-------
118 SCREENING
Table HI.
Organic Composition in Natural Waters as Determined by
the Pyrographic Method
Mountain Streams
Coweeta
Foothill Streams
Alcovy R.
Shoal Cr.
Coastal Plains
Rivers
Little R.
Mills Cr.
Swamps
Everglades
Moving Streams
Everglades Ponds
Total
Organic
Load (mg/l)
3.7
5.2
5.8
17.5
16.6
24.7
33.4
Organic Class
Concentration, mg/l
Carbohy- Proteins Lipids
drales
2.2
3.4
4.7
11.7
11.2
19.7
27.9
1.4
1.6
1.0
5.1
4.8
3.7
3.9
0.1
0.2
0.1
0.7
0.6
1.3
1.6
were supplied by 20 major industrial operations in Ala-
bama, Florida, Georgia, Mississippi and South Carolina
including: (1) Hercules, Inc., Hattiesburg, Mississippi; (2)
Air Products and Chemicals, Inc., Escambia Plant, Pen-
sacola, Florida; (3) Merck and Co., Inc., Medicinal Chem-
ical Plant, Albany, Georgia; (4) Monsanto Company
Plant, Pensacola, Florida; and (5) Delta Airlines, Inc.,
Atlanta, Georgia. The industry sources can be classified
as chemical, food, textile, paper and pulp, oil refinery,
brewing, pharmaceutical and aircraft maintenance.
Samples of industrial waste discharges were collected at
the points of discharge into receiving waters and were pre-
served by freezing in dry ice. The collected samples were
analyzed pyrographically under standard conditions.
The calibration constants for 19 industrial waste
effluents (one plant reclaimed its waste, and could not
technically be considered as a waste discharger) were ex-
amined with the object of determining the differences
between pyrographic patterns. The data are reported in
Table IV. A few representative patterns are presented
graphically in Figure 3.
Examination of reported data reveals that pyrographic
patterns of industrial waste are indeed different from pyro-
graphic patterns of natural unpolluted waters and hence
can be determined and differentiated from natural organic
background.
DETECTION OF WATER CONTAMINATION
Liquid and solid chemical wastes are usually either dis-
posed of in landfills, in open pits (impoundments) or
stored in 55-gallon drums. Contaminants enter ground-
waters from impoundments as a result of ground penetra-
tion. They can also enter surface waters as a result of run-
off. Chemical wastes stored in 55-gallon drums may leak
out with the passage of time and enter surface waters.
Chemicals that enter natural, unpolluted ground and/or
surface waters alter gross their chemical composition. In
addition to carbohydrates, proteins and lipids which are
found in natural waters, man-made refractory and per-
sistent chemicals appear. Due to dilution by receiving
waters, such man-made contaminants or pollutants are us-
ually found in trace quantities and in many instances could
be present in concentrations lower than that of organic
compounds present naturally in water.
The task is then to detect and quantify trace amounts of
contaminants in the presence of natural organic back-
ground of water, under conditions when natural organic
background of water can be greater than that of the con-
taminants. Hence, there could be two distinct and differ-
ent situations:
•The nature of hazardous waste contamination is un-
known.
•The nature of hazardous waste contamination is known.
Depending on the situation, two different approaches
to detection and quantification of water contamination by
the hazardous waste can be implemented.
Detection of Unknown Contaminants
In many instances the exact nature of chemicals disposed
of at a given hazardous waste site is unknown and exten-
sive trace analytical work must be undertaken to establish
whether surrounding water sources are contaminated or
not.
Using the pyrographic method of analysis, it is possible,
however, to develop an empirical pattern, visual examina-
tion of which will indicate whether or not contamination
of ground or surface water supplies took place. As shown
in Table II and graphically displayed in Figure 2, pyro-
graphic representations of unpolluted natural waters are
distinct and characteristic. In cases where water is polluted
by man-made chemicals, pyrographic patterns will become
.1. .
Figure 3.
Pyrographic Patterns of Industrial Wastewaters
-------
SCREENING 119
Table IV.
Pyrographic Fingerprints of Industrial Waste Effluents
(Peak Magnitude x KH)
Food and Beveraoe
Average
Peak Retention
No. Time
1 1.9
2 2.2
3 3.3
4 4.0
5 6.5
6 14.0
Various Industrial
Average
Peak Retention
No. Time
1 1.9
2 2.2
3 3.3
4 4.0
5 6.5
6 14.0
Textiles
Average
Peak Retention
No. Time
1 1.9
2 2.2
3 3.3
4 4.0
5 6.5
6 14.0
Chemical Industrial
Average
Peak Retention
No. Time
1 1.9
2 2.2
3 3.3
4 4.0
5 6.5
6 14.0
Food-A
67.9
2.1
13.3
5.9
0.7
3.8
Paper
and
Pulp
96.6
4.5
32.9
25.0
5.1
29.0
Textile
A
607.0
110.9
576.5
569.2
95.4
156.1
Chemical
A
38.2
4.2
84.3
39.4
10.6
6.5
Food-B
22.5
1.4
31.6
12.0
2.8
8.4
Oil
Food-C
342.1
29.5
147.0
144.6
21.2
61.0
Refining Pharmaceutical
34.6
2.2
46.5
20.9
6.2
16.8
Textile Textile
B C
669.6 2,289.5
83.9 555.2
293.4 3,213.5
152.8 2,364.0
279.0 392.8
117.9 1,252.0
Chemical Chemical
B C
85.9 4008.0
7.2 328.0
75.1 330.5
35.4 1444.7
5.8 27.1
38.8 327.0
2320.7
115.7
87.8
19.5
60.5
Textile
D
1,229.0
328.4
2,050.0
1,571.0
210.1
565.4
Chemical
D
49.6
5.0
42.7
26.2
2.2
12.3
Brewing
248.5
39.4
129.9
110.8
15.8
54.4
Aircraft
Maintenance
237.7
17.6
324.6
122.9
7.6
25.4
Tpxt.il e
E
681.5
102.2
363.1
228.2
36.5
171.5
Chemical Chemical
E F
11.7 45.2
2.4 2.3
8.7 32.6
3.6 31.2
1.1 1.6
3.7 14.7
distorted and the ratio and magnitude of peaks will
change.
In practice, it should be possible to match visually a py-
rogram of a water sample that is suspected to be polluted
by hazardous waste with a series of pyrograms typical of
known types of natural waters. Pyrographic patterns
shown in Figure 2 are quite representative of the majority
of types of unpolluted natural water found in the coun-
try.
If no pollution (intrusion of hazardous waste) took place
in ground or surface water, the pyrographic pattern devel-
oped will match one of the types of natural, unpolluted
waters, as shown in Figure 2. If ground or surface waters
have been contaminated by man-made hazardous wastes,
then the pyrographic patterns or representations of water
sample will be distorted and the ratios and magnitudes
of the peaks will be different from those of natural un-
polluted waters.
-------
120 SCREENING
A visual examination of pyrographic patterns provides
the simplest form of detection of the presence of haz-
ardous waste (or other man-made pollutant) in the natural
waters. The magnitude of contamination can be estimated
visually from a degree of pyrogram distortion or computed
mathematically as described below.
The pyrographic instrument can be calibrated for detec-
tion of principal natural water organics: carbohydrates,
proteins and lipids. By solving a series of simultaneous
linear equations, the concentration of each class of com-
pound can be determined. The computation logic is also
provided with a statistical measure of the quality of such
computations: these are "residual value" and "percent
fit."
The residual value reflects the quantity of information
that could not be fitted into the calibrated model and "per-
cent fit" reflects a degree of agreement between calibra-
tion factors and observed, unknown data. "Low residual
values" and high "percent fits" indicate that calibration
factors and analyzed substances match, and no other than
calibrated compounds are present in the sample. In case of
natural water, it means that only carbohydrates, proteins
and lipids are found.
High "residual values" and low "percent fits" indicate
that materials other than natural organics (carbohydrates,
proteins and lipids) are present in the sample. The higher
is "residual value" and the lower is "percent fit," greater
is the probability of contamination of water.
In case of a natural water the instrumentation system
is calibrated for carbohydrates, proteins and lipids using
known chemicals: glucose, albumin, and oleic acid. An ex-
ample pyrographic analysis of two synthetic solutions of
standard solutions is shown in Table V. Here one can see
that observed and actual values for each compound are
close, "percent fit" is high, and "residuals" are low.
Thus, no other than natural organics are present in this
sample. In practice, when "percent fit" falls below 75%
and "residual value" increases over 1000, then one may as-
sume that the water sample is contaminated by other than
natural organic matter.
Table V.
Pyrographic Analysis of Synthetic Solutions
Run*
Cbemlcml Concentration (mg/l)
Olek
Albumin Glucose Acid Relldtlll
Percent Fit
e
i
r 2
i 3
S Avg Found
" Actual
s
e
r
e
s
Presenl
1
2
3
A\g. Found
9.69
7.12
8.04
8.28
10
19.14
18.69
19.52
19 12
22.47
20.88
19.44
20.93
20
9.3
9.41
9 59
943
3.17
3.4
3.3
3.29
3
.92
99
.97
.96
117.449 98
343.97 95
140.839 99
128.886 %
149.194 96
152.734 96
Detection of Known Contaminants
The nature of chemical waste present at a given haz-
ardous waste site might be known and a sample of waste
matter might be available for calibration. In such cases,
an accurate determination of the natural organics present
in the water and a calculation of amounts of hazardous
waste can be made. The instrumental system is calibrated
in the usual manner for carbohydrates, proteins and lipids,
but is additionally calibrated for one or more suspected
hazardous waste compositions. After the analysis is per-
formed, the concentrations of natural organic and contam-
inants are computed as shown in Table VI.
Table VI.
Analysis of Mixed Industrial Wastes in Natural Water Matrix
Natural Background
Proteins
Carbohydrates
Lipids
Waste Source Run:
Chemical
Textile Finishing
Chemical
Textile Finishing
1.22 mg/l
7.47 mg/l
.46 mg/l
Volume Percent From Each Source
2 3
0.48 0.42 .44 .46
0.66 0.66 .87 .91
Conc'n. of Organic Carbon
From Each Source (mg/l)
.5
Average Actuil
0.46 0.5
.98 0.86
1.0
0.91
2.85 2.67
0.86
2.7
0.87
2.82
0.95 0.87
3.84 2.67
Here the technician has used a sample of natural water
from the Okonee River, Georgia, contaminated with two
types of industrial waste. The natural organic background
of the water was 9.15 mg carbon/1, to which small quan-
tities of chemical and textile waste were added (equal to
approximately 3.5 mg carbon/1). The man-made contam-
inants represented approximately one-fourth of the organ-
ic matter present in the water, a situation likely to be found
in real life. The water sample was analyzed five times and
results are shown in Table VI. The computer analysis of
the pyrogram revealed its natural organic composition and
identified and quantified industrial waste present. This ex-
ample illustrates that hazardous organic wasle present in
quantities smaller than natural organic background of
water can be detected and accurately measured.
CONCLUSIONS
An instrumental method is described for rapid screening
of water samples suspected of being contaminated by haz-
ardous wastes. The method can be used for analysis of
water samples contaminated by unknown organic pollu-
tants or for the accurate quantification of pollutants, tM
nature of which is known. The technique is designed for
field use and can be applied for empirical characterization
of the sample or for accurate quantitative chemical an-
alysis. The described procedure is fast (15 to 20 minutes)
and utilizes water samples as received without pretreat-
ment or other preparations.
-------
SCREENING 121
ACKNOWLEDGMENT
The material in this paper was extracted in part from the
EPA reports references (Nos. 7 and 9) (Contract 14-12-
802). The authors express our gratitude to Dr. H. Page
Nicholson, EPA Athens, Ga., for his support and en-
couragement in this work.
REFERENCES
1. Proceedings, U.S. EPA National Conference on Man-
agement of Uncontrolled Hazardous Waste Sites.,
October 15-17, 1980. Washington, DC. Hazardous
Materials Control Research Institute, Silver Spring,
Md., 1980.
2. Ibid, p. 45.
3. Birge, E.D. and Juday, C. Ecological Monographs 4,
No. 9, 1930, 63-80.
4. Lysyj, I., "Pyrographic Analysis of Waste Waters,"
Environmental Science and Technology, 8, 1974, 31.
5. Lysyj, I, "A Pyrographic Instrument for Analysis of
Water Pollutants. "American Laboratory. July 1971.
6. Lysyj, I., Nelson, K.H. and Webb, S.R. "Determina-
ation of Multicomponent Organic Composition in
Aqueous Media," Water Research, 4, 1970; 157-163.
7. Pyrographic Gross Characterization of Water Contam-
inants. "U.S. EPA Report R-2-72-227. Office of Re-
search and Monitoring, U.S. Environmental Proltection
Agency. U.S. Government Printing Office, Wash-
ington, DC 20402.
8. Lysyj, I, "Pyrography - A New Hydrochemical Tool,"
International Symposium on Identification and Meas-
urement of Environmental Pollutants. Ottawa, Canada
(1971).
9. "Pyrographic Gross Characterization of Water Con-
taminants. Interim Report prepared for the EPA under
Contract 14-12-802 (1970).
-------
AMBIENT MONITORING FOR SPECIFIC VOLATILE
ORGANICS USING A SENSITIVE PORTABLE PID GC
T.M. SPITTLER
ALAN W. OI
U.S. Environmental Protection Agency
Region I
Lexington, Massachusetts
ABSTRACT
With the intensifying concern over organic chemical
contamination at spills, dump-sites and chemical manu-
facturing facilities, a need has emerged for rapid assess-
ment of ambient air pollution by volatile chemicals. We
have recently tested a portable photoionization detector
(PID) gas chromatograph, in several field incidents. The
unit (Photovac 10A10, Thornhill, Ont.) is lightweight,
completely portable, and requires only compressed air for
operation. It has a measured sensitivity in the .1 to 10 ppb
range for a wide variety of commonly occurring volatile
organic contaminants such as vinyl chloride, benzene,
chlorinated methanes, ethanes and ethylenes and even low
volatility compounds like nitrobenzene and xylenes. A
characteristic of the PID system is enhanced sensitivity to
aromatic and unsaturated molecules over the FID detector.
However, the Photovac PID has an uncharacteristic sen-
sitivity to halogenated methanes and ethanes. This unusual
and theoretically unexpected sensitivity (e.g., CHC13 can
be detected at the 5-10 ppb level and 1, 2-di chloroethane
at the 1-5 ppb level in l-5c.c.air samples) makes the in-
strument even more useful for general air pollution mon-
itoring. There is no need to have different detectors (such
as Electron Capture) in order to determine all the volatile
haloalkenes and other alkenes of interest.
The paper describes several incidents involving specific
volatile organics.
(1) We used the instrument to measure vinyl chloride in
the vicinity of a plastics manufacturer. Levels of VCM
below 0.2 ppb were easily detected and semi-quanti-
tated in the field.
(2) Nitrobenzene was monitored in the vicinity of a ten-
year-old spill site which was being dredged from a river
bottom. Maximum ambient contaminant levels never
exceeded 20 ppb of nitrobenzene 100-300 feet down-
wind of the dredge operation. Organic quantitation
and calibration were conducted in the field.
(3) Acrolein was monitored for an OSHA study against a
background of 15 or more solvent peaks in another
manufacturing facility. The OSHA safety level of 0.1
ppm was never exceeded and sensitivity was found to
exceed 30 ppb even in the presence of interfering sol-
vent peaks.
In the above studies, and several others that will be dis-
cussed, a key feature is the ability to make in-field meas-
urements. Techniques for field calibration will be dis-
cussed in detail. One promising technique is the use of the
two-column option which has two distinct field use bene-
fits.
a. A short column can be used to enable rapid 'screen-
ing' type analysis with very short retention times at am-
bient temperature (e.g., 4 min. for detection of nitro-
benzene).
b. Two-column referencing of known retention times.
This technique of using a second column (and field stand-
ards) to identify retention times, provides a high level of
confidence in the positive identification of unknown am-
bient contaminants before one collects time-integrated
field samples for laboratory analysis using GC/MS as an
identification tool.
The combination of field screening and measurements
on a portacle GC, with integrated field samples and
follow-up GC/MS analysis, provides a powerful tool to
settle questions of ambient organic contamination. Ow
technique for doing such work will be discussed in de-
tail.
122
-------
MEASUREMENT OF FUGITIVE HYDROCARBON EMISSIONS
FROM A CHEMICAL WASTE DISPOSAL SITE
JAMES A. PETERS
KEITH M. TACKETT
EDWARD C. EIMUTIS
Monsanto Research Corporation
Dayton, Ohio
INTRODUCTION
One of the primary concerns is air quality near chem-
ical waste disposal sites, especially existing chemical land-
fills that were not constructed and operated according to
criteria now required by the Resource Conservation and
Recovery Act (RCRA). Since hydrocarbon emissions from
landfills and disposal sites are fugitive emissions, they are
difficult to monitor. Quantifying the emission rates of air
contaminants from landfills has not been practical because
of the complexity and costs. The transfer behavior of
organic pollutants between the solid-air interface has been
studied in detail for pesticide volatilization'1' and theo-
retical estimation techniques have been reported for the air
pollution aspects of volatilized organic compounds from
waste disposal sites.(2> 3> 4) However, at present the extent
of air contamination from chemical waste disposal sites
remains largely unknown.
In this paper, the authors describe a technique that can
be used to measure air quality close to waste disposal sites
and to calculate the emission rate of a fugitive emission
area source (open source) at minimum cost.
SOURCE DESCRIPTION
The Hooker Chemicals and Plastics Corporation is cur-
rently building a clay-lined waste containment facility at a
chlorine production complex in Montague, Michigan, to
contain chemical residues and contaminated soil from
various locations at this site. The Michigan Department of
Natural Resources (DNR) established several air qual-
ity conditions to be met during and after the transfer
period. Before beginning restoration activities, it was
necessary to determine background ambient concentra-
tions and emission levels of the perchlorinated organic
compounds of concern. The DNR had chosen hexa-
chlorocyclopentadiene (C-56) to represent those com-
pounds.
C-56 is a nonflammable liquid with a very pungent odor;
the odor threshold is as low as 0.33 ppb; some individ-
uals can detect as little as 0.15ppb.<5) It is a key intermed-
iate in the production of the c>dodiene group of chlor-
inated pesticides and flame retardants. Unlike some of the
pesticides derived from it, C-56 degrades rapidly by photo-
lysis, giving water-soluble degradation products. Tests on
its stability towards hydrolysis at ambient temperature in-
dicated a half life of about 11 days at pH 3-6, which was
reduced to 6 days at pH 9. The present TLV for industrial
exposure is 0.01 ppm, or about 7% of the lowest vapor
concentration shown to produce chronic toxic effects in
laboratory animals.(5)
Two inactive sites contained C-56 wastes that were to be
relocated to the new waste containment facility. The first
site is a roughly triangular above-grade dump partially
covered with soil, varying in height from 0 to 15 ft above
a flat ground surface. Trees were located on two opposite
sides of the site. The second site was an abandoned pro-
duction facility for chlorinated hydrocarbons, approx-
imately 200 ft square with a brine lake to the north. The
remaining areas surrounding the site were open.
MEASUREMENTS AND METHODS
Measurements of pollutant concentration and meteoro-
logical variables were made at the waste disposal sites in
June 1980. Descriptions of the equipment and methodol-
ogy follow.
Open Source Sampling
The major difference between open sources and point
sources is that open sources do not have a definable point
of emissions. There is no established protocol for sampling
open sources; the methods employed depend upon the type
of information needed, the physical configuration of the
source, meteorological conditions at the time of sampling,
and interferences due to other point and fugitive sources.
The following methods have been developed and were con-
sidered for sampling fugitive and open sources: 1) quasi-
stack approach, 2) grid/flux approach and 3) upwind/
downwind approach.(6> The quasi-stack approach was
unsuitable because the two sites were not isolable and the
grid/flux approach was eliminated because the disposal
site was heterogeneously soil-covered and the abandoned
production facility was not amenable to gridding.
The upwind/downwind approach is generally favored
because it preserves the integrity of the source and any
process associated with it. It also eliminates any construc-
tion to capture emissions, which may prove expensive.
Continuity of the emissions is generally of secondary im-
portance since the magnitude of the ambient air volume
into which the emissions are dispersed is large enough to
123
-------
124 AIR MONITORING
provide a degree of smoothing to cyclic or intermittent
emissions. Best of all, the upwind and downwind measured
ambient concentrations truly reflect the actual pollutant
levels at the sampler locations.
Meteorological Variables and
Atmospheric Stability
Dispersion models, normally used to predict con-
centrations surrounding a point source of known emis-
sion rate, are used in reverse for open source sampling
calculations. Several downwind concentration readings
are taken to calculate the source emission rate (after sub-
traction of the upwind or background concentration).
Because of the relative ease of application, the Gaussian
plume model was used in conjunction with Turner's(7) de-
termination of the dispersion coefficients ay and az, the
crosswind and vertical standard deviations of plume
spread.
Comparison of predicted and measured pollutant con-
centrations from point sources indicate that, with suffic-
ient meteorological information, the predictions of the
Gaussian plume model can be generally within a factor of
two of measured values. The accuracy achieved using the
Gaussian plume model depends to a great extent on the
accuracy of predicting or measuring ay and az
The detailed meteorological measurements required for
the most accurate determinations of ay and az are gen-
erally not available for inexpensive, rapid sampling and
analysis projects. The Pasquill-Gifford determinations of
cry and az as given by Turner, where ay and a are
functions of the distance from source and an atmosphere
stability parameter only, are strictly applicable only to dis-
persion over relatively smooth surfaces, which fortunately
was the case in this study.
Pasquill-Gifford atmospheric stability categories were
determined using the protocol described in Figure 1. The
upwind-downwind method is strongly influenced by
meteorological conditions, requiring a wind consistent in
direction and velocity throughout the sampling period.
For this reason upwind-downwind sampling is futile dur-
ing stable air (stability classes E and F) and very unstable
air (stability class A). Typically then, weather with strong
frontal activity and conditions conducive to inversions
are avoided, and upwind-downwind sampling is best prac-
ticed between 9 a.m. and 4 p.m. By confining the sampling
to periods of moderately unstable to neutral conditions,
DlOltTIOH INDD :
•t
Figure 1.
Flow Chart for Determining Atmospheric Stability Class.
-------
AIR MONITORING 125
the accuracy of the Gaussian plume model may be en-
hanced.
Given the stability class and the downwind distance of
the sampler from the source, continuous functions are then
used to calculate values for ay and CTZ. (8)
Wind speed and direction were monitored continuously
with a Climatronics model EWS portable weather station
with a 10-ft tower. An example of its output is shown later
in Figure 4.
Sampling and Analysis of C-56
Pollutant concentrations of C-56 were measured by
collecting air samples in a packed porous polymer tube
at a height of about four feet above ground level. Eight
samplers were used simultaneously; one sampling posi-
tion was upwind from the source and seven were position-
ed in an array downwind from the source.
The C-56 sampling and analysis procedure followed
NIOSH Method No. P&CAM 308 for hexachlorocyclo-
pentadiene in air.(8) In this procedure a known volume of
air is drawn through a tube containing Porapak T (an
ethyleneglycodimethylacrylate monomer), followed by in-
field, ultrasonic-assisted desorption with hexane, subse-
quent gas chromatographic analysis with a "Ni electron
capture detector and comparison of peak heights with
those obtained from injection of standards.
The sorbent tube used for collection of C-56 in air is
shown in Figure 2. The pumps used to pull air through the
sorbent tubes were MSA Model G personnel pumps, which
are rechargeable-battery-operated, diaphragm-type
with a sample flow indicator and a sample rate control
valve. They were operated for 90-113 minutes at a rate of
1.6-2.0 1/min. During sampling, the flowmeter float
position of each pump was checked periodically to ensure
that the proper flow rate was maintained. The combined
sampling/analysis detection limit is 0.001 ppb.
During the C-56 sampling period, sulfur hexafluoride
(SF6) tracer gas was released at two known rates at an up-
wind location that approximated the virtual point of
emission for the area source being sampled, based on the
atmospheric stability class and the width of the area
source. Ground level release rates of 100% SF6 were mea-
sured with a soap bubble flowmeter.
Downwind samples were collected in Mylar bags in a
sealed metal container. To collect an air sample, the bag
was opened to the atmosphere and a pump connected to
the sealed container. The pump, operating at about 3
1/min, created a light vacuum which pulled the ambient
SF6 sample into the bag. The sampling duration was
about three minutes per sample.
SF6 analysis was conducted in the field with an AID
Model 511 portable gas chromatograph equipped with a
6 ft x 1/8 in. Molecular Sieve 5A column operable at
104°F and an electron capture detector. Analysis followed
a procedure supplied by the equipment manufacturer00'
Standards of SF6 were prepared in the field by dilution of
a 1.00 ppm SF6 standard furnished by a commercial sup-
plier. The combined sampling/analysis detection limit was
0.02 ppb.
PLASTIC CAP
SILYLATED GLASS WOOL
75 mg PORAPAK T
25 mg PORAPAK T
4 mm i.d./6 mm o.d.
TEFLON TAPE
PLASTIC CAP
Figure 2.
Sorbent Tube for Collection of Hexachlorocyclopentadiene
(C-56) in Air.
-------
126 AIR MONITORING
DATA ANALYSIS AND RESULTS
In this section, the discussion focuses on the use of the
Gaussian plume model for area sources in combination
with the ratio equation technique for estimating the emis-
sion rate of an inactive chemical waste disposal site.
Plume Dispersion Model
For a continuous point source, the Gaussian plume dis-
persion model is a rather simple form of calculation. Be-
cause the source under study was at ground level with lit-
tle or no plume rise (H = 0), and since ground level con-
centrations were taken (z = 0), the model was simplified
to:
X =
nayazU
Where X is the mean concentration at the point of mea-
surement; y is the distance from the sampler to the plume
centerline; Q is the source emission rate; u is the mean
horizontal wind speed at the height of emission; and cry
and crz are the Gaussian standard deviations in the pollu-
tant concentration distribution in the horizontal and ver-
tical, respectively, at the point of measurement, and are
both functions of the distance from the source (X).
For fugitive source sampling the concentrations are mea-
sured and the emission rate Q is of interest. The model
then becomes:
Q =
a u exp [~
When sampling can be done along the plume center-
line (y = 0), the model is further simplified:
Q = xnayazu
In dealing with dispersion of pollutants from a defined
area rather than a point of emission, an approximation
can be made by treating the area as a source having an in-
itial horizontal standard deviation of plume spread, a .
A virtual distance \j. can then be found that will give this
horizontal dispersion coefficient, depending upon stability
class. Then equations for a point source may be used, de-
termining av as a function of x + x^. How this is done
is shown in Figure 3.
In this procedure, the area source is treated as a cross-
wind line source with an initial normal distribution, a
fairly good approximation for the distribution across an
area source. The initial standard deviation for a roughly
square area source can be approximated by a = S/4.3,
where S is the length of a side of the area.17' For a ground-
level area source, the vertical dispersion coefficient a is
left untreated as having originated from the actual source
PUMK
STATION
Figure 3.
Schematic Diagram of the Gaussian Point Source Plume Model
Applied to an Area Source.
rather than the virtual point.
The dispersion model then becomes:
Q =
yo
*>1/2a.
u
In this sampling program the area source dispersion
model concept was used only to find the virtual point of
emission in order to site the SF6 tracer gas release point.
By knowing the disposal site side length S and the atmos-
pheric stability class, a distance xy upwind from the dis-
posal site edge was found at which to release the tracer
gas and allow the SF6 plume to travel across the area
source to be measured at several downwind locations.
During C-56 sampling, two grab samples were collected
at each of three downwind plume centerline locations
for on-site analysis of SF6. Emission rates of C-56 were
then calculated by a ratio equation technique:
, = X(C-S6)
JC-56 x(SF6)
X ESF6 X 24.45
_
P T
Where EC.J6 is the emission rate of C-56 in g/hr; x(C-56)
is the concentration of C-56 at the sample point; x(SF«) "
the concentration of SF, in 1/hr, MWC.J6 is the molecular
weight of C-56 (212.11 f; and 24.45 is the molar volume of
a gas at STP in 1 /mole.
The measured concentrations of C-56 and SF6 should be
a mean concentration over the same time interval as the
time interval for which the a's and /* are representative,
typically three minutes. An often-forgotten calculation in
plume dispersion modeling is to correct for time-averaging
differences due to increased meander of wind direction
with time. When estimating emission rates (or concentra-
tions) from a single source for the time intervals greater
than a few minutes; the best estimate apparently can be ob-
tained from07':
xs =
Where xs is the desired concentration estimate for the
sampling time ts; Xk is the concentration estimate for the
-------
AIR MONITORING 127
Table I.
Calculation of C-56 Emission Rate by SF6 Tracer Gas Ratio
Method, Run 1, Abandoned Production Area.
Corrected
C-56
Sample concentration,
point ppb
SFe concentration, ppb
97.0 cc/min,94.0 cc/min,
SF» release SF» release
rate rate
C-56 emission rate, g/hr
97.0 cc/min,94.0 cc/min,
SF« release SFS release
rate rate
S*
s,
s.
0.053
0.047
0.032
0.71
1.13
0.75
0.46
0.57
0.30
0.35
0.20
0.20
0.31
0.22
0.27
Average C-56 emission rate » 0.26 g/hr
Confidence limits (95%) • 10.05
shorter sampling time tk; and p should be between 0.17
and 0.2. Equation (6) is most appropriately applied to
sampling times less than two hours.
Ground Level Concentrations
An example of the average ground-level concentrations
of C-56 and SF6 tracer gas taken during one of the emis-
sion test runs is given in Table I. The form of the meteor-
ological data collected during each run is shown in Figure
4. When the wind direction is shown to be steady (< 45 °
shift) over the sampling period, only the plume centerline
samplers need be used in estimating the emission rate of
the source. Of course, the precision of the calculated emis-
sion rate usually will improve with less wind direction
meandering. The ratio of predicted to observed concentra-
tions of SF6 ranged from 0.4 to 0.8 for all runs, which were
conducted under stability class C.
The estimated emission rate of C-56 shown in Table I
applies only to the ambient conditions experienced during
the sampling period. Any factor that could change the
volatilization rate, such as increased temperature or distur-
bance of the soil cover, will cause a corresponding change
in emission rates from a chemical waste disposal site.
CONCLUSIONS
The results of the calculated emission rates from
sampling chemical waste disposal area sources indicate
that a tracer gas-ratio equation technique can estimate
emission rates with good precision, provided a few open
source sampling rules-of-thumb are followed. Seasonal or
weather-dependent trends on the emission rate can be
monitored with relative ease. The technique as described is
fully portable and can be assembled rapidly. Knowledge of
gas chromatograph operation in the field is the only diffi-
cult skill necessary. Two people can make two or three
runs per day, weather permitting.
Regardless of how inaccurate the emission rate estimata-
tion is, the ambient concentrations measured are real
values and can have an accuracy ascribed to them. The
concentration of C-56 in test atmospheres was determined
by NIOSH in control experiments by sampling with bub-
blers containing hexane and subsequently analyzing the
bubbler solutions by GC/ECD. The determinations with
Porapak T sorbent sampling gave values averaging 100%
of those found by bubbler sampling over the range of the
method. Sorbent sampling for C-56 has a precision of 8%
over the concentration range of 13-865 ng/m}; the pre-
cision of the analytical method is 3 %.
While the collection and analysis method described here-
in is specific for C-56, methods are also available to quan-
tify emission rates of many organic chemicals simultan-
eously using the tracer gas/area source technique. Low
concentrations of a wide variety of pollutants are expected
at chemical landfills and sorbent trap sampling is consid-
ered to be more reliable, accurate, precise, and sensitive
than grab sampling.0'112)
Monsanto Research Corporation has developed a com-
bination sorbent system based on Tenax-GC, Porapak R
and Ambersorb XE-340 for use in collecting a broad range
of organic compounds.03' Over 30 different sorbent ma-
terials were evaluated in this EPA-sponsored program
(Contract No. 68-02-2774), whose goal was to choose a
collection of commercially available materials that, when
used in combination, showed promise for quantitatively
trapping compounds varying widely in volatilities and
polarities.
360
315
210
225
180
125
90
45
0
1:00 1:30 2:00 2:30 3:00 3:30 4:00 4:30 5:00
TIME Of DAY (PMI
5:30 6:00
Figure 4.
Meteorological Data for Run 1, Abandoned Production Area.
-------
128 AIR MONITORING
Tenax-GC is the only high-temperature (750 °F) adsor-
bent (2, 6-diphenyl-p-phenylene oxide) available which
allows the quantitative thermal desorption of low volatility
organic compounds. Porapak R is one of the highest-ca-
pacity polymeric adsorbents (n-vinyl pyrrolidone) with a
reasonable background level and with an overlap in the
range of utility with Tenax-GC.
Ambersorb XE-340 (a carbonized styrenedivinyl ben-
zene) gives less difficulty with the desorption of com-
pounds of intermediate volatility, fewer detrimental effects
by water, and less reactivity with collected samples than
with charcoal adsorbents. Also, its range of utility leaves
the smallest gap between polymeric and carbonaceous ad-
sorbents in the types of compounds collected. These sor-
bent materials are packed into three separate tapered glass
tubes for sampling, with flow being directed either in series
or in parallel. If sampling is done with the sorbent tubes
in series, the arrangement is Tenax-GC at the air intake,
Porapak R in the middle and Ambersorb XE-340 last.
ACKNOWLEDGEMENTS
The authors are grateful to Hooker Chemical and Plas-
tics Corporation for the opportunity to present the results
of this work.
REFERENCES
1. Spencer, W.F. and Cliath, M.M., "Transfer of
Organic Pollutants Between the Solid-Air Interface,"
Fate of Pollutants in the Air and Water Environ-
ments, Part I, I.H. Saffet, ed., John Wiley & Sons,
Inc., New York, NY, 1977, 107.
2. Shen, T.T. and Tofflemire, T.J., "Air Pollution
Aspects of Land Disposal of Toxic Waste," In: Proc.
of the 1979 National Conference on Hazardous Ma-
terial Risk Assessment, Disposal and Management.
Information Transfer Inc., Silver Spring, MD, 1979,
153.
3. Shen, T.T., "Emission Estimation of Hazardous Or-
ganic Compounds from Waste Disposal Sites," Pre-
sented at the 73rd Annual Meeting of the Air Pollu-
tion Control Association, June 22-27, 1980, Montreal,
Quebec. Paper No. 80-68.8. 24 p.
4. Thibodeaux, L.J., "Estimating the Air Emissions of
Chemicals from Hazardous Waste Landfills," /. Haz-
ard. Mater., 4, 1981,235.
5. Bell, M.A., Ewing, R.A., and Lutz, G.A., "Reviews
of the Environmental Effects of Pollutants: XII Hex-
achlorocyclopentadiene," U.S. EPA, Health Effects
Research Laboratory, EPA-600/1-78-047, December
1978. 94 p.
6. Kolnsberg, H.J., "Technical Manual for Measure-
ment of Fugitive Emissions: Upwind/Downwind
Sampling Method for Industrial Emissions," U.S.
EPA, Industrial Environmental Research Laboratory
EPA-600/2-76-089a, April 1976. 75 p.
7. Turner, D.B., "Workbook of Atmospheric Disper-
sion Estimates," U.S. DHEW, National Air Pollution
Control Administration, EPA-AP-26, January 1974.
84 p.
8. Eimutis, E.G., and M.G. Konicek, "Deviations of
Continuous Functions of the Lateral and Vertical At-
mospheric Dispersion Coefficients," Atmos. En-
viron, 6, 1972, 859.
9. "NIOSH Manual of Analytical Methods, Volume 5,"
U.S. Department of Health, Education, and Wel-
fare, DHEW-NIOSH Publication No. 79-141, August
1979.
10. "Analysis of Trace SF6 in Air," Analytical Note (AN)
116B of Application Information Data, Analytical In-
strument Development, Inc.
11. Esposito, M.P., and S.M. Bromberg, "Fugitive Or-
ganic Emissions from Chemical Waste Dumpsites,"
Presented at the 74th Annual Meeting of the Air
Pollution Control Association, June 22-26, 1981,
Philadelphia, PA. Paper No. 81-41.2. 14 p.
12. Hwang, S.T., "Hazardous Air Emissions from Land
Disposal/Treatment Facilities," Presented at the
74th Annual Meeting of the Air Pollution Control
Association, June 22-26, 1981, Philadelphia, PA.
Paper No. 81-41.4. 12 p.
13. Brooks, J.J., D.S. West, D. J. David, and J.D. Mulik,
"A Combination Sorbent System for Broad Range
Organic Sampling in Air," Proc. of the Symposium
on the Development and Usage of Personal Monitors
for Exposure and Health Effect Studies, U.S. EPA,
Environmental Sciences Research Laboratory, EPA-
600/9-79-032, June 1971. 383.
-------
A STUDY OF THE EMISSION RATE OF
VOLATILE COMPOUNDS FROM LAGOONS
ANDREW T. McCORD
Director of Research
Recra Research Institute
Amherst, N.Y.
INTRODUCTION
Theory
Nusselt1 has described the rate of evaporation of water
from a lagoon as follows:
,0.2
. (CW-C0).
.(1)
G = kilograms of water evaporated per hour
Am = surface area of the lagoon in square meters
D = diffusivity coefficient of water in square meters per
hour
L = a dimension of the lagoon surface. For a circular la-
goon, L, in meters, is the diameter. For an irregular
shaped lagoon, such as square, oblong or elliptical,
etc., it is the diameter, in meters of a circle of area
equal to the lagoon surface. For long narrow lagoons,
L will be the smaller dimension in meters.
WQ = wind velocity in meters per second
Co = humidity content of the air at a distance above the
lagoon surface, or the humidity of the air blowing
across the lagoon.
g water
dry air containing the water
Nusselt's equation may be applied to the evaporation rate
of a volatile compound which is dissolved in the lagoon water
if one replaces Cw, the moisture content of the air at the
immediate surface of the lagoon by Cv, the volatile content of
the air at the immediate surface of the lagoon. Assuming that
there is no volatile compound in the incoming air, Nusselt's
C0 for the volatile compound is zero. Nusselt's equation for
a volatile compound dissolved in the lagoon water then be-
comes, in pounds per hour
GA = 39A
m
•(C2
(W0)°-78 • (Cv) • (2.205) (2)
Cv is derived as follows:
C, the molar concentration of a volatile compound in the
lagoon water
_%
Mv
100-% +%
Mw My
where % is the weight percent of the compound in the lagoon
water. Because this value is most frequently less than one, the
above equation can be reduced, with negligible error to:
C =
100
, M
M.
•(3)
Mw = molecular weight of water =18
Mv = molecular weight of compound = 28.96
The mole-fraction of a volatile compound P* at the lagoon
surface (in the air) is the product of the molar concentration
and the vapor pressure (V.P.) of the pure compound, in
atmospheres, at any temperature, i.e.,
P* = (C)(V.P.)
From equation (3) P* = I—^
(4)
(V.P.)
(5)
_ /Pounds of compound in the lagoon
_ /
\
also% =1' "p"" 7 "-•—---- "• "" -e—" )100
Pounds of waste in the lagoon /
-I^. 100
~ V 8.34 (sp.gr)
.(6)
L0 = Pounds of volatile compound in the lagoon
V = total gallons of waste in the lagoon
sp .gr = specific gravity of the waste
129
-------
130 AIR MONITORING
Combining equations (5) and (6):
, _ (LnXMw)-(V.P.)
V-(8.34MMv)(sp.grK
.(7)
Cv -
wt. of compound in the air at the lagoon surface
wt. of air containing this compound
_ (P*)MV
.(8)
(1-P*)MA
MA = molecular weight of air
Again, because the concentration of a volatile compound
in wastewater is usually less than 1%, P* is small compared
with unity so equation (8) can be reduced to:
Cv M
P*MV
Combining equations (7) and (8)
_ (Ln)(Mw)(V.P.) Jk
v (VX8.34)Mv(sp.gr) MA
= (Ln)(Mw)(V.P.)
v (VX8.34)MA(sp.gr) ' '
•(9)
Thus
M,,
IA 8.34
Therefore
0.07453 /
= 0.07453
E0 = Initial evaporation rate of volatile compound.
APPLICATION
In order to evaluate the emissions in real-life operation in
the lagoon, one assumes the following arbitrary conditions:
• The volume of waste in the lagoon is maintained
constant
• The feed volume and composition is constant
• The discharge volume is equal to the feed volume
• Temperature and wind velocity are constant
• Steady state condition.
Table I.
Symbols used in material balance
V = lagoon volume in U.S. gal
F = feed rate in gal/hr
= discharge rate in gal/hr
K = volatile compound in the feed, Ib/hr
L0 = amount of volatile compound in the lagoon at start,
i.e., time-zero, Ib
E0 = initial evaporation rate of a volatile compound in
Ib/hr as computed from equation (12)
Table II.
Material Balance Over a Lagoon for
"v (sp.gr) W/"'' vw One Volatile Component
The analogous Nusselt expression now becomes m ,
in Feed Disch Evap. Am t
Time Lagoon Rate Rate Rate in
r (*-'hr)-39\ /D\°'22 (0-07453) Lfl Hrs Ibs Ibs/hr Ibs/hr Ibs/hr Lagoon
(W0)°.78(2.205) (11) ° Lo
but L0 = (0.0834)(V)(%)(sp.gr)- - - -from equation (6) lst L°
substituting for L and combining the constants ^ 1
+K -K -EO =L]
+K -L,F -E0L! =1-2
0.22
Note, the sp.gr. of the waste is eliminated.
iO.78
•(12)
3rd
nth
L-
n-l
+K
+K
-O •
L
-------
AIR MONITORING 131
summarizing,
F F
let X = 1 "TT^f0, a constant because F, V, EQ and L0 are
constants. °
Then Ln = (L^) X + K (13)
Using this approach, outlined in Tables 1 and 2, one can
write:
Time from start Volatile Content of Lagoon
n = 0 Ln
n=lhr
n = 2 hrs
L0X + K
= L,
n = 3 hrs
n = n hrs
(L0X + K) X + K =
L0X2 + KX +K
(L0X2 + KX +K)X +K =
L0X3 + KX2 + KX + K
L0Xn + K + KX + KX2 +
KX3
= L
~ Ln
Because K + KX + KX2 + KX3 . + KXn-1 is a geometric
V
progression with X less than 1, the sum
1-X
.(14)
K
1-K"
Where n is very large, Xn becomes very small and so L0Xn
may be neglected.
Equation (14) becomes Ln = —-
K
1-X
.(15)
Equation 15, represents the steady state and Ln will be now
designated Ls
In arriving at a steady state, i.e., when n is many hours, the
following equations are used:
initial pounds of a volatile compound in a lagoon
L0 = (0.0834(V)(%)(sp.gr.) 16
pounds of a volatile compound in lagoon at a steady state
_ K pounds
s ~ 1-K
pounds of a volatile compound in feed to lagoon
K = 0.0834(F)(%)(sp.gr.)lb/hr
E F
1 - X =-r-2 + T73 fraction less than unity
evaporation rate of a volatile compound in 16/hr.
Ib/hr
E =
S L
percent of a volatile compound in a lagoon at the steady
rate
%s = (% at start)Ls
Initial evaporation rate of a volatile compound in a lagoon
is given in equation (12). D and V.P. are calculated for the
specific operating temperature.
EMISSIONS OF A VOLATILE COMPOUND IN AN
AERATED LAGOON
Consider the case of a lagoon containing less than 1% of a
volatile compound in water. At the center of the lagoon, a
water pump withdraws lagoon water and directs the stream
vertically against a deflector plate to produce a cone of drop-
lets. The maximum trajectory is attained when the initial
direction of the droplets is 45° from horizontal.
-R/4-
LAOOON SURFACE S~ U PUMP
'DEFLECTION
CONE
Vj = initial velocity of the droplets in ft/sec
R = maximum diameter of the spray falling back into the
lagoon
The horizontal and vertical components of the velocity
Vh and Vv of a droplet are Vj cos 45° and V0 sin 45° respec-
tively and are both equal to 0.7071Vj. The maximum height
-------
132 AIR MONITORING
attained by a droplet will be, theoretically, at a distance
D
— from the deflector plate.
Let t seconds be the time for a droplet to reach maximum
height, i.e., to reach the point A.
Then :
From Newton's Laws of Gravity:
e = time in seconds = 0.17678R^
TT = 3.1416
Fv = correction factor, which approaches unit for dilute
solutions of volatile compounds
L0 = amount of volatile compound in the lagoon. Ib.
VL = volume of liquid waste in the lagoon, U.S. gal
Mw = Mol. wt. of water = 18
.(16) Mv = Mol. wt. of volatile component
V.P.= vapor pressure in atmospheres of the volatile com-
pound at T°F sp.gr. = specific gravity of the liquid
waste
32
•0 7)
Combining equations (16) and (17) and letting Vh = Vv =
0.7071V,:
(18)
and, from equation (16):
t = 0.08839Rl/i (19)
The total flight time for a droplet = 2t sec:
2t = 0.17678R'/i (20)
For a spherical water droplet:
surface area _ 4 ir r-* _ 3
volume 4_77 r^ r
3~
H = water delivered by the pump comp.
or.(H)'n3785)=63.083Hcm3/sec
60
The surface area of this mass of water, as spherical drops, r
cm radius
= --63.083H
r
= 189.25-cm2 .
r
(21)
Arnold- has studied the diffusion of a volatile compound
from a liquid surface into air and arrived at this conclusion:
Substituting these values in equation (22) one obtains:
_2H /189.25V L^ . MW . (V.P.)
r V 8.347 VL Mv (sp.gr;
The emission in Ib/hr
= Q_ 3600 sec x Mv
0 e ' HR
LB.
22400 x(460+5) 454 gm
460+60
substituting for Q from equation (23), and simplifying:
H T rvP"» d 1
.(24)
=
= . . .(25)
.(22)
0 (460+T) r VL (sp.gr) TT V0.17678R"
When the value of E0 is obtained, Es may be calculated by the
procedure described previously.
NOMENCLATURE
Am c Surface area, m = m2; c = cm2
C Molar concentration of a compound in water
Cv Gms compound as gas divided by gms dry air
containing the compound
Cw Humidity content of air = gms water/gmw dry air
containing the water
D Diffusivity coefficient of a compound in air in
m2/hour
d Diffusivity coefficient of a compound in air in
cm2/second
E0 s Evaporation rate of a compound in pounds/hour;
o - initial, s = steady state computed from eqi*
tion (12)
F Feed rate, gal/hr = discharge rate (in a lagoon)
-------
AIR MONITORING 133
G,GA G = Kg/hr;GA = lb/hr
*H gpm
K Amount of a compound in a feed stream, Ib/hr
L A lagoon dimension in meters
Los Pounds of a compound in a lagoon; o=at state,
s=at steady state
MwvA Molecular weights of w = water, v = compound,
A = air
P* Mole-fraction of compounds in air at lagoon
surface
Q Gas volume, vol. of compounds emitted to air in
6 seconds, cm^/sec
sp.gr. Specific gravity of a liquid waste
V Operating volume of a lagoon in gal
V.P. Vapor pressure of pure compounds, atm
W0 Wind velocity m/sec
F F
X I-£.-±Q
V L
y* partial pressure in atmospheres of a compound in
air surrounding water droplets =
LO.MW.V.P.
IT
F,,
Weight percent of compounds in water or lagoon
liquid
Time in sec
3.1416
Arnold's correction factor, becomes 1 for dilute
systems.
VL Mv 8.34(sp.gr)
REFERENCES
1. Nusselt, W., Heat Transfer, Diffusion and Evapora-
tion, Technical Memorandum 1367, National Advisory
Committee for Aeronautics, 1975.
2. Arnold, J.H., "Studies in Diffusion. Ill Unsteady
State Vaporization and Adsorption," Trans. Am. Inst.
Chem. Engrs. 40, 1944, 361-379.
3. Schlessinger, G.G., "Vapor Pressures, Critical Tem-
peratures, and Critical Pressures of Organic Com-
pounds." Handbook of Chemistry and Physics, 54th
Ed. D-162toD-188.
4. Fuller, Atomic and Structural Diffusion Volume In-
crements. Perry 5th Edition (3-233) CRC.
Appendix 1.
Vapor Pressure of Selected Compounds Based on Reference (3)
OCPDB
Ident.No.
90
110
240
380
790
810
890
930
1120
1244
1660
1670
1990
2360
2500
2510
2620
2640
2860
2910
3230
3349
3395
3411
3430
3541
Compound
Acetone
Acetonitrile
Amyl Acetate
Benzene
Carbon dlsulphlde
Carbon tetrachloride
Chlorobenzene
Chlorform
Cyclohexane
dichloroethane
Ethanl
Ethyl Acetate
di-Ethyl ether
iso-Propyl Alcohol
Methanol
Methyl Acetate
Methylene Chloride
Methyl ethyl ketone
Perchloroethylene
Phenol
Styrene
Toluene
1,1,1-trichloroe thane
1,1, 2-trichloroethylene
Freon
Xylene(s)
70°F
0.2229
0.0972
0.0056
0.0796
0.3813
0.1095
0.0094
0.1923
0.0924
0.2171
0.0578
0.0902
0.5214
0.0438
0.1233
0.2065
0.4773
0.1062
0.0192
0.0006
0.0083
0.0310
0.1331
0.0795
0.3280
0.0085
90°F
0.3586
0.1617
0.0108
0.1327
0.5817
0.1761
0.0165
0.3066
0.1504
0.3416
0.1056
0.1513
0.8032
0.0819
0.2155
0.3342
0.7646
0.1763
0.0341
0.0013
0.0151
0.0556
0.2192
0.1334
0.5108
0.0158
110°F
0.5580
0.2595
0.0199
0.2126
0.8614
0.2738
0.0278
0.4732
0.2366
0.5208
0.1848
0.2445
1.2006
0.1467
0.3622
0.5228
1.1849
0.2826
0.0582
0.0026
0.0264
0.0955
0.3485
0.2158
0.7790
0.0280
130°F
0.8426
0.4032
0.0348
0.3288
1.1386
0.4118
0.0452
0.7091
0.3609
0.7715
0.3113
0.3826
1.7461
0.2524
0.5879
0.7933
1.7826
0.4385
0.0957
0.0050
0.0444
0.1583
0.5369
0.3379
1.1315
0.0478
150°F
1.2383
0.6087
0.0594
0.4974
1.7487
0.6068
0.0716
1.0349
0.5355
1.1139
0.5069
0.5813
0.4191
0.9243
1.1714
2.6108
0.6612
0.1525
0.0091
0.0722
0.2538
0.8039
0.5138
0.6195
0.0787
-------
134 AIR MONITORING
APPENDIX 2
Diffusivity Coefficient
The diffusivity coefficient "d", is measured in cm2/sec. In
Arnold's equation, the diffusivity coefficient d is in cm2sec.
In Nusselt's equation, the diffusivity coefficient "D" is meas-
ured in m-/hr.
D = 0.36d
Fuller14' has derived values for d, using the relationship
d.0.OOIT,,5(L i
1 M2' ->/
1 *• cm "/sec
D values for a selected group of hazardous compounds is
presented for a range of temperatures in Appendix (2).
Table III
Values of Diffusivity Coefficient D (mVhr)
Compound
70 °F
90 °F
110 °F
Melhanol
Elhanol
Phenol
Acetone
M.E.K.
Methyl Acetate
Ethyl Acetate
Methyl Aminc
di-Melhyl Amine
Ethyl Amine
Hexane
Xylene
Toluene
Benzene
Di-Ethyl Ether
Melhylene Chloride
Chloroform
Carbon Telrachloride
FreonCCjF.CI.)
PCB(lcfilorine)
0.05687
0.04338
0.03000
0.03724
0.03231
0.03475
0.03067
0.05585
0.04282
0.04284
0.02837
0.02587
0.02827
0.03146
0.03159
0.03689
0.03203
0.02878
0.02578
0.02068
0.06069
0.04630
0.03199
0.03974
0.03448
0.03708
0.03273
0.05960
0.04572
0.04572
0.03027
0.02761
0.03017
0.03357
0.03371
0.03936
0.03418
0.03071
0.02751
0.02207
0.06460
0.04928
0.03405
0.04231
0.03671
0.03948
0.03484
0.06345
0.04867
0.04867
0.03222
0.02939
0.03212
0.03573
0.03589
0.04191
0.03639
0.03270
0.02929
0.02350
Example 1
To determine the steady state emissions of methanol
from a receiving lagoon.
Basis for Model
The receiving lagoon has an operating volume of 600,000
gallons.
The feed rate is constant at 6667 gph
The methanol concentration in the feed averages 0.09%.
The feed weight is 8.54 Ib/gal
The temperature is constant at 70°F
The wind velocity is stead) at 4 m/sec.
The initial evaporation rate E0 will be determined usi
Nusselt's modified equation;
% • (V.P.) • Wg.78 ..... (,2)
E0 = 0.53452 Am •
The the steady state evaporation rate Es will be determined
from the relationship;
Es=E0-i
The parameters needed are:
L0 = the pounds of methanol in the lagoon at zero time
= 600,000 gallons X
gal
X
= 4503 Ib Methanol
F = 6667 gph
V = 600,000 gal
L = 40m
4503 X 6667
K =
600,000
= 50 Ib/hr
Am = 0.7854 L2= 1256.64m2
W0 = 4 m/sec
% = 0.09
V.P.= 0.12324 Atm(App 1)
D = 0.05687,n2/hr @ 70°F
Accordingly,
E0 = (0.53452X1256.64)(0.23638)(0.09(0.1234) (2.94854)
= 5.193 Ib/hr Methanol.
To determine the emissions at the steady state, Es, first deter-
mine Ls, the methanol content of the lagoon at the steady
state
K
K = 50 Ib
!-X=
_
V 4503 600,000
-2
Ls = 1.2264410-2 = 4076.7 Ib Methanol
= 5.193 (4?Jf;?) = 4.7 Ib/hr Methanol
Es = E0
-------
AIR MONITORING 135
Example 2.
Ammonia emissions from an aerated lagoon at the steady
state. An aeration lagoon is 45 ft in diameter, has an operating
volume (V) of 300,000 gal.
A feed stream at 10.5 pH, containing 0.1% NH3 enters the
lagoon at a rate (F) of 210 gph.
Mass of ammonia in the lagoon at zero time :
Mass of ammonia entering the lagoon each hour:
•v -m , v 8.34 lbv 0.1 IbNH? ,,,,,,,,
is K = 210 gal X _, X , ----- •* = 1.7514 Ib.
gal
lOOlb
An aeration pump transfers 3390 gpm of lagoon water to a
deflector which produces an umbrella-like spray cover which
completely covers the lagoon.
The flight time of droplet in seconds will be;
V
lagoon diameter in feet
5.6568
In this case, the lagoon diameter is 45 ft, therefore flight
time (G) 2t = 1 . 1 86 seconds
Assume the water droplets (r) average 0.25 'cms radius.
The weight of water suspended as droplets at any time is:
min
gal
Ib
60 sec
Total surface of the droplets, Ac = 253720
X 4?rr2 (droplet area)
_4_
37rr3 (droplet wt.)
= 253720 X— cm2
r
when r = 0.25 cm Ac = 3,044,642 cm2
The vapor pressure (V.P.) of pure ammonia (NH3) at 70 °F
is 8. 55 13 Atmospheres
The molar concentration of 0.1% ammonia is 1.06X 10~~3
y* = (molar concentration)(vapor pressure) = 9.05x10 3
The diffusivity coefficient (d) of ammonia from water to air
at 70°F = 0.23785 cm2/sec.
The volume of ammonia gas entering the atmosphere from the
aeration lagoon V cm3 = 2Acy*l / according to Arnold.
V = 16513cm3/!.186 sec
Amount of ammonia emitted, Ib/hr:
3 mole
,=16513cr
3600 sec
hr
X
1
1.186 sec
moeH
= 83.8 Ib Ammonia/Hr.
To determine the emission rate of ammonia at the steady state
proceed as follows:
L0 = 2502 Ib
Ls = Ib ammonia in the lagoon at the steady state
T =
s
K
K = 1.7514 Ib
1-K= *&+L
83.8 .
2502 300,000
0.0342
- 0.0342= 5L211b
Emissions at the steady state
Es = EoJ ii =1.715 Ib/hr.
Steady state emissions of ammonia is 1.75 Ib/hr.
Thus, the ammonia concentration leaving the lagoon in
the liquid, at the steady state is
L^=_^ =2X10-5
(8.34)V (8.34)(300,000)
= 20 mg/1
-------
AIR MONITORING OF A HAZARDOUS WASTE SITE
DAVID A. SULLIVAN
JEROME B. STRAUSS
Versar Inc.
Springfield, Virginia
INTRODUCTION
Versar Inc., as a contractor to the U.S. Environmental
Protection Agency for air quality emergency response,
performs ambient monitoring, dispersion modeling and
other support functions as requested by the Division of
Stationary Source Enforcement. In this capacity, Versar
and GEOMET, Inc. (prime contractor) responded to a
request to perform ambient air sampling at Rollins Envir-
onmental Services' (RES) waste treatment and disposal
facility near Baton Rouge, Louisiana. This paper focuses
on Versar's approach to the identification and quantifi-
cation of airborne pollutants from a hazardous waste
facility.
Community residents had complained of odors and
respiratory impairment which they attributed to the RES
waste treatment and disposal facility. To determine what
action could be supported by air quality impacts, ambient
air concentrations were needed for hazardous pollutants
emitted from the site. Sampling and analysis were compli-
cated by the wide range of compounds which are brought
to RES for treatment and/or disposal. It was anticipated
that emissions would be variable and that reactions could
occur among components of the waste in liquid or solid
form, or among gaseous pollutants emitted from the fa-
cility. Since the airborne pollutants were not well identi-
fied at the outset of this program, a two-phase approach
was designed to identify and quantify organic air pol-
lutants. The study was intended to provide a quick re-
sponse assessment of airborne emissions from the facil-
ity, as opposed to providing a long-term ambient air
quality evaluation.
Using solid adsorption columns composed of Tenax
or activated charcoal for sampling, and using gas chroma-
tography (GC) and gas chromatography/mass spectrome-
try for analysis (GC/MS), the Phase I program was in-
tended to identify pollutants that were present in the
highest concentrations. RES management placed restric-
tions on sampling which allowed Versar to collect only
around the perimeter of the site (i.e., "fenceline" sam-
ples).
Upwind samples did not contain detectable concentra-
tions of any pollutants. Therefore, downwind concentra-
tions presented in this report also reflect downwind
minus upwind concentrations.
Benzene and toluene were present in the highest con-
centrations and were quantified during Phase I of the
project. The maximum concentrations at the fenceline
were 0.5 and 2.4 ppm, for benzene and toluene, re-
spectively. Compounds that were only tentatively identi-
fied and therefore not quantitied, included ethyl ben-
zene, 1,1,1-trichloroethane, methylene chloride, tetra-
chloroethylene, chloroform, and meta-xylene. Since the
impact to the general public was an important considera-
tion for this study, fenceline sampling did allow for a
suitable data base to determine impacts from the facility.
The Phase II program was designed to provide quanti-
tative air quality data and utilized a portable GC and cali-
bration standards for benzene, toluene and the tentatively
identified compounds of Phase I. The use of a portable
GC greatly assisted in the selection of areas of expected
maximum concentrations and was useful in providing
feedback on appropriate sampling times. Maximum con-
centrations were again found for benzene and toluene,
with maximum fenceline values of 0.2 and 0.4 ppm, re-
spectively. Tentative identification included several
compounds that were not observed during Phase I, in-
cluding pentenenitrile and methyl pentene.
DESCRIPTION OF RES FACILITY
The RES facility is located 12 miles north of Baton
Rouge on 165 acres of land. It is designed to treat and/or
dispose of liquid and solid wastes generated by area in-
dustries, such as from chemical, petrochemical, refining
and synthetic polymers plants. Treatment and disposal
practices include landfarming, landfilling, incineration,
biological treatment and physical/chemical treatment.
Area residents had filed numerous complaints alleging
that activities at RES were creating an odor problem and
were adversely affecting the health of the community. It
appears that many of the complaints were associated with
odors from the biotreatment area and the landfarm area.
Prior to the Phase I program, however, RES had
changed to a subsurface injection system for landfarm
application, which reduced odors and organic emissions;
also, the emissions from the biotreatment area were sub-
stantially reduced to meet a State-issued order. Modifi-
cations to the facility between Phase I and II sampling
included closing a number of landfill pits to reduce
odors, as requested by the State.
The data collected during the sampling programs were
considered to accurately represent ambient conditions li
the time they were obtained, (Phase I—September 25-26,
136
-------
AIR MONITORING 137
1980, and Phase II—November 12-13, 1980). Versar's
experience with modifications of the type used at the fa-
cility would suggest that the ambient concentrations had
been substantially higher before the corrective action had
been taken during the summer and fall of 1980. It is
therefore possible that complaints about the facility prior
to these modifications were the result of higher concen-
trations than those observed during the sampling program.
By the time the Phase II program began, the most
distinctive and unpleasant odors were found to be as-
sociated with the mixing pit rather than the landfarm or
biotreatment areas. The mixing pit combines ash with
liquid wastes as a treatment method. Gaseous and par-
ticulate emissions are emitted from this source.
SAMPLING
Air Quality
Phase I of the two-phase sampling program involved
collecting 46 air and eight liquid samples. Activated char-
coal and Tenax were each used to collect 23 air samples.
The height of the intake in all samples was approxi-
mately four feet, to represent the approximate breathing
level. In Phase I, the air sample volumes through the ad-
sorption columns were from 1 to 150 liters. During this
phase, the strongest odors were detected downwind of
the mixing pit and the biotreatment area.
These areas were evaluated by selecting upwind and
downwind sampling sites, where concurrent upwind and
downwind sampling was performed. As previously stated,
detectable concentrations were not found in any of the
upwind samples. Phase I was characterized by fairly vari-
able winds. The highest concentrations and the largest
variety of compounds were generally observed downwind
of the mixing pit. The sampling sites for Phase I are
shown in Figure 1.
In Phase II, the same basic sampling approach was
used as in Phase I, but feedback was obtained on site se-
lection and optimum sample volumes through the use of
an on-site Hewlett Packard 5880 dual channel GC. Sample
volumes were increased over Phase I since breakthrough of
pollutants into the back sections of the adsorption col-
umns did not occur. Volumes in the range of 150-400
liters were obtained during Phase II. Wind flow was fairly
consistently out of the east through east-southeast during
Phase II, which was the optimum wind flow to evaluate
fenceline concentrations downwind of the mixing pit and
biotreatment area. The location of the sampling sites for
Phase II is shown in Figure 2.
Sampling was performed from 3:00 to 6:00 a.m. in order
to coincide with minimum dispersion conditions and
therefore relatively high ambient air concentrations.
Afternoon sampling was also performed to coincide with
maximum activity periods. The early morning hours of
November 12, 1980, were particularly suitable for samp-
ling since stable conditions (relatively poor dispersion)
occurred with clear skies and light but reasonably steady
winds, 3.5 to 5.0 mi/hr between 1:00 and 6:00 a.m.
The sampling protocol was essentially the same for
Phases I and II, with the exception that a GC was not in
the field during Phase I. The basic protocol was as fol-
lows:
(1) Based on current wind direction data, downwind
and upwind sites were selected.
(2) A steel support post was used to provide a sample
intake height of approximately four feet.
(3) The seals on the adsorption tube were broken, and
the tube was attached to the end of approximately
one foot of neoprene tubing. The pump and tub-
ing were downstream of the sample.
(4) Sampling times generally were two hours; however,
some samples were taken for as long as four hours.
(5) The sample was stored in the GC trailer at 4°C un-
til it was analyzed. Ice was contained in sealed
plastic bags to minimize sample exposure to mois-
ture.
Figure 1.
Air Sample Locations—Phase I
Figure 2.
RES Phase II Air Sampling Locations
(All Sampling Sites Were Within Fenceline, Except 15,16 and 23)
-------
138 AIR MONITORING
Blanks were taken for quality control purposes by break-
ing the seals on the glass adsorption tubes, capping and
then handling the same as the samples. No contamination
was found during either Phase I or II.
The information associated with each sample including
the sampling period, sample volume, site location and
other specifications associated with each sample is sum-
marized in Tables I and II. During Phase I, the results
of the analyses indicated that the charcoal and Tenax ad-
sorption tubes provides similar results. During Phase II
most of the samples were collected on charcoal.
Liquid Samples
Liquid samples were taken at eight holding pits during
Phase I to assist in the evaluation of emissions from this
facility. GC qualitative analysis of the volatiles from these
samples did not indicate any compounds different from
those identified or tentatively identified during the air
sampling program. For selecting appropriate adsorption
Table I.
Sample Log Summary, Phase I
Sample
Number
5
6
7
9
10
II
12
17
IB
19
20
21
22
23
24
27
28
29
30
31
32
33
34
39
40
4!
3?
36
37
38
47
48
49
50
$topc>*3 b»tt*r»
Pu«p
Number
Charcoal
Oarcoal
Tenax
Tenax
Charcoal
Charcoal
Tenax
Tenax
Tenax
Tenax
Charcoal
Charcoal
Charcoal
Charcoal
Ten ax
Tenax
Charcoal
Charcoal
Tenax
Tenax
Charcoal
Charcoal
Tenax
Tenax
Charcoal
Charcoal
Tenax
Tenax
Tenax
Tenax
Charcoal
Charcoal
Tenax
Charcoal
Charcoal
Charcoal
Charcoal
Tenax
Tenax
Tenax
Tenox
Charcoal
Charcoal
Tenax
Tenax
Charcoa I
Charcoal
Flo. Volume
(liters!
51.7
1.88
98.7
11.3
11.6
143.0
2.65
136.0
14.4
126.0
66.0
2.40
9.86
122.0
116.0
2.26
1.29
52.8
95.7
5.61
1.36
37.4
71.4
8.16
2.40
126.0
10.2
16.8
140.0
147.0
2.73
3.28
13.9
1 14.8
2.34
126.0
120.0
14.3
10.5
130.0
68.2
2.48
16.3
136.0
143.0
2.65
Start
Tine
3:40p«
4: 27pm
3:l5p» 4:23p«
9/25/80
4:42p« 5:40p«
6:00p» 6:33pn
6:04pn 6:38p»
5e> 7:45a>i
7:50ae 9:!2a>
tubes and to aid in the selection of calibration standards
for field analysis, liquid samples can be a useful prelim-
inary step in an overall evaluation. By performing GC/MS
and/or GC analysis on the volatile gases of the liquid
samples, an indication of probable airborne pollutants
can be made.
Meteorological Monitoring
A portable wind system was set up at the same, well-
exposed location for both phases, as shown in Figures 1
and 2. This station was expected to represent general flow
for the entire facility except for conditions associated
with light winds (e.g., less than two miles per hour.) The
terrain is quite flat. The wind vane was oriented by com-
pass readings to the nearest degree, and the reduced data
corrected for the 4 degree east magnetic declination.
The height of the wind vane and anemometer was ap-
proximately eight feet above ground level which is rela-
tively low for monitoring wind data. Although wind
speeds were relatively low and wind flow was somewhat
more variable than at the more standard ten meter level,
sufficient documentation for general flow was obtained.
Relative humidity values were obtained from wet bulb
and dry bulb temperature readings from a sling physchro-
meter. On the sample log sheets, cloud cover data and
weather events were recorded to provide an indication of
atmosphere stability and possible interference from pre-
cipitation . These data are documented in Tables HI and IV.
ANALYSIS
The procedures followed in analyzing the samples in
the field (Phase II) and in the Versar analytical labora-
tory are discussed in the following section. This descrip-
tion pertains specifically to Phase II; however, similar
laboratory procedures were used for Phase I.
GC Analysis
GC analysis in the field was performed using an HP
5880 gas chromatograph equipped with a flame ioniza-
tion detector. GC analysis in the analytical laboratory
was performed using an HP 5700 gas chromatograph
equipped with a flame ionization detector. GC analytical
conditions were as follows:
Gas Chromatograph
Detector
Column
Carrier Gas
Hydrogen
Hydrogen
Air
Oven Temp., initial
program rate
final
Field
HP 5880
FID
6.!ms.s. 10WSP-1000
30 ml/min N2
30 ml/min
30 ml/min
350 ml/min
100°Cfor8min
4°C/min
200°Cibrl6min
Laboratory
HP 5700
FID
6.1ms.s. 10WSP-1000
30 ml/min Nj
30 ml/min
30 ml/min
250 ml/min
105°Cfor8mto
4°C/min
200"Cforl6min
Charcoal samples were analyzed using NIOSH Method
P + CAM 127 as outlined in the NIOSH Manual of
Analytical Methods. The use of carbon disulfide as the
-------
AIR MONITORING 139
Table II.
Sample Log Summary, Phase II*
Table III.
Meteorological Data, Phase I
September 25-26,1981
Sample No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
IB
19
Pump No.
4806
4815
5934
4806
4815
5934
5949
4816
5195
5942
5156
5942
5949
5195
4816
5156
5934
5156
5949
Start Time
(CST)
0447
0436
0434
0643
0630
0637
1410
1426
1416
1400
1440
1606
1619
1627
1610
1617
0412
0422
0356
Stop Time
(CST)
0641
0629
0625
0841
0831
0830
1618
1609
1625
1604
1616
1753
1802
1806
1757
1805
0547
0550
—
Operator
Sul I Ivan
Sul llvan
Feldman
Sul llvan
Sul 1 Ivan
Felthan
Koch/AI len
Su 1 1 1 van
Koch/AI len
Koch/AI len
Sul 1 Ivan
Al len
Al len
Al len
Koch/
Su 1 1 1 van
Koch/
Sullivan
Koch
Koch
Al len
Date
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/12
11/13
11/13
11/13
Site
Number/
Comments
Site *I3
Site f!4
Site *15
Site *13
Site 114
Site /16
Site f!8
Sits *19
Site (21
Site 116
Site *20
Site »16
Site f18
Site *21
Site 119
Site »20
Site »20
SI1B <22
Site *16,
Samp le"*
Volume
(liters)
199
210
207
206
225
210
271
176
273
202
188
174
219
211
183
211
175
176
—
pump stopped
20
21
22
23
5942
4816
5942
4815
0406
0611
0557
0404
0556
0734
0727
0615
Al len
Al len
Al len
Kbch
1 1/13
11/13
11/13
11/13
Site »23
Site »16
Site «3
Site *17
180
146
150
244
Date
9/25
9/26
Time
(cst)
1600
1700
1800
0700
0900
1000
Wind
Speed
(mph)
4
4
5
6
9
8
Average
Wind
Direction
(degrees)
250 (variable)
200 (variable)
040
050
055
055
Relative
Humidity
(§)
66
70
91
88
75
GC/MS Analysis
Prior to the qualitative analysis of the air samples
collected on Tenax and charcoal tubes, the GC/MS
calibration was verified by a direct injection of para-
bromo-fluorobenzene and the characteristic ions and rela-
tive ion abundances were verified. For samples collected
on charcoal, a 2-microliter injection of the methanol char-
coal tube extract was chromatographed into the mass
spectrometer. For samples collected on Tenax, the Tenax
tube was thermally desorbed at 180°C for 4 minutes using
Matheson purity helium (99.9999%) to purge the de-
sorbed components onto the GC column. All samples
were chromatographed and analyzed using the following
30
31
32
33
34t
35t
36t
37
38
39
40
* Except
tubes.
** Correc
t Tenax
4816
5156
5949
5942
4606
5195
4815
5156
4816
5942
5934
where Indicated
ted to standard t
sample ( low f low)
1741
1738
1754
1605
1934
1935
1945
2121
2132
2149
2218
as a Tenax
emperature
.
2132
2120
—
2143
2234
2235
2244
0015
0018
0034
0000
sample, al
and pressu
Sullivan
Sul llvan
Su 1 1 1 van
Su 1 1 1 van
Al len
Al len
Al len
Koch
Koch
Koch
Allen
1 samp les H
re.
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
11/13
Site
Site
Site
pump
Site
Site
Site
Site
Site
Site
Site
Site
'ere col lected
120
124
(18,
stopped
*17
124
124
f18
124
120
125
123
427
469
—
371
109
397
331
367
307
261
195
on charcoal
GC/MS conditions:
Column
Oven Temp.
Scan Range
.initial
initial hold
program rate
final
Ion Source Voltags
Scan Rate
1 % SP-1000 on 60/80 Carbopack B 2.44 m x 2
mm ID glass
45 °C
t min
j mm.
8"C/min
220 °C
34-257 AMU
70 eV
2.7 seconds
Electron Multiplier
Voltage
Run Time
2400V
35 minutes
desorption solvent unfortunately precluded its quantifica-
tion as a possible contaminant.
During Phase II, three compounds were chosen as in-
dicators for ambient onsite analysis. These were benzene,
toluene and meta-xylene. Work in Phase I had identified
these as compounds of interest to be quantified onsite.
All Phase II samples were screened by GC to determine
those with the highest concentrations. The three samples
with the highest concentrations were selected for GC/MS
analysis to qualitatively identify a wide range of pollut-
ants. After GC/MS analysis, GC analysis was then used to
quantify the identified pollutants.
Standards were prepared in carbon disulfide for those
identified components for which reference materials were
available. These were benzene, toluene, ethyl benzene,
1,1,1-tetrochloroethane, methylene chloride, tetrachloro-
ethylene, chloroform and meta-xylene. Ambient air con-
centrations were calculated from the amount measured in
the extract and the volume of air sampled.
Background-subtracted spectra were obtained for all
peaks and compared to in-house reference spectra and
spectra obtained from EPA/NIH/MSDC Library Search
System. Confirmation of identified compounds was ob-
tained by analysis of standard reference materials under
the same conditions as the samples. The standards and
samples were matched for retention times and mass
spectra.
The following criteria are the minimum requirements
that were used for the interpretation of the mass spectra
to confirm identification of any pollutant:
•The mass spectra contained all ions present above 10
percent relative abundance in the mass spectrum of the
reference with the general agreement of ±20% of the
relative abundance in the reference mass spectrum.
•Ions present in the experimental mass spectrum that are
not present in the reference spectrum must not exceed
10% of the total ion abundance in the experimental mass
spectrum.
-------
140 AIR MONITORING
Table IV.
RES Onsile Meteorological Data, Phase II
November 12,1980
Sunny skies in the morning, 3/10 cloud cover, high cloud (thin)
becoming 8/10 high cloud cover by sunset.
Average Relative
Wind Dir.- Range" Humidity
(Degrees)
085
075
065
055
055
060
060
045
055
070
065
060
065
050
055
100
095
050
Time
(CSD
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
Wind Speed
(mph)
3.5
4.0
5.0
4.5
4.5
5.0
6.5
7.0
7.5
10.0
9.0
9.0
9.0
7.0
6.5
2.5
2.5
2.5
Calm
Calm
2.5
3.0
2.5
2.5
035
055
040
040
Range"
(Degrees)
70
60
40
40
40
50
70
60
55
65
65
75
65
80
85
60
20
75
65
55
75
60
67
44
32
26
41
53
58
' Wind direction was corrected for magnetic declination of 4° East, associated with compass
orientation of wind vane. Wind speed and wind direction averages are based on the period
60 minutes preceding the indicated hour.
"The second highest maximum and second lowest minimum were determined and used to
estimate the range or the wind direction for the period. This provides a rough indication of
(he horizontal variability or wind direction which is a function of atmospheric stability.
—Data not available.
November 13 and 14,1980
9/10 high cloud cover in the morning becoming overcast by afternoon;
trace of precipitation occurred between 1900-2000 CST
Average
Time
(CST)
0100
0200
0300
0400
0500
0600
0700
0800
0900
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
Wind Speed
(mph)
4.5
5.0
5.0
5.5
5.5
5.0
4.5
5.0
7.0
7.5
8.0
8.0
12.5
12.0
12.0
8.0
9.0
9.5
9.0
8.0
9.0
12.5
13.0
10.0
Wind Dir.*
(Degrees)
070
065
070
085
075
075
090
055
085
090
080
075
105
110
110
110
105
105
095
095
105
110
110
095
Range**
(Degrees)
55
50
50
50
70
55
50
80
70
60
55
75
55
35
70
60
55
40
40
40
50
40
35
40
R.H.
<*)
—
—
59
...
56
_
—
56
--
._
...
...
...
...
—
.„
...
77
...
86
...
82
86
81
0100
7.0
November 14,1980
095 50
•The retention time of the experimental mass spectrum
must be within ± 5 scans or ± 30 seconds of the reference
compound.
Detection limits were 10 mg/1 in solution, and all com-
ponents that did not meet these criteria are considered
tentative identification.
Two charcoal tube samples (Nos. 9 and 38) and one
Tenax tube sample (No. 35) were analyzed by GC/MS
for identification of major organic compounds. The
presence of methylene chloride, acetone, carbon disul-
fide, chloroform, 1,2-dichloroethane, 1,1,1-trichloro-
ethane, benzene, hexane, toluene and ethyl benzene was
confirmed by mass spectra in the Tenax sample (No. 35).
Other trace components that were tentatively identified by
library search were:
2-pentenenitrile ethane, 1,2-dichloro
acetone isopropyl ether
isopropyl alcohol 1-pentene, 2-methyl
tert-butanol 2-pentenenitrile
4-penten-2-ol hexane
cyclopentanone, o-methyloxine butanoic acid, 3-3 dimethyl
hexanedinitrile 2-butanone
Also, tentatively identified by in-house spectra were
tetrachloroethylene and trichlorotrifluoroethane. Due to
the desorption procedures used for GC/MS analysis,
quantification of these samples (9, 35 and 38) was not
feasible by GC or GC/MS.
Table V.
Concentrations of Benzene and Toluene in Air as Calculated
from the Charcoal Tube Extracts, Phase I
Site
2
2
4
4
8
10
Field No.
1
2
9
10
33
36
Laboratory
Sample
Number
6171
6172
6175
6176
6203
6206
Benzene
(ppm)
0.05
0.05
0.03
0.02
0.004
0.007
TLV = 10
TolncM
(ppm)
2.3
2.4
1.4
1.0
0.02
0.05
TLV-MO
-------
Table VI.
Phase II GC Results Summary (Concentrations in ppm)*
AIR MONITORING 141
1 13 .006 -006
2 14 .056 .021 .084
3 15 (UPWIND SAMPLE )
4 13 .003 .001
5 14
6 16 ( UPWIND SAMPLE )
7 IB
8 19 .004
10 16 ( UPWIND SAMPLE )
11 20 .089 .240
12
13 IB .060 .063 .055 .059
14 21 .012 .034
15 19 .057 .052 .0132
TLV** 10 200 100
16 20 .215 .428
17 20
IB 22 .031 .108
20 23 (UPWIND SAMPLE )
21 16 ( UPWIND SAMPLE)
22 23 (UPWIND SAMPLE)
23 17
31 24
33 17 .016 .030 .022 .026
37 24
39 25
40 23 (UPWIND SAMPLE >
GC Run Onstte
.019 .241 GC Run Onslte
No significant
peaks
GC Run Onslte
No significant
peaks
GC Run Onslte
No significant
peaks
GC Run Onstte
Early peak
GC Run Ons Its
GC Run Onslte
No significant
peaks
.01 1 GC Run Onslte
No significant
peaks
.064
.020 GC Run Onslte
.042 .138
100 100 50
.026 GC Run Onstte
No significant
peaks
GC Run Onslte
GC Run Onslte
No significant
peaks
No significant
peaks
No significant
peaks
No significant
peaks
GC Run Onslte
No significant
peaks
.015 .039
No significant
peaks
GC Run Onslte
Some smal 1 peaks
No significant
peaks
' Blanks Indicate that pollutant concentrations
h Threshold limit value.
100
i below detection limits.
RESULTS AND CONCLUSIONS
The quantitative data obtained during the Phase I and
II programs, are summarized in Tables V and VI, re-
spectively. For each compound, threshold limit values
(TLVs) are provided only to indicate a point of compari-
son. However, since TLVs are associated with an eight-
hour average applicable to the workplace, they are not
suitable for exposure assessments of the general popula-
tion. Although the observed concentrations for which
TLVs were in place were substantially below these values,
it is beyond the scope of this paper to make any evalua-
tion regarding health effects associated with exposure of
the general public to these concentrations.
For both the Phase I and Phase II air sampling pro-
grams at Rollins Environmental Services, many organic
compounds were identified by GC and GC/MS analyses.
Samples upwind of RES did not indicate significant con-
-------
142 AIR MONITORING
centrations of organic pollutants and odors were generally
not noticeable at the upwind sampling sites. Benzene and
toluene were usually found in the highest concentrations
in the downwind samples. During Phase I, the highest
values for benzene and toluene were found to be 0.05
ppm and 2.4 ppm, respectively. Phase II sampling indi-
cated a maximum of 0.2 ppm for benzene and 0.4 ppm
for toluene. A number of organic compounds were
present, but in concentrations not high enough to allow
for positive identification. For example, pentenenitrile
which could possible be toxic at relatively low levels, was
tentatively identified.
To perform a more detailed identification and quanti-
fication of the pollutants emitted from this facility, a
longer-term program would be necessary. A sampling pro-
gram during the summer quarter, when the highest emis-
sion of volatile organics could be expected, may result in
detection of substantially greater pollutant concentrations.
Caution must be exercised in drawing conclusions from a
short-term program such as performed during Phases I
and II of the study. Although in Phase II, conditions
were such that relatively high concentrations would be
expected (i.e., sampling was performed from three to six
a.m. with fairly light winds and clear skies), a short-term
sampling program would not be expected to contain
values near the annual maximums. The variability of waste
being treated and disposed of is a factor that makes a
long-term program more meaningful.
The RES incinerator was not operational during either
Phase I or Phase II of the Versar sampling program. This
could explain some differences between these results and
those of a previously performed short-term program. Al-
though that program produced results that were generally
of the same order of magnitude as our study for benzene
and toluene,01 one sample from the previous program in-
dicated a fenceline concentration of 6.1 ppm for benzene,
which is substantially higher than any concentrations
measured during our program. In addition, the qualitative
identifications of other compounds were different from
those observed during this study.
Although short-term sampling programs have limita-
tions, the results provide an indication of ambient concen-
trations to evaluate the need to collect a more long-term
data base or to take enforcement action. To complete the
evaluation, a health effects review of the data obtained
from the sampling programs would be necessary. This
will be performed by the Environmental Protection
Agency (Region VI). At the time of this writing, the health
effects review had not been completed.
GENERAL RECOMMENDATIONS FOR
RELATED PROGRAMS
(1) Perform analysis of volatiles from liquid samples, if
possible.
(2) Choose adsorption medium on the basis of pollut-
ants anticipated to be present.|4i5-6>
(3) Use screening techniques to choose areas of maxi-
mum expected concentrations, such as portable Gas
Chromatography equipment with a flame ionization
detector to measure total hydrocarbons, obtaining
grab samples in collection bulbs in conjunction with
an onsite Gas Chromotrograph, etc.
(4) Sample collection periods should be chosen to en-
sure that a sufficient quantity of sample is obtained
to allow for identification and quantification, but
not so large that saturation of the absorption
medium occurs and breakthrough problems result
or that the collection efficiency is significantly re-
duced by high flow rates.(4|5)
(5) Calibration of portable pumps used to draw the
samples should be performed at least on a daily
basis.
(6) Proper handling, such as refrigeration of samples,
chain-of-custody documentation, etc., are essential
to ensure defensibility of the data collected.
(7) The use of an onsite wind system is highly recom-
mended in lieu of the use of offsite data sources.
Upwind and downwind samples should always be
taken concurrently with properly documented wind
data. Cloud cover and weather events should also be
documented.
(8) The field log should include an indication of site
activities associated with the sampling periods (e.g.,
incinerator operation, unloading of waste, truck
traffic, landfarm activity, etc.)
(9) If possible, sampling periods should include those
associated with maximum anticipated emission
terms and poor dispersion conditions, e.g., high
activity periods, maximum ambient temperature,
poor dispersion conditions, wind flow toward sen-
sitive receptors.
REFERENCES
1. "Phase I Air Monitoring Program/Rollins Environ-
mental Services," Versar Inc., Subcontract No. 6156-
12, November 4, 1980.
2. "Phase II Air Monitoring Program/Rollins Environ-
mental Services," Versar Inc., Subcontract No. 6156-
12, January 19,19981.
3. "Analysis Report, Rollins Environmental Services,"
Contract No. 21660-80-01, Enviro-Med Laboratories,
Inc., Baton Rouge, Louisiana, May 27, 1980.
4. "Characterization of Sorbent Resins for Use in Envir-
onmental Sampling," EPA-600/7-78-054, March
1978, IERL/RTP.
5. "Selection and Evaluation of Sorbent Resins for the
Collection of Organic Compounds," EPA-600/7-77-
044, April 1977, IERL/RTP.
6. Katz, M., ed., "Methods of Air Sampling and Analy-
sis," American Public Health Association, Inter-
society Committee, Byrd Pre-Press, Inc., Springfield,
Va., 1977.
-------
INFLUENCE OF SAMPLING TECHNIQUES ON
ORGANIC WATER QUALITY ANALYSES
ARTHUR M. SEANOR
LARRY K. BRANNAKA
Dames & Moore
Baldwinsville, New York
INTRODUCTION
In the past, the role of a groundwater geologist was
generally confined to the development of techniques to
obtain or remove sufficient quantities of water to satisfy
the objectives of a particular project. However, the re-
cent inclusion of approximately 113 organic compounds
on the priority pollutant lists has opened a new concern
for groundwater geologists. This concern directly relates
to the problem of obtaining a sample of groundwater that
one can be confident is truly representative of the sub-
surface fluids. The need to supplement the existing
knowledge currently available pertaining to the fate and
movement of organic constituents in the subsurface has
been documented by several authors.(1'2)
The presence of many trace organics in a groundwater
supply system even in very low concentrations may cre-
ate near-term or long-term health hazards. The complexity
of these chemicals has further confused attempts at un-
derstanding the fate of most trace organics once they enter
the subsurface. Many times compounds not produced or
known in a facility's waste stream have been identified in
samples of groundwater collected from a particular fa-
cility.
Because of the potential acute hazards posed by many of
the trace organic compounds (even in the ug/1 range), the
techniques for obtaining and storing of groundwater
samples containing these compounds may differ sub-
stantially from classical field sampling techniques. Even
with the vast amount of research currently underway with
respect to sampling of organics, the state-of-the-art for
obtaining samples still lags far behind that of the analytical
laboratory techniques.
Because the advanced laboratory techniques are so
costly and labor consuming, the necessity for accurate,
cost effective sampling techniques becomes more appar-
ent. The money and time involved in analyzing a suite of
samples for trace organics can be completely wasted if the
samples were incorrectly samples or improperly stored.
AREAS OF SPECIFIC
ORGANIC CONTAMINANTS
During a series of recent groundwater investigations,
Dames & Moore has attempted to identify potential prob-
lem areas in order to improve the efficiency of obtaining
and storing samples. Five major areas of concern were
identified whereby specific contaminants may be allowed
to either enter the groundwater or contaminate the samples
before analyses can be instituted. The areas of concern
addressed include the following:
•The drilling operations
•Piezometer or well construction materials
•Sampling equipment
•Sampling methods
•Sample preservation
Each of these areas of concern are addressed in more de-
tail in the following sections of this paper.
Drilling
Generally the drilling of a well is required to obtain
groundwater samples for analyses. A well should be in-
stalled wherever possible to minimize the potential for
organic contamination. However, this one operation,
which may in fact be the most costly part of a ground-
water study, has the potential to introduce the greatest
error! Additionally, if error is considered, then it can be
very costly from an economical and labor standpoint to
redrill a new well. During a drilling operation several op-
erations may lead, directly or indirectly, to the existence
of contaminants within the groundwater.
Where depths in excess of about 50 to 75 ft are neces-
ary to drill, rotary drilling or percussion drilling tech-
niques are normally employed. Rotary drilling is gen-
erally favored over percussion drilling because it is much
faster. However, circulation of drilling fluids or muds
during rotary drilling may significantly alter the chemical,
mircobiological or hydrogeological makeup of the well
walls or some depth into the formation.
The drilling equipment itself may present a source of
contamination. Safeguards, such as steam cleaning of
the drill rig and equipment, may not be sufficient when
investigating trace organics and buying all new equipment
for each hole is ludricous. Although hydrologists insist
upon all downhole equipment being steam cleaned be-
tween each boring when drilling in a hazardous or po-
tentially hazardous site, the potential exists for some areas
of the drilling equipment to be missed during any reason-
able steam cleaning operation.
For this reason, investigators generally attempt to drill
in the upgradient zone first and progress toward the areas
which are assumed to contain the greatest potential for
143
-------
144 SAMPLING
contamination. Additionally, it is a very rare occurrence
that a drill rig utilizing hydraulic systems does not leak
considerable quantities of hydraulic fluids onto the
ground or into the well bore.
