202
Lowe, G. W., 84-560
Lowrance, S. K., 83-1
Lucas, R. A., 82-187
Lucia, S. M., 89-298
Lueckel, E. B., 83-326
Lundy, D. A., 82-136
Lunney, P., 82-70
Lupo, M. J., 89-570
Lybarger, J. A., 86467
Lynch, D. R., 84-386
Lynch, E. R., 81-215
Lynch, J. W., 8042; 85-323
Lysyj, 1., 81-114; 83446
MacDonald, James R., 87-306
Mack, J., 84-107
MacPhee, C., 89-289
MacRoberts, P. B., 82-289
Madison, M. T., 89-95
Magee, A. D., 85-209
Magee, Brian, 87-166
Mahan, J. S., 82-136
Mahannah, Janet L., 88-152
Maher, Thomas F., 87-296
Makris, J., 86-11
Malhotra, C. C. J., 89455
Malone, P. G., 80-180; 82-220
Maloney, S. W., 85456
Malot, James J., 87-273
Mandel, R. M., 80-21
Mandel, Robert M., 88424
Manderino, L. A., 89-600
Mangan, Chuck, 88-598
Manko, J. M., 81-387
Mann, M. J., 85-374
Mansoor, Yardena, 8741
Manuel, E. M., 85-249
Margolis, S., 85403
Mark, D. L., 89436
Markey, Patricia,. 87-300
Marlowe, Christopher S. E., 88-546
Marlowe, Christopher S. E., 88-567
Marquardt, George D., 87-284
Marsh, Deborah T., 88-251
Marshall, T. C., 84-261
Marshall, T. R., 89-345
Marszalkowski, Robert A., 88-219
Martin, J. D., 89-251, 512
Martin, W. J., 82-198; 86-277
Martin, W. F., 83-322; 84-248
Martyn, S., 89430
Martz, M. K., 86-1
Maser, K. R., 85-362
Mashni, C. L, 86-237
Maslansky, S. P., 82-319
Maslia, M. L., 83-145
Mason, B. J., 84-94
Mason, R., 86-52
Mason, Robert J., 88-23
Mason, Robert J., 84-339; 87-34; 87-520
Massey, T. I., 80-250
Mateo, J., 86-14
Mateo, M., 83413
Matey, Janet, 88-598
Mathamel, Martin S., 81-280; 86472;
87-162; 88-546, 557, 567
Matson, C., 89-273
Mattejat, P., 89-152
Mattern, Charles, 87-268
Matthews, R. T., 83-362
Mauch, S. C., 89-157
Maughan, A. D., 84-239
Mavraganis, P. J., 83449
May, I., 89-152
Mays, M. K., 89-298
Maziarz, Thomas P., 88-395
Mazzacca, A. J., 83-242; 85-239
McAneny, C. C., 85-331
McArdle, J., 84486
McAvoy, David R., 88-142
McBride, R. E., 89-348
McCartney, G. J., 89-392
McCloskey, M. H., 82-372
McClure, A. F., 84452
McCord, A. T., 81-129
McCracken, W. E., 86-380
McDevitt, Nancy P., 87453
McDonald, Ann M., 88-145
McDonald, S., 89-190
McEnery, C. L., 82-306
McGarry, F. J., 82-291
McGinnis, J. T., 82-380
McGinnis, Roger N., 87-107
McGlew, P. J., 84-150; 85-142; 86403
McGovern, D., 84469
McGowan, T. F., 89-387
McGrath, Richard A., 87420, 87426
McKee, C. R., 84-162
McKnight, Robert, 87-111
McKown, G. L., 81-300, 306; 84-283
McLaughlin, D. B., 80-66
McLaughlin, Michael W., 87-296
Mclelwain, T. A., 89497
McLeod, D. S., 84-350
McLeod, R. S., 84-114
McMillan, K. S., 85-269
McMillion, L. G, 82-100
McNeil), J. D., 82-1
Me'ade, J. P., 84407
Meegoda, Namunu J., 87-385
Mehdiratta, G. R,, 89-512
Mehran, M., 83-94
Meier, E. P., 8245
Meier, Marina P., 88413
Melchior, Daniel C., 87-502
Melvold, R. W., 81-269
Menke, J. L., 80-147
Menzie, Charles A., 87-138
Mercer, J. W., 82-159
Mercer, Mark L., 87-143
Merkhofer, Miley W., 88-39
Merkhofer, Miley W., 8844
Mernitz, S., 85-107
Messick, J. V., 81-263
Messinger, D. J., 86-110
Meyer, J., 80-275
Meyers, T. E., 80-180
Michaud, G. R., 89-377
Michelsen, D. L., 84-398; 85-291
Michelsen, Donald L., 88455
Miklas, M. P., 89-35
AUTHOR INDEX 679
-------
Milbrath, L, W , 81-41S
Militana, L M., 86-152; 89-157
Miller. D O., Jr., 82-107; 83-221
Miller, K. R., 85-136; 86-158
Miller, Keith R., 88-103
Miller. M. A., 89-168
Millison, Dan, 88-269
Mills, W, J., 89-497
Mills. William B., 87^44
Millspaugh, Mark P., 88-60
Mindock, R. A , 86-105
Mineo, 1 O.. 89-286
Mischgofsky. F. I!., 88479
Mitchell. F. L.. 84-259; 85-406
Mittleman, A. 1.. 84-213
Moaycr, Masoud, 88-245
Molton, Peter M.. 87-183
Monsccs, M., 85-88
Monserratc, M , 86-14
Montgomery, R. J.. 86-292
Montgomery, V. J.. 83-8
Montgomery. Vema. 88-32
Moon. R. I- . 89-137
Mooncy. Ci A . 84-35
Moore, James B.. 87-27
Moore, S. F. 80-66
Morahan, T .1 . 83-310
Moran. B V 83 17
Morey. R. M . 81-158
Morgan. C H., 80-202
Morgan. R. C., 82-366; 84-213; 85-396
Morgensiem, Karl A . 88-84
Monn. J 0., 85-97
Momingstar, Maiy P.. 87-»71
Morson. B J . 84-535
Monensen. B K.. 86-74
Monoo. E. S., 86-213
Moscaii. A. F.. Jr., 86-164, 420
Moslehi. J., 85-326
Mote. Peter A.. 87-371
Mott, H V 89-526
Mott. R. M . 80-269; 83-433
Morwani. J N . 86-105
Mousa. J. J 83-86
Moy. C. S. 89-19
Moycr. I, II, 85-209
Moytan. C A. 85-71
Mueller. Susan I 88-528
Mullcr. B. W.. 82-268
Mullcr-Kirchcnbaucr. If. 83-169
MuUins, J W., 8S-J42
Mundy. P A.. 89-609
Mungcr. Robert 87-453
Mungin-Davii. Ouecoic, 88-208
Munoz. H., 84-U6
Murphy. Bnan 1.., 82-331. 396; 83 H
87-138; 87-153
Murphy. C B . Jr , 83-195. 84-221
Murphy. J. R . 84-213
Murphy, J , 89-152
Murphy, Vincent P., 87-390
Murray, J G.. 85-464
Musser. D. I , 85-231
Mutch, R. D . Jr., 83-2%; 89-562
Myers, F. 89-267
Myers. R. S , 89-459
Myers, V. B„ 82-295; 83-354
Mynck. J.. 84-253
Nadeau, P F, 82-124; 83-313
Nadeau, Paul F., 88-15
Nadeau, R. J., 85-128
Naglc, I . 83-370
Nafcid, D S.. 89-555
Narang. R., 80-212
Nauglc, D. P, 85-2f.
Na/ar. A , 82-187; 84-356
Nccdham, L. I... 84-153; 86-78
Neely, James M.. 88-561
Necly, N. S, 80-125
Neithcrcut, Peicr D., Kl-W)
Nelson, A. B., 81-52
Nelson. D D., 85-32
Nelson, Jerome S , 87-371
Neumann, C.. 82-350
Newman. J. R., 84-150
Ncwion, C. I 86-420
Newton, Jeffrey P . 87-187
Nichols, 1 D, 84-504
Sickens. Dan. 84-116; 87-268
\icKon. 1) M.. 86-460
Niclion. M 81-374
Niemelc, V. F. 82-437
Nilunanesh. J . 89 190
Nimmon*. M. J., 83-94
Nishcl. I C. T. 82-406
Noel, M R.. R3-7I
Noel, T M . 83-266
Noland. John W. 84-176. 203; 87-453
Norman, M.. 86-318
Norman, Michael A.. 88-313
Norman, W R. 82-111. 85-261
North. B E., 81-103
Nowell. Craig A., 87-179
Nunno, Thoma* J.. 88-193
Nybcrg, P. C., 84-504
Nygaard, D D.. 83-79
O'Connor. Ralph ( Jr.. 88-53''
O'Dea. D. 83-331
O'FUhcriy. P. M 84-535
O'Hara. Patrick K. 88-594. 86-126; 87-
367. 87-J96, 87-499
O'Kccfc, P. 80-212
O'Malley. R . 85-58
CTNcil. L. Jean. 88-435
O'Reilly. Kathlenc. 87-87-355
OToolc, M M . 85-116
Obascki. S.. 84-598
OfTutl. Carolyn K. 88-J29
Ofluli. Carolyn K.. 88-12
Ogg, R. N 83 202. 358; 86-356
Ohonba. I . 84-598
Oi. A W 81-122
Okcke, A. C. 8S-1.S?.
Oldenburg, Kmien L , 87-251
Oldham. 1, 89-306
Olsen, R. 1.. 85-107. 86-115, 313. 386.