When a drilling program is designed to drill through a
contaminated zone into a lower, possibly less contam-
inated zone, interaquifer communication may create an
avenue of contaminant transfer. In this case extreme cau-
tion and careful placement of intermediate casings will be
required. It is for this reason that for all investigations
on hazardous wastes sites, the authors feel, the addi-
tional expense of maintaining an experienced contaminant
hydrogeologist on the rig at all times is warranted.
Another common source of well contamination occurs
as a result of surface contamination. Wells that are not
completed or sealed sufficiently may allow surface water
to be transported down along the casing material to the
sampling point. Even good development techniques which
may be sufficient to remove all traces of inorganic con-
taminant may not be sufficient to eliminate the existence
of trace organics.
Many authors have discussed the merits and problems
associated with the various drilling techniques and as such
will not be examined here. Although certain techniques
may be favored by field personnel, no technique can be
exempt from being a contaminant source. Consequently,
it is imperative that a detailed log be maintained at all
times during the drilling operations. The drilling log
should note all unusual factors that may ultimately lead
to inconsistent results.
Piezometer or Well Construction Materials
Discussions by Pettyjohn, et al.(]> have compared the
merits of preferred well construction material versus the
cost of construction versus the quality of data. However,
notwithstanding the arguments regarding preferred ma-
terials, contaminants may also be introduced into wells
during the installation of the piezometer or casing ma-
terials.
A common practice during installation of a well is to lay-
out the well materials along the ground and either install
it en masse or in discreet sections. This generally results in
a considerable quantity of surface material being installed
into the borehole along with the piezometer material. The
authors have observed several cases in the field where
chemically inert well materials and thermally welded
joints were utilized in an attempt to limit the introduction
of contaminants. However, before installing these ma-
terials they were stored along the drill rig prior to installa-
tion. This combined with the use of sand filter packs
from local quarries can result in very confusing results
being obtained from samples collected in these "so
called" monitoring wells.
The preferred well materials were described by Petty-
john, et al. as being, in order of acceptance:
•Glass 'Polypropylene
•Teflon Other plastics and metal
•Stainless steel "Rubber
This list of preferred materials must also be considered
from a cost benefit point of view. Each particular site
must be examined as to what level of risk may be intro-
duced when one utilizes a more reactive material for well
construction. This is particularly significant for a consul-
tant who must justify to his client the additional costs in-
volved in utilizing a more inert material, such as glass or
teflon.
It is essential to evaluate each material type against the
possible effect on the trace organics being evaluated. In
any case where a more reactive material is utilized in well
construction, one well of the same construction material
should be placed outside the contaminated zone to provide
background data on the input of trace organics from the
uncontaminated aquifer and the influence of the well ma-
terial. Moreover, the well drilling and well construction
methods for the background well should be comparable to
that used to install the monitoring well.
During the production of steel casing, considerable
quantities of oils and solvents are necessary at various
stages. If trace amounts of these materials remain ad-
hered to the casing during installation, contaminants may
be introduced either directly or indirectly. Besides the ob-
vious case of oil from the casing getting in the sample, an
oil-coated pipe may prevent the complete sealing of the
aquifer from either the influence of surface contaminants
or intercommunication with another aquifer.
Sampling Equipment
The materials used to construct organic sampling
equipment also must be considered. Where possible, rela-
., ADSORBENT TUBE
CLASS TUBE-
i" PRESSURE TUBE
TEFLON PISTON -
36"
TEFLON OR GLASS
PISTON CHAMBER
TEFLON CONNECTOR
1" RETURN TUBE (TEFLON)
-)" PIEZOMETER TUBE
(PLASTIC)
• CHECK VALVES
(TELFON)
SCREEN (TEFLON)
Figure 1.
Dames & Moore Positive Displacement Pump
for Volatile Organics
-------
SAMPLING 145
lively inert materials are recommended to limit the pos-
sibilities of bleeding organics into the sample or adsorbing
organic solutes. The preferred materials are similar to
those outlined above for piezometer materials. Pump and
sampling devices that allow groundwater to contact metal,
rubber or lubricants should not be considered when
sampling for organics.
Some sampling equipment, by virtue of its design, may
create problems with respect to interwell contamination.
Bailers, for example, especially those with triggering de-
vices, are extremely difficult to ensure that they are com-
pletely cleaned after each well has been sampled. Further,
for many clients it cannot be considered cost effective to
dedicate separate samplers for each well.
The uses of adsorbing tubes and positive displacement
(in-hole pumps) can greatly improve the efficiency of the
sampling equipment. Pettyjohn, et a/.(" and others have
investigated the use of various adsorbing materials for the
extraction of nonpurgeable organics from groundwaters.
Dames & Moore is currently investigating the use of ad-
sorbing columns on the return tube of a positive displace-
ment type pump (Figure 1). The authors feel this would
essentially eliminate the contamination of groundwater
samples that are subjected to various sampling devices
and storage jars, before arriving at the laboratory for
analysis. Additionally, the sample would remain under a
positive pressure until being adsorbed. This positive dis-
placement pump will be described in more detail elsewhere
in this paper.
Sampling Technique
Water sampling techniques should be designed to meet
the primary objective of a sampling program: to obtain a
representative sample of the fluid flowing in the aquifer.
(That is, obtain a sample retaining the physical and chem-
ical properties of the aquifer fluid.) Preservation methods
and in situ tests will aid in measuring these properties but
are meaningless if the sample of well water is not repre-
sentative of the actual aquifer.
Water which has stagnated in the well bore has had
ample opportunity to change physical and chemical
properties such as temperature, pH, dissolved oxygen,
total dissolved solids, etc.<3) Volatile organic contaminants
and dissolved gasses may have either volatilized or effer-
vesced in the well bore. Stratification of the water ac-
cording to temperature and specific gravity may lead to
different results being obtained from samples collected at
different depths, none of which may necessarily be repre-
sentative of the aquifer.
Consequently, prior to taking a sample, it is necessary to
evacuate the well of the stored water in order to allow
aquifer water to flow into the well. The EPA recom-
mends evacuating at least two exchanges of water prior
to sampling.(4)
The well storage water may be removed by bailing,
pumping from the surface, air lift pumping, or by pump-
ing with a submersible pump. However, when pumping,
care must be taken not to overpump the well and allow
excessive fines to be drawn into the well. This results in
increased turbidity and possible damage to the gravel
pack.
Care must also be taken in the selection of the pump-
ing method and materials. This selection may be dependent
upon the parameters to be measured and on the fluid
composition.
Bailers that resemble long, narrow buckets are avail-
able, or they may be more sophisticated and allow. a
sample to be collected from a specific depth. For small
wells with low yield, evacuation by bailing is a feasible
method of obtaining aquifer water. Bailing, however,
becomes labor and time intensive for larger, deeper wells
with large volumes of storage. In such cases, some sort
of pumping arrangement is more economical with the
bailer only utilized in obtaining a sample.
A sample which is obtained by bailer is exposed to
the atmosphere during transfer to a sampling jar. This
may destroy the test credibility for volatile contaminants
and dissolved gasses. Additionally, the sample is exposed
to descending pressures as it approaches the surface.
Operation on the more sophisticated bailers becomes
difficult in freezing weather as the petlocks and check
valves freeze and often samples are lost. Bailers also in-
troduce a certain amount of dissolved oxygen which, in
turn, may affect parameters sensitive to dissolved gas com-
position such as pH, alkalinity and redox-dependent trace
metals.
Pumps used for well evacuation include suction lift,
airlift and submersibles. When using any type of pump, a
foot valve or check valve is required to prevent con-
tamination of the well water by water which has been in
contact with the suction hoses or the pump itself.
The suction lift pumps are limited as to the depth from
which they may pump (approximately 24 ft). When the
water level in the well is drawn down beyond the limit of
the pump, confidence that a complete exchange of water
has taken place is reduced.
Submersible pumps, on the other hand, are not limited
as to the depth of the water level from which they can
pump, as are the suction lift pumps. However, the use of
submersible pumps involves the installation and with-
drawal of the pump from each well during each sampling
episode, assuming that cost efficiency dictates that the
same pump be used for several wells.
In the case where the pump is used for multiple wells,
there may be contamination from surface soils which ad-
hered to the hose during handling. Withdrawal of a pump
with a check valve installed becomes laborious, especi-
ally for deep settings and requires the use of additional
supporting equipment and extra labor. Additionally, sub-
mersible pumps require the use of an outside power source
requiring additional bulk to be transported during each
sampling period.
Air lift pumping is a relatively old method that is
gaining increasing attention. This method involves forcing
air down a tube inside a casing with a resulting air-water
mixture rising up in the annulus.<5) Depending upon the
applications, this method is an ideal way to evacuate a
well in a very cost efficient timely manner. Equipment is
readily obtainable and relatively inexpensive. However,
testing by Dames & Moore has indicated that this method
generally results in increased dissolved oxygen (D.O.)
-------
146 SAMPLING
values which in many cases have taken several days to re-
turn to prepumping levels. Additionally, gas stripping and
pH changes may also occur.
Other pumping techniques may also alter the chemical
properties of the water through the introduction of dis-
solved gasses. Suction lift pumps subject the water to nega-
tive pressures which will affect concentrations of dis-
solved gasses. The water may also be contaminated
through contact with the pump and the hose.
An experiment was designed and performed by Dames
& Moore to evaluate the cross contamination potential by
using the same submersible pump for several wells. The
experiment and its findings are discussed in a later section.
A positive displacement pump of the type discussed later
in this paper has the advantage of being suited for samp-
ling both volatile contaminants and dissolved gasses. The
sample is maintained under a positive pressure and is kept
from the atmosphere until it reaches the surface. The dis-
charge tube may be connected to a syringe or an adsorb-
tion column as in Figure 1 to obtain a sample without at-
mospheric exposure.
The positive displacement pump has the added advan-
tage that it may be installed at various levels within the
same borehole, sampling from a small screened section
below the pump. The other sampling techniques discussed
above are not three-dimensional, that is, it cannot be de-
termined from what level the contaminants are entering the
well bore. By installing several positive displacement
pumps in a single borehole, the third dimension of con-
taminant plume can be measured. In Table I, adapted
from Barker'2' a comparison of several methods currently
employed for the sampling of organic compounds is made.
Sample Preservation
Once the problem of how to obtain a sample has been
solved, the next dilemma involves what to do with the
sample. Normally the samples are submitted to the analy-
tical laboratory in prerinsed amber glass bottles with tef-
lon lined lids. For many of the organics holding times are
critical and unless the laboratory receives the samples
within the allotted time frame, the quality of the resultant
analyses may be open to scrutiny.
Table I.
Characteristics and Evaluation of Various Methods Used to Sample Groundwaters
for Organic Contaminants [Adapted from Barker 1981]
Method
bailer
2
high-lift
pump
special
piston pump
4
Materials Contamination Depth
Problems Limit
(ft.)
T, M Min
T, G Min
M, R, T Min-Sli
Min.
Piez.
Dia. (in. )
1
2
2.5
1 C
Air
Contact
Y
N
N
Degassing-
Loss of
Volatiles
minor
minor
minor
air squeeze
M, R, T
air-lift pump T, M
piston pump M, R, T
jet pump M, R, T
Sli-Max
Min
Sli-Max
Sli-Max
1.5
1.25
2
1.5
minor
major
minor
minor
submersible
pump
6
suction pump
and flask
peristaltic
pump
positive
displacement
pump
M, R, T
T, G
T, P
P, T
Max 4
Min. 25 0.25
Min-Sli 25 0.25
Sli 1.5
Mjterials— T-teflon, M-melal; G-glass; P-plasuc; R-rubber.
1 teflon extruded tubing with teflon rod end; glass marble check valve.
: all glass-icflon. 2-siage pump driven bv high purnv N7; Tomson el al., 1980. Groundwater, IS. pp
N minor
N major
N major
N minor
Materials:
T - teflon
M - metal
G - glass
P • plastic
444-446. R - nlbbtr
1 metal with rubber O-rmg seals, posilne displacement pump driven by compressed gas; Signor (1980); similar to Bennett pump
4 rubber and metal positive displacement pump driven by compressed gas; Middelburg pump (Tole Devices Co Ltd., P.O. Box 456, New
Albam. Ind. )
? polveihvleneor teflon lubing and metal air lift pump; such as Johnson Water Sampler (Johnson Div., P.O. Box 3118, St. Paul. Minn.).
6 pump placed behind flask so thai \>aier contacts onh tubing and glass; hand vacuum pump or peristaltic pump can be used.
-------
SAMPLING
147
Basic holding times and recommended sample volumes
were published in the Federal Register*® and are repro-
duced here at Table II.
During the holding period for the samples, losses or al-
terations of organics may result from any of a number of
causes including:
•Adsorption onto glass jar walls or particulate matter
•Precipitation or co-precipitation
•Biological transformations
•Chemical reactions (oxidation)
Generally, filtering of samples prior to analysis is not
recommended. Precipitates that may have formed during
the holding period must be processed along with the water
during analyses. This limits the possibility of some con-
taminant being adsorbed onto particulate matter and being
lost during the filtration. Consequently, very careful
piezometer construction and development prior to any
sampling is necessary to reduce the concentrations of par-
ticulate matter that enters a well.
EXPERIMENT ON
PUMP CROSS-CONTAMINATION
An experiment was designed by Dames & Moore person-
nel specifically to identify and solve the problem associated
with cross-contamination resulting from the use of the
same submersible pump in several wells. The tests were
made on a site with known PCB contamination. PCB's are
a compound with known affinity for plastics. Two pumps
were used in the experiment, the first a 0.5 hp submersi-
ble pump with plastic impellers, the second was an all
stainless steel 0.75 hp submersible pump with stainless
Table II.<6)
Sample Volumes, Preservation and Holding Times
for Organic Priority Pollutants
(from Federal Register, December 18,1979)
Sample
Volume Preservative
Measurement (ml) @4°C
BOD
COD
chlorinated
organic
compounds
organic carbon
extractables
(except phenols)
extractables
(phenols)
purgeables (halo-
carbons and
aromatics)
purgeables(ac-
rolim and
acrylonitrite)
pesticides
cool
cool, H,SO4 to
pH<2
cool, 0.008%
1000
50
25 to 500 cool, h,SO, to
PH<1
1000 cool, 0.008%
5000
cool, H-SO. to
pH<2;0.008%N2S203
Max. Holding Time,
Before Analyses or
Extraction
(days)
2
28
28
7
7
20 to 500 cool, 0.008%
N?2S2°3
20 to 800 cool, 0.008%
1000
14
cool, 0.008%
a—should only be added in the presence of residual chlorine
steel impellers. Both pumps utilized the same collapsible
plastic hose for discharge.
The purpose of the experiment was to estimate the
amount, if any, of potential cross-contamination be-
tween wells which occurs by utilizing the same pump in
the sampling procedures. The contribution of the plastic
hose was also to be estimated. Both pumps were used in
the experiment to determine the advantages of using all
stainless steel parts, as opposed to plastic.
The pumps were used to evacuate a 120 ft deep well
known, to be contaminated with PCBs. The pumps were
then placed in a 55 gal drum of potable water. Samples of
the potable water were obtained to develop background
levels from the drum prior to placement of the pump. The
hose was connected and the discharge directed to recir-
culate water within the drum. The pump was then run for
1/2 hour, recirculating the potable water. The water from
the drum was subsequently sampled.
In order to isolate the contamination from the pump,
the above procedure was repeated, but without the hose
being connected to the pump in the drum. The first slug of
water discharged from the pump was subsequently sam-
pled. The hose was again connected, and the first slug of
water through the hose was then sampled. Once the ex-
periment was completed for both pumps, the samples were
transmitted immediately to a laboratory for analysis.
The results of the analyses are shown in Figure 2. A
significant amount of contamination was measured in the
test recirculating water using the plastic impeller pump
(approximately 11.4 ug/1). The samples isolated from this
pump showed approximately 3 ug/1 came from the pump,
and a similar amount from the hose. The stainless steel
pump produced insignificant levels considering the back-
ground levels of the "potable" water. When the hose was
connected, however, a marked contamination was noted,
approximately 3 pg/1. From this very limited experi-
ment, the authors have concluded that the plastic im-
pellers appear to have an affinity for PCBs and may be
responsible for a small amount of cross-contamination.
Additionally, a plastic discharge hose, although very con-
venient to use, does contribute to cross-contamination.
Therefore, where multiple wells must be sampled, the
choice of pump and hose materials can be significant for
low concentration contaminants.
DISPLACEMENT PUMP DEVELOPMENT
To limit the problems outlined in the above sections,
the authors believe that a monitoring program allowing an
individual positive displace pump to be installed into
each well is truly needed when dealing with the problems
of trace organics. Such a positive displacement pump is
currently being developed by Dames & Moore and is
shown in Figure 3.
This pump consists of a 1.75 in diameter tube with an
inside piston. Connected to the piston chamber is a short
section of well screen. A check valve between the well
screen and the piston chamber allows one way flow into
the chamber. When pressure is applied to the piston via
-------
148
SAMPLING
a 0.25 in tube at the top of the chamber, the check valve
closes and the fluid in the chamber is forced through a
second valve out a 0.5 in discharge tube. The present de-
sign will enable one liter of fluid to be pumped per stroke
of the piston.
A third tube is connected to the 0.5 in tube between
the screen and the piston chamber just below the check
valve. This permits monitoring of the hydrostatic pres-
sure within the screen.
The pump is suited for installation in various materials,
by attaching different well screen sections. By placing a
filter pack around the screen, problems from silt migra-
tion may be reduced, if not eliminated.
1" RETURN TUBE
15
10
0 "
13 6
50
BACK9ROUNO POTABLE WATER
1!!!!!!;=*: = 2.2
PUMP IMPELLER TYPE
Figure 2.
Cross Contamination Test Results for the Plastic Impeller
Pump and the Stainless Steel Impeller Pump
The pump may be constructed of teflon or plastic, de-
pendent upon proposed operations. All pressure connec-
tions are made with swaglok fittings, available in teflon.
CONCLUSIONS
This paper was prepared to assist the reader in under-
standing the complexities involved in sampling ground-
water for organic contaminants. Without careful attention
to all phases of a sampling program, the results of the
sampling program are worthless. This may result in a client
performing much additional investigative drilling and
sampling to answer questions that may have resulted from
erroneous procedures.
Additionally, in many areas low levels of trace or-
ganics may remain undetected because of poor sampling
i" PRESSURE TUBE
PISTON
36"
PISTON CHAMBER
•i" PIEZOMETER TUB£
• CHECK VALVES
SCREEN
Figure3.
Dames & Moore Positive Displacement Pump
programs. If allowed to remain undetected, these low
levels of trace organics may create a serious health hazard
in the long term or at least result in a much more costly
investigative program.
Although rapid developments are currently taking place
in the field of organic sampling, these are currently only
in developmental stages and as such, modifications and
improvements are constantly being made. Consequently,
proven standard cost effective techniques for sampling
organics still lags far behind the analytical laboratory
techniques.
REFERENCES
1. Pettyjohn, W.A., Dunlap, W.J., Cosby, R., andKeely,
J.W., Groundwater, 19, 1981, 180-9.
2. Barker, J.F., Short Course, Field Methods in Con-
taminant Hydrogeology. Department of Earth Sci-
ences, University of Waterloo, Waterloo, Ont., 1981.
3. Schuller, R.M., Gibb, J.P., and Griffin, R.A.,
Groundwater Monitoring Review 1 #1, 1981, p. 42.
4. U.S. Environmental Protection Agency, "Procedures
Manual for Groundwater Monitoring at Solid Waste
Disposal Sites," EPA-530/SW-611, 1977.
5. Morrison, R.D. and Brewer, P.E., Groundwater Mon-
itoring 1 #1, 1981, p. 52.
6. U.S. Environmental Protection Agency, Federal
Register, Dec. 18, 1979.
-------
SPECIAL SAMPLING TECHNIQUES USED FOR
INVESTIGATING UNCONTROLLED HAZARDOUS WASTE
SITES IN CALIFORNIA
HOWARD K. HATAYAMA
California Department of Health Services
Berkeley, California
INTRODUCTION
The primary purpose of this paper is to briefly dis-
cuss some special sampling techniques used to character-
ize the nature and extent of contamination at selected un-
controlled hazardous waste sites in California. These in-
clude techniques for sampling soil, waste, groundwater
and gases. The methods are discussed as they were applied
to specific sampling problems at four sites: (1) Occidental
Chemical Company, Lathrop, Ca., (2) McColl Site, Full-
erton, Ca., (3) Mola Development Site, Huntington Beach,
Ca. and (4) General Electric Company, Oakland, Ca.
The techniques for soil and waste sampling discussed
are: (1) Hollow-stem augering with drive sampling, (2)
Rotary drilling with drive sampling, (3) Bucket augering
and (4) Trenching. The two techniques for groundwater
sampling discussed are: (1) Single completion wells with
submersible pump sampling and (2) Multiple cased wells.
Installation of gas collection wells is covered under gas
sampling. Most of the information presented herein is de-
rived from the files of the California Department of
Health Services.
SITE DESCRIPTIONS
Occidental Chemical Company, Lathrop, Ca.
The facility is located in the Central Valley of Cal-
ifornia approximately 50 miles south of Sacramento, the
State capitol. It is built on the flood plain of the San Joa-
quin River with useable groundwater from 20 feet down
to 300 ± feet. It is a fertilizer manufacturing and pes-
ticide formulating plant handling a wide variety of organo-
halogen, organophosphate and carbamate type pest con-
trol agents. For a period of 10-15 years, the nemnatocide
DBCP, (1, l-dibromo-2-chloro-propane), were also man-
ufactured there.
Wastes from fertilizer manufacturing consisted largely
of a gypsum slurry which was ponded until just recently.
Wastes from the pesticide formulating and manufactur-
ing activities consisted of equipment washes, unrecover-
able bad batches, run-off, raw material containers and re-
turned obsolete off-specification consumer products. Until
the mid-1970s, all the liquid wastes were discharged to
an unlined ditch and pond.
In the 1950s and 1960s, the solid wastes were either
burned in open pits or buried in trenches. These trenches
were not apparent from observation of the ground surface
at the start of investigations. These practices resulted in
gross contamination of surface and near surface soils, and
of the ground water to depths greater than 200 ft.
McColl Site, Fullerton, CA.
This site is located in Southern California about 17 miles
southeast of downtown Los Angeles. It is a seven acre site
bounded on three sides by homes and on the fourth by a
golf course. It consists of six sumps covering about four
acres containing an estimated 50,000 yd3 of acid petrol-
eum tars.
The wastes were generated by the local petroleum re-
fineries during the 1940's from high octane fuel produc-
tion. The material is a black tarry substance with a firm
crust that softens upon warming by the sun. On very hot
days, the waste flows and up-wells. It contains large quan-
tities of sulfur dioxide, benzene and other hydrocarbons,
tetrahydrothiophenes and arsenic. It is highly acidic (pH of
the slurry is less than one). When the crust is broken, large
quantities of sulfur dioxide and other extremely odorous
compounds are released.
Mola Development Site, Huntington Breach, Ca.
This site is located along the Southern California coast
25 miles southeast of downtown Los Angeles on a plateau
overlooking a wildlife preserve. It is bounded on four sides
by homes and apartment houses.
It was originally a stream bed used as a sand quarry.
Subsequently, it was used as a disposal site for acid petrol-
eum tars and other petroleum wastes during the 1940s.
Thereafter, it was operated as a permitted site for drill-
ing muds and building demolition wastes. It was estimated
to contain approximately 100,000 yd3 of petroleum wastes
and 200,000 yd3 of drilling muds and building demolition
wastes.
The petroleum waste characteristics are very similar to
those of the McColl Site wastes, but somewhat diluted by
water. Similar types of gases were being generated by the
wastes along with significant quantities of methane and
carbon dioxide. The rate of gas generation was slower
although the gases were no less odorous. The absorptive
149
-------
150 SAMPLING
capacity of loosely compacted building demolition wastes
and the pooling of water on the layers of drilling mud
caused the tarry petroleum material to up-well in pools on
the surface. The up-welled material became more flowable
on warming.
General Electric Company, Oakland. CA.
This facility is located in Northern California, near
downtown Oakland. It is bounded on one side by homes
and on another by a food processing firm that draws water
from 150-200 foot depth. The water table aquifer is at
10-20 ft. The facility was a transformer repair shop that
handled PCB (polychlorinated biphenyl) and other types
of transformer oils in bulk quantities.
Spills and leaks of PCB's and other transformer oils
resulted in significant soil contamination and a layer of
PCB containing oil up to 6 inches thick floating on the
water table.
SOIL AND WASTE SAMPLING TECHNIQUES
Hollow-stem Angering With Drive Sampling
This technique (Figure 1) employs a continuous flight
hollow-stem auger to bore to the desired depth in the soil
or waste. The sample is then obtained using various types
of drive sampling devices. The auger is usually plugged
during drilling and the plug is removed to advance the
sampler ahead of the auger.
The unlined ditch and ponds at Occidental were sampled
at surface, 2, 5, and 10 ft depths using this technique.
The hollow-stem auger not only provided support for the
bore holes but also controlled sloughing of the more high-
ly contaminated soils into depths of less contamination.
A split-barrel sampling device was selected for obtaining
samples because it allowed the taking of a relatively un-
disturbed core, the physical inspection of the sample, and
efficient transfer of the sample for splitting and con-
tainerization. The top 1-2 in. were generally discarded as
a further step to control the effects of sloughing.
The sampler was thoroughly cleaned with scap, water
and acetone between each use. Sampler decontamination
procedures generally depends on the type of waste and the
contaminants of concern. In this case, since contamina-
tion of soils with organics and heavy metals was antici-
pated, both an aqueous and organic decontamination
solutions were used.
Rotary Drilling with Drive Sampling
This technique (Figure 2) is different from the pre-
vious one only in the drilling method. The desired depth
is reached by boring with a rotary drilling rig using some
type of fluid to circulate the cuttings and support the bore
holes. The sampler is advanced ahead of the drill bit.
This system was used to obtain discrete core samples
from the McColl Site petroleum waste sumps. A "dry"
drilling technique (augering or compressed air rotary)
could not be used because of the large quantities of sul-
Soll or M««tt
Figure 1.
Hollow Stem Auger with Drive Sampler.
fur dioxide and other noxious gases released by
ing the crust and maintaining an open hole. The ..._
ing winds carried these gases directly into the tdjoiaiaf
homes. Water was found to be an adequate fluid for
trolling these emissions.
Because of the flowable nature of the wastes, •
weight trailer mounted rig with large surface supports
necessary. The cleanliness of the split barrel
could not be maintained because of the use of th
fluid. The drilling fluid may also have penetrated the'
disturbed waste material around the bore holes.
effects were controlled to a certain extent by _
ing the top 1-2 in. of the sample. Any significant _
ture in the sample was noted. The petroleum waste
terial was found to be relatively impermeable under the
circumstances of the sampling.
Bucket Auger Drilling and Sampling
A bucket auger (Figure 3) is a cylindrical (12-18 in.
diameter) bucket with a swing-open bottom. A set of
-------
SAMPLING 151
Figure 2.
Rotary Drilling with Drive Sampling.
Figures.
Bucket Auger.
angled teeth is fixed to the bottom for augering into the
soil. The teeth guide the sample into the bucket as the
bucket is rotated by a turntable. Depending on the length
of the bucket, samples can be collected at 1 to 2 ft inter-
vals. Once a sample is obtained, the bucket is withdrawn
from the bore hole, swung out to the side and the sample
is released onto the ground surface by opening the bottom.
The large sample size generally allows taking several sub-
samples from each depth.
This technique was used at the Mola Development site
with the primary objective of obtaining samples of the
waste material deposited below the building demolition
wastes and drilling muds. From previous drilling exper-
ience at the site, it was determined that other techniques
were not adequate for penetrating the stubborn over-
burden and obtaining samples of the waste. With much
effort, it was possible to remove the concrete rubble over-
burden and obtain depth specific samples of the solidi-
fied petroleum waste material.
In some cases, the waste was in a semi-solid or slurry
form and it was not possible to obtain representative depth
specific samples. A composite sample was taken of these
borings and the bottom of the waste was roughly de-
termined by the sandy stream bottom.
No attempt was made at decontaminating the auger
between samples because of the time involved and the large
sample size. Sloughing of the walls of the bore hole could
not be adequately controlled. The rate of gas genera-
tion was significantly lower, here than at the McColl Site
and the prevailing winds provided higher dilution such that
gas control measures were not necessary during sampling.
Trenching
Trenching for purposes of locating and sampling buried
hazardous waste materials usually entails the use of an ex-
tendable back-hoe with a fully enclosed operator com-
partment (Figure 4). The unknown nature of the buried
material requires strict attention to the health and safety
of the sampling personnel. It is critical that as much in-
formation as possible on the location and possible nature
of the buried material be gathered before trenching is be-
gun. A series of cross-trenches is usually dug to intersect
the suspected burial area. The trench walls and diggings
are logged and sampled using manual devices.
Investigation of historical records at Occidental revealed
the existence and general locations of the burning and bur-
ial trenches. After marking the approximate locations,
trenching began with the intention of locating the trenches
more precisely and sampling their contents. The 3-4 ft of
relatively uncontaminated overburden was placed in sep-
arate piles while the contaminated soil and wastes were
placed on polyethylene sheeting.
A significant fraction of the wastes were found in the
buried and saturated zone. The wastes consisted of par-
tially full pesticide containers, empty drums and other sol-
id wastes related to pesticide formulation. The volume of
waste uncovered and the fact that much of the pesti-
cides were leaking directly into the groundwater man-
dated a change of strategy.
-------
152 SAMPLING
The wastes were loaded into trucks for immediate dis-
posal classified as extremely hazardous pesticide wastes
without benefit of analytical results. Containers with re-
sidual pesticides were overpacked and empty containers
were placed in dumpsters. Contaminated soil was sampled
and analyzed prior to disposal.
Although circumstances in this case dictated that the
exploration and removal process occur concurrently, it is
generally more appropriate to separate these two. This
allows time to develop as much information as possible to
safely and properly remove the wastes without incurring
undue risks to workers, the surrounding community, and
causing further contamination of the soils and ground-
water.
GROUNDWATER SAMPLING
Single Completion Wells with
Submersible Pump and Packer
As shown (Figure 5), this technique involves using a
submersible pump attached to an air inflatable rubber
tube to obtain a representative groundwater sample. After
positioning the pump/packer assembly at the desired
depth, the packer is inflated to form a seal between the
screened section and the rest of the water column. The
system which also includes a Teflon ® tube for sampling
is then flushed with 3-5 volumes (volume of the screened
section) before a sample is taken. The sampling tube is
not, however, shown in Figure 5.
Air, Electrical, Control Lines
Steel Casing
Cement Grout
Air-Inflatable Packer
Gravel Packing
Submersible Centrifugal
Pump
Figures.
Pump and Packer Assembly.
Figure 4.
Trenching.
Oil Layer
I
Shallow Aquifer
.Impermeable -;
* C. Stratum _*_"
Aquifer * »
&
:•;»
$
,PVC Well Casing
m
-Cement Grout Casing II
Cement Grout Casing 11
-Hell Screen
-Gravel Packing
Figure 6.
Multi-Cased Well.
-------
SAMPLING 153
Gas Sampling Valves
,— Metal Cap
Cement Grout Cap
Gas Collection Port
Figure?.
Gas Sampling Well.
The permeability of the soils in the area, analytical data
from surrounding wells, and the nature of the operations
at Occidental's facility indicated that groundwater was
contaminated with organics and inorganics to significant
depths. Based on available hydrogeological and pilot well
data, further sampling locations, each with three single
completion wells at various depths down to approximate-
ly 200 ft were selected. This type of well construction
allowed isolation of the most significant water bearing
strata. The packer further isolated the specific zone of in-
terest while limiting the amount of water flushed and con-
trolling any mixing of stagnant water in the well column
with water which is more representative of the aquifer.
The submersible centrifugal pump with relatively inert
component parts allowed sampling for purgeable organ-
ics such as DBCP and EDB (1, 2-dibromoethane) along
with other less volatile constituents. The depth of sampling
and these volatile components prohibited the use of other
sampling methods (bailing, vacuum pumps, and air lift
pumps). The entire sampling assembly was flushed with
distilled, deionized water between each location instead
of providing dedicated tubing for each location to provide
further control over the possibility of cross-contamination.
Multiple Cased Wells
This is more a method of well construction rather than
of sampling (Figure 6). A large diameter boring is made
through the upper, highly contaminated aquifer into the
first impermeable zone. This boring can then be cased and
grouted with cement or plugged with bentonite clay. A
smaller boring is then made through the casing or the seal
into the next aquifer where a single completion well is
constructed.
Multiple cased wells were employed at the General Elec-
tric facility where the PCB contaminated oil layer on the
water table threatened to cause contamination of the deep-
er aquifer during drilling. Representative depth specific
soil samples were also obtained from zones below the
water table aquifer using this drilling technique. Such an
approach should be considered in all cases where the prob-
ability exists of contaminating to a significant extent
previously uncontaminated zones.
AIR SAMPLING
Gas Well Installation and Sampling
The landfill gas sampling probes shown in Figure 7 are
encased in a perforated metal casing and lowered into a
bore hole. The annulus is packed with gravel and sealed
with cement. All parts are constructed of steel or brass to
minimize adsorption and desorption effects.
Gas collection parts are provided at two levels for char-
acterizing gases at these levels and to provide for the poss-
ibility of inundating the deeper probe with leachates.
Samples of the collected gases are obtained by pumping
through various types of adsorbents (i.e., activated car-
bon, or Tenax ® , impingers with selective solutions or by
filling Tedlar ® bags.
This approach was necessary at the Mola Develop-
ment site to evaluate the rate of generation of the noxious
gases, characterize them more fully and evaluate the po-
tential for subsurface migration to the surrounding com-
munity. The information was also necessary to develop
appropriate mitigation measures. These gas collection de-
vices were installed in some of the borings made by the
bucket auger and the well was allowed to equilibrate for
several weeks. Composite gas samples representing the en-
tire column were taken using Tenax ® adsorbent for
organics and impingers for inorganic vapors. This method
allowed extensive characterization of the gases generated
on-site although it was not sufficient to fully evaluate the
rate of gas generation for purposes of planning excava-
tion.
CONCLUSIONS
Technology has generally been available for sampling at
the sites discussed and other sites. Investigations at un-
controlled sites often require novel applications of these
technologies.
Maintaining quality control is a key factor in any sampl-
ing method selected, especially when addressing low levels
of highly toxic contaminants. However, compromises
must be made in some cases when sampling waste material.
Care must be taken in any investigation that the act of
sampling does not cause significant contamination of pre-
viously uncontaminated soils, aquifers or air space.
-------
THE STRATIFIED SAMPLE THIEF— A DEVICE
FOR SAMPLING UNKNOWN FLUIDS
MICHAEL G. JOHNSON
Mason & Hanger-Silas Mason Co., Inc.
Leonardo, New Jersey
INTRODUCTION
The primary reason for the development of the strati-
fied sample thief was the problem of sampling the fluids
recovered by the oil spill recovery devices tested at the
EPA OHMSETT test facility. The method used since the
facility opened and continued until the 1980 test season
was as follows:
•The test fluids were collected in containers on the auxili-
ary bridge during a test
•The container fluid level was measured to obtain total
volume collected
•The free water was decanted from the bottom of the con-
tainer
•The remaining fluid was remeasured
•The sample was mixed (emulsified) using an electric
motor and propeller
•A small sample of the fluids was taken while mixing
continued
•The sample was taken to the lab where it was centri-
fuged to determine the amount of water in the oil.
Some of the problems associated with this procedure
were:
•The measuring, draining and remeasuring took time and
produced more numbers that could be lost of misinter-
preted
•The mixing caused severe emulsification that made the
refurbishment of the oil for reuse difficult and expensive
•The plastic containers used for the recovered fluids were
subject to damage and degradation by sunlight.
To eliminate some of these problems, many methods
for sampling were discussed and/or tried including:
•Discrete sampling using a rotary table with tubing and a
solenoid valve
•Discrete grab samples (this was similar to the rotary table
except that all operations were manual).
The rotary table discrete sampler was less than satis-
factory because:
•The fluid in the discharge line from a device could strati-
fy and allow a non-representative sample to enter the
tubing
•High viscosity fluids would not flow through the sampler
tubing fast enough to fill the sample bottles in the time
allowed for each discrete sample
•Debris sometimes plugged the tubing or solenoid valve
•In some test configurations, there was not enough pres-
sure to allow sufficient flow in the sampler system
•The electrical system constantly failed due to corrosion
and mechanical damage.
The manual discrete sampling was better from the
mechanical/electrical standpoint but it had problems with
stratification, low pressure, debris and high viscosity. An-
other disadvantage was the large amount of manpower
required.
Some of the sampling methods discussed for possible
use were:
•Vacuum bottles (similar to those used for blood sampling)
•Multiple valved tube
•Grain sampler tube
•Cork plugged tube (COLWASA).
These methods were all rejected because of anticipated
problems such as:
•Cost
•Complexity
•Debris problems
•Viscosity effects
•Sealing
•Susceptibility to damage.
STRATIFIED SAMPLER DEVELOPMENT
A solution to the problems inherent with the other
sampling devices or methods was finally arrived at after
eliminating numerous other ideas. It is called a strati-
fied sample thief.<'•2'3)
This sampler uses discs and wipers to hold the strati-
fied liquids in position while the tube is slipped past them.
(Figures 1 and 2.) The wipers keep the inside of the tube
from carrying portions of the upper fluid down into other
layers. They seal well enough so that the samplers can be
carried to and stored in the laboratory horizontally.
This procedure allows the transfer of the sample from
the sample thief to another container to be accomplished
in the laboratory. This is done by clamping the sample
thief vertically in a rack and removing the sheath.
Advantages
•Simplicity
•Inexpensive
•Ease of use
•Speed of operation
•Representative sample obtained (see Table I)
154
-------
SAMPLING 155
i
2
3
4
5
6
7
8
Spacer
Supporting Washers
Wiper
Sheath
Center Rod
Extension
Bottom Stopper
O-ring
Figure 1.
Stratified Sample Thief Construction
-------
156 SAMPLING
1
B. The outer sheath
is withdrawn to
expose the center
section.
A. The sampler with the
extension rod is placed
in the barrel through
the pour spout.
c.
The outer sheath
is slid down the
center section,
trapping the liquid.
D. The entire
sampler is
withdrawn
from the drum
with a repre-
sentative
sample enclosed
ffl
Figure 2.
Use of the stratified sampler
-------
SAMPLING 157
•Horizontal storage
•Useable by personnel wearing bulky safety gear
•Can be made in lengths to eight feet for deep containers.
Disadvantages
•Hard to clean
•Sheath material may be damaged by organic solvents
(glass sheaths may be better)
•Use in low overhead area restricted by rod extension
•Bottom o-ring swells after repeated usage, requiring re-
placement.
COLWASA SAMPLER
This device is similar to the COLWASA(4) device, which
is a tube with a cork or plug in the lower end which can be
pulled into the lower end by a line or rod attached to the
cork. The line or rod runs up through the inside of the
tube and extends far enough out of the upper end of the
tube so that it can be grasped. A sample is taken by lower-
ing the assembly slowly into the fluid with the cork lowered
so liquid can flow into the tube. When the assembly
reaches the bottom of the fluid container, the line or rod is
pulled up to seal the liquid inside the tube. On some
models, a locking mechanism has been added to lock the
rod in place.
The COLWASA sampler works well for homogeneous
fluids or for stratified liquids of low viscosity. Stratified
fluids with high viscosity can cause a problem due to the
rather thick coating that can build up inside the tube. This
coating can prevent the taking of representative samples.
The stratified sample thief offers a solution to this prob-
lem.
The 2.22-cm (inside diameter) sampler was evaluated in
22 different tests in a 500-gal barrel of salt water with
measured quantities of oil varying from 10 to 90 percent.
The results of tests with the Johnson sampler and (for
comparison) with grab samples taken after homogenizing
barrel contents with an electric mixer are shown in Table I.
ACKNOWLEDGEMENT
The work upon which this presentation is based was
performed by Mason & Hanger-Silas Mason Co., Inc.
pursuant to Contract No. 68-30-2642 with the U.S.
Environmental Protection Agency. The mention of trade
names or commercial products does not necessarily con-
stitute endorsement or recommendation for use. The opin-
ions or assertions contained herein are those of the au-
thors and do not necessarily represent the views of the
U.S. Government.
Table I.
Johnson Sampler Test Results on Oil/Water Samples
Actual Johnson
Test Cone. sampler
number (% Oil) (% Oil)
1 74.2 73.1
2 64.3 60.0
3 53.9 59.8
4 67.2 66.3
5 45.9 46.1
6 56.0 54.7
7 76.5 71.2
8 86.9 83.1
9 66.4 64.9
10 81.8 77.8
11 40.3 39.7
12 49.7 48.3
13 44.3 44.9
14 37.0 35.2
15 39.2 38.8
16 30.8 29.0
17 23.2 21.1
18 13.3 8.8
19 18.8 18.0
20 10.9 10.3
21 14.5 9.7
22 89.5 90.5
Grab
sample
(% Oil)
81.0
73.1
59.8
65.5
44.5
54.7
75.2
85.1
64.9
79.7
39.7
50.5
42.6
36.2
41.4
29.4
23.7
9.0
16.3
9.0
11.2
90.0
REFERENCES
1. Borst, M. "Johnson Sampler Accreditation Test" (un-
published) Report to U.S. Environmental Protection
Agency MERL-Ci, Cincinnati, Ohio, 1981.
2. Nash, J.H. "OHMSETT Evaluation of the Clean At-
lantic Associates; Fast Response Open Sea Skimming
System." (Unpublished Report to Clean Atlantic As-
sociates, c/o Halliburton Services, P.O. Box 1431,
Duncan, Oklahoma 73533) 1980.
3. Farlow, J.S., Griffiths, R.A., "OHMSETT Research
Overview, 1979-1980", In: Proceedings of the 1981
Oil Spill Conference, American Petroleum Institute,
Washington, D.C. 664-665.
4. deVera, e.R., Simmons, B.P., Stephens, R.D., and
Storm, D.L., "Samplers and Sampling Procedures for
Hazardous Waste Streams", U.S. Environmental Pro-
tection Agency (MERL-Ci) Report No. EPA-600/2-
80-018, Cincinnati, Ohio, 1980.
-------
THE COMPLEMENTARY NATURE OF GEOPHYSICAL
TECHNIQUES FOR MAPPING CHEMICAL WASTE
DISPOSAL SITES: IMPULSE RADAR AND RESISTIVITY
KEITH A. HORTON
REXFORD M. MOREY
LOUIS ISAACSON
RICHARD H. BEERS, Ph.D.
Geo-Centers, Inc.
Newton Upper Falls, Massachusetts
INTRODUCTION
Geophysical mapping at waste disposal sites has pro-
vided a direct comparison of results obtained from the
application of two different geophysical techniques: (1)
electrical resistivity (ER) and (2) electromagnetic impulse
or ground penetrating radar (GPR). A correlation be-
tween data acquired by these two methods is reported
in this paper. The two types of data can be used in a com-
plementary manner since the radar signatures give geo-
logical meaning to the resistivity and the resistivity can be
used to estimate radar system capabilities and predict the
depth of penetration of the radar signals in the site
geology."1
GROUND PENETRATING RADAR
The ground penetrating radar system, illustrated in
block form in Figure 1, is composed of a transmitter, a
receiver, an antenna (or antennas), an analog magnetic
tape recorder and graphic display unit. The transmitter
generates repetitive extremely short-time duration pulses
which are radiated into the ground from a broadband an-
tenna placed in close proximity and electromagnetically
coupled to the ground. The radar operates as an echo
sounding device. Electromagnetic pulses, reflected from
targets or interfaces within the ground, are detected by the
antenna, converted into voltage as a function of time
waveforms in the receiver, and displayed on a graphic re-
corder.
As the antenna is scanned over the ground surface, a
continuous profile of subsurface electromagnetic condi-
tions is printed on the graphic recorder. The location and
depth of subsurface targets are inferred from the graphic
record. Depth of penetration and velocity of propagation
in the ground are a function of the constitutive electro-
magnetic parameters of the soil. Maximum depth of pene-
tration ranges from tens of centimeters to hundreds of
meters.
The transmitter and high frequency receiving elec-
tronics are mounted in the antenna. This allows the use of
several different antenna configurations operating at dif-
ferent frequencies with the same radar control unit and
recording equipment. Even though each antenna config-
uration radiates a broad spectrum of frequencies, a par-
ticular antenna unit is designated by the center frequency
GROUND SURFACE
TRANSMITTED PULSE
TARGET
REFLECTED PULSE
Figure 1.
Block Diagram of a Ground Penetrating Radar System
of its spectrum. Typical radar center frequencies vary be-
tween 10 and 1000 MHz.
The recording equipment and control unit are mounted
in a van or other appropriate vehicle and the antenna is
towed behind the vehicle. Data are recorded on magnetic
tape and on the greyscale graphic recorder; the latter in-
formation being compressed dur to the high input data
rate. After completion of a survey, the magnetic tape data
are played back to generate full resolution hard copy for
analysis.
An example of a typical transmitted and received pulse
train is shown in Figure 2a while the graphic recorder
display of the same is shown in Figure 2b. The dark bands
occur at signal peaks, (both positive and negative), while
the narrow white lines occur at the zero crossings be-
tween peaks. Areas to be surveyed are defined by driving
stakes into the ground to establish a reference grid sys-
tem. This grid is used as the coordinate system for the
radar data and for any other geophysical surveys also
undertaken. Positional information is accurate to within
a few feet. Grass and weeds are usually mowed and brush
cleared to ease actual travel over the surface.
158
-------
REMOTE SENSING 159
Several preliminary scans are made at eacri site to cali-
brate the antennas with respect to the electromagnetic
characteristics of the specific soils. In this way, each sys-
tem configuration is optimized to provide the desired
depth of penetration and/or resolution. In addition, the
velocity of propagation of the EM signal and the dielec-
tric constant of the soil are calculated by measuring the
two-way travel time of the signal to objects at known
depths or by a series of wide-angle reflection from an ob-
ject or interface at an unknown depth.
SIGNAL
AMPLITUDE
I RECORDER PRINT
I THRESHOLDS
HORIZONTAL
TRAVEL
TRANSMITTED PULSE'
SURFACE •
•1
^^^^^^^"•^^^^^^^^•^^^^B^^^^W^^W^WWn^^^^^"^^*^^^^^^
._•£;«:
INTERFACE
SIGNAL
H
Figure 2a.
Sketch of a Single GPR Pulse and Reflections
as Seen by the Receiver.
Figure 2b.
Example of Profile Information as Displayed by
the Graphic Recorder.
-------
160 REMOTE SENSING
RESISTIVITY
Earth resistivity surveys have been used for many years
in exploration for groundwater and mineral deposits, and
in the study of engineering properties of earth materials.<2)
Equipment to measure apparent resistivity consists of a
controlled source of electric current and a device for
measuring the potential differences generated by the cur-
rent passing through the earth.
Typically, four electrodes are used in resistivity measure-
ments. The volume of a subsurface material influencing
the resistivity measurement is controlled by the spacing
and geometric arrangement of the electrodes. While any
array of four or more electrode contacts can be used in
studying earth resistivity, a relatively few electrode con-
figurations have become accepted as standard arrays in
practice. The most common electrode arrays used in the
resistivity method are shown in Figure 3.
Many factors go into the choice of array configura-
tion for a given problem. Susceptibility to geological noise,
ease of array movement and the nature of the assumed
structure are a few of the factors to be considered.
(a) Wenner Spread
(b) Schlumberger Spread
J f r r r f f r rrrrtrrrr—r~r—7—7—*
(c) Double-dipole Spread
Figures.
Common Electrode Configurations for Resistivity Arrays
RADAR/RESISTIVITY CORRELATION
Examples of radar signatures obtained at a frequency of
120 MHz are shown in Figure 4, for "broken slag" and
"clay" present at an uncontrolled chemical waste disposal
site; the corresponding resistivity profile superimposed
on the radar profile is also illustrated. Apparent resistivity
is plotted downward in order that conductivity may be
visualized as increasing upward. The effective depth of
penetration is greater in "slag" with a corresponding lower
conductivity and less in "clay" with a higher conduc-
tivity. Although conductivity is frequency dependent and
the two methods are operating at frequencies differing
about seven orders of magnitude, the relative effects of
radar signal attenuation and measured d.c. conductivity
(or resistivity) are shown to be similar when moving from
one material to another.
85 + 30 85 + 00
STATIONS
84 + 50 84 + 00
83 + 60
-10
50
-100
14-11
"BROKEN-
SLAG"
ft
GC
500
-1000
• "CLAY"- -I
A RESISTIVITY DATA POINTS
Figure 4.
Radar and Resistivity Profiles.
Radar Profile Shows Characteristic Signals of
"Broken Slag" and "Clay".
The relationship between the radar depth of penetra-
tion and the d.c. conductivity (resistivity) is illustrated in
Figure 5. Based on the radar range equation and a com-
plex index of refraction soil model, the maximum depth
of penetration of the radar system is calculated for fre-
quencies of 10, 100, and 300 MHz.(M'5)
Also plotted in this figure are the radar penetration
depths achieved at 120 MHz and d.c. conductivity (resis-
tivity) measurements obtained at the chemical waste site
illustrated in Figure 4. Considering the complexity of this
-------
f
sr
£
1000
II
r* ^^
•< a-
» a
§ 5
O r-t-
II
3- 2
w 3
jo 51
p C
O. B
Q3 O
S a
••
o.
'
100 -
L(m)
COMPLEX REFRACTIVE INDEX MODEL
(ROUGH PLANE REFLECTOR)
WATER
CONTENT
30%
C. (dry) = 4
SLAG 8 GRAVEL
CLAYS
-------
I—
o
z §
o °
S" 2
< S
y o
1000
O O
—i 3
>
.
O C
CD S
S 3
tfl g
2c
» >
100 -
-s
QJ
3
70
m
I
m
C/5
tfl
co
o
H
11
10
10
10
A (dB/m)
10
-------
REMOTE SENSING 163
site and the extent of modeling involved, excellent agree-
ment is obtained between the soil and radar models and
the experimental data.
In mapping the substructure of a number of low-level
nuclear waste disposal sites, Geo-Centers has had the op-
portunity to test the model comparing both the calculated
and the measured depth of penetration and attenuation of
the radar system. Such a correlation is illustrated in Figure
6 in which the depth of penetration and attenuation
measured with radar frequencies between 10 and 300 MHz
are compared with calculated values. The range of expect-
ed attenuation for a number of typical soils and rocks is
also indicated on the figure. The data indicate that at-
tenuations range from 2.5dB/m to 25dB/m for the soils
encountered at the sites. The relationship of attenuation to
depth of penetration encountered in the field surveys
agrees well with that calculated using the model.(1)
In order to simulate a typical uncontrolled disposal site
in New England and obtain radar signatures from an as-
sortment of objects in known orientations and configura-
tions and at known depths, a site was developed near
Harvard, Massachusetts (about 25 miles west of Bos-
ton).(1) The site chosen was in an undisturbed portion of an
active sand mining operation.
A trench 150 feet long and 20 feet wide was excavated to
a depth of 10 feet, except for one 20 foot section, 20 feet
deep, in a glacial outwash deposit of well-sorted, cross-
bedded sand. Sixteen 55-gallon metal barrels, two plastic
drums and a one meter cubic box were buried at the site.
GROUND SURFACE
Figure?.
Portion of a Radar Profile Taken at 300 MHz Over a Vertical
Metal Barrel (A) and a Horizontal Metal Barrel (B) in a Trench
10 Feet Deep.
W
w
Q
hi
2-
I
I
I
3-
I
I
i
GROUND SURFACE
• • • • JJUUIJIP • • • • •JLJLJULP • • •
! <-
TRENCHED AREA
UNDISTURBED
Figure 8.
Apparent Resistivity Pseudo-section Across Trench. Resistivity Values Range From 190 ohm-m at the Location of a Metal Barrel (White
Area Indicated as "B") to 2400 ohm-m (Black). Horizontal Distance is 15 Meters.
-------
164 REMOTE SENSING
The metal barrels were representative of typical objects
found at waste disposal sites. The two plastic barrels and
wooden box were representative of voids of different size
and shape within the soil.
Comprehensive data were taken with three antenna sys-
tems at 10 MHz, 80 MHz and 300 MHz. Radar scans
were made along the length of the trench at approxi-
mately five foot intervals. In addition, one scan perpen-
dicular to the length of the trench and one diagonal scan
were made over each of the targets at each frequency.
Electrical resistivity data were obtained to supplement the
radar data.
An example of a radar profile taken over two buried
55-gallon metal drums, one on its end (A) and one on
its side (B) is given in Figure 7. All fifteen target config-
urations, both metallic and non-metallic, were detected
with the radar. Resistivity data at this site are presented
in Figure 8 as a computer plotted pseudo-section show-
ing position across the trench and apparent depth. The
resistivity values differ markedly between the disturbed
area within the trench and the undisturbed geology. In
addition, the effect of the presence of a metal barrel on
the resistivity values is also demonstrated.
SUMMARY
A correlation has been found between the data ob-
tained using ground penetrating radar and electrical re-
sistivity. These two geophysical techniques provide com-
plementary geophysical information.
Ground penetrating radar, coupled ^with ground-
truth, can be used to give greater meaning to apparent
resistivity values, both with respect to geological substruc-
ture identification and to target or object location and
identification. Measurements of bulk soil resistivity can be
used to predict the expected depth of penetration of the
ground penetrating radar. Other geophysical techniques,
such as induced electro-magnetism and induced polariza-
tion are expected to show similar correlations with ground
penetrating radar.
REFERENCES
1. Horton, K.A., Morey, R.M., Beers, R.H., Jordan, V.,
Sandier, S.S., and Isaacson, L., U.S. Nuclear Regula-
tory Commission, "An Evaluation of Ground Pene-
trating Radar for Assessment of Low-Level Nuclear
Waste Disposal Sites," NUREG/CR-2212 (to be pub-
lished).
2. Mooney, H.M., "Handbook of Engineering Geo-
physics, Vol. 2: Electrical Resistivity," Bison Instru-
ments, 1980.
3. Annan, A.P. and J.L. Davis, "Radar Range Analysis
for Geological Materials," Report of Activities, PartB,
Geological Survey Canada, 1977.
4. Ridenour, L.N. (ed.), "Radar Systems Engineering,"
McGraw-Hill, New York, N.Y. 1947.
5. Birchak, J.R., Gardner, C.G., Hipp, J.E., and Vistor,
J.M., "High Dielectric Constant Microwave Probes
for Sensing Soil Moisture," Proc. IEEE, 62, No. 1,
1974.
-------
UTILIZATION AND ASSESSMENT OF A PULSED RF SYSTEM
TO MONITOR SUBSURFACE LIQUIDS
ROBERT M. KOERNER, Ph.D.
ARTHUR E. LORD, JR., Ph.D.
JOHN J. BOWDERS
Department of Civil Engineering and Physics
Drexel University
Philadelphia, Pennsylvania
INTRODUCTION
Hazardous liquids infiltrating the ground from shallow
or deep injections, storage lagoons, leaking containers,
tank car spills, etc., often reach the groundwater table
where they then flow downstream to open water, streams
or wells. The necessity of locating and monitoring such
seepage trajectories is of paramount importance. Obvi-
ously, this can be done by core borings and monitoring
wells, but the economics of the situation begs for a more
suitable and cost effective technique. It seems that one of a
number of nondestructive testing (NDT) methods might be
adopted for such monitoring. Candidate techniques,
grouped according to their basic operating principle, are
as follows:
•Thermally Induced Methods
infrared
heat pulse
•Elastic Wave Methods
seismic refraction
seismic reflection
ultrasonics
acoustic emission
sonar (active and passive)
•Nuclear Methods
neutron
•Electromagnetic Methods
resistivity
very low frequency
eddy current (metal detector)
pulsed radio frequency
continuous microwave
magnetometer
penetrating radiation
These methods have been reviewed as far as their
overall applicability to hazardous materials problems,01
and the electromagnetic methods have been categorized
and compared in a unified manner for location of subsur-
face anomalies, mainly buried drums.(2) This paper is an
extension of one of these electromagnetic methods, the
pulsed radio frequency (RF) method, for the purpose of
locating the water table in different soil types and over
the complete seasonal change of one year. Water surface
measurements via water observation pipes were also made
so as to obtain "ground truth" information and thus pro-
vide an actual comparison.
OVERVIEW OF THE PULSED RF TECHNIQUE
A considerable amount of subsurface probing at shal-
low depths has been based on the transmitting of pulsed
RF waves in the frequency range from about 1 MHz to
200 MHz. The transmitted pulse travels through the soil
until it infringes upon an object or material with dissimilar
electrical characteristics. Part of it is then reflected back to
the ground surface where it is received and the time of
Table I.
Details of Pulsed Ratio Frequency (RF) Methods
Approx.
Freq.
Range
Investigator(s) (MHz)
Cook (4, 5)
Rosetta (6)
Morey (7)
1-100
100-200
100-200
Dolphin,
etal. (8) 15-50
Unterberger (9) 230
Harrison (10) 35
Rubin
etal. (11) 100-200
Rubin and
Fowler (12) 100-200
Benson and
Glaccum
(13,14) 100-200
Max.
Depth
(meters)
225
15
15
40
500
2000
10
15
10
Sandness
etal. (15) 100-200 10
Alongi(16) 1,000 3
Moffat and
Puskar (17) 6,000 3
Major
Application
Area
locating faults, walls, holes
locating faults, caverns,
water, utilities
locating faults, caverns,
water, utilities
locating rock cavities
salt thickness measurement
determining ice thickness
detecting subway tunnels
drilling guidance, subway
tunnel monitoring, coal
thickness
general subsurface probing,
locate and follow pollutants
in ground, detection of
buried containers of in-
dustrial wastes
general subsurface probing
as described above
locating mines, pavement
thickness, shallow voids,
pipelines
locating faults, joints,
cavities, pipelines
165
-------
166 REMOTE SENSING
travel is measured. The depth "d" to the interface is then
calculated from d = (vt)/2 where "v" is the wave velocity
(which is equal to c/y/T, where "c" is the velocity of
light and " er" is the relative dielectric constant of the
material in which the wave is propagating) and "t" is the
pulse travel time. The relative dielectric constant of num-
erous soils at different water contents has been evaluated
by many investigators, including Okrasinski, et a/.(3) As
shown in Table I, this technique has been studied by many
workers.
Additional systems of a similar type as just described are
also available at a higher frequency range, i.e., above 1
GHz. These are also listed in Table I. They could, how-
ever, better be classified as microwave methods. The ad-
vantage of the higher frequency range is that a shorter
wave length gives greater definition to the subsurface ob-
ject. However, the attenuation of the waves is higher, re-
sulting in lower penetration depths, see Lord, et a/.(ll)
The authors have been active in adapting the pulsed RF
method to problems of interest in hazardous materials
spills and incidents. Using a commercially available sys-
tem which transmits RF pulses in the 100-300 MHz fre-
quency range and presenting the received signal in a real
time printout, a visual subsurface profile is developed.
Strong return signals appear as the black areas, while weak
(or no) return signals appear as white. Gray areas require
appropriate interpretation. The particular system used in
this study transmitted short ( 10~" sec. long) pulses of
carrier frequency about 120 MHz into the ground. Re-
flected patterns were recorded in real time printout as pre-
viously discussed. Maximum depth of penetration at the
sites under investigation was approximately 6 meters. The
specific goal was to reflect the RF pulses off the water
table surface, thereby determining its depth under the wid-
est possible set of conditions and circumstances.
DETAILS OF TEST SETUP
AND SITE CONDITIONS
Each of the three sites selected for monitoring were un-
paved and originally thought to be relatively homogeneous
within the upper 3 m of interest. To guarantee the near-
surface existence of the water table, each site began at a
creek or river at one end of the traverse and proceeded
inland. One site went from exposed creek to exposed creek.
Shown in Figure 1 are photographs of the three sites with
the water bodies visible in the background.
Wharton Site
The Wharton site extends from a 2 m deep cedar water
creek onto the land for a distance of 100 m. The ground
surface rises 3 m within this length. Nine water observa-
tion wells were placed in the ground, each deep enough to
intercept the water table. The water observation wells
were 2.5 cm diameter open ended pipes with removable
surface caps to prevent damage or vandalism during time
of non-use. The spacing varied from 3 m to IS m.
Subsurface sampling from the site showed that the soil
beneath the upper IS cm was a poorly graded sand which
contained a well defined water table, below which the soil
was saturated and above which the soil was dry or parti-
ally saturated depending on the recent weather conditions.
Physical properties of the soil are given in Table II. Geo-
technical properties calculated from the date of Table II
show an in situ porosity of 0.38 and a saturation of 9%
for the soil above the water table.
Springfield Site
The Springfield site extends between two shallow fresh-
water creeks spaced 90 m apart. Between these two creeks
the land surface rises approximately 1.5 m. The ground
(a) Wharton Site
(b) Springfield Site
(c) Penrose Site
Figure 1.
General Surface Conditions of Sites Being Monitored
-------
REMOTE SENSING
167
Table II.
Physical Properties of Soils Encountered at Various Test Sites
Location
and
Stratum
Wharton Site
Upper
Lower
Springfield Site
Upper
Lower
Penrose Site
Riverside
Lands id e
Unit
Weight
PCF (g/cc)
111 (1.78)
105 (1.68)
107 (1.72)
125 (2.00)
108 (1.73)
Water
Content
%
2
2
13
23
29
Specific
Gravity
g/cc
2.68
2.68
2.70
2.67
2.20
Effective
Size
mm
0.25
0.10
0.065
N/A
0.19
0.15
Coef. of
Uniformity
18
5
11
N/A
130
20
Liquid
Limit
N/A
N/A
N/A
56
N/A
51
Plastic
Limit
N/A
N/A
N/A
36
N/A
51
Plasticity
Index
N/A
N/A
N/A
20
N/A
0
Classi-
fication*
SW-GP
SP
SW
OH
GW/SW
MH
N/A = not applicable
•refers to Unified Soils Classification System
surface is an asphalt paved parking lot with a crushed
stone base. Total pavement thickness is approximately
15 cm. Eleven water observation wells, of the type pre-
viously described, were placed along the traverse at in-
tervals of 4 to 15m.
Soil samples indicated two different soil types: The
upper soil (i.e. beneath the asphalt/stone base pavement)
is a well graded silty sand and the lower soil is an organic
clay of high plasticity. Physical properties of the soils are
given in Table II. Calculated values from these properties
show the upper sand to have a porosity of 0.44 and per-
cent saturation of 45 %, while the lower clay has a porosity
of 0.39 and is fully saturated.
Penrose Site
The Penrose site extends from the edge of a major river
(brackish water) up a relatively sharp ridge to gradually
upward sloping land. The traverse selected was 50 m long
and four water observation pipes were placed along its
length at spacings from 5 to 10 m. Soil samples indicated
two different soil types: Closest to the river is a random
fill consisting of cobbles, gravel and miscellaneous fines
which was too coarse to obtain an undisturbed sample.
Awaf from the river the soil abruptly changed due to past
CQirstiSiction activity. The soil here was identified as a
high plasticity silt whose physical properties are given in
Table II. The resulting in situ properties of the landside
silt were a porosity of 0.39 and a saturation of 100%, i.e.,
fuUy saturated. It was observed that the riverside random
fill was also high in water content and obviously close to
complete saturation.
TEST RESULTS
During the course of this 12 month study, 20 separate
scans were taken of each test site. Temperatures ranged
from summer high of 29 °C to winter lows of -6°C. Sur-
face conditions varied from dry, to wet, to snow covered
and included varying amounts of frost penetration. Time
of monitoring varied from morning, to noon, to late af-
ternoon.
Actual water table elevations were taken by placing a
stiff coaxial wire down the water observation pipe until it
came into contact with the water within the pipe. The
shielding of the wire was drawn back from the central lead
about 2 mm so that a shorted condition existed until con-
tact was made with the water surface within the pipe. The
cable was connected to a voltmeter at the ground surface
which clearly registered a completed circuit, hence the
water surface. The length of wire was then measured to
determine actual elevations from the top of the water ob-
servation pipes.
Pulsed RF traces of the various sites were made im-
mediately after water table measurements. The entire pro-
cess at each site took less than one hour, thus changes in
water table elevation between the two sets of readings were
not considered significant.
Wharton Site
A typical pulsed RF trace at the Wharton site is shown
in Figure 2. As with all of the traces it appears as though
the ground surface is level while, in reality, it rises from
left (the creek) to right (landside). Also noted is a heavy
black band immediately beneath the ground surface where
no data can be obtained. This "main bang" effect is due
to the interaction of the three pulses traveling directly
across the transceiver. Its thickness depends upon the
frequency of the pulses and varies from 20 to 40 cm.
Seen reasonably clearly on the trace is a separation from
light (above) to dark (below), which indicates the location
of the water table. Superimposed on the figure are the lo-
cations of the water surfaces within the water observation
-------
168 REMOTE SENSING
WHARTON SITE
WATER OBSERVATION PIPES
(J) (8)
DEPTH
—rO»
In
2m
'irtW i.j J
IOO m
Figure 2.
Pulsed RF Survey of Wharton Site made on April 3,1981
with actual water table elevations indicated by
horizontal line in circle
pipes. Correlation has generally been good with accuracy
of 14% during dry ground conditions and 27% during
conditions of high surface moisture. This latter situation
occurred usually during winter readings when the ground
was frozen in the upper 20 to 50 cm. Interestingly, the
ground surface having a snow or ice covering did not ser-
iously influence the readings. Readings taken during dif-
ferent times of the day and temperatures above freezing
did not appear to influence results.
Springfield Site
A typical pulsed RF trace of the Springfield site is
shown in Figure 3. Here a very different situation is seen to
occur where no correlation exists between the RF trace
and the actual water table elevation. This is undoubtedly
due to the capillary zone which draws moisture above the
water table in fine grained soils. Since the degree of satur-
ation in this zone is very high, e.g., completely saturated in
the lower soil, no distinct boundary exists for the pulsed
RF system to define a distinct water table elevation. How-
ever, within the water observation pipe, the capillary zone
does not exist, thus the actual water table surface can be
located.
What is dominant on the pulsed RF trace of Figure 3 is
an undulating discontinuity showing two well defined
areas. This discontinuity maps out the distinction between
the upper granular soil and the lower fine grained soil.
It was present in this exact location every time that a tra-
verse was made. The only different noted from time to
time was in contrast, which was probably due to varying
performance of the printer.
Penrose Site
A typical pulsed RF trace of the Penrose site is given in
Figure 4. As with the Springfield site, the capillary zone,
with its high moisture content soils, completely masks
out the presence of the actual water table as determined by
the water observation pipes. What is clearly seen is that the
soils on the left side (adjacent to the river) are very dif-
ferent than the soils on the right side (the inland soils).
Subsequent soil testing showed these two soil types to be
very different from one another and placed in their respec-
tive locations during past construction activity. The boun-
dary was always present, in the same location, with con-
trast being the only difference between various readings.
SUMMARY AND CONCLUSIONS
Of the many candidate techniques available to nonde-
structively test for subsurface water, as in seepage plume
detection, the use of pulsed RF methods are likely
methods. Indeed, the reflection from a well defined water
table can lead to reasonable accuracy. Results of 12
months of testing show that the greatest accuracy, approx-
imately 14%, is during dry summer weather where there
is no surface moisture. The accuracy decreases to 27%
during wet winter months when moisture is contained
within the soil above the water table. These results, how-
ever, are only valid for sandy (or gravelly) soils which do
not contain a capillary zone. This capillary zone which
SPRINGFIELD SITE
WATER OBSERVATION PIPES
£"££JLs
^^^^^W^^^^p^ JM. JH^^^^III^^^I
Mw'Vh'X
M;«^j K.MJ
'•fcl
lift'Mil
SO m
FigureS.
Pulsed RF Survey of Springfield Site made on April 3,1981
with actual water table elevations indicated by
horizontal line in circle
-------
REMOTE SENSING 169
PENROSE SITE
WATER OBSERVATION PIPES
Figure 4.
Puked RF Survey of Penrose Site made on April 3,1981
with actual water table elevations indicated by
horizontal line in circle
varies in thickness above the water table is saturated, or
nearly so, and has a theoretical height of:
T
h,. = cosaj^ 1
yu dv
where
h,. = height of capillary rise
ft = wetting angle of water to soil
T0 = surface tension force of water to soil
Tw = unit weight of water
d, = average soil void diameter
Thus, height of capillary rise is inversely proportional to
the average soil void diameter. Easily seen is that coarse
grained soils (gravels, sand and their mixtures), with large
particle sizes hence proportionately large void sizes, have
very little capillary zones, e.g., the soils at the Wharton
site.
This is not the case for the Springfield and Penrose
sites, both of which had predominantly fine grained soils
(silts, clays and their mixtures). At both of these sites the
actual water table was not possible to locate and was com-
pletely dominated by stratigraphic changes in soil layers.
Seen at the Springfield site was a undulating horizontal
layering, while at the Wharton site the distinction was
between two different soils placed beside one another.
In conclusion, the use of pulsed RF methods can be
used in seepage detection in granular soils, but not in fine
grained soils. This is not a particularly disastrous limi-
tation because it is the granular soils which have high per-
meability, hence seepage problems are more acute in these
soil types. For the fine grained soils a spill, or other inci-
dent involving liquids, will be contained relatively close
to the spill location due to the much lower soil permea-
bility. Thus, in its proper setting, pulsed RF methods can
be used for groundwater monitoring and the related prob-
lem of spill detection.
ACKNOWLEDGEMENTS
This project has been supported by the U.S. Environ-
mental Protection Agency Cooperative Agreement (No.
CR-804763) administered through the Municipal/Environ-
mental Research Laboratory in Edison, New Jersey. Our
sincere thanks to the Agency, and in particular, to the
Project Officer, John E. Brugger.
REFERENCES
1. Lord, A.E. Jr., Tyagi, S. and Koerner, R.M., "Non-
Destructive Testing (NOT) Methods Applied to En-
vironmental Problems Involving Hazardous Materi-
als Spills," Proc. 1980 Natl. Conf. on Control of
Hazardous Material Spills, May 13-15, 1980, Louis-
ville, Kentucky, 174-179.
2. Lord, A.E. Jr., Koerner, R.M. and Brugger, J.E.,
"Use of Electromagnetic Wave Methods to Locate
Subsurface Anomalies," Proc. U.S. EPA Natl. Conf.
on Management of Uncontrolled Hazardous Waste
Sites, October 15-17, Washington, DC, Hazardous
Materials Control Research Institute, Silver Spring,
Md., 1980, pp. 119-124.
3. Okrasinski, T.A., Koerner, R.M. and Lord, A.E. Jr.,
"Dielectric Constant Determination of Soils at L-
Band Microwave Frequencies," Geotechnical Testing
Journal, ASTM, 1, 1979,134-140.
4. Cook, J.C., "Seeing Through Rock With Radar," in
Proc. of the North American Rapid Excavation and
Tunneling Conference, Port City Press, Baltimore,
1972,89-101.
5. Cook, J.C., "Ground Probing Radar," Proc., Sub-
surface Exploration for Underground Excavation and
Heavy Construction, American Society of Civil En-
gineers, New York, 1974,172-194.
6. Rosetta, J.V., Jr., "Detection of Subsurface Cavi-
ties by Ground Probing Radar," Symp. on Detection
of Subsurface Cavities by Surface Remote Sensing
Techniques, U.S. Army Waterways Experiment Sta-
tion, Vicksburg, MS, 1977,120-127.
7. Morey, R.M., "Continuous Subsurface Profiling by
Impulse Radar," Proc: Subsurface Exploration for
Underground Excavation and Heavy Construction,
American Society of Civil Engineers, New York, N.Y.
1974, 213-232.
8. Dolphin, L.T., Bollen, R.L., and Oetzel, G.N., "An
Underground Electromagnetic Sounder Experiment,"
Geophysics, 39, 1974,49-55.
-------
170 REMOTE SENSING
9. Unterberger, R.R., "Overview of Electromagnetic
Methods in Geophysics and Application of Radar to
Detection of Cavities in Salt," Symp. on Detection of
Subsurface Cavities by Surface Remote Sensing
Techniques, U.S. Army Waterways Experiment Sta-
tion, Vicksburg, MS, 1977, 119.
10. Harrison, C.H., "Reconstruction of Subglacial Relief
from Radio Echo Sounding Records," Geophysics,
35, 1970,1099-1115.
11. Rubin, L.A., Griffin, J.N., and Still, W.L., "Sub-
surface Site Investigation by Electromagnetic Radar,
Phase I Report: Feasibility," Report NSF/RA-76-
0187, ENSCO, Inc., Springfield, VA., March 1976.
12. Rubin, L.A., and Fowler, J.C., "Ground-Probing
Radar for Delineation of Rock Features," Engineer-
ing Geology, 12, 1978, 163-170.
13. Benson, R.C. and Glaccum, R.A., "Remote Assess-
ment of Pollutants in Soil and Groundwater," Proc.
of the 1979 Conference on Hazardous Materials Risk
Assessment, Disposal and Management, Information
Transfer, Inc., April 1979, Miami Beach, FL, 188-
194.
14. Benson, R.C. and Glaccum, R.A., "Radar Surveys
for Geotechnical Site Assessment," Proc. ofConf. on
Geophysical Methods in Geotechnical Engineering,
Atlanta, Oct. 23-25,1979, Preprint #3794.
15. Sandness, G.A., Dawson, G.W., Mathieu, T.J. and
Rising, J.L., "The Application of Geophysical Sur-
vey Techniques to Mapping of Wastes in Abandoned
Landfills," Proc. of the 1979 Conference on Hazard-
ous Materials Risk Assessment, Disposal and Manage-
ment, Information Transfer, Inc., April, 1979, Miami
Beach, FL.
16. Alongi, A.V., "Pavement Thickness Measured, Voids
Detected by Downward-Looking Radar in New York
Test," Industrial Research News, Fall, 1973,36-39.
17. Moffat, D.L. and Puskar, R.J., "A Subsurface Elec-
tromagnetic Pulse Radar," Geophysics, 41, 1976, pp.
506-518.
18. Lord, A.E., Jr., Koerner, R.M. and Reif, J.S., "De-
termination of Attenuation and Penetration Depth of
Microwaves in Soil," Geotechnical Testing Journal,
ASTM, 2, 1979, 77-83.
-------
INTEGRATION OF REMOTE SENSING TECHNIQUES
WITH DIRECT ENVIRONMENTAL SAMPLING FOR
INVESTIGATING ABANDONED HAZARDOUS WASTE SITES
ROBERT W. PEASE, JR.
The MITRE Corporation
Resource Recovery and Energy Systems
Bedford, Massachusetts
STEPHEN C. JAMES
U.S. Environmental Protection Agency
Solid and Hazardous Waste Research Division
Cincinnati, Ohio
INTRODUCTION
Abandoned hazardous waste sites present varying de-
grees of difficulty to investigators. For example, aban-
doned sites which are really extensive, rural (with hinder-
ing vegetation) or in areas of complex geology and hy-
drology, represent troublesome environments for in-
vestigation. Therefore, it is important to develop ap-
proaches for thorough, but rapid and cost-effective as-
sessments of these difficult situations. In most cases, a
well designed and executed investigative program will in-
clude remote sensing techniques in addition to direct
measurement. Premature action to drill wells, collect and
analyze various air, water and soil samples or perform
excavations without careful planning and proper integra-
tion of available techniques may result in an unneces-
sary exposure to hazardous conditions and in an inac-
curate or incomplete understanding of the total problem.
Remote sensing techniques may be used to provide
reasonably accurate assessments of subsurface contamina-
tion, the location and extent of buried drums and other
data needs for determining -appropriate methods of
abatement. However, not all the critical information can
be obtained remotely, since each of the techniques has
limitations, both theoretical and site-specific and conse-
quently, direct sampling should be undertaken at every
uncontrolled hazardous waste site.
To accomplish site investigations in the most efficient
manner, a systematic approach is necessary to take ad-
vantage of the information that can be extracted from
remote sensing methods. In addition, a systematic ap-
proach allows a reduction in the time and cost and an in-
crease in the effectiveness of direct sampling.
In general, the following two objectives must be ad-
dressed by all investigations at uncontrolled hazardous
waste sites:
1. Determination of the nature and extent of the problem
and the resulting effects on public health (both actual
and potential)
2. Determination of environmentally sound and cost-
effective methods to effectively abate the problem (if
abatement is deemed necessary).
The first activity of any investigation should be the
identification of the specific data needed to meet each
objective. After this has been accomplished, the various
techniques available for data acquisition, both remote
and direct, can be evaluated with regard to the type of
information that can be obtained from each in relation
to the specific conditions at the site.
Although not always the case, it may be reasonably as-
sumed that remote sensing techniques should be used in
advance of the more direct data acquisition methods of
borings or excavations. This is not intended to imply, how-
ever, that all direct sampling should be held in abeyance.
There have been numerous instances in which emergency
action is dependent upon immediate results from air,
water and soil sampling, and for such cases remote sensing
techniques should be used secondarily.
The purpose, advantages and limitations of each of
the four remote sensing methods discussed in this paper
are summarized in Table I. This material should be con-
sulted prior to the development of an investigatory pro-
gram. Even though there are disadvantages inherent to
each technique, proper sequencing and phased studies
can potentially result in an overall optimized approach.
As the study progresses, preliminary conclusions will
necessarily be modified and the nature of direct sampling
activities will need to be evaluated continuously. Final
conclusions should not be drawn solely from the results of
remote sensing methods. Direct sample collection should
be undertaken for all studies.
Because no single procedure would be appropriate for
all site investigations, the conditions at an uncontrolled
hazardous waste site in Coventry, Rhode Island will be
used for the basis of development of systematic pro-
cedures with the expectation that some of the concepts
can be applied elsewhere, as appropriate.
The sequence of activities discussed on the following
pages represents the idealized case and may not neces-
sarily have been followed in actual practice during an in-
vestigation performed by MITRE at the above site. This
is due to the fact that one of the purposes of the site in-
vestigation was to evaluate the capabilities and limita-
tions of the selected remote sensing methods. The exact
171
-------
172 REMOTE SENSING
Table I.
Comparison of Remote Sensing Techniques
Technique
Purpose
Advancages
Limitations
Electrical Resistivity
Lateral Profiling
• determine lateral extent of
contaminated ground water
• facilitate placement of mon-
itoring wells and optimize
their number
• procedure less expensive than
drilling
• procedure more rapid than
drilling
* equipment light-weight, able
to be hand carried
• limited ability to detect
non-conductIv? pollutants
• technique unsuitable If no
sharp contrast between con-
taminated and natural ground
water
• monitor changes in plume
position and direction ^ ^^ ^ ^ conducted tn . Interpretation difficult If
vegetated anas water table la d«P
• Interpretation difficult If
lateral variations In stra-
tigraphy exist
• interpretation difficult if
radical changes in topogra-
phy are not accounted for it
choice of A-spaclng
* technique unsuitable in paved
ductlve objects
Depth Profiling * indicate change in contamina- same as above aame as above
tion with depth
• establish vertical control in
areas of complex stratigraphy
Seismic Refraction (Non- • determine depth and topogra- • procedure less expensive and • technique unsuitable if no
explosive Method) phy of bedrock safer than coring or excavation aharp velocity contrast be-
• determine depth of trench • procedure more rapid than twe*n units of Interest
containing buried drums coring or excavation <••*•• trench containing
burled drums and surrounding
• survey may be conducted In aoil)
vegetated areas
• survey requires access road
for vehicle
* depth of penetration varies
with strength of energy
source
• low velocity unit obscured
by overlying high velocity
units
• interpretation difficult in
regions of complex stratigra-
phy
Metal Detection • locate areas of high metal * procedure less expensive and • technique unsuitable for the
content (e.g., buried drums) safer thnn excavation or radar detection of non-metallic
* procedure more rapid than ex- o j c
cavation or radar • technique unsuitable for ob-
. equipment light-weight, able JectB bel°H flve feet
to be hand-carried • technique unsuitable for dt-
• survey may be conducted In termination of number or ar-
vegetated areas ranyement of buried objects
Ground-Penetrating Radar e locate burled objects (e.g., • procedure less expensive and * technique unsuitable for
buried drums) safer than excavation vegetated areas
• provide qualitative infor- • procedure more rapid than ex- • data requires sophisticated
nation regarding drum density cavation Interpretation
• detect Interfaces between • procedure deeper-penetrating • underlying object* obscured
disturbed and undisturbed than metal detection by those above
soil (e.g., bottom of tren- , ,,
Ches) • procedure yields more infor- • survey requires access road
macion than metal detection for vehicle
• detect plunes of high chemi- , .
c.l concentration ' P""-1-" "X *•« ««
-------
REMOTE SENSING 173
is helpful to keep them both in mind separately when
planning an investigation. The integration of remote and
direct sampling methods is therefore presented in the con-
text of meeting these two separate goals.
DETERMINATION OF NATURE AND
EXTENT OF PROBLEM
The following items had to be determined for a com-
prehensive understanding of the nature and extent of the
problem at the Coventry site:
•Direction, rate, and extent of subsurface migration of
contaminants
•Location of surface discharge areas of contaminated
groundwater and the subsequent fate of the contaminants
•Identification of contaminants
•Location of areas of contaminated soil and buried drums
and determination of the potential of this source of pol-
lutants for long-term release.
An example of a systematic approach to achieve the
above objectives is presented in Table II and the phasing of
site activities is shown by Figure 1. In Figure 1 a two-phase
monitoring-well installation and sampling program is
shown. The purpose of a phased investigation is to ob-
tain a preliminary understanding of the problem prior to
final planning of all direct sampling in order to more ef-
fectively guide subsequent activities. Additionally, this
figure shows that, in general, remote sensing precedes
direct sampling in order to reduce the time and cost of
the latter and to help ensure that the full extent of the
situation is identified.
For buried drums, metal detection, rather than ground-
penetrating radar should be used to locate the burial areas
because of the relatively lower cost and greater porta-
bility of the former technique. However, radar has great-
er penetration and should be used in all areas where
drums are suspected but not found by metal detection.
Limited excavation may then be required to gain in-
formation about the depth, condition and contents of
the drums so that the concept of total drum excavation
and chemical disposal can be evaluated with the other
abatement alternatives. Limited excavation was considered
feasible for the Coventry site since total drum excavation
did not appear significantly more expensive than other
abatement alternatives.
Surface water, groundwater and soil sampling are
necessary for any hazardous waste site investigation.
Monitoring wells are best located once the extent and di-
rection of the plume have been determined by a resistiv-
ity survey. Monitoring wells should be placed at the fol-
lowing locations:
•Upgradient from the source of contamination (to moni-
tor background conditions)
•Outside of the plume downgradient from the source of
contamination (to verify the extent of the plume and to
monitor its movement)
•Within the plume close to the source of contamination
(to obtain samples before extensive dispersion, dilution
or attenuation)
•Within the plume at the outer extent of contamination
(to observe dispersion, dilution and attenuation).
Table II.
Systematic Approach to Determine Nature and Extent of Problem at Coventry Site
Objective
Data Needs
Investigatory Methods
Comments
1. Direction, Rate,
and Extent of
Subsurface Con-
taminants
determine areal extent
of contaminated ground
water
determine vertical ex-
tent of contaminated
ground water
v assess hydrogeologic and geo-
logic settings of the site by
reconnaissance and study of
topographic maps, aerial pho-
tographs, and all existing
data
• conduct vertical resistivity
profiles in various locations
to determine approximate
depths of contamination
• conduct lateral resistivity
survey choosing A-spacing
based on results of vertical
resistivity profile
• Install monitoring wells in-
side and outalde of contami-
nated zones as defined by
resistivity survey for con-
firmational purposes
• conduct seismic refraction
survey over contaminated
area to determine depths to
bedrock and vertical subsur-
face profiles
• install cluster wells con-
sisting of borings screened
in soil and in bedrock both
inside and outside of con-
taminated zones
• establish permanent field
grid for use by all remote
sensing surveys; grid loca-
tions can be transferred to
site map by aerial photogra-
phy or land survey
• choose wide enough A-spaclng
for lateral survey to mini-
mize influence of variations
of anticipated depth to
ground water from ground
surface
• use established field grid
for seismic survey
• use results of seismic re-
fraction survey to select
most economic locations for
bedrock borings
• determine whether vertical
gradients exist and whether
contaminants are present in
bedrock fractures
• install bedrock wells as a
second-phase drilling effort
and install any additionally-
needed monitoring wells based
upon first-phase results
-------
174 REMOTE SENSING
Objective
Data Needs
Investigatory Methods
Commenta
drtc-rmine direction and
rote of subsurface ml-
prat ion
• Install minimum number of
monitoring wells to define
ground water flow; well lo-
cations should be based on
results of resistivity survey
and chemical analytical needs
(BCC third objective In this
cable)
• perform In-sltu permeability
ti-sts In selected monitoring
and bedrock wells
• because the ground surface
contours at Coventry site
Indicate a potentially large
arc of subsurface travel,
wells at edge of contaminated
zones were necessary to de-
tail direction of outer frlnpe
of pollutants
• install cluster well upgradi-
ont of pollutant source to
dut ermine background condi-
tions
Location of Sur-
face Discharge
Areas of Ground
Water Contamina-
t ion and Deter-
mlnathm of Pate
,»l P,,l lutiints
> locate surface discharge
areas
> determine fate of pollu-
tants
• reconnoiter site; locate dls- • use air quality measuring
charge areas by sight and air
quality measuring devices
, , . .
. use results of electrical re-
atsttvlty and water table map
to identify surface- receptors
of subsurface dls charge
• conduct downstream sampling,
including rate of flow of
surface water
. . . . .
• sample sediment and air around
discharge area
3. Chemical Identi-
fication of ill
Principal Contam-
inants
a determine principal con- a analyze composite samples
tamlnants from selected wells and sur-
face waters for priority
pollutants
e analyze selected wells and
surface waters for selected
compounds based upon priority
pollutant analysis
it. Location of Areas
of Burled Drums
and Contaminated
Soil and Deter-
mination of Life-
time of Future
Ke leant- of Chcmi-
cals
9 determine dispersion of
contaainants
i locate areas of burled
druma and contaminated
BOiL
• determine lifetime of
future release of chemi-
cals
9 select indicator analyses
based upon priority pollutant
screening: include general
water quality testa (pH, con-
ductivity, iron, chloride,
TOC)
• reconnoiter site; search for
areas of disturbed soil or
vegetation, or areas of dis-
colored soil
a Interview persons involved
with dumping activities for
information concerning loca-
tion of trenches and method
of operations
• study previously taken aerial
photographs to obtain histor-
ic information
• use results of resistivity
survey to locate source of
contaminated ground water
a conduct metal detection sur-
vey over all cleared or dis-
turbed areas of site
a conduct ground-penetrating
radar survey in burial areas
located by metal detection
survey
• conduct seismic refraction
survey over drum burial area
to determine depth of druma
a perform limited excavation
of burled druma to determine:
condition and contents of
druma, density of drums, lowar
boundary of drums
a sample soil In drum burial
areas and bulk chemical dis-
charge area; analyu extracted
leachate for specific chemical
compounds determined by pri-
ority pollutant screening
devices to determine areas of
poor quality. Indicating sur-
face discharge points: also
indlcatcs areas requiring
brcathlrlR protection devices
" K
• mass balances should be cal-
culated for principal pollu-
tants to determine ultimate
dispersion mechanlsm(s)
K
• upstream sampling necessary
for background conditions
• all water wells and surface
waters used for notable water
supplies within a mile radius
should be precautlonarlly
sampled
• composites made up of wells
close to source of pollutants
and surface discharge areas;
should limit each composite
to only two adjacent wells
J
• indicator analysis provides
relatively low coat method
for non-specific monitoring
of pollutant levels in moni-
toring well network
• use established field grid
for metal detection and ground-
penetrating radar survey
a metal detection used in ad-
vance of ground-penetrating
radar because of lower cost
and ease of use; however,
radar has greater penetration
and should be used in all
areas where drums are sus-
pected, but not found by
metal detection
a effectiveness of seismic re-
fraction method to determine
lower boundary of buried
druma remains subject to
verification
a drum excavation limited to
data gathering only and should
be terminated whan aufflclcnt
Information obtained; refer to
Phase II project raport (ref-
erenced In Section 1) for rec-
ommendations regarding excava-
tion procedures
• drum excavation should bt con-
ducted after all other eite
activities completed in order
to minimise personnel, on a tie
Coventry elte procedures used for illustrative purposes only; investigatory procedure* end sequence mav not
necessarily be directly spplicsble to other sites.