88-393, 261; 89-145
Olson. Kathlenc A.. 87-480
Oma. K- If, 84-191
Opcnshaw, I.-A. 83-326
Opu/ B 1 . 82-198; 86-277
Oraveiz, Andrew W.. Jr.. 88-129
Orr, J R., 85-349
Ortiz, M , 86-8-1
Osborn. J , 83-43
Osheka, J. W.. 80-184
Osier. J G., 86-138
Otis, Mark J., 88-347
Ottinelli, Luca, 87-476
Ounaman, D. W., 83-270
Owens, D W., 80-212
Owens, Victor, 87-228
Owens, William, 87-300
Owens. William W, 88-164
Ozbilgcn, Mchh M., 88-287
Ozbilgjn, Melih M., 88-125
Paczkowski. Michael 1 . 88-375
Page. Norbcrt P., 87-132
Page, R. A , 84-594
Paige. S F, 80-VJ, 202
Paine, U . 89-586
I'ajak. A. I'. 80-IS-l. 81-288
Palombo. I) A. 82-165
Pancoski. S , 89-292
1'ancoski. Stephen 1 88-403
Puncoski. Stephen I . 88-440
I'ankanin. J , 89-216
Paquclic. J Steven, 86-208, 393; 87-1
Parker. Frank 1... 81-313. 87-231
Parker. Frank I,. 88-119
Parker, J C , 84-213
Parker, W. R.. 84-72
Parks, (. A., 83-280
Pan-all, R. S . 83-195
I'arns, George li., 88-602
I'arnsh, C. S., 85-1
Parry, O. 1) R., 82-448. 84-588
Partridge, I.. J, 84-290; 85-319; 86-65
Partymillcr, K. G., 84-213; 89-413
Paschal, O.. 85-409
Paschke, R. A., 85-147
Pastor, S., 89-635
Patarcity, Jane M., 87-326
Patchin. P^ 89-267
Patel, M. A,, 89-455
Pateluna*. G. M., 89-78
Painode, Thomas L, 85-323; 87-193
Patrick, Cynthia D., 87-158
Patterson, D. G., Jr., 86-78
Paulson, Steven E. 88-413
Pearce, R. B.. 81-255; 83-320
Pcar&all, L. J., 86-242. 89-552
Pease, R. W., Jr.. 80-147; 81-171. 198
Pedersen, T A., 86-398
Pedersen, Tom A., 88-199
Pei. Phyllis C, 88-157
Pcnnington, D.. 85-253
Perkins. L. C. 89-137
Peril*. Randy. 88-97
Peters. J A., 81-123
Peters. N . II. 86-365
Peters. W R_. 82-31
Peterson. B.. 89-50
Peterson, J. M-, 85-199
Peterson, Sandy. 87-45
Pheiffer, Thomas, 88-193
Phelps. Donald K., 88-335
Philljps. C. R.. 89-198
Phillips, J. W.. 81-206
Pickelt. J S. 86-424
Pierson. T., 84-176; 89-152
Pike. Myron T.. 87-480
Piroentel. E. M-, 88-35, 241; 89-417
Pintcnich, J. U, 81-70
Phtmk, Marilyn A., 87-414
Plourd, K. P., 85-396
Plumb. R R, 84-45
Ponder. T C, 85-387
Popp. S. A-. 86-105
Porter. Don C, 87-436
Portier. R. J., 89-351
Possidento. M., 80-25
Possm. B. N-, 83-114
Potter. Thomas, 88-108
Powell, D H., 83-86
Prwm, R. S.. 89-111
Prater, R B., 89-91
Predpall, D. F., 84-16
Preston, J E, 84-39
Preuss, P W.. 86-167
Price. D E.. 84-478
Price, D R-, 82-94
Prickctt. T.. 89-152
Priznar. I J., 85-1. 74; 8634
Proko. K.. 85-11
Prothero, T G., 84-248
Prybyta, D. A., 85-468
Puglionesi, Peter S., 87-380
Pyles, D. G, 86-350
Quan. W., 81-380
Quimby. J. M., 82-36
Quinlivan. S.. 80-160
Quinn, K. J., 84-170; 85-157
Qumn, R D.. 86-393
Quintrcll. W. N., 85-36
Radcmacher, J. M., 84-189; 85-349
Rams, J M., 81-21
Ramsey, W. L. 80-259; 81-212
Ranney, Colleen A.. 88-103
Ransom. M.. 80-275
Rappapon, A., 81-411
Ratnaweera, Prasanna, 87-385
Raymond, Arthur, 88-403
Rea. K. H., 86-1
Rcbis, E N., 83-209
Redeker, Laurie A.. 87-21
Rcdford, David, 87-465
Reeme, T. L., 89-638
Reirsnyder, R. H.. 82-237
Rciter, G. A., 80-21
Remeia, D. P., 80-165; 81-96
Repa, H. W., 82-146; 85-164; 86-237
AUTHOR INDEX 680
-------
Reverand, J. M., 84-162
Reyes, J. J., 89-72
Rhoades, Sara E., 87-358
Riccio, R., 89-41
Rice, Craig W., 87-63
Rice, E. D., 85-84
Rice, J. M., 85-182
Rice, R. G., 84-600
Richards, A., 80-212
Richardson, S., 84-1
Richardson, W. S., 89-198
Richardson, W. K., Jr., 89-277
Richey, Maxine, 88-269
Rick, J., 84-469
Ridosh, M. H., 84-427; 85-243
Rikleen, L. S., 82-470; 85-275
Riley, John, 88-37
Riner, S. D., 82-228
Ring, George T., 87-320
Riojas, Arturo, 88-382
Rishel, H. L., 81-248
Ritthaler, W. E., 82-254
Rizzo, J., 82-17
Rizzo, W. J., Jr., 85-209
Robbins, J. G, 83431
Roberts, Andrew W., 88-313
Roberts, B. R., 83-135
Roberts, D. W., 86-78
Roberts, Paul V., 87-129
Robertson, J. Martin, 88-435
Rockas, E., 85-11
Rodenbeck, Sven E., 88-532
Rodricks, J. V., 83401
Roe, C, 89-246
Rogers, John, 88-503
Rogers, W., 84-16
Rogoshewski, P. J., 80-202; 82-131, 146;
84-62
Romanow, S., 85-255
Ronk, R. M., 86-471
Rood, A. S., 89-117
Roos, K. S., 83-285
Rosasco, P. V., 84-103
Rosbury, K. D., 84-265
Rosenberg, M. S., 89-202
Rosebrook, D. D., 84-326
Rosenkranz, W., 81-7
Rosenthal, Seymour, 88-513
Ross, Derek, 84-239; 87-315
Ross, Derek, 88-395
Ross, W. O., 89-592
Rothman, D. W., 84-435
Rothman, T., 82-363
Roy, A. J., 83-209
Roy, Mell J.-Branch, 87-48
Royer, M. D., 81-269
Rubenstein, P. L., 86-143
Ruda, F. D., 84-393
Rudy, Richard J., 88-219; 89-163
Ruikens, W. H., 82-442; 84-576
Rupp, G., 89-216
Rupp, M. J., 86-164
Ruta, Gwen S., 87-508
Ryan, C. R., 86-264
Ryan, Elizabeth A., 87-166
Ryan, Elizabeth A., 88-353
Ryan, F. B., 81-10
Rvan, John, 87-533
Ryan, M. J., 85-29
Ryan, R. M., 85-125
Ryckman, M. D., 84-420
Sabadell, Gabriel P., 88-177
Sachdev, Dev R., 87-341
Sackman, Annette R., 88-97
Sadat, M. M., 83-301, 413
Sale, Thomas C., 87-320, 87-358
Salisbury, Cynthia, 88-214
Salvesen, R. H., 84-11
Sanders, D. E., 82-461
Sanders, Thomas M., 87-218
Sandness, G. A., 81-300; 83-68
Sandrin, J. A., 89-348
Sandza, W. F., 85-255
Sanford, J. A., 84-435
Sanning, D. E., 81-201; 82-118, 386
Santos, Susan L., 87-166, 87-254
Santos, Susan L., 88-353
Sappington, D., 85-452
Saracina, Rocco, 88-214
Sarno, D. J., 85-234
Sarno, Douglas J., 88-255
Saunders, M. F., 89-111
Sawyer, Stephen, 88-504
Sawyer, Stephen, 88-508
Schafer, P. E., 85-192
Schalla, R., 83-117; 84-283
Schaper, L. T., 86-47
Schapker, D. R., 8647
Schauf, F. J., 80-125
Scheinfeld, Raymond A., 88-363
Scheppers, D. L., 84-544
Schilling, R., 84-239
Schleck, D. S., 89-642
Schlossnagle, G. W., 83-5, 304
Schmidt, C. E., 82-334; 83-293
Schmierer, Kurt E., 88-226
Schnabel, G. A., 80-107
Schneider, P., 80-282
Schneider, R., 80-71
Schnobrich, D. M., 85-147
Schoenberger, R. J., 82-156
Schofield, W. R., 84-382
Scholze, R. J., Jr., 85-456
Schomaker, N. B., 80-173; 82-233
Schuller, R. M., 82-94
Schultz, D. W., 82-244
Schultz, H. Lee, 87-143, 87-149
Schweitzer, G. E., 81-238; 82-399
Schweizer, J. W., 86-339
Scofield, P. A., 83-285
Scott, J. C., 81-255; 83-320
Scott, K. John, 88-335
Scott, M., 82-311; 83-376
Scrudato, R. J., 80-71
Sczurko, Joseph J., 88-113
Sczurko, Joseph J., 88-413
Seanor, A. M., 81-143
Sebastian, C, 86-14
Sebba, F., 84-398
Segal, H. L., 85-50
Selig, E. I., 82458; 83437
Sepesi, J. A., 85423, 438
Sergeant, Ann, 87431
Sevee, J. E., 82-280
Sewell, G. H., 82-76
Seymour, R. A., 82-107
Shafer, R., 89-519
Shafer, R. A., 84465
Shah, Ramesh J., 87414
Shannon, Sanuel, 87-300
Shapiro, Melissa F., 88-269; 89452
Shapot, R. M., 86-93
Sharkey, M. E., 84-525
Sharma, G. K., 81-185
Sharrow, D., 89-606
Shaw, E. A., 86-224
Shaw, Elizabeth A., 88-5
Shaw, L. G., 81415
Sheedy, K. A., 80-116
Shen, Thomas T., 82-70, 76; 84-68; 87-
471
Sheridan, D. B., 84-374
Sherman, Alan, 88-592
Sherman, J. S., 82-372
Sherman, Susan, 87-280
Sherwood, D. R., 82-198; 86-277
Shields, W., 89-130
Shih, C. S., 81-230; 82-390, 408; 83405;
89-012
Shih, Shia-Shun, 88-382
Shimmin, K. G., 86-143, 463
Shiver, R. L., 85-80
Shoor, S. K., 864
Shroads, A. L., 83-86
Shuckrow, A. J., 80-184; 81-288
Shugart, S. L., 86436
Shultz, D. W., 82-31
Sibold, L. P., 85-74
Sibold, Lucy, 87-14
Siebenberg, S., 84-546
Sigler, W. B., 89-9
Sikora, L., 89-298
Silbermann, P. T., 80-192
Silcox, M. F., 83-8
Silka, L. R., 8045; 82-159
Silka, Lyle R., 88-138
Sills, M. A., 80-192
Simanonok, S. H., 86-97
Simcoe, B., 81-21
Simmons, M. A., 84-85
Sims, L. M., 89-582
Sims, R. C., 83-226
Singer, G. L., 84-378
Singerman, Joel A., 87-341
Singh, J., 84-81
Singh, R., 83-147
Sirota, E. B., 83-94
Siscanaw, R., 82-57
Sisk, W. E., 84-203, 412
Skach, Robert F., 88-188
Skalski, J. R., 84-85
Skladany, George J., 87-208
Skoglund, T. W., 85-147
Slack, J., 80-212
Sladek, Susan J., 88-5
Slater, C. S., 82-203
Sloan, A., Ill, 85438
Sloan, Richard L., 88-273
Slocumb, R. C., 86-247
Smart, David A., 88-67
Smart, R. F., 84-509
Smiley, D., 84-66
Smith, C., 84-546
Smith, Craig W., 88-188
Smith, E. T., 80-8
Smith, J. R., 89-331
Smith, J. S., 84-53
Smith, Jeffrey W., 88455
Smith, John J., 87492
Smith, John, 88-214
Smith, Lee A., 85-396; 87-158; 88-208
Smith, M. 0., 86430
Smith, Michael A., 82431; 84-549; 87-
264
Smith, P., 86-313
Smith, Philip G., 87-101
Smith, R., 80-212
Smith, R. L., 85-231
Smith, S., 86462
Smith, Stephen M., 88-304
Smith, W., 86-333
Smith, William C, 88-594
Smith, William C, 87-367, 87496
Snow, M., 85-67
Snyder, A. J., 81-359
Snyder, M., 80-255
Sokal. D.. 84-239
Solyom, P., 83-342
Sosebee, J. B., 84-350
Sovinee, B., 85-58
Spatarella, J. J., 84440
Spear, R. D., 8941
Spear, R., 81-89
Spencer, R. W, 82-237
Spittler, T. M., 81-122; 8240, 57;
83-100, 105; 85-93
Spooner, P. A., 80-30, 202; 82-191;
85-214, 234
Springer, C., 82-70
Springer, S. D., 86-350
Sresty, Guggilam C., 88498
Srivastava, V. K., 83-231
St. Clair, A. E., 82-372
St. John, John P., 88-359
Stadler, Gerald J., 87-7
Staible, T., 85-107
Staley, L. J., 89421
Stamatov, J. R,, 89443
Stammler, M., 83-68
Stanfill, D. F., HI, 85-269
Stanford, R, L., 81-198; 84498; 85-275
Stankunas, A. R., 82-326
Stanley, E. G., 83-1
Starr, R. C., 80-53
Steelier, Eugene F., 87-334
Stecik, Robert E., Jr., 87-28
Steele, J. R., 84-269
AUTHOR INDEX 681
-------
Stcelman, B. E, 85-432
Stehr-Green, P. A., 86-78
Stehr, P. S., 84-287
Steimle, R. R, 81-212
Stein, G. F., 84-287
Steinberg, K, K., 84-253
Steinhauer, William G , 87-420, 87-42(,
Stephens, R. D . 80-15; 82^28; 85-102
Sterling, Sheny, 87-61
Stet2, Elizabeth, 88-269
Steward, K., 89-430
Slief, K., 82^134; 84-565
Stinson, Mary K., 88-504
Stilts, Hugh M., 88-300
Stockmger. Siegfried L., 87-420
Stoclunger, Siegfried L.. 88-343
Stokely, P. M.. 84-6
Sloller. P. J., 80-239; 81-198
Stoloff, S. W , 89-443
Stone, K. J L, 89-537
Stone, Manlyn E, 88-fi
Stone. T, 85-128
Stone, W. L, 81-188
Stoner, R, 84-66
Strandbergh, D , 84-81
Strattan, L. W., 81-103
Strauss. J B.. 81-136
Strenge, Dennis L.. 88-539
Sirenge. Dennis L.. 85-432; 87-109
Strickfaden, M. E. 85-7
Strobel, G., 89-163
Strong. T M., 85-473
Stroo, H. F... 89-331
Slroud. F B., 82-274
Struttraann, T., 89-27
Struzziery, J. J., 80-192
Suffci. Irwin H , 88-132
Sukol. Roxannc B.. 88-419
Sullivan. D. A., 81-136
Sullivan, J. M Jr.. 84-386
Sullivan. J H,, 83-37
Sullivan. Jeffrey A,. 88-274
Sullivan. Kevin M.. 87-208
Sunada. Daniel K., 88-177
Sutdj. R. W., 89-468
Sunon, C, 89-41
Sutton, P. M,. 86-253
Swaroop, A., 84-90
Swaroop. Ram. 87-258
Swatek. M A.. 8S-7.SS
Swenson. G A., IJJ, 83-123
Swibas, C M., 84-39
Swichkow, D.. 89-592
Sydow, Wendy L. 86-393, 398; 87-1
Syvcrson. Timothy L, 88-84
Tackett, K. M., 81-123
Tafuri. A. N.. 81-188; 82-169; 84-407;
89-202
Tanaka, John C, 87-330
Tanzer, M. S, 81-10
Tapscott. G., 82-420
Tarlton, S I , 84-445
Tarlton. Sieve, 87-355
Tasca, J J.. 89-111
Tate. C: L. Jr.. 84-232
Taylor, Alison C.. 87-153; 89-108
Taylor, B. 83-304
Taylor, Larry R., 88-158
Taylor, M D., 86-88