-------
REMOTE SENSING 175
H:
Phaae I
Monitoring Welle
Bedrock Boring*
Pnaae I
Chewlcal Analyala
- preliminary dtuer-
- Jc I •mill
Jtlulfer
Intlta
Data Havlav
Haalatlvlty
SElBBlc Refraction
Survey
Ion of
roprlate Indi-
- alte »eonnalaa«nea
- ravlen of prior ra-
porta and chealcal
- Interview of par-
anal aMoclatad
with duaplng
- atudy of topo-
iraphlc Mpa end
oarlal photagrapha
- aatabllahMot of
panaoent field grid
- production of photo-
letrlc tup of
rroundlng
>cetlon of cant a
lated ground wet I
•It* end a
- dataralnetlon of
depth to bedrock
- determination of
lower boundary of
dtuoa and prella-
Inary eeelnate of
nunber of druni
Well Inatallatton
and Chenlcil Analyala
- determination of cc
tenta, depth, and
- determination of
final eatlMte of
number of burled
- location of burled
- diceralMcton of hor-
liontal boundarl**
of burled druM
Figure 1.
Recommended Sequence of Activities at Coventry Site
The depth and topography of the bedrock should be
defined, not only to evaluate certain abatement methods,
but also to economically locate the deep borings that may
be needed for bedrock sampling, permeability testing and
water sampling. Seismic refraction is a very effective tool
for providing remote information on the configuration of
subsurface strata. Additionally, the results can be used to
locate bedrock wells, the position of which will depend
upon the objectives of the investigation. For example,
some investigators may wish to install bedrock wells only
to determine formation integrity as an acceptable base for
certain abatement alternatives. Other investigators may be
interested in locating bedrock wells to determine the
presence of contaminants in specific regions or channels.
DETERMINATION OF METHODS
TO ABATE THE PROBLEM
The data needs for selecting and determining the cost
of abatement alternatives (Table III) are similar to those
required for understanding the nature and extent of the
problem at the Coventry site. However, the two objec-
tives are best addressed separately, since the location for
direct sampling and remote sensing may differ between
the two, as may the use of the information obtained.
It is possible to use remote sensing techniques as a
"negative screening" step in the evaluation of certain
abatement options. An example of this concept is given in
the case of evaluating source encapsulation as an abate-
ment technique for the problem at Coventry. For encap-
sulation to be feasible it is necessary to have a low perme-
ability base within a relatively shallow depth from the
Table III
Major Informational Needs for Implementation of
Certain Abatement Activities at Coventry, Rhode Island
Alternative
Informational Needs
Removal of Buried Drum Condition
Drums and Dis- Drum Number
posal of Chemicals Drum Contents
Trench Location and Geometry
Encapsulation of
Source
Collection and
Treatment of
Leachate (trenches
and/or wells)
No Action
Alternative
Drum Contents
Imperviousness of Underlying Strata
Level(s) of Contamination (soil and/or ground-
water and/or bedrock)
Trench Location and Geometry
Areal Extent of Contamination
Type of Contamination
Concentration of Contaminants
Imperviousness of Underlying Strata
Aquifer Characteristics
Drum Contents
Drum Condition
Level(s) of Contamination (soil and/or ground-
water and/or bedrock)
Type of Contamination
ground surface. Seismic refraction has the potential to
determine whether the bedrock underlying the source of
contamination should be ruled out (negative screen) or
consideration as an acceptable base.
If, after a seismic refraction survey it is found that the
bedrock is either too deep or too fractured to function as
-------
176 REMOTE SENSING
an effective base, then rock coring is unnecessary. The
information obtained by this method is not sufficient,
however, to prove that the bedrock is sufficiently sound
for encapsulation without actual test borings and field-
permeability tests.
There is one investigational method for estimation of
the number of buried drums which is listed in Table II
and not included in the discussion of the preceding sec-
tion. The remote sensing technique of seismic refraction
can determine the lower boundary of buried drums. The
results of an exploratory excavation can be compared to
the seismic profiles obtained at a particular location and
if valid, the seismic profiles for all other buried drum
areas may be used for drum number estimates. If there is
no correlation found between the seismic profile and the
excavation (of if an exploratory excavation is not desired),
it is recommended that the maximum feasible excavation
depth or depth to bedrock be used for the lower boundary
of the buried drums.
ACKNOWLEDGEMENTS
The work on which this paper is based was supported
under MITRE contracts with the U.S. Environmental
Protection Agency, Solid and Hazardous Waste Research
Division, and the State of Rhode Island, Department of
Environmental Management. The encouragement and in-
terest of Mr. Donald Sanning of the EPA is especially
noted. The contributions to the site investigation of Mr.
Paul Stoller and Dr. Harold Yaffe of MITRE and Ms.
Nancy Cichowicz, formerly of MITRE and presently
with ERT, Inc., are also acknowledged.
The actual results of the Coventry investigation are
found in a paper by Yaffe et al.(4)
REFERENCES
1. Pease, R.W., et al., "Hazardous Waste Investigation:
Picillo Property, Coventry, Rhode Island," MTR-
80W00032, The MITRE Corporation, Bedford,
Massachusetts, 1980.
2. Cichowicz, N.L., Pease, R.W., Stoller, P.J., and
Yaffe, H.J., "Evaluation of Abatement Alternatives:
Picillo Property, Coventry, Rhode Island," MTR-
80W00253, The MITRE Corporation, Bedford,
Massachusetts, 1980.
3. Cichowicz, N.L., Pease, R.W., Stoller, P.J., and
Yaffe, H.J., "Use of Remote Sensing Techniques in
a Systematic Investigation of an Uncontrolled Haz-
ardous Waste Site", MTR-80W00244, The MITRE
Corporation, Bedford, Massachusetts, 1980.
4. Yaffe, H.J., Cichowicz, N.L., and Stoller, P.J.,
"Remote Sensing for Investigating Buried Drums and
Subsurface Contamination at Coventry, Rhode Island,
Proc. U.S. EPA National Conference on Manage'
ment of Uncontrolled Waste Sites, Oct. 15-17, 1980,
Washington, D.C., Hazardous Materials Control Re-
search Institute, Silver Spring, Md., 239-249.
-------
SURVEY AND ANALYSIS OF PRESENT/POTENTIAL
ENVIRONMENTAL IMPACT SITES
IN WOBURN, MASSACHUSETTS
SUSAN E. TITUS
The Bionetics Corporation
Warrenton, Virginia
INTRODUCTION
The U.S. Environmental Protection Agency, Region I,
initiated a request through EPA Headquarters to the En-
vironmental Photographic Interpretation Center (EPIC)
for a survey and analysis of a heavily industrialized area
(approximately 16 square miles) surrounding North Wo-
burn, Massachusetts. The purpose of the study was to es-
tablish the ambient environmental condition of this area,
where various industries have produced a high output of
industrial and chemical products for more than a century.
Many of these industries are alleged to have dumped or
buried spent chemicals and by-products, which are a sus-
pected source of endemic health effects in the vicinity.
Analysis of historical cartographic and photographic
sources and current aerial imagery can supply continu-
ous information to search efforts in the field, thus reduc-
ing the time anc cost of locating test sites, abandoned or
unpermitted dumps, and areas of environmental damage.
Through the utilization of imagery, sites can be located
and analyzed in a matter or hours or days by a single
analyst, as opposed to time-consuming ground searches
by inspection crews themselves.
Currently, field tests are being scheduled to coincide
with the receipt of incremental site assessments forwarded
from EPIC upon completion.
Located north of Boston, Woburn is a town with a long
history of industrial activity. Since the 1850's, industries
have manufactured products which include: chemicals for
textile, paper and leather processing; explosives, hide and
bone glue and arsenic-based pesticides. The wastes from
these products were frequently disposed of on grounds
surrounding the factories. A tract of approximately 120
acres containing abandoned dump sites, building ruins,
waste pits and a new industrial park is the main focus of an
area-wide survey to determine the extent of present or po-
tential environmental hazards and ambient environmental
conditions.
Presently owned and controlled by prominent area de-
velopers, this tract is a valuable piece of commercial real
estate. Its location in close proximity to 1-93 and Route 128
resulted in a surge of expansive redevelopment creating
the industrial facility known as Industri-Plex 128. Pur-
chased in 1968, the tract has yielded vast amounts of in-
dustrial by-products during excavation for future com-
mercial sites. Public attention was drawn to this area in
the mid-1970's when hydrogen sulfide odors were emit-
ted by rotting hides and animal wastes unearthed during
construction. Complaints were made to the State of
Massachusetts by residents of neighboring communities af-
fected by the fumes. In a separate but coinciding inci-
dent, an EPA employee reported a possible case of illegal
wetlands filling. Unpermitted, the filling was halted; sub-
sequent testing by the EPA and the U.S. Army Corps of
Engineers revealed concentrations of arsenic, hexavalent
chromium and lead. Research into prior land use along this
strip indicated past utilization for chemical waste disposal
in ponds and lagoons, as well as for a dump site by
various tanners and Tenderers.
Subsequent sampling of water throughout the Woburn
area revealed two municipal wells in North Woburn to
be contaminated with the solvent trichlorethylene (TCE),
a suspected carcinogen. The source of the contamination
by this compound is unknown.
Woburn has experienced a higher than normal rate of
cancer, approximately 13% higher than Department of
Health Statistics predicted, especially in a neighborhood
known as Walnut Hill located less than a mile south of
the industrialized corridor. Eight of fourteen reported
cases of leukemia have occurred within a half-mile radius
of one another in this small residential development/0 As
yet, no cause-and-effect correlations can be drawn for this
occurrence.
AREAS OF CONCERN
Seven sites and surrounding areas, including the indus-
trial park, were chosen for in-depth analysis of present
and historical land use. Known industrial and commercial
uses of these tracts include: the sites for tanneries and glue
manufacture, chemical processing, and extraction and
landfilling operations.
As of July 1981, sets of prints, overlay series and texts
concerning the Woburn study area have been forwarded
incrementally upon completion to Region I for use by EPA
Massachusetts Department of Environmental Quality
Engineering (DEQE) and the field inspection team (FIT).
The focus of critical concern from impact by uncon-
tained wastes is the quality of surface and subsurface water
in the study area. A review of historical aerial photography
shows that major changes have occurred in surface water
paths, wetlands and terrain throughout the Woburn area.
177
-------
178 REMOTE SENSING
Although alteration of these features through excavation,
filling, dredging or construction may re-route surface
drainage or re-shape impoundments, subsurface water will
often resume its flow in the original stream bed, along the
path of least resistance. Fluid or viscous waste and leachate
from buried contaminants will flow under gravity to
groundwater levels and form plumes along the stream.
These plumes may extend under many acres of ground if
the source is unchecked.(2) Locating the point source of
subsurface water contamination can be facilitated by ex-
amining historical land use in the drainage basin.
Slightly less critical, because of its more stationary na-
ture, is soil contamination. Sites with possibly contaminat-
ed fill material can also be identified through historical
imagery analysis.
In the realm of air quality, past and present offenders
may be noted by the presence of vents, stacks, fans and
similar exhaust structures.
METHODOLOGY
Use of Historical Cartographic Records
Since aerial imagery of the United States is scarce or
non-existent prior to the late 1920's, land use before this
period may be largely unrecorded. Human memory is sub-
ject to error; interviews with area residents on historic
usage may yield conflicting data or place sites far from
where they actually existed. In regions such as Woburn, it
is apparent that many major alterations probably occurred
between the inception of industrial land use in the 1800's
and the first available photography in the 1930's. In order
to assess past environmental influences, research was un-
dertaken at the U.S. Library of Congress for plots, docu-
ments or cartographic records of properties which were
recorded in the late 1800's or early 1900's.
In the Geography and Maps Division, the Sanborn Fire
Insurance Maps(3) were the focus of attention. Developed
to record safety precautions and hazards associated with
structures in the towns, they provide valuable informa-
tion on conditions in an area beginning in the 1800's.
Coverage of Woburn begins in 1888 and ends in a series
of 1926 maps corrected by paste-ons in the 1930's and
again in 1947. An example of the type and extent of in-
formation found on the maps is the Merrimac Chemical
Company sheet. In 1888, Merrimac Chemical occupied a
site in North Woburn adjacent to the railroad (Figure 1).
Each structure is plotted, and the building contents or
process described. At this time, the major compounds
present were sulphur, lead, muriatic acid and alum. By
1904 (Figure 2), the number of buildings had increased
and substances such as chloride of aluminum, bauxite
and pyrite had been added to the inventory. Moving
chronologically through the maps, it is possible to see the
transition of Merrimac Chemical (1888) to include New
England Manufacturing Company Chemical Works
(1918) and its evolution into Consolidated Chemical In-
dustries Co. Glue Plant (by the 1940's), along with de-
tails in process changes and on-site materials. The most
recent transitions were revealed by aerial imagery. The
site remains a functioning chemical process plant in 1963
(Figure 3), and is finally seen as a razed, abandoned site
in 1980 (Figure 4).
An example of the value of these cartographic records
in extreme land-use changes is depicted by the reproduc-
tion of a block in the town of Woburn in 1894 (Figure 5).
This lot had been the site of a leather company since 1888.
In 1904, it is recorded as American Hide and Leather
Co. Factory H (closed); in 1918, the lot contains scat-
tered residences and open ground. By 1926, the block con-
tains many residential units. This block, as it appeared
in 1963, is centered in Figure 6.
In a rough preview count, 14 of 48 selected sites traced
through the maps became residential, commercial or va-
cant areas by the 1940's. Each of the 48 sites will be
checked against the aerial imagery for 1938 and 1963 to
determine the continued development of these historic
industrial grounds.
The utilization of Sanborn Maps in hazardous waste
studies is also being pioneered in a current project for
Stroudsburg, Pa. Concerned with contamination from no
apparent sources, Sanborn Maps checked for the years
1905, 1912, 1923, 1930 and 1950 disclosed the location of
a coaltar injection well and two storage tanks which no
longer exist. No previous knowledge of the location of
these features was apparent in the region.
Use of Historical Aerial Photography
In order to accomplish a comprehensive, detailed data
base on locations of present and former excavation and
fill sites, streams, bogs, wetlands and impoundments, as
well as industrial locations, a survey utilizing historical
and current aerial photography was undertaken. His-
torical black-and-white imagery was obtained for De-
cember 1938, April 1963 and April 1978, from the U.S.
Geological Survey and the National Archives. Current
true-color imagery was obtained in November 1980 un-
der contract by EPIC.
The photo frames were chosen at the scale of 1:24,000
for the historical, and obtained at 1:10,000 for the color.
Each of the seven primary areas of concern was plotted on
overlays to USGS quadrangle maps (scale 1:24,000) and
analyzed from the imagery for both minute and massive al-
terations which may have occurred over the 42-year span.
With the location illustrated on the map overlays, the
specific details of site variation were plotted on clear mylar
overlays to enlarged photos of each year of coverage per
site. The scale of the enlargements varied from site to
site. An example of drastic change, illustrating the type of
details found, is evident in Figures 7 and 8. In 1938,
(Figure 7) large lagoons and trenches for liquid waste run
along the western boundary of the plant property. This
same section in 1963 (Figure 8) is the site of additional
plant structures, an access road and a filled lot with only
a single narrow trench visible. The surrounding locale
has experienced development and the addition of major
transportation routes.
In order to obtain and evaluate information such as
this from aerial imagery, film transparencies are backlit
on a standard Richards light table and examined in stereo
through optics capable of magnification up to 30x. A
-------
REMOTE SENSING 179
ffC CHEMICflL CO.
flc/os.
Figure 1.
Merrimac Chemical, Sanborn Maps, 1888
-------
180 REMOTE SENSING
[11 EH EH G
Figure 2.
Merrimac Chemical, Sanborn Maps, 1904
-------
REMOTE SENSING
181
Figure3.
Old Merrimac Chemical Site, 1963
trained analyst examines each site for details concerning
terrain, structures, ground disturbance, impacted drain-
age, discarded material, land use of the locale and "sig-
natures"* of other elements of environmental influence.
In general, examination of terrain features, structures,
drainage and land use is fairly straightforward; the de-
tection of drums, rubble, offal or evidence of under-
ground structures requires close scrutiny and reliance on
surrounding information for detection. All pertinent in-
formation regarding the site is recorded on the overlays
for location; a corresponding text contains details and
conditions of all possible past and present impact points.
In cases of confirmed or suspected waste disposal or
other pertinent environmental factors, comparisons will
be made by engineers in the region against local hydro-
logic, geological and air, studies to determine possible
cause-and-effect correlations, the extent of any effects,
and other related factors. From this information, fields
checks will be performed on soil and water quality. At
EPIC, when the photographic analyses have been per-
formed, a computer system is utilized to locate and chart
the exact location of sites to guide field sampling efforts.
EPIC's Imagery Analysis System (IAS), designed by
Cab"8, is capable of rectifying photo-to-map scales in
'The term "signature" is used to describe a particular pattern, shape,
tone or color which consistently indicates the presence of an object,
material, etc. on aerial imagery, even though the object itself may be in-
distinguishable.
Figure 4.
Old Merrimac Chemical Site, 1980
order to plot points directly from imagery to a standard
map. The system is also able to determine the exact
geocoordinates for an individual site or set of points and
to compute the area of a chosen feature. Using this in-
formation, the field inspection team can quickly and ef-
fectively locate precise points from which to extract
samples.
In addition to the site-by-site analysis, a survey of the
entire Woburn study area was performed to locate any
dumps, impoundments, fill mounds, extractions, in-
dustries or junkyards in each year of imagery coverage.
The sites were located numerically on a series of clear
overlays to USGS 1:24,000 quadrangle maps, and a brief
description, by site, for each year was included in an in-
ventory list (Figure 9). The transformation of a site from
a hillside farm in 1938 to an extraction site and back to
a filled meadow by 1980 can thus be quickly surveyed.
A survey is also being performed at 10-year intervals
from the Sanborn Maps of the study area between 1888
and 1947.
To accompany the site survey, two additional series of
overlays were developed to illustrate historical to present
drainage transformation and land-use patterns. Included
on the drainage overlays were streams, bogs, wetlands,
natural lakes and ponds, and man-made features includ-
ing dams, industrial lagoons and impoundments. This
series can provide much needed data on original water
courses in sections of disturbed drainage. When used in
-------
182 REMOTE SENSING
Figures.
Block of Chestnut and Scott Streets
SanbornMaps, 1894
Figure 6.
Block of Chestnut and Scott Street. 1963
-------
REMOTE SENSING 183
Figure 7.
Waste Lagoons; Industrial Site, 1938
conjunction with the site overlays, possible point sources
of subsurface water impaction may be located, and
abatement procedures formulated.
The land-use series employed a system of mapping
land usage categorized down to Level HI, modified from
the USGS Land Use Classification System/4' For example,
Category 2 in this system covers agricultural land use; at
Level III, sub-categories are determined and broken out
for crops (211), pasture (212), fallow (213) and orchards
(22). The drainage series and the land-use series together
can show historical patterns of agricultural run-off, pos-
Sitel—
1938: Agricultural use; barn, residence
1963: Extraction site
1978: Extraction site
1980: Filled, revegatated meadow
Site 2—
1938: Agricultural field
1963: Auto junkyard
1978: Construction site
1980: Residential development
Figure 9.
Sample inventory list for site comparison
Figure 8.
Waste Trench; Industrial Site, 1963
sible paths of industrial waste or leachate flows, and lo-
cations of liquid dump sites. Findings such as these can
assist in evaluating deposits of contaminants not associ-
ated with present land usage and drainage tendencies.
VERIFICATION OF FINDINGS
Upon the completion of the analyses of the seven spe-
cific sites and at mid-point in the area-wide site inventory,
a ground-truthing meeting was held by the imagery
analyst and representatives of the Massachusetts DEQE
and EPA Region I office. The comparison of data ob-
tained from the four sets of imagery at EPIC and informa-
tion researched by the DEQE (including different sets of
aerial imagery and historical information on zoning uses of
some properties) confirmed a high level of accuracy. Field
inspection of several sites also yielded confirmation of
suspected historic use not found on the present color
imagery. An example of this occurred during an inspec-
tion of a tannery site. Historical imagery revealed plumes
of sludge assumed to be chromium wastes along a slope at
the rear of the property. The 1980 color imagery showed
only natural-toned earth in this section. From the
ground at close range, however, a very light tone of pale
yellow was observed emerging in several places. Pale and
-------
184 REMOTE SENSING
bright yellow tones are sometimes indicative of chromate
leachates.
In another case, historic analysis showed "soft"
mounds of waste, which appeared to be hides, on the lot of
a plant. Although the dumping of hides was not reported
or suspected at this site, a walk through the abandoned
site turned up several scraps of old hide in a historically
filled area. Some information given to the state representa-
tive by local residents who recalled past dump sites and
industrial uses also indicates a correlation with informa-
tion gained from the imagery. The results of samples
gathered by the field inspection team will determine the
accuracy of selecting sites by imagery analysis.
ASSESSMENT OF MODE OF ANALYSIS
The procedure of site and area analysis via remote sens-
ing is an effective, time-saving and economical method
for determining environmental conditions in both the past
and the present. Although it has been employed in Woburn
as a locating device, imagery analysis is extremely useful in
monitoring control and clean-up operations and in emer-
gency response for assessing areas of immediate danger.
The utilization of historical imagery for a current na-
tional concern emphasizes the importance of maintaining
imagery obtained in the past. It is invaluable for research
in this mode; it does not describe what may have been but
shows what actually existed at a given time and place.
The use of historic cartographic records such as the San-
born maps, unprecedented in hazardous waste assessment,
also illustrates the necessity for preserving historical
records. The existence of such a detailed record of ma-
terials and industrial site locations long before the hazards
of chemical wastes were realized has contributed much
information on potential impact sites where no immediate
signs of danger may otherwise have been noted.
CONCLUSIONS
To date, the impact of uncontrolled hazardous wastes
upon Woburn has not been as spectacular nor as immedi-
ately devastating as at Love Canal although the essence of
the situation is the same. It is the intent of those involved
in the aerial analysis of the area to provide data which
may prevent damage from dormant wastes, and aid in the
control and clean-up of sites known to be a problem.
In combination, historical imagery and cartographic
references can supply information for all areas in the
United States where past land usage may pose a potential
threat to the quality of the environment, as has already
occurred in the vicinity of the town of Woburn, Massa-
chusetts, in Memphis, Tennessee, and at the Love Canal,
in New York.
As proven by studies already completed, this methodol-
ogy is fast, effective, low in cost and it has a high rate of
accuracy when tested in the field. It is hoped that utiliza-
tion of this technology will become more widespread as
cities throughout the United States examine the potential
dangers of unsound, past environmental practices.
ACKNOWLEDGEMENTS
The author wishes to acknowledge the assistance of the
following persons in obtaining information on the Woburn
area and developing the methodologies through which the
study is being accomplished: Richard T. Leighton, U.S.
Environmental Protection Agency, Region I; Thomas R.
Osberg, U.S. Environmental Protection Agency, Environ-
mental Photographic Interpretation Center; and Robert
Cleary, Massachusetts Department of Environmental
Quality Engineering.
REFERENCES
1. Tangner, P., "Hazardous Wastes: Ghosts of a Prodigal
Past" Technology Review, 8.2 (8), Aug/Sept, 1980.
2. Geraghty, J. J., "Evaluation of Hydrogeologic Con-
ditions" U.S. EPA Conference on Management of
Uncontrolled Hazardous Waste Sites, Oct. 15-17,1980,
Washington, D.C., Hazardous Materials Control Re-
search Institute, Silver Spring, Md., 49-52.
3. Sanborn Map Co., Inc., The Sanborn Fire Insurance
Map Series for Woburn, Ma., Feb 1888, June 1884,
Aug 1899, May 1904, June 1910, May 1918, May 1926,
May 1926-1947.
4. Anderson, J. R., Hardy, E.E., Roach, J.T. and
Witmer, R.E., A Land Use and Land Cover Classifica-
tion System for Use With Remote Sensor Data, Geo-
logical Survey Professional Paper 964,1976.
-------
PLANTS AS BIOINDICATORS
OF ENVIRONMENTAL POLLUTION
G.K. SHARMA, PH.D.
Department of Biological Sciences
University of Tennessee at Martin
Martin, Tennessee
CHRISTY COOPER
University of Tennessee
Center for the Health Sciences
Memphis, Tennessee
INTRODUCTION
Plant morphological and cuticular features have been
used for over a century in the interpretation of taxo-
nomic, paleobotanical, and phylogenetic relationships.
However, their use in the understanding of ecological
data is fairly recent. In addition, plant cuticular features
have recently been found to be extremely useful in estab-
lishing relationships between environmental pollution and
various taxa.
In view of the presence of acid rain and other pollutants
throughout the world, the use of plants as bioindicators
of environmental pollution in various habitats is an ex-
tremely significant contribution to the understanding of
our ecosystems. With detailed studies on a wide variety
of plant taxa, it is possible to suggest a specific relation-
ship between environmental pollution and a plant taxon
affected by it.
Numerous studies have revealed the detrimental effects
of environmental pollution on plants under natural and
controlled conditions. Scheffer and Hedgcock's study(6)
of the forests of the Northwestern United States revealed
the characteristic effects of sulfur dioxide injury on leaves.
Solberg and Adams(9) reported that fluoride and sulfur
dioxide destroyed the spongy mesophyll and the lower leaf
epidermis of plants. This was followed by chloroplast dis-
tortion, and palisade and upper leaf epidermis damage.
Chamberlain0' observed that dirt, smoke, and the gases of
a large city were fatal to conifers, especially Pinus bank-
siana. The plants illustrated chlorosis and necrosis.
Pyatt(5) investigated lichens as possible indicators of air
pollution in a steel-producing town in Wales and discov-
ered that generally the lichen flora decreased in number of
species present with increasing proximity of the source of
pollution.
Although several studies*2-4' indicate the effects of en-
vironmental pollution on the gross morphological features
of plants, relatively little work has been done to determine
the relationship between environmental pollution and the
cuticular features of plant leaves. Preliminary studies on
the subject'7-8' indicate the usefulness of leaf cuticular fea-
tures as indicators of environmental pollution in some
plant taxa. The purpose of the present study was to deter-
mine the gross morphological and leaf cuticular varia-
tions in American sycamore and their possible relation-
ship to environmental pollution. The present study is,
therefore, a continuation of a comprehensive project in-
volving plants as indicators of environmental pollution.
METHODS
American sycamore (Platanus occidentalis) is one of the
largest trees in the eastern deciduous forest of the United
States. It may grow to fifty meters in height and trunks
may be as large as four meters in diameter. The sycamore
family (Platanaceae) is composed of a single genus
(Platanus) of about seven species occurring in the North-
ern Hemisphere. The tree is characterized by broadly ovate
leaves which are 3-5 lobed. The bark of sycamore tree is
deciduous and is in the form of broad, thin brittle plates.
Sycamore grows well in rich soil such as bottom-lands. The
wood of sycamore tree is utilized for many specialized pro-
ducts such as boxes, crates, and furniture. Due to the
fungal disease caused by Gnomia veneta that often invades
sycamore trees, planting of the tree is currently not ad-
vised(3).
Four populations (A, B, C, D) of sycamore tree were
collected from different sites exhibiting varying degrees of
environmental pollution (Table I). Populations A and B
Tablet.
Distribution and Habitat Features of Sycamore Populations in
Northwest Tennessee, Middle Tennessee, and Northern Missouri,
U.S.A.
Loca-
Pop'n tion
A Reelfoot Lake,
Tenn.
B Overton Park,
Memphis, Tenn.
C Nashville, Tenn.
D St. Louis, Mo.
•+ + + +, highest level; +, lowest level.
Rel.
Pol'n
Level
Source
agricultural
operations
automobiles
automobiles
industry
automobiles,
heavy
industry
185
-------
186 REMOTE SENSING
were collected respectively in the semi-protected, semi-
wooded vicinity of Reelfoot Lake and Overton Park in
Tennessee.
Reelfoot Lake represents an extremely small, rural com-
munity in northwest Tennessee with a minimum level of
pollution. Although the Overton Park area is in the city
of Memphis, the tree samples were collected from a
wooded, semi-protected microhabitat unlike the polluted
macrohabitat of Memphis.
Populations C and D were collected from Nashville,
Tennessee and St. Louis, Missouri respectively, where
automobile traffic was an obvious source of environmental
pollution. The tree samples comprising populations C and
D were taken from the industrialized downtown areas of
Nashville and St. Louis, respectively. In both of these
large metropolitan areas, environmental pollution has re-
sulted from industrialization and urbanization.
Each tree population sample consisted of 20 leaves
collected at random from sycamore trees. These leaf sam-
ples were collected in mid-autumn to ensure their matur-
ity at the time of collection. Gross morphological features
were recorded and analyzed for the four populations
(Table II).
Table II.
Gross Morphological Characteristics*
of Platanus Occidentals
Population
Pattern A B
D
Leaflength 22.5±3.48 22.0±2.12 14.7±2.46 8.5±1.68
(cm)
Leaf width 24.3 ±3.87 22.6±1.84 I5.2±2.01 9.0±1.68
(cm)
Petiole 7.6±1.04 7.5±0.87 6.5±1.23 1.2±0.63
length (cm)
Internodal 6.9±1.48 6.0±1.01 5.1 ±0.92 2.9±0.78
distance (cm)
Length: 0.9 1.0 1.0 0.9
width
ratio
The values represent means of 20 measurements ± standard deviation.
Cuticular impressions of the upper and lower leaf sur-
faces were prepared by applying Duco® cement to the
washed and air dried leaves'10'. A small portion from the
central area of the leaf imprints was used to make cuticular
slides for the adaxial and abaxial leaf surfaces of each pop-
ulation. Cuticular features of these four populations were
analyzed by selecting at random 20 fields (n = 20) from
each microscope slide. The microscopic analysis was per-
formed using a 40x objective and lOx oculars.
RESULTS
No stomata occurred on the upper epidermis, as is gen-
erally the case in woody dicots. The number of undula-
tions in the epidermal cells were counted only on the lower
leaf surface since the undulations were not easily discern-
ible on the upper epidermis. The statistical analysis of the
data is given in Table III.
Populations A and B growing in relatively less polluted
habitats had longer leaves (22.5 and 22.0 cm) compared to
the leaf lengths (14.7 and 8.5 cm) in populations C and D
respectively. The latter populations exhibited a high de-
gree of environmental pollution in their habitats.
Leaf width measurements for populations A and B were
also higher (24.3 and 22.6 cm) than for populations C
and D (15.2 and 9.0 cm), respectively. Petioled length
and internodal distance were also greater in plant popula-
tions of less polluted habitats. The mean length-width
ratio in all the populations studied showed a very in-
significant variation (0.9-1.0) and can be regarded as con-
stant for the taxon under investigation. It is quite ap-
parent from the measurements of the gross morphological
features of the four sycamore populations, that polluted
environmental conditions had an adverse effect on the
growth of the plant.
Stomata were absent on the upper epidermis of syca-
more leaves, as is the case in many woody dicotyledons.
The trichome frequency was slightly higher on the upper
leaf surface as compared to the lower leaf surface. The
highest stomatal frequencies of 30.2 and 29.2 per unit area
(0.152 mm2) were found in populations A and B. These
two populations illustrated a minimum level of environ-
mental pollution. The leaves of sycamore tree seem to
have low stomatal frequency (26.1 and 23.3) in polluted
areas, an adaptation which may be of ecotypic signif-
icance in regulating the limited and controlled entry of
harmful gaseous pollutants into the plant tissues, espec-
ially when the tree grows in a polluted area.
Trichome frequency values of 3.3 and 3.5 for the upper
leaf surfaces of populations C and D respectively were
the highest. These two populations were exposed to the
highest levels of environmental pollution. Populations A
and B, however, had fewer trichomes. All of the trichomes
were unicellular and the majority of the trichomes were
found along the veins rather than the lamina.
The trichome shape was variable. Most of the tri-
chomes were star-shaped with 5-6 rays radiating from the
case of the trichome. These star-shaped trichomes were
prominent on the midvein of sycamore leaves. Also, sev-
eral simple pointed trichomes were noted in each popula-
tion. Populations C and D possessed the longest trichomes
(106.7 and HO.Oum), respectively. It seems quite apparent
that longer and more numerous trichomes were asso-
ciated with a high degree of environmental pollution.
These modifications or adaptations in sycamore may be of
ecotypic or even evolutionary significance. A pubescent
leaf surface may act as an insulator in a polluted environ-
ment. Leaf hairs are also known to be of value in shading
the living cells and thus reducing the temperature of the
leaf tissues.
The size of the largest and smallest stomata on the low-
er leaf surface in all the populations sampled showed
little variation. The smallest stomatal size ranged from
7.6um to 8.3 urn while the largest size ranged from 12.7p
to 14.3pm. Because of this relatively insignificant varia-
tion in the size of the two extremes within populations,
-------
REMOTE SENSING
187
Table III.
Leaf Cuticular Characteristics* of Populations
of Platanus occidentalis
Pattern
Stomatal
frequencyt
Stomatal
length (um)
Largest
stoma
Smallest
stoma
Trichome
density + +
Trichome
length (um)
Longest
trichome
Shortest
trichome
Epidermal
cell
undulations**
(number)
Subsidiary
cell complex
(number)
Population
A B
D
U
30.2±1.12 29.2±2.91 26.1 ±2.46 23.3 ±1.11
12.9±0.89 12.8±0.67 12.7±0.89 14.3±1.11
7.6±0.89 7.6±0.45 7.9±0.89 8.3 ±0.45
1.3 ±0.45 2.1 ±0.45 3.3 ±0.62 3.5 ±0.89
0.8±0.78 1.7±0.82 2.7±0.89 2.9±0.76
U 90.3±4.47 91.1 ±5.59 106.7±3.21 110.0±4.47
L
U
89.4 ±1.41 90.2 ±1.22 100.9 ±2.41 104.2 + 2.42
40.9±2.12 -56.3±4.47 62.7±4.47 80.0±1.01
39.2±2.11 53.2 + 2.23 60.2±2.82 76.4±1.84
3.5±0.89 3.6±0.78 3.1 ±0.89 3.2±0.68
3-4
3-4
3-4
3-4
'The values represent means of 20 measurements ± standard deviation.
"Mean epidermal cell undulations = undulations of the lower surface of leaves.
tMean Stomatal frequency = stomata of the lower surface of leaves observed through a 40x
objective and lOx oculars (field area = 0.152 mm1). U= upper surface of leaf; L = lower
surface of leaf.
+ +per unit area (0.152 mm!).
it may be suggested that the stomatal size range remains
fairly uniform in sycamore.
A count of the number of undulations in the epidermal
cells reveals that environmental pollution had no signifi-
cant effect on this cuticular feature. In all four sycamore
populations, the mean value for the number of undula-
tions ranged from 3.1 to 3.6. The subsidiary cell complex
remained constant, with 3-4 cells adjacent to the guard
cells. The constancy of the subsidiary cell complex in all
the populations sampled may be of taxonomic significance
for the taxon.
CONCLUSIONS
The study points out leaf cuticular and gross morpho-
logical variations in American sycamore as evidenced by
modification in leaf length, leaf width, petiole length, in-
ternodal distance, stomatal frequency, and trichome den-
sity in the four populations studied. Stomatal frequency
was low in the highly polluted habitats. The data sug-
gest that high trichome frequency may be correlated with
assumed high environmental pollution.
The number of undulations in the epidermal cells as
well as the stomatal size ranges were not affected by en-
vironmental pollution to any significant degree. Leaf
length-width ratio and the subsidiary cell complex re-
mained the same in all the populations. The latter is no
doubt a species trait and hence of taxonomic significance.
Similar results have been found in other plant taxa studied
by the author. With detailed investigations on additional
herbaceous and woody plants growing in polluted habi-
tats, it may be possible to suggest a more precise correla-
tion between plants and environmental pollution and
hence their use as indicators of environmental pollution.
REFERENCES
1. Chamberlain, C.J., "Gymnosperms: Structure and
Evolution," Dover, New York, 1934.
2. Feder, W.A., "Plant Response to Chronic Exposure
to Low Levels of Oxidant Type Air Pollution,"
Environmental Pollution, 1, 1970.
3. Gray, A., "Gray's Manual of Botany," American
Book Co., New York, 1950.
4. Mathis, P.M. and Tomlinson, G., "Lichens: Bio-
assay for Air Pollution in a Metropolitan Area (Nash-
ville, Tenn)," Journal Tenn. Academy of Science,
47, 1972.
5. Pyatt, B.F., "Lichens as Indicators of Air Pollution
in a Steel Producing Town in South Wales," En-
vironmental Pollution, 1, 1970.
6. Scheffer, T.C. and Hedgcock, G.C., "Injury to North-
western Forest Trees by Sulfur Dioxide from Smelt-
ers," U.S. Dept. Agr. Tech. Bull., 1117, 1955.
7. Sharma, O.K., "Cuticular Features as Indicators of
Environmental Pollution," Water, Air, and Soil Pollu-
tion, 8, 1977.
8. Sharma, O.K. and Butler, J., "Leaf Cuticular Var-
iation in Trifolium repens L. as Indicators of En-
vironmental Pollution," Environmental Pollution 5,
1973.
9. Solberg, R.A. and Adams, D.F., "Histological Re-
sponses of some Plant Leaves to Hydrogen Fluoride
and Sulfur Dioxide," American Journal of Botany, 43,
1956.
10. Williams, J.A., "A Considerably Improved Method
for Preparing Plastic Epidermal Imprints," Botani-
cal Gazette, 134, 1973.
-------
RATIONALE FOR DETERMINING PRIORITIES AND EXTENT
OF CLEANUP OF UNCONTROLLED HAZARDOUS
WASTE SITES
WALTER UNTERBERG, Ph.D.
WAYNE L. STONE
Rockwell International Corporation
Newbury Park, California
ANTHONY N. TAFURI
U.S. Environmental Protection Agency
Municipal Environmental Research Laboratory
Edison, New Jersey
INTRODUCTION
The Comprehensive Environmental Response, Compen-
sation and Liability Act of 1980, PL 96-510, requires re-
vision of the National Contingency Plan with a section
known as the National Hazardous Substance Response-
Plan and requires that the plan establish procedures and
standards for responding to releases of hazardous sub-
stances, pollutants, and contaminants. One aspect of this
new plan is defined as:
"105 (3) methods and criteria for determining the ap-
propriate extent of removal, remedy, and other measures
as required by this Act."
In this paper the authors attempt to develop a rationale
for determining "appropriate extent of removal" for un-
controlled hazardous waste sites. This involves three types
of decisions:
•Determination of Cleanup Priorities—To what area or
areas to direct the cleanup effort when time or resources
are limited and/or the affected area is too extensive to
clean up the entire release.
•Evaluation of Alternative Cleanup Methods and Selec-
tion of Optimum Methods—Once the priority cleanup
areas have been selected, (a) to evaluate cleanup alterna-
tives considering availability and cost of manpower and
equipment, effectiveness and speed of deployment, effect
on the environment and other applicable parameters;
and (b) to select the optimum method or combination
of methods for the priority cleanup operations.
•Determination of Extent of Cleanup (How Clean Is
Clean)—Once the optimum cleanup methods have been
selected, how far and how long should cleanup proceed?
The desirable "extent of removal" depends on (a) reach-
ing acceptably low levels of residuals, (b) the (usually
rising) cost of removal per unit mass of contaminant as
lower and lower contamination levels are achieved, (c>
environmental impact of cleaning methods themselves;
and other suitable parameters.
The authors deal, in order, with these three questions
and provide decision-making methodologies in these areas
to assist those charged with directing cleanup operations.
However, "removal" (which is primarily concerned with
cleanup) as addressed here excludes the initial response
to a hazardous substance release usually designated as
first-response actions, which protect human populations
against immediate dangers such as fires, explosions, and
ingestion of toxic materials. Likewise excluded are sub-
sequent remedial actions that stabilize the environment
after removal.
HAZARDOUS SUBSTANCES
The CERCLA legislation, in Section 101 (14), desig-
nates the substances covered by the Act. More than 600
chemical compounds and elements are now listed while
the exact number is still being decided. The physical be-
havior of hazardous substances (HS) on release can be
classified into 11 categories, (Table I), including releases
into water, into air and onto land. The principal dis-
tinguishing characteristics are density, vapor pressure,
and reactivity.
The hazards they present are categorized and coded in
Table II, and are mainly in the areas of toxicity and com-
bustibility. Figure 1 is a typical page from an alphabet-
ical computation which lists the hundreds of substances
by Chemical Abstract Service (CAS) number, the physical
behavior categories of Table I, and the hazardous effects
coded in Table II. This provides an initial guide to the
physical nature and danger of any known hazardous sub-
stance.
CLEANUP PRIORITIES
Even though first response actions have concentrated
primarily on safeguarding the human population, the haz-
ards to humans are still the paramount consideration even
in the removal phase. This is the key issue in responding
to hazardous substance incidents, in contrast to oil spills
where normally no human lives are endangered. There-
fore, oil spill response procedures are of limited use in this
discussion.
The proposed cleanup priority decision methodology
follows the flow chart of Figure 2. Inputs based on the
hazardous substance release, the first response actions,
188
-------
REMEDIAL RESPONSE 189
Table I.
Physical Behavior of Hazardous
Substances on Release
Water Spill
1. Liquids and solids dissolving in water (solubles)
2. As above, but hazardous gases/vapors result
3. Insoluble substances lighter than water (floaters)
4. As above, but hazardous gases/vapors result
5. Insoluble substances heavier than water (sinkers)
6. As above, but hazardous gases/vapors result
7. Liquids or solids that react with water
Land Spill
8. Liquids or solids that are self-reactive
9. Compressed gas that could be released
10. Liquids or solids below their boiling/evaporation point
11. As above, but hazardous gases/vapors result
Air Spill
2. Hazardous gases/vapors from solubles/water
4. Hazardous gases/vapors from floaters/water
6. Hazardous gases/vapors from sinkers/water
8. Self-reactive liquids or solids
9. Compressed gases expanding
11. Hazardous gases/vapors from slow evaporators
Table II.
Hazardous Effects Categories and Codes
a toxicity data not available
b combustible with ignition source (low vapor pressure)
c potentially toxic combustion/reaction products
d Highly toxic combustion/reaction products
e Do not drink contaminated water
E Explosion hazard
f Do not inhale toxic vapor/dust
g Readily flammable (high vapor pressure)
h Corrosive
T Toxicity hazard
() Parentheses indicate a lower hazard or property value
Note: Italicized and capitalized letters indicate serious hazards.
climatic conditions and the nature of the affected or
threatened environmental constituents are entered into
the Hazardous Substance Incident Profile Form from
which long- and short-term damage estimates are de-
rived. This form is dependent on the time and date and
may be filled out at different times during the cleanup ac-
tion. The form serves as a record of the overall event and
its consequences.
The generalized environmental response priorities are
shown on Table III. Normally, the priorities are to guard
against, in order, human, animal, ecological, socioeco-
nomic and social impacts. Also affected constituents
should be protected before threatened constituents. Priori-
ties, however, may be changed if special circumstances
indicate that the situation requires it.
A numerical Vulnerability Rating System was devel-
oped to establish cleanup priorities among environmental
constituents. This is illustrated in the worksheet of Figure
3. There are six additive vulnerability parameters, with a
maximum possible score of 100. Parameters 1 through 3
are hazardous substance-related while parameters 4
through 6 are constituent-related. Further, there are three
subtractive constituent-related parameters (numbers 7
through 9) indicative of the background contamination
level and assimilative capacity and any self-cleaning abil-
ity.
The numerical ratings are guides to cleanup priorities
(higher vulnerability ratings mean higher cleanup priority)
subject to two considerations: the priorities among classes
of constituents (axiom 1 on Table III), i.e., whether they
are in the order A, B, C, etc. as shown on Figure 3; and
the priority of affected constituents over threatened con-
stituents (axion 2 on Table III). The judgment of the user
must make the final priority assignment.
SELECTION OF CLEANUP TECHNIQUES
The number of hazardous substances cleanup strate-
gies that can be applied to different situations is almost
limitless due to the complexity of the parameters in-
volved. Hazardous substances themselves possess a va-
Tablelll.
Generalized Environmental Response Priorities
Axiom 1: Normally protect environmental constituents in the
following order of priority:
a. Human populations and habitats—human impact
Cleanup crews
General population
b. Fauna—animal/ecological impact
Marine
Land
Birds, etc.
c. Flora—ecological impact
Forests
Food crops
Other crops
Marshes
Marine, etc.
d. Property—socioeconomic impact
Industry
Business
Municipal
Fisheries, etc.
e. Aesthetic and recreational areas—social impact
Parks
Beaches
Wilderness
Water use and sports areas, etc.
Axiom 2: Protect affected areas before threatened areas.
Axiom 3: Use judgment and common sense in applying Axioms
1 and 2.
-------
190 REMEDIAL RESPONSE
COMPOUND
Acenaphthene (83-32-9)
Acenaphthylene (208-96-8)
Acetaldehyde (75-07-0)
Acetic acid (64-19-7)
Acetic anhydride (108-24-7)
Acetone cyanohydrin (75-86-5)
Acetyl bromide (506-96-7)
Acetyl chloride (75-36-5)
Acrolein (107-02-8)
Acrylonitrile (107-13-1)
Adipic acid (124-04-9)
Aldrin (309-00-2)
Allyl alcohol (107-18-6)
Allyl chloride (107-05-1)
Aluminum sulfate (10043-01-3)
Amnonia (7664-41-7)
Ammonium acetate (631-61-8)
Ammonium benzoate (1863-63-4)
Ammonium bicarbonate (1066-33-7)
Ammonium bifluoride (1341-49-7)
Ammonium bisulfite (10192-30-0)
Ammonium carbamate (1111-78-0)
Ammonium carbonate (506-87-6)
Ammonium chloride (12125-02-9)
Ammonium chromate (7788-98-9)
Ammonium citrate, dibasic(3012-65-5
Ammonium di chromate (7789-09-5)
Ammonium fluoborate (13826-83-0)
Ammonium fluoride (12125-01-8)
Ammonium hydroxide (1336-21-6)
WATER SPILL
1
o
v»
O
o>
(X)
(X)
2
1C
CU
N
a.
o
c
CU
o
X E,T
X T
(X)'E,T
X T
X T
X ' T
X
X
(X) 1
X '
(X)
X
X
X
X
X
X
X '
X j
X
X I
x ,
X '
X i
X
X 1
X .
1
E,T
E,T
E,T
E,T
T
T
3 4
-n 1C
— ' Ol
o n
rt -!
v> O-
O
C
at
-a
o
-s
1
X 1
X E,T
X T
1
X 'E,T
X |E,T
1
X E.T
X 'E,T
i
1
5
GO
3
l/l
X
I
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
(x)
6
JC
O)
N
-i
a.
o
c
I/I
P)
o
~*
T
E,T
T
T
7
jO
n>
Ol
£
m
X
X
LAUD SPILL
8
,-,
CD
Ol
£
X
x
9
0
o
-1
-s
n>
CL
o
in
x
10
in
o
m
Ol
o
2
ft-
0
13
X
X
X
X
X
X
X
X
X
X
11
-r-
Ol
N
-I
o.
o
c
I/I
Ol
•o
o
-1
E.T
T
E.T
T
I
E,T
E.T
E,T
E,T
1
1
X T
NOTES
b, c, e, f
b, c, e_, f
f, g_
skin, eye hazard, f_, %, h
skin, eye hazard, f, g_, F
poison (cyanide), d_, e, f
^violent reaction with water, eye haz-
j ard, skin burns, irritant, £, £, h^
skin, eye hazard, §., £, g_
Poison (cyanide), can explode u,e,f,£
b, f, irritant
skin poison, b, c, e_, f
eye hazard, £, g_
eye hazard, skin poison, f_, g_
f, tu choking gas
e, I
etches glass, e_, f, ^
sulfur dioxide formed with acids, £
e , I
oxidant, £_, £, g_, h^
e, f., li
etches glass e_, f, h.
f , h
Figure 1.
Typical Page From a Listing of Hazardous Substances
riety of physical, chemical and toxicological combina-
tions so as to make a fairly standard cleanup and treat-
ment methodology, sueh as for oil spills, not appropriate
or applicable. It is therefore impractical to attempt to
develop a specific cleanup methodology for each haz-
ardous substance and spill situation combination. Instead,
an attempt is made to present for each hazardous sub-
stance physical behavior release class (Table I) a variety
of spill treatment techniques that may be useful in clean-
ing up hazardous substances falling within a particular
class.
The treatment techniques, which are listed in Table IV,
fall under the classifications of mechanical, chemical,
and sorbents, gels and foam methods. The treatment tech-
niques are geared toward both containment and removal
actions and include the following cleanup strategies:
(1) Dilution and dispersion into the environment
(2) In-situ treatment
(3) Offsite treatment and disposal
(4) Treatment on site in an offstream treatment system
VULNERABILITY RATINGS OF
AFFECTED/THREATENED
ENVIRONMENTAL
CONSTITUENTS
(Table 3)
(F19. 3)
Figure 2.
Cleanup Priority Decision Flow Chart
-------
REMEDIAL RESPONSE 191
Method
Use/ Application
Table IV.
Hazardous Substances Cleanup Techniques
Advantages
A. MECHANICAL TECHNIQUES
1. Dispersion
2. Dilution (Water)
3. Dilution (Air)
4. Booms
5. Weirs
6. Spill Herding
Employ water jet, mechanical
mixer, aerator, etc. to promote
dilution. Should be used where
adequate dilution can be attained.
Consider water use classifica-
tion, and the ability for natural
or forced dilution.
Use blower equipment to intro-
duce clean air to lower hazardous
substance vapors to less than
TLV or lower flammability limit.
Use in calm and sheltered areas.
Need deployment device. Use
in fairly calm waters.
Need a weir and boat. Use in
calm, large waterbody.
Apply chemical herder to pro-
tect shore or other facilities.
Inexpensive, easily implemented
Inexpensive and timely
Can direct air away from
populated areas
Can be used over a larger area.
Many varieties.
Easily used.
Useful in rough water.
Disadvantages
Pollution may be spread over
a larger area
Pollution may be spread over a
larger area.
Pollution may be spread over a
larger area. Need large-capacity
blowers.
Cannot be used in rough or
rapidly rapidly moving water.
Cannot be used in rough water.
Not always available.
Not easily obtainable; not 100%
effective.
7. Dredging
8. Burial
9. Dikes (Earth)
10. Dikes (Foamed
Polyurethane)
11. Trenches
12. Diversion
13. Dam (Underflow)
14. Mist Knockdown
15. Vacuum Skimmer
Uses in hazardous substances
spills is largely untested.
Use only as a temporary meas-
ure to retard spread of spill un-
til recovery, or when spill sub-
stance is harmless.
Create barrier with bulldozer
or earthmoving equipment.
Need trained personnel to con-
struct on hard, dry surfaces.
Construct with bulldozer or
earthmoving equipment.
Use earthmoving equipment to
divert channel and sandbags,
dikes, etc. to block entrance
of pollution.
Create dam and underflow drain
in relatively small stream.
Use fine mist spray to suppress
water-soluble low-lying
hazardous vapors.
Use vacuum tank to generate
draining force for removing
small spills.
B. CHEMICAL TECHNIQUES
1. Neutralization
Can effectively remove
pollutants.
Easily implemented.
Material on site; common
equipment.
Can hold up to several feet
of water.
Construct with common
equipment.
Separates contamination.
Simple to construct.
Reduces air pollution potential.
Easily used.
Apply acid or base in powder
or slurry form.
Effective in restoring pH to
proper range.
Equipment not readily avail-
able; potential for environmen-
tal damage; generation of large
amounts of contaminated
material.
Not an environmentally sound
practice.
Composition of soil may not be
suitable (possible leakage).
Equipment hard to obtain;
leaks on wet ground.
Move large amounts of material;
highly permeable soil may not
contain spill.
Large amounts of earth must be
moved. Clear area and proper
soil conditions required.
Can only be used in small
streams.
Can create a water pollution
problem, runoff must be con-
tained.
Equipment not always obtain-
spills only.
Added chemicals may be a
threat to the environment.
Difficulty in determining cor-
rect amount of chemical
addition.
-------
192 REMEDIAL RESPONSE
2. Precipitation
3. Chelation
Add chemicals to clarify or pre-
cipitate out soluble substances.
Chalating agents are added to
water to bind (and deactivate)
metal ions.
Can be performed in situ or in
field treatment unit.
Effective in removing spilled
metal ions.
C. SORBENTS, GELS AND FOAMS
1. Activated Carbon
2. Polyurethane
Foam
3. Ion Exchange
4. Gelling Agent
Implement treatment by (1) drag-
ging AC-filled bag through water,
(2) passing water through AC
beds or columns, or (3) employ-
ing mobile treatment unit. Used
mostly for organics.
Apply via foam-laden belts or
shredded foam broadcast over
the spill to absorb pollutants.
Water must be pumped through
bed or column used mostly for
acid, base, metal, or toxic salt spills.
Adds gels through dispersion de-
vice to reduce viscosity (and
immobilize) spill.
Most universal sorbent; read-
ily available.
Can be squeezed and reused.
Effective removal method for spe-
cific pollutants
Effective containment method;
can reduce vapors.
Ultimate reaction, including
paniculate matter formed, may
pose a new threat to environ-
ment.
Misapplication of treatment
may cause environmental dam-
age. Essential metal nutrients
may be inadvertently removed.
Costly if carbon is not regen-
erated.
Bulky and difficult to transport.
A large amount of equipment
must be deployed. Resins must
be regenerated.
Practical use for small volumes
only; difficult to obtain; creates
a disposal problem; heat may be
generated in the reaction.
^^^^^ Environmental
^"^~^^^ Constituents
Vulnerability ^"~"-^-^__^
Parameters — .
1. Susceptibility to this
HS via pathways of:
a. Air (0-4)
b. Surface Water (0-4)
c. Groundwater (0-4)
d. Land Surface (0-4
e. Subsurface (0-4)
2. Proximity to Release Pt. (0-12)
3. Contamination Vector (0-12)
4. Population/Size/Extent (0-16)
5. Intrinsic Values
a. Hater Quality (0-4)
b. Air Quality 0-4)
c. Food Chain 0-4)
d. Ecological (0-4)
e. Economic (0-4)
f. Aesthetic (0-4)
g. Recreational (0-4)
6. Time Restrictions (0-12)
INTERIM TOTALS (100 MAX)
SUBTRACT (for this HS):
7. Assimilative Capacity (0-12)
8. Self-Cleaning Ability (0-12
9. Background HS Level (0-12)
RATING TOTALS
A. Human Populations
and Habitats
Al
A?
A3
A4
AS
A6
A7
A8
B. Fauna
Bl
B?
B3
B4
B5
B6
C. Flora
Cl
C2
C3
C4
C5
C6
D. Property
°1
D2
D3
D4
D5
D6
E. Aestheti
tecreationa
El
E2
3
E4
c and
Areas
E5
6
FigureS.
Vulnerability Ratings of Environmental Constituents
-------
REMEDIAL RESPONSE 193
Table V.
Hazardous Substances Cleanup Selections
Physical Behavior Class
(From Table I)
Potential Cleanup Altern-
atives (From Table IV)
Physical Behavior Class
(From Table I)
Potential Cleanup Altern-
atives (From Table IV)
1. Liquids and Solids Dissolving in
Water (Solubles)
2. As Above, but Hazardous
Gases/Vapors Result
3. Insoluble Substances Lighter
Than Water (Floaters)
4. As Above, but Hazardous
Gases/Vapors Result
5. Insoluble Substances Heavier
Than Water (Sinkers)
6. As Above, but Hazardous
Gases/Vapors Result
7. Liquids or Solids that React
with Water
A.I
A.2
A.12
B.I
B.2
B.3
C.I
C.3
As above, but add
A.3
A. 14
C.2
C.4
A.I
A.2
A.4
A.5
A.6
A.12
A.13
A.15
C.2
C.4
As above, but add
A.3
A.14
A.7
As above, but add
A.3
A.14
A.I
A.2
A.3
A.4
A.5
A.6
A.7
A.12
8. Liquids or Solids that are
Self-Reactive
9. Compressed Gas that Could
be Released
10. Liquids or Solids Below Their
Boiling/Evaporation Point
11. As Above, but Hazardous
Gases/Vapors Result
A.13
A.14
A.15
B.I
B.2
B.3
C.I
C.2
C.3
C.4
A.3
A.8
A.9
A. 10
A.ll
A.14
A.15
B.I
C.2
C.4
A.3
A.8
A.9
A. 10
A.ll
A.14
A.15
B.I
A.8
A.9
A.10
A.ll
A.15
B.I
C.2
C.4
As above, but add
A.3
A.14
The potential cleanup alternatives that may be applic-
able, depending on the particular substance spilled and
spill situation, for each physical behavior class are listed
in Table V. Figure 1 will aid in determining into which
physical behavior class a particular hazardous substance
falls.
Once the potential cleanup alternative (or combination
of alternatives) is selected, several options will probably be
able to be eliminated for favored due to obvious limiting
factors such as available equipment, manpower, financial
resources, etc. Also, the spill cleanup advantages/dis-
advantages listed in Table IV should be seriously con-
sidered. If no cleanup selection is feasible or if significant
environmental damage will result from cleanup actions,
no action at all may be the appropriate response.
DETERMINATION OF ALTERNATIVE
CLEANUP TECHNIQUE COSTS
After a potential spill cleanup technique or combina-
tion of techniques has been chosen, costs for each selec-
tion can be determined from past experience, on-scene
advisors, cleanup contractors or any other references at
hand. The costs to be included, where applicable, are:
•Labor
•Capital or rental costs
•Material and supplies
Daily cleanup costs ($/day) for each cleanup technique
can be computed by adding the appropriate cost compon-
ents.
•Packaging and transportation
•Ultimate disposal
-------
194 REMEDIAL RESPONSE
For each cleanup technique or techniques for which a
daily cleanup cost ($/day) was calculated, an attempt can
be made to try and obtain an estimate of the amount of
hazardous substance each cleanup technique is capable of
extracting on a mass per day (Ib/day) basis. This par-
ameter in most cases cannot be easily identified. There-
fore, estimates of extraction performances should be made
from past experiences of the OSCs or their advisors, from
cleanup contractors, or from cleanup equipment manu-
facturers' data.
Unit cost ($/lb) is defined as the amount of dollars re-
quired to extract one pound of hazardous substance. This
can be calculated by dividing the daily cleanup cost rate
($/day) by the extraction performance rate (ibs/day). Al-
ternative cleanup techniques can now be compared on a
unit cost basis to get an indication of relative economic
merit.
DETERMINATION OF MOST EFFECTIVE
CLEANUP TECHNIQUE
In many spill situations, it may not be realistic to choose
cleanup techniques solely on a unit cost ($/lb) basis. Al-
ternative cleanup techniques can be ranked on a unit cost
basis but the listing priority should be qualified with cer-
tain cleanup effectiveness criteria that will be affected by
the peculiarities of a specific spill situation and may out-
weigh any cost considerations. The cleanup effectiveness
criteria include:
•Completeness of removal—Can the desired or required
cleanness levels or standards be attained?
•Rate of removal—Can the substance(s) be cleaned up in
time to protect critical habitats or reduce evacuation
times, etc?
•Logistics—Can the techniques be easily deployed? Neces-
sary equipment? Is it available?
•Operational requirements—What levels of labor, skill,
etc., are required? Is it affected by weather?
•Environmental impact—Will the environment be adverse-
ly affected?
•Personnel protection—Will response personnel be un-
duly subjected to hazards?
•Disposal problems—Will large amounts of wastes be gen-
erated? Will this result in transportation and ultimate dis-
posal problems?
After consideration of the preceding cleanup effective-
ness criteria, the cleanup technique priority list can be re-
ordered to obtain the effective cleanup strategy for a par-
ticular hazardous substance incident.
EXTENT OF CLEANUP (HOW CLEAN IS CLEAN?)
Based on the results of the two previous sections, a spe-
cific strategy will have been selected to clean up the en-
vironmental constituents having the highest priorities. This
strategy will consist of one or more cleanup actions at
the same time, followed sequentially by other actions. For
instance, simultaneous mechanical and chemical tech-
niques may be used to extract and treat the hazardous
substance at the incident site, followed by transportation
0.9-
~ 0.7-
1 0.3-
3
o.e-
O.L
PUT KEEH.V.
comma)
CUMULATIVE
CLEANUP COSTS
~l
STATUTORY
ENVELOPE
751
" ENVELOPE
12 16
Time, Weeks
Figure 4.
Control Graph for Cleanup Time and Cost
HS hazardous substance
A larger constant effectiveness
B smaller constant effectiveness
C effectiveness decreases as
cost rises
ACCEPTABLE
LEVEL
CLEANUP COST
Figures.
Cleanup Effectiveness Curves
Table VI.
Factors Influencing Extent of Cleanup
1. Acceptably low final contamination levels
2. Change in effectiveness of cleanup techniques as contamina-
tion levels decrease
3. Change in incremental cleanup costs as contamination levels
decrease
4. Environmental impact of cleanup techniques themselves
5. Effect of simultaneous vs. sequential cleanup of different en-
vironmental constituents
6. Overall and ultimate environmental consequences—assimila-
tive capacity of environment
7. Public awareness and political pressures
8. Budget and time constraints (Superfund: $1 million and 6
months)
-------
REMEDIAL RESPONSE 195
of the extracted hazardous matter to another location for
disposal. These various techniques may be combined in
simultaneous/sequential/intermittent ways, as needed.
During the removal or cleanup phase one must initially
project and later keep track of three main parameters:
•Daily cost of individual cleanup techniques and of the
total cleanup effort, including cost of measurements in
item 3 below
•Daily amount of hazardous substance extracted
•Daily or periodic change in the hazardous substance re-
sidual contamination level in the environmental con-
stituents affected by the incident, based on standardized
meaningful measurements
The factors influencing the extent of cleanup during the
removal phase are listed in Table VI. Unfortunately, all
factors, except budget and time constraints which are dic-
tated by outside considerations, are difficult to pin down.
Clearly, the basic goal of the cleanup operation is to re-
duce the level of toxic substances to an environmentally
safe level, i.e., a level that does not present a hazard to
human health or welfare, including water supplies and
crops. Unfortunately, there are virtually no quantitative
guidelines concerning acceptable residual levels of toxic
substances for spill cleanup operations. Many of the
metals whose salts are considered hazardous occur natural-
ly in soils at low concentrations and are important trace
elements in the biosphere. As a basic guideline, ideal clean-
up operations should reduce the levels of toxic substances
back to the natural levels found in the general area of the
spill.
In certain cases, specific numbers for "safe" levels of
contaminants are given in various regulations, particularly
for drinking water and substances regulated under the
Clean Air Act.
Although quantitative guidelines for cleanup standards
have not yet been established, it is possible to go through
the list of toxic substances and identify those that will re-
quire extensive cleanup because they present severe en-
vironmental hazards. Thus polychlorinated biphenyls,
polynuclear armomatic compounds, various halogenated
aromatic and aliphatic hydrocarbons, nitrosoamines and
certain heavy metal salts of mercury, beryllium, and cad-
mium, among others, would merit special attention.
From a general consideration of the chemical, physical,
and toxicological properties of the toxic substances listed,
it is possible to perform a crude stratification into sub-
stances that are likely to present long-term environmental
hazards and require extensive cleanup and substances with
higher vapor pressures or greater chemical reactivity that
present a relatively lower long-term environmental hazard
and may require less extensive cleanup. A tentative strat-
ification scheme along those lines involving four groups
Date:
COST PARAMETERS (CP)
1. CLEANUP LABOR (0-3)
2. EQUIPMENT (0-3)
3. SUPPLIES & CHEMICALS (including fuels, lubes, power) (0-3)
4. CLEANUP IMPACT (allow for persistence and (0-3)
assimilative capacity)
5. AREA USE INTERFERENCE (due to continued cleanup) (0-3)
TOTALS ICP:
BENEFIT PARAMETERS (BP)
6. ECONOMIC (normalization of income from area) (0-3)
7. ENVIRONMENTAL (protection of ecosystems and habitats) (0-3)
8. AESTHETIC (preservation of visual quality & scenery) (0-3)
9. RECREATIONAL (normalization of indoor & outdoor areas) (0-3)
10. PUBLIC (normalization of public access & use) (0-3)
EBP:
TOTALS
EBP ZCP:
ENVIRONMENTAL CONSTITUENTS
A.
CONTINUE CLEANUP? YES/NO:
(YES IF X»0. NO IF X«0. GET MORE FACTS IF X IS BETWEEN +3 AND -3)
Figure 6.
Cost/Benefit Analysis of Incremental Cleanup
-------
1% REMEDIAL RESPONSE
of inorganic and seven groups or organic compounds is
presented in Table VII.
The factors listed in Table VI can be condensed into
three indicators for use in deciding the extent of removal,
or, "How Clean is Clean?"
•Ability to approach or reach acceptable residual con-
tamination levels
Table YD.
Classification Scheme for Long-Term Environmental
Hazards of Inorganic and Organic Compounds
PART 1: INORGANIC MATERIALS
GROUP II: Inorganic metal salts likely to require extensive
cleanup. Long-term environmental hazards. Drink-
ing water hazards. Likely to bioaccumulate.
GROUP 12: Inorganic nonmetal compounds likely to require ex-
tensive cleanup. Long-term environmental hazards.
Drinking water hazards. Likely to bioaccumulate.
GROUP 13: Inorganic metal salts presenting long-term en-
vironmental hazards, but likely to require less ex-
tensive cleanup than Group II.
GROUP 14: Inorganic materials less likely to present long-term
environmental hazards requiring extensive cleanup
operations. May present severe short-term toxic or
explosion hazards.
PART 2: ORGANIC MATERIALS
GROUP 01: Pesticides, likely to present long-term environ-
mental hazards, drinking water hazards and bioac-
cumulation hazards. Extensive cleanup operations
likely to be required.
GROUP 02: Nitrosoamines: Highly potent carcinogens, severe
drinking water hazard, although long-term biode-
gradation likely. Extensive cleanup operations like-
ly to be required.
GROUP 03: Polynuclear aromatics (PNAs): Potent carcin-
ogens, low volatility, long-term environmental haz-
ard, water and bioaccumulation hazards. Likely to
require extensive cleanup.
GROUP 04: Halogenated hydrocarbons and related com-
pounds: Known or likely carcinogens, probable bio-
accumulation, dkrinking water hazards, likely to re-
quire extensive cleanup.
GROUP 05: Aromatic phenols, amines, nitro derivatives and re-
lated compounds: Known or likely carcinogens,
probable bioaccumulation, drinking water hazard.
Biodegradation possible, may present less long-
term hazard than Groups 01 through 04.
GROUP 06: Amines, nitroles, aromatics and other compounds
presenting environmental hazards, known or sus-
pected carcinogens, drinking water hazards, but
more reactive or volatile than previous groups and
less likely to present long-term environmental
hazards.
GROUP 07: Hydrocarbons, acids, esters and other compounds
that are volatile or reactive enough to be biode-
gradable. Present lower long-term hazards and less
likely to require extensive cleanup. May present
severe short-term explosion or toxic vapor hazard.
•Comparison of costs versus benefits of incremental clean-
up as a test of whether to continue cleanup
•Time and/or total cleanup cost limitations
Cost Benefit Analysis
It is desirable to reach acceptable residual contamina-
tion levels in the contaminated environmental constituents
and prevent threatened constituents from being affected.
Further, cleanup costs should not exceed the benefits that
they provide. Lastly, time and/or total cleanup cost limita-
tions, statutory or otherwise, should not be exceeded. In
principle, all three conditions should be satisfied simultan-
eously. In practice, the connection between increasing
costs, decreasing contamination levels, and increasing
benefits can be tracked on an elapsed time basis. If all
three conditions cannot be satisfied, then a compromise
must be reached. The procedure detailed below is designed
to furnish the user the needed data to make a decision.
First, daily cleanup costs and cumulative committed
weekly cleanup costs are charted, for instance on a graph
like Figure 4. This gives the overall picture for removal
costs. Two rectangular envelopes are pre-drawn on this
graph: the Superfund statutory limitation of $1 million
and 6 months, and (in dashed lines) the envelope at 75% of
these amounts for advance planning and control purposes.
Next, a plot of effectiveness of cleanup should be made.
This is a plot of Residual HS Level vs Total Cleanup
Cost, separate curves and acceptable residual level lines
being drawn for each pathway or environmental constit-
uent, like Figure 5. If the cleanup effectiveness is con-
stant over the cleanup period, then the curve becomes an
inclined straight line, showing that equal increments in
cost produce equal decrements in contamination level. If
the decrements become smaller while cost increments stay
the same, i.e., the curve becomes flatter, then the clean-
up effectiveness for the particular pathway drops as lower
contamination levels are reached.
Third, make a Cost Benefit Analysis of Incremental
Cleanup as shown in Figure 6 which is a form on which
both cost parameters and benefit parameters are rated
for the environmental constituents that are benefitting
from cleanup. This analysis is similar to the one used in
another study.'" If the benefit score exceeds the cost score,
then it is useful to continue cleanup; if the difference is 3
points or less, more facts are needed for a decision; and
if the cost score exceeds the benefit score, then cleanup
should be stopped. The analysis should be repeated at later
times to see if the situation has changed.
To make the decision on How Clean is Clean, or when to
stop cleanup, at different points in time:
•Determine if the residual HS contamination is at the ac-
ceptable level for the constituents of interest (Figure 5)
•Make the cost/benefit analysis whether or not to continue
cleanup for the constituents of interest (Figure 6)
•Determine if the total cleanup costs are within the speci-
fied cost-time envelope (Figure 4)
If the residual contamination is at the acceptable level
within the specified cost-time envelope, cleanup obviously
should be stopped. If the acceptable level has not been
reached, but the costs of continuing outweigh the benefits,
-------
REMEDIAL RESPONSE 197
though time and funds remain, then a closer look must be
taken as to the impact of a nonacceptable residual level. If
funds and time run out before an acceptable level is
reached and the cost/benefit analysis indicates that clean-
up should continue, strong consideration should be given
to continuing cleanup as part of remedial action. The user
of this rationale must find a solution which is environ-
mentally, economically, and legally acceptable. Often a
compromise is the only possible way.
CONCLUSIONS
This study was prepared to assist on-scene decision-mak-
ing personnel during removal activities following acciden-
tal releases of hazardous substances, as required by Public
Law 96-510, or CERCLA. Decisions to be made are on
cleanup priorities (what to clean up first?), on the selec-
tion of cleanup techniques (how to do it?), and on the ex-
tent of removal (how clean is clean?). The section on
Cleanup Priorities contains forms on which to record the
hazardous substance incident profile, and to work out the
vulnerability ratings of all the environmental constituents
affected or threatened by the incident. From these the
cleanup priorities by environmental constituents are de-
rived. The section on Selection of Cleanup Techniques
lists various mechanical, chemical and biological tech-
niques and indicates how choices should be made in par-
ticular cases. The section on Extent of Cleanup (How
Clean is Clean?) contains forms and graphs on which to
enter projected and actual cleanup costs, the progress in
cleanup as evidenced by decreasing residual hazardous
substance levels, and cost-benefit analyses of incremental
cleanup actions. The user derives the environmental, eco-
nomic, and legal decision indicators that must be com-
bined to determine the appropriate extent of cleanup.
ACKNOWLEDGMENTS
This study was prepared as part of Contract No. 68-
03-3014 with the Environmental Protection Agency.
Special thanks is given to Mr. Frank Freestone, EPA
Oil and Hazardous Materials Spills Branch, for providing
technical direction for this project. Appreciation is also
given to Dr. Roy Clark, Rockwell International, for his aid
in preparing Figure 1 and Tables II and VII.
REFERENCES
1. Versar, Inc. "Handbook for Oil Spill Protection
Cleanup Priorities." EPA-600-8-81-002. USEPA, Cin-
cinnati, Oh. February 1981.
2. S&D Engineering Services, Inc. "Training Course Out-
line: Hazardous Material Incident Cleanup Decision-
Making." USEPA, Edison, N.J. February 1981.
3. Robinson, J.S. "Hazardous Chemical Spill Cleanup.
Noyes Data Corporation." Park Ridge, N.J. 1979.
-------
HAZARDOUS SUBSTANCE RESPONSE
MANAGEMENT MODEL
J. BILL HANSON
RICHARD STANFORD
U.S. Environmental Protection Agency
Office of Emergency and Remedial Response
Washington, D.C.
ROBERT W. PEASE, JR.
PAUL J. STOLLER
The MITRE Corporation
Bedford, Massachusetts
INTRODUCTION
The Environmental Protection Agency (EPA) is now
using on an interim basis a Hazardous Substance Re-
sponse Model as a management tool to lead govern-
mental project managers through the activities related to
cleaning uncontrolled releases of hazardous substances.
Response funds are available primarily from the Com-
prehensive Environmental Response Compensation and
Liability Act of 1980, P.O. 96-510 ("CERCLA" or
"Superfund"). Superfund was enacted on December 11,
1980 and establishes broad Federal authority to respond to
releases or threats of releases of hazardous substances,
pollutants or contaminants from vessels and facilities.
The Federal and State governments under the Act may
take response actions whenever there is a release or a
substantial threat of a release which may endanger public
health or welfare or the environment. The Model pre-
sents in Flow Chart form (Figure 1) a suggested master
network of activities designed to orchestrate a cleanup
project in a manner that approximates the process of re-
sponse currently being used on an interim basis by the
EPA. The Model, in its completed form, is currently be-
ing developed and will include management-related ac-
tivities, community relations, government/agency coordi-
nation, contractor procurement, contractor monitoring
and technical decision-making.
The goals of the Model are: (1) to describe the ac-
tivities necessary to manage cleanup of an uncontrolled
hazardous waste site; and (2) to present this information
in a form that is easy to understand, and which will
assist those responsible for the day-to-day management
of a project. In short, it is comprehensive in scope and
designed for practical use. The Model is a vehicle in-
corporating the requirements of CERCLA with the ex-
perience of the people who are well-versed in evaluat-
ing, planning and cleaning releases.
The use of a management model provides the follow-
ing benefits:
Information Transfer
The model allows new projects to benefit from past
experience and to conform to CERCLA requirements by
identifying the critical decisions in the project which must
be made before succeeding activities can begin. Addi-
tionally, it provides a checklist of items which should be
considered (but not necessarily performed, depending
upon the circumstances of the specific project), thus
ensuring that important components of the cleanup are
not overlooked.
Funding Decisions
The model organizes cleanup operations into discrete
phases with funding and decision points between each
phase. Therefore, the performance of a project and the
justifications for continuation come under pre-established
review by the funding authorities. This allows project
tracking on a large scale and improved cost estimation
because of sharper definition and limitation of subse-
quent activities.
Goal-Oriented Sequencing of Activities
The model arranges all site responses into activities
which accomplish the immediate needs of protecting the
public health while providing support for the development
of permanent remedial measures. Use of the model helps
to avoid unnecessary or counterproductive activities (and
their associated cost) and serves to ensure that neces-
sary activities or data collection are not overlooked.
Public Support and Participation
The model serves to facilitate public support and par-
ticipation by laying out in advance the purposes and
limitations of each operational phase. In this way, pro-
ject managers avoid generating unrealistic expectations.
Project Continuity
The use of the model will define the role of State and
Federal decision makers and thereby ensure that the con-
tinuity of a project is not lost if there is a changeover to
personnel. An additional benefit arises from the standardi-
zation of projects and provision of program consistency.
198
-------
REMEDIAL RESPONSE 199
PHASES OF RESPONSE
There are two general types of removal response-
immediate removal and planned removal. An immediate
removal may be taken when action must be taken within
hours or days to prevent significant harm to human life
or health, or the environment. This is the type of re-
sponse which has been traditionally undertaken for oil
and hazardous substance removal. Planned removals are
taken to stabilize releases where several weeks or months
are available to identify responsible parties, plan the re-
sponse and conduct procurement procedures. Planned re-
movals bridge the gap between immediate response (which
must occur within hours or days) and remedial response
(which can take a year or more to institute).
Phase I—Discovery and Notification
Sites requiring response are identified in several ways.
First, owners and operators of inactive hazardous waste
sites were required to notify EPA by June 9, 1981.
Second, the general public can inform the National Re-
sponse Center or the EPA of potentially hazardous
sites. EPA is also using aerial photography to identify
potential sites.
Phase II—Preliminary Assessment
Based on the available information, preliminary site
assessments can be conducted immediately or deferred
to be included as part of an area-wide search.
The purpose of the preliminary assessment is to:
•identify the source and general nature of hazardous sub-
stances;
•estimate the potential of the threat;
•initially determine whether potential responsible parties
will undertake site evaluation and response activities;
•determine whether immediate measures are required.
To make these determinations, the site manager evaluates
previous disposal practices, information from generators,
aerial and other photographs, and literature searches.
Personal interviews are useful at this phase. A perimeter
inspection may be needed to determine the potential for
releases. Finally, an on-site inspection may be performed
if the release may present an immediate threat to public
health and if sophisticated safety equipment is not needed.
Phase III—Immediate Removal
Immediate removal actions can be initiated at any time
during response to address acute situations. These are
situations where response actions must begin within hours
or days to mitigate fire or explosion hazards, to protect
drinking water supplies, or otherwise to prevent or miti-
gate harm to human life or health or to the environment.
Immediate removal actions should be conducted by
the responsible party. Only if the responsible party is not
immediately available or cannot rapidly respond, can
CERCLA funds be used.
By statute, immediate removal actions should be com-
pleted within six months and with less than $1 million.
However, if acute harm still threatens the environment or
public health, immediate removal actions may continue
until the situation is stabilized.
Phase IV—Evaluation and Determination
of Appropriate Response
To determine the nature and extent of the release and
the priority of the site for response, greater emphasis is
given to on-site sampling and analysis, especially when
there is an apparent risk to public health or environ-
ment. Activities during this phase should include:
•assessing amount, types and location of hazardous ma-
terials
|. DISCOVERY Cr
NOTIFICATION
||- PRELIMINARY
ASSESSMENT
HI- IMMEDIATE REMOVAL
IV- EVALUATION & DETERMINATION
OF APPROPRIATE RESPONSE
• BASED ON AVAILABLE
INFORMATION
• INITIAL DETERMINATION
OF SOURCE ft NATURE OF
POLLUTANTS
• SEARCH FOR RESPONSIBLE
PARTY
V- PLANNED REMOVAL
• PRIORITY SITE OR TO PREVENT
RAPIDLY DETERIORATING CONDITIONS
• CONSIST OF OPERABLE UNITS
• COST LESS THAN 11 MILLION
• TIME REQUIRED LESS THAN 6 MONTHS
VI-REMEDIAL RESPONSE PLANNING
» ACTION NEEDED IN HOURS
OR DAYS
• PREVENT/MITIGATE
IMMEDIATE THREAT TO
HUMAN LIFE Er HEALTH
OR ACUTE ENVIRONMENTAL
HARM
• PREVENT/MITIGATE HARM
TO REAL OR PERSONAL PROPERTY
• SITE INSPECTION -
SURVEY, DETERMINE NATURE
AND EXTENT OF RELEASE
• DETERMINATION OF RESPONSIBLE PARTY
• RANK SITE FOR PRIORITY LIST
• PREVENT CONTAMINATION
OF DRINKING WATER SUPPLIES
VM- REMEDIAL
IMPLEMENTATION
VIM- MONITORING ft
MAINTENANCE
-o-
-o
) CONSTRUCTION
• MAINTENANCE & OPERATION
OF REMEDY
• MONITORING FOR EFFECTIVENESS
OF REMEDY
• STATE/EPA AGREEMENT
• REMEDIAL INVESTIGATION
- COMPREHENSIVELY DETERMINE
NATURE AND EXTENT OF CONTAMINATION
• FEASIBILITY STUDY
- SELECT COST-EFFECTIVE ALTERNATIVE
- COMMUNITY RELATIONS PLAN
• CONCEPTUAL DESIGN OF REMEDY
• FINAL DESIGN OF REMEDY
Figure 1.
Phases of Response
-------
200 REMEDIAL RESPONSE
•determining or documenting immediate threats to the
public or the environment
•reviewing records
•determining sampling protocols for any subsequent re-
medial investigations
The search for responsibile parties will continue through:
•search of local records
•interviews with local persons
•review of waste disposal records and labels.
If a responsible party does not exist, cannot be identi-
fied, will not act in a timely fashion, or is likely to be
"judgment-proof" for all or a significant portion of the
costs of the release, the site is considered for further
fund-financed removal or remedial response.
At this point a decision is made to proceed with a
planned removal and/or remedial response depending
upon the need for action (as determined by the immedi-
acy of the threat and the site priority ranking). The con-
ditions under which these alternatives are selected and
characteristic activities of each are described in detail
below.
Phase V—Planned Removal
A planned removal action is taken in response to a re-
lease of hazardous substances which, while requiring
some near-term actions to stabilize a threat to public
health or the environment, does not require immediate
removal actions. Such a situation allows some technical
planning to more carefully select appropriate response
activities.
Restrictions are placed on planned removal activities in
order to conserve fund monies. All planned removals must
consist of operable units, that is, self-contained compon-
ents which will achieve some interim level of release miti-
gation or elimination without relying on future response
actions. Further, planned removals can be taken only at
priority sites or in situations where they are necessary to
prevent rapidly deteriorating conditions. Finally, to the ex-
tent practicable, planned removal actions should conform
to the cost-effective analysis required for full remedial
activities.
Phase VI—Remedial Response Planning
Remedial actions are most appropriate for long-term
and costly cleanups. The first step of remedial response
planning is the remedial investigation. The purpose of the
remedial investigation is to determine comprehensively
the extent and nature of the problem and to support the
development of alternative measures. Typical remedial
investigations include:
•geophysical and hydro-geological investigations
•soil sampling and analysis
•hazardous waste characterization
•groundwater, surface water and air monitoring.
The second step in remedial response planning is the
feasibility study. The feasibility study uses the results of
the remedial investigation to develop and evaluate alter-
native remedies, assess the environmental affects of the
release and alternative remedies, determine the cost-
effective remedial action, and prepare a conceptual design
of the approved action. A typical feasibility study in-
cludes:
•definition of response objectives and criteria
•laboratory studies
•development and evaluation of alternative remedial ac-
tions
•environmental assessment
•selection of an alternative
•conceptual design.
The selected remedial measure should meet public
health and environment requirements. If the resulting
remedies are more expensive than a predetermined cost
guideline, then other remedies, including non-cleanup
alternatives, shall be developed and evaluated.
The third activity in remedial response planning is the
remedial design. In this step the approved remedial ac-
tion will be clearly defined for implementation. A typical
remedial design may consist of the following elements:
•site response plan
•relocation plan
•engineering drawings and specifications
•contract documents.
During the remedial planning process and before
Phase VII, several technical, institutional and legal re-
quirements must be met by the States. The State must
show that it will be able to contribute its share of the re-
medial response planning and implementation costs, as
well as future operations and maintenance costs. The
State must also identify any required off-site disposal fa-
cilities required. Agreements between the State and EPA
are signed prior to initiating Phase VI, VII and VIII ac-
tions. These agreements set forth the State's and EPA's
responsibilities, define the scope of the project, and speci-
fy funding. When the State assumes the lead and funds
are transferred to the State, a cooperative agreement is
prepared. A less formal state agreement (or memo of
understanding) is signed when EPA maintains the con-
tractual lead.
Phase VII—Remedial Implementation
Several preparatory activities are necessary before con-
struction begins. Final permits must be obtained and
funding authorization is required from the State and
EPA. The State must also demonstrate its ability to as-
sume the responsibility for Phase VIII activities.
Remedial implementation—particularly for large pro-
jects—may be segmented. Each segment must be an op-
erable unit and would remedy or mitigate at least part
of the problem caused by the uncontrolled site. Project
segmentation allows the EPA Fund Manager to preserve
the size of the Fund and to balance response actions
across more sites.
Phase VIII—Monitoring and Maintenance
After the remedial action has been completed, the
States monitor the effectiveness of the response and pro-
vide any operation and maintenance necessary for the
continued effectiveness of the permanent remedy.
Elements of Phase VIII can include:
•operation of treatment and collection systems
•reports of monitoring and maintenance program
•documentation of the determination that the substantial
danger has been successfully reduced to design criteria.
-------
SURFACE SEALING TO MINIMIZE LEACHATE GENERATION
AT UNCONTROLLED HAZARDOUS WASTE SITES
DONALD E. BANNING
U.S. Environmental Protection Agency
Municipal Environmental Research Laboratory
Cincinnati, Ohio
INTRODUCTION
Many existing technologies, such as those currently be-
ing used for construction, hydrologic investigation, waste-
water treatment, spill cleanup and chemical sampling and
analysis, can be applied to uncontrolled hazardous waste
sites. The minimization of surface infiltration will, in al-
most all cases, be an integral part of the remedial steps at
those sites where the waste has been buried and the cost of
removal is prohibitive.
Minimizing surface infiltration typically consists of re-
grading, diverting surface water runoff and preventing or
eliminating infiltration. The effectiveness of surface seal-
ing depends upon the contribution of surface infiltration
to the total problem at the site. From a cost effective-
ness standpoint and ease of applicability, minimizing sur-
face infiltration poses marked advantages over other types
of remedial action unit operations.
The Solid and Hazardous Waste Research Division of
EPA has been involved either directly, through actual
EPA funding or indirectly through technically supported
efforts at several sites where minimizing surface infiltra-
tion has been implemented. Two of those sites are dis-
cussed in this paper.
WINDHAM, CONNECTICUT
After a site selection process that involved considera-
tion of over 400 sites, Windham, Connecticut, was selected
for remedial work in 1977. The Windham, Connecticut,
landfill is located in east central Connecticut, is approxi-
mately 25 acres in size and consists of two sections (Figure
1).
The landfill is located in an area of sand and gravel
immediately adjacent to the City of Willimantic water
supply reservoir. The eastern (old) half of the site is ap-
proximately 10 acres with its lower portion in the ground-
water. The western (new) landfill is approximately 15
acres and is above the groundwater. The site was oper-
ated from approximately 1945 to 1978.
A series of monitoring points emplaced through and
around the site defined the area of groundwater contam-
ination and its movement with time (Figure 2). The mon-
itoring system would also be useful in determining the
effectiveness of the remedial action after it was imple-
mented. This monitoring network included:
o
HARTFORD
OMANCHESTER
GLASTONBURY
CONNECTI CUT
TOWN OF WINDHAM I
Figure 1.
Location of the Windham Landfill, Windham, Connecticut,
from Approximately 1945 to 1978
Figure 2.
Location of Monitoring Points at the Windham Landfill,
Windham, Connecticut
(1) Suction and pan lysimeters in the refuse above the
water table and in the surrounding soils to de-
termine moisture infiltration
(2) The emplacement of piezometers into the water
table below and around the landfill to determine
the quantity and quality of leachate being gen-
erated and its movement through the subsurface
201
-------
202 REMEDIAL RESPONSE
(3) The establishment of surface water sampling points
(4) The placement of suction lysimeters to obtain leach-
ate in the refuse above the water table
(5) The implementation of a bimonthly sampling pro-
gram to determine changes in the geometry of the
pollutional plume with time
Between February and June 1979, wells, ponds and ly-
simeters were sampled on a monthly basis to determine
water quality. The water levels in the wells and ponds
were measured weekly and the volume of water collected
in the pan lysimeters was measured.
The leachate, of moderate strength, was generated by
the old fill whereas leachate of somewhat higher strength
was generated by the newer portion of the landfill. Leach-
ate degraded the groundwater down-gradient of the fill.
There was a significant contrast between the background
water quality measured upgradient of the site and that
of contaminated groundwater found down-gradient of
the landfill.
An analysis of the results of the weekly monitoring of
groundwater and surface water elevations and daily pre-
cipitation data verified significant infiltration was taking
place into the landfill. As much as 85% of the water pass-
ing through the refuse was derived from infiltration of pre-
cipitation.
The mass loadings to the Windham Reservoir from the
disposal site prior to any remedial action are shown in
Table I.
Table I.
Mass Loading of the Windham Reservoir
by the Windham Landfill*
(Before Surface Sealing)
Old Landfill
raj/1 Ib/day
New Landfill
mg/l Ib/day
Total Load
Ib/day Ib/year
TOC (total
org. carbon)
Total Ions"
Sodium
Chloride
Iron
Manganese
Hydraulic
Load
242 54
4,212 948
85 19
80 18
61.3 14
4.5 1
15,000 gal/day
ground water
663 127
8,736 1677
580 111
510 98
270 52
3.5 0.7
23 ,000 gal/day
ground water
181
2.625
130
116
66
1.7
661,000
958,000
47,600
42,300
24,000
615
*Based on chemical analyses for March 79 for MP 024 and MP (C26; background con-
centration for all constituents negligible (MP «8)
"Specific Conductance x 1.56 = Total Ions
Remedial actions proposed to abate and prevent pollu-
tion from the landfill included:
•Regrading of the landfill to maximize surface water run-
off and minimize infiltration
•The placement of a 20-mil PVC top seal
•Covering the top seal with approximately 18 inches of
final cover
•Revegetation (see Figure 3).
The remedial actions were designed to be passive to in-
sure minimum future maintenance. From the late summer
to the early winter of 1979, these remedial actions were
implemented at the site and final revegetation was accom-
plished in the spring of 1979.
Beginning in the fall of 1979, with the installation of
the seal, bimonthly sampling was instituted at selected
Table II.
Mass Loading of the Windham Reservoir
by the Windham Landfill*
(After Surface Sealing)
Constituent
TOC
Total Ions"
Sodium
Chloride
Iron
Manganese
Hydraulic
Load:
Old Landfill
rag/1 Ib/day
43.3
2808
32.2
73
14.1
4.33
280.61
3.22
7.30
1.41
0.20 0.02
12,000 gal/day
ground water
New Landfill
mg/l Ib/day
3.6
7800
290
312
51.2
0.11
0
0
0
0
0
0
Total Load
Ib/day Ib/ycar
4.33
280.61
3.22
7.30
1.41
0.02
1580.45
102422.65
1175.30
2664.5
514.65
7.3
(no ground water)
Total*
99.76
89.31
97.53
83.70
97.78
98.81
Averaje -
94%
*Based on latest chemical analyses for each monitoring Point (March 1981 for MP 124 ind
May 1981 forMP#26).
"Specific Conductance x 1.56 = Total Ions.
monitoring points. Results of the sampling showed that the
impact of the site on the groundwater has been mitigated
between November 1979 and March 1981. The current
mass loadings subsequent to the surface sealing, as well
as, the reductions achieved to date are presented in Table
II.
CHARLES CITY, IOWA
This site was a chemical waste disposal dump from 1952
to 1977 by a manufacturer of feed additives and veterin-
ary Pharmaceuticals for the poultry industry. It is located
in the floodplain of the Cedar River (Figure 4).
Since 1977, detailed studies undertaken there revealed
that the site contains over three million cubic feet of waste
material containing 28 priority pollutants. It lies over a
major water supply aquifer for municipal and rural areas
of northeastern Iowa. The following are calculated esti-
mates of some of the major compounds present:
Estimate of the Major Chemical Components Contained In
The LaBounty Dump Site, Charles City, la.
Chemical
Arsenic
1,1, 2-Trichloroethane
Nitrobenzene
Orthonitroaniline
Phenol
Amount Ib.
6,044,000
70,000
280,000
1,500,000
27,000
Early in 1979, the Solid and Hazardous Waste Re-
search Division (SHWRD) was requested by both the
Office of Solid Waste (OSW) and EPA, Region VII,
Kansas City, to implement some form of remedial action
research at the site. To date, ORD effort at this site has
consisted primarily of providing guidance on the place-
ment of effective monitoring system, assisting in the
development of final closure plans, which included a cap-
ping operation and conducting an on site waste sampling
study to ascertain the feasibility of in situ stabilization. If
the capping operation (clay) does not sufficiently reduce
the discharge of contaminates to the Cedar River, the over-
all closure plan calls for the construction of an upgrad-
ient cut-off wall to reduce or eliminate horizontal ground-
water flows into the refuse mass.
-------
REMEDIAL RESPONSE 203
^VEGETATED FINAL COVER
r-PVC MEMBRANE SEAL
REGRAOEO LANDFILL
REFUSE
Figures.
Typical Section Through the Windham Landfill
Between 10/2/79 and 10/9/80, an elaborate monitor-
ing system was used to quantitate the discharge of three
pimary/indicator contaminates from the site to the Cedar
River (Table III).
The surface infiltration minimization plan for the 13.3
acre area of the Charles City Site consisted of a two foot
thick clay cap, ground cover, erosion control at the site
toe, storm sewer rerouting, surface runoff controls and
post installation monitoring. The capping was completed
in November 1980. Monitoring data taken to determine
its effectiveness are presented in Table IV.
A comparison of the calculated LaBounty load con-
tribution to the Cedar River before and after the capping
operation is contained in Table V.
The wide range of concentrations for the contaminants
before and after sealing makes it difficult to conclu-
sively say whether or not the capping operation was
effective.
An alternative conceptual approach was evaluated by
the National Enforcement Investigation Center in Denver,
Colorado. In this analysis, the investigators made a basic
assumption that near-site and on-site rainfall are the pri-
mary driving forces for producing leachate. By this meth-
od, anomalies in correlated data might indicate effects of
other possible derived forces such as upward leakage from
the Cedar Valley aquifer or bank recharge from the Cedar
River.
The analysis suggested that shortly after a rain, a leach-
ate pulse forms and begins moving toward the Cedar
River. The pulse is characterized by a higher flow volume
• SEE BELOW
Figure 4.
Site Location, LaBounty Disposal Site
-------
204 REMEDIAL RESPONSE
Table HI.
LaBounty Disposal Site
Cedar River Monitoring
(Before Surface Sealing)
STATION 11
STATION 12
CALCULATED
LABOUNTY CONTRIBUTION
Day-Yr
10/2/79
1/14/79
2/13/79
4/16/80
5/08/80
6/13/80
7/11/80
8/7/80
9/10/80
10/9/80
Note:
Arsenic
Cone. Load
2.40
<2.00
4.30
6.4
7.4
4.2
8.2
4.1
4.0
^2
Cone.
Load
*Measured flow
4.7
<9.5
7.2
24.7
13.5
15.0
24.1
6.2
18.4
6.47
PPb
Ib/day
ONA
Cone.
0.29
12.
0.32
0.049
0.14
0.158
0.07
0.07
0.07
0.14
Load
0.6
56.8
0.5
0.19
0.25
0.56
0.21
0.11
0.32
0.45
at Station 11 used in
1,1,2-TCE Arsenic
Cone. Load Cone. Load
6.1 12.0 20.9 40.8
5.4 25.6 9.2 43.6
9.3 15.7 43.4 73.1
<5 19.3 15.5 59.9
^5 9.1 33.5 60.9
<:5 ^17.8 33.8 21.
-------
REMEDIAL RESPONSE 205
Table V.
Loadings to Cedar River from LaBounty
Before and After Capping
Before
LaBounty Contribution
After
LaBounty Contribution
Arsenic
Load
36.1
34.1-43.6
65.9
35.2
47.4
106
49.3
38.4
102
35.6
ONA
Load
9.7
80.2
11.5
13.9
7.05
25.9
19.6
3.55
11.2
10.6
1,1,2-TCE
Load
11.4
12.3
23.6
3.9-23.2
5.4-14.5
21.5-39
11.7-26.4
4.6-12.2
9.3-32.3
0-16.1
Arsenic
Load
28.2
33.2-36.1
48.0
64.2
82.5
48.0
ONA
Load
5.42
3.11
3.53
4.11
6.65
3.92
1,1,2-TCE
Load
9.0
7.0
5.0-13
3.0
0.0-2.9
0.0-18
Q= 4x13.3x0.052x1,440 = 3,984 gal/day
The net effect of the site capping should be to reduce
infiltration through the site by 80%. As a result mass
loadings/migration due to infiltration will be reduced by
CONCLUSIONS
A synthetic capping remedial action at Windham,
Connecticut under rather ideal hydrological conditions
lead to an overall 94.0% reduction of the mass loadings
between March 1979 and March 1981.
In a complex hydrologic situation 6 months of post re-
medial action monitoring is inadequate to determine effec-
tiveness. Conceptual approaches to predicting the effec-
tiveness of capping operation even in complex hydrologic
settings, are available, but the verification of the extent of
their accuracy remains to be proven.
ACKNOWLEDGEMENTS
Steven W. Sisk, Hydrologist, EPA, Office of Enforce-
ment, National Enforcement Investigation Center, Den-
ver, Colorado, for his contribution "Hypothetical Assess-
ment of Rainfall Effects on Leachate Production at the
LaBounty Site, Charles City, Iowa."
-------
APPLYING TECHNIQUES FOR SOLIDIFICATION AND
TRANSPORTATION OF RADIOACTIVE WASTES
TO HAZARDOUS WASTES
J. W. PHILLIPS
Hittman Nuclear and Development Corporation
Columbia, Maryland
INTRODUCTION
The equipment discussed in this paper was originally
developed in response to the needs of the nuclear power
industry for processing and transporting radioactive
wastes. Prior to 1979, liquid wastes, primarily in the form
of boric acid and miscellaneous chemical drains, could be
transported for burial in their liquid state. In mid-1979,
the licenses for the three burial sites were revised to pro-
hibit acceptance of liquid wastes. Utilities whose plants
did not contain systems capable of solidifying liquid
wastes were faced with a crucial decision. They could cut
back on the volume of liquid wastes being generated and
treat them with deep bed demineralizers or they could hire
an outside firm to come onto the site and, with portable
equipment, solidify the liquids prior to shipment. For
many utilities, the first option was not feasible due to the
large quantities of ion exchange resins that would be re-
quired to treat the liquid wastes involved.
Both mobile processing equipment, and equipment that
is permanently installed on-site are discussed in this paper.
The transportation equipment was designed and built to
conform to specific requirements pertaining to radioactive
wastes. The purpose of the following discussion is to
identify how these types of equipment could be modified
and used to accommodate hazardous chemical wastes.
SOLIDIFICATION OF LOW-LEVEL
RADIOACTIVE WASTES
As stated previously, the solidification of low-level
radioactive wastes can either be handled with mobile
equipment, leased and operated by a company that
specializes in this work or the equipment can be pur-
chased and installed for operation by a utility. Numerous
current suppliers exist for both mobile solidification
services and supply of systems for permanent in-plant
installation.
The types of wastes being solidified are spent ion ex-
change resins, concentrated solutions of boric acid and
sodium sulfate and calcium fluoride sludges. Various
lubricating and turbine oils have also been successfully
solidified. Most of these wastes have, for the past sev-
eral years, been solidified with Portland Cement using
various additives to condition the wastes based on their
chemical characteristics. A urea-formaldehyde (UF) based
material has also been extensively used over the past 5
to 7 years.
The use of UF has come under increasing attack over
the last two years because it is a condensing polymer
and produces liquid as it polymerizes. When the resin
hardens, it shrinks, forcing liquid out of the matrix.
This liquid is very low in pH as a result of the acidic
catalyst used to initiate the polymerization process.
A third product, which is only beginning to be used, is
a material produced by Dow Chemical Company re-
ferred to as the "Dow media." The Dow process is a
three part process which uses a binder, a promoter and
a catalyst. For certain wastes, an extender is also re-
quired. The proportions of materials used, and the chem-
ical formuli of the ingredients, are proprietary to Dow
Chemical Company.
Bitumen, or asphalt, has been used for the solidifica-
tion of radioactive wastes in Europe for several years.
While a few systems using bitumen have been sold in
this country for installation in nuclear power plants, none
are in operation at this time.
Due to numerous recent changes in the burial site li-
cense in Barn well, South Carolina, there has been a
significant increase in the number of mobile systems in
use over the last two years. Power plant operators have
turned to mobile solidification services in order to satisfy
evolving criteria. This increased use makes it very diffi-
cult to determine the exact number of power plants using
mobile systems at any one time. However, it is estimated
that 60 to 70 percent of the operating units were contract-
ing for some form of mobile solidification service as of
July of this year.
Typical System
Since almost all solidification of low-level radioactive
wastes is performed using cement, a typical solidification
system currently in use will be described. The compon-
ents of the system include a disposable container with in-
ternal mixing blades, a mixing head, a mixing motor, a
cement storage hopper, a cement feed system, a dust col-
lector, a pump skid and a control panel.
When the radiation level of the waste materials is high
enough to require shielding for the operator, the solidi-
fication process is performed with the disposable liner
either in a process shield or in a shielded shipping cask
(which can be used for subsequent transport of the waste
206
-------
REMEDIAL RESPONSE 207
to the burial site). After the liner has been placed in its
desired location, the mixing head with mixing motor is
bolted to the liner. A thick rubber gasket between the mix-
ing head and the liner neck provides a seal between the
inside of the liner and the outside atmosphere. All other
connections to the liner are made through the mixing
head. A prefilled cement hopper is positioned on a scaf-
fold and connected to the cement inlet valve on the mixing
head via a flexible screw feeder. The cement feed motor is
located on the mixing head. An induced draft dust col-
lector is connected to the vent connection on the mixing
head to catch any cement dust. A central control panel
contains the necessary stop/start switches, indicators and
control logic to operate the system.
Mobile systems are leased to a utility for periods rang-
ing from the duration of a single solidification opera-
tion to several years of service. Operator services are sup-
plied with the equipment. When the duration of the con-
tract is long enough and the frequency of use is high
enough, operating personnel are permanently located at
the site. Upon termination of the contract, the equipment
is removed from the site.
The procedure is quite different for permanently in-
stalled systems which are owned, operated and main-
tained by the utility. They are located within a building
designed especially to afford the operator the maximum
possible radiation shielding. These shield walls, however,
tend to make the equipment less accessible. In-plant sys-
tems are substantially more automated than the mobile
systems. A simplified process flow diagram of the in-
plant solidification system being installed at the three-
unit Palo Verde plant west of Phoenix, Arizona, for
Arizona Public Service is shown in Figure 1.
Aside from the degree of automation involved, the
primary difference between a mobile system and an in-
plant system is now and where the mixing of wastes and
solidification medium takes place. In an in-plant system,
the mixing takes place outside the disposal container, be
it a 55-gallon drum or a large liner. In the system shown
in Figure 1, wastes to be solidified are transferred to the
waste batch tank, where samples are taken and appropri-
ate chemical adjustments are made prior to solidification.
The batch tank is equipped with level indicators and an
electric mixer. The waste transfer pump transfers the
WASTE
INLET
WASTE
BATCH
TANK
CHEMICAL
ADDITION
TANK
BULK
METSO
STORAGE
SILO
BULK
CEMENT
STORAGE
SILO
WASTE
FEED
PUMP
CHEMICAL
ADDITION
PUMP
SCREW CONVEYOR
MIXER
FLUSH
MODULE
DISPOSABLE
CONTAINER
[ TRANSFER CART
o
Figure 1.
Simplified Palo Verde PFD
-------
208 REMEDIAL RESPONSE
wastes at a predetermined flow rate from the batch tank
to the waste mixer.
The cement and Metso* are stored in silos, designed to
hold large quantities of these materials to take advantage
of bulk deliveries. Cement and Metso are discharged from
the silos at feed rates controlled by horizontal screw con-
veyor. Direct current drive motors are used which allow
the feed rates to be adjusted depending on the desired
mixing ratios of waste, cement and Metso for each par-
ticular waste form. The cement/Metso mixture enters the
unit from the top at the back end of the mixing screws.
As the cement is moved forward, the pumped wastes en-
ter the mixing barrel and are mixed with the cement. The
discharge valve is opened and the cement paste drops into
the disposable container.
At the end of each cycle, the waste feed system and the
mixer are completely flushed. Extra dry cement added to
the liner, or to the last drum, adsorbs the flush water. The
primary flush module contains an air accumulator bottle
which provides the motive force for flushing the mixer in
the event of a loss of electrical power. The mixer also
includes a hand crank on the motor so that it can be
cleaned out manually if necessary. After the container is
filled it is moved to a capping station where a lid is auto-
matically affixed. The container is checked for external
contamination, decontaminated if necessary, and moved
into a shielded storage area to await transportation to the
burial site.
TRANSPORTATION
Obviously, the primary design consideration of radio-
active waste shipping casks are their radiation shielding
capability. This concern is of little consequence in trans-
porting hazardous chemical wastes. However, the struc-
tural integrity required in a design that must support
thick steel and lead walls, and that conforms to the NRC
regulations, results in a package whose ability to with-
stand foreseeable transportation accidents is not easily
matched.
Several types of shipping containers exist including
standard truck vans, shielded vans, box casks for hand-
ling small numbers of drums, and large cylindrical casks
for shipping liners or drums. Figure 2 shows a box cask
which holds 12 drums. A roller assembly pulls out from
the end of the box for ease in loading and unloading
drums. A cylindrical cask which will hold either a 170-ft3
steel liner (6 ft high by 6 ft in diameter) or 14 SS-gallon
drums is shown in Figure 3. The SS-gallon drums used in
the nuclear industry are designed and tested in accordance
with DOT Specification 17H and are certified to hold 840
pounds of waste.
The tie down connection of both the box casks and
cylindrical casks are designed to withstand the following
acceleration loadings:
•Front to back 10 g "Vertical 2g
•Side to side 5 g
•Registered trade name for sodium metasilicate anhydrous, manu-
factured by Diamond Shamrock, Soda Products Division. The full
product name is Metso Beads 2048.
Thus, the tie down connections on a typical shipping cask
weighing 50,000 pounds are designed to withstand simul-
taneous loads of 500,000 pounds forward, 250,000 pounds
sideways and 100,000 pounds upwards. Although the
trailers used to transport the cask cannot withstand these
forces, the strength of the containers insures that the con-
tents will remain undamaged and the cask integrity will
not be breached.
Figure 2.
Reusable Cask Which Holds 12 Drums
Figure 3.
Reusable Cask for a 170 ft' Steel Liner or 14-55 Gallon Drurro
-------
REMEDIAL RESPONSE 209
Drop tests must also be performed on the casks. The
cask must be dropped onto an unyielding surface with no
resultant damage to its contents or to the cask integrity
when dropped in every orientation. Two drop tests exist:
a one-foot drop test and a thirty-foot drop test. Most
casks are designed to withstand the one-foot test, al-
though in all probability these casks would not yield in a
drop test of as much as ten feet. Only specialized casks,
designed to handle "large quantities" of radioactive ma-
terial, will withstand the thirty-foot drop test.
CHEMICAL VERSUS RADIOACTIVE WASTES
Before the use of the cement solidification technique
for toxic and hazardous wastes is examined, the advan-
tages and disadvantages of handling chemical wastes as
compared to radioactive wastes should be noted. This is
done in Table I.
APPLICATION OF EQUIPMENT TO SPECIFIC USES
In considering using the equipment described above
for radioactive wastes for the immobilization of hazardous
wastes, one must first define the conditions under which
hazardous wastes are found. It is expected that the con-
tents of many of the drums will be unknown, lagoons
containing undefined substances will be encountered and
the quantities of contaminated soils and sludges will be
considerable.
Drums containing solids are not candidates for solidifi-
cation. The emphasis in this paper is on the solidification
of liquids and pumpable sludges. Containers that are in-
tact must be opened and the contents inspected. Breached
drums must also be inspected to determine the status of
their contents. Once the containers with liquids are
identified, their contents must be transferred to a batch
tank or a disposable liner depending on the type of system
to be used. This transfer from the dumping grounds to
the solidification area will probably be the most critical
and most dangerous step in the process.
Table I.
Advantages and Disadvantages of Handling of
Chemical Wastes Compared to Radioactive Wastes
Advantages
1. The material is not radioactive,
therefore proximity of the worker to
the waste, or thicknesses of walls and
material density, are not a problem.
2. Contact of the waste with protec-
tive clothing can be permitted.
3. Glass and/or plastic viewing de-
vices can be built into the system
wherever deemed appropriate.
Disadvantages
1. Small exposures to some wastes
through inhalation or physical contact
could be fatal even in very small doses.
2. Contents of most containers will be
unknown and identification nearly im-
possible.
3. Some of the wastes may be explosive.
4. Procedures for the transfer of the
liquids to the batch tank or a disposal
liner must be carefully planned.
5. Synergistic effects may result when
different wastes are mixed.
6. Chemical analysis of the contents of
each container will have to be made.
7. Process control plans will have to be
developed for each site.
ADAPTATION OF MOBILE
SOLIDIFICATION SYSTEMS
First Example
In order to assess how the equipment could be used, two
scenarios were developed and studied. The first situation
is the best case for a mobile system. It is a small site con-
taining drummed wastes and requires that no liquids
leave the site. It is assumed that the number of drums
present is less than a thousand, with most still intact.
Before the actual solidification process can start, it is
recommended that a prefabricated enclosure be erected
over a prepared surface. The prepared surface prevents
any spilled liquid from further contaminating the sur-
rounding soil. The prefabricated enclosure will protect
the equipment and personnel from the elements and es-
tablishes a controlled environment around the process to
minimize airborne hazards.
Drums are brought into the enclosure where they are
opened, inspected and samples taken for analysis. The
chemical analysis can be best accomplished in a centralized
laboratory equipped to handle toxic and hazardous ma-
terials. It should be assumed that a particular drum could
be in the enclosure for as long as two weeks while the
chemical analyses are being performed. When the con-
tents of several drums have been selected as a combined
batch for solidification, additional samples must be taken.
Samples of equal size from each drum are combined and
the proper pretreatment, as determined by the chemical
analysis, is made. Typical pretreatments include pH ad-
justment, addition of emulsifiers or addition of solvents.
The goal of the pretreatment process is to produce a
mixture which will dissolve or remain suspended in water.
This is essential since the setting of the cement is de-
pendent on the presence of water. Upon completion of
the pretreatment process a small quantity, approximately
100 to 400 ml of the mixed wastes, are solidified as a test
sample. Several samples may actually be made using dif-
ferent ratios of waste and cement.
In the final analysis, two samples are selected to repre-
sent a range of waste to cement ratios that will result in
an acceptable product. These ratios are then used to de-
termine the optimum quantity of wastes to be solidified in
a single operation and to estimate the quantity of cement
necessary to successful solidification. This procedure, or
process control program, is repeated for each batch of
wastes that differs significantly from previous batches.
When the successful solidification tests are completed,
the actual waste processing can be initiated. Since the
chemical analysis previously performed has identified the
wastes, the appropriate precautions can be taken regard-
ing proper clothing and the necessity for breathing ap-
paratus. The operating personnel, properly dressed, would
then transfer the contents of the drums to the mixing liner
and the designated pretreatment would be accomplished.
Liquids would be transferred by pump. Sludges and
pasteous solids might have to be hand transferred. For
large quantities df nonpumpable materials, the drums
might better be emptied into a mixing tank with the top
opened or half-opened. The material could then be pre-
-------
210 REMEDIAL RESPONSE
treated and mixed until a slurry is developed that can be
pumped into the disposable liner for solidification.
Once the waste is in the liner, the operation is identical
to the operation for radioactive wastes. Cement and any
designated dry additives are put into the liner through a
screw feeder at approximately 100 Ib/min. During this
phase of the operation, a dust collector continuously pulls
air from the liner and exhausts it through a baghouse
filter. Additional off gas controls may have to be in-
cluded depending on the type of materials handled, and
gas generation expected due to the pretreatment opera-
tion or to the reaction of the wastes to the solidification
media.
There are obviously certain materials that do not lend
themselves to this on-site solidification. Materials too
hazardous to risk direct handling could be handled re-
motely. However, the type of equipment necessary for
remote handling is not conducive to the limited scope of
this scenario. These materials would have to be packaged
and sent to a centralized facility.
With the solidification process complete, the mixing
motor is removed and all other connections are broken.
The container is capped with either a standard 55-gallon
drum lid or with a snap tight, nonremovable lid. At this
time the liner is moved to a suitable storage location to
await transportation.
Second Example
The second scenario also describes the use of a mobile
system and is, again, a small site, but one in which broken
and leaking drums litter the site causing contaminated
soil and sludges. Damaged and/or leaking drums could be
handled in a centralized processing facility if they are
considered too dangerous for on-site processing. These
drums could be repackaged in an overpack such as an
83-gallon drum for transportation. The basic setup de-
scribed in the first scenario for handling intact drums
could still be used, although precautions would be neces-
sary due to the high probability of exterior contamina-
tion. Decontamination facilities could be provided with
subsequent solidification of the decontamination solu-
tions.
Small quantities of sludge and contaminated soil are
best handled onsite in loading drums or liners (depending
on the viscosity of the material), which are then moved to
the solidification facility. In-drum mixing systems can be
used for soils to keep the motor sizes reasonable.
ADAPTATION OF IN-PLANT
SOLIDIFICATION SYSTEMS
The significant question regarding toxic and hazardous
wastes is how to handle large uncontrolled sites where
thousands of drums and millions of gallons of chemical
wastes have been indiscriminately dumped. Cleanup of
these sites will take years. No simple solutions are avail-
able. The costs are going to be high and the progress slow.
In most instances complex processing facilities built at
the existing dump sites will be the technical, if not the
political, solution. The ultimate disposal of the newly pro-
cessed wastes will depend on the specifics of each site.
Once the decision is made to process existing wastes by
immobilization in an inert binder such as cement, a de-
termination of the exact nature of the wastes to be pro-
cessed and the support facilities required must be made.
The forms which these wastes may take include:
•Liquids from lagoons, settling ponds and drums
•Sludges from lagoons, settling ponds and leaking drums
or other containers
•Contaminated soils caused by leaking containers or di-
rect dumping of liquids and sludges
•Pasteous solids from breached and/or intact containers
•Solids in drums or in other containers or from open con-
taminated sites
Facility Design
Using these five waste forms as a basis for facility de-
sign and considering the possible hazardous nature of the
materials, a conceptual facility flow pattern can be made.
This flow pattern, shown in Figure 4, includes several
features commonly used in the nuclear power industry for
waste handling and personnel access control. Depending
on the specifics of the site and the wastes, many of these
operations can be performed remotely. These remote op-
erations are performed using TV monitors mounted
throughout the facility and overhead bridge cranes to
move containers about.
As shown in Figure 4, materials from the site are
brought into the facility at the receiving point, probably by
truck. The containers are unloaded and moved into an
inspection room where the container is opened and the
contents inspected. If the contents are stable, a decision
can be made to send the container to the storage area for
shipment or to the repackaging area for placement in an
overpack. In either case, a sample may be taken for
analysis to aid in determining the proper method of ulti-
mate disposal.
Containers, whose contents upon inspection are iden-
tified for solidification, are first sampled. While the
samples are being analyzed, the containers are sent to a
contaminated storage area after appropriate labeling.
Containers with similar contents are identified and sample
solidifications are performed to develop specific
process parameters and determine necessary additives.
These containers are then retrieved from storage and
moved into the area with the batch tank. The contents of
each container are transferred to the batch tank by pump
or by picking up the container using special handling
equipment and pouring. The batch tank is sealed and the
contents mixed with appropriate solvents, emulsiilers and
chemicals for pH adjustment. At the end of the mixing
cycle a sample can be drawn for verification purposes.
Remote Operations
This section discusses a solidification operation that is
controlled remotely and automatically from a central con-
trol panel or control room. Dry cement feed is fed along
with the waste feed but enters the mixing device prior to
the wastes. Feed rates are controlled and can be adjusted
if necessary. The cement paste free falls into a disposal
-------
REMEDIAL RESPONSE 211
r
UNCONTAMINATED
MATERIAL FLOW
BATCH TANK
& PRETREATMENT
T T L
— Personnel Access
— Old Containers from Site
— New Containers
— Materials
CONTROL ROOM
Figure 4.
Hazardous Waste Solidification Facility Flow Diagram
container until a preset level is reached. The interface be-
tween the mixer and the container is sealed with appropri-
ate vent connections. Ultrasonic level probes are used to
prevent contamination of the instruments.
When the container is full, waste feed is terminated by
either diverting it back to the batch tank or by completely
flushing the lines. The diversion mode is selected when
additional containers are to be processed within a reason-
able time frame. If the batch tank is empty, or insuffici-
ent material is available to process another container, the
feed line is flushed. Additional dry cement is fed through
the mixer and into the container to hydrate the flush
water. Once the container is full it moves out of the fill
position into the capping station where it is capped and
sealed.
The solidification process itself has several fail-safe
design features built into it. First, the motor for the mixer
is located outside the solidification area and contains a
hand crank. This back-up system permits emptying of the
mixer should the process be stopped in mid-stream due
to motor failure or loss of electrical power. Maintenance
on the motor can also be performed without entering a
contaminated area.
Second, the system flush is controlled through a flush
module mounted outside the solidification area, again for
maintenance purposes. The flush water is kept under
pneumatic pressure at all times so it is available even dur-
ing a loss of electrical power. This module is used for
normal operations to ensure that the proper quantity of
flush water is used and to ensure that the unit is operable
if needed in an emergency.
Capped containers are inspected and tested for ex-
ternal contamination and decontaminated if necessary.
With the container certified as free of contamination it
should be labeled as to its contents and status and stored
until final transportation for disposal.
REFERENCES
1. Phillips, J., Feizollahi, F., Martineit, R., Bell, W. and
Stouky, R. "A Waste Inventory Report for Reactor
and Fuel-Fabrication Facility Wastes," ONWI-20.
Prepared for U.S. Department of Energy, Office of
Waste Management and Battelle Memorial Institute,
March 1979.
2. Guilbeault, B.D., "The 1979 State-by-State Assessment
of Low-Level Radioactive Wastes Shipped to Burial
Grounds," NUS-3440, Rev. 1. Prepared for EG&G
Idaho, Inc., November 1980.
-------
RENOVATION OF A WOOD TREATING FACILITY
W. LAWRENCE RAMSEY
O'Brien and Gere Engineers, Inc.
Washington, D.C.
RICHARD R. STEIMLE
Maryland State Office of Environmental Programs
Baltimore, Maryland
JAMES T. CHACONAS
Maryland Environmental Services
Annapolis, Maryland
INTRODUCTION
In August 1981, work began on the renovation of an
out-of-service wood processing facility located on a 130
acre site in St. Mary's County, Maryland. The project is
a cooperative effort between regulatory and service agen-
cies and the private sector to clean up a significant haz-
ardous waste problem in a fairly commonplace industrial
operation.
BACKGROUND
In the summer of 1975, inspectors of the Maryland
WRA determined that the waste practices at a wood treat-
ing facility in St. Mary's County, Maryland, might be
impacting ground water. Subsequent investigation indi-
cated that wastewater from a creosote and pentachloro-
phenol (penta) wood preserving operation was being dis-
charged to unlined "evaporation lagoons." While the
practice was obviously unacceptable, the regulatory pro-
grams at that time were not designed to handle ground-
water discharges.
After much discussion, it was decided that the best re-
vision to this facility would be to construct a recycle/
treatment facility utilizing spray irrigation for excess
wastewaters. Using this system, most of the penta and
creosote would be recycled. Wastewater containing tan-
nins and lignins in addition to waste preservatives would
be lagooned and disposed of by spray irrigation. The esti-
mated discharge would be about 1,000 gal/day.
In addition to the ongoing problem, the company would
be required to renovate the existing lagoons and clean up
the area. In August of 1977, a State Discharge Permit
was issued requiring the development of a plan for site
renovation and requiring the construction of a wastewater
treatment facility. During the summer of 1977, the State
obtained some composted sewage sludge to incorporate
into the contaminated soils to determine if biodegrada-
tion by land farming was possible.
When the test proved successful, the original plan was
modified to include on-site renovation of soils contam-
inated by the waste ponds. The idea was to construct
three spray irrigation fields. One would be in use, one as
a back-up and the final one to reclaim the contaminated
soils. Each year a portion of the contaminated soils
would be mixed with compost, tilled and seeded. The fol-
lowing year, that site would be rotated into spray irriga-
tion use. Any remaining liquid waste in the ponds would
be treated in the first year with the normal plant waste.
SITE INVESTIGATION
In 1975, samplings of waters in a spring fed "fresh
water pond" near the wood treating facility showed
phenolic concentrations of 1.80 mg/1. Analyses of oily
material floating on the pond was determined to contain
77.5 mg/1 of phenolics. There was no evidence of a sur-
face connection between the plant's waste ponds and the
fresh water pond. It was observed at the site that the oily
material was "bubbling" up from the spring end of the
pond.
These observations made it necessary to consider the
imposition of site renovation conditions on the property
owner. To support the conditions, a field investigation
was begun.
On May 6, 1975, three monitoring wells were installed
in augered soil borings. The boring logs showed that a
coarse to fine silty sand formation existing from the sur-
face to a depth of 15 to 20 ft. Below this is a sandy, blue/
black clay which extended to a depth of at least SO ft.
Field investigators noticed strong phenolic odors in soil
samples 9 to 13 ft below the surface. The monitoring
wells were constructed with 1.25 in. PVC schedule WO
pipe with 5 ft slotted screens. The screens were located
approximately 15 to 20 ft below the surface. The water
table was between 1.5 to 3.5 ft from the surface.
Water samples obtained from these three wells on
June 11, 1975, showed concentrations of phenols of 1.0f
9.0 and 14.4 mg/1. The pH of the groundwater ranged
from 5.5 to 5.7 The groundwater flow determined from
water elevations in the monitoring wells was from the
wastewater holding ponds to the fresh water pond. The
conclusions of this interdepartmental study was that waste
phenolics in the ponds had percolated into the ground*
water and then migrated to the fresh water pond.
The stream which drains the fresh water pond WM
studied by a private consultant. In that study, a water
sample taken approximately 900 ft below the pond con-
tained 0.12 mg/1 phenol. The consultant reported that
few aquatic insects were found in this section of the
stream. However, phenol concentrations decrease and the
biomass increases substantially within 0.25 mile from the
212
-------
REMEDIAL RESPONSE 213
first sample point. The data indicate a significant improve-
ment of water quality as the distance from the fresh water
pond increases.
To support the pending legal actions, a more detailed
hydrogeologic investigation was conducted. On December
7, 1978, WRA investigators sampled ground and surface
waters. A sample of the groundwater from one of the
wells installed in 1975 showed creosote concentrations of
1.6 mg/1 and pentachlorophenol concentrations of 1.34
jig/1. Analyses of liquid samples in the holding ponds
showed PCP concentrations ranging from 0.62 to 66.06
/ig/1 and creosote concentrations ranging from 0.654 to
70.72 mg/1. Analyses of the stream samples showed PCP
ranging from 0.01 to 27.0 Mg/1 and creosote concentra-
tions ranging from 0.001 to 1.769 mg/1. Pentachloro-
phenol and creosote concentrations in the fresh water
ponds were 0.01 Mg/1 and 0.308 mg/1 respectively.0'
Samples taken at various monitoring wells on the site
indicated that the phenolic concentrations increased with
depth until the water table was intercepted. At the water
table, analyses showed a band of high phenolic concentra-
tion that indicated that the wood preservatives were
"floating" on the groundwater.
Sampling of shallow dug residential wells in the area
failed to substantiate the County Health Department's
sampling that showed high phenolics. Samples also showed
phenolics in monitoring wells up the hydrologic gradient
from the facility. It was proposed that an air vector was
responsible for carrying vapors from the venting of the
pressure vessels to these sites. This theory will never be
tested as the site is shut down.
TECHNICAL BASIS FOR
THE RENOVATION PLAN
Renovation of the wood preserving plant incorporates
basic biological wastewater treatment processes currently
used in the wood preserving industry.® For this project,
freshly composted sludge from the Washington, D.C.
Wastewater Treatment Plant was mixed with the creosote
and penta contaminated soils to accelerate biological de-
gradation. The composted sludge provides a biologically
active soil addition containing up to 25 % dead and living
microorganisms. Bacteria, actinomycetes and fungi com-
prise most of the active microbial population.(3)
Since the contaminants are by nature, bactericides, the
microorganisms introduced to the contaminated soils will
undergo some acclimation. Following acclimation, how-
ever, creosote reductions exceeding 90% have been ob-
served in activated sludge and soil irrigation wastewater
treatment processes. Additionally, penta reductions ex-
ceeding 99%, during a 50-day period, have been reported
in a bench-scale study utilizing activated sludge treat-
ment/4' Increased soil organic matter has also been shown
to expedite penta degradation within the soil medium.(5)
The composted sludge to be used for site renovation
will be composted for 21 days and in lieu of the usual cur-
ingprocess will be applied to the site. Incorporation of the
composted sludge into the waste contaminated soils will
not only restore the microbial population but will also
create optimum soil conditions for biological activity.
These soil conditions include pH, structure, moisture
and temperature. The compost contains approximately
15% limestone and has a pH near neutral. At the planned
application rate 125-130 tons/acre, the composted sludge
should adjust the previously acid soils to a pH more con-
ducive to microbial activity.
The subsoils on the site range from sandy to clayey.
Incorporation of the composted sludge into sandy or
droughtly soils increases the moisture holding capacity of
the soil. In a clayey or compacted soil, the incorporated
compost enhances soil structure allowing better water and
gas exchange within the soil and thus promoting a more
aerobic environment.
Soil temperature will be affected not only by microbial
activity, but also by its ability to hold solar radiation. The
existing soil color will be darkened by the composted
sludge additions. The darker soil color should maintain
warmer soil temperatures later in the fall and early in the
spring and consequently enhance biological activity during
colder months of the year. In addition to microbial de-
gradation of the creosote and penta contaminants, sun-
light should contribute to the destruction of other phen-
olics and wood chemicals particularly in existing lagoon
and "fresh water" pond areas.
Contaminants which have leached past the zone of mi-
crobial activity in the soil, are expected to eventually dis-
appear via dilution. Following seeding, the composted
amended soils should be more than adequate to promote
revegetation of the site.(6)
METHOD OF SITE RESTORATION
Site restoration for this facility is to be executed ac-
cording to plans prepared by Lyon Associates, Inc. of
Baltimore, Md. Following general site inspection, all
supply and test wells are to be abandoned and sealed. The
lagoons are to be drained and spray irrigated in a desig-
nated disposal area. Following this procedure all con-
tainers will be cleaned and stored on site for removal. At
the same time the lagoons will be breached to prevent
further ponding of water.
Following dewatering, all contaminated structures in
the pond area such as cement foundations, steel pipes,
etc., will be removed and hauled to an approved hazard-
ous waste facility. Proper erosion and sediment controls
will be utilized during the land clearing.
The "fresh water" pond will be dewatered by breach-
ing the embankment. Sediment control will be maintained
downstream from the pond by using soil erosion control
fencing. Since the pond is spring fed, spray irrigation may
be utilized at the discretion of the contractor.
The earthwork will involve uniform regrading of the
lagoon area maintaining slopes less than 1 to 2. The fresh
water pond embankment will also be removed.
Following regrading, composted sludge will be applied
to the entire site at the rate of 125-130 tons/acre (approxi-
mately 2 in.) and tilled into the top 4 to 6 in. of soil. The
site will then be hydroseeded and the grass mowed as re-
quired during the growing season.
-------
214 REMEDIAL RESPONSE
CONCLUSIONS
When this project was begun five years ago, the State
had in mind several objectives. The first was to demon-
strate the feasibility of reclaiming a hazardous waste
site. At the time this renovation began, RCRA and other
laws governing safe disposal of hazardous wastes were
just being discussed; the techniques applied here were
innovative.
The second objective of the cleanup was the develop-
ment of low cost, unsophisticated solutions to waste prob-
lems to encourage cooperation and participation by in-
dustry in cleaning up their own problems. Too often,
over-zealous enforcement combined with unclear and
overly restrictive technical support have caused activities
to occur in the court room with technically ill-prepared
judges making decisions instead of renovation of a site by
experts.
Finally, from a management standpoint, the State de-
sired to treat or destroy hazardous wastes at, or as close
to, the point of generation as possible. This technique
eliminates possible transportation accidents and the
prospect of creating two hazardous waste problems in-
stead of one.
REFERENCES
1. Steimle, et al., WRA Field Investigation Report: "A
Subsurface Investigation of Creosote and Penta-
chlorophenol Concentrations in the Soils, Hollywood
Md., 1978."
2. Ramsey, W.L., In-House Memorandum. Water Re-
sources Administration, Maryland Department of Na-
tural Resources, 1979.
3. Poincelot, R.P., "The Biochemistry of Composting,"
in Proc. 1977, National Conference on Composting of
Municipal Residues and Sludges.
4. Dust, J.V. and Thompson, W.S., 1973, "Pollution
Control in the Wood Preserving Industry," Part 4,
Biological Methods of Treating Wastewater, Forest
Products J. 23(9):59-66.
5. Kuwatsuka, S., "Degradation of Several Herbicides in
Soils Under Different Conditions," Environmental
Toxicology of Pesticides, F. Matsumura, G.M. Bouch
and T. Masato, Eds. Academic Press, New York, N.Y.,
1972, 385-400.
6. Hornick, S.B., et al., "Use of Sewage Sludge Compost
For Soil Improvement and Plant Growth," U.S. De-
partment of Agriculture Science and Education Ad-
ministration, Agricultural Reviews and Manuals.
ARM-NE-6, August, 1979.
-------
THE FORT MILLER SITE:
REMEDIAL PROGRAM FOR SECUREMENT OF AN
INACTIVE DISPOSAL SITE CONTAINING PCB'S
WARREN V. BLASLAND, JR.
WILLIAM H. BOUCK
EDWARD R. LYNCH
ROBERT K. GOLDMAN
O'Brien & Gere Engineers, Inc.
Syracuse, New York
INTRODUCTION
In September 1980, the General Electric Company and
the New York State Department of Environmental Con-
servation (NYSDEC) signed the first major agreement in
the nation in which a corporation agreed to pay for re-
medial action to clean up abandoned hazardous waste
dump sites where, in the past, it had disposed of its in-
dustrial waste. The agreement involves remedial actions
at seven land disposal sites in Saratoga, Rensselaer, and
Washington Counties. General Electric has agreed to carry
out all engineering studies and the necessary remedial ac-
tion at four locations and has agreed to pay a percentage
of the cost of the engineering studies and necessary
remedial action at the three other sites.
One of the sites included in the agreement is known as
the Fort Miller Site. It is located in the Town of Fort Ed-
ward, Washington County, approximately 7.5 miles
south of Hudson Falls, New York. (Figure 1.)
GE/NYDEC AGREEMENT
Terms of Agreement—Fort Miller Site
General Electric, under the terms of the agreement,
conducted field investigations at the site to determine the
current condition of the site, including the hydrogeology,
the areal and vertical extent of waste present, the physical
state of the waste and the means by which wastes have
been released, may be released, have migrated or may mi-
grate from the site.
At the completion of the field investigations, an engi-
neering report was prepared to include:
•All data from the field investigations
•Identification of alternative remedial programs
•Selection of a recommended remedial program to meet
established goals. (The goal of the remedial program
was to abate any significant current and future releases
of hazardous wastes from the site.)
•Implementation schedule and preliminary plan including
the identification of all property to which access would
be required to implement the remedial program
•Program for continuing site maintenance
•Establishment of monitoring plan to evaluate the success
of the remedial program
Following approval of the engineering report from
NYSDEC, General Electric submitted a final plan for im-
plementation of the recommended remedial program.
Following approval of the final plan, General Electric
will complete the construction elements of the remedial
program at the site.
Upon completion of the construction elements of the
remedial program and following certification by NYSDEC
that the construction elements are in accordance with the
approved engineering report and final plan, General Elec-
tric will maintain and monitor the site for a 30 year period.
Current Status
The Fort Miller Site Engineering Report and final plan
have been approved by NYSDEC. The construction ele-
ments of the recommended remedial program will be be-
gun in March, 1982 and will be completed during 1982.
FORT MILLER SITE
General
The Fort Miller Site was an open municipal/industrial
dump site of 2.5 acres which was reported to have oper-
\ in-
AREA OF INVESTIGATION
Figure 1.
General Site Location Plan
215
-------
216 REMEDIAL RESPONSE
ated as a burning dump between 1953 and 1965. During
the period in which dumping was taking place at the site,
waste products containing polychlorinated biphenyls
(PCBs) were deposited at the site. These waste products
included reject electrical components which contained
PCBs.
Background Information
Prior to developing a field investigation work plan at
the site, all background regional and site specific historical
information was reviewed including aerial photographs of
the site prior to and during the years in which the site was
operated. The aerial photographs aided in identifying the
manner and mode in which the site was operated. Coun-
ty-wide soil and groundwater surveys, local groundwater
well logs and USGS topographic mapping were also re-
viewed.
Field Investigations
A field investigation was undertaken to determine the
three-dimensional limits of the site, and to defined the
ways and extent by which PCBs were migrating from the
site. The field techniques utilized during the field investi-
gation included:
(1) Topographic survey
(2) Magnetometer survey
(3) Subsurface exploration (test borings and test pit
trenching)
(4) Groundwater monitoring
(5) Surface water monitoring
Topographic Survey
A topographic survey was conducted at the site and im-
mediate surrounding area to determine the location and
elevation of significant features at the site. Elevations
ranged from about 215 feet (USGS datum) at the south
end of the site to 160 feet at the toe of the site. Surface
water runoff from the site and adjacent areas flows into
intermittent drainage ways which ultimately empty into
the Moses Kill, a tributary of the Hudson River.
Magnetometer Survey
A magnetometer survey was conducted at the site to de-
fine the southern, western and northwestern limits of
dumping at the site. The steeply sloped northern and eas-
tern limits were determined by visual observation. It was
necessary to accurately define the limits to insure that al-
ternative remedial programs considered would encompass
the entire area containing dumped refuse.
Subsurface Exploration
Subsurface explorations included the drilling of test bor-
ings and the digging of test pit trenches. Fourteen (#1,
3, 7, 8, 9, 11, 12, 13, 13A, 14, 19, 20, 21, 22) test bor-
ings were drilled at the site as shown in Figure 2. Seven
(#1, 3, 9, 13, 19, 20, 22) of the test borings were drilled to
refusal and in two of them (#1, 20), rock core samples
were obtained. In the seven test borings drilled to refusal,
soil samples were taken at five-foot intervals. In the re-
maining seven, no samples were taken and soil descrip-
tion were made from the driller's observation of auger
cuttings as they came from the hole.
The subsurface materials encountered at the site were
consistent with those reported in the county soil survey.
They were divided into three basic soil types: A medium
stiff to very stiff brown to gray brown clay. Generally this
material, within the limits of the dumping area, extended
to bedrock. Outside the limits of the dumping area the
brown to gray clay material was underlain by a medium to
stiff gray varved clay. The brown to gray, medium to stiff
gray varved clay was generally encountered at a depth of
10 to 15 feet below grade. This material was softer and
moister than the brown to gray brown clay. The bedrock
encountered, underlying the above clay soil, was a gray,
very dense, silty weathered shale. Depth to bedrock under
the site and along the periphery ranged from 23 to 35 feet.
Test pit trenching operations were conducted to verify
horizontal limits of the site as determined by the magneto-
meter survey, to determined the vertical limits of the
dumping at the site and to determine the limits of capac-
itor dumping at the site, which if the dumping was isolated
to a smaller area(s), would allow the remedial program to
address a smaller area(s). The location of the test pit
trenches are shown in Figure 2.
With the exception of trench No. 1, all trenches had
evidence of various quantities of capacitors or capacitor
paper. Based on these data, it was decided that the remed-
ial program would encompass the entire site. The hori-
zontal limits of the site defined by the magnetometer sur-
vey were verified by the test pit trenching. The vertical
limits of dumping at the site were defined by data obtained
from the trenching operations in conjunction with test bor-
ings installed through the site to the clay. The estimated
volume of refuse was 20,000 yd3.
During the trenching operations, water was observed
at the clay/refuse interface in trenches 4, 5, and 6. In each
of these locations a stainless steel groundwater observa-
tion well was installed for sampling prior to backfilling.
A cross section through the site is shown in Figure 3.
Groundwater Monitoring
Groundwater monitoring wells were installed in each of
the fourteen test boring holes. In addition, five (#2,4,5,6,
10) five-foot long stainless steel wellpoints were installed
at the base of the dumping area. Three others (#16, 17,
18) were installed in test pit trenches #4, 5, 6, respectively,
prior to backfilling. The location of all groundwater
monitoring wells is shown in Figure 2.
Several sets of groundwater elevations were taken at the
site. Wells #16, 17, 18, and 21 were shallow monitors (5 to
10 feet below grade) which extended through the dumped
refuse to the clay. The measured groundwater elevations
in these wells, along with the visual observations made
during the test pit trenching operations, indicated the
-------
REMEDIAL RESPONSE 217
V
LEGEND
-0. BORROW UREA TEST BORINGS
0 TEST BORINGS «/ GROUNOWATER MONITORS
B SURFACE WATER SAMPLES /
O WELLPOINT MONITORS
A STREAM SEDIMENT SAMPLES
O FLOW MONITOR
t^^m TRANCHES
LIMITS OF REFUSE
Figure 2.
Fort Miller Site—Existing Site Plan
presence of a local zone of saturation (or perched water
table) at the refuse/clay interface. Discharge of this perch-
ed water table does occur along the exposed faces of the
dumping area.
Well Nos. 19, 20, 22 are deep monitors (23-25 feet below
grade) which extend into the underlying clay. Based on
groundwater elevations taken from these wells, the perm-
anent water table generally occurs at a depth of from
3-15 feet below the clay. A profile of the permanent
groundwater table is shown in Figure 4. Based on the pro-
file, it was determined that groundwater moves radially
from the site towards the intermittent streams.
Groundwater samples were collected from 21 of the 22
wells (#21 was dry). A strict sampling protocol was im-
plemented to eliminate inadvertent introduction into the
well (or sample) of substances which lead to inaccurate
results during subsequent laboratory analysis. A quality
control procedure was also implemented to insure accuracy
of the field and laboratory procedures and analyses. The
collected samples were analyzed for pH, conductivity,
PCB, Total Organic Carbon (TOC) and selected heavy
metals (lead, chromium, nickel, copper, selenium and
zinc). For verification purposes, several of the wells were
resampled and reanalyzed. All groundwater samples were
filtered prior to analysis.
Of primary interest to this field investigative effort was
the concentration of PCB migrating from the site. A gen-
eral summary of the data is as follows:
•No PCB was detected in the background well (#1).
•The concentration of PCB in the wells installed around
the periphery of the fill area ranged from less than 0.01 to
3.7 ug/1 averaging0.7 jig/1.
•The concentration of PCB in the wells (#16, 17, 18) in-
stalled in the perched groundwater table ranged from
1.23 to 46 ug/1 averaging 19.74 ug/1.
Surface Water Monitoring
Surface water runoff from the Fort Miller Site is
collected in a system of drainage swales which converge,
at the northwest corner of the site, into an unnamed in-
termittent stream, tributary to the Moses Kill. A V-notch
weir was installed across this stream at the location shown
on Figure 2.
Flow measurements were taken at the weir location for a
two month period. Also, rainfall intensities were moni-
tored over the same period of time.
Stream samples were collected at peak and base flow
conditions at the weir and at background locations and
analyzed for PCB. The concentration of PCB at the weir
location (#6) (Figure 2) during the peak and base flow
conditions were 0.30 and 0.80 jig/1 respectively. The con-
centration of PCB at the background locations (#1, 2, 3,
4, 5) (See Figure 2) during the peak and base flow con-
-------
218 REMEDIAL RESPONSE
EXISTING
GR10E
400
SECTION A
Figure 3.
Existing Site Plan—Cross-Section A
ditions ranged from 0.04-0.58 yg/1 and 0.08-0.96 ug/1,
respectively. Surface water samples were not filtered prior
to analysis.
Air Monitoring
Prior to the initiation of the Fort Miller Site, it was
acknowledged, because of previous air monitoring con-
ducted by NYSDEC, that volatilization and airborne dis-
persion of PCB from the site was a problem which would
have to be addressed by the proposed remedial program.
For that reason air sampling and analysis was not con-
ducted at the site.
Safety Protocol
The implementation of the various field investigative
techniques in an area known to contain PCB wastes re-
quired the use of a field safety protocol. The safety pro-
tocol implemented was:
•During the drilling of test borings, the installation of
groundwater monitoring wells, the digging of surface
trenches and the sampling from groundwater wells, a dual
carbon respirator, disposable rubber gloves and boots
and a disposable acid-resistant suit were worn.
•During other field investigations, the safety equipment
listed above was optional, unless a distinct chemical odor
was noted, in which case the dual carbon respirator was
worn.
With the exception of the dual carbon respiratot^new
safety equipment was worn each day (when used) and was
buried at the site at the end of each day's work. Field
personnel were instructed to replace filters on a weekly
basis. A hard hat was worn at all times and a Scott Air
Pack and emergency eyewash station were available on-site
for use if needed.
PCB TRANSPORT MODES
As stated previously, the goal of the remedial program
is to abate any significant current and future releases of
wastes from the site. The sampling and analyses which
were accomplished indicate that the contaminant of pri-
mary concern is PCB. Identified transport mechanisms
by which PCB may be released from the site included
groundwater migration, erosion transport and volatiliza-
tion.
Groundwater Migration
From the field investigations, it has been determined
that there are two groundwater zones at the site, a shallow
perched groundwater zone at the clay/refuse interface
and a permanent groundwater zone at a depth of 3-20
feet below grade within the dumping area.
The groundwater discharge flow rates from the shallow
perched groundwater zone and the clay permanent ground-
water zone were calculated to be 500 gal/day and 5,000
gal/day, respectively.
-------
REMEDIAL RESPONSE 219
Results of sampling and analysis from monitoring wells
installed within the dumping area indicate PCB concen-
tration in the leachate ranging from 1.23 to 46 ug/1. At
the maximum flow rate of 500 gal/day in the perched
water table, the peak quantity of PCB discharged from the
site in the leachate flow would be in the range of 0.002
to 0.070 Ib/year. Similarly, measured PCB concentrations
in the permanent groundwater zone at the periphery of
the site range from 0.01 to 0.37 Ug/1, indicating a peak
quantity of PCB discharged from the site through the
groundwater system (maximum flow rate of 5,000 gal/
day) would be on the range of 0.0002 to 0.0056 Ib/year.
Erosion Transport
As discussed previously, samples of surface water run-
off were collected from the drainage channels during a
rainfall event and analyzed for PCB. Although these data
are insufficient to accurately quantify PCB losses, they in-
dicate that the actual transport of wastes through erosion
was on the magnitude of 1.0 Ib/year.
Volatilization
As discussed previously, volatilization of PCB from the
site was a known phenomena which would have to be
addressed by the recommended remedial program.
ALTERNATIVES
Based upon a review of the PCB transport modes and
their significance at the Fort Miller Site, potential secure-
ment alternatives were identified and evaluated. The three
alternatives which were considered in the engineering re-
port were:
•In-place containment of wastes
•Removal of wastes to new on-site secure landfill
•Removal of wastes to off site secure landfill
Under each of these alternatives there were a number
of options pertaining to construction materials, surface
drainage, groundwater control, leachate collection and
control of volatilization. The criteria used in evaluating
the alternatives included technical feasibility, ease of im-
plementation, compliance with regulatory requirements
and costs.
Each of the three alternatives considered, if properly
implemented, would satisfy the goal of the Fort Miller Re-
medial Program. Also, each alternative is compatible with
existing State and Federal regulations and would be like-
ly to receive regulatory agency acceptance.
Within certain limitations, each alternative is technically
feasible. However, there are certain advantages and dis-
advantages to each of the alternatives. The primary ad-
vantage of in-place containment was that existing solid
waste deposits would not be disturbed. Under the other
two alternatives, the excavation and transfer of the refuse
greatly increases the exposure, and therefore the potential
for release, of PCB to the environment.
The critical period of exposure would be during the
excavation phase, when the main PCB deposits now buried
below the surface would be uncovered. An intense rain-
storm during this period could result in a large loss of
PCB to the environment through erosion. The potential
for volatilization losses during this period would also be
greatly increased during removal to a new on- or off-site
facility. The period of potential exposure includes the
transportation from the site to the secure facility. Based
on the above and cost estimates, in-place containment of
wastes was recommended for the Fort Miller Remedial
Program.
RECOMMENDED REMEDIAL PROGRAM
The recommended remedial program is in-place contain-
ment of wastes, to include a clay cap with vegetative cov-
er, a relocated surface drainage system, a leachate collec-
tion system and a gas venting system. The purpose of
these facilities is to isolate the waste materials from the
surrounding environment. The plan also includes pro-
visions for maintenance of the facilities, and monitoring
to measure the effectiveness of the remedial program.
For the purpose of describing the remedial program and
monitoring plan, the site will be defined as the vertical
and horizontal area containing dumped refuse.
Site Preparation
All trees and brush within the area to be capped will be
removed, and stumps will be ground to grade. The wood
chips will be spread evenly over the site prior to capping.
Tree stumps within the area to be capped will be chem-
ically treated to kill roots and prevent resprouting. To
minimize future vertical displacement of the final cap, the
surface of the existing site will be proof rolled.
The site area will be regraded to reduce the slopes at
the northern and eastern sides. The regrading will consist
of placing fill at the base of the site to extend to toe of the
slope northward and eastward.
The material to be used as fill will be on-site clay which
will be excavated from borrow pits located east and north
of the existing drainage swales. The material will be in-
stalled in lifts not to exceed 12 ft at a standard proctor den-
sity of 80-85%.
Final Cover
Following site preparation, the site will be covered with
a clay cap, using the same on-site soils from which the
embankments are to be formed. The cap will consist of a
minimum 3.5 ft thick clay layer. The bottom 2 feet layer
will be installed in 6" lifts with each lift achieving stand-
ard proctor density of 90%. The top 1.5 ft (minimum)
clay layer will be installed in 0.5 ft lifts with each lift
achieving a standard proctor density of 80-85%. The cap
will be covered with a minimum of 0.5 ft of topsoil and
seeded to promote a grass cover.
Each winter, the upper zone of the cap will be sub-
jected to freezing. In frost-susceptible soils, such as clays,
the formation of ice lenses, and the subsequent thawing of
these lenses in the spring, can leave cracks in the clay ma-
-------
220 REMEDIAL RESPONSE
terial which could raise the average permeability of the soil
mass. The degree of ice lens formation is related to the
depth of protective cover, the availability of the fresh
water and the size of the pore spaces in the soil.
At the Fort Miller Site, ice lens formation should not
be a serious problem because the on-site silty clays are
generally less frost susceptible than other silty types. In
the Fort Miller area, the average depth of frost pene-
tration is about 2.5 ft. Construction of a minimum 3.5 ft
thick clay cap, covered with 0.5 ft topsoil will leave a com-
pacted clay layer extending 1.5 ft below the average frost
line, minimizing the possibility of cracking the impervious
layer due to annual freeze/thaw cycles.
In dry summer months, an exposed clay layer would be
subject to shrinkage cracking. The placement of the topsoil
layer along with the vegetative cover will provide an in-
sulating effect which would prevent evaporation and sub-
sequent shrinkage. If minor cracks do develop, the rein-
troduction of moisture into the clay will result in swelling
which will close the cracks.
The clay cap will be installed on a 5% grade on top,
and a 1:6 slope over the horizontal limits of the site. In
areas outside the limits of the cap, the clay materials will
be installed on a 1:4 slope in fill areas and a 1:3 in cut
areas.
Surface Drainage
At the present time, all surface runoff from the site is
collected in a series of drainage channels along the toe of
the site. The fill placed during regrading operations will
cover the existing channels along the north and east faces
of the site, and new channels will be constructed at the base
of the proposed slopes.
Groundwater Control System
The surface drainage channel will be installed to a depth
of a minimum of ten feet below the toe of the refuse to
control the level of the permanent groundwater table.
Leachate Control System
After placement of the final cover, the infiltration of
surface water into the refuse layer and the subsequent
leaching of this water through the face of the landfill will
be eliminated. Leachate could continue to discharge from
the site, however, until the local zone of saturation at
the interface of the refuse and underlying clay is drained.
To eliminate this potential uncontrolled discharge, a leach-
ate collection system will be installed around the site.
The collection system will consist of perforated drain-
age pipe installed along the base of the landfill. The pipe
will be laid in a shallow trench backfilled with granular
material up to the elevation of 1-2 ft below the bottom of
the refuse. The side of the trench downgradient of the
landfill will contain a barrier of compacted clay fill to
prevent leachate from migrating beyond the drain.
The collection system will discharge to a pumping sta-
tion which will pump the leachate to an underground
holding tank, the content of which will be periodically
pumped to a portable tanker for transport to a permitted
treatment facility. Since the impermeable cap over the site
will eliminate percolation of surface water into the refuse,
the source of additional leachate will be removed and the
existing leachate should be collected and disposed of with-
in an estimated six months to one year following cap in-
stallation.
Gas Control System
In a typical municipal landfill, methane gas is pro-
duced by the biological degradation of organic materials
under anaerobic conditions. Following the installation of
the cap over the site, the generation of gases from the
refuse could result in pressure-induced stresses which could
damage the cap.
The length of time the site has been closed, along with
the burning of refuse which has been reported, would in-
dicate that most organic materials have already decom-
posed, and that the potential for future generation of
significant quantities of methane is therefore quite lim-
ited. However, to ensure that the cap is not damaged by
gas pressures, the recommended remedial program will in-
clude a passive collection and treatment system.
The gas collection system will consist of a network of
perforated pipes laid in shallow trenches excavated into
the surface of the refuse. The trenches will be backfilled
with gravel and enclosed at the surface by the final cov-
er material. The gas collection pipes will lead to a common
above-ground vent. The vent will be fitted to an adsor-
bent-type treatment unit and will remain open to the at-
mosphere.
Site Security
Site security will include fencing sufficient to restrict
unauthorized access and warning signs to discourage tres-
passers. The Final Plan showing the recommended alterna-
tives is shown in Figure 4. A cross-section taken through
the site is shown in Figure 5.
MAINTENANCE PROGRAM
A major portion of the long-term maintenance effort
will involve the vegetative cover on the completed cap. No
trees, shrubs, brush or deep rooting weeds should be
allowed to germinate or establish on the site. If visual ob-
servations made during the performance of other main-
tenance activities indicate that low growing, deep rooting
weeds have germinated from the cap, a herbicidal weed
control program would be initiated. Periodic inspection of
the site will also reveal any problems of erosion, disease,
or thinning of grasses which would then be corrected. The
grasses on the site will also be moved periodically as re-
quired.
The estimated 50,000 gal of leachate will be collected
and treated as required. The gas collection facilities will
be operating continuously. The adsorbent-type treatment
unit will be replaced as required. All mechanical equip-
-------
REMEDIAL RESPONSE 221
Figure 4.
Fort Miller Site—Final Plan
ment (i.e., pumps) will be maintained as suggested by the
manufacturer.
MONITORING PROGRAM
The purposes of undertaking monitoring activities are
to measure the effectiveness of the remedial program.
A 30 year monitoring program was developed to deter-
mine failure of the remedial program. Failure of the re-
medial program is deemed to have occurred if any of the
following conditions are observed:
(1) Any portion of the cap is eroded, allowing wastes to
be carried away by surface runoff
(2) Surface water percolates through the cap and into
the refuse, creating a source of leachate
(3) Waste material is volatilized through the cap in
excessive quantities
(4) The groundwater table rises above the bottom of
the refuse, resulting in water saturation of the refuse
and the possibility of the formation of leachate
The monitoring activity which will determine if the
above failure modes have occurred, respectively, is as fol-
lows:
(1) Cap Erosion—Periodic inspection of the cap
(2) Surface Water Infiltration—Flow is observed in the
leachate collection system (after the shallow perched
groundwater has been drained).
(3) Volatilization—PCB concentration rise significantly
above baseline concentration. The baseline concen-
tration will be established following cap installation.
(4) Groundwater Table Interrupting Site—Ground-
water levels rise to intercept the site. The ground-
water table will be monitored continuously in the
vicinity of the toe of the site.
By undertaking the monitoring activities outlined above, it
will be possible to measure the effectiveness of the re-
medial program and to detect any failure of the system
to abate the release of material from the site.
Failure of the remedial program will not be deter-
mined by analyzing groundwater or surface water samples
for PCB. The very low PCB concentration levels in the
groundwater, presently below the refuse layer, may differ
at different sampling locations or differ at the same samp-
ling location at different times for many years.
A variation in PCB concentration at any downgrad-
ient groundwater monitoring or surface water location
could simply be the result of a higher or lower zone
of PCB concentration moving slowly past the sampling
point. Because these PCB concentrations may vary con-
siderably and since such variation would not necessarily
be the result of a failure of the remedial program, it
was not attempted to define failure of the remedial pro-
gram by attempting to interpret such results.
-------
222 REMEDIAL RESPONSE
s
/?!"
EXISTING GB40E
LEGEND
* \ * APPROX.
\» REFUSE
LIMITS
100
SECTION A
Figures.
Final Plan—Cross-Section A
CONCLUSIONS
The work accomplished under this program and the re-
sults can be briefly summarized as follows:
(1) Field investigations were undertaken to determine
the condition of the Fort Miller site and the physical
extent of the wastes. It was determined that there
are about 20,000 yd3 of refuse which have been
dumped in an area of approximately 2.5 acres.
(2) Sampling and analyses were undertaken to define
groundwater flow patterns and rates, and to esti-
mate the magnitude of migration of wastes (PCB)
from the immediate dumping area via groundwater
flow. It has been shown that the loss of PCB by this
mechanism is insignificant.
(3) The primary means by which PCB could be trans-
ported from the dumping area are volatilization
and erosion.
(4) Three alternative remedial programs for the Fort
Miller site were evaluated: in-place containment,
removal to a new on-site landburial facility and re-
moval to an off-site landburial facility. The criteria
used to evaluate these alternatives include compat-
ibility with program goals, technical feasibility, ease
of implementation, compliance with regulatory re-
quirements and costs. In-place containment has
been designated as the recommended remedial pro-
gram for the Fort Miller Site.
(5) A recommended Fort Miller remedial program has
been presented. The program is based on in-place
containment of the wastes, including a clay cap with
vegetative cover, relocated surface drainage system,
leachate collection system, and gas venting system.
The plan also includes provisions for maintenance
of the facilities and monitoring to measure the pro-
gram's effectiveness.
(6) The preliminary implementation schedule calls for
construction activities to be completed during 1982.
REFERENCES
1. Blasland, W.V., Knowles, G.D., Lynch, E.R., and
Goldman, R.K., "Remediation at an Inactive Munic-
ipal Disposal Site Containing PCB's", Proc. 13th
Mid-Atlantic Industrial Waste, June 29, 30, 1981;
476.
-------
ORGANIC LEACHATE EFFECTS ON THE
PERMEABILITY OF CLAY LINERS
D.C. ANDERSON
K.W. BROWN, Ph.D.
J. GREEN
Texas A&M University
Texas Agricultural Experiment Station
College Station, Texas
INTRODUCTION
Saturated conductivity or permeability is the primary
laboratory measurement made on compacted clay soils to
assess their suitability for use in constructing compacted
clay liners for hazardous waste landfills or surface im-
poundments. The value obtained for the permeability is
used to judge whether a compacted clay soil liner will pre-
vent the movement of leachates into water bodies below
or adjacent to the disposal facility.
There is, however, little information available on the
impact of organic fluids, likely to be placed in such con-
finements, on the permeability of the liner. Also, there
has been on simple permeability test method developed
that could be suitable for use with the range of possible
waste fluids. Furthermore, a study evaluating the permea-
bility of a range of typical clay soils with a spectrum of
potential waste fluids would generate a valuable data base,
useful for hazardous waste disposal permit applicants,
writers and reviewers. Consequently this study was un-
dertaken with the following objectives:
•To construct from available information a delineation of
the physical classes of fluid-bearing hazardous waste,
the leachates they generate, and the predominant fluids
in these leachates.
•To develop a simple, inexpensive and rapid method for
the comparative permeability testing of compacted clay
soils, that would be suitable for use with a wide range
of possible waste fluids.
•To evaluate the permeability of a range of typical clay
soils to a spectrum of potential waste fluids.
PHYSICAL CLASSES OF HAZARDOUS WASTE
Land disposed hazardous wastes generally fall into the
following four physical classes: aqueous-inorganic, aque-
ous-organic, organic, and sludges(1) (Table I).
Aqueous-inorganic is the class of wastes in which water
is the solvent (dominent fluid) and the solutes are mostly
inorganic. Examples of these solutes are inorganic salts,
metals dissolved in inorganic acids and basic materials
such as caustic soda. Examples of wastes in this category
are brines, electroplating wastes, metal etching wastes and
caustic rinse solutions.
Aqueous-organic is the class of wastes in which water is
the solvent and the solutes are predominantly organic.
These solutes are organic chemicals that are polar or
charged, as is inferred by their water solubilities. Examples
of this class of waste are wood preserving wastes, water-
based dye wastes, pesticide container rinse water and
ethylene glycol production wastes.
Organic is that class of wastes in which an organic fluid
is the solvent or dominant fluid and the solutes are other
organic chemicals dissolved in the organic solvent. Ex-
amples of this class of wastes are oil-based paint wastes,
pesticide manufacturing wastes, spent motor oil, spent
cleaning solvents and solvent refining and reprocessing
wastes.
Sludges represent the fourth class of wastes. They are
generated when a waste stream is dewatered, filtered or
treated for solvent recovery. Sludges are characterized
by high solids content such as that found in settled matter
or filter cakes and consists largely of clay minerals, silt,
precipitates, fine solids and high molecular weight hydro-
carbons. Examples of this waste are API separator sludge,
storage tank bottoms, treatment plant sludge or any fil-
terable solid from any production or pollution control
process.
Both economic and pollution control pressures con-
tinue to mandate solvent recovery and reductions in dis-
charges of aqueous waste streams. These factors have
made, and will continue to make sludges the fastest grow-
ing class of wastes. After placement of sludges in a waste
disposal facility, leachates migrate out of the sludge due
to gravitational forces, overburden pressures and hy-
draulic gradients. These leachates are similar in physical
form to the first three classes of waste shown in Table I.
LEACHATE GENERATED BY HAZARDOUS WASTE
To determine the effect of a specific waste on the per-
meability of a specific clay liner, two unique leachates
Table I.
Physical Classes of Hazardous Wastes
Waste Class
Aqueous-inorganic
Aqueous-organic
Organic
Sludges
Solvent Phase
Water
Water
Organic Fluid
Organic Fluid or
Water
Solute Phase
Inorganic
Organic
Organic
Organic and
Inorganic
223
-------
224 REMEDIAL RESPONSE
HYDRAULIC AND
BEARING PRESSURE
INFILTRATION OF
OUTSIDE WATER
FLUID PORTION OF
THE WASTE
HAZARDOUS
WASTE
WATER SOLUBLE
PORTION
OF THE WASTE
PRIMARY LEACHATE
CLAY
LINER
SECONDARY
LEACHATE
i
UNDERLYING '
STRATA
Figure 1.
Sources of Leachate that May Come in Contact with Clay Liners
LEACHATE
[ SOLVENT or FLUID PHASE | | SOLUTE or DISSOLVED PHASE |
[ORGANIC FLUIDS
ORGANIC
[INORGANIC]
Figure 2.
Composition of the Leachate of a Waste
DISSOLVED
PHASE
PREDOMINANTLY
ORGANIC
PREDOMNANTY
AQUEOUS
INORGANIC
COMPONENTS
ORGANIC
COMPONENTS
Figures.
Primary Leachate Generated at Disposal Sites
FLUID
PHASE
DISSOLVED
PHASE
Figure 4.
Secondary Leachate Generated at Disposal Sites
must be investigated. These leachates are the flowable
constituents of the waste and the flowables generated from
percolating water leaching through the waste (Figure 1).
Flowable constituents of a waste, hereafter referred to
as the primary leachate, includes both fluids in the waste,
the solvent and all components dissolved in these fluids,
the solutes. Depending on the physical class of a waste, its
primary leachate may be aqueous-organic, aqueous-
inorganic or organic. Leachate generated from water per-
colating through a disposal facility is composed of water,
the solvent and all the components dissolved or carried
with it in the water, the solutes. This flowable mixture,
hereafter referred to as the secondary leachate, may be
aqueous-organic or aqueous-inorganic, depending on the
composition of a waste.
The predominant fluid or solvent phase of a leachate
may be water or any organic fluid (Figure 2). Corre-
sponding solutes in a leachate are any chemicals that dis-
solve in the solvent phase. As in the physical classes of
wastes, both primary and secondary leachates are di-
vided into a solvent phase (the predominant fluid compon-
ent) and a solute phase (components dissolved in the
solvent). (Figures 3 and 4). While the solutes in the leach-
ate may affect the permeability of a clay liner, the solvent
phase will usually exert a dominating influence on per-
meability.
Essentially all available literature describing leachates
generated by hazardous waste disposal considers only
water as the predominant fluid.(2>3) Water is viewed as the
carrier fluid and the organic chemicals are considered to be
present in only trace quantities. While this is probably the
case at the interface of a secondary leachate and a water
table, the fluid phase at the interface of a primary leachate
and a clay liner will depend on the physical class of the
waste being disposed. An organic waste or sludge with an
organic fluid phase will most probably expose the clay
liner to the concentrated organic fluids contained in the
waste.
FLUID PHASE
Pin
Figures.
Solvent Phase of the Leachate of a Waste
FLUIDS IN HAZARDOUS WASTE LEACHATE
As stated previously, organic fluids that have been
placed in waste impoundments cover the spectrum of
chemical species. For the purpose of experimentally evalu-
ating the effects organic fluids may have on the permea-
bility of clay soils, the fluids have been classified into four
groups. These groups are based on the physical and chem-
ical properties that will govern their interactions with clay
minerals. These properties include acidity, basicity and
polarity (Figure 5). All clay liners are initially wetted with
water. Consequently, the interaction of organic fluids
with clay minerals must be viewed in the context of fully
hydrated clay minerals.
-------
REMEDIAL RESPONSE 225
Table II.
Properties of the Four Organic Fluids and Water
Fluids
Acetic Acid
Aniline
Xylene
Methanol
Water
Temp.
Fluid
Melting
Point
17
-6
-47
-98
0
Range of
State (°C)
Boiling
Point
113
184
139
65
100
Density
at 20°C
g/cm3
1.05
1.02
0.87
0.79
0.98
Viscosity
at 20°C
Centipoise
1.28
4.40
0.81
0.54
1.0
Water Solubility
at 20°C
g/1
00
34.0
0.2
0=
»
Dipole
Moment
(debyes)
1.04
1.55
0.40
1.66
1.83
Dielectric
Cons t an t
at 20°C
6.2
6.9
2.4
31.2
80.4
Regardless of the type of organic fluid, if it has a low
viscosity, it will be leachable. Fluids with low viscosity are
"by their very nature" leachable and able to extract or-
ganic components from otherwise dry waste.(4) However,
several other fluid properties effect the resultant soil per-
meability values. Some of these properties are discussed
with respect to the four groups of organic fluids and
water in the following sections.
Organic Fluids Studied
Organic fluid representatives of the four classes of or-
ganic fluids (Figure 5) were selected for use in the compar-
ative permeability studies. The four classes studied were
acidic (acetic acid), basic (aniline), neutral polar (methan-
ol) and neutral nonpolar (xylene) organic fluids. The four
organic fluids and water along with their relevant physi-
cal and chemical properties are shown in Table II.
Water containing CaSo4 at a concentration of 0.01N was
used to establish the baseline permeability of each soil
core and as the control fluid. The calcium salt was se-
lected for use due to its stabilizing effect on permeabil-
ity. A concentration of 0.01N was selected because it ap-
proximates the salt concentration typically found in soils.
Clay Soils Studied
Four clay soils typical of those used for the lining of
hazardous waste landfills and surface impoundments were
selected for these studies. These clay soils are referred to
by the following names: noncalcareous smectite, cal-
careous smectite, mixed cation kaolinite and mixed cation
illite. The soils are described in Table III.
MATERIALS AND METHODS
To assess the suitability of compacted clay soils for the
lining of waste disposal facilities, the primary laboratory
measurement made is saturated conductivity or permea-
bility. These tests often take months to complete for two
reasons. Compacted clay soils often have permeability
values lower than 1(T8 cm/sec, and it is often necessary to
pass a pore volume of water through a soil before a stable
baseline permeability value can be obtained. After the
permeability baseline is established, the passage of at
least a pore volume of the organic test fluid may be neces-
sary to fully determine the effects an organic fluid may
have on the permeability of the compacted clay soil. Con-
sequently a pressurized air source has been used to increase
the hydraulic gradient and reduce the time needed for
testing.15'6'
In this study, the authors utilized a pressurized, air-
induced, elevated, hydraulic gradient in a rapid compara-
tive permeability method for use on compacted clay soils.
A hydraulic gradient of 361.6 (equivalent to a hydraulic
head of 42.2 m of water) was used for the two smectite
clay soils. A hydraulic gradient of 61.1 (equivalent to a
hydraulic head of 7 m of water) was used for the illitic
and kaolinitic clay soils.
Table III.
Properties of the Four Clay Soils
Clay SoH
Description
% Sand
( 50nm)
% Silt
( 50-2.0nm)
% Clay
( 2.0nm)
Predominant
clay
minerals*
Shrink-Swell
Potential
Corrosivity
(Steel)
Cation Exch.
Capacity
(meg/lOOg)
Total
Alkalinity
(meg/lOOg)
Noncalcareous
Smectite
35-37
26-28
36-38
S
M
K
Very high
High
24.2
3.3
Calcareous
Smectite
7-8
42-44
48-50
S
K
Very high
High
36.8
129.3
Mxd Cation
Kaolinite
39-41
17-18
42
K
M
Mod.
High
8.6
0.8
Mxd Cation
Illite
14-15
38-39
47
I
S
Mod.
N.D.
18.3
4.2
•In order of descending quantity in the soil.
N.D. Not determined
S—Smectite, M—Mica, K-Kaolinite, I—Illite
-------
226 REMEDIAL RESPONSE
The equation used for calculating the permeability of
the compacted clay soils was as follows:
K =
ATH
(1)
where:
K = permeability constant (cm/sec)
V = volume of liquid (cm3)
A = cross-sectional area of liquid flow (cm2)
T = time (sec.)
H = hydraulic gradient
Compacted soil cores used to evaluate the permeability
of the soils to organic fluids were prepared at or above
the optimum water content. After the clay soils to be test-
ed were compacted into the soil chambers, they were
mounted in permeameter base plates and fitted with a
fluid chamber and permeameter top plates. Each top plate
was fitted with a pressure inlet to connect it to the pres-
surized air source via the pressure distribution manifold
(Figure 6).
A moisture trap, pressure regulator and pressure gauge
was placed between the pressurized air source and mani-
fold. The moisture trap was positioned between the air
source and regulator to prevent the buildup of debris on
the membrane in the regulator. The pressure gauge went
between the regulator and manifold so that the hydraulic
pressure being applied to the permeameters could be
monitored.
Cutoff valves for the pressure leading to each permea-
meter were placed between the manifold and permeameter
top plates. These valves allowed the placement or removal
of individual permeameters without depressurizing the
entire test system.
To avoid channel formation, the compacted clay soils
were seated at low pressure. By letting 10 cm of the stan-
dard leachate stand on the soils for 48 hours, an effective
seat was obtained for the top few millimeters of the soil.
The thin layer prevented bulk flow, thereby permitting
the rest of the soil to adequately seal the permeameter
sidewalls at elevated pressures.
All gaskets used in the permeameters were teflon to pre-
vent their deterioration and possible blow out from con-
tact with the various organic fluids. To avoid leakage
around the hard teflon gaskets, all metal surfaces against
which the gaskets seated were wiped clean of grit. In gen-
eral, all permeameter components were found to with-
stand continuous operational use at pressures equivalent
to 50m of water.
Directly under the soil chambers and in the permea-
meter base plates were porous stones to permit seepage
of the leachate to the outlet in the base plate. To limit
the extent of leachate mixing after passage through the
compacted soil, the outlet was fitted with an adapter to
0.3 cm inside diameter teflon tubing. The use of trans-
lucent teflon at the outlet on the base plate provided a
convenient window with which to monitor the expulsion
of trapped air. Standard leachate (0.01N CaSo4) was
passed through all soil chambers until there were no air
bubbles visible in the outlet tubing. Where soil piping oc-
curred in soil samples, eluded soil clay particles were vis-
ible both clinging to the inside walls of the outlet tubing
and as a suspension in the collected flow samples.
Teflon tubing carried the leachate to an automatic frac-
tion collector which collected leachate samples simultane-
ously from ten permeameters at specific time increments.
Since there was a potential for volatile losses occurring
during leachate delivery from the tubing to the sample
bottles in the fraction collector, the top of each sample
bottle was fitted with a long stem funnel and the fraction
collector was placed in an air tight cooled chamber. Addi-
tionally, the entire test apparatus was fitted into a vented
hood (Figure 6). This extra precaution was taken as insur-
ance against worker injury in the event of accidental spills
or system leaks.
Figure 6.
Schematic of the Permeability Test Apparatus
After the cores were seated in the permeameters at low
pressure, the selected air pressure was applied to the fluid
chamber until stable permeability values were obtained
with the standard leachate. At this point, the pressure was
released and the permeameters were disassembled to per-
mit examination of the core for signs of swelling or de-
terioration of any kind. If the clay had expanded out of
its mold, the excess was removed with a straight edge
while trying not to smear the surface of the soil. All ma-
terial that had expanded out of the cores was oven dried
and weighed to estimate the percent swelling that had
taken place.
With the three soil types that had swollen, additional
standard leachate was passed to ensure that the permea-
bility was not affected by the excess soil removal. This
extra step was not necessary with the kaolinitic soil since
it experienced no swell after passage of the standard
leachate.
Next, the remaining standard leachate was removed
from the fluid chambers and replaced with the organic
fluids. After passage of the simulated primary leachates
(organic fluids), the permeameters were depressurized,
disassembled and the cores dissected to determine if struc-
tural changes had taken place in the compacted clay soils.
-------
REMEDIAL RESPONSE 227
8 :
10'
IOO
NONCALCAREOUS SMECTITE a
CALCAREOUS SMECTITE A
MIXED CATION KAOLINITE o
MIXED CATION ILLITE •
WATER (0.01 N CaSO4)
0.5 ttO 0.5 I.O I.5 2.0 2.5 3,0
PORE VOLUMES
Figure 7.
Permeability of the Four Clay Soils to Water (0.01N CaSo4)
II :
NONCALCAREOUS SMECTITE a
CALCAREOUS SMECTITE A
MIXED CATION KAOLINITE O
MIXED CATION ILLITE •
05
0.0
0.5 I.O l!5
PORE VOLUMES
2.0 2.5 3.0
NONCALCAREOUS SMECTITE
CALCAREOUS SMECTITE A
MIXED CATION KAOLINITE o
MIXED CATION ILLITE •
10'
0.5 00
OS 10 1.5 2.0 2.5 3.0
PORE VOLUMES
Figure 9.
Permeability and Breakthrough Curves of the Four
Clay Soils with Aniline
Permeability values for the clay soils have been plotted
against the cumulative pore volumes of the test fluids that
passed through the compacted cores. The volume of
"fluid out" is divided by the volume of the pore space in
a given core to obtain the fraction of a pore volume
passed at each permeability value. Recorded data values
are shown graphically in Figures 7 through 11.
In several of the figures representing primary leachate
treated cores, the increase in organic fluid as a percentage
of the incremental leachate volume ("fluid out") has
been depicted across the top of the permeability graphs.
For these breakthrough curves, the values for "fluid
out" were determined by one of two methods depending
on the fluid to be analyzed. Fluids that were insoluble
(xylene) and minimally soluble (aniline) were determined
simply by recording the volume of the organic and aque-
ous layers in the sample collection bottles. The only
miscible fluid for which determinations were made was
methanol. The percentage of methanol in water was de-
termined using a thermoconductivity gas chromatograph.
RESULTS AND DISCUSSION
Figure 8.
Permeability of the Four Clay Soils Treated with Acetic Acid
Permeabilities of the four compacted clay soils to the
standard leachate (0.01N CaSo4) are depicted in Figure 7.
-------
228 REMEDIAL RESPONSE
100
NONCALCAREOUS SMECTITE a
CALCAREOUS SMECTITE A
MIXED CATION KAOLINITE o
MIXED CATION I LUTE •
METHANOL <»*»
00
0.5 10 1.5
PORE VOLUMES
2.0
2.5
30
NONCALCAREOUS SMECTITE
CALCAREOUS SMECTITE A
MIXED CATION KAOLINITE 0
MIXED CATION ILLITE •
00
0.5 1.0 1.5
PORE VOLUMES
2.0 2.5 3.0)
Figure 10.
Permeability of the Four Clay Soils Treated with Methanol and
the Breakthrough Curve for the Mixed Cation Illite Soil
The vertical dashed line represents the point at which the
permeameters were depressurized for the final time prior
to placement of the organic fluids in the non-control
permeameters. There appears to have been little effect of
depressurization on the permeability of the control cores.
Permeability of the noncalcareous smectite and mixed
cation kaolinite soils were essentially constant during the
passage of approximately two pore volumes of the stan-
dard leachate. In contrast, the permeability of the calcar-
eous smectite decreased slowly while the permeability of
the mixed cation illite increased slowly.
Traditionally, the permeability testing of prospective
clay liners for hazardous waste landfills and surface im-
poundments, has used only standard aqueous leachates
(such as 0.01N CaSo4 or CaCL) as the permeant fluid. All
four of the clay soils used in this study, if only evaluated
by this traditional test, would qualify as adequate for lin-
ing hazardous waste disposal facilities on the basis of their
having permeabilities lower than 1 x 10"7 cm/sec.
All four clay soils permeated with acetic acid showed
initial decreases in permeability (Figure 8). However, there
was also a significant amount of soil piping occurring in
these cores as was shown by the presence of soil particles
both clinging to the inside walls of the outlet tubing and
Figure 11.
Permeability and Breakthrough Curves of the Four Clay Soils
Treated with Xylene
deposited on the bottom of the leachate collection bottles.
In addition, the leachate from these cores was usually
tinted (red, creamy or black) indicating that soil compon-
ents were dissolved by the acid. The initial decrease in
permeability is likely due to the partial dissolution and
subsequent migration of soil particles. These migrating
particle fragments could then lodge in the fluid conduct-
ing pores, thus decreasing the crossectional area available
for fluid flow.
Two of the soils treated with acetic acid (calcareous
smectite and mixed cation kaolinite) showed continuous
permeability decreases throughout the test period. Both
the noncalcareous smectite and the mixed cation illite
eventually began to rebound from the initial permeability
decreases, but not until passage of 39% and 62% of a pore
volume respectively. The permeability increases on both
of these soils were probably due to the progressive soil
piping that eventually cleared the initially clogged pores.
Permeabilities and breakthrough curves for the four clay
soils treated with aniline are given in Figure 9. While all
four clay soils showed significant permeability increases,
the calcareous smectite soil showed the least.
Both the noncalcareous smectite and mixed cation illite
exhibited early breakthrough of aniline with concurrent
-------
REMEDIAL RESPONSE 229
permeability increases. There was some indication that the
rate of permeability increase for these two clay soils was
leveling out just above the 1 x 10~7 cm/sec benchmark.
While the permeability of the mixed cation kaolinite soil
eventually exceeded 1 x 10"7 cm/sec, the breakthrough of
aniline and concurrent permeability increases took sig-
nificantly longer to occur. Only the calcareous smectite
clay maintained a permeability value below 1 x 10~7 cm/
sec.
Permeabilities and a breakthrough curve (mixed cation
illite) of four clay soils treated with methanol are given in
Figure 10. All four clay soils eventually attained
permeabilities substantially in excess of 1 x 10~7 cm/sec.
The permeability of the noncalcareous smectite increased
the fastest and reached the highest permeability value of
the four methanol treated cores. Examination of the meth-
anol treated soil cores revealed structural rearrangements
of the soil particles resulting in the formation of large
pores and cracks visible in the surface of the soils.
Methanol breakthrough was monitored on the mixed
cation illite soil. Figure 10 depicts the step by step con-
current increase in the permeability and percent methanol
in the "fluid out" for the illitic soil. Since no particle mi;
gration was detected in the methanol treated cores, soil
piping was ruled out as a mechanism for the observed
permeability increases. There was likewise no leveling of
the rate of permeability increases as might have been ex-
pected if the increases were solely due to the lower viscos-
ity and density of methanol compared to water.
Permeabilities and breakthrough curves of the four clay
soils treated with xylene are given in Figure 11. All four
xylene treated soils showed rapid permeability increases
followed by leveling at a permeability roughly two orders
of magnitude above their permeability rates of water.
These permeability increases are far greater than could be
explained in terms of the lower viscosity and density of
xylene compared to water, indicating that other factors
such as structural changes in the soil are responsible for
the large increases observed.
CONCLUSIONS AND RECOMMENDATIONS
The time needed for the comparative permeability test-
ing of clay soils can be significantly decreased by the use
of elevated hydraulic gradients. A hydraulic gradient of
between 50 and 100 seems optimal for such tests in that the
tests can be completed in 2-3 months, while permeability
increases occur slowly enough to obtain several points
along a changing permeability curve.