Taylor, Michael L, 88-419
Tecpen, Krislina L, 88-274
Teets, R. W., 83-310
Teller, J., 84-517
Testa, Stephen M., 88-375
Tetta, D, 89-130, 259, 301
Tewhey, J D , 82-280; 84-452
Thibodcaux, L J,, 82-70
Thiesen, H M., 82-285
Thorn, J. E., 89-479
Thomas, A., 84-176
Thomas, C. M., 85-112
Thomas, G. A., 80-226
Thomas, J. E., Jr., 84-150; 85-142
Thomas, J. M., 84-85
Thomas, S. R., 85-476
Thompson, <• M., 84-20
Thompson, S N., 83-331
Thompson. W F- 84-469; 85-387
Thomsen, K 0., 86-138. 220
Thornc, D. J., 89-117
Thorecn, J. W., 81-42, 259; 82-156
Thorslund, T W, 86-193
Threlfall. D.. 80-131; 82-187
Tifft, E C. Jr.. 84-221
Tillinghast, V., 85-93
Timmcrman. C, I... 84-191; 89-309
Tinto. I 85-243
Titus. S II. 81 177
long. Pcicr. 87-141
Tope, Timothy J.. 88-119
Topudurti, 1C. 89-407
I'orpy. M F. 89-331
Towarnicky, J . 89-380
Towers, D S, 89-313
Town-scnd. R. W 82-67
Travcr, R. P.. 89-202
Travis. Daniel S, 88-119
Trees. I) P 84-49
Trcmblay. J W . 83-423
Trezek, G, J 86-303
Tricgel, 1 K., 83-270
Trojan, M., 89-503
Tro«l:r. W. L, 85-460
Irueu, J. B., 82-451
Truitt, Duane, 87-449
Tucker. W. A., 84-306
Tuor. N. R., 83-389, 84-368
Turkeltaub. Robert H . 88-569
Turner. J. R. 83-17
Turnham, B., 85-423
TurofT. B.. 80-282
Turpin R D.. 81-110. 277; 83-82;
rn
Tusa, W K., 81 2. 82-27
Iwedell. A. M . 80-233
Twedell. D. B, 80-30, 202
Tvagi. S.. 82-12
Tvburski, T E. 85-3%
Ulirsch. Gregory V.. 88-532; 89-72
Unger. M., 89-503
Unites. D. P. 80-25; 81-398; 83-13
Umcrberg, W., 81-188
Urban. M J . 84-53
Urban, N. W., 82-414; 83-5, 304
Vais. C. 84427
Valcntmcm, Richard A., 88-77; 89-404
Van Amam, David G., 87-223
van dc Velde. J L.. 88-479
van der Mcer. J P., 8&4T9
Van lie. J. J . 83-28
van Hpp, T. D , 86-361
Van Gemert, W J. Th.. 82-442
van Munster, Joan, 87-330
Van Slyke, D.. 83-442
Van Tascel. Richard, 87-396
VanAmam, D. G.. 89-313
Vandcrlaan, G. A., 81-348; 82-321;
83-366; 86-107
VanderVoori. J D , 86-269
Vandervort. R., 81-263
Vasudevan, C, 89-623
Vcgu. Ivetic,'88-37
Velazquey, l.uis A., 87-453
Vclcz, V G., 86-93
Vias, C.. 84-273
Virgin. John J , 88-226
Viste, D. R., 84-217
Vitale, Joseph, 88-199
Vocke, R. W., 86-1
Vogel, G. A., 82-214
Voltaggio, Thomas C., 88-317
von Braun, M. C.. 86-200; 89-430
von Undent, I., 86-31, 200; 89-430
von Stackelberg, Katherine, 88-550; 89-
82
Voorhees, M. L, 85-182
Vora, K. H., 84-81
Vrablc, D. L.., 85-378
Wagner, J.. 84-97
Wagner, K., 82-169; 83-226; 84-62;
85-221
Walker, K. D., 84-321
Wall, Howard O., 88-513
Wallace, J. R., 83-358
Wallace. Kenneth A., 87-213
Wallace. E P.. 83-322
Wallace, Robert C. 88-195
Wallace. William A., 88-259
Wallen. Douglas A., 88-138
Waller. M. J.. 83-147
Wallis, D A., 84-398; 85-291
Walsh, J. J.. 80-125; 81-248; 83-376
Walsh, J P.. 8243
Walsh. J., 82-311
Walter. Mama B., 87-409
Walther. E G.. 83-28
WardelJ. J., 81-374
Warner. R. C. 86-365
Warren, S. D.. 89-485
Waster. M. B., 85-307
Watkin. Geoffrey W., 87-508
Watson. K. S.. 85-307
Way. S. C, 84-162
Weathington, B. Chris, 87-93
Weaver, TR. E C. 85-464
Webb. K. B., 84-287; 86-78
Weber. D. D., 83-28, 86-132. 217
Weber, W. J., Jr., 89-526
Wehner. D. E, 89-194
Weiner. P H., 81-37
Weingart, M. D.. 87-405
Weiss, C. 84-546
Wcissman, Arthur B.. 88-8
Went, F. C. 83-175
Welks, K. E.. 80-147
Wells. Suzanne. 88-269
Wentz. John A.. 88-419
Werte. C P.. 89-596
Werner. J D., 83-370; 86-69
West, M. L, 89-586
Western, R. F. 89-99, 157
Wetzel. R. S^ 80-30. 202; 82-169, 191;
85-234
Wheatcraft, S W., 83-108
Wbelan, Gene, 88-295
Whelan, Gene, 85-432; 87-409
Whelan, Gene, 88-539
White, D., 89-497
While, D C. 86-356, 361
White, L A., 85-281
White. M., 80-275
White, R. M., 82-91
White, R. J., 89-41
WhiUock, S. A.. 8346
Whitmyre, Gary K., 87-143
Whitney. H. T^ 86436
Whiuaker, K. P., 82-262
Widmann, W.. 89-163
Wiehl, Christopher D., 88-569
Wieland. Karen A., 88-274
Wiggans, K. E. 85-314
Wilbourn, R. G., 89-3%
Wildeman. Thomas R.. 88-261
Wilder, L. 80-173; 82-233
Wiley, J. B., 85-58
Wilkinson, R. R.. 80-255
Williams, R. J . 89-78
Williams. R C, 86-467
Williamson. J. A., 89-9
Williamson, S. J., 84-77
Willis, N M., 86-35
Willis, N.. 89-606
Wilson, D. C, 804
Wilson, D. J., 89-562
Wilson, L. G., 82-100
Wilson. S. B., 89-227
Wine, J.. 83-428
WinkJehaus, C, 85-423
Wirth, P. K., 84-141
Wise, K. T., 84-330
Witherow, W. E, 84-122
Witmer, K. A., 85-357
Witt, Ann, 88-79
Witten, Alan J., 88-152
AUTHOR INDEX 682
-------
Wittmann, S. G., 85-157
Woelfel, G. C, 85-192
Wohlford, W. P., 89-463
Wolbach, C. D., 83-54
Wolf, F., 83-43
Wolfe, S. P., 85-88
Wolff, Scott K, 87-138
Wong, J., 81-374
Woo, Nancy, 88-145
Wood, D. K., 89-631
Wood, J. G., 89-198
Woodhouse, D., 85-374
Woodson, L., 86-208
Woodward, Richard E., 88-273
Worden, M. H., 84-273
Worden, R., 89-41
Worobel, R., 89-488
Worobel, Roman, 88-424
Worst, N. R., 84-374
Wotherspoon, J., 86-303
Wright, A. P., 80-42
Wright, Brad, 88-55
Wu, B. C, 86-350
Wuslich, M. G., 82-224
Wyeth, R. K., 81-107
Wyman, J., 83-395
Yaffe, H. J., 80-239
Yancheski, Tad B., 88-265
Yang, E. J., 81-393; 83-370; 84-335;
86-52
Yaniga, P. M., 86-333; 89-273
Yaohua, Z., 84-604
Yare, Bruce S., 87-315
Yen, Hsin H., 87-341
Yeskis, Douglas J., 87-213
Yezzi, J. J., Jr., 81-285
Young, C. P., 89-638
Young, L., 80-275
Young, R. A., 81-52
Youzhi, G., 84-604
Yu, K., 80-160
Yuhr, L. B., 85-112; 86-465
Zachowski, Michael S., 87-85
Zaffiro, Alan D., 87-457
Zamuda, Craig, 85412, 419; 86457; 87-
56, 87-61
Zamuda, Craig D., 88-304
Zappi, M. E., 89-519
Zarlinski, S., 89-543
Zatezalo, Mark P., 87499
Zeff, J. D., 89-264
Zhang, Jinrong, 88-467
Ziegenfus, L. M., 84-521
Ziegler, F. G., 81-70; 85-349
Ziemba, W. L., 89436
Zimmerman, P. M., 84-326
Zumberge, J., 89-41
Zuras, A. D., 85-1
AUTHOR INDEX 683
-------
KEY WORD/SUBJECT INDEX
1980 to 1989
Above Ground Closure, 83-275
Accuracy, 88-157
Acid
Extractable Screening,87-107
Mine Drainage, 88-261
Oil Sludges, 88-395
Acidic Waste Site, 85-326
Activated Carbon, 81-374; 82-259, 262;
83-209, 248, 253, 342; 88-409; 89-479
Administrative Order, 88-72
Adsorbent Traps, 87-459
Adsorption, 84-393
Clays, 89-543
Advanced Technologies, 84-412
Aerosol, 88-546
Agency for Toxic Substances and
Diseases
Registry, see ATSDR
Agricultural Fire Residue, 84420
Air
Dispersion, 89-570
Modeling, 82-331; 84-66
Monitoring, 82-67, 268, 299, 306, 331;
83-82, 85; 86-15; 2; 88-335, 557,
561, 567; 89-15
Ambient, 81-280; 83-293; 85-125; 87-284
Cleanup Site, 84-72
Design, 86-152
Emissions, 82-70
Nitrogen Compounds, 83-100
Real Time, 83-98
Sampling, 88-557
Techniques, 82-334; 86-152
Two-Stage Tube, 83-85; 84-81
Photos, 80-116; 85-116
Quality, 82-63
Assessment, 82-76; 87-284
Sampling, 88-567
Pump (SP), 88-567
Stripper, 88-188, 395; 89-479
Stripping, 83-209, 313, 354; 84-170;
88-125, 446; 89-558
Emissions Control, 84-176
In Situ, 89-313
Soils, 86-322
Toxics
Modeling, 89-157
Alara, 87-403
Allied Barrel & Container, 88-32
Alternative
Concentration Limits, 86-173
Financial Mechanisms, 89-600
Hazardous Waste Management, 88-5
Soil Treatment, 88-484
Strategy, 88-214
Treatment Technologies, 86-361
Alternative Remedial Contracts Strategy
(ARCS), 88-15
Ambient, 88-282
Air Quality, 89-157
Ammunition Waste, 88-569
Anaerobic, 88-451
Biodegradation, 88495
Analysis, 8245; 88-145
Attributive Utility, 8844
Drum Samples, 84-39
Environmental, 88-97
Field, 88-251; 8941
Geostatistical, 88-274
Lower Detection Limits, 87-280
Metals, 83-79
Mobile, 86-120
PCBs, 87420
Portable Instruments, 82-36, 40, 57
Pyrographic, 81-114
Quality Control, 84-29
Screening, 83-86; 85-97
Site Data Base, 8449
Soil, 88-251
Spectrometer, 83-291
Analytical Methods
Precision and Accuracy, 89-50
Annuity, 88-23
Antimony, 89-298
Aquatic Ecosystem, 88-119
Aquifer
Alluvial, 87444
Bedrock, 86403
Gravel, 88-219
Response Test, 87-213
Restoration Program, 87-238
ARARs, 87-436; 88-8, 12, 35, 241, 295,
304, 435
Asbestos, 89-547
Compliance, 88-12
Arizona
TCE Contamination, 82424
Arnold Air Force Base, Tennessee,
89-309
Arsenic Waste, 84469; 85409
Asbestos, 85-21; 88-145; 89-547
ASCE, 81-2
Ashland Oil, 88-317
Assessment, 82-17. 27; 83-37
Areal Photography, 85-116
Biological, 82-52
Cold Weather, 82-254
Endangerment, 84-213, 226; 88-295,
539
Environmental, 86-1
Exposure, 86-69; 87476; 88-300, 353
Health, 88-528, 532
Effects, 84-253
Risk, 84-230, 261
Management, 81-348, 351
Mathematical Modeling, 81-306, 313
Mercury Contamination, 82-81
Methods, 81-79
Multi-Attributive Utility, 88-39
Pesticide Plant, 82-7
Petitioned Health, 88-528
Public Health, 88-353
Remedial Action, 88-338
Risk, 86-69; 87485; 88-35, 241, 277,
287, 295, 304, 353, 382, 484, 539,
550, 602; 89-78
Public Health, 89-78
Quantitative, 88-277; 89-78
Site, 85-209; 88-60, 152
Technical Risk, 88-602
Wetland, 87431
Assessments, Type A & B, 88-605
Asset Liquidation, 89-600
Assignment of Obligations, 88-23
ASTSWMO, 88-77
ATSDR, 86467; 88-524, 528, 532, 537
Attributive Utility Analysis, 8844
Audit, 81-398
Environmental, 88-60
Compliance Monitoring, 88-93
Austria, 88-219
Automobile Shredder Fluff (Auto
Fluff), 89-216
Background, 88-282
Baird & McGuire Site, 85-261; 87-138
Bankruptcy, 89-600
Banks and Lending Institutions, 88-60
Bar Code Inventory, 89485
Barriers, 82-249
Bentonite, 82-191; 89-519, 526
Cement, 84-126
Gelatinous, 82-198
Geomembrane, 86-282
Leachate Compatability, 84-131
Sorptive Admix, 86-277
Basic Extraction Sludge Treatment,
86-318
Battery Casings, 89-301
BOAT, 88-12
Bedrock Aquifers, 85-142
Contaminant Movement, 82-111;
85-202
Contamination, 89468
Fractured, 84-150; 87-213; 89468
Fracturing, 89468
Bench-Scale
Study, 81-288
Testing, 80-184; 88-329
Beneficial Use, 84-560
Beneficiation, 88413
Benthic Organism, 88-323
Bentonite, 89-543
Barrier, 89-519, 526
KEY WORD/SUBJECT INDEX 685
-------
-Cement Mixtures
Durability, 85-3-15
Slurry Wall, 89-313
Bcnlomlc-Soil
Mixture Resistance, 84-131
Slurry Walls, 85-357, 369
Benzene. 88-202, 451; 89-570
Berlin & Farro, 81-205
B.E.S.T., 89-348
Bid Protests, 84-520
Bidding, 89-181
Cleanup Contracts, 84-509
Bioassay, 87-66; 88-323; 89-23
Microfax, 88-323
Sediment, 88-323
Bioassessment, 88-72
Bioavailability, 88-142
Biodecomposition, 88-265
Biodegradation, 82-203; 84-393, 85 234;
88^*44, 446, 467, 495
Anaerobic, 88-495
In situ. 88-495
Bioindicators. 81-185
Biological, 88-455
Monitoring, 81-238; 89-75
Technical Assistance Cimup. 89-613
Treatment, 86-253; 87-208
Bio-polymer Slurry Drain. 88-462
Bioreclamation. 85-239; 87-193, 315. 533
Bioremedialion, 88-273, 395, 429, 446,
490; 89-10, 325, 331, 338
Biota, 88-72
Biotechnology. 88-273
Biotransfomiation. 88-138
Blasting. 89-468
Block Displacement Method. 82-249
Borehole
Geophysics, 89-277
Logging. 88-363
Bottom Barrier, 84-135
Bridgeport Rental and Oil Services Site,
85-299
Bno Refining. 87-315
Bromine
Organic. 82-442
BTX. 89-642
Building Decontamination, 84-486
Bureau of Reclamation, 89-652
Buna!