Three classes of organic fluids (basic, neutral polar and
neutral nonpolar) may cause substantial increases in the
permeability of clay liners. Organic acids may also degrade
the effectiveness of clay liners although not as quickly as
the other three organic fluid classes examined.
Consequently, permeability of a clay liner may be
effected by the primary leachate (free fluids) of a waste.
Whenever organic liquid bearing wastes are to be con-
tained in a landfill or surface impoundment by a clay liner,
the permeability of the liner should be evaluated with a
standard leachate (such as 0.01 N CaSo4), and also with
the primary leachate of a waste.
REFERENCES
1. U.S. EPA, "Report to Congress: Disposal of Haz-
ardous Waste", USEPA #SW. 115. Washington, D.C.
1974.
2. McDougall, W.J., Fusco, R.A., and Obrien, R.P.,
"Containment and Treatment of the Love Canal Land-
fill Leachate." Presented at the Annual Water Pollu-
tion Control Federation Meeting held Oct. 11, 1979.
3. Chian, E.S.K. and DeWalle, F.B., "Evaluation of
Leachate Treatment." Volume I and II. USEPA #EPA
600/2-77-186 a & b. Cincinnati, Ohio. 1977.
4. Walstenholme, R.M., "Disposal of Solvent Waste." In
the Second Solvent Symposium held by the University
of Manchester Institute of Science and Technology,
Manchester, England, 1977, 138.
5. Bennett, J.P., "Permeability of Soils at Elevated
Permanent Pressures." Master's Thesis at Colorado
State University, Fort Collins, Colorado. 1966.
6. Jones, C.W., "Permeability Tests with the Perma-
nent Water Under Pressure." Earth Laboratory Report
#EM-559. Division of Engineering Laboratories, Com-
missioner's Office, Denver, Colorado. 1960.
-------
MULTIATTRIBUTE DECISION MAKING FOR
REMEDIAL ACTION AT HAZARDOUS WASTE SITES
TERRY H. ESS
Hazardous Materials Control Research Institute
Silver Spring, Maryland
CHIA S. SHIH, Ph.D.
University of Texas at San Antonio
San Antonio, Texas
INTRODUCTION
Many modern technology related problems, especially
those involving man created risks, are extremely complex
and uncertain. Large volumes of data and multiple con-
flicting objectives lead to the requirement for incorporat-
ing subjective judgments into the decision process. If the
existence of adversary positions (i.e., a politicized con-
flict) is involved, then something other than just intuition
must be relied on to find generally acceptable solutions.
The problem area of hazardous waste management is just
such a situation. It would be beneficial to find a sys-
tematic technique which would allow decision makers to
adequately address the complex issues involved and de-
velop viable solutions.
Multiattribute decision analysis is an apparently ef-
fective tool for this type of problem. It is highly flexible,
incorporates methods to handle uncertainty, multiple ob-
jectives, etc. and is a well developed technique. However,
it is not without faults. The most glaring as pointed out
by Rowe(3) is its inability to properly treat the public's sub-
jective perception of risk. In addition, conventional de-
cision analysis can become extremely cumbersome if the
problem being analyzed is complex and no readily avail-
able means of tree "pruning" exists.
An obvious answer to this is to somehow integrate multi-
attribute decision analysis with a quantitative risk assess-
ment technique which addresses public perception. The
basic steps entailed in such a unified approach are deline-
ated below:
1. Construct a decision tree for the potential course of
remedial actions.
2. Complete a detailed risk analysis.
a. Determine the objective risk of each decision
branch. (Note: According to some authorities01
this would be termed modeled risk.)
b. Determine the appropriate risk referents.
c. Use an objective risk vs. risk referent comparison
to determine if a decision branch requires modi-
fication or should be eliminated from considera-
tion (i.e., pruned).
3. Conduct a sensitivity analysis of the risk comparison
to determine which branches are only "marginally"
acceptable.
4. Complete the multiattribute decision analysis with
the pruned tree.
5. Conduct a sensitivity analysis of the "solution.''
This paper will explore this integrated technique in more
detail.
DECISION TREE CONSTRUCTION
The organization of decision and consequence se-
quences is the first task in the decision making process.
Generally the following steps should be utilized to con-
struct a tree:
1. Generate an objective hierarchy which terminates in
the desired attributes and attribute measurements
(Figure 1 diagrams a possible hierarchy for hazard-
ous waste problems).
2. Determine the viable courses of action available.
3. Determine the possible chance events (i.e., hazards,
outcome, etc.) resulting from a decision.
4. Arrange the decision options and resulting chance
events in chronological order (Table I provides a
generalized structure for problems which largely
involve risk).
5. Evaluate the specific probabilities for each chance
event.
6. Evaluate the magnitude of each attribute.
A number of key areas in this process require a more
detailed explanation. First it is necessary to take a closer
look at the meaning of "viable courses of action." The
Protect biological
systems (especially
man) in the most ef-
ficent acceptable
manner
i
1
Acceptable protection Efficent manner
Human
r~
i
i
1
effects Non-human effects Colt
J
1$)
1
Lud ti
(yr)
Risk Aesthetics * induced anomalies
in indicator species
(animal & plant>
fatal- acres
ities set a-
etc. side for
waste
Figure 1.
Hazardous Waste Objective Hierarchy
230
-------
REMEDIAL RESPONSE 231
Table I.
Possible Risk Classification Scheme
Risk Description
Class of Consequence
Fatalities Morbidity Property
Damage
I mediate
Catastrophic
Involuntary
Regulated Voluntary
Ordinary
Involuntary
Regulated Voluntary
Delayed
Catastrophic
Involuntary
Regulated Voluntary
Ordianry
Involuntary
Regulated Voluntary
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Decision Hazard
Outcome Exposure
Conse-
quence
Attribute
Magnitude
\^) AJCOf
Figure 2.
Generalized Decision Tree Structure
term implies that some pre-decision tree criteria is used to
eliminate "non-viable" alternatives from entering into
the decision making process. In the hazardous waste area
the principal criteria to serve this purpose is implementa-
tion time. The discovery of an uncontrolled dump site
for highly toxic substances would require initial positive
action be implemented in a minimum of time. The num-
ber of viable options would probably be small. In the case
of planning a new controlled disposal/storage site this
constraint would be greatly reduced, therefore, allowing
for the consideration of a much wider scope of options.
A keener look is also required in the area of hazard
event space. A quick glance at Figure 2 would cause
some to think that what we are talking about is one
"success" event and a few (one or more) "failure"
event(s). This is of course far from the truth in anything
but the most simplistic cases. In most "real-life" prob-
lems the failure pathways shown correspond to the top
event of an appropriate fault tree. A fault tree is an-
other type of analytical tree which allows us to depict the
logical interrelationships between basic events that lead to
a undesired event (i.e. the failure event). Using fault
trees, Boolean algebra and various statistical techniques'6'
it is possible to determine failure event probabilities. In
many cases the probabilities determined are only order of
magnitude estimates. The success events in Figure 2 repre-
sent the summation of all possible success events for a
specific decision. In practical terms the probability of suc-
cess is equal to one less the sum of all the computed
failure event probabilities for a specific hazard chance
node.
This leads us right into the next area of decision tree
construction which requires elaboration, the significant
potential inaccuracies in both probability and attribute
assessments. As indicated in the prior paragraph even
when "objective" methods are used, the information ob-
tained may be only accurate within an order of magnitude.
When intuitive judgments are involved, which is often the
case in these assessments, then even more inaccuracies
can be expected. In fact, the level of accuracy could be so
low that a single "correct" answer cannot be determined.
This does not negate the value of using quantitative tech-
niques but should caution us to refrain from making un-
justifiable claims of accuracy. Because of this, it is almost
as important to obtain variance information about as-
sessments as it is to know their mean. This variance in-
formation becomes especially important in the subsequent
step of sensitivity analysis.
Table II.
Denney Farm Risk Data Summary
Alternative Joint Prob- Involuntary Reg. Vol.
ability Fatal. Morb. Fatal. Morb.
(Term)
1. Leave buried
2. Install & maintain a
groundwater monitor-
ing system
3. Excavate & store mater-
ial on site
.01
(long)
.9
(long}
3.3'10'^
(long)
.45
(long)
.2
(short)
3.2>10"5
(short)
.04
(long )
.1
(short)
2.5*10~Z
(short)
145 1301
12 107
38 341
12 107
5
7
12
45
60
36
4. Excavate s transport
liquids and residues
via truck to Syntax
same as 3 plus:
3. 5*10 "7
(short)
(short)
-------
232 REMEDIAL RESPONSE
Table ID.
Risk Referent Assumptions
Indirect gain-loss balance:
Class
[public)
Regulated Voluntary
( workers J
Alt
^
2
3/4
3/4
Balance
Marginally unfavorable
Indecisive
Marginally favorable
Favorable
Value
.001
.01
.1
1.0
Controllability:
Alt
1
2
3/4
Control
Approach
.1
.3
1.0
Degree of
Control
.1
.3
1.0
State of
Implement.
.5
1.0
Basis of
Effect.
.5
1.0
RISK ANALYSIS
Risk cannot be meaningfully analyzed in an aggregate/3'
It must be separated into classes which are commensurate
with the factors leading to subjective perception. A pos-
sible classification scheme for hazardous waste problems is
provided in Table III. This scheme is modified slightly
from that advocated by Rowe in order to specifically fit
the area of hazardous waste. The principal changes are the
deletion of the following risk classes: identifiable risk,
voluntary risk, natural risk and man-triggered risk. Since
generally acceptable solutions are our interest, not the
acceptability to specific groups, the identifiable classifica-
tion is not needed. Only the involuntary and regulated
voluntary classes are required since the worker's environ-
ment is regulated. Finally, since we can usually classify all
hazardous waste problems as man-originated, there is no
need for the other origination classes.
After agreeing on an appropriate classification scheme,
the objective (estimated) magnitude of risk of each class on
each decision tree must be determined. In general this
would be formulated in the following manner for each
decision branch when a decision tree format is used:
Where j a specific path in a decision branch
Pj - j oint probability along path j
Ai a specific risk class
a^ - the consequence magnitude of risk Ai on
path j (i.e. number of fatalities, etc.)
E total population exposed to risk Ai
T - time in years
R0 Objective risk magnitude
(Note: This provides a measure of risk commensurate with
the manner of data presentation used by Rowe(3).)
The next step is to determine appropriate risk referents.
The purpose of a risk referent is to serve as the measure of
risk acceptability (incorporating subjective perspectives).
Rowe has proposed a methodology of calculating risk
referents which is composed of two steps:
1. Using historical data as a base, absolute risk re-
ferences are determined for each risk class.
2. These absolute references are modified to fit the spe-
cific situation being analyzed producing risk re-
ferents.
A clearer understanding of the purpose of this process can
be gained by examining Figure 3. In this exhibit the risk
referent calculation steps are related to the transforma-
tion factors responsible for the subjective perception of
risk. It should be noted that Rowe's assessment of the ac-
curacy of this method indicated potential variances in ex-
cess of one order of magnitude. Prospect theory(5) provides
a formal explanation of the psychological rationale be-
hind the determination of risk referents and gives some
hope for future refinement of this process.
The final phase of the risk analysis can now be accom-
plished. This is nothing more complicated than a compari-
son of a decision branch's estimated risk (for each class of
risk) with the appropriate risk referent. If the risk does not
exceed the referent by more than one order of magnitude
then the risk is considered publicly acceptable. This order
of magnitude comparison is used due to the inherent in-
accuracies in the risk information. If this criteria cannot
be met then the choice remains to either modify the de-
cision branch (which will probably effect some or all of
the non-risk attributes being considered such as cost) or to
eliminate that branch. It should be noted that one of the
modification steps possible when the difference between
objective and subjective risk is large is to attempt to edu-
cate the public to the "actual" risk. In many cases this
may be a difficult, costly and time consuming path but in
some cases it may be the only option other than complete
abandonment of a project.
Factors involving type of consequence:
•Voluntary or involuntary (1)
•Discounting of time (1)
•Controllability (2)
Factors involving nature of consequence:
•Position in hierarchy of consequence (1)
•Ordinary of catastrophic (1)
Other factors:
•Magnitude of probability of occurrence
•Propensity for risk taking (2)
(1) Explicitly included in determination of absolute risk reference.
(2) Explicitly included in determination of risk referent.
FigureS.
Transformation Factors Utilization in Risk Referents
MULTIATTRIBUTE DECISION ANALYSIS
This process can be broken into three principal parts:
1) the determination of utility functions for each attri-
bute, 2) the development of a multiattribute utility func-
tion and 3) the calculation of expected utilities. The first
two are by far the most complicated. The determination of
an attribute's utility function is a well developed analyst-
decision maker interrogation procedure. It normally in-
cludes questioning both the qualitative and quantitative
-------
REMEDIAL RESPONSE 233
Table IV.
Risk Referent Calculation Factors
Risk Classification
ic, fatal
Involuntary, ordinary,
fatal
Involuntary, catastroph-
ic, health effect
Involuntary, ordinary,
health effect
Regulated voluntary,
ordinary, fatal
Regulated voluntary,
nrrtinarv. health r*f?»r*t-
Risk Ref
_7
S'lO-''
5-flO""7
3*10-=
ino-4
6*10-'
Risk
Prop
Fac.
1
.1
.1
.1
1.0
1.0
Proportion.
Derating
Factor
ait- i no i
A J.t j, . UU1
2 .01
Alt 1 .001
2 .01
3/4 .1
Alt 1 .001
2 .01
Alt 1 .001
2 .01
3/4 .1
1.0
1.0
Control.
Factor
Alt 1 . 01
Alt 2 .015
Alt 1 .01
2 .015
3/4 1.0
Alt 1 .01
2 .015
Alt 1 .01
2 .015
3/4 1.0
1.0
1.0
Note: All risks are treated as immediate
characteristics of the decision makers preferences. The re-
sult is a function which transforms attribute magnitudes
into a cardinal measurement, utility, structured so that the
best course of action is always the alternative with the
highest expected utility.
The development of a multiattribute utility function
basically involves the assessment of attribute independ-
ence and then the development of an appropriate mathe-
matical formulation. This formulation allows the analyst
to combine each attribute's utility function into a single
utility function. A simplified summary of the independ-
ence assumptions and resulting formulations in multiat-
tribute utility theory is provided in Figure 4. It should
be noted that the concept of preferential independence
involves attributes under the condition of certainty while
utility independence is specifically concerned with uncer-
tainty. This process is developed according to the decision
maker's perception of the attributes' relationships, not
according to some established standard rules. Some gen-
eral observations about the independence perceptions
likely to be held by most decision makers in hazardous
waste problems seems in order though. First, the risk at-
tributes normally will be both preferential and utility in-
dependent of the other attributes. Second, each risk at-
tribute would normally be both preferential and utility
independent of each other.
The third part of the procedure, the calculation of
expected utilities, is basically just a mechanical process of
"averaging out" and "folding back." Averaging out in-
volves nothing more than the computation of PjUj (p-
joint probability along decision path j; Uj-utility magni-
tude for the jth path) for a decision node. Folding back
entails the elimination of the less desirable paths at a de-
cision node. Unlike most of the other aspects of this inte-
grated analytical procedure, this portion of the process
is purely "objective."
SENSITIVITY ANALYSIS
Sensitivity analysis is used at two points in this inte-
grated procedure: 1) after the risk analysis and 2) after
completing the decision analysis. A good question at this
point is what is this sensitivity analysis about and why is
it necessary? As expressed previously, a significant level of
inaccuracy is inherent in the type of problems we are
dealing with. Sensitivity analysis is a post solution investi-
gation that indicates how much trust should be placed in
solutions when we know that all or most of our parameter
values (probabilities, attribute assessments, etc.) are not
certain. Therefore, sensitivity analysis has a very im-
portant part to play in the overall analysis. Using this
technique after risk analysis allows us to ascertain if any of
the "acceptable" decision branches is really only margin-
ally so. There would be a great deal of doubt if one of
these marginally acceptable branches was the solution or
a part of the solution chosen during decision analysis. In
that case we would probably be inclined to look at the
problem with even greater intensity. In addition this tech-
nique will help us identify options whichJiave no realistic
chance of being accepted. The same type of function is
provided by the sensitivity analysis conducted after de-
cision analysis.
Sensitivity analysis does not have any rigid rules about
what specific methods to use. Within the context of the
problems we are addressing, two possibilities present
themselves. The first is simply to change the value of
selected parameters and see if the "solution" changes.
A more useful approach involves the substitution of
variables for selected parameters (normally one at a time),
then solving the decision problem in terms of the vari-
able. By changing the variable we can graphically depict
the effect it has on the desirability of each possible
decision. This allows us to quickly determine which
parameters are significant and therefore require close
attention.
Independence Definitions:
Preferential Independence (PI)— attribute X is PI of attribute Y if
preference for consequence (x.y1) with y' held fixed do not depend on
the amount of y'.
Utility Independence (UI)— attribute X is UI of attribute Y if prefer-
ence for lotteries on (x,y') with y1 fixed do not depend on the amount
ofy'.
Additive Independence (AI) — attribute X and Y are AI preferences
for lotteries (x,y) depend only on the marginal probability distribu-
tions on x and y.
Utility Formulations:
IfXiUiXi.i = 1,2..., n then
u(x) = EkjUj(x) + kjjUjWUj) +...k, nnu, (x,)...un(xn)
If jx,xi| PIX
< > ™AU,i=2,3 .....
n and X^IX, then either
1)1+ ku(x) = III
2)u(x) = £kiui(Xj)
IfXiAiXi, i = 1,2,... ,n then
u(x) = Ekiui(xi)
Source: (1)
Figure 4.
Summary of Multiattribute Utility Theory
-------
234 REMEDIAL RESPONSE
APPLICATION
In order to illustrate the use of the integrated method-
ology, the case study of Denny Farm 1(1) will be used. In
this case an uncontrolled chemical dump site containing
TCDD along with other substances was discovered in
Missouri. The problem posed is how best to eliminate the
health risk at and around this site. Four viable alternatives
were suggested: (1) leave the site as is, (2) install and main-
tain a groundwater monitoring system, (3) excavate the
dump and restore the waste in a controlled manner on
the current site, and (4) excavate the site and transport
liquids and residues via truck to Syntex, an approved
hazardous waste storage site. Since joint probabilities for
each chance path have been delineated in the study, the
simplified decision tree shown in Figure 5 is used. Due to
the limited information provided in the report only four
attributes can be used: two classes of human risk, fatali-
ties and morbidity, cost and lead time. The division of
human impacts between fatalities and morbidity is made
possible by adopting the arbitrary assumption that ten
percent of all possible harmful human exposures de-
termined in the study will result in fatalities. In reality
this assumption would have to be verified and revised as
necessary.
The next step is to proceed with a risk analysis. The risk
data provided by the study is summarized in Table II. Use
of this data requires that it be expressed in terms commen-
surate with the risk referents that will be calculated. This
requires that a time duration in years for long term risk be
established and that estimates of the total population ex-
posed be determined. For the purposes of this paper the
long-term risk duration is assumed to be evenly distributed
over a span of 30 years. Estimates of the total popula-
tion exposed were calculated using data provided in the
study. In some but not all cases this could be reasonably
assumed to be equal to the maximum exposed numbers
derived in the study. The objective risk was then calculated
using the equation provided in the "Risk Analysis" sec-
tion of this paper. These risk estimates are summarized
in Table V. With objective risk determined, the next step
was to calculate risk referents. This was accomplished us-
ing the procedure outlined by Rowe.(3) In order to do this
some reasonable assumptions about the public's and
worker's perception of the indirect gain-loss balance and
the controllability of each alternative had to be made (see
Table III). A summarization of the factors used in the
risk referent calculation is shown in Table IV. A com-
parison of objective risk and risk referents is provided in
Table V. A quick glance at this comparison indicates that
none of the proposed alternatives is "acceptable" to the
public and that alternatives 3 and 4 are only marginally
"acceptable" to the workers in terms of fatalities.
Regretfully, no information was provided in the study
with which to conduct a sensitivity analysis of the above
risk assessment. Illustrating this procedure will require
that some reasonable variance information be postulated.
It will be assumed that:
•The parameters used in the decision tree, joint proba-
bilities and attribute magnitudes, can vary by ± 50%
Table V.
Risk Comparison
Alt
1
2
/4
/4
od
Risk classification
Involuntary, catastrophic, health
Involuntary, catastrophic, health
Involuntary, ordinary, health
Involuntary, ordinary, fatal
Involuntary, ordinary, health
Reg voluntary, ordinary, health
Involuntary, ordinary, fatal
Involuntary, ordinary, health
Reg. voluntary, ordinary, fatal
Reg. voluntary, ordinary, health
Objective
Risk
3.3.KT"
3.0<1
7.5'10">-5.0»10'1
7.5«10°-5.0.10J'
7.6'10"'-3.3 10"'
4.5xlo"°-4.0'10'"'
3.»-io"-2.e.ifl"
7.9.10""-5.2-10"J
8. 3vlO"J-7.5'10'Z
1.0-10'°-9.0»10"5
9.0.10''-6.0«10"'
Same as 3
Risk Referent
Range
1.0.1flH-l.
s.o.io"'1-?.
s-O'id'^-s.
3.0«-lo"l-3.
5.0'lo"'-5.
3.0»lo"*-3.
1.0,H>'-1.
6.0«10"3- .
Sane as
°'l°"
o 10"
0 10'
o.io"1
OrtO"
ono"4
one"
06
3
Reg. voluntary, ordinary,
health
within the limits imposed by the exposed population size
and the maximum possible probability of 1.0.
•The risk referents can vary by ± one order of magnitude.
Using these assumptions, the information calculated In
Table VI indicates the possible range of variance in both
objective risk estimates and risk referents. This provides a
sound basis for assessing which alternatives have a realis-
tic chance of being acceptable.
At this point it is fairly clear that alternatives 1 and 2
will under no foreseeable circumstance approach an accep-
table level of risk so they can be eliminated. It appears
potentially feasible and worthwhile to modify alternatives
3 and 4 by eliminating the risk to the public of signifi-
cant release of residual TCDD after excavation by either
eliminating most of this residual during the clean-up or by
-------
REMEDIAL RESPONSE 235
Human Risk
Fatal Morbid.
Cost Lead Time
(10 $) (mo)
Human Risk Cost Lead Time
Fatal Morbid. (10 S) (io>)
0
3
45
60
36
108
2
9
0
0
0
.75
.75
.75
3.5
3.5
3.5
3.5
3.5
3.5
3.5
3.5
0
0
5
7
4
12
0
3
45
60
36
108
3.0
3.0
3.0
3.0
3.0
3.0
8
8
8
8
8
8
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
0
3
45
36
108
0
3
45
36
108
2
9
3.0
3.0
3.0
3.0
3.0
3.5
3.5
3.5
3.5
3.5
3.5
3.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
Figure 6.
Revised Decision Tree
FigureS.
Initial Decision Tree
some form of encapsulation. This modification will
cause both the cost and lead time of each alternative to
be increased. No other risk modifications seem realistic
with the information given. As indicated in Tables V and
VI, even with the suggested modification alternatives 3 and
4 would still not be acceptable. This would suggest that it
is time to "go back to the drawing board" and assemble
some other alternatives. Possible other approaches
would need to include means to either mitigate or eliminate
the public exposure due to tornados and contaminated
workers. Since no in-depth information is provided by the
study for other approaches or modifications, the heroic
assumption that alternatives 3 and 4 (modified) are the
best that can be offered must be made to proceed with
the analysis. In real life this is a possible outcome; in
which case someone is stuck in the unenviable position of
explaining this to the public.
With the risk assessment completed a revised decision
tree including only alternatives 3 and 4 (modified) is all
that is necessary (see Figure 6). Completion of the analysis
requires utility functions for each individual attribute and
a multiattribute utility function. The utility functions used
for each of the attributes are diagramed in Figures 7-10.
Figures 7 and 8, Fatalities and Morbidity, provide an ex-
ample of risk neutral functions while Figure 9, Costs, is
0)
O
C
01
^ CP
-------
236 REMEDIAL RESPONSE
tl U
U «
o. a
1.0-
.6-
.4-
30
60
Morbidity
90
120
.2-
Lead Time (months)
Figure 8.
Morbidity Utility Function
Figure 10.
Lead Time Utility Function
a u
u n
a, a
1.0-
.8-
.6-
.4-
.2-
Ruman Risk Cost Lead Total
Fatal Morbid. (10 S) Time
(no)
Relative Heights .38
1
.12
1
1.0 2.0 3.0 4.0
Cost ($ Millions)
Figure 9.
Cost Utility Function
data. In order to combine the four separate utilities into a
single utility value a multiattribute utility function is
necessary. As in the case of the individual utility functions
above, this function is directly related to the perceptions of
the decision maker in question. For the purposes of this
paper the use of an additive function seems reasonable
due to the general observations mentioned previously and
the small relative difference between the two alternatives
being considered.
Finally expected utility for each path is calculated us-
ing the mechanics of decision analysis to arrive at a "solu-
tion." The decision tree with all calculated values is dia-
gramed in Figure 11. Alternative 3 (modified) is obviously
preferable to 4 (modified) in this particular case. Since
both alternatives are so similar it is unlikely that a sensi-
tivity analysis would show this to change within any
reasonable limits.
CONCLUSIONS:
The proposed methodology provides a quantitative tool
that is systematic but flexible; is capable of handling un-
certainty, multiple conflicting objectives and the subjec-
.25 .25 1.0
.3 .1 .6
.3 .1 .6
.3 .1 .39
.3 .1 .44
.3 .1 .10
.92 .92
0 .S
0 .5
0 .29
0 .34
0 0
0 .SO
0 .46
Figure 11.
Problem Solution
live judgments of decision makers; and addresses the
highly subjective nature of public risk perception. With
the prudent use of a "viable option" criteria and risk
analysis, the number of options that must be fully con-
sidered can be effectively limited. On the other hand,
this approach will help pinpoint requirements for consid-
ering a wider scope of alternatives when necessary. The
process is obviously powerful and, therefore, inappropri-
ate for use with single attribute problems. Given the
potential complexities inherent in hazardous waste prob-
-------
REMEDIAL RESPONSE 237
lems the use of this technique appears justified. This
method of problem solving does not try to eliminate the
use of subjective value judgments, an impossible and
counterproductive task, but does require these judgments
to be openly scrutinized.
Finally, it should be obvious that this methodology is
by no means perfect. It is an initial attempt to bring to-
gether a number of powerful quantitative tools and apply
them to the area of hazardous waste. A great deal of re-
finement and improvement is still required in this area.
REFERENCES
1. Buchanan, J. et al., "Technical Study and Remedial
Action for Denney Farm Site 1, Auroria, Mo. (Final
Report)." Ecology and Environment, Inc., 1980.
2. Keeney, R. & Raiffa, H., Decisions with Multiple Ob-
jectives: Preferences and Value Tradeoffs, John Wiley
& Sons, 1976.
3. Rowe, W., An Anatomy of Risk, John Wiley & Sons,
1977.
4. Shih, C., "Decision Analysis and Utility Theory,"
Seminar on Risk and Safety Assessment, HMCRI,
1981.
5. Tversky, A. and Kahneman, D., "The Framing of De-
cisions and the Psychology of Choice," Science, Jan-
uary 30, 1981.
6. Vesely, W. et al., Fault Tree Handbook, U.S. Nuclear
Regulatory Commission, NUREG-0492,1980.
-------
RISK ASSESSMENT NEAR UNCONTROLLED HAZARDOUS
WASTE SITES: ROLE OF MONITORING DATA
GLENN E. SCHWEITZER, DIRECTOR
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
Las Vegas, Nevada
PREVIOUS EXPERIENCE WITH RISK ASSESSMENT
Many assessments of the risks associated with human
exposure to industrial chemicals have been carried out dur-
ing the last decade. Most studies have been conducted
by Government agencies, particularly OSHA and EPA or
by industry in response to regulatory requirements. Pub-
lic concern over exposures to carcinogens, in particular,
has been a significant stimulant in encouraging such as-
sessments.
Risk assessments have considered a number of technical
approaches to several types of exposure scenarios. For
example, clinical and epidemiological studies to investi-
gate human health effects as the pivotal concern in risk
assessments have been triggered by reports of worker ex-
posures in certain manufacturing facilities. Modelling
techniques, using limited data on chemical properties,
have been employed to predict future risks that might re-
sult from the manufacture and use of newly developed
chemicals. Finally, materials balance studies have been
important in estimating environmental discharges from
certain production processes and certain uses as a basis for
risk assessments of industrial chemicals such as vinyl
chloride and formaldehyde. While monitoring data have
often been used in such assessments, only on a few occa-
sions were the monitoring data collected in a systematic
manner designed to support authoritative risk assess-
ments.
The most widely used methodologies for assessing chem-
ical risks have emphasized a chemical-by-chemical analy-
sis. One approach has been to estimate the toxicity of
the chemical as a function of dose, usually drawing on
the results of laboratory experiments. A second approach
relies on direct observations of the biological effects of
the chemical exposure, effects either on human popula-
tions or more frequently on biological surrogate popula-
tions. In either approach the dose levels or the effects to
critical receptor populations are estimated, with the risk
then determined to be the probability that a person with-
in the exposed population of concern will suffer an ad-
verse effect from the likely exposure.
The uncertainties in these methodologies as illustrated
in Figure 1 are well known. They include the problems
on extrapolating effects from laboratory animals to man,
in determining the critical receptor populations and their
activity patterns, and in estimating dose levels. Never-
theless, these approaches have been used to support many
regulatory decisions.
NEW FACTORS TO BE CONSIDERED
AT UNCONTROLLED SITES
While past experience provides a point of departure for
developing methodologies appropriate for assessing risks
near hazardous waste sites, and specifically uncontrolled
sites, different constraints and uncertainties must be
addressed. Many types of chemical mixtures are involved,
including mixtures of unknown composition. Thus, the
chemical-by-chemical approach is not sufficient to esti-
mate risks. Also, risks to future generations may over-
shadow the nearer-term risks that have often dominated
past risk assessments. This longer time horizon requires
greater attention to the chemical loadings of environ-
mental compartments which serve as temporary pollutant
reservoirs such as ground water, sediments, and soil. Final-
ly, containment uncertainties require the taking into ac-
count of unanticipated future chemical leakages from the
site which over time could dominate concerns over risks.
Risk assessment methodologies must be flexible in order
to accommodate a wide variety of site-specific charac-
teristics. They should be adaptable to varying time limita-
tions for carrying out assessments. This time span could
range from a few months (which limits sampling to one
time) to several years (which permits phased sampling and
also seasonal sampling to take into account changes in
weather and hydrological conditions) depending on both
the urgency of the problem and the public perceptions of
the urgency. With regard to budget constraints, different
approaches may be required depending on the number of
samples that can be processed.
The starting point for any risk assessment should be the
purpose of the assessment. The purpose should shape the
application of the methodology and the rigor of the assess-
ment. In general, risk assessments are intended to clarify
the degree of hazard and thereby help determine: (1) the
priority for conducting detailed environmental evaluations
and for undertaking preventive or remedial action or (2)
the extent and character of the preventive or remedial
action. The former objective usually requires a less exten-
sive assessment. Often intertwined with these objectives
are: (1) liability and related issues raised at legal proceed-
ings and (2) concerns over the immediate hazard to area
238
-------
REMEDIAL RESPONSE 239
residents. Indeed, risk assessment will often be used by
several parties for somewhat different purposes, and there
will frequently be a blue in distinguishing between assess-
ments to set priorities and assessments to determine the
magnitude of remedial action.
ADAPTING THE
CHEMICAL-BY-CHEMICAL APPROACH
The assessment of the risks resulting from exposures to
individual chemicals will undoubtedly continue to be an
important consideration in assessing total risks near haz-
ardous waste sites. Two key issues, however, are the selec-
tion of the chemicals of greatest concern and the aggrega-
tion of the risks posed by individual chemicals.
Several factors are important in selecting the chemicals
for individual risk assessment. For example, certain chem-
icals may have been deposited at the site in very large
quantities and therefore they are of principal concern
simply because of their volume. Secondly, one or more of
the chemicals known to be present at the site may be so
toxic that the threat even at low volumes is obviously a
major concern (e.g., dioxin). Certain "indicator" chem-
icals commonly found at waste sites may behave in the en-
vironment in a manner characteristic of many other chem-
icals as well and determining the environmental distribu-
tion of such "indicator" chemicals would be indicative of
broader contamination problems. Analysis of leachate
from the edge of the site or preliminary monitoring close
to the site might identify chemicals that are escaping from
the site. Finally, sampling for a group of chemicals which
are usually analyzed as a package such as the 129 priority
pollutants might result in considerable savings per chem-
ical in analytical costs. All of these factors should have a
bearing on the selection of the chemicals for intensive in-
vestigation.
After the chemicals of interest are determined, the classi-
cal approach of estimating source strengths, pathway load-
ings, location of critical receptor populations, and finally
exposure levels can be pursued (Figure 2). In this regard,
pathway loadings, including movement of the chemicals
from one media to another, are of considerable impor-
tance in estimating both current exposures and exposures
Identify Critical Pollutants
Determine Source Strengths
Identify Critical Receptors
Determine Receptor Locations
Identify Critical Pathways
Determine Pathway Loadings
Determine Contamination Patterns
Overlay Population Activity Patterns
Estimate Exposure Levels
Figure 2.
Classical Approach to Estimating Exposure Levels
RISK
t.
Extrapolations
Population at Risk
Sensitivity/Activity Patterns
Estimate
of
Dose / Estimate
of
Effects
Measurements
of chemicals in
micro-environ-
ment of human
receptors
Measurements
of chemical
accumulation in
biological surro-
gates for man
Human body4 ^Medical
burden exams
measurements
Bioassays of
biological sur-
rogates for man
Epidemiological
studies
t.
Figure 1.
Extrapolations Required in Determining Risk
-------
240
REMEDIAL RESPONSE
Direct Human Exposure:
Via Dermal Contact
& Inhalation
Indirect Human Exposure:
Deposition on Crops &
Ingestion-Bioaccumulation
in Grazing Animals
Gaseous &
Airborne Particles
Surface Runoff
Human Exposure:
Recreation Contact
& Water Supply
Recharge
to
Stream
Downward Movement
in Soil Moisture
Figure3.
Environmental Pathways from a Waste Site
of future generations. In addition to the natural pathways
of air, water, soil and groundwater (Figure 3), man-made
pathways such as sewers, drainage ditches and dump
trucks can be important.
In most cases, monitoring is the only reliable approach
to pathway analysis. However, frequently models (e.g.,
air pollution dispersion, hydrological and aquatic fate
models) and analyses of the physical and other proper-
ties of the chemicals can assist in determining the best loca-
tions and time intervals for monitoring activities. Also,
models can often provide a framework for snythesizing
and analyzing monitoring data.
In recent years, the kinetic analysis model has become
increasingly popular for providing a general framework
for multi-media monitoring programs. This model empha-
sizes the determination and use of inter-media transfer co-
efficients as a basis for estimating how specific chemicals
partition among different environmental compartments.
The interpretation of monitoring data as an indicator of
exposure of subpopulations near waste sites to specific
chemicals is more of an art than a science. Exposure levels,
now and in the near term via air, food and drinking water
and to a lesser degree through recreational uses of water
and through soil contact, are usually of priority con-
cern. Determining such exposure levels requires innova-
tive approaches to matching the multi-media contamina-
tion patterns in the area with the habits and movement
of resident populations. Even greater innovation is re-
quired with regard to future exposures, for the trends in
both pathway loadings and population patterns must be es-
timated.
With the exposure patterns for individual chemicals in
hand, estimates can be made of the dose that will reach
the population. A variety of coefficients are needed, how-
ever, to convert chemical concentrations in the micro-en-
vironment of the receptor into effective dose. Inhalation,
ingestion, and absorption coefficients in particular are
needed for each chemical of interest. Finally, after de-
termining the dose and the toxicity properties of the chem-
icals, estimates of the health effects can be made.
Methods do not now exist for aggregating the health
effects of individual chemicals that reach human popula-
tions into valid causal health effects descriptions. The
presence of one or more highly toxic chemicals might total-
ly dominate, although such a case will probably be the ex-
ception. If several chemicals of comparable potency reach
the receptor at comparable dose levels, an additive ap-
proach may be the only course although there is little
scientific basis for judging whether simple addition will
overestimate or underestimate the effects.
Approaches to determining the toxicity of complex mis-
tures have been a subject of research for several years in
-------
REMEDIAL RESPONSE 241
relation to discharges of industrial wastewater. Some of
the existing mutagenic screening tests and aquatic bio-
assay procedures seem appropriate for classifying the rela-
tive hazard of specific waste samples, particularly in rela-
tion to safe handling in the field or in the laboratory. How-
ever, comparable approaches to estimating risks to a gen-
eral population that is being subjected to a large num-
ber of chemicals are not available.
AN APPROACH TO
COMPARATIVE CONTAMINATION LEVELS
Given the many uncertainties in attempting to assess
risks near hazardous waste sites on a chemical-by-chem-
ical basis, a complementary approach of comparing rela-
tive contamination levels can often be very helpful. The
approach is intended to give a degree of perspective to
the chemical contamination that is present in an area of
environmental concern. The objective is to provide a basis
for determining relative risk although little insight is pro-
vided as to absolute risk. Two approaches employing
different types of "controls" are suggested below.
A control area with characteristics similar to the area of
environmental concern is highly desirable. Of special in-
terest would be a control area near the waste site—but in-
sulated by geographic features from the direct influence of
the site. This area would be impacted by all of the indus-
trial emissions and effluents that permeate the region as
well as other common sources of contamination such as
I Heavily Populated Areas
Sampling sites on Grand Island are
not shown on this map.
• Air ^ Drinking Water
OSoil • Surface Water &
• Sewer Stream Sediment
(7T3 Biota
Scale in Mile)
Figure 4.
Control Areas at Love Canal
agricultural chemicals. Thus, in comparing the contam-
ination near the site with the contamination in the con-
trol area, it should be possible to attribute any higher
levels found near the site to the influence of the site itself
and not to the background characteristics of the region.
The selection of control areas during investigations of
Love Canal is shown in Figure 4.
A second type of control can be provided by national
baseline data. Such data indicating the levels of ambient
or background contamination usually encountered in
different types of demographic settings can help clarify
the significance of environmental measurements near a
hazardous waste site. In Table I, an initial attempt is
made to provide such comparative data. However, several
problems were encountered in developing these data, in-
cluding:
•Many environmental measurements reported in the litera-
ture are not based on sound quality assurance programs
and therefore may be unreliable. Data of "research"
quality, a relatively high quality level are presented in
the table. Background data of this quality are in short
supply.
•Many measurements are made in suspected hot spot areas
of a highly localized character. Care must be taken to pre-
sent only ambient data which are characteristic of a
general area. Ambient data which are characteristic of
several types of urban areas are given in the table.
•Data are available on only a few chemicals in most media.
For some media, such as sewer sediments, soils and in-
door air, little representative data are readily available.
Comparing data from different sampling programs in-
volves considerable uncertainty. For example, seldom are
the analytical precision and accuracy of the data reported
and even less often is the representativeness of the
sampling approach known. Even when comparing two
areas using data from the same data set questions con-
cerning comparability of data can arise.
Love Canal investigations involved comparisons of data
from four areas:
(1) the canal itself,
(2) the adjacent Declaration Area,
(3) a nearby study area and
(4) the Control Areas.
For soil and sediment samples, the precision and accuracy
of each data point were considered to be about a factor
of three. Thus, if less than an order of magnitude sep-
arated two numbers, these two measurements might be
considered to be essentially the same. When differences
exceeded an order of magnitude, however, the measure-
ments probably indicated significant differences in pollu-
tant levels.
When comparing the sets of data collected in different
areas, statistical techniques play a significant role. How-
ever, uncertainties in interpretation techniques remain.
Should comparisons be made between maximum values,
mean values or some other numbers? Is the frequency of
occurrence of a chemical the significant aspect for com-
parison or is the frequency of occurrence overshadowed
by major differences in levels? How are trace quantities to
be handled? Finally, even this data set, which is much
-------
242 REMEDIAL RESPONSE
Table I.
Comparative Nation-Wide Monitoring Data
in Commercial, Industrial and Chemical Cities
Volatile Organics Measured in Air and Water
Air (ug/m3)
Comm Ind
Chem
Surface Water (pg/l)
Comm. Ind. Chem.
Chloroform
1.2-Dichloroethane
1,1.1 -Trichloroethane
Carbon Tetrachlonde
Bromodichloromethane
Trichloroethene
Dibromochloromethane
Bromoform
Tetrachloroethene
Chlorobenzene
(1)
ND-0 44
0.3
ND-1 07
ND-0 09
T-7.8
.02-1.1
T-1.9
0.98-2.5
ND-0 11 097-2.96
(2)
1-43
4-7
4-8
1/87
1-9
1-8
Drinking Water (
Comm. Ind. Chem.
(2)
0.4-311 4-93 0.6-86
T T-0.4 T-6
0.9-29 0.8-28 T-16
T-16 T-17 T-5
2-3 T-1
ND-0 066
ND-0 03
0.02-1.8
ND-0 45
Research quality data If only one value shown, only one detectable sample was reported.
ND = Nol delected, T - Trace
(I (Range of mean values. (2) Range of single samples
•6-8 cities per category
1 25 Inch
Diameter
6
Feet
Deep
Cores for chemicals that can evaporate quickly
' (2 separate samples)
, Cores for other chemicals
(5 cores which are mixed together into one sample)
Figure 5.
An Approach to Obtaining
Representative Soil Samples
larger than the data sets usually available, could be crit-
icized as being too small to allow authoritative conclu-
sions.
When comparing levels between sampling sites, care is
needed to insure that the contamination of the samples
taken at each site are representative of the overall con-
tamination at the site. This problem is particularly acute
when sampling soil and sediments. One approach to ob-
taining a representative sample at a soil sampling site is
shown in Figure 5. This approach is useful if there is a
likelihood that the contamination extends to significant
soil depths. Obviously, if only surface contamination is of
interest, a different approach would be appropriate.
Similarly, in comparing levels between sampling areas
involving multiple sites, the sample sets should be repre-
sentative of the areas as noted above. Ideally, the num-
ber of samples taken in each area of interest should be
about the same. However, given limited budgets, fewer
samples will undoubtedly be taken in control areas. A sta-
tistical approach to determining the minimum number of
samples to be taken within an area is possible if the stand-
ard deviation of the individual samples can be estimated
in advance. When the standard deviation cannot be es-
timated in advance, a preliminary sampling program to es-
tablish an estimate of this parameter can be particularly
useful.
If the foregoing concerns are adequately taken into ac-
count in the design of the monitoring program, realistic
comparisons of levels can be very useful as a basis for es-
timating relative risks. This approach can be of partic-
ular value in determining the most important environ-
mental pathways. Indeed, the uncertainties associates with
this approach may be of less magnitude than the extrap-
olation uncertainties usually encountered in the chemical-
by-chemical approach described above.
COUPLING MONITORING DATA
WITH POPULATION PATTERNS
The tendency in designing monitoring systems is to focus
sampling in the immediate areas of suspected contamina-
tion and in areas near population concentrations. In a
simple geographical situation involving a single site and a
-------
REMEDIAL RESPONSE 243
Urban Area
cc
Urban Subarea
Industrial Complex
Waste Site
i
Suspected Hot Spot
1. Grid entire area (ABCD).
2. Determine sources, likely hot spots, and populations in each square.
3. Obtain monitoring data for each square. Emphasize squares with high populations and
those with likely hot spots. Determine average concentration in each square.
4. Determine squares with highest level of:
Average Concentration x Persons Hours per day in Square.
5. Subgrid squares identified in 4. (WXYZ).
6. Repeat 2,3,4 for subsquares.
Figure 6.
Linking Monitoring Data and Population Distribution
relatively small adjacent population, this approach seems
adequate. However, frequently waste sites will be inter-
mingled with other sources of chemical pollutants. More
sophisticated approaches may be necessary to design the
most cost-effective monitoring system that adequately
characterizes the area while also providing insights as to
which sources are the principal contributors to contamina-
tion levels of concern.
A systematic approach based on a geographical grid is
depicted in Figure 6. The concept calls for determining
the sources, the population characteristics and the con-
centration levels in each media of interest in each grid
square. Squares of particular concern—due to suspected
hot spots or high population densities—are further sub-
divided, and each subdivision is then investigated in more
detail.
This approach, of relating all parameters of interest to
a fixed grid, has special appeal when analyzing the effec-
tiveness of proposed remedial action. Computer models
can help test the sensitivity of exposure levels in each
square to specific source reduction scenarios.
Certain population groups may be of greater concern
than others. Some groups may be more susceptible biolog-
ically to adverse chemical effects, or their lifestyles may
increase their exposures to chemicals. In Figure 1, some
of the population characteristics that should be taken into
account in detailed population studies are listed.
Types of Exposed Populations Near Uncontrolled Sites
Population Characteristics
J General Populations / Special Populatic
Distance to Sue
Distance to natural or
man-made pathways frc
Consi
uildmg Characteristic
Use of con;
material frc
Air circulation
Figure 1.
Population Characteristics
INVESTIGATING A WASTE SITE
Risk assessment data requirements should be a prin-
cipal factor in developing strategies for investigating a
specific hazardous waste site. A combination of the chem-
ical-by-chemical approach and a comparative assessment
of general contamination of the area appears to be a prac-
tical method for addressing risks at the types of sites most
commonly encountered. For large or complicated geo-
graphical areas, the sampling plan should be tied to a geo-
graphic grid to facilitate analyses of contributing sources
and of exposure problems. Four activities used to imple-
ment such an approach are shown in Figure 8 and dis-
cussed below. They are:
•An analysis of current and archival aerial photography.
Photo interpretation can be useful in identifying and
-------
244 REMEDIAL RESPONSE
Aerial Imagery
Geophysical Investigation
Ground
Water
Pollution
Plume
rj'JTT . O 0s TH 7^-~ry P "v ^*T—"^ "0*0 v^-cr-
Zoning Board-
County Health Dept,
On-Site Observations
Company
Headquarters
Records
FigureS.
Techniques for Targetting Monitoring Activities
OLANDFILL/DUMP
D K A S T F.
Figure 9.
Use of Aerial Imagery to
Identify Possible Waste Sites
-------
REMEDIAL RESPONSE 245
Figure 10.
Use of Aerial Imagery to Delineate Problems Near Waste Site
characterizing the site, pinpointing containment breaches
and environmental pathways away from the site, and
delineating nearby populations and valuable natural re-
sources (see Figures 9 and 10).
•Geophysical surveys. Using a combination of ground
based remote sensing devices (e.g., seismometers, mag-
netometers, radar, resistivity instruments), insights can
be gained concerning the boundaries of the site, densities
within the site, and anomalies in soil and water patterns
around the site. In the future it may be possible to de-
lineate subsurface plume patterns in certain types of
terrain.
•A review of previously compiled Government and in-
dustry reports. Most major sites have been investigated
by State or local Government agencies and, in addition,
industry has some records for many sites. Of particular
importance are records on the content of the site and sus-
pected containment problems.
•On-site inspection. Even cursory on-site inspection of the
site and of the surrounding area can provide insights as
to possible problems that are not easily obtained from
examination of records and photographs. Aerial photog-
raphy can be helpful in guiding an on-site inspection.
Within this framework, a monitoring program can be
effectively designed to provide data needed for risk assess-
ment. The program should probably concentrate on the
presence and movement of a limited number of specific
chemicals and also screen for a larger set of chemicals in
a variety of media. Certain environmental pathways of
concern (e.g., locally grown food and drinking water con-
tamination) may be obvious; others will probably warrant
exploratory monitoring. In any event, a preliminary mon-
itoring effort should help limit a more intensive moni-
toring effort so that within a limited budget a sufficient
number of samples can be taken in the media and in the
geographical areas of most concern to establish statistical
credibility.
Political pressures to carry out a hurried program should
be resisted and promises to deliver early results avoided.
Four practical considerations argue against tight time con-
straints. First, a phased approach to sound and cost-ef-
fective monitoring is recommended. Secondly, monitor-
ing during different seasons is usually important to take
into account weather and runoff conditions. Third, tech-
nical problems inevitably arise in the taking and analyz-
ing of samples with attendant delays of days and some-
times weeks or months. Finally, data formatting and in-
terpretation almost always take much longer than antic-
ipated.
There is no standard procedure for presenting monitor-
ing data for risk assessment purposes. Both the site char-
acteristics and the specific discoveries at the site will prob-
ably dictate somewhat unique formats that present the
most information in an understandable manner. The re-
sults will seldom be clear cut, and adequate time should
be allowed for debate prior to settling on an interpreta-
tion.
BIOLOGICAL MONITORING—
THE TREND OF THE FUTURE?
The problems in extrapolating from exposure monitor-
ing data to effective dose and then to human health
effects have been noted above. Biological monitoring
addresses some of these problems. In particular, this ap-
proach attempts to reduce the uncertainties concerning
the biological availability of chemicals in the environ-
ment and the integration of exposures to chemicals
through multiple environmental pathways. It provides a
method of predicting human health effects while also
directly measuring environmental effects.
As indicated in Figure 11, one type of biological mon-
itoring is simply to measure the chemicals that accumu-
late in species indigenous to the local area. This type of
program that was undertaken at Love Canal is shown in
Figure 12. The species selected for the program are com-
monly found in many areas of the country. The chem-
icals of principal interest were those chemicals that were
known to have been deposited in the Canal. The approach
did not attempt to discriminate between the original chem-
icals that were taken up by the biota and the metabolites
of these chemicals.
Biomonitoring was an additional dimension to increase
the sensitivity of the overall monitoring program at Love
Canal. It was hypothesized that chemicals not detected at
significant levels in air, water or soil might accumulate to
higher levels in biota because of the multiple routes of
-------
246 REMEDIAL RESPONSE
Bioaccum ulation
Whole Body
Selected Organs and
Tissues
Body Fluids
Btoeffects
Biosystem Responses
Pathological Changes
Physical Condition
Figure 11.
Biomonitoring Using Selected Animal Species
Field Mice:
Carcass
Hair
Dogs
Crayfish
Earthworms
Silver Maple
6
2
6
8
35
12
' 20
33
21
14
32
13
15
10
5
14
Organics
Inorganics
Inorganics in
Hair
Organics in
Whole Body
Organics in
Whole Body
Inorganics in
Leaves
Figure 12.
Biological Monitoring Program at Love Canal
exposure. At the same time it was recognized that re-
liable uptake coefficients would be needed for both the
chemicals that were found and for some that were not
found before sound conclusions could be reached.
A second type of biological monitoring is the measure-
ment of biological responses to chemical contaminants
using either indigenous biological species or species intro-
duced into the area of concern. There are many research
reports indicating that certain species respond in various
and consistent ways to different types of chemical ex-
posures. However, operational use of this technique will
require baseline data on response/exposure correlations
developed in controlled laboratory experiments. In the ab-
sence of extensive filed on the correlations of such re-
sponses with different types of chemical exposures, such
responses cannot be interpreted in a meaningful way.
Finally, the most direct approach of all may be medical
investigations and human surveillance to identify possible
health impacts on nearby populations. A variety of tech-
niques have been tried in other types of environments ex-
posed to air pollutants, food contaminants, and pesticide
applications. These techniques have ranged from routine
chemical analyses of blood, urine and breath to investiga-
tions of impacts on responses of the nervous and immuno-
logical systems.
Biological monitoring techniques are o'nly now begin-
ning to emerge from the research stage. Conceptually,
these direct measures of chemical-biological interactions
can provide persuasive evidence of the problems or lack of
problems near hazardous waste sites. It seems likely that
during the next few years they will be increasingly em-
ployed near sites. In the near future, however, they will
complement and not replace our traditional reliance on
conventional monitoring.
PRACTICAL ASPECTS OF
IMPLEMENTING A MONITORING PROGRAM
In conclusion, the monitoring data will only be useful
if the detailed aspects of collecting and analyzing the data
are sound. Several practical suggestions in this regard in-
clude:
•At the outset the objectives of the monitoring program
should be clearly defined. The users of the monitoring
data should be satisfied that the data will be collected,
analyzed, and presented in a manner which is respon-
sive to their needs.
•Monitoring programs can be helpful both in identifying
hot spots of near term concern and in assessing the
longer term habitability of the area. Each of these ob-
jectives may require a different program orientation,
however, as shown in Figure 13. Usually both objectives
will be important in varying degrees, and the program
emphasis should be adjusted accordingly.
•The entire array of monitoring opportunities should be
reviewed at the outset. Many of the pathways that might
be considered, particularly in situations where sites are
Assessing Exposures Near Uncontrolled Sites
Objective
Chemicals of concern
Levels of concern
Likely sources
Pathways of primary
concern
Populations of concern
Assessment emphasis
Monitoring approach
Awning
Hal Spoil
Identify near term
exposure problems
High toxicity chemicals
suspected to be in area
"High" levels
Past and current leakages
at site
Air, drinking water, food
Susceptible individuals
near hot spots
Determine highest inter-
grated exposure levels for
specific groups of indi-
viduals
Limited multimedia sampling
broad areas and intensive
single media monitoring of
suspected hot spots near
populations. Emphasize mon-
itoring and interface of re-
ceptor
Aliening
Chemical Saturation
Determine long term
habitabilily of area
Many chemicals including
degradation products
All levels including trace leveli
Past, current, and future leakages
Also, soil, sediment, biota,
surface water, sewers, ground-
water
Entire population
Document highest and median
concentration levels in individual
media and compare such levels
with control areas
General multimedia sampling of
broad areas with bias toward
natural and man-made path-
ways from the site. Concur-
rent monitoring in nearby con-
trol areas. Monitor along entire
environment pathway
Figure 13.
Dual Approach to Risk Assessment
-------
REMEDIAL RESPONSE 247
Outdoor Air
Indoor Air
Public Buildings
living Quarters
Storage Areas
Food
Locally Grown
Supermarket
Drinking Water
City System
Wells/Springs
Surface Water
and Sediment
Streams/Ponds
Swales/Sinks
Transportation
Arteries
Soil
Sewer Systems
Sanitary
Storm
Basement Sumps
Groundwater
Vegetation
Figure 14.
Framework for Monitoring System Design
intermingled with other industrial activities in popu-
lated areas are shown in Figure 14.
•Statistical aspects are an important factor in the design
of a monitoring program. A statistician on the planning
team can help insure that adequate consideration is given
to these aspects both in designing the program and in
formatting and interpreting the data.
•A photo interpreter can also be an important member of
the planning team, both in selecting monitoring sites and
in designing the approach for relating concentration lev-
els to population activities.
•Before any samples are taken, a complete sampling plan
should be developed with built-in check points for ad-
justing the plan as sampling results are obtained. Devia-
tions from the plan should be resisted other than to ad-
just the plan in response to unexpected data results.
•As a rule of thumb, on the order of 10% of the sampling
might be for screening to help clarify hypotheses as to
possible gradients and hot spots around the site, 80% di-
rected to investigating the hypotheses and 10% reserved
for supplementary sampling of neglected areas that arise
late in the program
•The sampling scheme should include provisions to con-
firm or reject previously reported findings of a contro-
versial nature.
•Access to preferred sampling sites is not always possible.
The sampling plan should be sufficiently flexible to com-
pensate for such problems.
•Geophysical investigations can be very cost-effective in
targetting groundwater and soil sampling sites.
•A quality assurance program involving surrogate recov-
eries, inter and intra laboratory duplicates and field
and laboratory blanks is essential. Each data point should
be individually validated as acceptable data, and pre-
cision and accuracy data should be developed for each
data set. The quality assurance program may account for
0% to 20% of the monitoring costs.
•Special efforts are needed to minimize holding times be-
tween sampling and analysis. However, extended hold-
ing times beyond two weeks may be unavoidable. In that
event appropriate storage procedures are particularly im-
portant to prevent excessive decay of the samples.
•Contaminants associated with the sampling and analy-
tical techniques are difficult to avoid, and data sus-
pected of such contamination should be considered for
discarding. Of particular concern, for example, are ben-
zene and toluene when using Tenax traps, methylene
chloride and phthalates that are present in laboratores,
isophorone which can be a derivative of the laboratory
solvent acetone and the high pH in groundwater asso-
ciated with grouting of sampling wells that may result in
artifacts being observed.
•Data formatting and presentation are of critical impor-
tance. Plotting each data point on maps is probably the
safest way to insure a totally objective presentation of
findings.
•Monitoring data may not provide a definitive portrayal of
pollutant gradients or pollution patterns but may only be
suggestive of general pollutant distribution. Interpreta-
tions of the data may be controversial, and efforts should
be made to isolate criticisms of the quality of the data
from criticisms of the interpretations.
-------
COSTS OF REMEDIAL ACTIONS AT UNCONTROLLED SITES
HOWARD L. RISHEL
SHEILA M. KENNEDY
SCS Engineers
Long Beach, California
JAMES J. WALSH, P.E.
DENNIS P. GILLESPIE
SCS Engineers
Covington, Kentucky
OSCAR W. ALBRECHT
U.S. Environmental Protection Agency
Cincinnati, Ohio
INTRODUCTION
During 1980, the U.S. Congress enacted the Compre-
hensive Environmental Response Compensation and Lia-
bility Act (CERCLA, also known as Superfund, P.L. 96-
510) which was proposed to provide funds for the U.S.
Environmental Protection Agency (EPA) to assist in the
mitigation of pollution problems at uncontrolled waste
disposal sites through remedial actions. The responsible
offices within EPA (the Oil and Special Materials Division,
and the Office of Environment) requested the Office of
Research and Development to provide technical informa-
tion to support this process. As a part of this effort, SCS
Engineers was contracted to review, compile, update and
integrate existing data on the costs of such remedial ac-
tions, in terms of discrete unit operations which could
then be combined to construct conceptual remedial ac-
tion scenarios.
This type of review-and-update approach was consid-
ered more appropriate than additional conceptual design
efforts because much conceptual design work had already
been done. The design work which exists, however, is scat-
tered, incomplete and inconsistent in methodology; much
of it is out of date and vague either about the methods
used to arrive at a cost figure or about what components
the cost figure included.
Through the review-and-update approach, a consistent
methodology on the existing data in terms of scope, lo-
cation, time frame and cost computations was imposed.
In addition, the missing details were supplied and results
presented in a uniform format, with a minimum of over-
lap between the individual unit operations. The resulting
document presents these data in a framework of a broad
and consistent methodology, with enhanced detail.
No new conceptual design work was done for this pro-
ject. Where data were incomplete, some detailed in-
formation was provided, but the thrust of this work was
to enhance previously existing conceptual design data and
make them more available and useful to enforcement
personnel responsible for overseeing the retrofit opera-
tions.
Because the document was intended for use in the re-
trofitting of uncontrolled sites, the unit operations ex-
amined are limited to those appropriate for the clean-up
of closed or abandoned sites.
Methodology
In developing and characterizing unit operations for
remedial action at waste disposal sites, hypothetical site
profiles and their associated unit operation profiles were
developed. Costs for each unit operation were then com-
puted at three cost levels: high and low U.S. averages,
and the price estimate for a single city, Newark, NJ
(mid-1980 dollars).
SITE PROFILES
Site profiles, or hypothetical disposal sites, were de-
veloped for landfill disposal and for surface impound-
ments (disposal ponds). Each of these was portrayed at
five scales of daily operation. The resulting site profiles
were configured to conform to uniform sets of design cri-
teria and environmental conditions. For both landfills
and surface impoundments, the selected scales of opera-
tion were developed in terms of daily input. This em-
phasis on daily input is consistent with the usual view of
landfill practices and the assumptions that surface im-
poundments are intended as temporary storage, with rela-
tively short retention times.
Landfills
The size of the five hypothetical landfill disposal sites is
shown in Table I. The range of scale sizes was developed
from data presented in References (2) and (7). The fol-
lowing assumptions were made:
(1) The surface area for each landfill is square.
(2) All landfills are cut and cover operations, with cut
slopes at a 2:1 ratio and fill slopes at a 3:1 ratio.
(3) Operation at each landfill was 260 days/year for
ten years before the site was closed.
(4) The compaction rate was 0.596 tonnes/com-
pacted m3.
The layout of the hypothetical landfill site, without ref-
erence to the scale of operation is shown in Figures 1 and 2.
248
-------
REMEDIAL RESPONSE 249
To allow comparison of remedial actions between sites
operating at different scales, the following environmental
conditions were also held constant:
(1) Ground surface and groundwater gradient are at a
1% slope.
(2) Groundwater is 4.0 m below the ground surface.
(3) Low permeability strata (<10~6 cm/sec) is 15 m
from the ground surface.
(4) Unconsolidated earth materials have a permeability
of 10 ~5 cm/sec or greater.
Surface Impoundments
The scales of operation for surface impoundments are
shown in Table II. The range of scale sizes was developed
from data in References (3) and (8). In this case, the scale
of operation is given in terms of the daily volumetric flow
of influent.
Table I.
Size of Five Hypothetical Landfill Disposal Sites
Weight of Volume of Area of Waste
Waste Material Waste Material Waste Material to
Contained Contained Contained Soil
(Tonnes) (m3) (ha) Ratio
24,000 40,000 0.99 1:1
120,000 200,000 3.33 1.5:1
240,000 400,000 5.41 2:1
700,000 1,200,000 12.80 3:1
1,200,000 2,000,000 18.51 4:1
1 tonne = 1.1 ton
skip that
1 tonne = 1.1 tons
1m3 = 1.31 yd3
1 ha = 2.47 acre
The plan of the hypothetical surface impoundment with-
out reference to the scale of operation is shown in Figure
3. The following design criteria were common to all of the
surface impoundment site profiles:
•The ponds are square and unlined.
•Berms were constructed from soils excavated during pond
construction, and have 3:1 side slopes.
•The site operated 365 days/year for ten years before
closure or abandonment.
•Sediment was removed from pond bottom every two
years.
•Wastewater contained 100 mg/1 settleable solids.
•Density of solids was 2 g/ml.
•Sludge is 70% moisture by weight when removed.
•Wastewater was recirculated after allowing three to ten
days for solids settling.
•Because of short detention time and sludge on bottom,
precipitation, evaporation and percolation losses are
considered negligible when compared to the volumes
involved.
•Average percolation <14 1/day/m2 (0.2 gal/day/ft2)
To help compare remedial actions for each of the scales
of operation, the following environmental conditions sur-
rounding each pond were developed:
(1) Ground surface and groundwater gradient are at a
one percent slope.
GROUND SURFACt ( IX^-^"^ 83
r ^V
hi
hz
"•^j 1 1
/%
SLOPE
I)'"
GROUND «ATER
FLOW OF GROUND WATER
Low Permeability Strata
Total volume of refuse tm )
Total volume of soil (mj)
Height of landfill above ground surface (m)
Depth of landfill below ground surface (m)
Top side of landfill (m)
Bottom side of landfill (m)
Side of landfill at ground surface (ml
Area of landfill at ground surface (m )
Figure 1.
Typical Side View of Landfill
FLOW OF
GROUND WATER
Figure 2.
Typical Top View of Landfill
(2) Groundwater is 4.0 m below the ground surface.
(3) Low permeability strata (aquiclude or aquitard;
K<10~6 cm/sec) located at 15 m from ground
surface.
(4) Unconsolidated earth materials have a permeability
of >10~5 cm/sec.
Table II.
Scale of Operation for Five Hypothetical Surface Impoundments
Influent Volume
of Waste
-------
250 REMEDIAL RESPONSE
TOTAL
LENGTH
BERM
POND WATER SURFACE
GROUND SURFACE
A
FREEBOARD
GROUND WATER TABLE
4 ID
ism
Low Permeability Strata (Aquiclude or Aquitard)
Where:
•(Total length) = impoundment surface area.
•Berm side slopes constructed at 3 horizontal to 1 vertical.
UNIT OPERATION PROFILES
•Berm top width nominal 2 m, maximum 3 m rs.
•0.5 m freeboard designed into all ponds.
•h = pond depth.
Figures.
Side View of Surface Impoundment
Thirty-five unit operations were identified which
might be used as part of a remedial action effort at an
uncontrolled hazardous waste disposal site. These opera-
tions are shown in Table III for both impoundments and
landfills. Although there appears to be some overlap be-
tween the two lists, each operation was configured sep-
arately for the distinctly different design criteria and en-
vironmental conditions assumed in each site profile.
AH of the unit operations shown were taken from the
literature and only those for which the literature con-
tained adequate conceptual designs and cost data were
addressed. In some cases, a part of the necessary data
was missing and was supplied by SCS.
Alternative Unit Operations
When a pollution problem exists, a number of unit
operations may be used interchangeably. The following
list gives some unit operations which may be used either
conjunctively or as alternatives for each other:
For elimination of contaminated site runoff, and preven-
tion of precipitation from entering a landfill or closed
impoundment, the following are conjunctive or alterna-
tive unit operations:
•Contour grading and surface water diversion
•Surface sealing
•Revegatation
•Berm construction/reconstruction
For minimizing leachate formation, the following are
conjunctive/alternative unit operations:
•Contour grading and surface water diversion
•Surface sealing
•Grout curtain
•Sheet piling
•Slurry trench
•Well extraction
•Well point system
•Chemical fixation
For the control of leachate/contaminated groundwater
migration, the following are conjunctive/alternative unit
operations:
•Grout curtain
•Grout bottom seal
•Sheet piling
•Well point system
•Well extraction
•Well injection
•Underdrains
For gas (methane and other volatile hydrocarbons), the
following are conjunctive/alternative unit operations:
•Perimeter gravel trench
•Gas migration control—passive
•Gas migration control—active
COST COMPILATION
Once a remedial action unit operation had been de-
fined in terms of its intended use (with respect to the
landfill or surface impoundment site profiles) and the ex-
tent to which it can be used in conjunction with or as an
alternative to other unit operations, the cost of the opera-
tion can be defined for each scale of operation associated
with the site profile. To do this, the unit operation is
broken down into component requirements. Each com-
ponent is further defined in terms of sub-components
(labor, materials and equipment) and costs are assigned
to each component and sub-component. The assigned
costs are in terms of mid-1980 dollars for U.S. upper and
lower cost overages (for the continental 48 states), as well
as for the example location of Newark, NJ.
-------
REMEDIAL RESPONSE 251
Table III.
Unit Operations Used as Remedial Actions
Landfills
1. Contour grading and surface
water diversion
2, Surface sealing
3. Revegetation
4. Bentonite slurry trench cut-
off wall
5. Grout curtain
6. Sheet piling cutoff wall
7. Bottom sealing
8. Drains
9. Well point system
10. Deep well system
11. Injection
12. Leachate recirculation
13. Chemical fixation
14. Chemical injection
15. Excavation and reburial
16. Ponding
17. Dike construction
18. Perimeter gravel trench vents
19. Treatment of contaminated
waste
20. Gas migration control—
passive
21. Gas migration control—
active
Impoundments
22. Pond closure and contour
grading of surface
23. Surface sealing of closed im-
poundments
24. Revegetation
25. Slurry trench cutoff wall
26. Grout curtain
27. Sheet piling cutoff wall
28. Grout bottom seal
29. Toe and underdrains
30. Well point system
31. Well extraction system
32. Well injection system
33. Leachate treatment
34. Berm reconstruction
35. Excavation and disposal at
secure landfill
After costs are assigned to each component, conceptual
design capital and operating cost estimates are accumu-
lated and allowances for overhead and contingencies are
applied. Total and average life cycle costs are then com-
puted for each unit operation.
Estimation of Component Costs
For the most part, the 1980 Dodge and Means Guides'4'6'
were used to obtain the needed costs. The costs were then
expressed in terms of metric units.
Regional adjustments indexes presented in the Dodge
Guide(4) were used to modify the metric versions of the
cost estimates for geographical differences. These indexes
were applied to obtain revised material and labor costs for
the U.S. low, U.S. high and Newark, NJ estimates. No
index was applied to equipment costs, since it was assumed
that equipment costs are the same nationwide.
Because the Dodge Guide"0 and Means'6' present costs
differently, assumptions were made so that the regional
adjustment indexes could be used for both texts. For ex-
ample, in the Means, labor costs were not identified as a
separate entry, but were included as part of installation.
Thus, whenever the Means was used to present costs, the
Dodge Guide Regional Adjustment Index for Labor was
applied to installation costs.
Frequently, neither guide itemized costs into categories
of labor, material and equipment but simply presented a
"total" estimate. Depending upon which reference was
used, the following rules were applied: in the Dodge
Guide, if only a total cost was presented, an average
labor/material index was applied to the unit cost; in the
Means, the "total" costs include an overhead allowance
of 25%. This allowance was removed before the labor/
material index was applied. In all cases, costs were adjust-
ed so that overhead allowances were not included at the
subcomponent level.
As the scale of operation changed, the quantity of any
one component required for a unit operation also changed.
The cost of each component presented in the unit opera-
tion conceptual design cost tables typically includes the
sum of costs for any material, labor or equipment sub-
components. These total costs for each component do not
include overhead and contingencies.
Once all the components within a unit operation were
costed, the costs were summed, giving a subtotal capital
cost for the unit operation. This subtotal capital cost was
then used to obtain an overhead allowance (always 25 %),
and a contingency allowance (between 10 and 40%, de-
pending upon the unit operation). The subtotal capital
cost was added to the overhead and contingency allow-
ances to obtain the estimate of total unit operation capi-
tal cost. (See example below.) This method was used for
all scales of unit operation.
Life Cycle Costing
Once total capital and operating costs were-determined
for lower and upper U.S. averages and for Newark, total
and average life cycle costs were computed to ensure that
any subsequent cost comparisons of unit operations
could be equitably accomplished.
Although operation and maintenance (O&M) cost esti-
mates are for 1980, as the first year of operation, O&M
component quantity requirements were estimated to ac-
curately reflect requirements for each of the first ten years
of remedial operations. This ten-year life of the con-
ceptual designs means that life cycle evaluation of operat-
ing costs only addressed subsequent inflation and appropri-
ate discounting of these O&M component costs to their
mid-1980 present values.
It was further assumed that capital costs would not be
amortized and discounted, but would be considered as
fully incurred in the first year of operation. As a result of
these assumptions, average annual compounding inflation
rates for electricity and for all other O&M components
were derived using estimates from the April 1980 Survey of
Current Business'". These inflation rates were derived as
shown in Table IV.
In determining the present values of future expenditures,
the March 1980 Gross National Product Implicit Price
Deflator of 174.51(5) was similarly evaluated in terms of its
1972 base year to estimate an annual general inflation rate
of 7.4%. To this, an assumed 4% social time preference
rate was added to create a total annual discount rate of
11.4%. The life cycle cost methodology was then fol-
lowed, in which inflated operating costs were discounted
to their mid-1980 present values, and summed with total
capital costs, to determine total life cycle costs over the
ten-year life span of each unit operation. Average life
cycle costs were then computed by dividing this total by
the site profile's daily scale of operation.
The contract report of this work includes two tables for
each unit operation, giving the cost of each component,
plus overhead and contingency allowances, O&M costs
and life cycle costs for a medium size landfill and im-
poundment. One of these unit operations is given as part
of the costing example in Tables V and VI.
-------
252 REMEDIAL RESPONSE
EXAMPLE UNIT OPERATION
An example of remedial action unit operation can be
taken in the revegetation of a landfill (Unit operation 3,
Table III). All other unit operations were derived simi-
larly and reported in the same format in the contract re-
port.
One such unit operation is described in Table V in gen-
eral terms.
Table VI describes how the vegetation unit operation
would be achieved in a medium size landfill operation.
The component requirements and their associated costs
when the revegetation unit operation is applied to the
5.41 ha landfill site are shown in Table VI. In this table
both capital and O&M component requirements are identi-
fied and costed in terms of mid-1980 dollars.
The costs associated with each component may repre-
sent the sum of various labor, material and equipment
costs incurred in accomplishing that component. For ex-
ample, costs for the "mulching" component are for ap-
plying mulched hay over 5.41 ha. The labor costs, ac-
cording to the 1980 Dodge Guide, are typically $85.20/ha
($34.50/acre). This number was adjusted for regional
labor cost differentials and multiplied by the number of
hectares involved; these labor costs ranged from $50 to
$110. The value for Newark, NJ was $100. Material
(hay) and equipment costs were added to these labor cost
estimates seen in Table VI. A similar process was followed
in costing each capital and O&M component.
Once all of the capital components had been identified
and costed, allowances for overhead and contingencies
were added to complete the capital cost portions of these
estimates. In all unit operations, a 25% overhead allow-
ance was assumed. This assumption is partly based on
the fact that the Means construction cost guide(6) also as-
sumes a 25 % allowance for contractor's overhead. A con-
tingency allowance to cover unforeseen cost additions was
also applied to each capital cost subtotal. In general, the
contingency allowance ranged from 10 to 40% depending
on the extent to which the unit operation was expected to
encounter unforeseen difficulties.
Each of the O&M component costs were escalated for
future inflation and then discounted to their present values
(in mid-1980 dollars). These present values were summed
over a ten-year life cycle of future site maintenance and
then added to the capital cost total to determine total life
cycle cost. Average life cycle cost was computed by di-
viding this total by the landfill's former daily scale of
operation.
APPLICATION AND LIMITATIONS
The contract report document, produced as a result of
this review and update work, consists primarily of unit
operation costs at the component and subcomponent
levels, and average life cycle costs. By applying this cost
information to the report's discussion of alternative or
conjunctive unit operations, enforcement personnel will be
able to configure and cost complete remedial action
scenarios.
The following is an example intended to show how the
report may be used to help develop an estimated cost for
a remedial action scenario, involving a combination of
unit operations. The estimated cost for the hypothetical
remedial action is then calculated using the cost data pre-
sented in the report.
Hypothetical Problem
An abandoned hazardous waste site has been investi-
gated and is found to be contaminated surface and
groundwater. It is decided that the site must be isolated
by: (a) preventing surface runoff from entering the stored
hazardous waste, (b) preventing groundwater migration
through the site, and (c) implementing a monitoring pro-
gram to confirm the effectiveness of the steps taken. The
unit operations required are as follows:
•Contour grading and surface water diversion
•Surface sealing
•Bentonite slurry trench cutoff wall. Monitoring is in-
cluded in unit operation.
For example purposes, it is assumed that the hypotheti-
cal site has the following dimensions:
•Surface area = 4 ha
Table IV.
Derivation of Inflation Rates for O&M Inputs
Type of O&M Input Electricity
Related Published Electric
Cost Index: Power
March 1980 Index Value: 305.7
Average Annual Index % Increase
Since Base Year (i.e., 12.75 years
since mid-1967 when index
equaled 100): 9.160%
Assumed Future Inflation Rate 9.2%
All Other
O&M Inputs
Consumer
Price (CPI-W)
239.9
7.104%
7.1%
Table V.
Unit Operation 3—Revegetation
Use:
Revegetation helps to physically stabilize the earth material and re-
duce infiltration; it also serves to minimize erosion of the cover ma-
terial by wind and water.
Configuration:
Revegetation involves first grading the landfill, covering it with 1
suitable, fertile soil, adding soil supplements, and then seeding.
Conjunctive Uses:
Contour grading and gas migration control systems are used as colt
components.
Assume:
1. Entire surface of landfill is revegetated.
2. 0.6 m of clay and silt loam will be used for landfill cover.
3. Clay and silt loam are easily accessible; transportation costs tK
not included.
5. Native grasses will be used for seed.
-------
REMEDIAL RESPONSE 253
•Site is square, 200 ra each side
•Average depth of bentonite slurry trench cutoff wall
must be 10 m to impervious material.
•Hydrological investigation indicates that bentonite
slurry trench cutoff wall must extend around three sides
of site to cut off groundwater migration through the
site.
The procedure for estimating the cost of this hypotheti-
cal remedial action scenario is as follows:
1. Refer to the pertinent tables in the report, and list
the components of the unit operations selected. This
step is shown in Table VII, first column.
2. Refer to the price list in the appendix of the report,
and determine what units will be required to measure
the cost of each component. This step is shown in
Table VII, second column.
3. Calculate the number of units of each cost com-
ponent required for the site. This step is shown in
Table VII, third column, for the hypothetical site
used in this example.
4. Refer to the price list in the appendix of the report,
and list the unit cost for each cost component re-
quired. A decision will have to be made on whether
to use U.S. high, U.S. low, or Newark, NY costs.
For this example scenario, U.S. high costs were used,
and are shown in Table VII, fourth column.
5. The final cost calculation required multiplication of
the number of units (Step 3 above) by the unit
cost (step 4 above), as appropriate, and summation
of the cost components to arrive at a total cost. For
this example, see Table VII, fifth and sixth columns.
For the hypothetical remedial action scenario used here
as an example, the estimated costs were calculated as
follows:
Total capital cost $1,277,100
Total O&M Cost, during 10 years $ 128,700
Total 10-year life cycle cost $1,405,800
CONCLUSIONS
The literature review verified that little has been done in
estimating the cost of hazardous waste cleanup at un-
controlled or abandoned sites. Those sources which ad-
dressed remedial responses frequently followed a case-
study or national, industry-wide approach; cost informa-
tion, if provided, was too highly aggregated for costing
separate remedial unit operations. Those sources which
did develop cost estimates at the unit operation level fre-
quently omitted critical components or allowed substan-
tial overlap in the scope of each unit operation. This
study has attempted to overcome such deficiencies, as
well as to quantitatively bound the effects which scale
economies and regional price differentials are expected to
have on the costs of implementing 35 different remedial
response unit operations.
the primary product of this study has been a costing
methodology which was consistently applied to each of
these unit operations. The resulting cost estimates would
seem to lend themselves readily to: (1) comparing costs
for alternative unit operations which perform the same
Table VI.
Costs of Revegetation for Medium Size Landfill
(5.41 ha-27,685m3)
Capital Costs
Area Preparation
Excavation, Grading & Recon-
touringof site
Hydroseeding
Mulching
Capital cost (subtotal)
Overhead Allowance (25 %)
Contingency Allowance (10%)
Total Capital Cost
O&M Costs
Grass Mowing (6/yr)
Refertilization (1 time/yr)
Total O&M Cost
Total Life Cycle Cost'
(over 10 years)
Average Life Cycle Cost*
Per ha
Perac
Lower U.S.
3,710
43,820
4,900
1.450
53,880
13,470
5,390
72,740
320
250
570
77,550
14,335
5,796
Dollars
Upper U.S.
6,440
50,790
6,900
2,080
66,210
16,550
6,620
89,380
650
350
1,000
97,810
18,079
7,310
Newark, NJ
5,760
49,050
6,200
1.860
62,870
15,720
6,290
84,880
550
300
850
92,050
17,015
6,880
•See text for methodology and assumptions
function, or (2) computing combined cost estimates for
unit operations which comprise a complete remedial re-
sponse scenario. The problem with the first type of use is
that such simple cost comparisons do not address techni-
cal differences in the capabilities or efficiencies of alterna-
tive unit operations which accomplish the same goal. Such
differences depend on both the inherent configuration of
each respective unit operation, and the environmental
setting under which it is actually implemented.
As comprehensive as the site profiles may be, they
still represent a conceptual design of an environmental
setting and are no substitute for actual site conditions
when choosing among alternative unit operations. A sub-
sequent report may address some of these qualitative con-
siderations.
The problem with the second type of application is
that the simple addition of unit operation costs, when
configuring a complete remedial response scenario, ig-
nores both real-world influences on conceptually derived
costs and the minor joint effects of scale and agglomera-
tive economies. More specifically, some scale economies
may still be enjoyed when multiple unit operations all re-
quire similar component inputs. Agglomerative economies
resulting from minor component overlap or redundancy
may also occur when unit operations are combined
into remedial response scenarios. Subsequent research
should address these considerations.
Because complete remedial action scenarios for uncon-
trolled sites typically consist of several unit operations,
much remains to be done, even from a conceptual de-
sign cost perspective, in identifying the most appropriate
short- and long-term remedial scenarios. Such an effort
might include a systematic evaluation to determine the
most prevalent pollution problems occurring at uncon-
trolled landfill and impoundment sites. Once this set of
-------
254 REMEDIAL RESPONSE
"typical" pollution cases has been determined, likely
remedial action scenarios could be configured using unit
operations developed in this study. The resulting com-
posite cost estimates for these scenarios could be de-
termined using unit operations developed in this study and
then compared to determine the relative and absolute cost
advantages of each alternative scenario.
REFERENCES
1. Current Business Statistics, Survey of Current Busi-
ness, Vol. 60, No. 4 U.S. Department of Commerce,
Bureau of Economic Analysis. Apr. 1980.
2. Fred C. Hart Associates, Inc. Analysis of the Tech-
nology, Prevalence and Economics of Landfill Dis-
posal of Solid Waste in the United States—Volume II.
EPA Contract No. 68-01-4895, U.S. Environmental
Protection Agency, Office of Solid Waste, Washing-
ton, D.C. 1979.97pp.
3. Geraghty and Miller, Inc. Surface Impoundments and
Their Effects on Ground Water Quality in the United
States. EPA 570/9-78-004, U.S. Environmental Pro-
tection Agency, Office of Drinking Water, Washing-
ton, D.C., 1978.
4. McMahon, L.A., "1980 Dodge Guide to Public Works
and Heavy Construction Costs," McGraw-Hill, New
York, N.Y., 1979.
5. National Income and Product Tables, Survey of Cur-
rent Business, 60, No. 4. U.S. Department of Com-
merce, Bureau of Economic Analysis, Apr. 1980.
6. Robert S. Means Company, Inc., "Building Construc-
tion Cost Data 1980," 1979.
7. SCS Engineers, "Study of On-Going and Completed
Remedial Action Projects: Survey Results and Recom-
mended Case Studies," US EPA, Solid and Hazardous
Waste Research Division, Cincinnati, Ohio, 1980.
8. SCS Engineers, "Surface Impoundment Assessment in
California." US EPA Contract No. 68-01-5137, Of-
fice of Drinking Water, Washington, D.C. 1980.
-------
FORCED CLEANUP;
A POLICE ACTION OR A MONEY JUDGMENT?
JAMES C. SCOTT
ROBERT B. PEARCE
Black & Veatch of Michigan
Detroit, Michigan
INTRODUCTION
In September 1980, Black & Veatch of Michigan, Con-
sulting Engineers, was appointed to the unique position
of Circuit Court Receiver. Not only did this appointment
come as a surprise, but also as an agent of the court, it
gave the Engineer responsibility of undertaking compli-
ance with a previously entered Court Order. Included
were certain activities which had not been performed by
the Defendant in accordance with the direction of the
Court. These actions were in the case of the State of
Michigan, et al. vs. Berlin & Farro Liquid Incineration,
Inc., et al.; a case involving an uncontrolled hazardous
waste site in Genesee County, Michigan.
In addition to making the appointment of a Receiver,
the Court required the Defendant to deposit a cash bond
to ensure payment to the Engineer for services to be per-
formed. Work was to include isolation, stabilization and
confinement of liquid hazardous waste, with related long-
term monitoring to ensure compliance. The program was
to cover design, methodology for implementation and
performance of the work.
When the Engineers' recommended program was sub-
mitted to the Court for approval, the Defendant took ac-
tion to recover the bond and transfer jurisdiction in the
case to Federal District Court. There he was involved in
bankruptcy proceedings. At issue was the clarification of
whether the Circuit Court Order constituted enforcement
action under the State's police powers or constituted a
money judgment against the Defendant.
HISTORY
In 1971, the Berlin & Farro Liquid Incineration Com-
pany applied for and received from the State of Michigan
a license to haul liquid industrial waste, as well as an air
quality permit to install a liquid waste incinerator. The in-
cinerator was placed in operation in late 1972 and the
Company began accumulating wastes at the site.
Subsequently, community complaints led to an investi-
gation by the Michigan Department of Natural Resources
(DNR). The inspection revealed Berlin & Farro was op-
erating in violation of various State Regulations. Formal
Administrative Complaints were filed against the Com-
pany by the State in 1973.
In 1974, the DNR Air Quality Division conducted an
intensive investigation of the incinerator and concluded
it could continue to operate on an interim basis provided
particulate control equipment was installed and certain
procedures were followed during its installation. By this
time, Berlin & Farro had accumulated a substantial
amount of drummed liquid waste, installed several holding
tanks both above and below ground, and constructed
two large storage lagoons for liquid waste.
In March 1974, spills of industrial waste on the site were
documented by the DNR while investigating compliance
with previous orders. By May, a Consent Order was
drafted by the Michigan Air Pollution Control Commis-
sion and a Pollution Incident Prevention Plan was sub-
mitted to the Michigan Water Resources Commission.
In November, several spills of liquid industrial waste
again occurred on the property, resulting in contamina-
tion of a County Drain. But, by October, Berlin & Farro
was making some progress in implementing its Pollution
Prevention Plan and the Water Resources Commission
provided a permit to allow continued storage of liquid
waste on the site.
DNR investigations conducted early in 1975 showed
the Company was not continuing to make good faith pro-
gress in meeting air, waste storage and sampling require-
ments. By August, the Company had accumulated over
2,300 drums (55 gal each) of liquid and solid waste on the
site. An estimated 2.5 million gallons of bulk liquid waste
was also being stored in the lagoons and underground
storage tanks.
Several meetings were held by the DNR with the State
Attorney General and Berlin & Farro attorneys in an at-
tempt to reach some agreement on the air and water
quality problems at the site. These efforts failed and in
September 1975, an Emergency Order to Cease and De-
sist was issued against the Company by the DNR. This
Administrative Order was to close the incinerator and re-
voke the liquid industrial waste hauling license.
Ten days after the Emergency Order was issued, Berlin
& Farro attempted, but failed to have the Circuit Court
restrain the DNR from enforcing its Order. Two months
later the Company was served by the Department with an
Order for hearing an administrative complaint. Early in
December, a hearing took place and in February 1976,
the hearing examiner issued recommendations which in-
cluded reinstatement of the liquid waste hauler's license
255
-------
256 REMEDIAL RESPONSE
and a permanent Cease and Desist Order for operation of
the incinerator.
The Air Pollution Control Commission then acted,
but in a somewhat less positive fashion. They issued a
permanent Order to close the incinerator, unless proper
air pollution devices were installed. In April, the DNR
had restored the Berlin & Farro hauling license.
In June 1976, another spill occurred at the site, again
contaminating the County Drain and a downstream creek.
Various administrative and legal maneuvers continued
well into 1977, when Berlin & Farro was implicated in
other litigation as having received C-56 from the Hooker
Chemical and Plastics Corporation. DNR personnel then
attempted to sample soil, water, waste and sediment at
the Berlin & Farro site. The Company refused to grant
permission for entry and a search warrant was eventually
obtained. Samples were taken from the site, drains, la-
goons and nearby creek bed.
Results from the survey revealed that soils from the
lagoon and surrounding areas were contaminated with ex-
tremely high concentrations of heavy metals (lead, nickel,
chromium, copper and zinc), organochlorines and other
organic compounds. Sediments in the County Drain con-
tained concentrations of organochlorines higher than those
found in the lagoon, with some values exceeding 27,000
mg/kg. Contamination of sediment in the drain and creek
occurred downstream for over 15 miles.
Administrative efforts continued into 1979 when the
Michigan Attorney General finally filed suit against
Berlin & Farro et al. in Genesee County Circuit Court.
This litigation led to a Preliminary Injunctive Order but
the Company failed to comply with all its terms. As a re-
sult, in August 1980, the Genesse County Circuit Court
issued a Contempt Order and fined Berlin & Farro $250
(maximum under Michigan law). The Court also ordered
appointment of a Judicial Receiver to assure compliance
with the previous Consent Order.
To appoint a Receiver, the Court requested the State
to prepare a list of qualified firms and/or persons. The
DNR provided the names of three engineering firms and
one private individual, qualification and experience docu-
ments were solicited, and during a September 1980 hear-
ing, the Court ordered the appointment of Black & Veatch
of Michigan. Further, it ordered the defendant to post a
$10,000 cash bond to cover the affairs of the Receiver.
RECENT EVENTS
Black & Veatch was directed by the Court to develop a
program to assure compliance with a limited number of
specific items from the Preliminary Injunctive Order. The
items selected were determined during the September
hearing and were mutually agreed upon by the State and
Berlin & Farro.
For the most part, the expected results and to some ex-
tent, the methodology by which these items were to be
completed were stipulated in the Order, e.g., "...shall
erect a fence, at least six feet in height (including barbed
wire topping), along the perimeter..." or "...shall deposit
soils from the Site...on top of the sludges in order to
minimize any potential odors to the extent possible." At
the same time, some vagueness existed, i.e., was it to be a
cyclone fence or stranded wire and how much cover was
justified to minimize odor? The answers would have a
significant effect on the cost.