Shon-Tenn, 87-512
Buned
Drums. 80-239
Waste, 87-300; 89-27
California
Superfund Program, 82-428
Ranking System, 85-429
Callahan Site, 82-254
Canal Bottom Liner, 87-334
Cap
Clay, 89-181
Capacity Assurance Plan, 89-606
Capital Budget, 88-602
Capping, 83-123, 2%; 88-245
Cost, 83-370
Carbon Recovery System, 89-558
Carbon Teirachloride, 88-188
Carcinogens, 84-11
Rcporlablc Quantities. 86-162
Case
Histories, 88-395
Management Strategy, 88-79
Cell Model, 85-182
Cement
Asphalt Emulsion, 84-131
Bentomtc Slurry Wall, 86-264
Kiln Dust (CKD), 88-398
Centrifuge Tests
Clay Liners, 89-537
CERCLA (See Also Superfund), 88-295.
537, 539; 89-417
Cleanup Cost Data Base System,
89-186
Enforcement, 89-631
EPA/State Relations, 86-22
Facilities Settlements, 88-23
Options and Liabilities, 86-18
Program Objectives, 89-503
Remedies, 85-4
Settlements
Facilitating. 88-23
Litigation, 88-55
Policy. 89-600
Change Orders, 84-521
Characterization and Analysis, 88-567
Chcmical(s), 88-539
Analysis, Rapid, 80-165
Concentration, 88-282
Control, 81-341; 84-416
Fixation, 87-187
lla/ardnus Releases, 88-37
leaching. 8H-U3
Occurrence. 88-282
Oxidation, 83-253. 87-174
Plant
I-mcrgcncy Removal. 83-338
Ranking Method*. 88-282
Rcngent, 88-419
Specific Parameters. 85-412
Chcmomcinc Profiling. 86 242
Children
Arsenic Exposure, 85-409
China, 84-604
Chlorinated
Hydrocarbons, 88-219, 39S
Oroundwatcr. 89-519
Monitoring. 82-1
PhenoU, 89-325
Volatile Organic*. 88-164
Chlorobcnzenc, 89-570
Chromic Acid. 86-448
Chromium, 88-409, 413; 89-455
Recovery, 88-413
Sludge. 80-259
Circulating Bed
Combustion. 89-3%
Combusior, 85-378
Citizen Information Committees, 85-473
Claims. 84-521; 89-647
Clay. 88-440
Cap. 88-199; 89-181
Leachatc Interaction, 83-154
Liners. 89-512. 543
Deformation, 89-537
Organic t-eachatc Effect, 81-223
Organically Modified, 88-440
Plastic, 89-512
Cleanup, 80-147, 257; 88-317; 89-282.
286, 325
Activities, 88-313
Air Monitoring. 84-72
Alternative Levels, 88-287
Asbestos. 85-21
Assessment. 83-389; 85-116
Bioassay, 87-66
BT-KEMI Dumpsile, 83-342
Case Studies, 83-395; 84-»40
Coal Far. 83-331
Cold Weather. 82-254
Community Relations, 85-468
Contract Bids, 84-509; 87-496
Cost(s), 89-186
Allocation. 84-326
L-ffcitivencss, 86-193
PRP Ability to Pay, 89-600
Criteria, 83-301; 88-103
Degree, 87-436
Delays, 83-320
Drum Site, 83-354
Dual Purpose, 83-352
Effectiveness,
l
-------
Concrete, 88-419
Cone Penetration Test, 88-158
Confirmation Study, 88-208
Confined Disposal Facility. 88-338, 343,
347
Connecticut
Risk Evaluation, 80-25
Consent Decree, 89-592
Consistency, 88-79
Consultant
Liability, 86-47
Contained Aquatic Disposal, 88-338, 347
Container-Piles, 88-479
Contaminant, 88-245, 295
System Design, 82-175
Transport, 86-88; 88-539; 89-570
Volatilization, 88-498
Contaminated
Land, 84-549
Sediment, 88-338
Soil, 83-226, 231; 88-395, 409, 424,
435; 89-396
Cleanup, 83-354; 87-172
Contamination, 88-208, 300
Explosives, 88-569
Groundwater, 88-84, 113
Mapping, 83-71; 84-85
Contingency
Fund, 80-21
Plan
Massachusetts, 83-420
Remedial Sites, 84-489
Continuing Evaluation, 88-567
Contract, 88-214
Administration, 89-647
Laboratory Program, 87-43; 88-282
Contractors
Indemnification, 86-52; 87-521
Liability, 87-34, 520
Contracts
Bidding, 87-496
Construction, 87-496
Control, 87-492
FIT, 86-36
Remedial Planning, 86-35
REM/FIT, 83-313
Superfund, 86-40, 46
Technical Enforcement Support, 86-35
Cooperative Agreement, 84-103; 85-53
Copper Smelter
Arsenic Wastes, 85-409
Corporate Successor Liability, 87-48
Corrective Action Process, 89-503
Correlation, 88-103
Cost, 80-202; 81-248; 82-289; 83-209;
88-409, 598
Above Ground Waste Storage, 82-228
Air Stripping, 83-313
Analysis, 89-404
CERCLA Financed, 83-395
Cleanup, 82-262; 83-296, 366, 370;
84-341; 89-186, 282
Allocation Model, 84-326
Level, 83-398
Closure Apportionment, 86-56
Computer Models, 83-362
Cover, 82-187
Discounting Techniques, 86-61
Earned Value, 87-492
Effective, 88-594
Effective Screening, 85-93
Effectiveness, 89-404
Evaluation, 82-372; 84-290; 86-193
Estimates, 80-202; 84-330, 335; 88-594
Ground Freezing, 84-386
Groundwater Treatment, 83-248, 358
Health and Safety Impact, 83-376
Interest and Litigation, 88-55
Lackawana Refuse Site, 87-307
Leachate
Collection, 83-237
Monitoring, 82-97
Management, 84-339
Minimization, 81-84; 87-258, 326
Model, 87-376
Recovery, 84-313; 88-605; 89-600
Actions, 88-277
Documentation, 82-366
Private, 88-67
Reduction, 88-287
Remedial, 82-118
Remedial Action, 89-181
Risk Benefit Analysis, 88-484
Savings, 86-164, 420
Via Negotiation, 82-377
Treatment System, 81-294
Water Recoveiy System, 82-136
Counting Techniques, 88-145
Coventry, RI, 80-239
Covers (See Also Caps), 82-183, 187,
448; 84-588
Design, 89-4
and Construction, 85-331
Pesticide Disposal Site, 85-349
Credibility, 88-157
Creep Characteristics, 86-247
Creosote, 88-226; 89-642
Bioremediation, 87-193
Contamination, 89-130
Impoundment, 85-323
Incineration, 89-387
Cresol, 88-424
Criticism, 84-532
Cutoff Wall, 83-123, 296
Chemically Resistant, 83-169,, 179, 191
Cost, 83-362
Cyanides, 84-598, 600; 88-467
Cylinder, 88-183
Management, 87-268
Damage
Models, 88-15
Recovery, 81-393
Data
Bases, 83-304; 84-49, 59, 88-282
Problems, 86-213
Gathering, 88-259
Quality, 89-50
Objectives, 88-35
RI/FS, 86-398; 87-72
DC Resistivity, 86-227
De Minimis Settlement, 89-190
Debris, 88-12, 419
Decay Theory, 87-208
Dechlorination, 88-429
Decision, 88-55
Analysis, 88-44, 55
Making, 81-230
Tree Analysis, 82408
Decommissioning, 89-586
Decontamination, 80-226; 88-419, 557;
89421, 586
Buildings, 84-486
Waterway, 83-21
Defense Environmental Restoration
Program (DERP), 89-596
Defense Priority Model (DPM), 89-99
Deformation
Clay Liner, 89-537
Degradation, 88-108, 467
TNT Sludge, 83-270
VOCs, 84-217
Delaware Groundwater Management,
89-618
Demonstration, 88-521
Test, 88-504, 508
Denitrification, 88-451
Denney Farm, 81-326
Denver Radium Superfund Site, 89-652
Depth-Specific Samples, 87-320
Dermal Exposure, 87-166; 88-142
DERP (See Defense Environmental
Restoration Program)
Design, 88-594
Mathematical Modeling, 81-306
Preliminary, 80-202
Sample, 88-503
Detection, 88-152
Buried Drums, 84-158
Detonation, 84-200
Detoxification, 80-192; 84-382; 87-533
Fire Residues, 84-420
Dichloroethene, 88-138
1,1-dichloroethene, 88-108
Diesel Fuel, 86415; 88-317, 462
Diffusion
Effective Transport, 87-129
DIMP, 81-374
Dioxin,81-322,326;83-405;84-287; 85-261;
86-78, 97; 87-306; 88-255, 292, 479,
513, 587; 89-117, 286
Destruction, 89-380
Dipole Configurations, 88-84
Direct Reading Instrument (DRI),
88-567
Discovery Methods, 86-84
Dispersion, 88-455
Coefficients, 83-135
Modeling
Chemical Release, 87-525
Disposal, 81-329; 88-183, 335, 343, 575,
592
Above Ground, 83-275
Commercial Criteria, 82-224
Computer Cost Model, 83-362
Confined Facility, 88-347
Contained Aquatic, 88-338, 347
Liability, 83-431
Mine, 85-387
Salt Cavities, 83-266
Shock Sensitive Chemicals, 84-200
DNAPL Oil, 89-497
Documentation
Cost Recoveiy, 82-366
DOD (see U. S. Department of
Defense)
DOE (see U. S. Department of
Energy)
Dose-Response Assessment, 89-82
Downhole Sensing, 83-108; 87-320
Drain System, 83-237
Drainage
Acid Mine, 88-261
Nets, 86-247
Trench, 88462
Dredging, 88-335, 338, 343, 347
Disposal, 88-335, 338
DRF, 88-587
Drilling
Buried Drum Pit, 86-126
Dual Wall Drilling, 87-355, 358
Horizontal, 86-258
Drinking Water
Contamination, 84-600
Drum(s), 82-254
Analysis, 84-39
Electric Method, 87-385
Buried, 82-12; 84-158
Disposal Pit, 86-126
Handling, 82-169
Site Cleanup, 83-354
Tracking, 89485
Dust Control, 84-265
Ebonite Casings, 89-301
Ecoassessment, 88-72
Economic Aspects
Hazardous Waste Sites, 87-264
ECRA, 89-9
Effluent, 88-347
Electric Reactor, 84-382
Electric Utilities Site, 89-377
Electrical Leak Detection, 89-35, 56
Electrochemical Oxidation, 87-183
Electromagnetic
Conductivity, 89-27
Induction, 83-28, 68; 86-132, 227
Resistivity, 82-1
Survey, 80-59; 82-12; 88-84
Waves, 80-119
Emergency
Planning, 84-248; 88-565
Community Right-to-Know Act.
89443
Removal, 83-338
KEY WORD/SUBJECT INDEX 687
-------
Response, 88-37, 313
Emissions
Monitoring, 83-293
Rates, 84-68
Encapsulation, 87-405
Endangered Species, 88-435
Endangerment, 88-72
Assessments, 84-213, 85-396, 423, 438;
88-295, 539
Enforcement, 84-544; 85-21. 89-600
CERCLA
EPA/State Relations, 86-18
Cleanup. 84-478
Endangerment Assessments, 84-2H;
85-396
Information Management, 85-11
Environmental
Analysis. 88-97
Assessment. 89-9
Audit, 88-60, 65; 89-13
Cleanup Responsibility Act (I C'RA).
88-60
Compliance Monitoring, 88-93
Concerns, 84-592; 89-635
Evaluation
Manual, 89-609
Policy. 89^09
Impact. 81-177; 88-435; 89-194. 576
Liability. 87-45. 88-60
Modeling, 87-149
Pathways, 88-532
Risk Analysis, 82-380
Real Estate Transfer. 87-499
Sensitive Areas, 87-341
Torts, 87-»8
Hpidemiologic Study. 84-287, 87-532
Dioxm, 86-78
Estuary
PCB Analysis. 87-420
Etnyleoe Glycol, 89-298
European Technology, 88-193
Evaluation. 88-329, 504
Continuing, 88-567
Groundwatcr. 88-19
Public Health. 88-304
Evaporation. 88-424
EXAMS Model. 88-119
Excavation, 82-331; 88-479; 89-463
Executive Branch Dispute Resolution.
89-631
Exhumation, 82-150
Expedited Response Action (l-RA),
86-393; 88-188. 226
Expert
Judgment, 88-44
System, 88-93
Exploratory Drilling, 86-126
Explosives
-Contaminated Materials, 89-289
Contaminated Soils Incineration,
84-203
Contamination. 88-569; 89-493
Waste Disposal Sites, 84-141
Exposure, 88-119, 142, 528
Assessment, 86-69; 87-126, 153
88-300, 353; 89-82
Children, 84-239
Limit, 88-546, 567
Pathway, 88-300
Response Analysis, 82-386
Scenarios, (BW84
Extraction, 84-576; 89-479
Groundwaier, 89-241
Metals, 87-380
Soils, 89-348
Vacuum, 87-273
Wells. 88-125
Fast-'I racked
Design and Cleanup, 87-296. 362
llydrogeological Study. 85-136
Fate, 88-119
and Transport, 87-126
r-'aull Tree Analysis, 88-3K2
Faunal Species,, 89-576
Feasibility Study (FS). 88-113, 295, 338.
435, 484, 490, 89-436
Arsenic Waste. 84-469
Federal
Cleanup. 85-7
Compliance Program. 89-631
Facility
Compliance, 88-516, 565; 89-631
Coordinator, 85-32
Smie and Ixxral Jurisdiction, 87-53
State Cooperation, 82-420; 85-50
Field
Analysis, 88-251
Data Acquisition, 86-148
Identilicnlion. 85-88; 86-120
Investigations, 89-251
Operation Methods, 87-28
Quality Assurance. 86-143
I jhoratory, 87-93
Sampling, 84-85, 94
Screening. 86-105. 87-100. 107, 88-174;
89-197 11
Validation. 88-323
Financial
Ability to I'.iy. 89-600
Assessment, 89-600
Fire, 81-341. 82 299
Underground, 86-350
Firefighter
Toxic Exposure. 86-152
First Responder Training. 85-71
ITT
Contracts, 83-313 86-36
Health and Safety, 80-85
Fixation. 89-113
Solidification, 86-297; 87-187. 396
Flotation. 88-455
Homing Covers. 84-406
FlorcfTc, 88-317
Honda. 88-287
Remedial Activities. 82-2f*
Fluorescence. 86-370
X-Ray (XRF) Spectroscopy, 88-9-
Hushing
Soil. 89-207
Hy Ash
Bcnioniic Barrier Improvement,
89-526
Foams
Vapor Suppression, 87-480
Foodcham, 88-359
Fort Miller. 81-215
Foundry Wastcwater, 84-598
FT/IR, 86-371
Fuel Spill, 88-202
Fugaciry. 88-142
Fugitive
Dust Control, 84-265
Hydrocarbon Emission Monitoring.
81-123
Funding
Mixed, 89-592
Galvanizing Operation, 88-245
Cms. 88-183
Chromaiograph. 82-57, 58: 83-76
PCI1 Analysis, 87-420
Portable. 82-36; 83-105; 89-15
Screening. 86-386
ChromuiogrHphy/Thcrmal retraction,
89-41
Collection and Treatment, 86-380
Cylinder Management, 87-268
Migration, 88-265
Plants, 86-93
Subsurface, 89-251
Unknown, 84-416
Gasification Plant Site Contamination,
86-242
Gasoline, 85-269
l-jctraciion, 87-273
Gaussian Puff Model. 87-465
GC/MS. 82-57; 89-50
PCH. 87-420
Generator Cleanup, 85-7
Geochcmical
Control, 89-267
Modeling, 88-245
Geographic Information Systems
86-200; 89-430
Gcohydrology. 83-117; 89-259
Geologic Repositories, 87-502
Geomcmbranes, 86-269; 89-56
Barrier Technology. 86-282
Linen
Leak Detection, 89-35
Seam Testmg, 86-272
Geophysical, 83-68. 71
Diffraction Topography, 88-152
Investigation, 84-481; 86-217
lagging. 86-292; 87-320
Methods. 82-17
Modeling. 86-110
Monitoring, 83-28
Survey, 81-300
Techniques, 83-130; 86-465; 89-27
Geophysics. 81-84: 82-91; 88-363; 89-277
Characterizing Underground Wastes,
86-227; 87-300
Horizontal Radiate. 87-371
Geostaliiiical
Decision-Making. 89-146
Methods, 85-107: 86-217; 88-274
Geotechmcai Engineering, 89-436
Geoicchnology
Containment System, 82-175
Property Testing, 85-249
Techniques. 83-130
Germany. 84-565, 600
Gilson Road Site. 82-291
Glass Matrix. 89-309
Government
Local, 89-645
Relationships, 89-645
Ground
Engineering Equipment, 87-187
Freezing, 84-386
Penetrating Radar, 80-59, 116, 239;
81-158. 30ft 83-68; 86-227; 87-300
Groundwatcr. 88-108, 138, 164, 219, 234.
300, 375, 382; 89-122, 241, 246. 251,
259, 267. 277, 476. 479, 558
Activated Carbon Treatment, 86-361
Applied Modeling, 86430
Bedrock Aquifers, 86403
Biological Treatment, 86-253, 333
Biodegradation. 85-234; 87-208
Bioremediation. 89-273
Case Histories, 86-430
Chemical Oxidation. 87-174
Chrome Pollution. 86-448
Cleanup. 82-118, 159; 83-354; 84-176;
87-311. 348; 88-19: 89-W7, 313.
468, 534
Collection. 86-220
Computer Modeling, 87-111
Containment, 82-259; 83-169
Movement, 82-111; 85-147
Contamination, 81-329, 359;82-280;
83-13, 358; 84-103, 141, 145, 162
170, 336; 85-43, 157, 261; 88-«4,
113; 89-468
Cyanide, 84-600
Detection, 84-20
Liabilities, 83-437
Mapping, 83-71
Potential, 80-45
Control, 89-436, 468
Diffusion
Effect on Transport, 87-129
Dioxin, 89-117
Discharge to POTW, 89-137
Evaluation, 88-19
Hydrologic, of Landfill, 86-365
Extraction System, 87-330
Flow System, 83-114, 117
Flushing. 86-220
Halocarbon Removal, 85-456
Heavy Metals, 86-306
688 KEY WORD/SUBJECT INDEX
-------
Cleanup, 87-341
Transport, 87-444
HELP, 86-365
Horizontal Drilling, 86-258
Hydraulic
Assessment, 87-348
Evaluation, 83-123
Investigation, 80-78, 84-1, 107;
86-158
In Situ Biodegradation, 85-239
Lime Treatment, 86-306
Management Zone, 89-618
Mathematical Modeling, 81-306
Metal Finishing Contamination,
83-346
Microbial Treatment, 83-242
Migration, 80-71; 84-150, 210
Prevention, 83-179, 191; 84-114;
86-277
Mobility, 84-210' 87-444
Modeling, 82-118; 83-135, 140, 145;
84-145; 86-88; 89-163, 146, 152
Exposure Assessment,
87-153
Monitoring, 80-53; 82-17, 165; 88-363
Evaluation, 85-84
Interpretation, 82-86
Long-Term, 85-112
Post-Closure, 83446
Statistics, 84-346; 86-130
Well Design and Installation,
86460
Plume Definition, 85-128
Pollutant Fluxes, 87-231
Pollution Source, 81-317
Post-Closure Monitoring, 83446
Protection, 80-131, 84-565
Recharge, 86-220
Recovery
Cost, 82-136
Design, 82-136
Remedial Plans, 83-130
Remediation, 86-220; 87-213; 88-125,
446; 89468
Research Needs, 83-449
Restoration, 82-94; 84-162; 86-148;
87-204, 223
Sampling, 81-143, 149
Slurry Wall, 86-264
Interaction, 89-519
Studies, 86-431
Superfund Protection Goals, 86-224
SUTRA, 87-231
TCE Contamination, 82-424; 89-137
Treatability, 81-288
Treatment, 80-184; 82-259; 83-248,
253; 86-220; 87-218; 88-188, 226,
409; 89-246, 436
Trend-Surface Modeling, 87-120
Ultra Clean Wells, 86-158
VOC Biodegradation, 84-217
Well Abandonment, 87-439
Grout, 83-169, 175
Chemistry, 82-220
Grouting, 82-451
Silicates, 82-237
Guarantee Agreement, 88-23
Halocarbon Removal, 85-456
Halogen
Combustion Thermodynamics, 85460
Hanford Site, 89-417
Harbor Contamination, 89-130
HARM, 89-99
Harrisburg International Airport, 85-50
Hazard
Degree, 81-1
Potential, 80-30
Ranking, 81-188
Prioritizing, 81-52
Scoring, 85-74
System, 81-14; 82-396
U.S. Navy Sites, 83-326
Unknown, 81-371
vs Risk, 84-221
Hazardous Materials, 88-119
Identification, 85-88
Release, 87-525; 88-37
Storage
Spills, 82-357
Technical Center, 82-363
Hazardous Substances, 88-537
and Petroleum Products, 88-60
Hazardous-Toxic-Waste, 88-202
Hazardous Waste, 88-295, 446, 539;
89-606
Categorization, 89-488
Emergencies
Information Sources, 84-59
In situ Vitrification, 86-325
Expert Management System, 86-463
Land Treatment, 86-313
Management
Alternatives, 88-5
Facility Siting, 84-517
Policies, 84-546
Screening, 86-370
Short-Term Burial, 87-512
Site, 88-39, 532
Bioremediation, 87-533
Exposure Assessment, 87-153
Ranking, 88-44
Reuse, 84-363
Risk Analysis, 87-471
Safety, 87-162
Social, Psychological and
Economic Aspects, 87-264
Treatment, 86-303; 88-546; 89-298
Health and Safety (See Also Safety),
89-282
Assessments, 84-261, 85423; 88-528,
532; 89-72
Petitioned, 88-528; 89-72
Public Health, 88-353
Communication, 88-524
Community Concerns, 82-321
Concerns, 89-635
Cost Impact, 83-376
Evaluation
Public Health, 88-304
Exposure
Potential Ranking Model, 87-158
Significant Human Exposure
Levels, 88-537
Guidelines, 83-322
Hazardous Waste Site, 87-162
Hazards, 80-233
Potential, 88-567
Medical Surveillance, 87-532
Plan, 83-285
Program, 80-85, 91, 107
Recreational Exposure, 87-143
Training, 86-473
Health Risk Assessment, 84-230, 253;
87-143; 89-108, 582
Heart Stress Monitoring, 84-273
Heat Stress Monitoring. 88-546
Heavy
Black Liquor, 88-313
Metals, 88-12, 84, 261, 338, 343, 353,
359, 398, 508; 89-78, 222, 298
Analysis, 88-97
Cleanup, 87-341
Impoundment Closure, 83-195
Soil Treatment, 87-380
Treatment, 87-218
X-Ray Fluorescence, 86-114
Herbicide(s), 89-325
Dioxin, 89-117
Mixing, 86-97
Hexone Oxidation, 87-183
High-Pressure Liquid Chromatography,
83-86
Horizontal Drilling, 86-258; 87-371
Hot Gas Process, 89-289
HRS Revisions, 88-269
Human Exposure
Potential Ranking Model, 87-158
Significant Levels, 88-537
Human Health Evaluation Manual,
89-609
Hyde Park, 85-307; 88-479
Hydraulic Barrier, 89-259, 468
Deformation Effects, 89-537
Hydrocarbons, 85-269; 88-375; 89-392
Biodegradation, 86-333
Chlorinated, 88-219, 395
Contamination, 89-331
Extraction, 89-348
Field Screening, 87-174
Leaks, 82-107
Petroleum, 88-395
Recovery, 86-339
Hydrogen Peroxide, 89-264
UV Light, 87-174; 89-264
Hydrogeologic
Assessment, 87-348
Data, 84-6
Evaluation, 80-49
Fast-Track, 85-136
Investigation, 8145, 359; 83-346;
86-148, 403
Landfill, 85-182
Hydrogeology, 89-277
Hypothesis Tests, 88-503
Identification, 83-63; 88-329
Hazardous Material, 85-88
Reactivity, 83-54
Illinois
Closure/Post Closure, 83-459
Immediate Removal
Dioxin, 87-306
Immobilization, 82-220; 88429, 504;
89476
Impact
Analysis, 88409, 598
Assessment, 81-70
Impoundment, 8045
Closure, 83-195; 84-185; 85-323;
86-318
Leaks, 83-147
Membrane Retrofit, 82-244
Sampling, 85-80
Surface, 88-245
In Situ, 88455, 467, 504
Biodegradation, 85-234, 239, 291;
88495
Chemical Treatment, 85-253
Decontamination, 88498
Permeability/Hydraulic Conductivity,
88-199
Pesticide Treatment, 85-243
Remediation, 89-338
Soil Decontamination, 87-396
Solidification/Fixation, 85-231
Stabilization, 85-152
Steam Stripping, 87-390, 396
Treatment, 84-398; 85-221; 88446,
490
Vapor Stripping, 89-562
Vitrification, 84-195; 89-309
Volatilization, 88-177
Incineration, 82-214; 85-378, 383; 88-255,
292, 413, 513, 569, 575; 89-286, 374,
377, 387
Air Pollution Control, 87459
Dioxin, 89-380
Explosives Contaminated Soils, 84-203
Gaussian Puff Model, 87465
Halogens, 85460
Mobile, 80-208; 81-285; 87453, 459
Ocean, 87465
Oxygen Technology, 88-575
Performance Assessments, 85464
Research, 84-207
Safety, 864
Sampling, 87457
Sea, 80-224
Incinerator, 88-582
Infrared, 88-513, 582
Mobile, 88-582; 89-380
Portable, 88-587
Regulation, 88-592
Rotary Kiln, 89-374
Shirco, 88-513
Transportable, 89-387
Indemnification, 86-52; 87-520
KEY WORD/SUBJECT INDEX 689
-------
Indirect Heating, 89-421
Inductive Coupled Plasma Spectrometer,
83-79
Industrial
Hygiene, 88-546. 561, 567; 89-15, 75
Property, 89-9
Waste
Biological Treatment, 87-208
Information
Committees, 85-473
Management, 85-11
Infrared Incinerator, 85-383; 88-5K2
Innovative Technology, 88-35, 193, 241,
516, 521
Inorganics, 88-282
Installation Restoration Program,
88-300; 89-309, 596
McClcllan AFB, 84-511; 85-26
Insurance, 82-464; 88-60. 602
Integration, 88-79
Integrity, 88-504
Interagency Management Plans, 80-42
Interest/Discount Rates, 88-55
Interstate 70 Acid Spill. 88-32
Inventory Control, 89-485
Investigation
Hydrogeolomc, 82-280
Remedial, 88-
IRIP, 88-569