The Court also directed Black & Veatch to include an
implementation schedule in the program. Prompt enforce-
ment was the goal. While the program was to provide for
strict compliance, the Engineer was requested to consider
alternatives which might provide satisfactory compliance
at a lesser cost. A second hearing was to follow the En-
gineer's Report and provide a form for final negotiation of
method, cost, contractors and schedule. At this point,
Berlin & Farro continued to insist on the right to plan and
determine compliance activities. The Court, on the other
hand, appeared increasingly impatient and less responsive
to the Berlin & Farro demands.
Black & Veatch submitted a report to the Court within
three weeks. Since considerable earth work was to be in-
volved, the impending winter made it apparent that rapid
mobilization would be essential to ensure completion prior
to frost. Failure would delay completion for at least six
months.
But the Engineer apparently was not the only one busy.
Only a few days after filing its Report with the Court,
Black & Veatch was served with a Temporary Restraining
Order. This notice came from the Federal District Bank-
ruptcy Court in Detroit, Michigan. A hearing date was set
and the work stopped. Since that time, no real progress has
been made in resolving the continuing migration of pol-
lutants from the site. What has occurred has been a vari-
ety of legal and bureaucratic actions aimed at achieving
the long term goals of the participants.
SPECIAL CONCERNS
Rather than continue with the chronological discussion
of historical facts, attention will now be focused on some
of the more interesting factors which have evolved in this
case. Some will be typical to many cases. Others will
suggest a variety of complicated factors that make imagin-
ative investigation and thorough planning essential to
successful resolution of hazardous waste problems.
Money
The title of this paper makes reference to A Money
Judgment. It is of little consequence whether or not this
case involves a money judgment. Money has and will con-
tinue to be a key issue. It is probable the same will hold
true in most other cases. Corrective costs are high and
available sources of funds are limited. Defiance may have
played a major role in the inactivity of the Defendant in
this case but the cost of corrective action surely has been
a major concern. Some progress was possible when Com-
pany resources were used in remedial activities. When
cash distribution to the Receiver became an issue, the
creditors objected and quickly reduced the Court's flexi-
bility and enforcement.
The State frequently claimed that monies could easily be
made available from Company assets (land) and State or
-------
REMEDIAL RESPONSE 257
Federal funds. This unfortunately has not been the case.
The problem with the Company assets has been compli-
cated by the bankruptcy proceedings and even further by
a legal technicality which will be discussed below. Statu-
tory and political factors have also restricted ready alloca-
tion of governmental funds. Even appropriations to the
Superfund are not likely to be directly applied to this
project.
Jurisdiction
Jurisdictional questions have been far from simple.
What started as an incinerator-air pollution question has
developed into a water resources concern. Improper and
uncontrolled storage rather than burning of the materials
became the long-term issue. Action to rescind the incin-
erator license therefore had limited effect on environment-
al control.
Eventual litigation appeared to place the State Circuit
Court in a position to enforce permanent cleanup. The
role of the Federal Bankruptcy Court interfered with the
effectiveness of this effort and has significantly delayed
control of the pollutants.
Declaration of a Toxic Substances Emergency by the
State of Michigan (only the second such declaration in
the State's history) has shifted the responsibility for di-
recting actions from the DNR to the Michigan Toxics
Substances Control Commission (TSCC). This has result-
ed in the Commission initiating an entirely new program
for permanent site cleanup. At the same time, it has been
reported by the State that the Circuit Court continues to
claim jurisdiction in the case and that it has indicated any
program developed will be subject to review by its Re-
ceiver prior to authorization.
Liability
When initially considering cleanup responsibilities, the
TSCC resolved to solicit the services of a consultant to
develop various alternatives and recommend the "best"
alternative for a permanent solution at the site. Through
subsequent actions however, the Commission itself cre-
ated definition of the alternatives and had an evaluation
of these performed by an Environmental Protection Agen-
cy consultant.
An interesting liability question is raised. In perform-
ing as a Court Receiver, Black & Veatch concluded
that their liability was limited to the execution of the
Court Ordered tasks in a responsible and professional
manner. The assignment did not include creation or evalu-
ation of the final design concepts. Such was not the case
in the TSCC plan. Their original plan called for the crea-
tion of concepts which could have had far. reaching lia-
bility implications to the consultant who created them.
The eventual action, which called for independent evalua-
tion of the TSCC-created alternatives, still carries with it
a far greater liability potential than many firms might be
willing to prudently accept.
There is a similar concern for potential future liability
on the part of a prospective recipient of the hazardous
waste. One alternative calls for permanent burial of the
Berlin & Farro materials at the Hooker Chemical Com-
pany "vault" in western Michigan. While there is a
recognized presence of some Hooker wastes in this ma-
terial, it includes many hazardous materials derived from
other sources. If the Berlin & Farro waste were accepted,
the leachate system at the vault might collect some of
these other wastes and place Hooker in a position of fu-
ture responsibility for treatment and disposal of materials
it did not produce, a circumstance it would certainly
not welcome.
Time
As pointed out in the historical summary, difficulties
in this case were first identified in 1973. Over eight years
later, little has been done to correct the environmental
abuse at the site. Granted, action has been effective in
halting an escalation of the problems but a real solution
has yet to be implemented. This situation is extremely dis-
appointing in view of the major actions taken: state ad-
ministrative orders, Circuit Court Orders, appointment of
a Court Receiver and a State Declared Emergency.
Initial efforts appear to have been directed toward con-
trol of the facilities while still operational. As problems
became more pronounced, a shift in strategy seems to
have occurred. The procedures called for in the Court's
Preliminary Injunctive Order were clearly intended to
secure the site to minimize immediate environmental de-
gradation, rather than toward a permanent solution.
The more recent steps, in response to the emergency
declaration by the TSCC, are now directed toward the
original objective of a permanent solution. This direction
will, of course, be accompanied by an increased cost of
implementation. It would be speculative to suggest
whether this vacillation in objectives was premeditated or
circumstantial but it is probable that it will markedly af-
fect both the time at which contamination is controlled
and the point at which the case is eventually concluded.
Legal Constraints
The nature of the action often required in hazardous
waste cases indicates that legal considerations become of
paramount importance. In the case of Berlin & Farro,
the State operated for some time under the misguided
conception that funds would be reasonably available by
liquidating the site on which the Company was operating.
Unfortunately, the corporation held title to only a small
portion (approximately ten acres) with the remaining
property being held personally by the Company president,
a private individual and not a party in the litigation. As-
cess to the assets was further complicated by the bank-
ruptcy proceedings.
Enforcement power does not carry with it the automatic
right of access to the property of the Defendant. Con-
fusion and inaccurate public reporting developed regard-
ing the ease with which investigations could be per-
formed. This resulted in several instances of strained re-
lations and implied threats. At one point, the Company
president brandished a shotgun to prevent Court au-
thorized seizure of a Company vehicle.
Two bonds were posted in this case. The first, a cor-
porate guarantee bond has been placed on the list of
-------
258 REMEDIAL RESPONSE
debtors to receive its share of the liquidated assets of
Berlin & Farro preventing funds from becoming im-
mediately available. Even the second, a cash bond in the
hands of the Circuit Court, became temporarily unavail-
able for its intended purpose as a result of continued legal
maneuvering.
CONCLUSION
Certain conclusions can be drawn from the experiences
in this case. Hopefully they can be used beneficially' in
other hazardous waste projects.
•A change in certain statutes may be appropriate to better
define which agencies or authorities have jurisdiction in
various hazardous waste affairs. The need for protection
of personal rights and a balance of powers prevent a
clearly defined answer to this question.
•Participants in hazardous waste activities should give
careful consideration to potential liability. Long-term
risks could well overshadow the long-term benefits.
•A realistic assessment should be made at the onset re-
garding desired results and the time for accomplishment.
Wasted effort and lost time will otherwise result.
•The entire hazardous waste arena is constantly faced
with legal entanglements. Lack of attention to this fact
may well result in a failure to meet the desired objective.
Finally, the question of A Police Action vs. A Money
Judgment has yet to be clearly settled. While the bank-
ruptcy Court eventually directed payment of the Engi-
neer's fee in this case, the Order directing this action
stated only "...that Black & Veatch be paid...". It did
not provide a general determination of the issue. Rather
the Court reaffirmed during its hearing that all issues in-
volving disbursement of Company assets must come be-
fore an individual judgment, an action which certainly
will encumber enforcement activities and make the ulti-
mate decision academic.
-------
TECHNICAL AND FINANCIAL ASPECTS
OF CLOSURE AND POST CLOSURE CARE
JOHN W. THORSEN
Weston Designers/Consultants
West Chester, Pennsylvania
INTRODUCTION
Federal legislation and regulations have been developed
to protect public health and environmental quality from
harmful discharges of hazardous wastes. One of the spe-
cific requirements addressed by Federal law is the need to
provide closure and post closure plans with assurances of
financial capability to implement these plans. Federal law
requires that owners of hazardous waste management fa-
cilities provide for proper closure of. their facility. It also
requires that the owner of the facility provide for post
closure care, at a site where the waste remains after closure,
for a period of 30 years. The purpose of these requirements
is to assure the public that, if an incident occurs which
could force an operational hazardous waste management
facility to close, funds will be available to properly close
the facility and assure no environmental or health impacts
occur.
The regulations also specify that adequate funding must
be available to assure closure and post closure care, as
needed, and that such steps can be implemented and
carried out with no financial burden being incurred by the
public.
One aspect that is not addressed by this regulatory pro-
gram is post closure care after the owner's responsibility
has been terminated, i.e., what happens to a site after 30
years to assure environmental protection.
TECHNICAL ASPECTS
Closure is the period after wastes are no longer ac-
cepted, during which the owners or operators complete
treatment, storage and disposal operations, dispose or de-
contaminate equipment and apply cover to or cap land-
fills. The purpose of the closure standards is to assure
that all hazardous waste management facilities are closed
in a manner that:
•Minimizes the need for post closure maintenance, and
•Controls, minimizes or eliminates to the extent neces-
sary to protect human health and the environment, post
closure escape of waste, leachate, contaminated rainfall
or waste decomposition products to ground or surface
waters or to the atmosphere. There are two types of
closure and post closure requirements in the Federal
rules:
A. General requirements which are applicable to all facil-
ities, and
B. Specific technical requirements which are included in
the facility-specific regulations.
The closure standards, specifically for the plan and
cost estimate, assures the owner has considered the ac-
tions and costs that must be taken to properly close his
facility. The regulations specify timing for the develop-
ment and submittal of closure plans. Closure standards
require. that closure plans had to be developed by May
19, 1981 for existing hazardous waste management facil-
ities, or during the development of a plan of operation for
new facilities. During the interim status period, plans must
be maintained on-site. These plans need not be submitted
for regulatory agency review until they are submitted either
prior to closure or to obtain a final permit.
If a facility is to close before a final permit is issued, the
closure plan must be submitted to the EPA Regional Ad-
ministrator, or authorized State at least 180 days before
closure is anticipated to take place so the regulatory agen-
cy can review and approve the document and so the requi-
site public participation, with regard to plan actions, can
be accomplished.
For those facilities in existence on November 19, 1980,
the closure plan must be developed for all portions of the
facility that were active on that date. If some areas were
closed after this date and before the development of the
plan, these areas must be properly closed in accordance
with the closure standards. Partial closure is also a con-
cept that can be utilized to reduce the financial require-
ments for facilities, generally landfills. This allows an own-
er to sequentially close areas of his facility. By minimiz-
ing the active portion of the facility, the necessary ac-
tions needed to meet the closure requirements are less and,
hence, the costs needed to implement the plan are less.
The regulations also require the closure plans be revised
when there is a change in the plan, including changes in
activities, schedules or estimated cost for closing the facil-
ity. The cost estimates must be updated annually.
The closure plan is the cornerstone of the closure regula-
tions. Closure plans must address the following basic ele-
ments including:
•How and when a facility will be closed
•An estimate of the maximum inventory of waste on-
hand at any given time
•Decontamination activities
259
-------
260 REMEDIAL RESPONSE
•Schedule of activities over the closure period
•Cost estimate to implement the plan.
The schedule for closure must address, at a minimum,
three key dates:
•The final date of waste acceptance
•Dates for completion of specific activities
•The final date of closure completion.
A schedule would naturally begin with a 180 day notifi-
cation prior to the beginning of closure. The second date
would be the beginning of closure. An important require-
ment is all treatment and disposal of wastes on-site, must
be accomplished within 90 days of closure. Closure must
be complete within 180 days of the acceptance of the last
waste shipment. This 180-day requirement may be waived
if it can be shown that, realistically, more than one year is
needed. An example of this situation would be closure of
surface impoundments where stabilization of the sludge
must take place before cover and vegetation can be placed.
The final date of completed closure is the date, after all
activities have taken place, upon which separate certifi-
cation can be made by an independent registered pro-
fessional engineer and by the owner that closure has taken
place in accordance with the closure plans, and is com-
plete.
There are facility-specific requirements outlined in the
Federal regulations and these are further addressed in
guidance documents prepared by the U.S. EPA. These
documents explain the facility-specific requirements and
suggest methods for meeting them. They also provide out-
lines for the development of closure plans to meet the gen-
eral and facility-specific requirements.
Post Closure Care
While closure requirements impact all hazardous waste
management facilities, post closure care requirements im-
pact those only where the wastes remain after closure,
i.e., hazardous waste disposal facilities which include land-
fills, land application facilities and seepage facilities. Post
closure is the period after closure during which owners,
or operators of disposal facilities, must conduct certain
monitoring and maintenance activities.
Post closure care consists of, at least, groundwater
monitoring and reporting in accordance with the regula-
tions and maintenance of monitoring and waste contain-
ment systems. In addition, facility security provisions may
be required to remain operable during the post closure
care period or a portion thereof. The circumstances, dur-
ing which security would be required, is where an oppor-
tunity exists for the accidental encroachment on the facility
by people, or animals which could result in adverse health
or environmental impacts.
The post closure care period begins upon the certifica-
tion by an independent registered professional engineer
that closure has been completed. No agency action is re-
quired to initiate the post closure care period. As with the
closure requirements, the post closure care plan must be
submitted 180 days before closure is anticipated to begin.
Post closure owner responsibility extends for a period of
30 years. Activities must be structured over this 30-year
period to assure there are no adverse impacts resulting
from the site on human health and the environment.
There is a provision allowing application to the appro-
priate regulatory agency for a shortening of the 30-
year timeframe. Criteria have been established to evaluate
if foreshortening the post closure care period would still
continue to provide for no adverse human health, or en-
vironmental impacts.
One of the methods EPA has used to attempt to assure
the facility will be safe after post closure owner respon-
sibility has terminated, is land use controls consisting of a
notice to the local land authority, generally a zoning
board. This notice must include a survey plan with the
dimensions of the landfill cells or other disposal facility
specifications with respect to permanent benchmarks; it
must be submitted within 90 days after closure is com-
plete. This survey plan must be developed by a registered
land surveyor and must include the type, location and
quantity of hazardous waste disposed of in the facility.
In addition, the owner must record a notation on the deed
for the property that will notify any potential purchaser
of the past use and current and future use restrictions
for the facility as a result of the activities that have taken
place at the facility. The regulations require post closure
use of the facility cannot disturb the integrity of the final
cover, liners or other components of the containment sys-
tem.
As with the closure, the post closure plan is the corner-
stone of the post closure requirements. Its basic purpose
is to assure the owner has considered the actions and the
associated costs, to provide for post closure care.
Some of the plan development requirements are the
same as for post closure plans: the how and why must
be specified, the activities and frequencies to assure appro-
priate post closure care must be identified, a schedule must
be developed, and cost estimates must be provided. Plan
revisions are allowed and, in fact, needed where the sched-
ule activities, or cost estimates change.
The plan describes the basic activities that must take
place during post closure care. These include monitoring
and maintenance activities. Monitoring of groundwater is
the primary monitoring requirement. Continuation of the
groundwater monitoring system utilized during the active
life of the facility must be continued, sometimes at a less
frequent rate. In addition, inspection and maintenance
activities must be conducted to assure the integrity of the
facility. Specifically, the following activities must be in-
cluded in the inspections and maintained if the need arises:
•Run-off control structures
•Settlement (through maintenance and use of benchmarks)
•Erosion damage
•Gas control facilities
•Leachate management facilities
•Groundwater monitoring system
After the activities and schedules have been established,
cost estimates must be prepared to implement the plan.
Ground rules for cost estimates are briefly addressed in the
section on Financial Aspects.
-------
REMEDIAL RESPONSE 261
FINANCIAL ASPECTS
Four topics are discussed below under the financial as-
pects of closure and post closure care. These include:
•the basic rules for estimating the costs of the activities
necessary to implement the closure and post closure plans;
•the financial mechanisms that can be utilized to meet the
financial responsibility requirements;
•the timing for implementation of this RCRA require-
ment and,
•the current regulatory situation regarding the provision of
financial assurance to the regulatory agencies.
In developing estimates to implement the actions out-
lined in closure and post closure plans, four different
techniques can be used:
•The experience of the owner or operator
•Contractor estimates of the actions
•Cost estimating handbooks
•Work-ups from labor, materials and equipment require-
ments.
Experience to date has shown owner/operator exper-
ience generally provides the least costly of the different
techniques.
As indicated above, the estimates to implement the plans
are made in current year dollars and must be updated to
account for inflation. These updates must be annual and
for this purpose a standard tool to gage inflation over the
past year has been selected by the U.S. EPA. This is the
"Implicit Price Deflator for Gross National Product"
which is developed by the Bureau of Economic Analysis
in the U.S. Department of Commerce. This document is
published monthly, and is available from the Bureau of
Economic Analysis. Further, since it is a regulatory re-
quirement, U.S. EPA regional personnel should have
ready access to this information.
The actual index is developed in a ratio manner by
dividing the deflator for year "X" by the deflator for
year "X-l". This ratio is then multiplied by the estimate
to result in a revised total dollar figure needed for the
accomplishment of the closure and the post closure re-
quirements. The financial assurance mechanism must then
be modified to assure that adequate funds are available to
implement the closure and post closure plans.
Basic Cost Estimation Rules
Basic rules for estimating closure costs are outlined
below:
•Costs must be based on the activities described in the
closure plan
•The estimate must be made when the closure costs are
most expensive, i.e., when the maximum extent of opera-
tions is the greatest and there is the largest anticipated
on-site inventory of waste materials
•Costs can be based on the owner or operator conducting
the closure activities which allows the owner/operator to
take advantage of the depreciated value of equipment
used in closing a facility
•All pertinent costs must be identified and included in the
cost estimate
•Costs are estimated in the year of preparation and in-
flation factors are considered on an annual basis using
the index described above.
The ground rules for estimating costs for post closure
activities are only slightly different from those in the
closure sequence. The largest single difference is the cost
must be based on contracting for the provision of
services needed to implement and carry out the closure
plan. Other ground rules include:
•Cost must be based on the activities described in the post
closure plan
•Cost must reflect the care for the total area of haz-
ardous waste containment
•Cost must be complete and include all the pertinent
costs
•Cost estimate must cover the entire 30 year time period
required for post closure owner responsibility.
Financial Mechanisms
As required by the hazardous waste regulations, own-
ers or operators of hazardous waste management facil-
ities, must establish financial assurance for closure and,
where necessary, post closure to assure the plans can be
implemented at any point in time and the funding needed
for the plan implementation will be available. The follow-
ing options are available for providing financial assur-
ance for closure and post closure care:
•Trust fund
•Surety bond guaranteeing performance
•Stand-by letter of credit assuring funds
•Combination of the above.
These methods are outlined specifically in the Fed-
eral Register, January 12, 1981.(5) States' requirements
may vary from the Federal requirements, but they must be
at least as stringent as the Federal requirements in provid-
ing financial assurances. Examples of the mechanisms are
also provided in the January 12, 1981, Federal Register.
TIMING OF REGULATIONS
The regulations were scheduled to be effective on July
13, 1981. However, because of concerns raised regarding
the limited range of mechanisms promulgated by U.S.
EPA in January, the U.S. EPA has extended the comment
period and, subsequently, extended the effective date until
October 13, 1981.
Basic concerns regarding the mechanisms promulgated
in January, centered around the lack of a self-insurance
mechanism and a financial test mechanism to meet the
financial assurance, and liability requirements as required
by the law. The concerns expressed by the industrial com-
munity were very specific. The U.S. EPA has indicated
it is looking into, and most likely will develop, mecha-
nisms for self-insurance and for a financial test so that a
firm can provide for self-insurance to meet the financial
assurance, and liability requirements. EPA agreed there
was no point in initiating and acquiring much more costly
mechanisms outside a firm if, in fact, the ability to insure
oneself was going to be promulgated at some point in the
-------
262 REMEDIAL RESPONSE
future. There are other concerns regarding the need for
separation of, or duplicate closure and post closure mech-
anisms, as well as the ability for a corporate entity to pro-
vide financial assurances for multiple sites using one
"umbrella" mechanism.
One of the basic underlying concerns regarding the
mechanism selected and utilized under this program, re-
volves around tax considerations. If EPA allows insurance
to guarantee closure and post-closure funds, then the
funds set aside through insurance premiums would be tax
deductible as business expense. However, if these funds
were placed in a trust, there would be tax liability to this
money. The regulated community sees this as being an un-
desirable situation since they would be taxed for money
they could not utilize in their day-to-day business.
U.S. EPA anticipates the issue will be resolved in a
fashion so the revised regulations can be published before
the October 13, 1981 deadline.
CONCLUSION
The depth and breadth of events that must take place in
this one sub-rpogram of the hazardous waste program
illustrates the complexity of the entire program. Hence,
great care must be taken in analyzing and meeting the re-
quirements specified by the State and Federal programs.
The basic purpose of the Federal requirements for haz-
ardous waste sites is to show that the owner or operator
has adequately considered, and planned for the closure
and post closure technical and financial requirements.
The technical requirements are needed for both the in-
terim status standards, as well as for obtaining a RCRA
permit. The financial assurance requirements are also
needed for both interim status and RCRA permits; how-
ever, the implementation has been delayed until at least
October 13, 1981.
The importance of careful cost estimates for closure and
post closure care cannot be over-stressed. These require-
ments can have a direct financial impact on a firm and,
therefore-, must be analyzed in conjunction with the
closure and post closure plans to minimize exposure.
As specific as these rules appear to be, because each
facility is unique, there is a capability for innovative, and
creative rule interpretation to allow for minimization of
exposure under these regulations. Such decisions may
often mean the expenditure of capital costs as opposed to
long term operation and maintenance costs. Cost analysis
should be readily employed to assure the most cost effec-
tive long term solution for closure and post closure.
REFERENCES
1. Burt, R.E., et al., "Draft Guidance for Subpart H of
the Interim Status Standards for Owners and Operators
of Hazardous Waste Treatment, Storage and Disposal
Facilities, International Research and Technology
Corporation, Aug. 29, 1980.
2. "Hazardous Waste Management System; Standards
Applicable to Owners and Operators of Treatment
Storage and Disposal Facilities; and Permit Program,"
Federal Register, 46, No. 24, Feb. 5, 1981, 11126-
11177.
3. "Interim Status Standards for Owners and Operators
of Hazardous Waste Treatment, Storage and Disposal
Facilities, Federal Register, 45, No. 98, May 19, 1980,
33232-33258.
4. Severn, R.R., et al., "Draft Guidance for Subpart
G of the Interim Status Standards for Owners and Op-
erators of Hazardous Waste Treatment, Storage and
Disposal Facilities," International Research and Tech-
nology Corporation, Oct. 6, 1980.
5. "Standards Applicable to Owners and Operators of
Hazardous Waste Treatment, Storage and Disposal
Facilities; Consolidated Permit Regulations," Federal
Register, 46, No. 7, Jan. 12, 1981, 2802-2882.
6. "Standards Applicable to Owners and Operators of
Hazardous Waste Treatment, Storage and Disposal
Facilities; Consolidated Permit Regulations," Federal
Register, 46, No. Ill, June 10, 1981, 30624.
-------
PRACTICAL CONSIDERATIONS FOR THE PROTECTION OF
PERSONNEL DURING THE GATHERING, TRANSPORTATION,
STORAGE AND ANALYSIS OF SAMPLES FROM HAZARDOUS
WASTE SITES
GORDON A. ALLCOTT
ROBERT VANDERVORT
Radian Corporation
Salt Lake City, Utah
JOHN V. MESSICK
Radian Corporation
Austin, Texas
INTRODUCTION
Uncontrolled hazardous waste sites can present a spec-
trum of potential health, safety and environmental
hazards. Detailed information regarding each site is
necessary before comprehensive plans can be developed
and safely implemented to mitigate hazards presented by
the uncontrolled waste. Acquisition of detailed site spe-
cific information can involve serious potential risks.
In this paper, the authors present practical suggestions
for the protection of investigators in exploring, sampling
and evaluating uncontrolled hazardous waste sites. Also
considered are factors related to protection of the public.
First a generic approach to hazard control planning is pre-
sented. This is followed by discussion of a composite case
study based on real experiences in several waste explora-
tion and analogous hazard situations.
GENERIC APPROACH TO
HAZARD CONTROL PLANNING
Occupational exposures to hazards associated with
waste site exploration and evaluation can be controlled or
avoided. Success in prevention of adverse exposures de-
pends on detailed_planning, training and execution follow-
ing principles of industrial hygiene, safety and occupa-
tional medicine. The discussion which follows presents a
logical sequence of assessment and planning activities
which has been developed and applied in actual waste site
projects.
Preliminary Hazard Assessment
The first step in assessing potential hazards at a waste
site is to collect and evaluate all known and available
information regarding the site. Each site presents a
unique matrix of risks. Among the important variables
which define relative risk are: (1) materials deposited at
the site, (2) type and integrity of hazardous material con-
tainers, (3) hydrogeological and topographical features of
the site and (4) the period during which deposits were
made at the site. To help assure that all useful informa-
tion is acquired, a list of informational categories (Table I)
with some specific examples has been developed. In prac-
tice, Table I can be used as a basic checklist during
documentation, search and site visit activities.
The second step in preliminary hazard assessment is the
identification of information which has relevance to the
specific site but is uncertain or unavailable. The items in
Table I can also be useful to this analysis. The objective
here is to identify all potential hazards or gaps in known
information which could reasonably affect safety in ac-
tivities to be conducted on or off the site. Caution must
be observed to prevent undue biasing of this analysis by
facts which are known about the site. For example, at
some sites a great amount of descriptive information may
be available. The importance of missing information may
be difficult to appreciate.
To avoid overly conservative assessment of potential
hazards, a third closely related effort should accompany
step two. This effort should try to logically eliminate
potential hazards or data gaps from serious consideration
when there is good reason to believe they have no im-
Table I.
Important Features of Uncontrolled
Hazardous Waste Sites
Waste Parameters
Sources/volume/form
History of waste deposits
Dates
Sources of waste
Type and quantity of waste
Sources of additional information
Containment/confinement of waste
Waste containers
Types, age, condition
Designed confinements
Pits, lagoons, cells, trenches, cover
Uncontrolled practices
Open dumping
Open burning
Flooding/evaporation/percolation
Hazardous properties
Physical and chemical properties
Flammability, corrosivity, reactivity
Potential toxicity
Acute, chronic
Site Characteristics
Topography
Geology
Hydrology
Climatology
Wildlife (reptile, animal, insect)
Ground cover
History of sampling/exploration
Accessibility
Security
Adjacent tenants
Related Considerations
Nearby population centers
Hospital facilities
Ambulance service
Fire district
Utilities available/proximity
Industrial equipment rental
Law enforcement
263
-------
264 SITE & PERSONNEL SAFETY
portance. For example, a site may be known to contain
only hazardous wastes from building demolition (e.g. as-
bestos) and junk transformers (e.g. PCB's). In this case,
organophosphate pesticides could be eliminated from con-
sideration since there is no logical evidence that they
would be present at the site. Similarly, serious considera-
tion of buried drums of flammable solvent and other un-
likely hazards could be rejected.
The fourth step in the preliminary hazard assessment
is to obtain descriptions of the site exploration, waste
sampling and other activities which are proposed. A list
of several types of activities which may be required in
exploring and evaluating a waste site prior to development
of remedial action plans is given in Table II. The initial
focus of this effort should be the equipment ancf pro-
cedures to be employed on site. Next, this analysis should
consider offsite activities which will be required to sup-
port the proposed project. Transportation of waste sam-
ples to a laboratory and their analysis are examples. It
is important that anticipated procedures be described to-
gether with the rationale for their selection.
Table II.
Site Reconnaissance, Exploration,
Sampling and Evaluation Activities
Surveying
Plat, elevations, grid
Water table elevation
Metal detection
Surface probing
Marking, staking, flagging
Access
Timber and brush removal
Surface grading
Road construction
Water pumping/diversion
Security
Fencing, barricades, signing
Offshift site and equipment protection
Sampling/Exploration
Core/well drilling
Excavation/trenching
Drum/container sampling
Gathering of soil, waste, water, air,
vegetation, biological samples
Disposal of excess sample material,
washings, etc.
Sampling Handling/Analysis
Identification and logging
Packaging
Storage
Aliquoting
Extraction
Filtration
Chemical analysis
Waste disposal
Inspection/Maintenance
Personal protective items
Support system (compressors, gen-
erators)
Hygiene facilities
Exploration equipment
Shipping/Receiving
Labelling, packaging and shipment per
DOT regulations
Items handled
Equipment
Supplies
Samples
Clothing
Wastes
The fifth step in the preliminary hazard assessment is
the simultaneous consideration of potential hazards and
the activities proposed to be conducted on and off site.
The purpose of this exercise will be to identify situations
where hazardous exposures of research personnel, sub-
contractors or the public may occur. The interface be-
tween each activity and each known or accepted potential
hazard must be considered. Prediction of both "probable"
and "worst case" exposures is very important to this
analysis. Examples of several "worst case" incidents
which could occur both on and off the site are presented in
Table III.
An attempt should be made to describe the logical con-
sequences of uncontrolled exposure incidents. Could a fire
or explosion occur? Could personnel be exposed to high
airborne concentrations of toxic contaminants? Could
personnel be splashed with toxic or corrosive materials?
These are examples of questions which require probable
and worst case answers. Thoughtful organization of data
from this step in the analysis can help to highlight im-
portant areas of uncertainty and high hazard situations.
Exposure Control Program Development
Having identified probable and worst case exposures
which may occur in work on or off the site, plans can be
formulated for protection of personnel. Experience has
indicated that a broad range of protective measures may
be applicable to different sites or to distinct activities
planned for a single site. For each waste site scenario, the
objective is to provide a program of protection which of-
fers a reasonable margin of safety to exposed employees.
Table ID.
Examples of Worst Case Incidents
Fire and/or explosion of flammable or combustible solvents or pesticide
mixtures
Explosion of waste containers containing shock, pressure or heat sensi-
tive materials
Penetration or rupture of compressed gas cylinders (buried or at the
surface) containing toxic materials
Penetration of protective gear by toxic liquids, gases or vapors
Penetration of protective gear by equipment movement, flying debris, or
contact with sharp objects
Interruption or contamination of supplied breathing air
Excavation and surface cave-ins
Equipment rollovers
In transit leaking or rupture of sample containers
Rupture or leakage of sample containers while in storage
Violent reaction of waste samples with analytical reagents
Medical emergency in hazardous area (e.g. heart attack)
In waste site work as in other occupational settings,
feasible engineering, administrative and work practice
exposure controls should be applied to reduce reliance on
personal protective clothing and equipment. Contrary to
the opinions of some, the self contained "moonsuit" ap-
proach to employee protection should not always be neces-
sary to provide a reasonable margin of safety.
Job hazard analysis is a useful approach to identifying
potential exposures and required controls. Where serious
potential hazards are indicated, careful review of the
factors which dictate the need for maximum protection
should be made. In some cases, extreme caution may be
dictated by the absence of information which could be
safely acquired with reasonable cost and effort.
In other situations, incompatibilities may arise from
inappropriate combinations of equipment, procedures and
hazards. Substitution of alternate methods (e.g. a change
in drilling equipment and procedure) or limiting the use of
a method (e.g. no drilling at coordinates where metal de-
tectors indicate the possible presence of buried drums
and cylinders) may reduce or eliminate these incompati-
bilities. In some situations, freedom to substitute tech-
niques may be limited by legal or contractual constraints.
However, where improved safety may result, the impact of
-------
SITE & PERSONNEL SAFETY 265
Table IV.
Sample Schedule of Personal Protective Items by Work Activity
Work Activity
1
1. Initial survey of site
2. Metal detection survey
3. Site clearing and
preparation
4. Core drilling and core
handling
5. Mound exploration
6. Final survey of site
7. Rescue, repositioning
of support systems
|
X
X
X
X
*
*
*
X
X
X
X
X
X
X
*
X
X
*
X
*
X
X
*
X
X
X
X
X
X
X
X
•
X
X
X
X
I
*
X
X
*
*
X
*
X
'Clothing or equipment which may be required based on site conditions.
these constraints should be discussed with the sponsor-
ing group so that reasonable compromises can be achieved.
When personal protective clothing and equipment are
necessary to supplement other exposure controls, they
should be selected in a logical, objective manner. De-
cision logics to select respiratory protection devices are
available from the National Institute for Occupational
Safety and Health (NIOSH), the American National
Standards Institute (ANSI), respirator equipment manu-
facturers and others. Similar decision logics are becom-
ing generally available for other forms of protective
clothing and equipment (e.g. gloves, boots, body cover-
ings). Suggestions for protective clothing and equipment
ensembles usage in common waste site work are given in
Table IV.
Whenever heavy reliance is placed on protective clothing
and equipment, it must be remembered that no protec-
tive system is one hundred percent effective. A number of
probing questions must be answered. Will the scheme of
protection afford adequate safety in the event of prob-
able and worst case circumstances? Is the information
being sought worth the risk? How will erosion of the
Planned safety margin be detected?
In addition to selection and use of protective clothing
and equipment, a variety of support activities and fa-
cilities are necessary to complete the hazard abatement
plan. Categories of support items which may be required
at specific sites are given in Table V. Fortunately, many
field applicable emergency abatement and personal hy-
giene support items are now commercially available. Ex-
amples would be portable chemical toilets and hand
washing stations and portable eye-wash fountains. These
and other items can be relied on for control of routine
work hazards and minor emergencies.
Hazard control planning should also include provisions
for major accidents. Emergency plans should address seri-
ous injuries, fires, explosions, spills, releases and other
adverse events. A list of several elements which should be
included in emergency planning is given in Table VI.
Control Plan Implementation
Implementation of hazard control plans should involve
a variety of training and rehearsal sessions. Additionally,
physical examinations, clinical tests and other procedures
may be required. A list of pertinent training and examina-
tion items is found in Table VII.
On-site, full-time safety and health supervision is highly
recommended where significant potencial hazards exist.
No hazardous work should commence until all hazard
control systems and facilities are set up and checked out.
-------
266 SITE & PERSONNEL SAFETY
Table V.
Site Exploration Support Facilities Equipment and Services
Emergency Services
Fire department
Ambulance
Medical care
Law enforcement
Utilities
Electricity
Potable water
Telephone
Sanitary sewer
Site Security
Fences/gates
Lighting
Guard station
Security officers
Intruder alarm
Other
Freight pickup/delivery
Industrial equipment rental
Fuel delivery
Facilities
Chemical toilets
Handwashing station
Protective clothing and
equipment storage
First aid station
Lunch tables/benches
Temporary office
Lockers
Water coolers
Equipment
Emergency shower(s)
Eye-wash fountain(s)
Communications equipment
Telephone
Public address system
Siren/horn
CB radio
Voice actuated tape recorder
and radio communication
Fire alarm
SCBA units
Stretcher
Electrical generator
Air compressor
Clothes washer
Clothes dryer
Meteorological station
Water pumps
Fire extinguisher(s)
Each site exploration procedure should be tested in a
no hazard area with the scheduled hazard control measures
in effect. Any problems encountered should be satisfac-
torily resolved before work is performed on the waste
site. Work on the site should proceed from areas of rela-
tively low hazard to areas of greater hazard.
COMPOSITE CASE STUDY
This section describes practical lessons which have been
learned in waste site and other projects involving analo-
gous work situations. The information gained from sev-
eral projects has been combined to form a composite
case study so that a broad range of experience can be
discussed as well as protecting provileged information
regarding specific waste sites.
Site Description
The hypothetical waste site to be evaluated is located
approximately ten miles outside a small industrial center
in the western United States. The total site encompasses
320 acres of basically level land. A marsh extends from
within the site to the east. A drainage ditch defines the
northern boundary of the property. Vegetation at the site
consists of grasses, reeds, bushes to 4 ft and trees to 40 ft
high and 0.5 ft in diameter.
Waste disposal at the site began in the 1930s. Industrial
development accompanying World War II increased use of
the site during the 1940s. During the period 1940 to 1975,
the local municipality and several nearby industries in-
termittently used the site. One local company provided
records of waste which they deposited there. Other com-
panies used the site at various times; some of these are no
longer in business.
Table VI.
Elements of Emergency Plan
Anticipated emergencies
Weather extremes
Fire and explosion
Work accidents
Medical emergencies
Civil disobedience/unauthorized
entry to site
Established procedures/
responsibilities
Communications
Weather service
Fire department
Hospital/industrial clinic/
physician
Ambulance service
Law enforcement authority
Rescue/extrication of injured or
ill
Supervisor and employee action
Outside support
Fire suppression/control
Supervisor and employee action
Outside support
Medical care
First aid
Outside medical support
Site evacuation
Site security
Evacuation of adjacent properties
Table VII.
Examinations, Training and Rehearsals
Examinations/Histories
Hearing
Vision
Pulmonary function
Medical history
Employment history
Physical examination
Clinical tests
Training/Rehearsals
Emergency and routine communications
First aid (CPR, EMT)
Respiratory protection use, limitations, inspection, maintenance
Emergency equipment (deluge shower, eye-wash)
Rescue of ill or injured
Site evacuation
Use, limitations and care of protective clothing and equipment
Site surveying, sampling and exploration procedures
Sample identification, packaging, storage, shipment
Laboratory procedures
Disposal of hazardous materials
Fire fighting
Physical evidence indicates that wastes have been de-
posited in scattered locations on approximately 100 acres.
Waste has been buried on dry ground and submerged in
the marsh to depths of 15 ft. Liquid wastes have been
discharged into pits for evaporation, burning and/or per-
colation. Drummed waste litters an 0.5 acre area. It is
suspected that similar waste has been piled on the site and
covered with earth to form mounds.
Work Assignment
The site will be evaluated in phases. Results of each
phase will be considered in planning successive phases.
Phase I work, discussed here, involved the following tasks
and information:
(1) Location of all waste disposal areas
(2) Core drilling of dry land waste disposal areas and
adjacent ground using a 50 ft grid pattern
-------
SITE & PERSONNEL SAFETY 267
(3) Core sample analysis by RCRA procedures
(4) Sampling and analysis of groundwater and marsh
water
(5) Excavation of waste mounds with photographs and
description of their contents
(6) The small area where drummed waste resides on
the ground surface is to be roped off and left un-
disturbed.
Project Chronology
The following major activities were performed:
(1) An interdisciplinary project team was assigned and
project organization established
(2) Available information was reviewed and a list of de-
sired information prepared
(3) A preliminary reconnaissance visit to the site was
conducted including discussions with several waste
generators
(4) A comprehensive plan to perform the work assign-
ment was prepared including an integrated health
and safety plan
(5) Required equipment, clothing, etc. were assembled,
fitted and tested
(6) Project employees were examined, trained and re-
hearsed in planned activities
(7) A project base was established at the site
(8) Planned activities and safety procedures were re-
hearsed with all project personnel including sub-
contractors in clean areas at the site
(9) Preliminary and indepth site work was performed
(10) Core and water samples were shipped to the labora-
tory
(11) Laboratory analyses were performed
(12) A technical report was prepared.
In the remaining sections of this paper, highlights of
safety and health activities are presented.
Site Reconnaissance
The site was observed and photographed from undis-
turbed areas surrounding the waste site and from well es-
tablished roads through the waste disposal area. Recon-
naissance personnel wore impervious boots and gloves,
eye protection, safety helmets, and half-facepiece air-
purifying respirators while in or near waste disposal areas.
No samples of waste or water were gathered or any waste
disturbed: Detector tubes were used to indicate the pos-
sible presence of classes of vapors and gases.
Project Work and Hazard Control Planning
Project team members including safety and health per-
sonnel worked together to identify potential hazards,
formulate plans for conducting required work, and inte-
grating safety and health considerations into all phases.
Key elements of the comprehensive safety and health plan
were:
(1) Job hazard analysis for all project activities with
identification of activity specific control measures
(2) Consideration of routine and worst case exposure
possibilities and formulation of procedures for
normal work and emergency action
(3) Specification of field and laboratory support ser-
vices and equipment, facilities and safety supervision
(4) Purchasing, assembly and testing of all protective
systems before traveling to the site
(5) Medical examination and training of in-house and
contractor personnel prior to waste site work
(6) Establishment of a base facility at the site complete
with safety, hygiene and emergency systems prior to
initiating site work
(7) Full-time safety and health supervision of all work
(8) Rehearsal of all site exploration and sampling work
in clean areas before entering contaminated areas
(9) Rehearsal of all sample gathering, packaging, stor-
ing, aliquoting, analysis and waste disposal pro-
cedures prior to handling real samples
(10) Rehearsal of emergency actions
(11) Post medical examination and debriefing of all pro-
ject personnel
Description of Base Facility
A base facility was established on an access road upwind
and equidistant to most areas of waste exploration work.
The facility included a trailer-mounted emergency shower,
hygiene facilities, stores of protective clothing and equip-
ment, rescue equipment, first aid supplies, portable power
generators and communications equipment. Lunch tables,
chemical toilet, equipment cleanup, inspection and repair
facilities were also provided. Direct communication lines
and message relay systems were established with the local
hospital, fire department, ambulance service and private
emergency assistance group. The base facility was manned
by an experienced safety and health professional. This
person had authority to stop work at any time if unac-
ceptable risks or failures arose.
Preliminary Survey of Site
Prior to sampling, core drilling and excavation of the
site, a preliminary survey and mapping was performed
by employees working in pairs wearing impervious suits,
gloves, boots, safety helmets, eye protection and air-
purifying respirator. Metal detectors were used to locate
buried metal (possible waste drums). A narrow metal rod
was used to probe for "drum graves" or other areas of
weak surface support. Later, no core drilling would be
performed where metal was detected or infirm ground was
discovered. These precautions helped to reduce the po-
tential for penetrating a waste drum with the core drill
and also potential upsets of the drilling equipment.
Core Drilling and Mound Exploration
An independent drilling contractor was engaged to core
drill the site. A two-man team was thoroughly trained in
all sampling and safety procedures. A geologist sampled,
described and packaged core segments. During the ad-
vance into the site, drilling and retreat from the site these
men wore impervious boots, gloves and air-supplied hel-
met with vortex tube cooler. Air-supplied helmets were
selected after consideration of probable and worst case
core drilling incidents. This analysis indicated only a re-
-------
268 SITE & PERSONNEL SAFETY
mote chance for exposure to contaminant concentrations
near an IDLH (immediately dangerous to life and health).
Compressed air was supplied from a gasoline fueled,
rotary air compressor with 100 ftVmin capacity at 100
psig. The compressor was fitted with a high temperature
shutdown device. Inlet air was drawn from an elevated air
intake. Compressed air was regularly checked for compli-
ance with Class D breathing air requirements and continu-
ally filtered for removal of particulates and organics. Con-
densate was removed at the compressor and at multiple
drip legs in the pressure hose system. An air hose, 1 in. in
diameter connected the compressor to two dual outlet pres-
sure/flow regulation and filtration units. Each helmet was
supplied with 100 to 200 ft of air hose (3/8 in. diameter).
A microphone attached to a portable recorder was in-
stalled inside the geologists helmet so that descriptions of
core samples could be tape recorded. This microphone
was also connected to a two way radio. The on-site safety
supervisor maintained the air compressor, tested air qual-
ity, maintained radio contact with the geologist and direct-
ed advance and retreat activities. He also supervised dis-
robing and hygiene practices. Air-supplied helmets and
support equipment were thoroughly inspected, cleaned and
repaired as necessary each day.
The safety supervisor remained in sight of and in radio
contact with core drilling. Air horns were available for
signalling and alarm. The base camp was equipped with a
self-contained breathing apparatus (SCBA) which could
be used for rescue after calling for and confirming that
outside help was on the way. Back up SCBA rescue as-
sistance could be on-site in less than 15 min.
Mound exploration was performed using the same en-
semble of protective clothing and equipment. A backhoe
was used to open selected mounds from the upwind side.
Photographs were taken, descriptions recorded and over-
burden replaced.
Groundwater and Marsh Water Sampling
After all core drilling was complete and slotted screen
casings installed, water samples were obtained from core
holes. Depth to water measurements were also made and
later combined with survey information to describe the
water table on the site. During water sampling from core
holes the same protective ensemble used in preliminary
survey of the site was employed.
Marsh water samples were gathered by wading into the
marsh at the periphery of the waste deposit. Chest waders,
impervious gloves, goggles and half-facepiece air-purifying
respirator were used. A safety harness and attended life-
line were also worn.
Sample Handling, Transportation, Storage,
Analysis and Disposal
The waste and water sampling and analysis program
was planned, critiqued and rehearsed before any samples
were collected. Provisions for documented chain of cus-
tody were built into the system. Laboratory space for
sample storage and analysis was dedicated prior to site
work.
Selected segments of core samples were deposited in
numbered, one quart, glass jars with threaded, Teflon® -
lined closures. Similar jars were used for water samples.
After closure, jars were cleaned of surface contamination
and additional identification attached. Soil and waste
samples were stored in a freezer on site. Water samples
were refrigerated.
Sample jars were placed in protective plastic mesh and
packed in ice chests filled with vermiculite and "Blue
Ice." Chests were sealed and marked in compliance with
Department of Transportation regulations. Samples were
conveyed to the laboratory by Federal Express or as truck
cargo with returning team members.
At the laboratory, technicians wearing gloves, apron,
goggles and half-facepiece air-purifying respirators un-
loaded sample chests into freezers and refrigerators. This
ensemble of protection was worn during all work with the
samples in designated, restricted access laboratories. Later
individual samples were removed to a fume hood, opened,
an aliquot withdrawn for analysis and the remainder re-
turned to storage. All extractions, filtrations, etc. were
performed in laboratory fume hoods. Diluted wastes were
deposited in 55-gal drums and disposed as hazardous
waste.
CONCLUSIONS
Each waste site project has unique requirements for
personal protection. The basic approach to hazard control
planning presented here should have wide application. As
experience is gained in this area, more comprehensive
checklists and decision logics can be developed to facili-
tate hazard control planning.
The need to train personnel, rehearse proposed work
and critique work practices and protective systems cannot
be overemphasized. Well trained personnel appreciate
potential hazards and the limitations of protective sys-
tems. They will endeavor to minimize exposure and chal-
lenge to their protective equipment.
Field and laboratory supervision by qualified safety and
health professionals is vitally important. Both work situ-
ations demand full understanding of potential hazards and
protective options. Proper inspection and maintenance of
protective systems can only be assured by qualified per-
sons. Conditions at any site are subject to change. Meth-
ods to quantify changes in potential hazard are limited.
Persons who recognize the potential impact of changing
conditions and are willing to halt work until proper ac-
commodations can be made are indispensable.
Exploration of uncontrolled hazardous waste sites is
essential to development of remedial action plans. It can
be done safely with proper information, planning and ex-
ecution. Each waste site experience can contribute to im-
proved methods if properly evaluated at its conclusion.
-------
SAFETY PROCEDURES FOR
HAZARDOUS MATERIALS CLEANUP
ROBERT W. MELVOLD
STEVEN C. GIBSON
Rockwell International
Newbury Park, California
MICHAEL D. ROYER
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Edison, New Jersey
INTRODUCTION
EPA and contractor personnel at hazardous substance
cleanup operations face the risk of direct chemical ex-
posure or exposure to chemical-caused hazards such as
fire, explosion or oxygen depletion. Many of the perils of
entry, assessment and cleanup of hazardous waste sites
have been recognized and addressed by numerous authors,
but a number of important safety topics have yet to be
adequately covered.
This paper describes a project that: (1) identifies spe-
cific hazardous substance cleanup tasks that require im-
proved safety guidance and (2) provides monographs con-
taining the best available safety guidance for selected
topics. Two of the safety monographs produced as a re-
sult of the project are summarized. The monographs are:
(1) air quality monitors and (2) medical surveillance for
hazardous materials cleanup personnel.
IDENTIFICATION OF SAFETY TOPICS
The purpose of this project was to identify significant
safety topics related to hazardous materials cleanup that
had not been satisfactorily addressed by previous in-
vestigators or adequately covered in current or draft EPA
manuals. The topics identified were then reviewed and,
with the EPA Project Officer's concurrence, the most
worthy were researched and written up in the form of
monographs. The approach employed consisted of (1)
conducting an exhaustive survey of safety literature,
(2) evaluating EPA current or draft safety documents and
(3) soliciting comments from field personnel on potential
safety topics.
Safety Literature Survey
An extensive literature survey was conducted on safety
procedures, techniques and equipment used in the control
and/or cleanup of hazardous materials. In addition to
examining pertinent books, report, manuals and other
printed materials, there were numerous personal com-
munications. Staff members of government agencies, trade
organizations, professional associations and private com-
panies, who are knowledgeable in the area of controlling
hazardous materials or cleaning up uncontrolled hazardous
waste sites, were contacted.
EPA Safety Documents Review
Thirteen EPA draft and published safety guidance re-
ports were reviewed to develop a preliminary list of
safety topics that required additional attention (Table I).
A comparison matrix was prepared for the documents as
part of the reviewing process and a summary report was
written. Three of the manuals are clearly more complete
than the others:
(1) Personnel Protection and Safety Training Manual,
EPA Course No. 165.2 (A8)
(2) Hazardous Materials Incident Response Opera-
tions Training Manual, EPA Course No. 165.5 (A9)
(3) Hazardous Waste Site Investigation Training Man-
ual, (A10).
These manuals are collections of a number of mono-
graphs, reports,—sections of other manuals and ANSI,
OSHA/NIOSH, CFR documents as well as commercial
catalogs that were assembled for EPA training courses.
In fact, several EPA documents (or sections thereof) cited
in Table I can be found in Manuals A8, A9 and A10.
In addition to having produced the safety documents
cited in Table I, the EPA is involved in other efforts to
establish sound safety procedures for use during hazardous
waste site investigations and hazardous substance clean-
up activities. The EPA signed a Memorandum of Under-
standing (MOU) in December 1980 with the USCG,
OSHA and NIOSH. The MOU objectives include develop-
ment of a comprehensive guidance manual that will es-
tablish procedures to protect workers involved in investi-
gating and cleaning up hazardous waste sites and re-
sponding to emergencies involving hazardous substances.
Another EPA effort that will provide safety informa-
tion and guidance for worker health involves the prepara-
tion of 29 monographs on "Technical Methods for In-
vestigating Sites Containing Hazardous Substances." This
work is being sponsored by the EPA's Discovery and In-
vestigations Branch, Hazardous Sites Control Division.
Final Selection of Safety Monograph Topics
From the review of the cited EPA safety manuals/
memoranda and other available safety literature, a pre-
liminary list was assembled of safety topics to be empha-
sized in the study. The topics were:
•Air quality monitors
269
-------
270 SITE & PERSONNEL SAFETY
Table I.
Summary of Draft EPA Manuals/Procedures Relevant to Safety Procedures in Hazardous Materials Cleanup
Document Title
Prepared by
Status
Al. Agency-wide Policy Order 1440.1
Respiratory Protection
A2. EPA Occupational Health & Safety
(OHS) Manual, Chapters 1-7
A3. EPA OHS Manual, Chapter 8, Lab-
oratory Use of Toxic Substances
A4. EPA OHS Manual, Chapter 9, Haz-
ardous Waste Site Investigations
and Emergency Spill Responses
A5. EPA Safety Manual for Hazardous
Waste Site Investigations--
Draft, 1979
A6. Oil & Special Materials Con-
trol Division's Draft Interim
Safety Procedures
A7. Field Health and Safety Manual
A8. Personnel Protection and Safety
Training Manual for EPA Course
No. 165.2
Office of Occupational Health
and Safety (OOHS)
AA for Planning & Management,
OOHS
OOHS
OOHS
OOHS and National Enforcement
and Investigation Center
(NEIC)
Environmental Response Team
at the request
of the Superfund Implementa-
tion Task Group
U.S. EPA Region IV
U.S. EPA, Hazardous Response
Support Division and the
National Training and Opera-
tional Technology Center
Sent out for review by EPA sub-
units, 10/80
Issued to EPA employees on
September 12, 1977
Sent out for review by EPA sub-
units, 9/80
Sent out for review by EPA sub-
units, 9/80
Draft, 1979
Circulated to various EPA sub-
units for comment on January
2, 1981
Published October 1980
In use on recurring basis for
training course
A9. Hazardous Materials Incident
Response Operations Training
Manual, EPA Course No. 165.5
AID. Hazardous Waste Site Investi-
gation Training Manual
All. Hazardous Waste Site Safety
Procedures
A12. Memo on Update of OOHS/ERT
Activities
A13. Environmental Protection Agency
FY-1980-81, Medical Monitoring
Program Guidelines
U.S. Environmental Protection
Agency, Hazardous Response
Support Division and the U.S.
Coast Guard
Field Investigation Team (FIT)
National Project Management
Office, Ecology and Environ-
ment, Inc. under EPA Contract
No. 68-01-6056
Richard D. Spear, Chief
Surveillance & Monitoring
Branch, Region II, U.S. EPA
Robert C. Magor, Ph.D.,
Director, Office of Occupa-
tional Health & Safety
(PM-273)
OOHS
In use on recurring basis for
training course
Has been used in training
courses
Memo of 5/21/80 that outlines
Interim HWS Safety Procedures
This memo indicates level of
sophistication that has been
achieved in OOHS evaluation of
protective equipment (5/10/79)
Appended to AID
•Medical surveillance
•Criteria for selection of protective clothing
•Personnel and equipment decontamination
•Index of operational and procedural manuals (to simplify
location of appropriate safety information)
•Criteria for delineating the "hot line", and
•Odor detection for respirator canister failure (detection
of an odor typically signals either a leaking facepiece or
failure of the adsorbent canister.)
Based on the results of the survey of EPA and other
safety literature, it was determined that a monograph on
air quality monitors should be produced and work was
initiated. Due to time and budget constraints, some topics
could not be addressed in monographs. To prioritize the
remaining topics, comments were solicited from a selec-
tion of EPA Regional personnel and private contractors
with hazardous material handling and/or clean up experi-
ence and responsibilities. A synoptic matrix of their com-
ments and suggestions relative to the proposed mono-
graph topics appears in Table II. Two of the six sug-
gested topics—the index of operational and procedural
manuals, and medical surveillance—received a consensus
endorsement. Personnel and response equipment decon-
tamination received somewhat less support. The remain-
-------
SITE & PERSONNEL SAFETY 271
Table II.
Responses to Suggested Safety Topics
Topic
1
2
3
4
5
6
7
Topics:
A
Has reservations
about training
people to use
odor detection as
method, Desensiti
zation is a prob-
lem. Avoids
problem by using
fit test and
fresh canisters.
Hot line varies
with wind, con-
containment. NFPA
guidelines are
sufficient. Use
of programmable
Q
May be too dan-
gerous ; not too
excited about
this idea.
Every spil 1 is
different so
Also approaches
by government and
private compan-
ies vary. Not too
be nice. idea .
Full range of Uses almost ex-
clothing is cur- clusively throw-
rently employed by away clothing.
TATs. Development Doesn't think
of criteria is a sophisticated
more difficult criteria is re-
than it is worth. quired for his
region.
Must search a Believes this
multitude of mate- to be a good
rial s; should idea . Could in-
avai lable. Sympa-
thetic with
idea.
Uses annual pub-
vice checkups
for response
personnel .
Sample requisi-
tion by EPA staff
can result in
equipment contam-
ination; depot
decontamination
is suggested.
Interested in
idea.
Would like an
and for specific
materials not to
be flushed with
water.
of OHM-TADS and
list by
chemicals .
Doubts a medical
tical . Personnel
would not pay
proper atten-
Not much in-
volvement with
decontamination
and therefore
not much need for
this item.
Would like to see
expenditure for
resources to re-
spond to spil Is.
C
No need for study
of odor detection ;
fool ish to train
people.
Coast Guard HACS
program can be
distances for hot
1 ine and evacua-
tion within half
hour of input . Thi s
fore, redudant .
EXE has program
with NIOSH to de-
velop cri teria ;
therefore, prob-
ably not neces-
sary.
Uses computer
tie-in to OHM-TADS
to get ful 1 [irmt-
Also, TAT has I/?"
document from
4UCTR117. It
might be OK to
develop an index,
thou<|h.
Considers this
that should be
investigated.
Uses soap/water
washdown. Also uses
disposable cloth-
ing, whenever pos-
sible. Believes
further research
of topic is
indicated.
Interested in repre-
ing deteriorating
drums safely.
1 - Odor Detection of Respirator Canister Failure
2 - Criteria for Delineating the "Hot Line"
3 - Criteria for Selection of Protective Clothing
4 - Index of Operational and Procedural Manuals
D . E
Would like to see Foolish and imprac-
a cross-reference tical. S6me odors
chart or index for deaden the senses.
canisters. Not
excited about odor
detection.
Prefers to use
air moni toring
dispersion model-
ing can oe used for
spi 1 Is.
Assumes worst case
and uses "moon
suit" when hazard
unknown. Rarely
uses anything less
than splash pro-
tect ion , face
mask, hard hat and
Hyqienist makes
dec is ion on cloth-
ing protection.
Supports idea of
an index . Manual s
can be identified
various subject
tabula tions .
Bel ieves thi s
one.
Expressed no com-
ment on this
topic. (Propri-
etary considera-
tions may be
involved. )
Interested in
Also, likes info
on neutral ization
and gas/1 iquid
immobi 1 ization
for field opera-
tions.
test.
Has an SOP that
uses operational
parameters to prag-
matical ly del tneate
hot 1 me; not enthu-
puter program.
The currently used
4 levels of protec-
tion are adequate
not for cleanup.
Decision to down-
grade should be up
to hygipmst/
Only experience wi i 1
properly show the
advantages and dis-
many manual s .
determined by hygien-
ist/toxicologist for
easel ine and
fol lowup.
Kpcommends good med-
ical survei 1 lance
with depot-level de-
contamination and
No other topics
suggested.
r
Subjective; many
different odors,
many simi lar
Sensitization is
a factor. Some
chemicals do have
characteristic
odors ; perhaps
pilot study
called for.
Hot 1 ine is arbi-
trary in field; no
visible contamin-
ation oftentimes.
Safety manual for
s i te investiga-
teams . Not enthus-
iastic .
Uses clothing ap-
appropnate to
TLV, IDLH and
tors. Uses Noyes
chemical data
book , d repet i -
tion of CHRIS.
Considers present
techniques
adequate .
I ikes a modi f ica-
tion of this -
terse, brief ex-
manuals; abstracts
could be used.
tant, but not
often done, for
qui red.
Currently uses d
procedure for
scrub-down in con-
junction with wash
containment . Not
enthusvastic.
Woul d 1 i ke an
index for syno-
nyms and a cost
rate for disposal
of drummed wastes.
G
Not enthusi-
astic about
this idea.
Determination
of perimeter
should be
provided.
People do
not pay at-
tention to
clothing or
its selec-
tion until
it is too
late.
Johns Hopkins
University has
done some
work on this
subject.
Bel leves
search in
this area is
Bel ieves
this topic
requires
further
No other
topics
suggested.
Consensus
Negative
Mixed
Mixed-
Negative
Posi ti ve
Positive
Mixed-
Positive
5 - Medical Surveillance
6 - Personnel & Response Equipment Decontamination
7 - Others (to be suggested by interviewees)
ing topics received mixed reviews, except for odor detec-
tion of respirator canister failure, which received a uni-
formly negative response.
With the aid of the interviewees' comments, the final
list of safety monographs was developed. The list includes:
(1) Air Quality Monitors
(2) Medical Surveillance for Hazardous Materials
Cleanup
(3) Selection of Personnel Protective Clothing for
Hazardous Materials Cleanup
(4) Personnel & Equipment Decontamination
(5) Index of Operational Manuals for Hazardous Ma-
terials Cleanup
The "hot line" criteria topic received a positive response,
but it was not addressed by this project since another
EPA project* is investigating criteria for evacuation dis-
tances for hazardous substances derailment incidents.
The "hot line" is directly related to evacuation distances
and it was decided to address the topic only once, in
the other project.
*EPA Contract No. 68-03-3014, Task 10 (Phase II), TMS-2, is develop-
ing a Fire Control Handbook for Hazardous Substances.
-------
272 SITE & PERSONNEL SAFETY
Two safety monographs produced by this effort and
further described in abridged fashion in the following sec-
tions of the paper are Air Quality Monitors and Medical
Surveillance for Hazardous Materials Cleanup.
AIR QUALITY MONITORS
Combustible and/or toxic gases or vapors and oxygen-
deficient atmospheres may be encountered with lethal re-
sults if proper protection is not provided. Hazardous ma-
terials cleanup personnel must ascertain accurately and
efficiently what atmosphere conditions prevail in order to
take safe and corrective action. This responsibility re-
quires that response personnel use portable monitoring
devices that are extremely reliable.
A search of available product literature, however, re-
veals a bewildering array of different air quality moni-
tors, employing various principles of operation, packaged
in many sizes and shapes, with a host of operating char-
acteristics and selling for a variety of prices. To assist re-
sponse personnel in selecting an appropriate air quality
monitor, an evaluation of commercially available monitors
was conducted. The price of the air quality monitors was
not a consideration in the evaluation.
Only those devices capable of monitoring hazardous
gases or vapors on a continuous and quantitative basis
were considered. Thus, detectors that indicate the presence
of a substance by discrete sampling were eliminated,
whereas monitors that possess the ability to quantita-
tively sense fluctuations in gas or vapor concentrations
over a period of time were included.
Evaluation
A desk-top analysis was conducted of commercially
available devices for the monitoring of combustible
and/or toxic gases or vapors and oxygen levels. Thirty
basic design configurations, representing models cur-
rently produced by 18 different manufacturers, were
evaluated.
The gas-phase monitoring devices were segregated into
three categories characterized by the gas system being
monitored: (1) oxygen, (2) combustible gas or vapor and
(3) toxic gas or vapor (as represented by H^S and CO). The
toxic gas or vapor category was limited to H2S and CO
monitors for two reasons. First, many toxic gas monitors
are capable of detecting numerous gases. A complete desk-
top evaluation and comparison of all monitor/gas com-
binations was considered unworthy of the time and cost it
would require. Second, a fair comparison of toxic gas
monitors could only be made between monitors for which
there was performance data on the same gas(es).
Unfortunately, there was no gas for which performance
data were available for all toxic gas monitors. However,
data was available for FL,S and/or CO for each monitor,
so the toxic gas monitors were evaluated in two groups
depending upon whether CO or RS monitoring data was
available for the instruments. Those instruments for
which performance data existed for both CO and H2S were
evaluated in both groups.
Fifteen instrument parameters were identified as the
most important for the intended use of the monitors. A
tabulation of the 15 specifications and qualifications, in-
cluding solid-state electronics and the degree of skill re-
quired for operation, for each combustible gas vapor mon-
itor are given in Table III. Similar tables were prepared
for the toxic gas/vapor (IL,S and CO) and oxygen moni-
toring devices.
The evaluation was accomplished by applying to each
specification or qualification a factor that relates the im-
portance of that particular parameter to the features de-
sired in the final product. Thus, rating criteria were es-
tablished to reflect differences in the various devices and
weighting factors were assigned to each particular specifi-
cation so that each device was rated in accordance with
the criteria. (See Table IV for evaluation rating criteria.)
Conclusions
With the device specifications and qualifications identi-
fied and the evaluation criteria established, individual rat-
ings were determined for each device specification. By
summing all the individual ratings for a particular device,
an overall rating for the device was obtained. Overall
ratings for the top-rated devices in each of the four de-
vice categories are as follows. Among oxygen monitors,
the Dynamation O2-25, Energetic Ecolyzer 400 and Bio-
marine OM325R all received the highest ranking with
identical scores of 245. Ranked second with a score of 235
was the Gastech Model 1641.
Among combustible gas monitors, the Century OVA-
128 was ranked highest with a score of 258, followed by
the Dynamation LCD-Combo with a score of 240. Rating
totals in the H2S toxic gas monitor category show high
values of 225 for the HNU's PI-101 and 214 for the In-
terscan Model 1170. For the CO toxic gas monitor cate-
gory, the Enmet COS-100 was highest rated with 219,
while the Gastech EC-231 was second with 216.
Presumably, any of the highly rated devices can be used
for the specific monitoring application for which it was
intended. However, it must be remembered that the desk-
top evaluation relied almost exclusively on manufacturer's
product literature and on telephone conversations with the
manufacturers or their representatives. An independent ex-
perimental and/or field evaluation of the various moni-
tors would provide better data on which to base the evalu-
ation. It is quite possible that monitor ratings developed
from first-hand field evaluation could differ from those
derived from a desk-top or literature evaluation. More-
over the assembled information is dated and therefore sub-
ject to change with time.
MEDICAL SURVEILLANCE FOR
HAZARDOUS MATERIALS CLEANUP
A monograph has been produced describing a medical
surveillance program for hazardous materials cleanup per-
sonnel. The program is a test protocol designed to super-
vise the health and well-being of the hazardous material
cleanup team members. Medical surveillance programs for
hazardous materials cleanup personnel are directed at
-------
SITE & PERSONNEL SAFETY 273
Table III.
Combustible Gas/Vapor Monitors—Operating Specifications
Instrument* Vol umey Setup Temp . Opera t1 no
(Operating Weight Time Range Range (s) Accuracy
Principle) (in3/lbs) (win) {°C) (ppm) (X F.S.)
GasTech Inc.
Johnson
Instrument
Division
Ho del 1 314** 248/8 1 -8 to 0-1 001 +10
(CC) 43 LEL
Toluene
0-500 ppm
Organic
Vapor (OA)
Model 1641**'* 344/8 1.5 0 to 0-100% *5
(CC) 40 LEL
Methane
Dynamation 84/3 <2.5 -9 to 0-100% +2
Inc. LCD 49 LEL
Combo" (CC) Hethane
Enmet Corp. 224/6.5 10 -10 0-50*. *3
CGS-100*" to 50 LEL
(SS) Methane
(OA)
Century Sys- 425/12 1 0 to 0-10 ppm +1
terns Corp. 55 0-100 ppm
OVA-128 , 0-1000 ppm
(GC/FIOT
Infrared 227/9 5 0 to 0 100* +5
Industries, 50 LEL
Inc. IR-711 propane.
(IR)I1 0-1000 ppm
JP-5
International 167/5 0.5 o to 0-100; LEL *5
Sensor Tech- 40 Methane
nology
AG6000** (SS)
Scott Avia- 126/5 Instant 0 to 0-100'. LEL *10
tion/Div. of 43 Hexane
A-T-0 Inc.
Model 0-17 (CC)
Energetics 432/8 1 -18 0-lOOi LEL +1
Science/Oiv. to 40 Methane
of Bee ton
Dickinson
S Co. Eco-
lyzer 400**
(CC)
Appliances 40 Pentane
Co. Model
260" (CC)
Biomarine 561/9 5 < 5 15 0 100% LEL +5
Industries, to 40 Hexane
Inc. Model
902" (CC)
* All devices are of solid-state construction and require
bustion, SS • Solid State, GC/FID = Gas Chromatography/F
** Also equipped to monitor either toxic gases or oxygen.
"* Also equipped to monitor both toxic gases and oxygen.
/ This principle of operation is sensitive to the C-H bond
gases/vapors.
t A/V/F = Audio/Visual/Failsafe; Bat • Battery Check; Del
1 Not determined by independent assessment.
OA - Others available.
Minimum
Detection .
(I F.S.) Alarms'
< 1 A/V/F
(Bat,
Del)
2 A/V/F
(Bat,
Det)
1 A/V/F
(Sat,
Det)
<1 A/V/F
(Bat)
< 1 A/F
(Bat,
Det)
25 A/F
(Bat)
3 A/V/F
(Bat)
5 A/F
(Bat)
1 A/V/F
(Bat,
Det)
(Bat)
(Bat)
average skill on the
lame lonization Dete
and, therefore, the
Detector Check.
Intrln- Calibration
slcally Medium/
Safe Time (min)
5-10
Yes " Methane/
5-10
,
5-10
Yes Hethane/
5-10
Yes Methane/
5-10
JP-5/5-10
Yes Methane/
5-10
Yes Hexane/
5-10
Yes Methane/
5-10
5-10
5-10
part of the operator
ctor, IR = Infrared A
corresponding device
Response
Time,
Interferences (sec
cones, lead
vapors
Argon, 'hel ium 4
{filter will stop
most)
None (while de- 30
tecting total
comb. )
None (while 2
detecting total
comb . )
ing CH bond
H2, Methyl 20
Mercaptans
Poisons - sili- 5-10
ones, lead
apors
one (while de- 15
ecting total
omb. ) Poisons
ilicones, sili-
ates, lead
apors
cones , si 1 icates ,
lead vapors
cones, lead vapors
Operating Principle code
isorption.
will respond to most, but
901 Operating Remote
) Period (hrs) Readout
8 No
6 0-1 V
9 0-2 V
10-14 No
8 0-5 V
58 0-100
mv
10 No
4-6 No
10 No
8-9 No
°
: CC = Catalytic Corn-
not al 1 , combustible
employment qualifying, general health maintenance and
prevention or early detection of harmful effects of haz-
ardous substances. In the case of adversely affected mem-
ber^) of the cleanup team, this may require medical in-
tervention and removal from the cleanup team.<2) The ob-
jective of this monograph is to describe and document
the elements that comprise a reasonable medical surveil-
lance program for hazardous materials cleanup personnel.
The described elements are based on state-of-the-art medi-
cal methodology and the proposed medical surveillance
program is expected to be tailored to specific applications
as required by the user organization. Also, medical sur-
veillance programs should be designed to secure maxi-
mum preventive benefits at minimal cost and incon-
venience.
A medical advisory board sponsored by the organiza-
tion responsible for the hazardous materials cleanup
should have the responsibility of developing a medical
surveillance program and evaluating its results. The
board should be empowered to make recommendations
based on its review of the examinations performed on
personnel.
Two examinations that are commonly performed'2' are:
•Preplacement
•Periodic
Preplacement/Pre-Employment Examinations
Preplacement examinations serve an essential function
in health surveillance by providing an historical record of
-------
274 SITE & PERSONNEL SAFETY
previous exposures, information on the state of health
prior to joining the team and a baseline for comparisons
with later health observations. Preplacement examina-
tions are used to ensure that workers are physically able
to use personal protective equipment.12' Employment and
medical history and a physical examination consisting of
physical and biological monitoring are elements of the
preplacement examination and should be tailored to the
specific hazards of the job under consideration.
Periodic Examinations
A periodic examination is a tool that can be used to
detect incipient disease, physiological changes, biochemi-
cal deviations or evidences of absorption of toxic agents
and to establish interim reappraisals.<3) Periodic examina-
tions consist of periodic baseline examinations (compre-
hensive) and hazard-oriented surveillance examinations
(limited).
The periodic examination provides two primary services:
(1) maintains a continuing record of the general health
status of the team members and (2) provides an epidemi-
ological record, followup analyses and/or examination for
possible latent effects of exposure to some hazardous sub-
stance^).
Periodic examinations would be similar to pre-employ-
ment/preplacement examinations. However, the em-
phasis would be on possible exposures that had occurred
since the previous examination. An example of the physi-
cal/biological parameters that a periodic examination
could investigate'4' for possible exposure of a response
team member to selected hazardous substances is shown in
Table V.
The hazard-oriented surveillance examination is de-
veloped to screen for exposures that may occur during a
hazardous material cleanup operation, such as at an un-
controlled hazardous waste site. Hazardous materials
cleanup personnel may be exposed to toxic chemicals that
may or may not have been identified from the onset of the
response action.'2' Based on the premise that hazardous
materials response team members could be operating in
situations where the hazards are unknown, three points
should be made:
(1) It is not economically feasible to test the team mem-
bers for all hazardous substances if they were ex-
posed to unidentifiable numbers and types of chem-
icals in question.'2'
(2) There are no general tests that can be used to screen
for organics or inorganics generically. The tests are
chemical-specific and the biological pathway ex-
amined would be dependent on the type of chemical
in question.(2)
Table IV.
Evaluation Rating Criteria
Vol ^/Weight Setup Time
in Ibs Solid State (mm) Sk
0-750 15
751-1500 13
1501-2250 11
2251-3000 9
3001-3750 7
>3750 5
£N '••
16-30 w
31-45 No
46-60
>60
Minimum Detection
°1
0-0. St 10
0.6-1. OS 5
Oxygen
None
Interferents
Interferents
Toxic i Combust
GC/FID
Combustibles
2'< LEL 4
20
•2500 ppm 10
* 2500 ppn- 0
ible Gases*
30
None stipulated, .ery
specific if
1R absorption
muroOroceSSO
(Used bv the
SO)
gas tknown.
with
r 30
HI RAN-
Temperature Range
Lower Upper Operational Accuracy
lill Limit (°C) Limit (°C) Range(s) (* F.S.)
IS 0-10 15 Average <-30 15 >70 >2 Ranges 15 < +2 20
11-20 10 Ability 10 -20 to -30 12 60 to 69 2 Ranges 10 +3 to +5 15
0 21-30 5 -10 to -19 9 50 to 59 1 Range 5 +6 to +10 10
31-60 1 Greater 0 to -9 6 40 to 49 > +10 5
>60 0 Technical >0 0 < 40 Add 5 if factory
Ability 0 allows options
for range(s).
Alarms
H-S and CO Audio
1 ppm 10 Fai Isafe:
1-2 ppm 6 Detector
•> 2 ppm 4 Battery
Interferences
1R absorption 22
Hydrocarbons with sim-
ilar frequencies to
the material being
mom to red
Colorimetnc 22
Strong reducing
agents
Ctic combustion 20
Si 1 1 cones , lead
vapors, (poisons.)
Sol id state 17
Marcaptans, H^
Intrinsi- Calibration Procedure
5 cally Safe Contained gas
5 Y ?fl unnecessary 14
Direct cylinder
5 . attachment to
5 N° ° monitor 12
Calibration gas
introduced wi th
bag and cyl inder 10
Cal ibration gas
introduced with
syringe and
cylinder 8
Response
Time (Sec)
Photo ion izat ion 10 0-10 30
Many organics (mate- 11-20 24
rials with lonization 21-30 18
potential equal to or 31-60 12
less than the energy 61-90 6
level of the UV source) >90 0
Electrochemical 5
Unsaturated hydro-
carbons, C?HrOH, H,,
NO ° £
Calibration gas
introduced with
syringe and
glass container
Cal ibration gas
introduced with
permeation tube
and instrument
Cal ibration
gases plus ex-
ternal computer
for computations
Operating
Period (Hr)
>16 25
12-16 20
8-11 15
4-7 10
- 4 5
Calibration
Time (min)
0-5 6
6 6-10 5
21-30 3
2
Remote Readout
2 Ranges 10
1 Range 5
Because of tie «ide disparity in product information on Interferents. allowance has been made for Interferents generally associated with
specific principle o< operation by subtracting tne effect of the Interferents from the manmum allowable (30) according to the following
schedule major single interferent. -5. small interferent family. -8; large interferent family, -10.
-------
SITE & PERSONNEL SAFETY 275
Table V.
Biological Parameters Affected by Selected Hazardous Substances4
s-
o
-p c
CO O
• I—
in co
-o. — .
i— c ai
ro <1) -p
c a. S-
•i- "O
4-J — - Ol
ro S-
Q. E =!
13 rO CO
U X O
O LU Q.
O X
~^ U Hi
fO "O i—
U O ro
•1- •!- =3
-a s_ c
ai 01 c:
S: D. ro
Asbestos x x
Beryllium x x
Carcinogens (other x x
classified)
Lead x x
Ozone x x
p-tert-Butyl toluene x x
Styrene x
Toluene x x
Trichloroethylene x x
Vinyl Chloride x x
c
o
•r- O)
-P U
ro c
U ro
• r- E
<+- s-
•r- o
4_> 4 —
i- i.
QJ C1J
(_> >, D_
ro
C S- >>
ro 1 i-
•r- X ro
O c
•i- -P 0
co co E
>, CD O) r-
(— "^/ r~ -^
Q. LU O Q.
X XX
X XX
X
X
X XX
X X
X
X X
XX X
X XX
>^
en =-,
o en
i — O
0 r—
10 -P O
•r- >, -P
CO 0 >,
to >, <_>
E i— E
x c: -P c
LU 'r- 3 •!-
i- Q. i-
-Q rD oo ^
ro
-1
X
X X
X
X
X
X
01 ^
co O)
ro CO
f~ (O
-P (J •!-<_> C Ol
C '1— E T--I- CO
3 E ro E E ro
O ro t/l ro ro 4-^
CJ •• -p c -P CO ro
CO 3 ro 3 C -C
TD O) i — 1- i — ro D-
O -r— CD | — CD S- CO
OS- h- 0
r- -P E <-> E -C
CQ 00 I3-i- 3O Q_
•r- i- -P S- -r-
QJ E dj QJ d) i_ QJ
+J Ol 00 O OO O C
Oi -c > — - ro • — -.c: •!-
i— O O Q. ^~
D. |— r— (— oo ro
E "O O ro D- O -^
O O CD x CD -C i—
O O 00 O OO Q. ec
,^
CO
X XXX
X
X
X
X
XXX
o
E
rO
•r- Z3 O)
-0 1—00
3 CD ro
S- "O
•I- O •r-
i— E -P
•r- ro Q.
CQ CD Oj
* * Q_
1 CO
ro 0. C -0
•P 1— ro ro
0 CD S- Ol
r— CD r— _l
X X
X
X X
Table VI.
Health Effects and Medical Surveillance for
Selected Airborne Contaminants (Partial List)3
Airborne
Contaminant Health Effects
Acetone Irritation of respiratory tract
and mucous membranes;
narcosis; dermatitis
Alumina Lung overload; no fibrosis
Ammonia Irritation or respiratory tract
and mucous membranes
Antimony Gastrointestinal and heart
Pentachloride disorders; respiratory irritation
Asbestos Fibrosis of lung; bronchogenic
carcinoma; mesothelioma;
cancer of stomach, colon and
rectum
Bauxite (may Lung overload
contain silica)
Beryllium Fibrosis of lung; dermatitis
Medical
Surveillance
Interval history
Interval history
Chest X-ray; lung
function
EKG; Chest X-ray;
lung function
Chest X-ray; lung
function; sputum
cytology (OSHA re-
quirement — annual
medical surveillance)
Chest X-ray; lung
function
Chest X-ray; lung
function; skin in-
inspection
Airborne
Contaminant
n-Butyl alcohol
Butyl Cellosolve (2-
Butoxy Ethanol)
Cadmium Oxide
Carbon Dioxide
Carbon Monoxide
Health Effects
Irritation of respiratory tract
and mucous membranes;
narcosis; dermatitis
Irritation of respiratory tract
and mucous membranes;
narcosis; dermatitis; red cell
hemolysis
Pulmonary edema; fibrosis of
lung; kidney damage
Asphyxiation
Asphyxiation; disturbed
consciousness
Medical
Surveillance
interval history
Complete blood count
Chest X-ray; lung
function; cadmium
and protein in urine;
urinalysis
None
Interval exam; car-
carboxyhemoglobin
if intoxication
suspected
-------
276 SITE & PERSONNEL SAFETY
(3) Field monitors can be used to evaluate the type and
concentration of chemicals on site that the team
might be exposed to indirectly.
Personal monitoring and area monitoring (manual and
automatic) are the most common techniques used to de-
termine an exposure to atmospheric chemicals. However,
both are subject to significant errors. Biological monitor-
ing may be useful to validate the atmospheric monitoring.
Also, exhaled air analysis can be used to measure ex-
posure to industrial chemicals.(3> A partial list of potential
health effects, type of medical surveillance and industrial
hygiene considerations required for several specific air-
borne contaminants is given in Table VI.<3) The usual ap-
proach in determining that a chemical insult to the body
has occurred due to exposure to a hazardous material in-
volves medically observing changes, either from the norm
or a previous baseline, in a person's health.
During a response action, response team members
should be carefully observed by the on-scene coordinator,
on-site industrial hygienist or medical personnel and fel-
low team members. The first warning signs of toxic ef-
fects from a hazardous exposure are often observed by a
fellow worker.(5) Finally examinations on the termination
of employment are desirable since they document the
health status then and provide evidence of any changes
that have occurred during the employment period. This
does not rule out the possibility that effects may show up
at a later date.
ACKNOWLEDGEMENT
This project was sponsored by the U.S. EPA's Oil and
Hazardous Materials Spills Branch, Municipal Environ-
mental Research Laboratory-Ci, Edison, New Jersey,
under Contract No. 68-03-3014 with Rockwell Interna-
tional. The content of this publication does not necessarily
reflect the views or policies of the U.S. Environmental
Protection Agency, nor does mention of trade names,
commercial products or organizations imply endorsement
or recommendation by the U.S. Government or Rockwell
International.
The authors express their appreciation to EMSC staff
members Mr. Fitz Bush, Mr. David Cabrey, Ms. Patricia
Casey, Dr. George Schneider, and Ms. Patricia Scofield,
for their assistance to the project. Also, the authors
acknowledge the assistance rendered by Ms. Debra Sibert
and Mr. Eugene Port of Health Science Associates of Los
Alamitos, California.
REFERENCES
1. Pritchard, J. "A Guide to Industrial Respiratory Pro-
tection," LA-GG71-M, UC-41, Los Alamos Scientific
Laboratory, Los Alamos, New Mexico, March 1977.
2. Melius, J. and Halperin, W., "Medical Screening of
Workers at Hazardous Waste Disposal Sites," pre-
sented at the Society for Occupational and Environ-
mental Health Conference on Hazardous Waste,
Washington, D.C., December 7-10, 1980.
3. Galley, L.J. and Cralley, L.V., "Patty's Industrial
Hygiene and Toxicology, Volume III, Theory and Ra-
tionale of Industrial Hygiene Practice," John Wiley
and Sons, New York, N.Y., 1979.
4. NIOSH Pilot Study for Development of an Occupa-
tional Disease Surveillance Method, published May
1975.
5. Proctor, N.H. and Hughes, P., "Chemical Hazards of
the Workplace." J.B. Lippincott Company, Phila-
delphia, Pennsylvania, 1978.
-------
INITIAL SITE PERSONNEL
PROTECTION LEVELS BASED ON TOTAL VAPOR READINGS
RODNEY D. TURPIN
U.S. Environmental Protection Agency
Environmental Response Team
Edison, New Jersey
INTRODUCTION
The EPA Environmental Response Team (ERT) was
established in October 1978 to provide technical assistance
to On-Scene Coordinators (OSC), Regional Response
Teams (RRT), EPA Headquarters and other Regional
Offices, as well as other governmental agencies at spills of
hazardous materials and at uncontrolled hazardous waste
sites. In this paper, the author describes the procedures
used by ERT in selecting the appropriate level of personnel
protection during the initial assessment and operational
phase. Personnel protection is categorized into three levels
and criteria for selection is based on potential total vapor
exposure.
The objective of using total atmospheric gas/vapor con-
centrations for determining the appropriate personnel pro-
tection level is to provide a numerical criteria for selecting
Level A, B, or C protection. Until atmospheric contam-
inants are specifically identified and personnel protection
selected based on lexicological properties, total gas/vapor
concentrations provide a numerical value that can be used
as a guide for selecting personnel protection equipment.
Although total gas/vapor concentration measurements
are useful to a qualified professional for the selection of
protection equipment, caution should be used in the inter-
pretation of this data. The response of an instrument to
several gas/vapor contaminants does not provide the same
sensitivity as measurements involving a single contam-
inant. Since total vapor field instruments see all contam-
inants in relation to a specific calibration gas, the concen-
tration of an unknown environment may be over esti-
mated or under estimated.
When carcinogens or other highly hazardous substances
are suspected, the protection level should not be based
solely on the total gas/vapor criteria, rather the level
should be selected on a case-per-case basis weighing heav-
ily on potential exposure and chemical characteristics of
the suspected material.
FACTORS FOR CONSIDERATION
In utilizing atmospheric gas/vapor concentrations as a
guide for selecting a level of protection, a number of fac-
tors should be considered:
1. The uses, limitations, and operating characteristics of
the monitoring instruments must be recognized and
understood. Instruments such as the photoionizer and
Organic Vapor Analyzer (OVA) do not respond to all
substances that may be present or may respond differ-
ently to identical substances when compared to one an-
other. Therefore, experience, knowledge, and good
judgment must be used to compliment the data obtained
with this instrumentation.
2. Hazardous other than detectable toxic gases/vapors
such as non-detectable gases (i.e., phosgene, HCN,
chlorine, etc.,) liquid/solid particulates, explosives,
combustibles, radiation, oxygen deficiency, and other
harmful conditions may exist in the atmosphere or on
the site.
3. The risk to personnel entering an area must be eval-
uated against the need for entering. Although this as-
sessment is largely a value judgment, a conscientious
assessment of the variables involved and the risk to
personnel must be balanced against this need for site
entry.
4. The knowledge that carcinogens or other highly toxic
substances are involved or suspected requires that gross
levels not be the sole factor in determining the level of
protection. Other factors which must be taken into con-
sideration are exposure, chemical characteristics of
known/suspected materials, instruments, weather con-
ditions, etc.
5. Functions which need to be performed on-site must be
evaluated. Based upon total atmospheric gas/vapor con-
centrations, Level C protection may be judged ade-
quate. The work functions to be performed such as
moving drums, opening containers, bulking of materi-
als and other operations that increase the probability
of exposure may require a higher level of protection.
CRITERIA
The criteria for relating Levels of Protection to tptal
atmospheric concentrations are:
Level C 0-5 ppm above background
Level B 5-500 ppm above background
Level A 500-1000 ppm above background
Level A Protection
Level A protection provides the highest degree of pro-
tection against hazards due to inhalation, skin and eye
277
-------
278 SITE & PERSONNEL SAFETY
irritation, if the inherent limitations of the personnel pro-
tective equipment are not exceeded. Yet Level A protec-
tion does present the highest potential for heat stress.
Cooling vests/suits have demonstrated some degree of re-
lief to body temperature build-up. The range of 500 to
1000 ppm total gas/vapor concentrations in air is based on
the following criteria:
•Since Level A will provide protection against air concen-
trations greater than 1000 ppm for most substances, an
operational restriction of 1000 ppm is established as a
warning to:
•Take into consideration analytical instrumentation sen-
sitivity to wind velocity, humidity, temperature, etc.
•Evaluate the calibration and/or sensitivity error asso-
ciated with the instrument(s).
•Evaluate the need to enter environments greater than
1000 ppm.
•Identify the specific constituents contributing to the
total concentration and their associated toxic properties.
•More precisely determine the concentration of the in-
dividual constituents that make up the total concentra-
tion.
•A lower limit of 500 ppm (for Level A) total gas/vapor
concentration in air is selected as the value to upgrade
from Level B to Level A in order to fully protect the
skin and/or eyes until qualitative and quantitative de-
terminations can be made of the constituent products
and skin and/or eye hazards excluded.
•The range of 500-1000 ppm total concentration is suffic-
iently conservative to provide a safe margin of protec-
tion due to instrument error, calibration and sensitivity,
unanticipated transient concentrations, and protection
against highly hazardous substances that could account
for the total concentrations.
Experience with properly operating portable field equip-
ments for measuring total gases and/or vapors has demon-
strated levels approaching 500 ppm have not routinely
been encountered on hazardous waste sites. High con-
centrations have only been encountered in closed build-
ings, at openings to containers or when working in the
spilled contaminants.
A decision to require Level A should also consider the
negative aspects of this high level of protection, i.e.,
higher probability for accidents due to cumbersome equip-
ment, increased resources needed and the physical stress
caused by heat buildup in fully encapsulating suits.
These factors need to be carefully evaluated and balanced
against the reasons for utilizing Level A.
Level B Protection
Level B protection is the minimum level of protec-
tion required for initially entering an open site where the
type(s), concentration(s) and presence of airborne gas/
vapors are unknown. This level of protection provides a
high degree of inhalation, skin and eye irritation or ab-
sorption protection. Although a small portion of the body
(neck and head) are exposed, primarily liquid/solid ma-
terials have the highest potential of causing acute or
chronic effects due to exposure of this area. The use of a
hooded chemical resistant jacket would further reduce the
potential for exposure to this area of the body. Although
the potential for heat stress is not as high as Level A, it
can be a major problem. Cooling vests/suits have demon-
strated some degree of relief to body temperature build-up.