IRP, 88-569
295, 363, 539
KPEG Process, 88-474
Kriging. 80-66; 88-274; 89-146
Probability. 88-274
Laboratory
Data, 88-157
Management, 81-96
Mobile, 86-120; 89-19
Quality Assurance, 87-93
Regulated Access. 81-103
Screening, 88-174
La Bounty Site, 82-118
La Saile Electric Site, 89-447
Lackawana Refuse Site, 87-367
Lagoons. 81-129; 82-262
Closure, 89-642
Floating Cover, 84-406
Land Ban, 88-398
Land Disposal
Restrictions, 88-12, 429
Sites
Numeric Evaluation, 87-508
Land Treatment, 86-313
Systems, 89-345
Landfanmng, 88-490
Landfill, 88-164; 89-570
Closure. 80-255; 88-199
Covers, 86-365
Future Problems, 80-220
Gas, 88-164
Leachate. 89-122
Life Cycle. 88-164
Risk, 85-393
Test Cell. 88-199
Leach
Field, 88-409
Tests, 88-484
Leachate, 88-347
day Interaction, 83-154
Characterization, 86-237
Collection, 83-237; 85-192
Control, 84-114; 86-292
Drainage Nets, 86-247
Effects on Clay, 81-223
Generation Minimization, 80-135, 141
Landfill. 89-122
Migration, 82-437; 84-217
Minimization, 81-201
Modeling, 83-135; 84-97; 85-189
Monitonng Cost, 82-97
Plume Management, 85-164
Synthetic, 86-237
Treatment, 80-141; 82-203, 437;
83-202, 217; 84-393; 85-192
Leaching, 88-508; 89-222
Chemical, 88-413
Solid, 88-395
Soil, 88-424
Uad, 84-239; 85-442; 86-164, 200. 303;
89-413, 430
Contamination, 89-301
Recycling. 89-301
Leek Detection, 83-94, 147; 85-362;
87-523; 89-56
Legal Aspects
Extent of Cleanup, 83-433
legislation
Model Siting Law, 80-1
LEL. 88-265
Level of Protection, 88-546
Liability, 82-458, 461. 464, 474; 88-55,
65, 67; 89-13
Consultant, 86-47
Contractor, 87-34, 520
Corporate. 80-262
Successor, 87-48
Disposal, 83-431
Generator. 81-387
Groundwaicr Contamination, 83-437
Inactive Suet, 80-269
Supcrfund Cleanup Failure. 83-442
Supcrfund Minimization, 86-18
Trust Fund, 83-453
Lime, 88-398
Liner. 89-543
Breakthrough. 83-161
Canal Bottom, 87-334
Flexible. 84-122
Leak
Detection, 85-362. 89-35
Location, 82-31
Membrane, 89-56
Synthetic, 89-534
Membrane, 83-185
Testine. 86-237
Liquid Membrane. 89-318
Liquid/Solids Contact Reactors (LSCs),
89-331
lunation,
Expected Moniiary Value. 88-55
Lobsters, 88-359
Love Canal. 80-212, 220; 81-415, 82-159,
399; 86-424
Low Level VOC Analysis, 87-85
Low Occurrence Compounds, 85-130
Low Temperature Thermal Dcsorption,
88-429
Macroinvenebrate, 88-72
Magnetrometry, 80-59, 116; 81-300;
82-12; 83-68; 86-227; 87-300
Management, 88-15, 343
Capacity, 89-606
Plans
New Jersey, 83-413
Remedial Program, 88-15
Superfund, 88-15
Managing Conflict, 84-374
Marine Sediment, 87-485
Marsh Cleanups. 87-341
Mass Selective Detector, 85-102
Massachusetts Contingency Plan. 83-420;
85-67, 89-95
Mathematical Model. 88-119. 359
MCI., 88-8
M( LG, 88-8
Mcdellan AFB, 85-43; 87-204
Medical Surveillance, 84-251, 259;
86^*55; 87-532; 89-75, 91
Membrane-Like-Material, 89-318
MEPAS, 88-295
Mercury, 82-81
Metals, 82-183, 88-282; 89-476
Analysis, 83-79
Cleanup, 87-341
Detection, 80-239
Detector, 80-59; 81-300; 82-12
Finishing, 83-346
Screening, 85-93
Washing, 89-207
Methane, 88-265
Methanogenisis, 88-265
Mclhylene Chloride, 88-446
Microbial Degradation, 83-217, 231, 242
Microbubblc. 88-455
Microcomputer, 89-108
Microdicpersion, 84-398; 85-291
Microcncapsulation, 87-380
Microorganisms, 88-490
Microtox, 89-23
Bioassay, 88-323
Migration, 84-588; 88-132
Cutoff. 82-191
Prevention, 82-448
Sedimantary Channel Deposit, 87-414
Mill
Paper. 88-313
Mine
Disposal, 85-387
Drainage, 88-261
Heavy Metal Mobilization, 87-444
Mine/Mill Tailings, 85-107
Sues. 83-13; 87-436
Tailings Cleanup, 84-504
Waste Neutralization and
Attenuation. 86-277
Minimum Technology Requirement,
88-234
Mixed
Funding, 89-592
Waste. 87-403; 88-539; 89-417
Mobile
Incinerator. 85-378. 382; 87-153. 459;
88-582; 89-380
Laboratory. 80-165. 84-45; 86-120;
89-19
MS/MS. 84-53
Thermal Destruction, 89-377
Treatment. 86-345, 89-392
Waste Oil Recovery, 87-179
Model. 88-108, 142
Vacuum Stripping. 89-562
Modeling. 88-132, 234; 89-267. 570
Air Toxics. 89-157
Applied, 86-430
Cell, 85-182
COM, 87-376
Environmental, 87-149
Exposure Assessment, 87-153
Geochemical, 88-245
Geophysical Data. 86-110
Groundwater, 89-152, 241
Treatment, 83-248; 87-11
Human Exposure Potential Ranking
Model. 87-158
Leachate Migration, 82-437; 85-189
Management Options. 83-362
Plume, 89-146
Random Walk, 89-163
Remedial Action, 83-135
Sediment Movement, 87-426
Site Assessment, 81-306
Three-Dimensional, 89-152
Trend-Surface. 87-120
Molten Baths. 89-421
Monitoring, 88-113, 347
Air, 88335, 546, 561, 567
Ambient Air, 81-122, 136
Montana Pole, 88-32
Medical, 88-546
Radiological Exposure, 88-546
Wells. 88-202
Dual Wall Hammer Drilling
Technique, 87-358
Installation. 81-89
Integrity Testing, 86-233
Installation In Fn-Place Wastes.
8M24
Location. 81-63
State Regulation, 87-89
Monongahela, 88-317
Monte Carlo Technique, 88-550
MS/MS Mobile System, 84-53
Multi-Attribute Utility Analysis, 88-39
Multi-Media
Exposure Assessment, 87-476
690 KEY WORD/SUBJECT INDEX
-------
PCB Cleanup, 87-362
Risk Analysis, 87-471, 485.
Multiple Burner System, 89-374
Multi-Site/Multi-Activity Agreements,
85-53
Municipal Landfill(s), 89-251
RI, 87-72
m-Xylene, 88-451
National
Contingency Plan (NCP), 88-304
Revisions, 86-27
Contract Laboratory Program, 84-29
Priority List (NPL), 85-1; 88-537;
89-552
Deletion, 86-8
Mining Sites, 83-13
Resource Damage, 81-393
Response, 81-5
NATO/CCMS Study, 84-549
Natural
Attenuation, 88-113
Resources,
Damages, 87-517; 89-194
Definition, 88-605
Injury, 89-613
Restoration/Reclamation, 84-350
NCP (National Contingency Plan, see
National)
Negotiated Remedial Program, 84-525
Negotiating, 82-377, 470
Netherlands, 84-569
Neutral Validation RI/FS, 86-445
Neutralization, 83-63
New Bedford Harbor Site, 87-420, 426;
88-335, 338, 343, 353, 359
New Jersey, 88-77
Cleanup Plans, 83-413
DEP, 85-48
Reserve Fund, 85-58
New York City, 84-546
NIKE Missile, 88-202
Site, 88-208
Site Investigation, 86-436
NIOSH, 88-546
Nitrate(s), 89-267
No-Action Alternative, 85^49
"No Migration" Demonstration, 88-234
Non-Destructive
Assay System, 89-586
Testing Methods, 82-12, 84-158;
86-272
Non-Target Compound Identification,
89-86
North Hollywood Site, 84-452
Notification
Mass, 87-7
NPL (National Priorities List, see
National)
Numerical
Evaluation System, 87-508
Model, 88-55
Observational Method, 89-436, 459
Obsidian, 89-309
Occupational Health Programs, 84-251,
259
Ocean Incineration, 87-465
Odor, 82-326; 83-98
Oil
Pond Pollution, 86-415
Recovery, 85-374; 87-179
Retrieval, 89-318
Sludge
Best, 86-318
Spill, 88-317
Cleanup, 89-318
Oily Wastes, 89-318
Old Hardin County Brickyard, 82-274
Olmsted AFB, 85-50
OMC Site, 84-449
On-Site
Leachate Renovation, 84-393
Storage, 89-455
Water Treatment, 87-169
Optimization of Soil Treatment, 87-172
Organic(s), 88-12, 508
Chemical Oxidation, 87-174
Degradation, 89-338
Emissions, 82-70, 84-176
Land Treatment, 86-313
Sludge Stabilization, 84-189
Solvents Permeability, 84-131
Vapor
Analysis, 83-98
Field Screening, 83-76
Leak Detection, 83-94
Personnel Protection, 81-277
Wastes, 88-440
Characterization, 84-35
Fixation, 87-187
Organically Modified Clays, 88^140;
89-292, 543
Organism
Benthic, 88-317
OSHA, 88-546
Safety Requirements, 87-162
Training Requirements, 87-18
Ott/Story, 81-288
Oxidation, 88-467; 89-264, 407
Chemical, 87-174
Electrochemical, 87-183
Oxygen
Incineration Technology, 88-575
Supply, 89-338
o-Xylene, 88-85
Ozone, 89-264
Pacific Island Removal, 84^27
Paper Mill, 88-313
Parametric Analysis, 81-313
Passive Treatment, 88-261
PCBs, 81-215; 82-156, 284; 83-21, 326,
366, 370; 84-243, 277, 449; 86-420;
87-89; 88-241, 251, 329, 335, 338, 343,
353, 359, 419, 474, 508, 513, 575, 587;
89-67, 207, 313, 377, 396, 413, 447,
476
Analysis, 87-420
Cleanup, 87-362
Field
Screening, 89-19
Measurement, 83-105
Fractured Bedrock, 89-497
Land Disposal Site Evaluation, 87-508
Modeling Movement,87-426
Screening, 86-370
Soil
Extraction, 87-104
Treatment, 87-187
PEL (see Permissible Exposure Limit)
Pennsylvania Program, 81-42
Pentachlorophenol, 88-226
Analysis, 88-274
Performance Incentive, 88-15
Incentive, 88-214
Periphyton, 88-72
Permanent Remediation, 89-309
Permanent Remedy, 89-623
Permeability Coefficient Measurement,
84-584
Permissible Exposure Limit (PEL),
88-546
Permitting, 88-582
Persistence, 88-119
Personal Protection, 88-561
Personnel
Protection
Equipment (PPE), 88-546
Levels, 81-277
Safety Equipment, 86-471
Pesticides, 82-7; 85-255, 349; 86-386;
88-395; 89-325
Contamination, 88-495
In Situ Treatment, 85-243
Risk Assessment, 86-186
Petro Processors Site, 84-478
Petro-Chemical Systems Site, 89-282
Petroleum
-Contaminated Soil, 89-345
Contamination, 84-600
Hydrocarbons, 88-395
Sludges, 88-395; 89-292
Pharmacokinetic, 88-142
Phased Approach
Remedial Investigation, 87-326
Phenol, 88^24
Chlorinated, 89-325
Polychlorinated, 88-347
Treatment, 87-218
Photographic Interpretive Center, 84-6
Physical Chemical Data Use, 84-210
Physical/chemical Methods, 88-395
Picillo Farm Site, 82-268
Pilot Plant, 81-374
Bioremediation, 87-315
Pilot Study, 88-347
Pink Water, 88-569
PIRS, 82-357
Pittson, PA, 80-250
Plan Review, 86-143
Plant Bioindicators, 81-185
Plasma Reactor, 89-421
PLM, 88-145
Plugging Wells,87-439
Plume
Capture/Interception, 89-468
Modeling, 89-146
Plutonium Fabrication Facility, 89-586
Policy, 89-609
Pollution Abatement Site, 84-435
Polyaromatic Hydrocarbons, 84-11;
89-259
Polychlorinated
Biphenyls, 88-504
Phenols, 88-347
Polynuclear Aromatic Hydrocarbons,
86-242; 89-23, 130
Bioremediation, 87-193
Pond Closure, 88-245
Portable Incinerator, 88-587
Post-Closure
Care, 81-259
Failure, 83-453
Groundwater Monitoring, 83-446
Monitoring, 82-187
Monitoring Research, 83-449
Potential Health Hazard, 88-567
Potentially Responsible Party (PRP),
85-275; 89-190, 600
Risk Premium, 87^1
Search, 87-5; 89-600
Methodologies, 87-21
POTW
Groundwater Discharge To, 89-137
Leachate Treatment, 83-202
Power Curves, 88-503
Pozzolans, 88-398; 89-413, 476
Preauthorization Decision Document,
89-592
Precipitation, 88-398
Preliminary Off-Site Evaluation, 88-567
Pre-Remedial Programs, 87-14; 88-269
Pretreatment, 89-455
Price Landfill
Groundwater Computer Modeling,
87-111
Remedial Action, 83-358
Prioritization (See Also Hazard
Ranking), 81-188; 87-409; 88-79
Priorities, 88-32
Removal, 88-32
Private
Cleanups at Superfund Sites, 86-27
Cost Recovery, 88-67
Property Legal Issues, 86-31
Probabilistic Spatial Contouring, 85-442
Probability
Assignment, 88-55
Kriging, 88-274
Product Recovery, 88-226
Program Optimization System, 88-39
Property
Rjsk Assessment, 87-45
Transfer, 89-9, 13
KEY WORD/SUBJECT INDEX 691
-------
Protection
Level of, 88-546
Proton Magnetometer, 89-27
PRPs, 88-32
Public Cost, 89-181
Psychological Aspects of Ha/ardous
Waste Site, 87-264
Public
Awareness. 83-383
Health, 84-232; 85-438; 87-138; 88-524
Assessment, 88-353, 550
Risk, 89-78
Statement, 88-537
Information, 89-447
Program, 80-282; 84-3; 85-473
Needs, 84-368
Involvement, 85-476
Meetings, 88-269
Participation (Sec Also Community
Relations) 82-340,346, 350;
83-383; 88-400; 89-635
Communication, 87-254
Failures, 83-392
Policy
Cleanup Level. 83-398
Relations, 85-468
Pulsed Radio Frequency, 81-165
Punitive and Natural Resource Damage,
88-55
Purge and Trap, 88-174
Purgeable VOC, 88-174
p-Xylene, 88-451
Pyrolysis, 88-413: 89-309
Quality
Assurance
Audits. 84-94; 86-143
Field Laboratory, 87-93
Lower Detection Limits, 87-280
Monitoring Well Integrity. 86-233
Control, 82-45; 84-29; 86-287
Indicators. 89-50
Radar Mapping. 85-269
Quantitative Risk Assessment, 88-277.
89-78
Radials
Horizontal, 87-371
Radio
Frequency, 88-498
Radioactive
Health Risk, 89-582
Mine Tailings. 84-504
Mixed Wastes. 87^03
Naturally Occurring Material. 89-652
Site Assessment, 85-432
Wastes, 81-206; 87-405; 88-193; 89-4.
417
Radiological Exposure Monitoring,
88-546
Radionuclidcs, 86-306; 89-198, 576
Radium, 89-198
Concentrations, 88-103
-Contaminated Soil, 89-652
In Soil, 88-103
Processing Residues, 84-445
Radon, 89-198
Contamination, 84-457
Gas, 82-198
RAMP, 82-124
Love Canal, 82-159
Random Walk Model, 89-163
Ranking, 88-208
Chemical, 88-282
System, 81-14; 85^129
RCRA, 88-295, 539; 89-417
CERCI.A Integration, 89-631
Closure Options
Superfund Sites, 87-337
Hnforcemcnt, 89-631
Program Objectives, 89-503
Requirements, 85-4
Section 3012, 84-535, 544
Superfund
Interrelationship, 86-462
Response Impact, 87-515
RDX, 82-209; 88-569
Reactivity
Identification. 83-54
Real Estate
Hazardous Waste Implications, 82-474
Transfer, 87-499
Transact ionk. 88-60
Reclamation
Chromium Sludge, 80-259
Recognized Hazard, 88-567
Recommended Exposure Limit* (RliL).
88-546
Records Management System, 81-30
Recoverable Storage. 89-455
Recovery, 88-375
Chromium, 88-413
Hydrocarbons, 86-339
Orgamcs, 84-145
Recreational Exposure, 87-143
Recycling, 89-301
Regional Response Team, 80-6; 82-274
Regulatory Impact, 89-606
REL (s.cc Recommended I-jcpr»urc
Limits)
Releases. 88-32
REM Contracts, 83-313
Remedial
Action, 82-289; 88-35, 241, 245. 435
Alternatives, 84-35. 277. 290. 306,
321; 86-361. 87-258
Risk Assessment. 85 319.
86-65
Assessment. 88-338
Bedrock Aquiforv 86-403
Cisc Studies. 82-131
Const ruction Contract!.. 87-4%
Contingency Plans. 84-489
Cost. 84-335. 341; 89-181
Management. 84-339
Model. 87-376
Criteria, 88-5
Decision-Making, 84-66
Design, 80-202; 87-367
Pesticides, 85-255
Evaluation System, 87-238
Florida's Site, 82-295
Groundwater, 84-565
Horizontal Radials, 87-371
Investigation, 84-435
Data, 88-532
Guidance, 84-498
Lessons, 84-465
Negotiated, 84-525
Netherlands, 84-569
North Hollywood Site, 84-452
Objective. 88-19
Options. 80-131
Pesticides, 86-186
Planning, 85-281
Planning Contracts, 86-35; 87-492
Priority System. 85-432; 87-409
Program, 88-214
Management, 88-15
Progress Status. 80-125
Public Involvement, 85-476
Screening and Evaluation, 84-62
Selection, 84-493
Smelter Sue. 86-200
Soil Treatment, 87-172
Technologies, 85-285
Alternatives. 88-300
Assessment, 89-497
Construction
Safely Plans. 83-280
Cost luMimation Model, 84-330
Design
Groundwalcr, 83-123. 84-109, 35f>
Model Based Methodology,
83-135
OMC Site, 84-449
Thamcsmead, 84-560
Investigation, 88-295, 363. 539; 89-459,
555
Municipal Landfill, 87-72
Planning
Foreign, 88-219
Programs, 88-60, 594
Management, 88-15
Projects
Corps of Engineers, 83-17
Response
Role of U.S. Army, 82-414
Technologies
Cost, 89-186
Screening and Evaluation, 84-62
Remediation, 88-152, 251, 259. 375, 409
419, 594. 89-3%, 430, 623
Design
Sedimentary Channel Deposits,
87-415 ^
Discounting Techniques, 86-61
Innovative Approach, 85-307
Western Processing. 87-78
Remote Controlled Excavation, 89-463
Remote Sensing, 80-59. 239; 81-84, 158
165, 171. 88-152
Removal
Emergency, 88-32
Pnonlies. 88-32
Rcportable Quantities. 86-182
Reporting Requirements, 88-37
Research
Post-Closure Monitoring, 83-449
US EPA Program. 80-173
Reserve Fund, 85-58
Residual. 88-108
Resistivity, 80-239; 81-158; 82-31; 83-28
Resource
Damage. 89-194
Recovery. 81-380
Response
Costs. 88-32
Emergency, 88-13
Model 81-198
Procedures, 80-111
Restoration
Natural Resource, 89-613
Swansea Valley, 84-553
Resuspension. 88-347
Retardation Factor. 88-245
Retention Index. 89-86
Reusing Hazardous Waste Sites. 83-363
Reverse Osmosis, 82-203
Reversionary Trust 88-23
Rl/FS. 88-15, 55, 343; 89-552
Bridgeport Oil and Rental Services
Site. 85-299
Chromic Acid Leak. 86-448
Computerized Expert Systems, 86-208
Data Quality Objectives. 86-398
Guidance, 88-1
Neutral Validation. 86-445
New Bedford Site. 87-420
NIKE Missile Site, 86-436
Phased Approach, 87-326
Project Performance Improvement.
87-1
Site-Specific Values, 87-126
State Cooperation, 88-15
Wood Treating Site, 86-441
Right-to-Know, 86-4
Risk, 88-142, 145. 300
Acceptability, 83-405; 88-382
Analysis, 81-230; 83-37; 87-471
Computer, 84-300
Environmental, 82-380
Premium, 87-41
Assessment, 81-238; 82-23, 386, 390,
406, 408; 83-342; 84-283, 321;
85-393, 412, 449; 86-69, 74, 457;
87-61; 88-35, 65, 241, 277, 287,
292, 295, 304. 353, 382, 484, 539,
550, 602; 89-102, 67, 78, 82, 95,
108
Air Quality, 82-63
Communication, 87-254
692 KEY WORD/SUBJECT INDEX
-------
Comparative, 83-401
Data Problem, 86-213
Dermal Exposures, 87-166
Dioxin, 89-117
Environmental Modeling, 87-149
Foodchain, 89-13
Health, 84-230
Manual, 85-419
Modeling, 82-396
Multi-Media, 87485
Prioritizing, 85-433
Properties, 87-45
Public Health, 87-138
Quantitative, 84-290; 86-65, 186
Radioactive Chemicals, 89-582
Remedial Action Alternatives,
85-319
Scoping Level, 87-143
Underground Tanks, 84-16
U. S. EPA Guidelines, 86-167
Cleanup Level, 83-398
Concepts
Superfund Process, 87-251
Decision Analysis Module, 86-463
Design, 84-313
Estimation, 88-382
Evaluation, 80-25
Management, 89-91
Minimization, 81-84
Perception, 86-74
Superfund Sites, 87-56
Risk-Based Approach, 88-208
Roasting, 87-380
Rocky Mountain Arsenal, 81-374;
82-259; 85-36; 89-75
Rotary Kiln Incinerator, 89-286, 374
Routes of Exposure, 89-67
RRT, 88-317
Safety (See Also Health and Safety),
82-299, 306; 85-406; 89-75
Cost Impact, 82-311
Equipment, 86-471
Incineration, 86-4
Information, 84-59
Plans, 84-269
Procedures, 81-269
Remedial Construction, 83-280
Sampling and Analysis, 81-263
Tank Investigation and Removal,
85-198
Training, 82-319
Sample
Design, 88-503
Preparation, 88-145
Size, 88-503
Thief, 81-154
Sampling, 80-91
Air, 88-546, 567
Pump (SP), 88-567
Analysis
Safety, 81-263
Biological, 82-52
Drums, 81-154
Impoundments, 85-80
Screening, 81-103, 107, 114
Statistical-Based, 86-420
Strategy, 85-74
Subsampling, 84-90
Techniques, 81-143, 149
Sanitary Wastes, 88-164
SARA, 88-5, 269, 295, 409, 537, 539,
598
Title III, 89-443
Scoping Level Assessment, 87-107
Screening, 88-329; 89-41
Acid Extractables, 87-107
Analytical, 85-97
Field, 86-105
Mass Selective Detector, 85-102
Metals, 85-93
PCB. 86-420
Spectrometry, 83-291
Statistical, 86-164
X-Ray Fluorescence, 86-115
Sealed Double-Ring Infiltrometer,
88-199
Security, 83-310
Sediment , 88-353
Bioassay, 88-323
Contaminated, 88-338; 89-130
Toxicity, 89-130
Transport, 88-338
Sedimentary
Channel Deposits, 87-414
Movement, 87-426
Multi-Media Risk Assessment, 87-485
PCB Analysis, 87-420
Seismic
Boundary Waves, 85-362
Refraction, 80-239; 86-227
Sensing
Downhole, 83-108
Serum Reference Methods, 84-243
Settlement, 85-275; 89-190, 592
Agreements, 82-470
Hyde Park, 85-307
Authorities, 88-23
CERCLA Facilitation, 88-23
De Minimis, 89-190
Financing Mechanism, 8_-23
Inflation Hedge, 88-23
Offer, 88-55
Structural, 88-23
Specialist, 88-23
Sewer Line Decontamination, 89-493
Shenango, 80-233
Shirco Incinerator, 88-513
Shock Sensitive/Explosive Chemical
Detonation, 84-200
Shope's Landfill Cleanup, 83-296
Short-Term Burial, 87-508
Shotblasting, 88-419
Significant Risk, 89-95
Silicates, 82-237; 86-303
Grouts, 83-175
Silresim Site, 82-280
SITE, 88-77, 508, 513, 516, 521; 89-264,
396, 404, 407, 421
Site, 89-413
Assessment, 80-59, 91; 83-221; 84-221;
85-209; 88-60, 152; 89-9
Discovery, 83-37; 86-84
Entry, 88-567
Evaluation, 80-25, 30
First Year, 87-25
Hazard Rating, 80-30
Inspection, 88-269
Sampling Strategy, 85-74
Investigation, 85-48
Listing, 89-552
Location, 80-116; 81-52
Location Methodology, 80-275
Problems
Whales, 84-594
Program, 86-356
Ranking, 89-99
Remediation, 89-459
Reuse, 84-363, 560
Screening, 88-97
Siting, 80-1
Hazardous Waste Management
Facility, 84-517
Public Information Needs, 84-368
Slagging, 88-193
Sludge, 88-413; 89-292
B.E.S.T. Process, 86-318
Stabilization, 86-277
Slurry
Trench, 82-191; 88-462
Wall, 85-357, 374; 86-264; 89-181, 519
Small Quantity Generator, 85-14
Smelter, 89-430
Lead, 84-239; 85-442
Site Remediation, 86-200
Social Aspects
Hazardous Waste Site, 87-204
Soil, 88-12, 142, 145, 282, 467, 490, 546
Advanced Technologies, 84-412
Air Stripping, 86-322
Analysis, 88-251
-Bentonite
Barrier, 89-526
Slurry Wall, 89-519
Bioremediation, 87-533
Characterization
Electric Method, 87-385
Chemistry of Hazardous Materials,
86-453
Cleanup, 88-202, 495
Contamination, 82-399, 442; 83-43;
84-569, 576; 88-395, 409, 424, 435,
569; 89-345
Coal Tar, 89-642
International Study, 82-431
Pesticides, 85-243; 88-495