A limit of 500 ppm total atmospheric gas/vapor con-
centration readings on portable field instruments has been
selected as the upper restriction on the use of Level B.
Although Level B personnel protection would provide ade-
quate protection against most substances at concentra-
tions higher than 500 ppm, an upper limit of 500 ppm is
selected as the decision point for a careful evaluation of
the risks associated with higher concentrations.
Considerations:
•Analytical instrumentation sensitivity to wind velocity,
humidity, temperature, etc.
•Calibration and/or sensitivity error associated with the
instruments.
•The probability that substance(s) present are cutaneous or
percutaneous.
•The necessity for entering higher concentrations in Level
B.
•The work function to be done and the increased proba-
bility for exposure.
•Qualitative and quantitative identification of the specific
components.
Level C Protection
This level provides the same high degree of skin protec-
tion as Level B, but lesser inhalation and/or eye protec-
tion. A relatively low (0-5 ppm above background) am-
bient concentration has been established as the range for
wearing Level C protection equipment. An upper limit of
total vapor concentration of 5 ppm (above background)
has been selected primarily based on the use of a full-face
air purifying gas mask with canisters and requirements as
to its constraints and limitations. These are:
•MSHA/NIOSH approved air-purifying devices should
only be worn in atmospheres where the substances have
been identified or the potential of exposure is highly un-
likely.
•Substances must have good warning properties.
•Continuous air monitoring must occur for the atmos-
pheric contaminants identified.
•Appropriate, approved canisters must be used.
•Sufficient oxygen (at least 19.5% of air at sea level) must
be present.
Full-face, air purifying devices will provide respiratory
protection against most environmental vapors greater than
5 ppm; however, until qualitative and quantitative infor-
mation is available about the substances, concentrations
greater than 5 ppm indicate a higher level of respiratory
protection should be used.
PROTECTION EQUIPMENT
Since exposure factors vary from one situation to an-
other, it is impossible to develop a specific safety pro-
cedure that will be appropriate for all situations; ERT
-------
SITE & PERSONNEL SAFETY
279
limited the selection of protective equipment to three gen-
eral categories. Although total gas/vapor monitoring is the
primary screening mechanism used, the protection levels
indirectly address protection against three possible routes
of exposure (inhalation, skin absorption/irritation, and
ingestion).
Level A—Personal Protection Equipment
•Positive Pressure SCBA (MSHA/NIOSH approved) op-
erated in the positive pressure mode
•Totally Encapsulating Suit (boots and gloves attached,
cooling vest when applicable)
•Gloves—Inner (tight fitting and chemical-resistant)
•Boots—Chemical-protection, steel toes and shank. De-
pending on suit boot construction, worn over suit boot
•Gloves—Outer, chemical resistant. Depending on suit
construction, worn over suit gloves. May be replaced with
tight fitting, chemical-resistant gloves worn inside suit
gloves.
•Underwear—cotton, long-John type
•Hard Hat—(under suit)*
•Disposable protective suit, gloves, and boots. (Worn un-
der or over encapsulating suit)
•Coveralls (under suit)*
•Two-way radio communications
"Optional
Level B—Personal Protective Equipment
•Positive Pressure SCBA (MSHA/NIOSH approved),
operated in the positive pressure mode
•Hooded, two-piece chemical-resistant suit (cooling vest
when applicable)
•Gloves—Inner, tight fitting, chemical-resistant
•Boots—Outer (chemical-resistant, heavy rubber dispos-
ables)
•Boots—Inner (chemical-resistant, steel toe and shank)
•Two-way radio communications
•Hard Hats*
•Face Shield*
*Optional
Level C—Personal Protective Equipment
•Full-face gas mask, air-purifying respirator (MSHA/
NIOSH approved)
•Chemical-resistant clothing
•Overalls and long-sleeved jacket or coveralls; hooded
two-piece chemical splash suit (when applicable—hooded
disposable coveralls)
•Gloves—Outer (chemical-protective)
•Gloves—Inner (tight fitting, chemical-resistant type)
•Cloth or disposable coveralls—inside chemical protective
clothing, cooling vest when applicable*
•Escape pack
•Hard Hat* (face-shield, optional)
•Boots—Outer (chemical-protective heavy rubber throw-
aways)
•Boots—Inner (chemical-protective, steel toe and shank)
•Two-way radio communications
* Optional
ACKNOWLEDGEMENT
The author expresses his appreciation to all members of
EPA's Environmental Response Team, especially Tom
Sell, Training Course Director, for their many contribu-
tions and his indebtedness to David Weitzman, EPA, Of-
fice of Occupational Health and Safety, Washington,
D.C., Richard Costello, NIOSH, Hazardous Evaluation
and Technical Assistance Branch, Cincinnati, Ohio and
David Dahlstrom and David Schafer, E&E, ERT-TAT,
Cincinnati, Ohio for their critical review and comments.
-------
HAZARDOUS SUBSTANCE SITE AMBIENT AIR
CHARACTERIZATION TO EVALUATE ENTRY TEAM SAFETY
MARTIN S. MATHAMEL
Ecology and Environment, Inc.
Chicago, Illinois
INTRODUCTION
Because of the wide variety of ambient conditions that
may be encountered in investigation and remedial action
activities on hazardous substance sites, a means of de-
termining the potential dangers is essential in order to pro-
tect worker health and safety. A number of airborne toxic
chemicals have been found at the boundaries of hazardous
substance sites (Table I).(1)
The air emissions from hazardous substance sites are
complex because there is usually no single point source on
a site and because concentrations of toxic chemicals de-
crease rapidly as they diffuse and disperse in the atmos-
phere."1 In addition, explosive or oxygen-deficient atmos-
pheres may exist, and radiation may be encountered.
Depending on the nature of the site, the information
necessary to assess the hazard potential may be obtained
by researching existing data sources. However, in many
cases, a variety of equipment may be required to character-
ize and monitor the ambient air, both on-site and at the site
boundaries. The goals of such an ambient air characteriza-
tion scheme are as follows:
(1) To qualitatively and quantitatively define the haz-
ards due to airborne pollutants on hazardous sub-
stance sites
(2) To locate specific areas of a site upon which to focus
the investigation and remedial action activities
(3) To monitor the surrounding community for any ad-
verse environmental impact that work on the site
may cause
The data obtained from the air characterization study is
used to develop a personnel protection plan that allows
workers to perform site activities in the lowest level of pro-
tective gear consistent with maintaining their health and
safety.
In this paper, the author discusses the equipment that
can be used to define the hazards associated with hazard-
ous substance sites. The actual selection of specific per-
sonnel protective gear based on the analytical data gen-
erated by the air characterization is beyond the scope of
discussion but several references detailing selection criteria
are included.
PRELIMINARY HAZARD ASSESSMENT
Before work is initiated at a hazardous substance site,
detailed background research should be conducted to as-
certain what chemicals, hazards, special requirements,
etc., may be associated with the site. Possible sources of
information are state, county and municipal agencies, in-
cluding the federal and state EPA.
All site waste generator records should be reviewed, in-
cluding inventories, shipment manifests, and permits. Per-
sonal interviews with site personnel, public officials, and
private citizens may be helpful.
The most valuable information, of course, is the results
of any air monitoring or sampling that has been perform-
ed. In addition, data on the waste composition may pro-
vide information on potential airborne pollutants. It may
be desirable to perform an offsite preliminary inspec-
tion, since many times a simple visual reconnaissance can
go a long way in terms of defining probable hazards.
Table I.
Examples of Predominant Species Found in Air Near
Hazardous Waste Sites.(1)
Benzene
Toluene
o-xy lene
tn + p-Xylene
Aceta Idehyde
Benzaldehyde
laopropyl ether
Phenol
Ethyl acetate
Ethyl ether
Dimethyl ether
Naphthalene
Ethyl benzene
Methyl isobutyl ketone
n-Pentanal
Ch loroform
Methylene chloride
Dibromorae thane
Dichloroethane
Tetrach lor oe thane
Tr ich lor oe thane
Trich loroethylene
Tetrachloroethylene
Carbon tet rach lor ide
Vinyl methyl ether
Vinyl isopropyl ether
Vinyl ch lor ide
Ch lorobenzene
Ch loroto luene
Di ch lorotoluene
Tr ich lorobenzene
Tr ich lorotoluene
Bromoxy lene
PCfl'a
Maximum concentration observed at each dunpiite,
ug/n,3
Love Canal
Niagara Falls,
N.Y.
5703
1472
73
140
172
10
1140
73
270
1140
5
240
too
6700
159
74
44
Kin Hue,
Ediaon, N.J.
1550
2600
245
56
120
10
232
35
10
6100
444
38
266
1250
63
57
22
500
13
394
20
5000
13000
50
34
1
50
Elizabeth,
N.J.
234
325
79
225
218
95
16
Olh«r
10
64
20
41
11J
2(2
1.1
2.0
4.5
I.I
110000
118
100
Reference 1
280
-------
SITE & PERSONNEL SAFETY 281
Once the background data are assembled, a preliminary
hazard assessment is performed, generally requiring an in-
terdisciplinary scientific approach by individuals with
training in chemical, biological and radiological safety
and health and experiences in determining and implement-
ing a personnel protective equipment program. The task is
not entirely quantitative, since many subjective profess-
ional judgments are required.
Each site is unique, posing specific problems and re-
quiring individual attention. For example, the site may be
known to contain a hazardous chemical; however, if the
site is properly managed and the hazardous chemical ade-
quately contained, there may be only a minimal threat
to worker health and safety.
AIR MONITORING EQUIPMENT
In many cases, the information obtained from the back-
ground research either is not sufficient to define the site
hazards or indicates that the hazard potential is great
enough to require specific personnel protective equip-
ment. Both instances necessitate the use of a variety of
equipment to characterize and monitor the ambient air so
that the appropriate level of personnel protection can be
chosen.
Continuous monitoring is also required since workers
may encounter hazards such as explosive atmospheres or
high levels of radiation for which no protective equip-
ment is available. Continuous monitoring would also
allow the selection of the minimum amount of protec-
tive equipment consistent with maintaining worker health
and safety.
The air monitoring equipment described in this paper
can be conveniently divided into two categories:
•Direct reading instruments, which provide a "real
time" readout of the concentration of pollutants, and
•Collection media, which collect and concentrate the
pollutants for subsequent laboratory analysis.
A list of equipment which is sufficient to evaluate en-
try team safety and to determine and implement a per-
sonnel protection equipment program for hazardous sub-
stance site workers is given in Table II.
With the exception of explosive and/or oxygen-defic-
ient atmospheres, there is no way to reliably predict the
health effects of exposure to hazardous substances based
on the readout of a single direct reading instrument, un-
less the instrument can quantitatively respond to all of the
substances present at any given instant. This is especially
true for_entry into sites where little or no background
information exists, since the hazards are undefined and on
sites where "dynamic" activities such as drum opening,
sampling, drilling, etc., are in progress.
Hazardous substance sites may contain a number ot
pollutants in levels below the detection limits of direct
reading instruments but still can cause synergistic health
effects. Moreover, many toxic substances such as particu-
late organics and metals simply cannot be determined by
a direct reading instrument.
The effects of personnel exposure to these types of sub-
stances must be carefully determined, especially for long-
term site activities involving repeated exposure. For ex-
Table II.
List of Equipment for Ambient Air Characterization.
Hazard
Explosive
atmosphere
Oxygen-
deficient
atmosphere
Toxic
atmosphere
Radioactivity
Direct Reading
Combustible gas indicator
Oxygen level meter
1. Portable photoionization
detector (PID)
2. Portable flame ioniza-
tion detector (FID) w/gas
chromatograph (GC) option
3. Colorimetric tubes
1. Radiation survey meters
(alpha, beta, gamma)
2. Passive monitors (alarms)
Collection Syste
Not used
Not used
Sampling pumps in con-
junction with absorption
tubes, filters, and im-
pingers
Dosimeters (film badges)
See References 2 and 3 for criteria for use of combustible gas indicators and colorimetric tubes.
ample, gamma radiation is relatively easy to detect in the
field but low-level alpha or beta radiation is not. The
health effects of human exposure to low levels of alpha
and beta are potentially more severe than exposure to a
much higher level of gamma radiation. Thus, the deter-
mination of the effects of exposure must be made by a
health physicist and cannot be determined solely by a
single instrument reading.
As another example, direct reading instruments such as
the photoionization detector (PID) and .the flame ioniza-
tion detector (FID) are very useful in locating "hot spots"
—those areas of a site that contain significant levels of
volatile chemicals. However, in order to determine the
health effects of~direct exposure, and ultimately personnel
protection levels, the specific chemicals must be identi-
fied.
For a positive chemical identification, air samples are
collected for laboratory analysis; the air contaminants
are concentrated on a collection medium by passing am-
bient air through it. Since this type of air sampling is ex-
pensive, the value of an extensive background research is
apparent. For sites with insufficient background data, it
may be cost effective to perform work activities using full
protective gear, while observing personnel for signs of
stress. Once again, the subjective professional judgment of
an individual experienced in the implementation of a per-
sonnel protective program is essential.
Explosive and Oxygen-Deficient Atmospheres
Explosive and oxygen-deficient atmospheres are unique
in that the hazards associated with them can be deter-
mined by using direct reading instruments, provided that
the operator is experienced and trained in field use of
the instruments. For explosive atmospheres a combustible
gas indicator is used. Although there are many commer-
cially available units, most test for the concentration of
explosive gases and vapors by measuring the heat produc-
ed by the combustion of a test sample. The readout is
usually expressed as a percent of the lower explosive
limit (% LEL), a dimensionless quantity defined as the
lowest concentration of a gas or vapor by volume in air
-------
282 SITE & PERSONNEL SAFETY
which will explode or combust when there is an ignition
source present.
The National Institute for Occupational Safety and
Health (NIOSH) considers the action level for limitation
of work in explosive areas to be 10% LEL, unless pre-
cautions are taken to prevent ignition sources. A reading
of 20% LEL is considered IDLH (immediately dangerous
to life and health) and requires evacuation regardless of
what precautions have been taken. Also, work activities
cannot be performed in atmospheres where the concen-
trations of gases or vapors exceed the upper explosive
limit (UEL), which is the concentration of a gas or vapor
by volume in air above which an explosion will not occur
if there is an ignition source. Evacuation is necessary if
UEL concentrations are detected, since dilution by air
can easily create an explosive situation. Gear is not avail-
able for protection against the explosion hazard; immed-
iate evacuation is the only recourse.
Oxygen-deficient atmospheres can be determined by us-
ing a commercially available oxygen level meter. A typical
unit works on the principle of selective penetration of a
membrane by oxygen with a subsequent reaction with an
electrolytic reagent. The readout is generally directly in
percent oxygen (% 02).
NIOSH requires that an external air supply, usually a
self-contained breathing device (SCBA), be used for work
in atmospheres with less than 19.5% oxygen. Severely
oxygen-deficient atmospheres indicate the presence of sig-
nificant quantities of pollutants. Note that combustible
gas indicators may not function properly in severely oxy-
gen-deficient areas and should be used with an oxygen
meter.
Radioactivity
There are three types of radiation that can be encoun-
tered on hazardous substance sites: alpha, beta and
gamma. Human exposure to each type of radiation
causes unique health effects, which are related to the re-
ceived dose.
Sources of gamma on hazardous substance sites include
hospital wastes, which may contain materials such as ces-
ium, cobalt, and radium. Gamma radiation is easily de-
tected using a variety of instrumentation, since it is in the
form of electromagnetic energy and therefore travels rela-
tively long distances. Because gamma is penetrating ex-
ternal radiation, there is no convenient means of pro-
tecting workers from exposure; if gamma radiation is de-
tected, personnel should be evacuated immediately and
appropriate government agencies contacted.
Alpha and beta radiation, however, travel much shorter
distances than gamma and are therefore much more diffi-
cult to detect, since the monitoring device must be placed
close to the source. Alpha sources are common and in-
clude substances such as thorium, radium and uranium.
Biological laboratory wastes may contain beta sources,
including radioactive carbon, phosphorus and sulfur.
Since alpha and beta radiation are particulates and tend to
adhere to clothes and boots of site workers, monitoring
devices that will detect their presence should be used to
screen personnel as personnel leave the site.
Although the external hazards due to alpha and beta
radiation differ, protective clothing is available. The great-
est danger is in the inhalation or ingestion of alpha or beta
particles. An air-purifying respirator fitted with a radio-
nuclide canister will in most cases offer protection against
radioactive particulates. Adequate decontamination facil-
ities must be available, however.
For field work, personnel should be equipped with a
portable alpha/beta/gamma survey meter. Lapel-sized
passive radiation monitors which respond with an audible
alarm when the wearer encounters gamma are recom-
mended for at least one team member. In addition, a dos-
imeter, which provides an indication of the total whole
body exposure to radiation over an extended period of
time, should be worn by all onsite personnel. Dosimeters
are typically distributed to workers upon site entry and
are worn on the lapel, underneath any protective clothing.
Toxic Atmospheres
When workers are engaged in activities on a hazardous
substance site, Federal safety and health standards141 dic-
tate that they be protected against any respiratory danger.
Personnel must wear an SCBA until the ambient air is fully
characterized. It is desirable to use the lowest level of res-
piratory protection possible (i.e., the air-purifying respir-
ator or no respiratory protection at all), since the addition
of protective gear generally increases worker stress. Thus,
the air monitoring program must provide the data neces-
sary to determine respiratory protection.
The problem of cutaneous and percutaneous chemical
hazards must also be addressed. Skin protection must be
worn if there is any possibility of systemic injury, skin irri-
tation, or death resulting from contact with an airborne
gas, liquid, or paniculate. Typically, the following infor-
mation regarding all airborne pollutants is obtained in
order to specify protective gear:'5'
•The chemical identity, the airborne concentration, and
the physical state of each pollutant
•Permissible exposure limits of each pollutant
•The vapor pressure and equivalent ppm of each pollutant
•Warning properties for a gas or vapor
•Eye irritation potential for each pollutant
•The LEL, UEL, and IDLH concentration for each
pollutant
One of the methods of determining the need for respir-
atory protection is to compute the "equivalent ex-
posure"'61 based on the first three parameters in the table
above. The calculations involved in computing the equiv-
alent exposure are complex and are best left to the indus-
trial hygienist who has experience in determining accep-
table risk levels. Many factors in addition to those above
enter into the selection of specific personnel protection and
are beyond the scope of this discussion.*
Direct reading instruments such as the PID, the FID and
colorimetric tubes are used to locate "hot spots" (areas
of high air contamination) and in some instances provide
•See References 4, 5, and 7 through 14 for information on selecting
protective equipment and determining the toxicological effects of ex-
posure to hazardous substances.
-------
SITE & PERSONNEL SAFETY 283
a tentative identification of volatile inorganic and organic
chemicals. Hot spots are important since they define the
areas of the site that may contain significant quantities of
chemicals.
Air sampling pumps for volatile organics are usually
placed on the hot spots. Direct reading instruments can
also be used to monitor the ambient air for dramatic
changes in quality that may occur as a result of activ-
ities such as drilling, drum opening, etc. because these in-
struments provide a real time readout of pollutant levels.
The PID is a non-specific vapor/gas detector. It uses
the principle of photoionization to detect a wide variety
of volatile chemical compounds. The readout is in parts
per million (ppm), references to a specific calibration
standard. It does not respond to the common constit-
uents of air, including methane.
The flame ionization detector (FID) in conjunction with
a gas chromatographic (GC) column is used to detect and
to tentatively identify volatile organic chemical com-
pounds. In the "survey mode" it functions as a non-spe-
cific total hydrocarbon detector, with a readout in ppm,
referenced to a specific standard, usually methane.
In the "GC mode" the unit can determine the "reten-
tion time" of an unknown chemical compound. The re-
tention time is the length of time necessary for the chem-
ical to pass through the GC column. By comparing the
retention time of the unknown with the retention time of
known standards, a tentative qualitative/quantitative iden-
tification of the unknown can be made.
Since specific chemical standards and calibrated col-
umns are needed, an idea of the nature of the unknown
is needed prior to the GC analysis. Thus, the impor-
tance of background research is demonstrated, since a
prior knowledge of the air characteristics allows the selec-
tion of analytical standards for the GC.
Colorimetric tubes provide a means of qualitative/semi-
quantitative identification of volatile organic and inorgan-
ic chemicals. The sample is drawn into the tube by means
of a hand-operated pump. Since the reagents in each tube
react with a specific chemical compound, prior knowledge
of the chemicals expected is necessary for selection of the
tubes. At a minimum, tubes for hydrocyanic acid, phos-
gene and hydrogen sulfide should be used on sites inves-
tigated.
For positive identification of all the pollutants that may
be present, various collection media are used in conjunc-
tion with sampling pumps. Sampling pumps are commer-
cially available in a number of configurations, but for
general site characterization, an intrinsically safe
"NIOSH" pump is recommended. This type of pump is
portable, battery operated, and sufficiently lightweight so
that it can be worn by personnel. Typical collection med-
ia include activated carbon, Tenax GC, XAD-2, silica
gel and Florisil tubes, membrane and glass fiber filters,
polyurethane foam plugs and impingers containing se-
lected reagent solutions. Specific applications for these
media, including the typical laboratory analysis required
are given in Table III.*
*For a complete listing of methods and collection media for air sampling,
see References 15,16 and 17.
Table III.
Specific Applications for Collection Media Including
the Required Laboratory Analysis.
Pollutant
Volatile organics
Particulate
organics
Pesticides (in-
cluding PCBs)
PBBs
Metals
Volatile inorganics
Particulate
inorganics
Cyanides
Collection Media
Carbon tubes
Tenax tubes
XAD-2 tubes
Silica gel tubes
Glass fiber filters
Florisil tubes
Polyurethane plugs
Glass fiber filters
Glass fiber filters
Membrane filters
Impingers/reagent solutions
Membrane filters
Glass fiber filters
Filters/impingers
Laboratory Analysis
Gas chromatograph/mass
spectroscopy (GC/MS)
GC/MS
GC/MS
GC/Electron capture
GC/MS
Atomic absorption (AA)
Wet chemical methods
Wet chemical methods
Wet chemical methods
As previously mentioned, pump placement for volatile
materials can be determined by using the PID, the FID
and colorimetric tubes to survey the site for hot spots. In
general, pumps are placed in selected work areas, as well
as near drums, leachate ponds or streams, surface water
impoundments, lagoons, areas with visibly contaminated
soils and in the headspace of groundwater monitoring
wells. Upwind and downwind sampling points are recom-
mended.
Once personnel protection has been determined, air
monitoring is still necessary. For example, assume that an
air-purifying respirator has been determined to be ade-
quate for a worker involved in drilling operations. Since
drilling may result in a spontaneous release of toxic sub-
stances, a means to determine if the air quality has changed
is necessary to insure that the respirator will protect the
worker. In this situation, a FID or a PID can be used to
give a real time readout of pollutant levels. A combus-
tible gas indicator/oxygen meter combination should also
be used to determine if an explosive situation has been
created. If the drilling operations are long term, it would
be desirable for the worker to wear a sampling pump fitted
with a carbon or Tenax tube so that his/her total ex-
posure can be monitored. Real-time monitoring should
also be used for drum opening and staging, sampling, ex-
cavating, draining of tanks and lagoons and related re-
medial action activities.
Environmental Monitoring
Environmental monitoring also can be performed using
the equipment discussed in this paper. Of particular con-
cern is the impact that site work activities have on the
surrounding community, since new hazards may be
created. In general, the highest concentrations of toxic
pollutants will occur directly over the site, close to the
point source. A reasonable assumption is that the concen-
trations will decrease as the pollutants migrate offsite.
Thus, if a complete air monitoring program for personnel
protection has been initiated, the values measured onsite
can be considered worst case.
-------
284 SITE & PERSONNEL SAFETY
Several models have been developed to predict the air
emissions from sites.(l8- "• M) However, offsite monitoring
may be required, since in many cases workers spend a good
deal of time at the site boundaries. A common mistake is
to remove protective gear, as if the site boundary somehow
limits pollutant migration. This type of site boundary
monitoring is an integral part of evaluating overall site
safety.
CONCLUSION
Entry into a hazardous substance site involves poten-
tial exposure to a number of hazards. It is essential that the
risks associated with these hazards be determined such that
protective measures can be taken to maintain worker
health and safety. A hazard assessment is necessary and
should precede any site work. Background research for site
specific information as well as on-site air characterization
will supply the quantitative data needed to perform the
assessment. A personnel safety plan should be designed
based on these data.
In addition to the quantitative data, many qualitative
judgments enter into the determination of an effective
and efficient safety plan. Perhaps the most important
factor is determining if the physiological and psycholo-
gical stresses caused by wearing protective gear are great-
er than the risks of direct exposure to hazardous sub-
stances. Therefore, individuals with training and exper-
ience in hazard evaluation and the implementation of a
personnel protection plan are as necessary as the monitor-
ing equipment used to characterize the ambient air.
REFERENCES
1. Esposito, M. and Bromberg, S., "Fugitive Organic
Emissions from Chemical Waste Dumpsites," paper
presented at the 74th annual meeting of Air Pollu-
tion Control Association, Philadelphia, Pennsylvania,
June 21-26, 1981.
2. "Combustible Gas Indicators, A Manual of Recom-
mended Practice," American Industrial Hygiene
Association, Akron, Ohio, 1981.
3. "Direct Reading Colorimetric Tubes, A Manual of
Recommended Practice," American Industrial Hy-
giene Association, Akron, Ohio, 1976.
4. OSHA Safety and Health Standards (29 CFR 1910),
U.S. Department of Labor, Occupational Safety and
Health Administration, OSHA 2206, revised January,
1976.
5. Pritchard, J., "A Guide to Industrial Respiratory
Protection," Appendix F, "Joint NIOSH/OSHA
Standards Completion Program Respiratory Decision
Logic," Publication LA-6671-M, Los Alamos Scien-
tific Laboratory, Los Alamos, New Mexico.
6. American Conference of Governmental Industrial Hy-
gienists, "TLVs Threshold Limit Values for Chem-
ical Substances and Physical Agents in the Work-
room Environment," Appendices B, C, and D, and
"Documentation of the Threshold Limit Values for
Substances in the Workroom Air," ACGIH, Cin-
cinnati, Ohio, 1978.
7. Mackison, F., Stricoff, R., and Partridge, L., ed-
itors, "NIOSH/OSHA Pocket Guide to Chemical
Hazards," U.S. Dept. of Labor, Occupational Safe-
ty and Health Administration, August, 1980.
8. "Methods for Prevention and Control of Occupa-
tional Skin Disease," Bureau of National Affairs,
Washington, D.C.
9. Sansone, E. and Tewari, Y., "The Permeability of
Laboratory Gloves to Selected Solvents," American
Industrial Hygiene Journal, 39, Feb. 1978.
10. Sansone, E. and Tewari, Y., "Penetration of Pro-
tective Clothing Materials," American Industrial Hy-
giene Journal, 39, Nov. 1978.
11. Williams, J., "Permeation of Glove Materials By
Physiologically Harmful Chemicals," American In-
dustrial Hygiene Journal, 40, Oct. 1979.
12. Coletta, G., Swope, A., et al., "Development of Per-
formance Criteria for Protective Clothing Used
Against Carcinogenic Liquids," Occupational Safety
and Health Administration, Cincinnati, Ohio,
October, 1978.
13. Clayton, G. and Clayton, F., editors, "Patty's In-
dustrial Hygiene and Toxicology," Volumes 1 and 2,
John Wiley and Sons, New York, N.Y., 1978.
14. Sax, N., ed., "Dangerous Properties of Industrial
Materials," Fifth Edition, Van Nostrand Reinhold,
New York, N.Y., 1978.
15. "Industrial Hygiene Field Operations Manual,"
U.S. Department of Labor, OSHA, Jan. 1980.
16. "NIOSH Manual of Analytical Methods," U.S. De-
partment of Labor, OSHA, 1974.
17. "Analysis of Organic Air Pollutants by GC/MS,"
U.S.E.P.A., June 1977.
18. Shen T. and Tofflemire, T., "Air Pollution Aspects
of Land Disposal of Toxic Waste", Proc. of the 1979
National Conference on Hazardous Material Risk
Assessment, Disposal, and Management, Information
Transfer, Inc., Silver Spring, Maryland, 1979.
19. Shen, T., "Emission Estimation of Hazardous Or-
ganic Compounds from Waste Disposal Sites," paper
presented at the 73rd annual meeting of Air Pollu-
tion Control Association, Montreal, Quebec, June
22-27, 1980.
20. Hwang, S., "Land Disposal Toxic Air Emissions
Evaluation Guidelines," Office of Solid Waste,
U.S.E.P.A., December, 1980.
-------
SHAKEDOWN AND PERFORMANCE TESTING OF THE
EPA-ORD MOBILE INCINERATION SYSTEM
JOHN E. BRUGGER, Ph.D.
JAMES J. YEZZI, JR.
FRANK J. FREESTONE
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Edison, New Jersey
INTRODUCTION
The EPA's mobile incineration system was designed
and constructed under the Office of Research and De-
velopment's program to demonstrate the feasibility and
practicality of destroying/detoxifying hazardous sub-
stances and wastes that have been improperly discarded at
landfills, spilled onto soil or stored in above-ground con-
tainers. The basic purpose of the incinerator is to allow
cleanup to take place totally at the affected site, thus
avoiding both transport and the ever present institutional
barrier that an area will become a permanent destruction
site.
The system is purposely over-designed to accommodate
diverse situations and can be further modified to handle
unique situations. It is expected that private interests will
make use of the plans, specifications and reports to con-
struct and use versions that are more dedicated to par-
ticular, defined problems where the research potential is
not required.
Those skilled in the art will readily determine what
changes should be made and will, undoubtedly, be able to
reduce both construction and operating costs. As an ex-
ample, operating cost savings can be achieved by better
utilization of waste heat.
SYSTEM DESCRIPTION
The mobile incineration system is mounted on three
45-ft air-ride suspension trailers for permit-free hauling
over Interstate Highways and improved roadways, as well
as for short-distance travel over unimproved roads or ter-
rain to a suitable setup site. The system is self-con-
tained, except for consumable supplies (water, fuel,
chemicals and power, which preferentially will be sup-
plied by commercial lines, but can be furnished by auxili-
ary diesel generators).
To conform to applicable regulations and operating
permit requirements, gaseous emissions are continuously
monitored or intermittently sampled to the extent required
and all solid and liquid wastes are collected, analyzed
and treated, as necessary, before disposal. The system
was specifically designed to meet the stringent require-
ments (40-CFR-761) for the incineration of polychlori-
nated biphenyls (PCBs), namely, minimum residence time
of 2 sec at 2200 °F at 3% excess oxygen, and attainment
Figure 1.
Mobile Incineration System. Trailer #1 (lower right) showing
control booth, solid feed mechanism and kiln with ducting to
Trailer #2 (center) showing secondary combustion chamber and
flue gas quench with ducting to Trailer #3 showing paniculate
filter, mass transfer scrubber, induced draft fan with diesel
drive and base of stack, and diesel-electric unit.
of a combustion efficiency of 99.9% along with such re-
lated requirements as limitations on the release of par-
ticulates and acid gases, destruction and removal effici-
ency and rapid shutdown in the case of malfunction or
failure to meet the criteria of 40-CFR-761.
The principal unit on Trailer No. 1 is a 52-in. ID by
16 ft long A.P. Green "Kast-O-Lite 30 in.* refractory-
lined kiln having specially designed double layer, leaf
type, stainless steel seals packed with graphite-impreg-
nated packing and rotated by a hydraulically driven motor
with speed and pitch adjusted to regulate the residence
time of solids. The effective volume is 236 ft3. Firing is
with two tilted 4 in. Maxon burners with proportioned air
control. Controlled tuyere (excess) air is also provided to
accommodate atomizer injected liquids, pumpable sludges
and ram fed solids.
The hopper-fed ram has an adjustable stroke and fre-
quency. Liquid and solid wastes will be homogenized sep-
arately to ensure homogeneous composition and heat
value, since variations in feed composition can lead to
highly objectionable combustion transients. Liquid and
solid wastes can be fed into the kiln concurrently. Temper-
ing water is injected into the kiln to control temperatures
and maximize throughput.
285
-------
286 TREATMENT & DISPOSAL
The first trailer also has a motor-powered hydraulic
system, forced air blower and the required piping and in-
strumentation. Thermocouples are mounted to record tem-
perature (the maximum is 1800°F) and an oxygen con-
centration meter monitors the exit gas.
The second trailer is designed to support the secondary
combustion chamber (SCC) which has the same bore and
refractory as the kiln, but is 36 ft long. The SCC is sur-
rounded with a shroud through which cooling air is passed.
The second trailer also has two Maxon burners and is de-
signed to meet the 40-CFR-761 requirements for PCB
incineration.
Gases pass from the kiln, whose primary purpose is to
initiate combustion and to insure that hazardous sub-
stances are fully volatilized from any solids, through "in-
conel" ducting—designed to handle misalignment and
temperature-induced dimensional changes. A swirl plate at
the entrance to the SCC aids development of turbulence
(the design Reynolds number is greater than 30,000). The
temperature and oxygen concentration are continuously
monitored at the end of the SCC. The residence time is
computed from feed and temperature data. The SCC trail-
er has its own air supplies and the necessary instrumenta-
tion and control equipment.
The SCC terminates in a gas flow measuring venturi
alloy nozzle that conducts the flue gas to a quench elbow
equipped with water spray nozzles that drop the gas
temperature to approximately 180°F and remove
particulates and acid gases.
The gas/vapor is ducted to a ground-level, baffled
sump and then ducted upwards to the first element on
the third trailer, namely, a "CHEAP" paniculate re-
mover, also fabricated from an "Inconel" alloy. The
CHEAF collects particulates on a sprayed fiberglass bdt
that is pulled through and re-rolled automatically when
the pressure drop (approximately 40 in. W.C.) exceeds a
set-point as the belt becomes plugged with particulates
(down to the 0.3 /* range).
The gas/vapor then enters a cross-flow caustic mass
transfer unit ("Ceilcote") (MX) packed with "Telerettes"
and scrubbed with sodium carbonate or caustic soda as
required. Fluid levels in the quench tank, CHEAF and
MX sumps are controlled, pH is measured and makeup
water or alkaline fluid is added as required. A demister
section is incorporated into the MX. The ISO hp diesd-
driven induced draft fan can develop the equivalent of
70-90 in. W.C. for use at mile-high altitude sites and
consists of an "Inconel" shaft and 36 in. impeller with
aSS316Lhousing.
A controlled reflow duct connects the exit of the fan
to the inlet of the CHEAF to maintain the correct
pressure drop across the CHEAF. The flue gas from the
fan passes through a sound suppressor and the stack,
which has ports for the flue gas sampling that is re-
quired by regulations. The nominal gas flow is 7200
Figure 2.
Field assembly of Mobile Incineration System showing (left to
right) Trailer #1 with feed and kiln, Trailer #2 with secondary
combustion chamber and venturi quench and sump, and
Trailer 03 with paniculate filter, mass transfer unit (scrubber),
open control panel, draft fan and drive, and electric power
supply. (Back-hoe at far left used in assembly.)
-------
TREATMENT & DISPOSAL 287
scfm. The instrumentation and interlock system on
Trailer No. 3 is the most complex.
The system is designed to accept as feed those sub-
stances cited pursuant to Sections 311 and 307 of the
Clean Water Act, along with materials designated under
the Resource Conservation and Recovery Act. The feed-
ing of inorganic salts, which tend to flux the kiln re-
fractory, and of heavy metals, which form volatile oxides
or halides, must be controlled; data gathering on this as-
pect is an integral part of the continuing shakedown/
testing.
SYSTEM TESTING
At the time this paper was written, shakedown was es-
sentially complete, i.e., the pumps, valves, switches,
interlocks, etc. have been tested singly and as a system,
but without firing the system.
The current test plan has five major phases, each of
which involves extensive sampling and analysis:
(1) commercial fuel oil burn,
(2) burn of fuel oil spiked with iron oxide particles,
(3) incineration of fuel oil spiked at increasing levels of
dichlorobenzene,
(4) combustion of fuel oil spiked with increasing con-
centrations of PCBs (or Aroclors) and
(5) incineration of an oil-based waste leaching from a
local landfill.
Federal and State of New Jersey permits are both re-
quired, as is assent from the municipality in which the
test is conducted. Local residents have visited the facility
and will participate in a public hearing.
The main purpose of the fuel oil burn is to establish
baseline data. Extensive gas sampling, for example, with
the modified method 5 system and/or SASS train, will be
replicated. The presence of polychlorodibenzodioxins
(PCDD) and polychlorodibenzofurans (PCDF) will be
carried to the limit of detection. All wastes will be
analyzed and continuous or intermittent measurements
(as required) will be made on carbon dioxide, carbon
monoxide, hydrochloric acid, sulfur oxides, nitrogen
oxides, total hydrocarbons, polluting organic hazardous
cfreinicals (POHCs), particulates, etc.
;the test with iron oxide-spiked fuel oil will assess the
efficiency of the paniculate removal capability of the
system. Many factors were taken into account in choos-
ing avery fine iron oxide. Subsequently, testing is planned
With metal soaps, metal alkyls and metal aryls.
la the third phase, orthodichlorobenzene will be added
to fuel oil in concentrations of 1, 5 and 25% to assess
DREs, chiefly.
In the fourth phase, based on the outcome of the pre-
ceding chlorinated aromatic chemical tests, actual samples
of PCBs or PCB-containing fluids will be fed at concen-
trations of 0.05, 1, 10 and 20%. A planned fifth phase
will be based on the effective incinerability of an oily
waste leachate contaminated with priority pollutants.
In all these tests, the system will be first brought to
temperature, which requires 16-24 hours, and then op-
erated as long as is required to collect meaningful samples.
The system will then be idled (not fully cooled down) un-
til the next phase is to be intiated. Excluding tests on a
proposed leachate, the test series will require 45 days.
The data are expected to yield a definitive array of com-
bustion and operating conditions that can be adapted
for use in detoxifying other hazardous substances.
Since the mobile incineration system is a research unit,
provision has been made for the incorporation of modi-
fications, especially in air pollution control, should these
be indicated.
ACKNOWLEDGEMENTS
The basic system was fabricated by MB Associates (now
Tracor-MBA), San Ramon, California, under EPA con-
tract 68-03-2515. Especial mention should be made of
the contribution of R. Tenzer, Wm. Mattox, P. Kirke-
gaard, Wm. Ford, D. McKown, and their associates.
The major part of the system was delivered in Sep-
tember 1980 to an on-site operating unit in Edison, N.J.,
namely, the Environmental Emergency Response Unit
(EERU) (a contractor-operated facility), for completing
assembly, undertaking shakedown, conducting tests and
performing field demonstrations. EERU was operated by
Mason & Hanger-Silas Mason Company, Inc., under
Contract No. 68-03-2647 until June 1981. Thanks are due
to M. Sproul, G. Campbell, V. Manolio, F. Brown and
the Mason & Hanger home office staff (Lexington, KY)
that prepared most of the P&I diagrams and initiated in-
stallation of the instrumentation and control equipment
and the various lines (electric, hydraulic, fuel, compressed
air).
EERU is now being operated by IT Enviroscience
under Contract No. 68-03-3069. ITE has current re-
sponsibility for completion of assembly, shakedown,
field testing and operation. The contributions of C.
Steuwe, C. Pfrommer, S. Anicito, A. Sherman and R.
Miller, on site, and of R. Novak, R. Lovell and many
others at ITE headquarters (Knoxville, TN) are acknowl-
edged.
REFERENCES
1. Tenzer, R., Ford, B., Mattox, W. and Brugger, J.E.,
"Characteristics of the Mobile Field Use System for
the Detoxification/Incineration of Residuals from Oil
and Hazardous Material Spill Cleanup Operations,"
J. Haz. Materials, 3, 1979, 61-75.
2. Tenzer, R., Ford, B., Mattox, W. and Brugger, J.E.,
"Mobile System for the Detoxification/Incineration of
Cleanup Residuals from Hazardous Material Spills,"
in Disposal of Oil and Debris Resulting from a Spill
Cleanup Operation, Am. Soc. for Testing and Mater-
ials STP 703, 1980, 118-136.
3. Tenzer, R., Mattox, W., Brugger, J.E. and Free-
stone, F.J., "Design and Testing of a Mobile Incin-
eration System for Spilled or Waste Hazardous and
Toxic Materials," Proc. of the 1980 National Confer-
ence on Control of Hazardous Material Spills, May
13-15, Louisville, Kentucky, 1980,467-475.
-------
HISTORY AND BENCH SCALE STUDIES FOR THE
TREATMENT OF CONTAMINATED GROUNDWATER AT THE
OTT/STORY CHEMICAL SITE, MUSKEGON, MICHIGAN
STEPHEN C. JAMES
U.S. Environmental Protection Agency
Solid and Hazardous Waste Research Division
Cincinnati, Ohio
ALAN J. SHUCKROW, Ph.D.
ANDREW P. PAJAK
Touhill, Shuckrow and Associates, Inc.
Pittsburgh, Pennsylvania
INTRODUCTION
Hazardous leachates and contaminated ground and sur-
face waters are often associated with unsecured indus-
trial waste storage and disposal sites. Numerous prob-
lems are encountered in the cleanup of such a site. One
major problem is identifying the most effective treatment
technology for the contaminated stream. Contributing to
this problem are:
(1) The inability to characterize completely the con-
taminated stream due to technical and economical
limitations
(2) The paucity of information on the effectiveness of
techniques for treating the broad spectrum of or-
ganic and inorganic compounds frequently present
in these streams.
Therefore, the U.S. Environmental Protection Agency be-
gan a project now being conducted by Touhill, Shuck-
row and Associates, Inc. (TSA) to evaluate techniques
for concentrating hazardous constituents of aqueous waste
streams. A literature review, desktop evaluation and lab-
oratory bench scale experimental studies form the basis
for judging the potential of numerous candidate tech-
nologies.
During the course of this project, a part of the lab-
oratory experimental work was carried out at Cordova
Chemical Company, present owners of the Ott/Story Site,
using ground water contaminated by prior operations. This
site provided an opportunity to interface research activ-
ities with development of a treatment system for a high
priority problem site.
In this paper, the authors provide a brief background
on the site history, a method selection of candidate treat-
ment technologies and summary of results of the experi-
mental studies.
SITE HISTORY
During 1957, Ott Chemical Company began produc-
tion of various organic chemicals in Dalton Township,
Michigan, north of the City of Muskegon.'" Many of the
organic chemicals produced used phosgene as a raw ma-
terial. At the time of plant startup, Michigan State Water
Resources Commission permitted Ott Chemical Company
to discharge wastewaters into seepage beds, provided that
certain limitations on the types and amounts of substances
disposed were adhered to and that the discharge was mon-
itored regularly and results reported to the state.
Permit Violations
Beginning in 1957 and continuing through 1967, permit
violations were evident. Initially, chlorides and phenols
were detected in monitoring and water supply wells. Ott
Chemical Company attempted to mitigate pollution prob-
lems, by using a groundwater purge system to contain the
contaminated plume on company property, and by re-
ducing contaminant concentrations in their discharges to
the seepage beds. The State permitted discharge of purged
groundwater to a nearby creek, provided there was no im-
pairment of stream's beneficial use.
Although the remedial action program reduced contam-
inant concentrations going into groundwater and limited
pollution migration, other problems became evident. Fish
tainting because of purge water discharge was observed in
the nearby creek, and other groundwaters in the area were
polluted by stored or disposed wastes.
In recognition of the continuing and spreading problem
and the fact that remedial actions were not successful
completely, the State began, in 1965, water quality inves-
tigations to document the extent of contamination. At
least 40 organic chemicals, some of which are considered
to be hazardous, were identified in the groundwater.
Remedial actions continued but in 1968 the State
ordered discontinuance of direct discharge of process
wastewaters to the groundwater. Although cooling water
discharges were still permitted to be discharged to the
seepage beds, treatment of process waters prior to dis-
charge to the Muskegon River was required. Nevertheless,
wastes stored on site and spilled in the plant area continued
to exacerbate and spread the problem. In 1974, process
wastes and purged groundwater were connected to the
municipal wastewater treatment system.
288
-------
TREATMENT & DISPOSAL 289
Up to 1976, attempts to improve waste handling prac-
tices at the site continued. Some of these efforts included:
waste incineration, development of spill prevention
measures and plans for deep well disposal. However, con-
trol of the contaminated groundwater plume was not
accomplished and the pollutants continued to migrate to-
ward residential water supply wells down-gradient from
the plant site. A small nearby trout system became severe-
ly degraded by contaminated groundwater exfiltration.
Ownership Change
Solutions to the contamination problems were hin-
dered, to some extent, by operating entity and ownership
changes between 1965 and 1973. In the 1972-1973 period,
Story Chemical Company became owner of the facility.
Ultimately, during 1976, Story Chemical Company filed
for reorganization in Federal Bankruptcy Court, abandon-
ing the plant site in September 1977. Site abandonment
caused additional concern because of 23,000 pounds of
phosgene gas left in two containers on site.
In October 1977, Cordova Chemical Company pur-
chased the facility, and cooperated with the State in site
cleanup, including providing funds to furnish an al-
ternate water supply for the area. In return, the state
agreed to limit the new owner's liability for previous prob-
lems.
In 1979, groundwater contamination was found at a
depth of 30 m, and the contaminated plume was 400 m
wide. Furthermore, wells in the area were deemed unpot-
able and residents were advised to use bottled water for
drinking.
Ott/Story operations left a number of unresolved con-
cerns. In particular, large volumes of organic chemical mi-
grating through soil and ground and surface water contin-
ued to have an adverse environmental impact. It was with-
in this context that bench scale studies were conducted to
uncover suitable treatment methods for decontaminating
polluted groundwater.
Table I.
Treatment Process Trains
Process Train Sequence Ancillary Processes
Biological followed by
carbon sorption
pH adjustment, coagulation,
settling before biological for
heavy metal control. Clarifi-
cation, filtration before carbon
for solids removal
Waste Type
High TOC, low in toxics (to
biomass)
Carbon sorption followed pH adjustment, coagulation,
by biological settling before carbon to
reduce solids loading
Biological with
powdered activated
carbon included
Membrane followed by
biological
Stripping followed by
rartron sorption
Biophysical system pro-
ceeded by coagulation and
clarification to minimize
heavy metals and solids
loading
pH adjustment, clarifica-
tion and filtration before mem-
brane to prevent membrane
fouling
pH control before stripping.
Clarification and filtration
before carbon for solids
removal
High TOC, high in toxics (to
biomass
High TOC, high in toxics
Membrane removes high
molecular weight organics and
inorganic ions; biological re-
moves remaining organics
Stripping to remove volatile
organics and reduce organic
load to carbon; carbon to
provide further organic removal
LITERATURE REVIEW/DESKTOP EVALUATION
Results of the literature and desktop evaluation have
been reported elsewhere.<2'3> The following list of candi-
date concentration technologies was developed:
BIOLOGICAL TREATMENT
CARBON ADSORPTION
CATALYSIS
CENTRIFUGATION
CHEMICAL PRECIPITATION
CRYSTALLIZATION
DENSITY SEPARATION
DIALYSIS/ELECTRODIALYSIS
DISTILLATION
EVAPORATION
FILTRATION
FLOCCULATION
ION EXCHANGE
RESIN ADSORPTION
REVERSE OSMOSIS
SOLVENT EXTRACTION
STRIPPING (AIR AND STEAM)
ULTRAFILTRATION
The processes that had the greatest range of applicability
included:
BIOLOGICAL TREATMENT
CHEMICAL COAGULATION
CARBON ADSORPTION
RESIN ADSORPTION
STRIPPING
MEMBRANE PROCESSES
It also was concluded that these unit processes would
have to be assembled in process trains (containing more
than one of these unit operations) to achieve the degree of
treatment required at most sites. Five possible process
trains were formulated (Table I). These trains do not rep-
resent the only possible configurations; however, they do
provide broad range applicability to a wide variety of con-
taminants.
These process combinations were developed with data
from single compound studies. Actual treatability of
multi-compound streams generally will require bench-scale
studies using actual wastewater. Because of the specific
contaminants at the Ott/Story site, chemical coagulation
and membrane processes were judged to have limited ap-
plication and were not studied in the laboratory.
BENCH SCALE STUDIES OF UNIT PROCESSES
In laboratory bench scale studies, individual unit pro-
cesses were examined first; then process trains were eval-
uated. Wastewater used in these studies was contam-
inated groundwater at the Ott/Story Site collected from
two wells located in the contaminant plume. For a portion
of the studies, an equal volume composite of water from
these two wells was used; in other studies the water from
each well was used separately. Analyses of the composite
sample and water from the separate wells are shown in
Table II. Although the water from well W-17d has a much
lower total organic carbon (TOC) concentration, it con-
-------
290 TREATMENT & DISPOSAL
Table II.
Groundwater Quality at the OTT/Story Site
Conventional Pollutants:
pll
COD (mri/1)
TOC (m.'|/l)
NH3-N (mg/1)
Ma^or Priority Pollutants:
(in ug/1)
Vinyl chloride
1, 1-dichloroethylene
1,1-dichloroethane
1, 2-dichloroethane
Benzene
1,1,2,2-tetrachloroethane
Toluene
Composite Sample
(50% OW-9 & 50% W-17d)
9.3 - 12
5400
400 - 1500
64
ND - 32,500
5 - 6,570
60 - 19,850
350 - 111,000
ND 7,370
5 1,590
5 5,850
tains many of the same organic priority pollutants as well
OW-9, frequently at similar concentrations.
Results of many of these experimental studies have
been described in detail elsewhere.(2- "• 5| 6) In this paper,
a brief summary is presented.
Steam Stripping
Continuous flow, steam stripping experiments were con-
ducted using a composite sample. The TOC concentration
in the stripper bottoms ranged from 300 to 400 mg/1 and
appeared to be independent of the overhead: feed ratio.
Overhead TOC concentration approached 3500 mg/1 at an
overhead:feed ratio of 0.06. The average TOC reduction
between feed and bottoms was 34%.
These results indicate a major constraint associated with
steam stripping, i.e., it is necessary to further treat a bot-
tom waste stream having a flow only slightly less than the
feed flow. Because the TOC concentration appeared to be
independent of system flow rates and feed TOC levels over
the ranges examined and because additional treatment of
the bottoms would be necessary, steam stripping would
likely be a costly, yet only moderately effective, concen-
tration process.
Adsorption by Granular Activated Carbon and Resin
Batch isotherm studies were used to evaluate various
sorbents and operating conditions. Then continuous flow
granular activated carbon (GAC) studies were conducted
off pH 9.3 to 10.0, using Filtrasorb 300 GAC.
In continuous flow studies three or four glass columns
arranged in series with sampling ports located at the in-
fluent and effluent ends of each column were used. Each
column was 122 cm by 2.5 cm (ID) and contained 91 cm
of GAC. The system was operated in a downflow mode at
a loading rate of approximately 1.35 l/m3s (2 gal/min/ft).
This provided an empty bed contact time (EBCT) of
approximately 15 min. per column.
Influent TOC concentration varied substantially, rang-
ing from 316 to 950 mg/1. Generally, after only 3 to 10 bed
Well OW-9
9.6- 1105
5200 - 8300
1500 2400
40 163
ND
ND - 2600
ND - <0.1
>500 - 2000
600 - 900
ND - <1
3,300 5,600
Me 11 W-17d
9.5 9.6
1600 2600
171 730
7.9 16
ND - 5, 240
ND
ND 6,570
3,670 >1000
>5000 25,000
ND - 1, 590
3,610 - 7000
volumes (BVs), TOC removal decreased to 50%. TOC
leakage reached 90% after about 200 to 240 BVs were pro-
cessed and continued at this level until up to 500 BVs
had been processed. As would be expected, effluents from
columns 2 and 3 were intermediate between columns 1 and
4. Column performance data correspond to batch sorption
study data.
The difference between treatabilities of the composite
groundwater and waters from the individual wells is shown
in Figure 1. Organic materials in W-17d (as reflected by
TOC concentration) were more readily sorbed by GAC
than was TOC in the composite groundwater or OW-9
groundwater.
Continuous flow resin adsorption studies were con-
ducted on the composite sample using XE-347 carbon-
aceous resin. Again, these continuous runs were preceded
by batch isotherm studies.
Three columns similar to those used for GAC studies
were charged with 792 to 835 cm3 of resin and were op-
erated at loading rates of 2.95 to 3.79 BV/hr. EBCT
ranged from 6 to 20 min.
Breakthrough characteristics were similar to those of
the GAC studies except that XE-347 TOC removal de-
clined more rapidly. TOC removal diminished to 50%
after about 5 BV were processed and appeared to stabil-
ize at about 10% for at least 120 BV.
Biological Treatment
Several attempts were made to acclimate an activated
sludge culture to raw groundwater. All attempts, how-
ever, were minimally successful. Neither a conventional
activated sludge nor a commercial microbial culture could
be acclimated. A light colored, filamentous biomass which
settled poorly appeared to be indigenous to the water
and inhibited a healthy biomass.
Approximately 80% TOC reduction was achieved;
however, stripping due to aeration appeared to account
for about two-thirds of this removal. Addition of trace
elements and nutrients, and pH adjustment to pH 7.0 to
7.5 did not aid microbial acclimation to raw ground-
-------
TREATMENT & DISPOSAL 291
Composite Groundwater (OW9 L Ill/d)
Raw OH9 Groundwater
150 200 250
TOC Loaded (mg TOC/g GAC)
Figure 1.
TOC Adsorption by GAC
water. Addition of powdered activated carbon (PAC)
at aeration chamber concentrations of approximately
10,000 mg/1 also did not aid acclimation to raw ground-
water or improve TOC removal or mixed liquor appear-
ance.
BENCH SCALE STUDIES OF PROCESS TRAINS
Granular Activated Carbon/ Aerobic Biological Treatment
Because of the apparent toxicity of groundwater to
biological processes and the rapid breakthrough of TOC
in adsorption systems, a process train consisting of gran-
ular activated carbon (GAC) followed by activated sludge
(AS) biological treatment was investigated. It was ex-
pected that GAC would protect the biological system
from toxic materials while organics leaking from the GAC
would be degraded or stripped during the biological
process.
Carbon pretreatment of raw groundwater permitted
development of a culture of aerobic organisms capable
of further treating GAC effluent. In excess of 95 % TOC
removal was realized by this process during the period
when TOC removal by the GAC exceeded 30%. After
this initial period, process train performance declined as
carbon performance declined. Data indicate that:
(1) Some fraction of TOC which initially is sorbed by
carbon begins to leak through the system after a
short period of operation
(2) The fraction of TOC which leaks through the car-
bon system is not toxic to the biological system but
does not appear to be removed or reduced either
biologically or by air stripping associated with AS
aeration
($) Operation of the biological process at hydraulic re-
, tention times ranging from 4 to 16 hr, with or with-
out powdered activated carbon, or with or without
Phenobac seems to have little impact on process
performance (based upon TOC removal)
(4) Overall system performance was maintained at 75-
85% TOC removal (effluent TOC of 100 to 185
mg/1) for about 21 days (46 retention times for bio-
logical system and 110 BVs for carbon).
As shown in Table III, organic priority pollutant analy-
ses conducted during operation of the GAC/AS Sys-
tem indicated that almost all of the organic priority
pollutants detected in the raw groundwater were removed
consistently to less than the level of detection (0.01 mg/1)
by the system. Additionally, the AS process completely
removed the few organic priority pollutants leaking
through the GAC system even during the phase of oper-
ation when overall TOC removal was declining. Analysis
of biological sludge showed no organic priority pollutants
at a 0.01 mg/1 detection level.
Granular Activated Carbon/Anaerobic Treatment
A GAC/upflow anaerobic filter (UAF) system was in-
vestigated using the composite groundwater and water
from well OW-9 (both high TOC waters). UAF per-
formance declined as GAC performance declined.
Changes in UAF organic loading rate did not appear to
affect UAF performance.
Overall, the performance of the GAC/UAF process
train was slightly poorer than the GAC/AS process train
achieving a maximum TOC removal of 81% when treat-
ing the composite groundwater. During this period TOC
removal averaged 66% and ranged from 38 to 81%. The
average TOC removal by the GAC process was 31 % with
a range of 10-46%; the average TOC removal by the UAF
process was 50% with a range of 12-67%.
GAC/Trickling Filter Treatment
The GAC/TF process train was investigated because
operation was expected to be easier and less energy inten-
sive than activated sludge and also less likely to induce
stripping of volatile organic compounds.
An unseeded trickling filter provided little or no re-
moval of TOC from GAC pretreated groundwater. A new
filter, initially seeded with activated sludge recycled for
about four weeks, developed a significant biomass on the
filter media. Although TOC removal was substantially
better than for the unacclimated system, performance
was poorer than achieved by the GAC/AS process train.
Moreover, as GAC performance steadily decreased, an-
aerobic conditions developed in the TF.
Granular Activated Carbon/Anaerobic/
Aerobic Biological Treatment
The GAC/UAF treatment process discussed earlier was
modified by the addition of an activated sludge process.
When TOC leakage from the GAC process was con-
siderable, the aerobic process reduced the UAF effluent
TOC by 50% or more. Otherwise, little additional TOC
removal was achieved by the AS process. The process
train provided TOC removals of 86 to 100%. The im-
portance of GAC pretreatment was demonstrated when,
-------
292 TREATMENT & DISPOSAL
Table III.
TOC and Priority Pollutant Data for Granular Activated
Carbon/Activated Sludge Process Train (mg/l)
Compound
TOC
Total Cyanide
CN
Total Phenol
Methylene chloride
1, 1-Dichloroethene
1, 1-Dichloroethane
Trans-1,2-dichloro-
ethane
Chloroform
1 , 2-Dichloroethane
1,1, 1-Trichloroethane
Trichloroethylene
Benzene
1, 1,2-Trichloroethane
Perchloroethylene
Toluene
Chlorobenzene
Phenol
2-Chlorophenol
2,4-Dichlorophenol
1,2-Di chlorobenzene
Dibutyl phthalate
Raw
Ground-
water
9-16-80
637
NA
NA
NA
2.1
1.6
2.4
0.06
9.8
72
7.6
0.06
1.2
0.11
0.49
2.3
0.23
0.025
0.040
0.010
0.085
ND
GAC
Effl.
9-16-80
380
NA
NA
NA
0.029
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Raw
Ground-
water
9-23-80
929
0.11
0.05
16
14
0.06
0.17
0.04
0.70
25
0.39
0.03
1.5
0.07
1.9
0.97
0.29
0.028
0.036
0.010
0.077
ND
GAC
Effl.
9-23-80
604
0.21
0.05
0.16
0.01
0.01
0.02
ND
0.06
1.4
0.04
ND
0.02
ND
ND
0.05
ND
ND
ND
ND
ND
ND
AS
Effl.
9-24-80
90
0.23
0.05
0.10
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
0.05
GAC
Effl.
10-1-80
770
0.23
0.05
0.10
0.16
ND
ND
ND
ND
0.05
ND
ND
ND
ND
ND
0.01
ND
ND
ND
ND
ND
ND
AS
Effl.
10-1-80
183
0.20
0.05
0.10
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
NA—not analyzed
ND—not detected
No other priority pollutants detected at 0.01 mg/l detection limit
within one day after carbon pretreatment was terminated,
UAF process performance declined markedly. Results
from this series of tests indicated that:
(1) As the UAF process became acclimated, its per-
formance improved and it accounted for a larger
share of TOC removal.
(2) As the carbon column performance declined, UAF
performance also declined.
(3) Generally, the AS process provided minimal TOC
removal. Aerobic biodegradation, stripping or ad-
sorption onto mixed liquid solids were negligible.
However, a normal appearing, well-settling acti-
vated sludge biomass, albeit one with a relatively
low MLSS (mixed liquor suspended solids) and low
sludge production, was observed.
(4) As TOC leakage from the carbon column increased,
overall system performance declined; however, the
amount of removal attributable to the aerobic pro-
cess increased.
(5) The overall system, under the operating conditions
studied, generally did not perform as well as the
GAC/AS train, i.e., it did not maintain low effluent
TOC levels (< 50 mg/l) for as long as the GAC/AS
train. However, both systems appeared to be able to
maintain effluent TOC levels below 100 mg/l for
equivalent duration.
Ozonation Treatment Studies
The effects of ozone (03) pretreatment were evaluated
using O/AS and O3/GAC/AS process trains. Ozonation
conditions were as follows:
•Air served as ozone source.
•03 dose—2g/hr.
-------
TREATMENT* DISPOSAL 293
•Oi flow rate—2 liters/minute.
•Batch 63 contact time—1 hour/3 liters of groundwater.
The process trains involving ozone pretreatment did not
perform as effectively as the other trains studied thus,
suggesting little benefit to pre-ozonation (in terms of TOC
removal aside from that attributed to stripping during
ozqriation). TOC removal by the O3/AS train ranged
from 37 to 60% while removal by the O3/GAC/AS
train ranged from 42 to 88 %.
While little change in TOC concentration may occur as
the result of ozonation alone, ozonation might be ex-
pected to enhance either the biodegradability or adsorp-
tion of organics. However, pre-ozonation did not appear
to improve the performance nor increase the longevity of
carbon treatment on the basis of TOC breakthrough.
CONCLUSIONS
Desktop evaluations identified a number of concen-
tration technologies that may be applicable for treating
hazardous leachates and contaminated ground and sur-
face waters. The most promising, broadly applicable
technologies were:
•Adsorption: carbon and resin
•Biological treatment (aerobic and anaerobic)
•Biophysical treatment
•Chemical coagulation
•Membrane processes
•Stripping: air and steam
These unit operations then were studied separately and
in combinations using groundwater from an actual haz-
ardous waste contamination site.
Carbon pretreatment of the raw groundwater permitted*
development of both aerobic and anaerobic biological pro-
cesses apparently by removing pollutants toxic to the bio-
logical processes. Highest TOC removals were achieved
with an activated carbon/aerobic biological (activated
sludge) system. However, as the carbon performance de-
clined, so did that of the rest of the system. The GAC/an-
aerobic biological process train performed somewhat more
poorly than the GAC/activated sludge system. Studies of
ozonation did not imply any advantage as a pretreat-
ment operation. The process of ozonating yielded TOC
removals similar to that achieved by air stripping.
ACKNOWLEDGEMENT
The work upon which this paper is based was per-
formed by Touhill, Shuckrow and Associates, Inc., under
Environmental Protection Agency Contract #68-03-2766.
Cordova Chemical Company, the present owner of the
Ott/Story Site, has been extremely cooperative in the con-
duct of this study. Special thanks to Mr. Barrel Cardy of
Cordova Chemical Company for his assistance toward the
completion of this study.
REFERENCES
1. Klepper, G., "Groundwater Pollution from the Ott/
Story Chemical Company Operations, Dalton Town-
ship, Muskegon County, Michigan." Michigan De-
partment of Natural Resources, Water Quality
Division, June 1981.
2. Shuckrow, A.J., Pajak, A.P., Osheka, J.W. and
James, S.C., "Bench-Scale Assessment of Technol-
ogies for Contaminated Groundwater Treatment,"
Proc. US.EPA National Conference of Uncontrolled
Hazardous Waste Sites, Oct. 15-17, Washington, DC,
HMCRI, Silver Spring, Md., 1980,184.
3. Shuckrow, A.J., Pajak, A.P. and James, S.C.,
"Concentration Technologies for Hazardous Aqueous
Waste Treatment," EPA-600/2-81-019, U.S. EPA,
Cincinnati, Ohio, 1981.
4. Pajak, A.P., Shuckrow, A.J., Osheka, J.W. and
James, S.C., "Assessment of Technologies for Con-
taminated Groundwater Treatment," Proc. of the In-
dustrial Wastes Symposia, Las Vegas, Nevada,
September 28-October 3,1980.
5. Pajak, A.P., Shuckrow, A.J., Osheka, J.W. and
James, S.C., "Concentration of Hazardous Constitu-
ents of Contaminated Groundwater," Proc. of the
Twelfth Mid-Atlantic Industrial Waste Conference,
Bucknell University, Lewisburg, Pennsylvania,
July 13-15, 1980.
6. Shuckrow, A.J. and Pajak, A.P. "Bench Scale As-
sessment of Concentration Technologies for Haz-
ardous Aqueous Waste Treatment." EPA-600/9-81-
002 b, Land Disposal: Hazardous Waste., Proc. of
the Seventh Annual Research Symposium, Philadel-
phia, Pennsylvania, March 16-18,1981.
-------
DESIGN OF A TREATMENT SYSTEM FOR
HAZARDOUS RUN-OFF
PETER B. LEDERMAN, Ph.D.
JOHN W. HAMMOND
Roy F. Weston, Inc.
West Chester, Pennsylvania
JOSEPH P. LAFORNARA, Ph.D.
Emergency Response Team
U.S. Environmental Protection Agency
Edison, New Jersey
INTRODUCTION
On July 30, 1979, the State of Pennsylvania requested
that the U.S. Environmental Protection Agency (EPA)
conduct an on-site assessment of an abandoned mine
drainage tunnel, known as the Butler Tunnel, from which
an unknown discharge was entering the Susquehanna
River. Approximately 500 to 1,000 gal/day of oily sludge,
along with an assortment of cutting oils, aromatics,
substituted phenols, alkyl resins and short and branched
chain hydrocarbons were being discharged from the tun-
nel. This discharge was traced to an alleged illegal dump-
ing of millions of gallons of toxic wastes into an aban-
doned coal mine through a sewer borehole at Dupont,
Pennsylvania, two miles inland from the river.
Roy F. Weston, Inc. was retained by EPA to de-
termine alternative approaches to treatment and/or de-
contamination of the toxic oily material that might be
found in coal mine pools being drained by the tunnel,
as a result of exploratory drilling (by other EPA con-
tractors) at Pittston, Pennsylvania. In addition, Weston
was to conduct a similar investigation on the Butler
Tunnel. For both of these investigations, the consul-
tant's work was to include physical and chemical char-
acterizations of the contaminated material, and an evalu-
ation of disposal techniques for treating the wastes.
The project had four phases:
Phase I: pertinent available data on mine pool wastes
were to be collected and reviewed. Analytical protocols
for characterization of the mine pool wastes were to be
developed. Potential alternatives for concentration and
disposal of the wastes were to be determined.
Phase 2: data from treatment studies previously per-
formed on the effluent from the Butler Tunnel were to
be evaluated. Supplemental treatment studies, if neces-
sary, were to be recommended. Alternative methods
for temporarily treating the effluent were also to be
recommended.
Phase 3: waste samples collected from the mine polls
were to be analyzed. Cost estimates were to be de-
veloped for feasible treatment alternatives.
Phase 4: supplemental analytical and treatability studies
on the Butler Tunnel effluent were to be performed,
which were necessary to recommend a temporary
treatment scheme for the tunnel's effluent.
Reports summarizing the results of Weston's overall
project as well as an overview of the over all clean-up op-
eration have been presented elsewhere.ll>2)
The purpose of this paper is to present a description of
the treatment process train which was designed to remove
hazardous pollutants from the Butler Tunnel effluent. The
development of the design basis, results of treatability
studies, the design and costs for installation and operation
of the equipment are presented.
CHARACTERIZATION OF TUNNEL DISCHARGE
Since August of 1979, when the Butler Tunnel clean-
up program was initiated, samples of the Butler Tunnel
discharge were systematically collected and analyzed for
oil and several classes of toxic organic chemicals. These
included analysis for dichlorobenzenes, disctyephthalate,
alkylphenols, toluene, xylene, ethylbenzene and naphtha-
lene. After a period of preliminary screening of the efflu-
ent, dichlorobenzene (DCB) was chosen as an indicator
parameter in order to reduce the analytical load to a man-
ageable level. In addition to these analytical data, the dis-
charge flow rate from the tunnel to the Susquehanna was
measured.
Presented in Figure 1 is the cumulative distribution of
flow rates from -the Butler Tunnel from August 1979 to
April 1980. As indicated, the flow rates during this period
ranged from approximately 1 mgd to greater than 12 mgd.
Statistically, on 20% of the days the flow rate was less
than 3.0 mgd, on 40% of the days it was less than 6.0 mgd
and on 80% of the days it was less than 12.0 mgd.
Presented in Figures 2 and 3 is a comparison of the oil
and DCB loads (lb/day) discharged versus the flow rate
(mgd) of the discharge on the same day. During this per-
iod, the lowest flow which could be accurately measured
was 2 mgd, therefore this flow rate was assumed for days
on which the flow rate was reported as "less than 2 mgd."
A review of Figures 2 and 3 shows that the most signifi-
cant discharges of oil and DCB, in terms of lb/day to the
Susquehanna River, occur when the tunnel flow rate is
low. The sensitivity level for oil was estimated to be 1.0
to 1.5mg/l.
294
-------
TREATMENT & DISPOSAL 295
"O 12.0
Ol
5 10.0
o
There Were 31 Days Where the Flow:
Were Only Specified as > 12 mgd
This Point Represents 16 Days
Where Rows Were Estimated as*,
2 to 9 mgd
This Point Represents 52 Days
^-Were Rows Were Estimated as
t (o 2 mgd
Percentile
Figure 1.
Cumulative Distribution of Flows
Pittston, Pennsylvania (Butler Tunnel)
August 1979 to April 1980
The results of priority pollutant analysis of a sample of
the tunnel effluent are presented in Table I. On the day
that this sample was collected, the discharge was in excess
of 12 mgd. These data confirm that the Butler Tunnel has
been contaminated by various solvents and other organic
species. Review of the data in Table I shows that the base-
neutral extractable fraction represents approximately 90%
of the total organics reported on the table.
TREATMENT REQUIREMENTS
In general, the analysis of the data for the Butler Tun-
nel discharge confirmed what was previously suspected.
The principal contaminants in the discharge are oil and
dissolved organics.
Regulations governing the quality of this discharge are
found in Chapter 93—Water Quality Standards, of the
Pennsylvania Code, Title 25—Environmental Resources.
Review of this chapter indicated that there were no spe-
cific water quality criteria for the principal contaminants
in the Butler Tunnel discharge. Therefore, Section 93.8—
Development of Specific Water Quality Criteria for the
Protection of Aquatic Life, requires that 96-hour continu-
ous flow bioassays be run, using the methodologies re-
ferenced in Subsection 93.8e of Chapter 93. However,
during the analytical program, bioassays were not run.
Therefore, it was assumed, based on conversations with
EPA, that treatment of the Butler Tunnel wastes for oil
and organics, such as those in Table I, would be re-
quired to levels which represented reasonable reduction of
the initial concentrations.
TREAT ABILITY INVESTIGATION
While developing the preliminary treatability plan, vari-
ous specific aspects of the overall emergency program
were considered including:
1. In order to successfully treat, for all the contaminants
in Table I, it would probably be necessary to use acti-
vated carbon. Alternatives to activated carbon which
were initially considered, but then rejected, included:
•Air stripping of the organic species—while this might
be reasonably effective on those species found in the
volatile fraction in Table I, it would probably have
little effect on the base-neutral extractable fraction,
which represented around 90% of the total organics
present.
•A low-level biological system—this was rejected be-
cause of the low values for BOD5 in the raw waste;
also the significant area requirements that would be
needed to treat the probably flow rates required. (The
available space for on-site treatment was limited to an
area of 250 ft x 70' on a ledge overlooking the mouth
of the tunnel.)
2. If activated carbon was to be used, then it would be
necessary to minimize any fouling problems due to oil,
or material which periodically sloughed from inside the
tunnel. For this reason, Weston decided that solids
separation facilities were necessary before the activated
carbon system and that it would also be necessary to
Table I.
Organic Analysis (GS/MS)
Butler Tunnel
March 20,1980
Pittston, Pennsylvania
A. Base-Neutral Extractables (yug/L)
1,2-Dichlorobenzene
Naphthalene
Diethylphthalate
Bis (2-Ethylhexyl) Phthalate
B. Acid Extractables (/ug/L)
None detected
C. Volatile Fraction (/zg/L)
Methylene Chloride
Trans-l,2-Dichloroethylene
Chloroform
1,1,1 -Trichloroethane
1,1,2-Trichloroethane
Trichloroethylene
Benzene
2-Chloroethylvinyl Ether
D. Pesticides
None detected
164.0
7.6
13.4
192.9
1.6
6.5
1.3
3.4
16.7
6.8
1.3
2.1
-------
296 GEOHYDROLOGY MATH MODELING
periodically inject chlorine, for disinfection of the ac-
tivated carbon.
3. If activated carbon was going to be used, pH adjust-
ment of the raw wastewater may be necessary.
To implement the treatability plan, the following tests
were run:
•Activated carbon isotherms.
•Chlorine demand.
•Titration curves.
In order to economically conduct this preliminary work,
DCB was chosen as an indicator parameter for analysis of
the isotherms.
Presented in Figure 4 is an isotherm run on the tunnel
effluent. As indicated in the figure, activated carbon may
be used to treat the Butler Tunnel wastewater to DCB
equilibrium concentration of less than lOMg/1- To establish
chlorine residual of 2 mg/1, a chlorine dosage of 8 mg/1
was required. To adjust the pH of the raw wastewater
from 4 to neutral, approximately 40 mg/1 of caustic (50
mg/1 as CaCo3) was required.
&
X
ra
•o
1.
Flow (mgd)
Figure 2.
Oil Load
Pittston, Pennsylvania (Butler Tunnel)
August 1979 to April 1980
PROCESS DESIGN
As a basis for its concept design, Weston used the re-
sults of its data analysis and treatability tests and, in addi-
tion, incorporated the following factors:
•Based on conversations with the EPA project officer, it
was decided to investigate designs for two flow rates—
-The most significant discharges of contaminants from
the Butler Tunnel have occurred on days when the flow
rate was 3.0 mgd and 6.0 mgd. The reasons for this are:
-The construction of facilities much in excess of 6.0 mgd
might be prohibitive from an economic point-of-view.
•The limitations on available land area, for an on-site
facility.
Presented in Figure 5 is the general process flow dia-
gram for the Butler Tunnel treatment system. As indicated
I
I
Flow (mgd)
Figure 3.
DCB Load
Pittston, Pennsylvania (Butler Tunnel)
August 1979 to April 1980
in the diagram, the Butler Tunnel discharge would be
treated in the following sequence:
•Water from the tunnel will flow into a wet well through a
bar screen that will remove large solids. This wet well
will be constructed at the mouth of the Butler Tunnel,
below the ledge. The west well will have an emergency
overflow provision to the Susquehanna River, so that
flows in excess of the design flow can be bypassed.
•Water will be pumped from the wet well to a flow di-
vision box. The purpose of the flow division box is to
divide the flow before being fed to several corrugated
plate separators, operating in parallel, that will remove
suspended solids and floatable oils. It will be important
in the final design of the facility to seiect lilt pumps,
whose operation will keep oil/water emulsions to a
minimum.
•As described above, the corrugated plate separators will
remove solids and floatable oils; these two streams will be
sent to holding tanks prior to ultimate disposal. Cor-
rugated plate type separators were selected over other
types of separators because they will minimize the land
area required for the unit operation.
•After treatment by the separators, the wastewater will
go through a pH adjustment step (to a final pH near
6.5-7), flow through another flow division box and then
to parallel trains of activated carbon adsorbers.
•The preliminary design contraints for the carbon ad-
sorbers were 30 minutes contact time and 3 gal/min/ft2
application rate. Commercially available adsorbers were
considered first; however, in order to satisfy these de-
sign constraints and to allow for mobility and accessi-
bility at the site, more land area was required than was
available. For this reason, Weston recommended the con-
struction of four adsorbers—two trains with two stages
each. Each adsorber would be compartmented into
quadrants with a common centerwell. The compartment-
alization will allow for backwashing facilities of reason-
able capacity and size to be designed.
-------
GEOHYDROLOGY MATH MODELING 297
10,000i-
1,000
t
8
7
6
5
Plttston, Pennsylvania
Butler Tunnel
28 March 1980
C (1,2-Dichlorobenzene) - v g/L 100
Figure 4.
Carbon Isotherm for 1,2-Dichlorobenzene on Carbon
Filtersorb 300 in Butler Tunnel Wastewater
•After being treated in the carbon adsorbers, the flow
will go to an effluent sump which will be used to store
backwash water, and also to allow for chlorination of
effluent prior to discharging to the Susquehanna River.
Additionally, the operator will be able to chlorinate the
influent to the carbon adsorbers periodically, to mini-
mize bacterial growth on the adsorbers.
•Water used for backwash will go to a backwash receiv-
ing tank. This tank is necessary to handle the surges in
the flow during backwash. After entering the backwash
receiving tank, the water will be retained so that solids
and oil can be separated from the bulk of the water.
These solids and oil will be separated and sent to the
sludge and oil holding tanks respectively.
•Oil from the oil holding tank will be pumped on a peri-
odic basis into tank trucks for transport to an acceptable
waste oil disposal facility, such as Rollins in Bridgeport,
New Jersey.
•For the purpose of this design, the consultant has as-
sumed that solids to be disposed of, will also require
disposal as a hazardous waste. Such disposal is expensive
(estimates are typically $125/ton); therefore, a centri-
fuge has been included to dewater the solids prior to
disposal.
Presented in Figure 6 is a proposed equipment arrange-
ment site plan that Weston developed for the 6.0 mgd de-
sign case. As can be seen by this figure, the proposed
6.0 mgd facility will occupy all available space.
Corrugated
Plate Separators
Centrifuge
To Truck Disposal
River
Backwash Pump
Figure 5.
Pittston Butler Tunnel Process Flow Diagram
-------
298 GEOHYDROLOGY MATH MODELING
Carbon
Contact Tank
(30' ef)
Carbon
Contact Tank
, (30' 0)
Backwash
Receiving Tank
(40' 0)
fi\.— wet Well
Notes and Assumptions
1 Available Area is 250' x 100' with 30' Required Setback. Per Northeastern
Engineering Report - Page 2.
2 Corrugated Plate Separator Requires 25' x 50' Plan Area.
3 Trailers are Approximately 12' x 50'. and Allow 1/2 Length of Trailer
For Mobility and Accessibility.
4 Backwash Receiving Tank Sized to Receive Backwash from Two Contactor
Compartments, Each with Surface Area of Approximately 160 ft2,
and a Backwash Rate of 20 gpm/ft2 for 15 Minutes
5 Oil Tank Sized for 10.000 Gal. to Allow for Intermittent Tank Truck
Removal.
6 Solids Holding Tank Sized for 1,000 Ib/day of Solids (6 mgd, 20 mg/L TSS),
Two Days Storage Capacity at 1°0 TSS (10,000 mg/L).
7 Carbon Contactors Will Be Installed 10' Apart to Allow for Piping, Etc.
Figure 6.
Pittston, Butler Tunnel
Proposed Equipment Arrangement Site Plan
COST ESTIMATES
Estimates have been made for the installed capital
costs for the major equipment included in the treatment
system (Table II.) The estimates of the total installed
capital cost for the 3.0 mgd and Butler Tunnel Treatment
System are $2,279,400, and for the 6.0 mgd System, the
estimate is $3,434,700. These estimates are based on an
ENR Index of 3160, and do not include certain direct
costs associated with construction such as engineering,
construction management and startup. These cost esti-
mates are considered to be accurate to ±30%.
Estimates of the operating costs (Tables III and IV)
for the Butler Tunnel Treatment System were also de-
veloped. The estimate of the total operating cost for the
3.0 mgd System is $l,347,000/year, and for the 6.0 mgd
System, the estimate is $2,559,000/year.
Review of the data in these tables indicates that the car-
bon represents the most significant cost. The principal
factors which will affect these costs are:
•Raw Wastewater Strength: as the concentrations of con-
taminants in the raw wastewater decrease, the life of an
existing carbon bed increases. The value of raw waste-
water concentrations, used in making these estimates of
operating costs, was the median of the values for DCB.
If the concentration of DCB in the Butler Tunnel dis-
charge decreases, the life of the adsorption bed would
correspondingly increase.
•Allowable Effluent Concentration at Carbon Exhaustion:
the effluent concentration will directly influence ad-
sorption capacity of the carbon bed. As shown for the
carbon isotherm in Figure 4, as the equilibrium concen-
tration of DCB in solution increases, the Ib DCB ad-
sorbed/lb carbon increases. Increasing the allowable
final effluent concentration of DCB at breakthrough
from 20 /ug/1 to 400 /zg/1, reduces the estimated carbon
replacement cost by over 70%.
•Unit Cost of Replacement Carbon: while this may be ob-
vious, it is a significant point. Fundamentally, the cost
per pound for replacement of a particular carbon must
-------
GEOHYDROLOGY MATH MODELING 299
Table II.
Estimate of Installed Capital Costs for
Butler Tunnel Treatment System
(ENR Index = 3160)
Estimated $
Item
1. Life Pump Station
2. Flow Division Box
3. Corrugated Plate Oil Separators
4. pH Adjustment Facility
5. Flow Division Box
6. Carbon Adsorption Facility
7. Sludge Holding Tank
8. Oil Holding Tank
9. Chlorination Facility
10; Centrifuge
Subtotal
11. Miscellaneous1
12. Contingency
Total1
t. Includes piping, electrical, instrumentation and site work.
2. These estimates do not include certain direct costs associated with construction of the facilities
such as engineering, construction management and start-up.
Table in.
Estimate of Operating Costs for Butler Tunnel Treatment System
(3 mgd Design)
Pittston, Pennsylvania
(May 1980)
3 mgd
172,400
25,800
115,100
63,900
25,800
876,100
32,400
22,300
73,000
178,900
1,585,700
396,400
297,300
2,279,400
6 mgd
217,200
25,800
276,100
90,900
25,800
1,368,500
46,600
28,700
124,800
185,000
2,389,400
597,300
448,000
3,434,700
1. Labor
27 manhours per day
$11 per hour (salary and benefits)
52 weeks
7 days per week
2. Maintenance (mechanical, structural, electrical,
instrument)
3 % of capital costs other than carbon inventory
3. Electrical
HPat$0.50/kwh
4. Chemicals
Caustic: 350 tons/year at $200/ton
Chlorine: 36.5 tons/year at $400/ton
Carbon': 833 tons/year at $l,200/ton
5. Sludge and Waste Oil Disposal
Sludge Disposal: 600 tons/year at $125/ton
Waste Oil: 21,900 gal/year at $0.83/gal
6. Total Estimated Operating Costs
$ 108,100/year
$ 42,500/year
$ 17,400/year
$l,084,200/year
$ 94,400/year
$l,347,000/year
Table IV.
Estimate of Operating Costs for Butler Tunnel Treatment System
(6 mgd Design)
Pittston, Pennsylvania
(May 1980)
1. Labor
27 manhours per day
$11 per hour (salary and benefits)
52 weeks
7 days per week
2. Maintenance (mechanical, structural, electrical,
instrument)
3 % of capital costs other than carbon inventory
3. Electrical
HP at $0.05/kwh
4. Chemicals
Caustic: 700 tons/year at $200 ton
Chlorine: 75 tons/year at $400/ton
Carbon': 1,670 tons/year at $ 1,200/ton
5. Sludge and Waste Oil Disposal
Sludge Disposal: 1,215 tons/year at $125/ton
Waste Oil: 43,800 gal/year at $0.83/gal
6. Total Estimated Operating Costs
$ 108,100/year
$ 62,300/year
$ 26,300/year
$2,174,000/year
$ 188,300/year
$2,559,000/year
1. Actual quantities and costs for carbon replacement will depend on allowable effluent concen-
tration at breakthrough, and the actual adsorption capacity of the carbon which is finally
selected for use. The value indicated reflects the probably order of magnitude of these costs,
assuming effluent DCB concentration at breakthrough of less than 50ng/l.
1. Actual quantities and costs for carbon replacement will depend on allowable effluent concen-
tration at breakthrough, and the actual adsorption capacity of the carbon which is finally
selected for use. The value indicated reflects the probable order of magnitude of these costs,
assuming effluent DCB concentrations at breakthrough of less than 50 lig/1.
be combined with the adsorption capacity of that carbon
to, in effect, establish the cost per pound of contaminant
removed. In general, two types of carbon are available.
The first type of carbon is virgin carbon; the second type
is reactivated carbon. The costs and capacities of virgin
carbon are higher than for reactivated carbon. Virgin
carbon is available from a greater number of suppliers
than is reactivated carbon; therefore, purchases of virgin
carbon are done more in a "buyer's market" than are
purchases of reactivated carbon. Of course, in either
case, unless some form of service contract is established,
variations in spent carbon disposal costs/resale value
must also be considered.
REFERENCES
1. "Report to On-Scene Coordinator: Treatability Study
Major Pollution Incident Butler Tunnel, Pittston,
Pennsylvania", Roy F. Weston, Inc., West Chester,
Pennsylvania, 1980.
2. Lafornara, J.P., Hammond, J.W., Lederman, P.B.
and Massey, T.I., "The Pittston Story", Proc. of 1980
U.S. EPA National Conference on Management of
Uncontrolled Hazardous Waste Sites, Oct. 15-17, 1980,
Hazardous Materials Control Research Institute,
Silver Spring, Md., 250-254.
-------
COMPUTER-ENHANCED GEOPHYSICAL SURVEY
TECHNIQUES FOR EXPLORATION OF
HAZARDOUS WASTE SITES
G.L. McKOWN, Ph.D.
G.A. SANDNESS, Ph.D.
Battelle, Pacific Northwest Laboratory
Richland, Washington
INTRODUCTION
Collection of chemical and geological data during site
assessment activities is an integral and necessary part of
hazardous waste site investigations. Among the first in-
vestigative methods that might be employed at a site are
geophysical surveys. These ex situ studies are based on
geohydrologic changes induced by waste materials, and
can provide a more complete picture of subsurface
phenomena prior to in situ sampling and well drilling.
Included among the potential applications are:
Location and/or Delineation of the Boundaries of Sus-
pected Burial Sites.
Many of the techniques used depend on a change in
density, reflectivity, susceptibility, or permittivity of tar-
get materials, all of which are characteristic of buried
wastes. This type of underground location is readily
accomplished by a magnetometer/radar combination that
has been successfully employed by Battelle-Northwest in
many locations. The ability to accurately locate burial
trenches, individual targets such as drums, and concen-
trations of landfilled trash/wastes has been demonstrated.
Other techniques may be used if poor resolution is accep-
table.
Location of Waste Containers.
At many sites, the possibility exists that high con-
centrations of hazardous materials could be encountered
when probing underground. This is particularly true at
uncontrolled sites and locations where burning/burial has
occurred. To minimize the hazard of drilling or coring
into such deposits, a survey should be made before drill-
ing commences. Radar, magnetometry and metal detec-
tion are applicable tools since the targets do not provide
adequately defined acoustic reflection and resistivity
measurements are too difficult to interpret unless detailed
surveys are conducted.
Location of Underground Pipes and Cables.
Wells or cores must not be drilled through underground
utility conduits. This is a problem at any site where in-
dustrial operations are concentrated and where insufficient
information on underground utility routes exists.
Several techniques are available for locating or trac-
ing utility conduits. Acoustic or radiofrequency genera-
tors can be used to drive a metallic line and the route traced
by detectors from the surface. Resistivity measurements
can provide information on nests or concentrated runs of
metallic objects. For the most general case where either
metallic or nonmetallic conduits are possible, a combina-
tion of magnetometry and radar offers the greatest def-
inition. About the only situation that may not be iden-
tified with this combination is plastic gas lines and the
location of these are likely to be documented since plastic
pipe has not been used until relatively recent times.
Location of Discontinuities in Underground Strata.
Radar or acoustic systems can be employed to detect dis-
continuities, either natural or manmade, in an underlying
stratum or bedrock. Acoustic/seismic methods can be em-
ployed if the voids are large (on the order of 100 m2 or
greater), whereas radar can detect much smaller holes
(on the order of 1 m2 or less). However, the radar sys-
tem is limited in depth of penetration to 3 to 10 m for
typical lithology.
In either case, the presence of intervening strata con-
founds reflection patterns and limits the use of these meth-
ods. Other techniques may be used in special instances. For
example, magnetometry is an excellent tool for location
of voids/faults in magnetic bedrock such as basalt and
intervening layers are acceptable if they are spatially con-
gruent. Gavimetry has similar applicability, although
resolution is poorer and might be used to detect mine
shafts, limestone caverns and sinkholes that are evident on
the surface.
Location of Water-Bearing Strata.
Since the dielectric contant of water-bearing materials
differs considerably from other geologic structures, it
should be possible to use radar systems to map eleva-
tion and slope of the water table. On surveys to date, the
authors have had only moderate success in delineating such
features. The water table in one case, was quite obvious
in radar scans, but this represents a favorable case (Depth
of 2 to 3 m and salt water). In other cases, the depth to
ground water has been too great (> 10 m) or the aqui-
fers not sharply defined enough to allow meaningful in-
terpretation. In some instances, the presence of interven-
ing strata or discrete objects (rocks) totally masks the
effect.