Cover, 86-365
Decontamination, 87-396; 88-498
Dioxin Contaminated, 88-292
Extraction, 82-442; 89-348
Flushing, 89-207
Gas
Analysis, 86-138
Groundwater Survey, 88-158
Sampling, 84-20
Survey, 87-97, 523; 89-555
Gasoline Extraction, 87-273
Geotechnical Property Testing, 85-249
Heavy Metal Treatment, 87-380
Incineration, 89-387
Leaching, 88424
Liners, 89-512
Construction, 89-512
PCB Analysis, 89-19
Radium-contaminated, 88-103
Stabilization, 87-198
Solidification, 89-216
Steam Stripping, 87-390
Superfund, 88-429
Treatment, 88429, 474; 89-396
Alternatives, 88-484
Optimization, 87-172
Thermal, 84404
Vapor
Extraction, 89-479
Measurement, 85-128
Stripping, 89-562
Venting, 88-177
Washing, 85-452; 88-193, 424; 89-198,
207, 318
Soil-Bentonite Slurry Walls, 85-357, 369
Solid Waste Management
China, 84-604
Solidification, 81-206; 88-395, 440, 508;
89-216, 222, 413
Fixation, 86-247
Organics, 86-361
Silicates, 82-237
TNT Sludge, 83-270
Soliditech, 89-413
Solubility, 88-108
Solute
Migration Control, 89-526
Transport, 89-152
Solvent
Extraction, 88-429; 89-348
Mining, 83-231
Sorption, 88-132
Source Control, 88-188
South Valley San Jose 6 Site, 87-355
Spatial Contouring, 85442
Spectroscopy
X-Ray Fluorescence (XRF), 88-97
Spent Solvents, 88-164
Spill(s), 88-313, 317
Hazardous Materials Storage, 82-357
Stabilization, 80-192; 88-440: 89-216, 222
292, 476
Viscoelastic Polymer Waste, 85-152
Stabilization/Solidification, 80-180-
85-214, 231
Organic Sludge, 84-189
Quality Control, 86-287
Soil, 87-198
Startup
KEY WORD/SUBJECT INDEX 693
-------
Groundwatcr System, 87-223
State
Cooperation, 88-15
Criticism. 84-532
Enforcement 84-544
Participation, 82-418; 84-53
Plans
New Jersey, 83-413
Pennsylvania, 81-42
Statute
Natural Resource Injury, 89-613
Supcrfund Program, 82-428; 85-67
Statistical
Analysis
Air Toxics Data, 89-157
Methods. 84-243
Groundwaier Monitoring. 84-346;
86-132
Sampling, 86-426
Screening. 86-64
Modeling
Geophysical Data. 86-110
Statistics, 88-503
Steam Stripping, 82-289; 87-390, 396;
89-558
Storage Tank locates. 88-462
Strategic Planning, 88-79
Streamline, 89-488
Stnngfellow Site. 80-15, 21
Stripper
Air, 88-395
Structured Settlements, 89-600
Subsamplmg. 84-90
Subsurface Geophysical Investigation,
84-481
Supcrfund (See Also CERCLA), 88-108,
113. 145. 214, 338. 409. 419. 435, 503.
89-309
California, 81-37
Cleanup Failure liability. 83-442
Compliance. 88-12
Contractor
Indemnification, 86-56; 87-520
Liability, 87-34
Contracts, 86-40. 46
Drinking Water. 83-8
Federal/State Cooperation. 81-21;
83-428
Field Operations Methods. 87-28
Groundwaier Protection Goals.
86-224
Impact on Remedial Action. 86-407
Implementation. 83-1
Innovative Technology Programs.
86-356
Management. 83-5; 88-15
Natural Resources Damage. 87-517
Private Cleanup, 86-27
Private Property Cleanup, 86-31
Private Sector Concerns. 81-10
Programs
New Jersey. 83-413
Texas. 83-423
RCRA
Closure Options. 87-337
Interrelationship, 86-462
Response Impact, 87-509
Revisited, 86-412
Right-to-Know, 86-11
Risk Assessmcnl. 87-61
Risk-Based Policy, 87-251
Site
Management, 86-14
Risk. 87-56
State
Perspective, 84-532
Programs, 88-72
Strategy for Dealing With, 86-469
U. S EPA Research, 81-7
Surface
Geophysics, 87-300
Impoundment, 88-245
Sealing, 81-201
Water
Exposure, 87-143
Management, 80-152
SUTRA, 87-231
Swansea Valley, 84-553
Swedish Dump Site Cleanup. 83-342
Sweeney, 82-461
Sydney Mine Sue. 8S-28S
Sylvester Site 81-359
Synthetic
Liner, 89-534
Membrane Impoundment Retrofit,
82-244
Tailings. 85-107
Tank Investigation and Removal, 8S-198
Tar Creek Site. 87-439
TAI
Health and Safety. 80-85
2,1.7,8-TCDD. 88-292
TCI- Contamination. 82-424
Technical Enforcement Support
Contract, 86-38
Technology
Emerging. 88-516
European, 88-191
(•valuation. 82-233
Innovative, 88-193. 516
Treatment. 88-329
Icntativclv Identified Compound*. 89-86
I.l,2.2-letrachlt>roethanc. 88-138
Texas
Ambient Air Sampling, 8S-125
Supcrfund Program, 83-423
Thamesmcad. 84-560
Thermal
Destruction, 88-429
Extraction/Gas Chromatography.
89-U
Treatmeni
Soils, 84-104
Volatilization System, 89-392
Thermodynamics
Halogen Combustion, 85-100
Thin-Layer Chromatography, 86-420
Time Varying Parameters. 89-108
Times Beach, 88-255
Title III, 88-516, 565
Compliance, 89-443
TLV. 88-546
INT. 82-209; 85-314; 88-569 89-J93
Toluene. 88-451
tomography. 88-152
Top-Sealing, 80-135
Town Cias. 84-11. 86-93
Toxaphene, 88-495
Toxic Substances and Disease Rcgisir,
Agency. 85-403
Toxicity. 88-119
Sediments. 89-130
Toxicological
Data, 86-193
Profiles, 88-537
Toxin-Exposure. 89-91
Trace Atmospheric Gas Analyzer, 83-98,
100
Training, 88-546
Tirst Respondent 85-71
OSHA Requirements, 87-18
Resources, 83-304
'Transport, 88-132
Contaminant. 88-539
Heavy Metals. 87-144
Model. 88 125. 287
Transportable Incinerator. 89-387
Transuranic Waste, 89-586
Trcatability. 88-12
Study. 88-1, 484
Composting, 89-298
Tests. 88-413
Treatment. 88-ISS. 53]
effectiveness. 88-429
Groundwatcr, 89-241
In Snu. 82-451; 83-217, 221. 226, 231
Mobile. 86-345
On-Silc, 82-442
Passive, 88-261
System Design, 81-294
Technology, 88-329
Trench
Drainage, 88-462
Slurry, 88-462
Trend-Surface Modeling. 87-120
Tnchlorobenzcnc, 89-497
1,1,1-lnchloroethanc. 88-108
Tnchloroethcnc. 88-138
Tncbloroethytene. 89-313, 497
Groundwatcr Contamination, 89-137
Trieium. 89-576
Ultraviolet Ught. 89-264
fH,O2 oxidation. 87-174
UMTRA Proiect. 87-449
Uncertainty, 88-259
Analysis. 89-102, 82
Engineering, 89-436, 459
Underground Storage Tank, 88-202
Fuel. 86-350
Leak Detection. 87-523
Spill Risk Assessment, 84-16; 86-176
Tnchloroethylene, 86-138, 430
Waste Characterization, 86-227
United Kingdom. 80-8. 226
Unknown Gases, 84416
Unsaturated
Flow, 88-234
Zone, 88-132
U.S. Army
Corps of Engineers, 82-414; 83-17;
88-15
Installation Restoration Program.
84-511
U.S. Coast Guard (USCG), 80-6
U-S. Dept. of Defense (DOD), 89-596
Environmental Restoration Program.
82-128. 87-7
Hazardous Materials Technical
Center. 82-363
IRP. 85-26
Site Cleanup. 83-326
TNT Cleanup, 85-314
US. Depi. of Defense (DOD), 89-99
U.S. Depl. of Energy (DOE), 85-29; 88-
39; 89-582, 586. 652
CEARP, 86-1
L S Environmental Protection Agency
(EPA) ^
Expedited Response Action Program.
86-393
Mobile Incinerator, 81-285
Reponable Quantities, 86-182
Research. 81-7
Risk Assessment Guidelines, 86-167
U.S. Navy. 85^8
Uranium, 89-267
Tailings, 87-449
UV/Hydrogen Peroxide, 89-407
UV/Ozone, 89-264. 407
Study 85-456
Vacuum
Extraction, 87-273. 39ft 88-193
Stripping. 89-562
Vados Zone, 88-158. 164
Monitoring, 82-100
Value Engineering. 88-594
Vapor
Emission. 82-326
Extraction System. 88-188
Foam Suppression. 87-480
Soils, 85-128. 157
Variance. 88-234
Variogram, 88-274
Verona Well Kicld,87-330
Vienna Basin, 88-219
Vinyl Chloride. 88-138
Viscoelaslic Polymer Waste. 85-152
Vitrification, 87-405
In Situ, 84-191; 86-325
VOC (Volatile Organic Comoound),
88-125. 158, 174, 219, 287. 395, 409;
89-122, 277, 313. 468, 479, 555, 558,
562, 570
Air Stripping, 89-313
694 KEY WORD/SUBJECT INDEX
-------
Contamination, 89-558
Groundwater, 89-519
Purgeable, 88-174
Total, 88-174
Volatile
Nitrogen Compounds Monitoring,
83-100
Organics
Analysis, 87-85; 89-15
Chlorinated, 88-164
Emissions, 81-129; 84-68, 77
Foam Suppression, 87-480
Lower Detection Limits, 87-280
Monitoring,* 81-122; 84-72
Removal, 87-218
Sampling, 87-457
Screening, 86-386
Soil Gas Survey, 87-523
Stripping From Soils, 86-322
Volatilization, 88^67
Volume Estimation, 88-274
VOST, 87-457
Wales, 84-594
Walls
Design and Installation, 86^160
Gelatinous, 82-198
Slurry, 82-191
Washing, 89-198, 207
Waste
Management Facilities
Real Estate Transfer, 87^99
Minimization, 89-13, 606
Oil Recoveiy, 87-179
Radioactive, 88-193
Storage
Above Ground, 82-228
Geologic Repositories, 87-502
Wastewater
Disposal Ponds, 88-84
Treatment, 80-160; 84-598
Water Treatment
Cost, 83-370
On-Site, 87-169
Waterway Decontamination, 83-21
Well
Abandonment, 87-439
Monitoring, 88-202
Well Field Contamination, 87-320
Well-Point Systems Evaluation, 87-228
West Germany, 83-68
West Valley Demonstration Project,
87-405
Western Processing Site,87-78, 198;
89-645
WET Procedure, 86-303
Wetland, 88^35
Assessment Procedure, 87-431
Contamination, 85-261
Treatment, 88-261
Wilsonville Exhumation, 82-156
Winter Flounder, 88-359
Woburn, MA, 81-63, 177
Wood Treating, 88-226
Facility, 81-212
PAH, 86-242
Plant Bioremediation, 87-193
RI/FS, 86-441
X-Ray
Analyzer, 85-107
Fluorescence, 85-93; 86-115
Spectroscopy (XRF), 88-97
Xylene,
m-Xylene, 88-451
o-Xylene, 88-451
p-Xylene, 88-451
Zinc, 86-200; 89-430
KEY WORD/SUBJECT INDEX
695
-------