300
-------
GEOHYDROLOGY MATH MODELING 301
In favorable cases, survey techniques might be used to
define the limits of a contamination plume. The ideal case
would be highly contaminated sources infiltrating at a high
rate. Electrical resistivity would be most applicable in at-
tempting to delineate the plume; very low frequency
(VLF) resistivity techniques probably offer the greatest
chance for success.
Location of Abandoned Wells and Drill Sites.
^Occurrence of old wells, abandoned drill holes, burial
pits, sinkholes, and other penetrations of near surface im-
permeable layers provide potential infiltration points for
contaminated surface water. It has been shown that radar
works well in locating such areas of disturbed earth, es-
pecially considering that backfilling these holes would
rarely have been done according to good engineering prac-
tices.
Infrared imaging has also been used with less spectac-
ular success in locating old excavations. Acoustic holo-
graphy is a relatively new technique that is ideal for plot-
ting depth and characteristics of boreholes once they have
been located.
Location of Solution Channels. Cracks and Fissures.
Geophysical surveying in this application is less cer-
tain, and depends on characteristics of the specific site.
Acoustic/seismic reflection probably has the widest appli-
cation, although interpretation would be extremely diffi-
cult in complex matrices.
Radar probably is not generally applicable, although
specific instances may be amenable to this treatment.
Tracer methods are applicable, but require an extensive
experimental program.
Location of Objects of Historical Interest.
Some Army installations contain sites of great arch-
aeological interest. Although exploration is not an objec-
tive of site survey programs, geophysical methods pro-
vide aid in minimizing disturbance of archaeological sites
during conduct of on-site exercises.
Since most discrete objects and structural features of
archaeological interest are located within the first few
meters of the surface, a radar system that provides under-
ground map views is an excellent tool for these investiga-
tions. The potential for this application was demonstrated
during a radar survey of Fort Monroe for the U.S. Army.
Geophysical Logging.
Various down-hole analytical techniques may be applic-
able to site assessment programs. The porosity of media
down boreholes can be determined by neutron-epithermal
neutron methods, which in some cases can be used to de-
rive permeabilities. Natural gamma radiation logs can be
used to determine certain lithologic changes and estimates
of geochemistry. Electrical logs provide a more detailed
observation of water strata than can be obtained from in-
spection of cutting returns.
GEOPHYSICAL SURVEY SYSTEM
Researchers at Battelle-Northwest have selected metal
detection, magnetometry and ground-penetrating radar
for inclusion in an integrated survey unit for burial ground
and other subsurface investigations. Radar transceivers
are available for operation at frequencies of 100, 300, 500
and 900 mhz.
The vehicle was specially constructed of nonferrous ma-
terials to facilitate the use of a magnetometer mounted
alongside the radar unit. In addition to transporting the
radar and magnetometer, the vehicle contains a micro-
computer and a telemetry system which acquires, pre-
processes, and transmits radar/magnetic data to a larger
computer at the survey site. Recent additions include an
all-terrain Trackster transporter and a motor home for
transport between sites and to permanently house the
computer systems.
A subsurface survey is accomplished in four steps.
1. A grid covering the study area is marked on the ground
surface to guide subsequent measurements and to pro-
vide location coordinates.
2. Manual survey of the grid using hand-held metal de-
tectors, marking all identified targets and transferring
locations to an overlay map.
3. Operation of the survey vehicle over the same grid,
accumulating radar and magnetometer data. The data
produced by the survey instruments is transmitted to a
PDF 11-34 computer where it is processed, formatted,
and stored on either a magnetic tape or a disk storage
unit. Preliminary data manipulation and processing
functions can be performed on site to edit and com-
press the data for final storage or to facilitate data
monitoring or previewing during a survey. This near
real-time data handling capability can be of immeas-
urable value during conduct of a site survey.
4. Data superposition and analysis is done mainly by a
unique software package created to filter and enhance
output. Graphic display devices are used first to pro-
vide vertical profiles of each radar survey run. The
profiles are then combined in digital form and sliced
horizontally to generate a map view of a selected depth
interval in the area surveyed. When specialized en-
hancement programs are applied, the output appears
as color photographs with specific hues depicting bur-
ied targets. This system has been applied to surveys
conducted in Idaho, Washington, New York, Virginia,
Indiana, Pennsylvania, Utah and Nevada.
DATA PROCESSING AND DISPLAY
As received and recorded by the computer in the field,
the radar data are not in an optimal form to be displayed
and interpreted. Subsequent digital data programs can
enhance the quality and usefulness of the data. Several
computer programs have been developed at Battelle to
enhance and display the radar and magnetometer survey
date.
Figure 1 is a flow chart showing the computer pro-
grams involved in the data recording/processing/display
-------
302
GEOHYDROLOGY MATH MODELING
MACKTCMETHJJ 3ATA
jPWAfiO AMD , CALCULATION . REDACTION
DOWNWARD j Of SECOND TO
CC*JTI'.LAT!GN ' ZESlVA'IVES 1 POU
CONTOUR PVOTTISC
-.rfjsir' VOOULATIO
i Pv-GTQGftAPMlC OUTPUT
'SCALING FOR
COlCfi OUTrLT
Figure 1.
Computer Program Flow Chart for Radar
and Magnetometer Data
sequences for radar and magnetometer data. The
GEOSUR and ISA programs are for recording and manip-
ulating the data after it has been transmitted to the com-
puter from the survey vehicle, and are not described
further.
The stored digital radar data can be thought of as a
three-dimensional array containing 120 bytes/ping, N
pings/track, and M tracks. The data processing programs
REMVH and HYPER operate on two-dimensional radar
data; that is, they operate on each track in sequence. Pro-
gram SLICE operates on the entire three-dimensional ar-
ray to construct a two-dimensional array.
Program REMVH
A characteristic of the reflected radar signals is that they
contain components that are unchanging from ping to ping
and convey no information about subsurface reflectors.
These components are due to a combination of factors:
the oscillatory shape of the waveform, the presence of a
ground surface reflection, random reflections from the
survey vehicle and the radar unit itself, and electronic
effects in the radar unit. These unwanted signal com-
ponents appear in a radar image as alternately bright and
dark horizontal bands. The left half of Figure 2 illustrates
this effect. In each image of Figure 2, the horizontal di-
mension corresponds to distance along the traverse line,
and the vertical dimension is proportional to the travel
time of the radar signal (or to depth).
The purpose of program REMVH is to remove the
banding due to stationary signal components. Let I (1 < I
< 120) be the datum index in a given ping, and let J
(1 < J < N) be the ping index in a given track (N is the
number of pings in the track). For each fixed value of I,
the program computes a running histogram as it scans the
array in the J direction. At a given point, (I, J), in the
array, the running histogram includes all data with the
bounds II = 1-1 to 12 = 1+ 1 and Jl + J-JA/2 to J2 =
J + JA/2, where JA is a specified number. At the edges
of the array, the bounds are as follows: if I = 1, II = 1; if
I = 120, II = 120; if J = 1, Jl = 1; and if J = N, 32 =
N. The banding is removed by computing an average value
from the data within the bounds of the running histogram,
then subtracting the average from datum I, J. How-
ever, if the row contains values that are substantially high-
er or lower than the average value, this simple procedure
can yield local correction values that are either too high or
too low. This has the effect of introducing a spurious hor-
izontal oscillation in the data. The histogram provides a
simple means for excluding a range of high and low val-
ues. Improved corrections are typically obtained by ex-
cluding the highest and lowest 10% of the data from the
calculation of the average.
The right iialf of Figure 2 shows the effect of apply-
ing program REMVH to the raw radar data. This ap-
proach to radar image enhancement improves the detec-
tability of discrete buried objects or irregular masses of
waste material, but it is not effective in cases where it is
necessary to detect flat, horizontal interfaces.
Program HYPER
The function of this program is to remove hyperbolic
reflection patterns from radar images, or reflection pro-
files- Hyperbolic patterns occur because a radar output
beam has a finite angular width. Reflections are there-
fore received from a buried object both before and after
the sensing transducer passes directly over it. It is some-
times desirable to remove the hyperbolic patterns from the
radar profiles because they can increase the difficulty
of interpreting the data.
Program SLICE
Ground penetrating radar surveys normally involve
cumbersome methods for mapping detected buried ob-
jects. The locations of buried objects are usually deter-
mined visually from a collection of radar profiles, then
manually transcribed to a map. The purpose of program
SLICE is to utilize a digital computer to accomplish the
same task.
Program SLICE is essentially a routine to construct a
selected two-dimensional data array from the original
three-dimensional array. A depth range is specified in
terms of data indices which range from 1 to 120. Then for
each ping in the array, a new array element is derived
by averaging the ping over the specified depth range. The
dimensions of the new array are N x M, where N is the
number of pings per track and M is the number of tracks.
Because the original data spacing along a survey track is
much less than the track spacing, it is necessary to ex-
pand the new array in the across-track direction. This is
accomplished by linear interpolation. Interpolation in both
directions allows the array size to be completely selec-
table.
-------
GEOHYRDOLOGY MATH MODELING 303
Figure 2.
Two Forms of a Single Radar Profile Illustrating
Computer Processing Operations
In cases where the distribution of buried objects is suf-
ficiently sparse to allow radar reflections from a wide
range of depths, a single, thin, horizontal slice will show
only those objects contained in the corresponding narrow
depth range. Examples of the output of program SLICE
are shown in Section 5.
Program IDAPIC
Preliminary video monitor and hardcopy displays of raw
and processed radar data can be made almost in real time
by means of a data previewing capability included in pro-
gram IDA. Final output products, however, are made in
photographic form by utilizing a DICOMED D48, black
and white or color, digital film recorder.
The function of IDAPIC is to transmit IDA formatted
radar data to the film recorder. It provides options for
file subsetting, image contrast enhancement, and color-
coded level slicing.
Program PICT
This is a general purpose program for producing photo-
graphic images of raster-scan data from a wide variety
of sources. In particular, it accepts the file structure of
the output data produced by program SLICE. It is other-
wise similar in both form and function to program
IDAPIC.
Magnetometer Data
The data acquisition system on the survey vehicle is de-
signed to transmit either one or two magnetic field values
along with each transmitted radar ping. Two magnetic
field values are allowed to order to provide for the pos-
sible future use of two magnetometers. However, only one
has been used in past and current programs. The magneto-
meter data is, therefore, basically a two-dimensional layer
in the three-dimensional IDA data array.
Program MAG
Numerous methods for analyzing magnetic data have
been described in the literature. Commonly, these methods
involve data manipulation operations such as upward
and downward continuation of the magnetic field, calcula-
tion of vertical derivative and reduction to the pole. Pro-
gram MAG performs all three of these operations. Its
basic function is to systematically alter or enhance the
magnetic field patterns in such a way that they can be more
easily interpreted.
-------
304 GEOHYDROLOGY MATH MODELING
In a region free of magnetic materials, the total magnetic
field anomaly. AB, satisfies Laplace's equation;that is:
t 82(AB) | 32(AB)
dy- d/-
(1)
By the method of separation of variables, the solution of this
differential equation can be expressed as:
B(x.y./) = £ £ f 7s,
• z
where
(2)
mx
f (-—.yM =a two dimensional Fourier series expansion
\ x ' / of the magnetic field in the measurement
plane z = 0
Lx = MAx
L6 =NAy
M and N are the number of data in the x and y directions,
respectively, and Ax and Ay are the data spacings in the x and
y directions, respectively.
Upward and downward continuations of a magnetic
anomaly are calculated by inserting the appropriate values
of z into the above summation. Vertical derivatives of AB
are calculated by differentiating the exponent within the
summations.
Program CNT
Program CNT is a system routine on Battdle labora-
tory computers. Its function is to plot contour mapi
from various kinds of data arrays. Hardcopy output prod-
ucts are typically drawn on a CalComp plotter.
Program M AGP
It is often useful and desirable to display magnetic data
in the form of a color picture. This is conveniently accom-
plished by means of program PICT. However, the num-
ber of magnetic measurements made in a given area is
generally small in comparison with the number of picture
elements needed to form a pleasing image of the magnetic
field pattern in that area. A procedure is needed to effec-
tively increase the number of data. Program MAGP,
therefore, is basically a linear, two-dimensional interpola-
tion routine to expand the magnetic data file.
Program PICT divides the total range of magnetic field
values in the expanded file into a discrete number of lev-
els, assigns a color to each level, and produces a level-
sliced color picture of the magnetic field values.
EXAMPLES OF OUTPUT
An example of essentially raw radar data is given in
Figure 3. This scan represents a plot of depth (negative
vertical axis) versus track length (horizontal axis), with the
ground surface at the top of the raster. Dimensions of
the vertical profile correspond to a depth of 6 m and a
track length of 30 m. The return from a small spherical
object about 0.2 m in diameter is noted to illustrate the
Figure3.
A Radar Profile Showing the Reflection
from a Sphere (arrow)
-------
GEOHYRDOLOGY MATH MODELING 305
difficulty in interpreting such records for the location of
specific objects.
The results of an actual survey yield a series of vertical
profiles. Each record represents an excursion by the survey
vehicle along parallel tracks at some spacing. Interpreta-
tion of vertical profiles by visual analysis and transfer to a
grid map provide information on large underground struc-
tures, especially continuous forms such as pipes, sewers,
wall foundations, etc.
An example of color-enhanced horizontal slicing illus-
trating the utility of the processing is given in Figure 4.
In this case, slices from 0 to 1.5 m depth have been com-
bined and filtering introduced to enhance the location of
discrete, small targets. The number and distribution of
such objects are apparent.
The area shown covers about 3 acres. Interpretation of
some of the gross artifacts, few of which are apparent
from records of prior site activity or surface features, is
shown.
DIRECTIONS FOR NEW DEVELOPMENT
Several aspects of Ground Penetrating Radar Tech-
nology require further development in order to enhance
broad scale application of the method. Presented below
are some areas of improvement that Battelle is currently
pursuing:
•Transceiver-antenna design; improved waveshape, anten-
na coupling, increased peak power, tunable-frequency
systems
•Antenna arrays and associated electronics to improve
resolution and to provide increased areal coverage cap-
abilities
•Collection, transmission and manipulation of data at
faster rates and with improved reliability
•Review-refresh memory data systems; comparison of old
and new data to detect changes in underground systems
•Portable, handled radar systems for coverage of small,
remote, rough terrain areas with surface obstacles such
as rocks, trees, fences, etc.
•Target recognition through signature analysis, 3-D data
overlay systems, size and shape correlation, deconvolu-
tion of wave forms
•Integration of GPR data with acoustic profiling, sonic
holographic and/or electrical resistivity techniques to im-
prove detection of nonmetallic objects
•Improved, ruggedized transport system, including auto-
matic locator systems to define xy coordinates and attach
angle of the transporter
•Electromagnetic characterization of soils, subsurface
geologic strata, and representative targets
Success in expanding the technology will provide an en-
hanced capability for investigations of Hazardous Waste
sites.
PACKED CLAY
(OLD BASEBALL DIAMOND)
OLD DIRT ROAD BED
STORM SEWER
NUMEROUS
DISCRETE TARGETS
POSSIBLE OLD
TUNNEL STRUCTURE
GAS LINES
OR WATER PIPES
Figure 4.
Computer-Produced Map View of Radar Reflections at
a Survey Site with Filtering to Reduce Broad
Reflection Patterns
-------
THE USE OF MATHEMATICAL MODELS TO ASSESS AND
DESIGN REMEDIAL ACTION FOR CHEMICAL WASTE SITES
C.R. COLE
G.L. McKOWN, Ph.D.
Battelle, Pacific Northwest Laboratory
Richland, Washington
INTRODUCTION
It is estimated that there may be 30,000-50,000 old dis-
posal sites and sites contaminated from spills in the United
States, of which 1,000-2,000 are sufficiently contaminated
to pose a hazard to the public. Restoration costs are likely
to exceed $ 1,000,000 per site.
Faced with the complexity and high costs of address-
ing such a large number of sites, it is clear that formalized
methods are required to expedite the restoration process
and identify the most cost-effective means of site remedial
response. This paper describes a methodology which util-
izes mathematical models describing fluid flow and con-
taminant transport. These constructs have been developed
over a period of nearly two decades and employed to solve
a variety of surface and groundwater contamination prob-
lems. They offer the ability to organize, interpret and bet-
ter understand what is presently occurring, and in ad-
dition, provide the ability to predict what will occur in the
foreseeable future. These capabilities are of particular
value in designing remedial action since, through the use
of these tools, one can:
•determine when a specific site is sufficiently character-
ized and understood such that when implemented the
selected remediation program will be effective
•allow one to design a cost effective site characteriza-
tion and data gathering effort
•rapidly and cost effectively assess a wide variety of re-
medial options without moving any dirt
•design a cost effective site surveillance and monitoring
plan which will maximize the ability to detect any devia-
tion from the remediation plan.
THE SITE RESTORATION PROCESS
While it is the abandoned landfill or superfund problem
set has has stimulated the current interest in site restora-
tion, two major programs in the U.S. have preceded it.
The U.S. Department of Energy and its predecessors have
had an active Decontamination and Decommissioning
(D&D) program for many years now. Similarly, the U.S.
Army Toxic and Hazardous Materials Agency has con-
ducted its Installation Restoration (IR) Program since the
mid-seventies. As a result of these undertakings, a formal-
ized approach to site restoration has emerged. The ap-
proach is divided into three phases : (1) site characteriza-
tion, (2) site assessment and (3) site remediation.
Site Characterization
Work performed during site characterization is directed
to the description of the source(s) of contamination and
the surrounding environment. The objective is to enumer-
ate and quantify to the extent possible the physical, chem-
ical, and biological factors which will affect fate and mi-
gration of contaminants. For maximum efficiency, charac-
terization is performed in an integrated manner with as-
sessment so that only necessary data are collected.
The first element of characterization, the preliminary
survey, is focused on gathering all existing data on a site.
If data are sufficient, no further characterization efforts
are required. If the preliminary assessment reveals data
gaps which must be filled before a final assessment can be
made, then additional characterization efforts are imple-
mented. Hence, the preliminary assessment is employed to
determine the adequacy of existing data and the optimum
design for any subsequent data collection activities.
When active characterization is necessary, it is divided
among a number of field activities. Geophysical survey
techniques can be employed to map deposits. These efforts
are complimented by field geohydrologic investigations
which determine the nature and extent of geologic and hy-
drologic features which will affect the movement of water
and contaminants. Finally, sampling and analysis is per-
formed to detail the disposition of contaminants in the en-
vironment and the physical/chemical properties of geolog-
ic media which would attenuate movement. All charac-
terization activities may not be required at every site. As
noted previously, the type and extent of activities should
be identified during a preliminary assessment.
Post-closure characterization can actually be thought of
as two distinct activities: (1) surveillance of remedial
action implementation and (2) post construction mon-
itoring. The former is directed toward assuring safe per-
formance of work and conduct of that work in the man-
ner prescribed in the final plan. The second activity tries
to confirm the effectiveness of the remedial action on a
continuous basis and, in so doing, will provide a warn-
ing in the event of failure and provide for the collection
of a historical record that can be used to determine the
reason for failure.
Site Assessment
Activities conducted for the assessment are designed to
determine the implications of the available data. In this
306
-------
GEOHYRDOLOGY MATH MODELING 307
context, they are interpretative in nature. The first level
of assessment, preliminary assessment, is conducted us-
ing available data to determine if sufficient information is
on hand to characterize a site and quantify the risk posed
by that site. Two levels of output are possible. The first is
a determination that current data are insufficient. In this
case, the assessment identifies the minimum necessary data
that should be collected to allow conduct of a final assess-
ment. The second possible output is the result of a final
assessment itself.
The use of the mathematical modeling tools needs to be-
gin in this initial assessment phase. Even if sufficient data
for implementation of a mathematical model are not avail-
able at this stage the organization and approach required
to implement a mathematical model can greatly aid the
initial assessment process and improve ones ability to un-
derstand the system. When enough existing data are avail-
able to justify the implementation of a mathematical mod-
el in this initial assessment phase, the quality of the as-
sessment and the understanding of the system improves
even more.
The final assessment is conducted only after sufficient
data and system understanding are in hand. Narrowly
defined, the product of a final assessment is the descrip-
tion of the time frames, durations and concentrations of
contaminants that various segments of the environment
will be exposed to. In this context, the assessment pro-
duct is a prediction of what continuous monitors would
record if the situation were allowed to remain in its cur-
rent state. When coupled with data on the dose-response
function for the contaminants of concern, the assessment
output is converted to an estimate of the risk posed by
the site. Consequently, the assessment product is essential
to several key decision points:
1. Which sites pose unacceptable risks and therefore re-
quire remedial action? ,
2. How extensive must clean-up action be to bring a site
within the realm of acceptable risk?
When it is determined that remediation is required and
conceptual approaches are identified, assessment is again
conducted to predict the effectiveness of proposed alterna-
tives and to identify the most cost effective post-closure
monitoring plan.
Site Remediation
Remediation activities are the major end-product of site
restoration work. They are initiated whenever a final
assessment concludes that risks posed by a site are un-
acceptable. At that time, the preliminary work is begun
to identify all feasible alternatives. After a cursory screen-
ing, the most promising alternatives are described in a
conceptual framework and input to the post-closure assess-
ment where estimates are made of the effectiveness of each
alternative. Output from the post-closure assessment is
compared to output from the final site assessment to de-
termine the level of risk reduction offered by each alterna-
tive. Alternatives not capable of reducing risks to accep-
table levels are discarded. Remaining candidates are
ranked according to cost. The ratio of risk reduction to
cost is then employed to select the most cost-effective
approach.
The selected alternative is then taken into final design
and construction. Post-closure characterization is con-
ducted to assure that construction is performed in accor-
dance with final designs and to detect any releases or ac-
tivities which pose a hazard to the public as a result of
remediation.
The key element of the above approach to site restora-
tion is the assessment activity. When properly conducted,
an integrated assessment continues to focus attention on
the most cost-effective configuration for conduct of work.
In the early stages, assessment is employed to minimize
the extent of expensive field characterization work while
insuring that sufficient data are collected for evaluation.
In the body of the assessment, the objective is to obtain
as accurate a measure of risk posed by a site as possible
so that a determination can be made as to whether a site
requires remediation, and if so, how much. The post-
closure assessment is directed to quickly ascertaining the
likely effectiveness of remedial action alternatives and
their optimum design before resources are committed as
well as to identify a minimum, yet sufficient, configura-
tion for the monitoring network that will provide all the
desired information.
The strength of the assessment phase is directly related
to the technology employed to accomplish it. Develop-
ments over the last decade suggest that, when properly
applied, mathematical models for fluid flow and contam-
inant transport are the best tools currently available for
solving environmental problems. They provide the de-
cision maker with the extra insight needed to determine
the optimum solutions for a given problem.
MATHEMATICAL MODELING—AN OVERVIEW
It is important to briefly discuss in general terms what
a mathematical model is, its value, its limitations and its
range of application in order to illustrate its value as a
tool in the problem assessment and remedial action de-
sign process. Mathematical models, whether analytical
or numerical, can only be a limited representation of an
actual system or "cartoons of reality.(4) Models allow one
to examine an entire system by incorporating expert under-
standing of a complex system with the generally limited
field data in a form which can be tested against field ob-
servations of system response to various driving forces.
Disagreements between model and observation can be used
to guide further field studies or to cause initial interpreta-
tions of the system to be modified in plausible ways. ,
Processes that occur in the actual system are described
by appropriate mathematical equations which correlate the
system's physical and chemical conditions in a proper
manner such that as changes are made to the physical and
chemical parameters describing the natural conditions,
the appropriate response of the real system may be pre-
dicted by solving the mathematical equations.
In order to construct and effectively utilize a model it
is important to:
(1) Understand the physical, biological and chemical pro-
-------
308 GEOHYDROLOGY MATH MODELING
cesses which occur within the system of interest
(2) Identify and determine the important physical, biologi-
cal and chemical parameters and inputs which allow the
solution of the mathematical equations which govern
the physical and chemical processes that occur
(3) Characterize the current condition of the system and
identify its boundaries and boundary conditions
(4) Update model parameters as new system data or a bet-
ter understanding of the system are attained.
Models, in general, are not static; they must be dynamic
unless the system being modeled is very simple. A model
acts as a storehouse for all the pertinent information
assimilated on the system, and or the understanding of the
processes which occur in the system. A model is capable of
learning in the sense that as the data density and quality
improve, those improvements can be incorporated into
the model to improve its predictive capabilities. As the un-
derstanding of processes advances, the model equations or
parameter relationships can be changed to incorporate this
new knowledge and so improve ability of the model to pre-
dict. Model maintenance is an integral part of any effort
involving the use of models.
Mathematical hydrodynamic and transport models,
which can be constructed at various levels of complexity or
realism depending on the use to be made of them, attempt
to duplicate the real system by a series of complex equa-
tions that describe fluid flow and material transport. The
advent of high speed digital computers has made the solu-
Table I.
Characteristics of a Few of the Ground-Water Flow and
Transport Models Being Employed in Safety Assessment
of Nuclear Waste Repositories and for Other Ground-
Water Studies
HYDROLOGIC
PATHSfa)"4)
VTT
FE3DGWI8)
CFEST(b)
Two-dimensional, analytical/numerical method,
homogeneous geology, saturated flow.
Two-dimensional, finite difference numerical
method, heterogeneous geology, saturated flow,
multilayer systems with interaquifer transfer.
Three-dimensional, finite element numerical
method, heterogeneous geology, saturated flow.
Three-dimensional, finite element numerical
method, heterogeneous geology, coupled flow,
energy and solute transport.
CONTAMINANT TRANSPORT
GETOUT*6' One-dimensional, analytical method, chain de-
cay, single speciation, equilibrium sorption, con-
stant leach rate, dispersion.
MMT<'-2" One- and two-dimensional, discrete parcel ran-
dom walk numerical method, chain decay,
single speciation, equilibrium sorption, time
variant leach rate, dispersion, and arbitrary
source distribution.
CFEST Three-dimensional, finite element numerical
method, heterogeneous geology, coupled flow,
energy and solute transport.
(a) All the hydrologic models operate in a transient or steady state mode
and have streamline or pathlme options that can predict travel time.
(b) Code documentation in progress.
tion of these complex sets of equations practicable and
cost effective for:
(1) Developing an understanding of a complex system
(2) Determining sensitivity of the system to uncertainties
in data and understanding
(3) Providing insight on the system and making predic-
tions of system response to various perturbations
(4) Prediction of system response to various proposed
management alternatives
(5) Testing the adequacy of various engineering designs.
The steps involved in modeling are:
(1) Definition of the study objectives
(2) Collection of data on the system
(3) Formulation of a conceptual model of the system
(4) Translation of the conceptual model and data on the
system into the form required for the selected numer-
ical model
(5) Calibration and verification of the numerical model
(6) Check of the sensitivity of the model to variations in
input data
(7) Use as a predictive or investigative tool to attain the
study objectives.
Model output (or predictions) always needs careful eval-
uation because the results are generally only as accurate as
the data and the knowledge of the system upon which the
models are based. Model results should be considered as
a guide to the probable system response. In general, when
properly applied, models are the best tools available for
assistance in understanding and making decisions on com-
plex systems. Modeling alone, however, is no substitute for
routine monitoring in problem areas. A successful man-
agement program must include both modeling and con-
tinued monitoring.
Description of a Few of the
Many Available Mathematical Models
The effort to site nuclear waste repositories under the
sponsorship of the U.S. Department of Energy has been
responsible for the identification and development of a
suite of ground-water hydrologic and transport models.
The modeling technology involves: (1) hydrologic codes to
define the ground-water flow field and indicate water flow
paths and travel times and (2) transport codes to describe
the movement and concentrations of the contaminants in
the flow field and the rates of arrival'of contaminants at
points of biosphere uptake.
The groundwater hydrologic and transport codes iden-
tified and developed include several levels of complexity.
Code selection and use are determined by the study pur-
pose, the quantity and quality of input data and the pro-
cesses felt to be driving the system. There are hydrologic
codes at four levels and transport codes at three levels as
shown in Table I.
In addition to the saturated flow models there are
various levels of hydrologic codes for modeling unsatur-
ated hydrologic systems. These codes for unsaturated flow
modeling are required when the contaminant source is
located in the unsaturated hydrologic system. The codtt
which Battelle has had experience applying include:
-------
GEOHYRDOLOGY MATH MODELING 309
•UNSAT(9)—one dimensional, finite difference numerical
method, unsaturated flow, with crop water extraction.
•TRUST*00'1"—one or two dimensional, integrated finite
difference numerical method, unsaturated flow, deform-
able porous medium.
Numerous mathematical models have been developed
and applied to a variety of surface water contamination
problems. These models include a variety of surface water
hydrodynamic and transport models which have been util-
ized to understand these complex systems as well as predict
future surface water quality. These models include hydro-
dynamic, contaminant transport and sediment transport
models for river, lake and estuarine systems. A few of
these models which Battelle has had experience in apply-
ing are given in Table II.
APPLICATIONS OF MATHEMATICAL MODELS
Surface water and groundwater hydrodynamic and con-
taminant transport models are directly applicable to prob-
lems related to siting new facilities, assessing existing con-
taminated ones as well as evaluating remediation options
when required. The level of sophistication required for
any effort is dictated by data availability, major phenom-
ena of importance (i.e., saturated or unsaturated flow and
the level of resolution desired. The usefulness of models as
an aid in these assessment and remediation processes is
best illustrated by examples of past work with similar pur-
pose. Three such examples are given below.
Table II.
Characteristics of a Few of the Available Surface Water
Hydrodynamic and Contaminant Transport Codes
HYDROLOGIC
EXPLORED
DKWAV<2>
DWOPERO
Two-dimensional (x-y), dynamic wave, finite
difference code for modeling tidal or river flow in
estuary or river systems.
One-dimensional, kinematic wave, finite differ-
ence code for modeling river flow and account-
ing for rainfall and runoff contributions.
One-dimensional, dynamic wave, finite differ-
ence code for modeling river flow and account-
ing for rainfall and runoff contributions.
CONTAMINANT TRANSPORT
EXPLORE Same as above but also simulates contaminant
transport for dissolved contaminants.
DKWAV Same as above but also handles contaminated
sediment transport both overland and in rivers.
SERATRAW Two-dimensional (x-z), transient, finite ele-
ment code for sediment and contaminant trans-
port modeling in rivers, lakes and estuaries.
FETRAP) Two-dimensional (x-y), transient, finite element
code for sediment and contaminant transport
modeling in rivers, lakes and estuaries.
TODAMO One-dimensional, transient, finite element code
for sediment and contaminant transport model-
ing in rivers, lakes and estuaries.
'The capability of TRUST has been expanded to allow stream lines
and pathlines to be predicted from TRUST input and output streams.
Kepone in the James River
As a part of the Kepone mitigation feasibility study,(5) a
mathematical simulation of sediment and Kepone (a highly
chlorinated pesticide) transport in the James River estuary
was conducted by applying the sediment and contaminant
transport model, FETRA, to an 85-km river reach between
Bailey and Burwell Bays. The FETRA code consists of
three submodels coupled together to take into account sed-
iment-contaminant interaction. The submodels are:
1) sediment transport model, 2) dissolved contaminant
transport model, and 3) particulate contaminant (con-
taminants adsorbed by sediment) transport model. Trans-
port of sediment and particulate contaminants is simulated
for each sediment type or size fraction. The modeling pro-
cedure of the FETRA code involves simulation of sediment
transport. The results are then used to simulate dissolved
and particulate contaminants by accounting for interac-
tion with the sediments. Changes in river bed conditions
are calculated including: 1) river bed elevation change,
2) distributions of ratio of each bed sediment component
in the bed and bed armoring, and 3) distribution of con-
taminants within the river bed.
The FETRA code was applied to simulate the migra-
tion of sediment and Kepone for three river discharges.
Tidally influenced depth and velocity distributions in the
study area were obtained by the unsteady, one-dimensional
code, EXPLORE. These results were used by the two-
dimensional code, FETRA, to obtain longitudinal dis-
tributions of sediment and Kepone. Hence the results dis-
cussed here are cross-sectionally averaged values chang-
ing with tidal flow. Comparison of computed results and
field data for both sediment and Kepone concentration
indicates very good agreement, confirming the validity of
the model.
Mathematical simulation of Kepone transport under
most probable flow conditions yields an estimate of 89.1
kg/yr of Kepone transported seaward from Burwell Bay.
Of this 89.1 kg of total Kepone, 25.3% is carried by sed-
iment, while 74.7% is in a dissolved phase. Since it is esti-
mated that approximately 9600 kg of Kepone is present
in the top 0.3m of bottom sediment in the river, with this
flushing rate, it may take at least 108 yr for natural water-
sediment flushing mechanisms alone to clean the James
River.
The effectiveness of ten partial Kepone cleanup activ-
ities to reduce Kepone concentrations in the river was
studied by simulating ten hypothetical conditions by the
FETRA code. Through this cost effective evaluation pro-
cess, it was quickly determined that none of the partial
mitigation options would yield significant reductions in
total flushing times. Hence, ineffective expenditures were
avoided.
Assessment of Seepage From
Buried Uranium Mill Tailings
The advantages of returning uranium mill tailings to the
pits excavated during surface mining operations are being
more widely recognized and utilized to minimize any un-
-------
310 GEOHYRDOLOGY MATH MODELING
desirable environmental effects.03' Particularly notable
among the benefits is the better control of radon gas emis-
sion through appropriate mill tailing burial. This advan-
tage alone warrants burial of the tailings.
As plans are made for the burial of the tailings in pits,
(as opposed to using conventional tailings ponds above the
ground surface), concern arises about the potential for
contaminant seepage from the buried tailings. Disposal in
the mined out pits places the tailings nearer ground water
and may increase the potential for ground water contam-
ination.
The purpose of this study was to examine potential for
ground water contamination by seepage from buried tail-
ings under four alternatives of clay liners and tailings
placement. To accomplish this comparison of alternatives,
laboratory work and numerous measurements were made
on materials typical of the site. These measurements pro-
vided the data on soil characteristics necessary for input
to the hydrologic flow and transport models which were
used in this assessment study.
The four alternatives considered to minimize seepage of
contaminants from buried mine tailings included:
1. Placement of saturated tailings in a covered pit with a
clay liner in the bottom, but no side liners
2. Placement of dewatered tailings in a covered pit with
clay bottom liner, but no side liners
3. Placement of saturated tailings in a pit having a clay
bottom liner, no side liners, and drains to facilitate
pumping of drainage solution from sumps placed above
the bottom clay liner
4. Placement of saturated tailings in a pit with both bottom
and side clay liners.
The assessment involved combined subsurface fluid flow
and contaminant transport modeling of the four alterna-
tives for controlling seepage from buried tailings. The
assessment included combined saturated and partially sat-
urated flow and contaminant transport models for two-di-
mensional vertical, cross sections typical of the tailings
burial pits proposed for use at the site.
The results obtained from the models were the contam-
inant flow paths away from the tailings pits, the advancing
contaminant flow fronts for various sorbed and non-
sorbed constituents of major environmental concern, and
the associated quantities of contaminant flow for each of
the alternatives. These results enabled the comparison of
the environmental consequences of the four alternatives.
It was also possible to gain considerable insight about com-
bining the beneficial alternatives to obtain the best overall
method for controlling seepage from the buried pits.
Although four alternatives were considered in this study,
only the first three alternatives, which utilize a bottom clay
liner but no side liner, provided important results. All of
the results obtained in this study demonstrated that a side
liner on the upper 70 to 80% of the typical 30.5m
side wall height is of very little or no benefit in reducing
contaminant seepage; hence, high side wall liners are un-
necessary and use of such would moreover be economically
wasteful.
The study demonstrated the very real benefits of reduc-
ing the volume of fluid available to seep from the tailings.
The similarity fo contamination of consequences for sys-
tems that use dewatered tailings or underdrains tends to in-
dicate that the means of reducing the volume available for
seepage is less important than the actual volume reduction
of water that can be realized.
Based upon what was learned from this study, it was
felt that the best contaminant seepage control would be
provided through combining the desirable features of the
four alternatives considered. The recommendations would
include the accepted pit covering procedure already
planned for the site; specifically, the additional proposed
control alternative was:
1. To provide a well constructed bottom clay liner placed
at least 3.0 m above the regional water table and not
less than 0.9 m thick
2. To provide stub clay side liners continuing part way up
the pit side wall that form a continuous saucer-shaped
bottom and side liner for the pit
3. To install a network of gravity drains and pumping
sumps with the drains in the tailings sufficiently above
the bottom clay liner to provide very effective drainage
of the tailings
4. To pump tailings drainage effluent from sumps while
the pit is being filled with tailings and on an as needed
basis for the first 6 to 8 months after the filling has been
completed.
Additional studies were proposed to determine the ap-
plicability of these side liner recommendations at other
sites and to assure acceptable minimum contaminant seep-
age under other conditions. Specifically, further study is
needed for various soil materials and other pit configura-
tions using the combined approach of lining and reduc-
ing tailings water. Such studies should also be followed by
companion field observation as the recommended man-
agement practices are put into use at the Wyoming site or
elsewhere to check, verify, and improve upon the recom-
mended disposal control practices. In this way, it will be
possible to economically further reduce contaminant seep-
age and better assure the adequate disposal of uranium
tailings in burial pits.
Initial Assessment of Fixed Flue
Gas Desulf urization (FDG) Sludge
as a Landfill
The first full-scale application of IU Conversion Sys-
tems, Inc. (IUCS) fixation process (Poz-O-Tec* process)
for treating flue gas desulfurization (FGD) sludge is in
operation at the Conesville Power Station of the Colum-
bus and Southern Ohio Electric Company.'3' The Poz-O-
Tec® process blends fly ash, bottom ash (if desired), lime,
scrubber sludge and other additives. The Poz-O-Te^
sludge is landfilled in a surface disposal area where it
hardens to a material with relatively low permeability.
In this two phase project the movement of contam-
inants in the groundwater is being studied in order to pro-
vide a basis on which utilities can make decisions regard-
ing the Poz-O-Tec* method, in particular and the db-
• Registered Trade Mark.
-------
GEOHYRDOLOGY MATH MODELING 311
posal of FGD sludge, in general. The Phase I efforts dealt
with initial site characterization and preliminary assess-
ment, whereas Phase II will include extensive field mon-
itoring, model application, and a final assessment.
In Phase I, which will be discussed here, a model data
base was formulated and laboratory studies were con-
ducted to determine saturated and partially-saturated per-
meabilities of sludge and ash materials, and to describe the
composition and distribution of the sludge and ash leach-
ate. The data base and laboratory results were used to im-
plement, calibrate, and verify a two-dimensional, finite
difference, hydrologic flow model for the study area. The
model output was coupled with a technique of graphing
contaminant arrival distribution to predict the distribution
in time and space of contaminants traveling with the
groundwater within the study area.
The flow model was calibrated against a potential sur-
face representative of a high Muskingum River stage. As
a check on the accuracy of the model prediction, the
model was verified against a low river stage potential sur-
face. In both cases, the model prediction was in good
agreement with the field-measured potential surface.
A sensitivity analysis was conducted to determine the
sensitivity of the model predictions to changes in trans-
missivity (T) and recharge rate (Q). The findings indicated
that the model is fairly sensitive to changes in these
parameters; therefore, the T and Q distributions arrived
at in the calibration process cannot be changed apprec-
iably without significantly altering the model prediction.
It should be noted, however, that once calibrated and
verified, the final modified transmissivity distribution was
within plus or minus 10% of the initial transmissivity
distribution.
Hypothetical examples were used to demonstrate the
methodology for studying arrival distributions of contam-
inants at the Muskingum River. Attention was focused
on contaminant migration from two surface water ponds
(Pbz-0-Tec® and emergency sludge pond) within the Poz-
0-Tec® disposal area and not on seepage to the ground-
water over the entire disposal area. This emphasis was
chosen because:
•Laboratory studies and field investigation have shown
the Poz-O-Tec® material to be highly impermeable
(10s to 107 cm/sec), implying little, if any, water per-
colates through this material
•Man-made and natural channelization has been devel-
oped in the disposal area which encourages immediate
runoff of precipitation into the Poz-O-Tec® pond (i.e.,
water does not remain ponded on the Poz-O-Tec®
material)
•The Poz-O-Tec® pond is hydraulically connected to the
water table, thereby providing a direct path for Poz-
O-Tec® runoff to move into the groundwater
•Leakage from the emergency sludge pond (used to inter-
mittently store thickened unfixed FGD sludge) to the
groundwater could easily have occurred between January
1977 when the Poz-O-Tec® operation began and Sep-
tember 1978, at which time the pond was lined.
The results of the Phase I modeling, characteriza-
tion and data gathering efforts were used to provide in-
put to the monitoring system design of Phase II. At the
conclusion of these Phase I efforts it was not possible to
thoroughly assess the groundwater contamination result-
ing from the Poz-O-Tec® process for three reasons:
•Chemical analysis of groundwater obtained from below
the Poz-O-Tec® and ash pond found sulfate to be the
only water quality parameter above drinking water
standards (secondary drinking water standard)
•The source of the sulfate (Poz-O-Tec® , or fly ash, or
both) has not been identified
•If it can be assumed that contaminants enter the ground-
water in pulses in direct response to rainfall, arrival dis-
tribution techniques have shown that their concentra-
tions will be greatly reduced in the process of reaching
the Muskingum River.
Additional monitoring and water quality analysis of the
Phase II efforts will be necessary before the degree of
groundwater protection provided by Poz-O-Tec® can be
determined.
SUMMATION
The authors have briefly outlined the need and role of
mathematical models in the site characterization, site
assessment and remedial action design phases of a site
restoration process for a hazardous chemical waste dis-
posal site. In order to dispell some of the misunderstand-
ing and distrust regarding mathematical models and their
useage, a brief overview was presented along with sum-
mary descriptions of typical model application efforts.
Some of the basic concepts of mathematical modeling
and how the various kinds of currently available codes can
be effectively applied to problems of a type similar to
those of site restoration at a chemical waste site have
been shown.
The key points regarding mathematical models are:
(1) Mathematical models are simply tools and like many
other tools come in a variety of forms from simple to
complex and must be utilized by craftsmen skilled in
their appropriate use
(2) Models provide a systematic approach for organiza-
tion and interpretation of data on complex systems as
well as a means for achieving a better understanding
of these systems and lastly provide a means for com-
paring alternatives or making predictions
(3) Models can not be used in lieu of data for assess-
ment or as a substitute for monitoring
Mathematical models need to be utilized throughout
the site restoration process from the initial characteriza-
tion phase and continuing through the remedial action de-
sign, construction and final surveillance monitoring
phases. The capabilities of mathematical models can be
utilized in the site restoration process to:
(1) Help determine when a site is sufficiently character-
ized and understood for remediation to be imple-
mented when it is required
(2) Aid in the design of cost and time effective data gather-
ing efforts associated with site characterization
(3) Aid in the rapid and cost effective assessment of:
•whether remediation is required
-------
312 GEOHYRDOLOGY MATH MODELING
•a wide variety of remediation options when it is re-
quired
(4) Aid in the design of a cost effective site surveillance
and monitoring plan.
REFERENCES
1. Ahlstrom, S.W., Foote, H.P., Arnett, R.C., Cole,
C.R. and Serne, R.J. "Multicomponent Mass Trans-
port Model: Theory and Numerical Implementation
(Discrete-Parcel-Random-Walk Version).'' BNWL-
2127, Pacific Northwest Laboratory, Richland, Wash.,
1977.
2. Baca, R.G., Waddell, W.W., Cole, C.R., Brand-
stetter, A. and Cearlock, D.B. "EXPLORE-I: A
River Basin Water Quality Model." Prepared for
U.S. Environmental Proltection Agency by Battelle,
Pacific Northwest Laboratories, Richland, Wash.,
1973.
3. Bond, F.W. and Nelson, R.W. "Modeling the Fixed
FGD Sludge Landfill-Conesville, Ohio (Phase I)."
EPRI-CS-1355. Prepared for Electric Power Research
Institute by Battelle, Pacific Northwest Laboratories,
Richland, Wash., 1980.
4. Chow, V.T. "An Introduction to Systems Analysis
of Hydrological Problems." Proc. of the Second In-
ternational Seminar for Hydrology Professors. Logan,
Utah, 1970.
5. Dawson, G.W. "The Feasibility of Mitigating Kepone
Contamination in the James River Basin." EPA-
440/5-78/004A. Prepared for U.S. Environmental
Protection Agency by Battelle, Pacific Northwest
Laboratories, Richland, Wash., 1978.
6. DeMier, W.V., Cloninger, M.O., Burkholder, H.C.
and Liddell, P.J. "GETOUT-A Computer Program
for Predicting Radionuclide Decay Chain Transport
Through Geologic Media." PNL-2970, Pacific North-
west Laboratories, Richland, Wash., 1979.
7. Fread, D.L. "National Weather Service Operational
Dynamic Wave Model." Hydrologic Research Labora-
tory, National Weather Service, NOAA, Silver Spring,
Md., 1978.
8. Gupta, S.K., Cole, C.R. and Bond, F.W. "AEGIS.
Finite-Element Three-Dimensional Ground-Water
(FE3DGW) Flow Model—Formulation Program List-
ings and User's Manual." PNL-2939, Pacific North-
west Laboratories, Richland, Wash., 1979.
9. Gupta, S.K., Tanji, K.K., Nielsen, D.R., Biggar,
J.W., Simmons, C.S. and Maclntyre, J.L. "Field Sim-
ulation of Soil-Water Movement with Crop Water
Extraction." University of California, Department of
Land, Air and Water Resources, Water Science and
Engineering Paper No. 4013, 1978.
10. Narasimhan, T.N. and Witherspoon, P.A. "Numeri-
cal Model for Saturated-Unsaturated Flow in Deform-
able Porous Media 1. Theory." Water Res. Res. 13,
1977,657.
11. Narasimhan, T.N. and Witherspoon, P.A. "Numer-
ical Model for Saturated-Unsaturated Flow in De-
formable Porous Media 3. Applications." Water Res.
Res. 14, 1978, 1017.
12. Narasimhan, T.N., Witherspoon, P.A. and Edwards,
A.L. "Numerical Model for Saturated-Unsaturated
Flow in Deformable Porous Media 2. The Algorithm."
Water Res. Res. 14, 1978, 255.
13. Nelson, R.W., Reisenauer, A.E. and Gee, G.W.
"Model Assessment of Alternatives for Reducing
Seepage from Buried Uranium Mill Tailings at the
Morton Ranch Site in Central Wyoming." PNL-
3378, Pacific Northwest Laboratories, Richland,
Wash., 1980.
14. Nelson, R.W. and Schur, J.A. "AEGIS. PATHS
Groundwater Hydrologic Model." PNL-3162, Pacific
Northwest Laboratories, Richland, Wash., 1980.
15. Onishi, Y., Serne, R.J., Arnold, E.M., Cowan, C.E.
and Thompson, F.L. "Critical Review: Radionuclide
Transport, Sediment Transport, and Water Quality
Mathematical Modeling; and Radionuclide Adsorp-
tion/Desorption Mechanisms." PNL-2901, Pacific
Northwest Laboratories, Richland, Wash. 1980.
16. Onishi, Y. and Wise, S.E. "Mathematical Modeling
of Sediment and Contaminant Transport in the James
River Estuary." Proc. of the Specialty Conference on
Verification of Mathematical and Physical Models in
Hydraulic Engineering, ASCE, College Park, Mary-
land, August 9-11, 1980, 303-310.
17. Onishi, Y. and Wise, S.E. "Finite Element Model for
Sediment and Toxic Contaminant Transport in
Streams." Proc. of the Specialty Conference on Con-
servation and Utilization of Water and Energy Re-
sources, ASCE, San Francisco, California, August
8-11, 1979, 144-150.
18. Reisenauer, A.E. "AEGIS. Variable Thickness Tran-
sient Ground-Water Flow Model. Volume 1. Formula-
tion." PNL-3160-1, Pacific Northwest Laboratories,
Richland, Wash., 1979.
19. Reisenauer, A.E. "AEGIS. Variable Thickness Tran-
sient Ground-Water Flow Model. Volume 2. User's
Manual." PNL-3160-2, Pacific Northwest Labora-
tories, Richland, Wash., 1979.
20. Reisenauer, A.E. 1979. "AEGIS. Variable Thick-
ness Transient Ground-Water Flow Model. Volume
3. Program Listings." PNL-3160-3, Pacific Northwest
Laboratories, Richland, Wash., 1979.
21. Washburn, J.F., Kaszeta, F.E., Simmons, C.S. and
Cole, C.R. "AEGIS. Multicomponent Mass Trans-
port Model: A Model for Simulating Migration of
Radionuclides in Ground Water." PNL-3179, Pacific
Northwest Laboratories, Richland, Wash., 1980.
22. Whelan, G. "Distributed Model for Sediment Yield."
Master's Thesis Mechanics and Hydraulics, University
of Iowa, Iowa City, Iowa, 1978.
-------
PARAMETRIC ANALYSIS OF GEOLOGICAL
HAZARDOUS WASTE DISPOSAL
EDUARDO A. FIGUEROA
FRANK L. PARKER, PH.D.
Department of Civil and Environmental Engineering
Vanderbilt University
Nashville, Tennessee
INTRODUCTION
The uncontrolled discharge of hazardous material has
steadily increased not only because of increases in pop-
ulation, agriculture and industry, but also because law
and regulations, such as the Clean Air Act (as amended)
and the Marine Protection, Research and Sanctuary Act
(as amended) curtail the discharge of hazardous pollu-
tants in the air and water, leaving the alternative of land
disposal as the last or final disposal practice. And, as can
be seen from past experiences, the best available technol-
ogy has not always resulted in secure landfills. There is
growing evidence of the vulnerability of clay and plastic
liners to the effects of certain chemicals, there is also the
difficulty of installation and susceptibility of plastic liners
to puncture. Moreover there are little data to show that
the lifetime of the plastic liners, even if they stay intact,
will exceed the lifetime of the hazardous wastes they con-
tain.
Consequently, the authors have done a parametric an-
alysis to see how the waste disposal repository acts as a
total system and to determine what features need to be
improved to guarantee the safety of the disposal facility.
To study the problem, one has to look at the system as
a whole, where the various parameters interact, and allow-
ing variation in values of the parameters according to the
uncertainty of these values. To do this, the authors
adopted a simple conceptual model in which the repository
is located close to the ground surface (less than 10 meters),
the waste is packaged, is leached by groundwater, is sorb-
ed during its travel underground and finally reaches the
biosphere. Dilution in the biosphere has not been consid-
ered. Concentrations in the biosphere have been com-
pared to recommended limits.
In this initial screening of geological hazardous waste
disposal sites, one can afford to use simpler models, so
that with a minimum of information they can be satisfac-
torily used to predict the behavior of contaminants under
normal or extreme conditions. A simple one-dimensional
analytical solution to the diffusion-dispersion equation in
porous media was used to show that, in a multiparameter
system characterized by the packaging, the waste disposal
technique, the geology and hydrology of the site, there are
a number of combinations (multidimension response sur-
face) that will result in concentrations released to the bio-
sphere that are less than permissible values, but more im-
portant that there are many parameter combinations that
will not meet release criteria.
ANALYTICAL MODEL
The one-dimensional transport equation for horizon-
tal, single-phase flow in saturated, unconsolidated and
homogeneous media can be expressed as:
9c 9c _. 92c p9s.. n
— = - u—+ D r-7--r-+ kcn
9t 9x 9xz e9t
(1)
Ogata and Banks2 presented a general solution of the form:
l+i
— = tt[erfc[ !
L2(0/Pe)1/2
erfc
(2)
where
6 = (u*t)/X
Pe= (u*X)/D
Sorption is implicitly considered as "u" was used as the
velocity of the pollutant, under the assumption that the
mass transfer of solute to sorption sites is rapid and that
the sorption equilibrium is linear, ds/dc = Kd, where Kd
is the distribution coefficient and the term [1 + (Kd/e )]
is known as the retardation factor and gives the relative
velocity of movement of a sorbing solute to the movement
of water.
Input Data
The study considered the solubility limit as the initial
concentration of the pollutant in the aquifer and the longi-
tudinal dispersion coefficient with a fixed value of 0.9 m/
day. The parameters that were varied were: groundwater
velocity 0.1, 0.01, 0.001, and 0.0001 m/day, partitioning
coefficients of 1, 10, 100, 1000, and 10,000 1/g and dis-
tance to the biosphere of 100, 1000, and 10,000 meters.
The data are summarized in Table I. Biological or chem-
ical degradation were not considered, since the pollutants
under study are "conservative" substances.
313
-------
314 GEOHYDROLOGY MATH MODELING
Table I.
Chemical Characteristics of Pollutants
Partition
coefficient
d/g)
10M04
100-250
1-100
Permissible
ambient cone.
Solubility
limit
(mg/D
0.001
0.004
0.005
0.012
0.26
1.75
Pollutant
AROCLOR 1254
ENDRIN
TOXAPHENE
RESULTS
Three different pollutants were studies: PCB, toxa-
phene, and Endrin. Each has different sorption charac-
teristics, different solubility limits and different permiss-
ible water concentrations.
Aroclor 1254 (PCB), has high partitioning coefficients
from 10* to 1061/g, low solubility limit of 0.012 mg/1 and
permissible ambient concentration of 0.001 ug/1. Endrin
has partitioning coefficients of 100 to 250 1/g, solubility
limit of 0.26 mg/1, and permissible ambient concentra-
tions of 0.004 ug/1. Toxaphene has partitioning coeffic-
ients of 1 to 1001/g, solubility limit of 1.75 mg/1, and per-
missible ambient concentration of 0.005 ug/1. The ambient
concentration values were those recommended by EPA in
Quality Criteria for Water3 and the solubility values are
those in the treatability manual.(4) Partitioning coefficients
have been obtained from different publications and from a
Vanderbilt University study.<5>
The results of the analyses are presented in Table II and
in normalized form in Figures 1, 2, and 3, so the analysis
will be valid for any pollutant if we know the initial con-
centration, the release criteria concentration, and the par-
titioning coefficient.
At 1000 meters from the source, which is considered as a
realistic distance in humid climates for the pollutant to be
released to the biosphere in the case of near surface geo-
logical repositories, Aroclor 1254 will reach the release cri-
teria limit with groundwater velocities of 0.1, 0.01, 0.001,
and 0.0001 m/day, at 6.25 x 103, 23.0 x 103, 35.0 x 105
and 37.0 x 103 days, respectively. For the same ground-
water flow rates, Endrin will reach the biosphere in 8.7
x 103, 22 x 103, 29 x 103, and 31 x 103 days, and toxaphene
at 22.5 x 103,24 x 103,25 x 103, and 25.5 x 103 days.
As one can see, pollutants can be expected to appear in
concentrations higher than recommended values in about
10
-4
o
O
O
-5
10
-6
10
-7
10
ax,
IIM
ay
Kd 100
Velocity
(m/d)
a 0.
b 0.
c 0.
d 0.
1
01
001
0001
Dist. from
source
Tm)
x
y
z
100
1000
10000
[dx
bx'
'
c
10*
10J
10"
105 TIME (d) lo6
Figure 1.
Time to Reach Ambient Criteria Concentrations
-------
GEOHYDROLOGY MATH MODELING 315
Table II.
Time to Reach Ambient Criteria Concentration (d)
Velocity (m/d)
Distance (m)
AROCLOR 1254
ENDRIN
TOXAPHENE
0.1
100
240
200
230
1000
6250
8700
22500
10000
88000
110000
1650000
0.01
100
350
225
250
1000
23000
22000
24000
10000
600000
740000
2400000
Velocity(m/d)
Distance (d)
AROCLOR 1254
ENDRIN
TOXAPHENE
0.001
100
370
240
255
1000
35000
29000
25000
10000
2300000
2 100000
2600000
0.0001
100
390
245
260
1000
37000
31000
25500
10000
3500000
2900000
2700000
20 years, with high groundwater velocities and in 100
years, with low velocities. These times are not sufficiently
long to prevent contamination of the biosphere from these
contaminants.
Even if one considers the best possible conditions, high
retardation factors, low groundwater velocities and high
ratios of release concentration over solubility (10-4), the
time involved is only 120 years. If one considers low re-
tardation factors, high groundwater velocities and mod-
erately low ratios of release concentration to solubility
(10-7), at 1000 meters from the source, the release criter-
ion will be reached in a decade.
CONCLUSIONS
This simple model shows how recommended maximum
release concentrations to the biosphere can be reached in
short periods of time if the leachate is permitted to reach
solubility limit values in the aquifers underlying hazardous
waste sites. Natural existing conditions will not secure the
pollutants for extensive periods of time. Only in very lim-
ited situations, with low concentrations of hazardous ma-
terials located in extremely favorable hydrogeological con-
ditions, (low precipitation and thick clay formations), will
the pollutants be secure in near surface geological repos-
itories. If these calculations are correct,.for near surface
storage one can assume that one can only substantially de-
crease the groundwater velocity with great effort and the
distance to the biosphere with equally great effort.
-4
10'
o
0
0
r5
n-7
ax
(c,d)x
ay
Kd 1
a
Velocity
(m/d)
a 0
b 0
c 0
d 0
.1
.01
.001
.0001
Dist. from
source
(m)
X
y
z
100
1000
10000
b
10H
10J
time (d)
Figure 2.
Time to Reach Ambient Criteria Concentrations
-------
316
-4
GEOHYDROLOGY MATH MODELING
I I I I
10
o
O
O
-5
10
10
-6
| I I I M I j
axil (b,c,d)x
| I I I I I I
I I I | I I I I
ay ,(b,c,d)y
TTT
aa bz
Kd 10000
Velocity
(m/d)
Dist. from
source
(m)
a 0.1 x 100
b 0.01 y 3000
c 0.001 z 10000
d 0.0001
(c,d)z
10J
10
105 *:^.^ /w\ 106
Figure3.
Time to Reach Ambient Criteria Concentrations
time (d)
Consequently, the major parameters that are suscep-
tible to revision are the solubility limit and the partition
coefficient. To decrease the solubility limit requires an
alternation in the form of the waste. This, of course,
would differ for specific waste types. One may conclude
that, for certain wastes, only total destruction, by incin-
eration, for example, is acceptable. Whereas for the
others, it is possible that encapsulation methods may be
sufficient. It may also be possible to change by altering
the natural chemistry of the site. However, this can only
be a temporary solution.
Therefore, it may be necessary to examine alternatives
to near surface disposal. Disposal at intermediate depths
(>50 m) would in general, decrease the groundwater flow
and increase the distanceto the biosphere.
NOMENCLATURE
co —initial concentration mg 1 -'
c —final concentration mg 1 -'
D —dispersion coefficient m'd -'
£ —soil void fraction %
k —reaction constant
n —order of reaction
s —mass of solute sorbed per unit dry mass of solid
t —time
u —average velocity m d -'
x —distance in flow direction m
p —bulk density of soil g m -3
REFERENCES
1. Council of Environmental Quality, "Contamination of
Groundwater by Toxic Organic Chemicals," U.S. Gov-
ernment Printing Office, Washington, D.C. 20402,
January, 1981.
2. Ogata, A., Banks, R.B., "A Solution of the Differ-
ential Equation of Longitudinal Dispersion in Porous
Media," U.S. Geological Survey Prof. Paper 411-A,
Washington, D.C., 1961.
3. U.S. Environmental Protection Agency, "Quality Cri-
teria for Water," U.S. EPA, Washington, D.C. 20460,
1976.
4. U.S. Environmental Protection Agency, "Treatability
Manual, Volume I. Treatability Data," EPA-600/8-
80-042a, July, 1980.
5. Jaffe, P.R., Parker, F.L., Wilson, D.J., "The Fate of
Toxic Substances in Rivers," Proc. of the ASCE
National Conference of Environmental Engineering,
Atlanta, Georgia, July, 1981.
-------
LOCATING OF GROUNDWATER POLLUTION
SOURCES FROM LIMITED FIELD DATA
JACK C. HWANG, Ph.D.
ROBERT M. KOERNER, Ph.D.
Department of Civil Engineering
Drexel University
Philadelphia, Pennsylvania
INTRODUCTION
Conventional methods of locating an unknown ground-
water pollution source require an extensive number of
observation wells from which sufficient data can hope-
fully be obtained to plot concentration contours for the
identification of the pollution source.(1) Such an operation
usually is quite expensive and furthermore, with limited
field data, results are approximate, at best. However, by
assuming constant dispersivity of the contaminant and the
steady state groundwater flow field, the mass transport
equation for the contaminant becomes linear. Thus, to
locate the pollution source is now equivalent to solving
an identification problem of a linear dynamic system.
The aquifer parameter identification problems or so
called inverse problems, have been investigated for the past
10-15 years. Faust and Mercer(2) gave a brief account of
its recent development. In general, the calculation pro-
cedure for groundwater flow consists of finding a parame-
ter set (transmissivity, storage coefficient, sources, etc.)
that minimizes deviations between observed and calcu-
lated values of hydraulic head. Thus, for steady state flow
problems, the criterion adopted to define the optimal set
of parameters is to minimize:
m
(1)
where Hj and hj are the observed and calculated heads at
node i for a total of m observations. For the unsteady state
flow problems the criterion function is the minimization
of:
(2)
where hjjj.j and h,,,^ are the calculated and the observed
heads made from O to T at mth discretized point cor-
responding to time tj, where j = 1,2, ...T.
Several methods are available in order to achieve the
minimization. Cooley(3) derived the modified Gauss-New-
ton procedures after linearizing the steady state equation
using a similar technique to that used by Yeh and Tauxe.(4)
Frind and Pinder(4) solved the inverse problem for trans-
missivity by the Galerkin finite element approach using
steady state hydraulic heads as input data. For the dynam-
ic system, Yeh(6) presents a good review on the methods of
aquifer parameter estimation for unsteady state flow in an
unconfined aquifer. Methods such as quasi-linearization,
maximum principle and gradient method, influence coef-
ficient method and the minimax-linear programing ap-
proach are presented in detail for achieving the minimi-
zation.
The application of the system's approach, however, for
the identification of a groundwater pollution source simi-
lar to those for aquifer parameter identification seems to
have been overlooked by previous researchers. In fact, if
a steady-state flow field is assumed, the system control
equation is linear. The criterion is then the minimization
of the following expression:
m
Jt = S(yi>t-yin)2
i = 1
(3)
in which y ,t and y-t t are the calculated and observed heads
of concentration made at ith well corresponding to time t.
The matrix equation resulting from the finite element dis-
cretization for the mass transport equation can be written:
c = Bc + g
(4)
where the location and the strength of the pollution source
are contained in the vector g. The estimate, g, approaches
the true value of g as Jt approaches a minimum value. The
method adopted in this study for achieving the minimi-
zation is based on system sensitivity theory which will be
described in the following sections. As written, the paper
presents the theoretical foundations for the study which
will be presented at a subsequent time.
ANALYTIC FORMULATION
The governing equation of two-dimensional mass
transport of a contaminant in groundwater flow can be
written as follows:
3c_
9t
T^+ v -^
3x 3y
(5)
317
-------
318 GEOHYRDOLOGY MATH MODELING
where
c =
u,v =
Dx,Dy =
q
concentration of the contaminant
velocity components of groundwater
components of dispersivity
source term
The source term q can be formulated according to the
physical behavior of the pollution sources. Adey and Con-
nor(7) outlined some expressions of the source term includ-
ing sources with time decay, rate, etc. In order not to ob-
scure the trend of thought in a jungle of equations, a sim-
plified, but typical, formulation will be described.
The pollution source considered here is of a point source
(or sources) with constant strength. Thus, the source term
is as follows:
q = Eqit?(x-x*i)(y-y*i) (6)
where
t? = Dirac Delta function
qi = strength of the ith source
(x*j,y*i) = coordinates of the ith source
The major axes of the dispersivity tensor coincide with
the global axes of the aquifer and, furthermore, the
velocity field of the groundwater is steady but may vary
spatially.
With proper initial and boundary conditions pre-
scribed, the governing equation can be solved numerically
to determine the concentration distribution of the pollut-
ant over a period of time. Using the finite element method,
it can be shown"" that the Galerkin's expression for an
aquifer element of depth b and fluid density p is:
where M, K and A are Jhe assembled matrices of M, K
and A, respectively, and c and q are assembled vectors of c
and q, respectively.
For convenience, we can drop the notation "--" and (8)
to obtain the following:
c = Be + g
where
B = -M"1 (K + A)
g = M" q
(9)
(10)
(11)
If one assumes a zero normal gradient of concentration
across the whole boundary, the entries in the vector q all
vanish except those corresponding to source locations. For
example, for a source of strength r located at the j node of
the descretized finite element domain then the entry of the
j th row of vector q is r.
The problem of locating a pollution source is an inverse
problem of that just being described. The concentration is
considered known, in fact, they are given from the concen-
tration measurements taken at equal time intervals at ob-
servation wells. The problem is to identify the unknown
vector g which contains the information of where the
source (or sources) is located and what is its strength (or
strengths).
c {(K + A) c - Me - Q} = 0
in which the superscript T indicates transpose and,
c = concentration vector of nodal unknown
c = time derivative of c
0 - interpolation function
(7)
+ D
dx dy
dx dy
K =
A = //pb[u+ (f£)T +
m
M = //pb if> dx dy
Q - / q^ds + Z q± 6 (x - x*)(y - y*)
q = normal derivative of c prescribed on segment Sj of
the boundary
By assembling the element matrices into global matrices
for the whole solution domain and applying the boundary
conditions, one arrives at the final form of finite element
formulation of the problem as follows:
______ (8)
Me + (K + A) c = q
METHOD OF SYSTEM SENSITIVITY
Let yt be a measurement vector at time t then:
y, = Fct (12)
where F is the location matrix of the observation wells.
F is an m x n matrix with zero entries except at well nodes:
(13)
0 ••
0 -
0 .
• 0
•10
m
(14)
If yt is an estimate of y, and
J, = (y,-y,)TR(y,-yt)
then J, represents the magnitude of error between the es-
timated and observed concentration at time t. (R is a
weight matrix of the observation wells.) In particular,
for R = I (identity matrix) then
m
E (yi
1=1
-------
GEOHYDROLOGY MATH MODELING 319
The system sensitivity theory provides a recurrence
formula for g (estimate of g) for the next time step in a
manner so as to minimize Jt. When Jt approaches a
minimum value, g approaches the real g.
agi is the so called trajectory sensitivity vector which
should be obtained from the governing equation of the
system. From equation (9) it can easily be shown that
St+At
9g
(23)
in which A is the step size, ranging between O and 1 for
normalized elements. ^ will be evaluated through the
governing equation of the physical system and the mea-
surement matrices.
For convenience,
let
= yt - yt
(17)
Furthermore, if the time derivative of the concentration
is approximated by a finite difference formulation, i.e.,
Ct+At " Ct
At
(24)
the combination of equations (23) and (24) gives the de-
sired expression of the trajectory sensitivity vector in
iterative form:
thus
Jt = et R et
and
(18)
3c
= (B
3c
t+At
(25)
i = 1.2,... n
9J
now
3J
and
3J
t 9et ' 3*d
i =1.2 n (19)
(efc R et) = 2R et
(20)
(yt - yt)
(21)
in which
a.
•*- i th row
In short, equations (16), (22) and (25) compose the itera-
tive numerical system for obtaining the converged value
ofg.
OVERVIEW OF COMPUTATIONAL PROCEDURE
T~
.... n
thus
3J
1.2, .... n
(22)
For illustration, a generalized example will be discussed.
Consider a given suspect site that is arbitrarily subdivided
into a finite element pattern of quadrilateral elements.
The groundwater flow pattern can be in any direction with
respect to this pattern but must be known or, at least, as-
sumed. The observation wells are located on the grid and
are numerically generated through equation (8). With data
from these wells, the problem becomes one of locating
the pollution source which is arbitrarily shown in Figure
1. (Note that multi-source problems can also be handled
by this method.)
The computer programs are written in such a way that
the assembled matrices, M, K, A can be called out for
matrix operation. Arbitrarily select 1/3 of the total nodes
-------
320 GEOHYDROLOGY MATH MODELING
• Observation wells X Source
Figure 1.
Example Problem Showing Typical FEM Grid and
Source and Observation Wells
and regard them as observation wells. The generated time
series of concentrations at these nodes then serve as
measurements at observation wells.
The flow chart of a sample computer program is given
in Figure 2. The convergence criterion currently being
used is:
I (gI+1 - g^/gj- I 1 io"3 (26)
i.e., when the relative error in each entry of the g vector is
less or equal to 10~3, it is considered to be a convergent
system.
During the initial iterations, adjustments are needed
for selection of proper step size (A). To big a step size may
cause the process to diverge. It, in general, reflects on
Supply initial values
Main Iteration scheme
exit
Figure 2.
Computer Flow Chart for Program Currently in Use
-------
GEOHYROLOGY MATH MODELING 321
the growing residue. The step size should be cut in half if
such a situation occurs. As noted in the introduction, pro-
grams are currently being run using the method in a wide
variety of situations. Results will be forthcoming shortly.
They are in the form of pollution contours, the focus on
which clearly identifies the pollution source (or sources).
SUMMARY AND CONCLUSIONS
The general problem of locating a pollution source by
having data from a limited number of observation wells is
quite challenging, yet worthwhile, since its environmental
impact is extremely important. One potentially useful
analytic technique for this purpose is the use of the finite
element method (FEM). While a good deal of information
exists on this subject as far as taking a known source and
mapping out its migration, the inverse problem of finding
the source from a limited number of observation points is
more difficult.
Still more important is that of the available methodolo-
gies, the application of a systems approach (as described
herein) is a new solution technique. With a steady-state
flow field assumed, the system control equation is linear
and the criterion function can be minimized to an arbi-
trarily selected value.
The method of system sensitivity was described in the
paper along with related theoretical considerations. This
analytic formulation was the basic objective in this study.
While work is still ongoing, the basic programs are de-
veloped (see flow chart of Figure 2) and various problems
are currently being evaluated.
REFERENCES
1. Roux, P.H. and Althoff, W.F., "Investigation of Or-
ganic Contamination of Groundwater in South Bruns-
wick Township, New Jersey," Groundwater, 18, No. 5,
1980.
2. Faust, C.R. and Mercer, J.W., "Groundwater Model-
ing: Recent Developments," Groundwater, 18, No. 6,
1980.
3. Cooley, R.L., "A Method of Estimating Parameters
and Assessing Reliability for Models of Steady State
Groundwater Flow," Water Resource Research, 13,
No. 2, 1977.
4. Yeh, W. W-G and Tauxe, G.W., "Optimal Identifi-
cation of Aquifer Diffusivity Using Quasilineariza-
tion," Water Resource Research, 7, No. 4, 1981.
5. Frind, E.G. and Finder. G.F., "Galerkin Solution of
Inverse Problem for Aquifer Transmissivity," Water
Resource Research, 9, No. 5,1973.
6. Yeh, W. W-G, "Aquifer Parameter Identification,"
Journal of Hydraulic Division, ASCE, 101, HY9,1975.
7. Adey, R. and Brebbia, C.A. "Numerical Method in
Fluid Dynamics," Pentech Press, 1974.
8. Connor, J.J. and Brebbia, C.A., "Finute Element
Techniques for Fluid Flow," Newnes-Butterworths,
1976.
9. Frank, P.M., "Introduction to System Sensitivity
Theory," Academic Press, 1978.
-------
DIOXIN INVESTIGATIONS IN SOUTHWEST MISSOURI
DANIEL J.HARRIS
U.S. Environmental Protection Agency, Region VII
Surveillance and Analysis Division
Kansas City, Kansas
INTRODUCTION
During 1971, in a well-known incident of waste misman-
agement, three horse arenas in Missouri were sprayed for
dust control with a mixture of used oil and chemical waste
still bottoms. In the ensuing three years, 64 horses died and
ten adults and children became ill as a result of contact
with the contaminated soil.
It was not until 1974, and after excavation and removal
of the arena soils, that tetrachlorodibenzo-p-dioxin
(TCDD) was identified as the causative agent. Investiga-
tion by the Missouri Division of Health and the Center for
Disease Control (CDC) traced the source of this substance
to the defunct Northeastern Pharmaceutical and Chemical
Company (NEPACCO), which had leased manufacturing
facilities at a chemical plant in Verona, a small com-
munity in southwest Missouri.
As a result of an anonymous phone call from a former
NEPACCO employee, the Environmental Protection
Agency (EPA) became interested in the past activities of
this company in 1979. The caller made allegations con-
cerning other sites where NEPACCO wastes were disposed
of and provided names of prospective interviewees. The
historical record of the company and the specter of addi-
tional TCDD sites, coupled with the Agency's burgeon-
ing interest in hazardous wastes, prompted an immed-
iate investigation. This paper summarizes the results of
that ongoing effort.
HISTORY AND BACKGROUND
In November of 1969, NEPACCO leased a specific man-
ufacturing area of the Hoffman-Taff, Incorporated chem-
ical plant in Verona for the purpose of making hexachlor-
ophene. Hoffman-Taff had previously used this area in
the manufacture of "Herbicide Orange" for the military.
As a result of curtailed usage in Viet Nam, Hoffman-
Taft lost its contract and the facilities sat idle for a num-
ber of months. Shortly after NEPACCO leased the man-
ufacturing line, the entire plant was acquired by Syntex
Agribusiness, Inc.
The beginning processes of the herbicide and hexachlor-
ophene operations were similar, and apparently the idle
facilities represented an opportunity for NEPACCO to
begin manufacturing with minimum capital expenditure.
Both processes involved the intermediate production of
2, 4, 5-trichlorophenol (2, 4, 5-TCP). A major difference
in the two operations, and of central interest to this in-
vestigation, was the subsequent treatment of the 2, 4, 5-
TCP. Hoffman-Taff produced a technical grade herbi-
cide which permitted impurities, including TCDD, to go
forward into the final product. The pharmaceutical grade
hexachlorophene necessitated distillation of the 2, 4, 5-
TCP, which resulted in a still bottom waste stream with
TCDD concentrations in the range of 350 mg/1.
Between approximately April 1970 and January of
1972, NEPACCO is reported to have produced 328
batches of 2, 4, 5-TCP. Shortly thereafter, the company
lost its market and ceased operations as a result of the lim-
ited use restriction imposed on hexachlorophene by the
Food and Drug Administration (FDA).
EPA INVESTIGATION
Agency investigation of the past activities of NE-
PACCO began in October of 1979 and resulted from an
anonymous phone call from a disgruntled employee who
had been terminated by Syntex; he had worked in the
NEPACCO area. This individual identified other former
employees who allegedly had knowledge of NEPACCO
operations, waste management practices, and disposal
sites in the area.
In response to this information, a field team spent two
weeks in southwest Missouri conducting interviews with
former NEPACCO and Hoffman-Taff employees, inde-
pendent trash haulers and landowners and in reconnais-
sance of reputed disposal sites. A parallel effort of the
investigation included retrieval and review of all available
historical information on the company which encompassed
FDA files, Missouri Department of Natural Resource and
Division of Health files, court depositions from the horse
arena incident, as well as in-house documents.
This initial investigation resulted in the identification
of three potential sites which included Baldwin Park in
Aurora, the Syntex trenches west of the Verona plant,
and the Denney Farm Site, which was sampled by the
Agency in April of 1980. As a result of the publicity
surrounding the Denney site investigation and the num-
erous contacts between EPA employees and area resi-
dents, other rumors came to the Agency's attention. These
leads were pursued in August of 1980 and resulted in the
identification of additional sites.
322
-------
CASE HISTORIES 323
To date, the Agency has interviewed approximately 75
individuals and has identified seven proven or potential
NEPACCO waste disposal sites which are in various stages
of investigation or cleanup. In significance, these sites vary
from major undertakings such as the Denney Farm Site,
where remedial work is underway, to farms having small
volumes of NEPACCO wastes with TCDD concentra-
tions in the parts per trillion (ppt) range.
SITE STATUS
Baldwin Park
Located in Aurora in Karst topography, this 170-
acre site was at one time a lead and zinc mining area
with shafts, reportedly, up to 380 ft. deep. Following ex-
haustion of mineral resources, the tract, or at least a 30-
acre portion of it, was used as a municipal dump by
Aurora, as well as surrounding communities.
According to interviewees, large quantities of expended
filter cake material, used by NEPACCO for product
purification, was disposed of in the dump along with
municipal trash. In 1973, the site was closed and reclaimed
and given its present name. Although a sample of the
NEPACCO filter cake discovered at another location was
found to contain TCDD at a level of 8 ppb, samples of
runoff and municipal and private wells surrounding Bald-
win Park have not revealed detectable levels of NEPACCO
contaminants. No final disposition of the site has been
made.
Denny Farm Site
Identified during the October 1979 reconnaissance, the
Denney farm is the largest proven site discovered to date.
The site is in a rural wooded area of Karst topography
on a ridge top approximately one third mile from the near-
est surface water which is a losing stream. Visual evidence
of the existence of the site included a depression in the
ground about 10 ft. wide and 50 ft. long, with a mound of
soil at the west end. Interviewees indicated that between
30 and 150 drums of waste were buried there in 1971. The
drums were dumped from the back of a truck, lay as
they fell, and were in marginal condition at the time of
burial. Although eyewitnesses were unanimous in their be-
lief that the trench contained NEPACCO waste, specific
information as to drum contents was completely lacking.
In the winter of 1980, the Agency decided to mount a
field sampling investigation with the objective of docu-
menting site contents and any lateral migration of these
contents. Although there was no direct evidence to indi-
cate the presence of TCDD, the knowledge that the drums
came from NEPACCO and the company's history of
waste mismanagement, dictated that the study plan be de-
signed to provide the highest levels of personnel protec-
tion. Consequently, a draft plan was developed and wide-
ly circulated for comment within the Agency as well as
without. Many excellent suggestions were received and
incorporated into the final plan.
Preparatory activities undertaken in support of the in-
vestigation included removal of site vegetation, the erec-
tion of an eight-foot security fence, personnel baseline
physical examinations, a town meeting in Verona to ex-
plain EPA activities and, finally, a field trip to the site for
the news media immediately prior to imposing access re-
strictions.
Sampling, which included collection of perimeter soil
samples as well as exposure and sampling of drums, took
place in April of 1980 with the highly capable assistance
and support of the EPA National Emergency Response
Team. Initial on-site activities included setting up a de-
contamination station for men and equipment, putting up
a wind speed and direction indicator, and the surveying in
of perimeter sample locations.
Protective clothing and safety gear, depending upon the
particular work task, ranged from disposable splash suits
with rubber boots, gloves and hard hats, to the encapsul-
ating "moon suits" for those individuals working in the
trench and sampling drum contents. During the time that
the trench was open, the rural fire department was on
site, and aircraft were restricted from the area.
Working from one side of the trench, a back hoe was
used to excavate down to a depth of 1.5 to 2.0 ft. Manual
shoveling completed the removal of soil from around the
drums in preparation for sampling. Thirteen drums, some
of them crushed and empty, were exposed. Eight drums
with visually varying contents were selected for sampling.
Following completion of sampling, the trench was immed-
iately closed and samples packed for shipment.
Analytical results received in June 1980 ranged from 1,4
ppb TCDD in some drums up to 319 ppm in others. A
single composite soil sample taken in the trench from
around the exposed drums had a TCDD concentration of
72 ppb.
Following confirmation of TCDD and based upon ex-
pert testimony that migration would occur, and that the
location was neither suitable for hazardous waste nor
could be made suitable, a temporary cap was put over the
trench to eliminate percolation of precipitation. Concur-
rently, EPA initiated a weekly monitoring program of pri-
vate wells and surface waters in the area, and directed
Ecology and Environment, Inc. to prepare an engineer-
ing report evaluating appropriate remedial measures. The
Agency also entered into discussions with Syntex Agri-
business.
Effective September 12, 1980, Syntex, while denying
knowledge or responsibility for any acts of NEPACCO or
others in placing wastes in the trench, entered into a con-
sent decree to undertake remedial measures at the site.
The company's site restoration plan, comprehensive in
scope and sophisticated in design, was approved by the
Agency on February 25, 1981. Following construction
and stationing of numerous on-site structures and ancillary
equipment, Syntex began removal of trench contents on
June 15, 1981. As of the writing of this paper, remedial
work is still underway.
Syntex Trenches
Located on Syntex property west of the company's
plant in Verona are a number of trenches, now inac-
-------
324 CASE HISTORIES
tive and covered, which were ostensibly used by Hoff-
man-Tate for the disposal of uncontaminated trash. Two
of these trenches were reportedly open during the time
NEPACCO was active. Some interviewees have alleged
that NEPACCO waste, both drummed and in bulk, was
put in these trenches, apparently without the knowledge of
company management. Syntex and the Agency have open-
ed discussions on this potential site and exploratory sampl-
ing is anticipated in the near future.
Farm Site A
A former NEPACCO employee rented a house on this
farm located a few miles south of Verona. Six 30-gallon
open top drums, originating from the Verona plant and
containing minor volumes of a dark residue, had been left
in a barn on the farm by the employee. The drums came
to the attention of the Agency in late 1979 and were
sampled in the winter of 1980. Analysis indicated 7 ppb
TCDD. Syntex subsequently agreed to take the drums
along with limited quantities of hay, corn and potatoes
which were also stored in the barn. A barn dust sample was
collected following removal of these materials and final
disposition of this site is awaiting the analytical result.
Water and Wastewater Technical School
The hexachlorophene operation in Verona also resulted
in the production of a high-strength refractory waste-
water which was initially discharged to evaporative la-
goons owned and operated by Hoffman-Taft. In Septem-
ber of 1971, Hoffman-Taff refused to accept anymore of
this wastewater and NEPACCO subsequently entered in-
to a contractual arrangement with the Water and Waste-
water School in Neosho, Missouri for acceptance of this
wastewater and for the conduct of treatability studies.
For convenience, a portion of this wastewater was put in
a small open top steel tank on the school grounds for
analysis and the treatability studies.
As a result of the publicity surrounding the Agency's
ongoing investigation in the area, school officials became
suspicious of the asphaltic-like residue remaining in this
tank and collected a sample in the fall of 1980. Analysis
by a commercial laboratory indicated a TCDD concen-
tration of approximately 2 ppm.
In response to this finding, EPA personnel assisted in
a cleanup of this solid material on school grounds which
had accumulated as a result of students operating a valve
on the tank. Following confirmation of TCDD by the
EPA Regional Laboratory, which also reported a resid-
ual of 62 ppb in the spill area soil after cleanup, the tank
and 15 drums of residue mixed with attached soil were
removed to a remote bunker on school property for tem-
porary secure storage.
In addition, Syntex Agribusiness assisted by putting up
an eight-foot security fence around the spill area while
EPA personnel assisted the school in putting a temporary
cap over the area to eliminate the possibility of migra-
tion of contaminated soil particles. Although a final solu-
tion for the spill area and the drummed wastes has not
been determined, the public health and environment is be-
ing protected in the meantime.
Neosho Digester
Between October of 1971 and February 1972, 225,000
gallons of NEPACCO wastewater was trucked to Neo-
sho, about 30 miles west of Verona. The waste was placed
in an open top abandoned wastewater treatment plant di-
gester for storage, apparently with the expectation that the
Technical School, which operated the plant at the time,
would find some means for treating it.
After NEPACCO went out of business, the waste re-
mained in the digester. Starting in 1977, after the city of
Neosho resumed operational control of the wastewater
treatment plant, the digester was filled in with rock, soil,
and various other fill materials. In March of 1981, an EPA
team, using a drilling rig, collected a core sample from the
bottom (20 foot depth) of the digester. Analysis of the
sample revealed 2,500 ppm of 2, 4, 5-TCP and 60 ppb
TCDD. Following this finding, down gradient stream and
private well water samples were collected. The analytical
results are not available as of the writing of this paper and
disposition of the site is pending.
Farm Site B
In addition to Baldwin Park, two farms are known to
the Agency as having received the NEPACCO filter cake.
Although the source of the rumor is not known, there was
a belief among some local farmers that the filter cake,
presumably because of residual hexachlorophene, was
good for controlling hoof rot in cattle. In an imaginative
approach to waste disposal, NEPACCO permitted fann-
ers to take this material, which was subsequently spread
in feedlots and gate areas in such a manner that the cat-
tle would walk through it. One feedlot sampled in Aug-
ust of 1980 was found to have 0.4 ppb. In addition, 8 ppb
was found in a sample of the filter cake itself, which be-
cause of a leaking drum, was dumped on the ground and
never spread or mixed. Although the landowner has been
cautioned to stay away from the two areas, final dispo-
sition of the site is pending.
DISCUSSION
At this point in time, all NEPACCO sites are suspect.
The objective of the investigation in southwest Missouri is
to determine where all NEPACCO wastes, of whatever
description, were disposed of, and to take appropriate re-
medial measures at those sites judged to pose some hazard
to public health or environment. This objective is being
accomplished by a number of parallel paths which in-
clude interviews of former NEPACCO employees and
area residents, sampling and technical evaluation of avail-
able documents and information, with the aim of recon-
structing the NEPACCO process and developing esti-
mates of wastes quantities.
Approximately 35,400 gal. of NEPACCO still bottoms
have been accounted for to date. This figure does not in-
-------
CASE HISTORIES 325
dude the bottoms recovered from the Denney site which
have not been entirely quantified as of the writing of this
paper. Even less quantification is possible with other
NEPACCO waste streams, and there is no assurance that
other sites will not surface. Indeed, at this time, the Agen-
cy is aware of rumors and hints of other sites too vague to
be addressed in this paper.
Some ten years after the event, this investigation has
proven to be extremely resource intensive. Time has
dulled the memory, some individuals have moved away,
others have died, and still others have expressed concern
about publicity and are reluctant to talk with EPA inves-
tigators. The investigation has been further hampered by
unavoidable delays in obtaining NEPACCO records, by
limited access to low analytical detection capabilities for
TCDD and finally, by the absence of guidance criteria in
determining environmentally acceptable levels for TCDD.
The activities of NEPACCO in the handling and dis-
posal of their various waste streams are classic examples
of hazardous waste mismanagement. These practices have
adversely affected the health and welfare of some individ-
uals and have produced a sense of fear and uncertainty in
the minds of others.
In bright contrast, the residents of southwest Missouri,
and the Agency, are fortunate in having the problem solv-
ing capabilities of Syntex Agribusiness brought to bear
upon the legacy left by NEPACCO. The commitment of
Syntex to the Denney Farm Site has resulted in a model
project which is testimony to the company's positive en-
vironmental philosophy.
-------
THE DENNEY FARM SITE REMEDIAL PROJECT:
A MODEL FOR THE SAFE EXCAVATION,
STORAGE, AND ELIMINATION OF DIOXIN
RAY FORRESTER
Syntax Agribusiness
Springfield, Missouri
INTRODUCTION
Syntex' project to clean up and permanently dispose of
hazardous wastes, abandoned some 10 years ago by an-
other company, may be considered a model of industrial
governmental cooperation. Its success is due to the efforts
of Syntex' sophisticated multi-disciplinary team and to
the commitment, cooperation, and sense of mutual respect
that developed among Syntex Agribusiness, the United
States Environmental Protection Agency (USEPA) and the
Missouri Department of Natural Resources (MDNR).
This joint effort was particularly important because
Syntex did not generate the problem but inherited it.
Syntex neither created nor disposed of the wastes in the
disposal trench, and the disposal site is not on Syntex
property. Syntex found itself involved in this matter only
because it purchased a chemical plant at Verona,
Missouri, from another company which had earlier leased
part of its facility to the North Eastern Pharmaceutical
& Chemical Co., Inc. ("NEPACCO") for the manufac-
ture of hexachlorophene. In the summer of 1971
NEPACCO, in violation of its lease and unknown to
Syntex, arranged with the owner of the Denney farm, lo-
cated in Southwest Missouri, to have barrels of its waste
transported to the farm and disposed of in a trench lo-
cated on the property. It is this trench and its contents
that Syntex, the USEPA and the MDNR have been con-
cerned with since.
In 1979-80 the disposal trench was found by an EPA
investigation team and determined to contain highly toxic
wastes, including 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin,
commonly known as dioxin. Syntex, the EPA and the
MDNR joined together to develop a plan designed to clean
up safely the toxic wastes found in the disposal trench.
Unfortunately, the task was seriously complicated by
the absence of crucial facts such as the size of the trench,
the number of barrels in it and the identity of the wastes in
those barrels. Testing was undertaken to ascertain this in-
formation using ground penetrating radar, metal detec-
tors, electromagnetic conductivity and resistivity. How-
ever, the results, while providing an estimate, were not as
helpful as one might have wished. For example, after
various tests over a number of months, estimates of the
size of the trench ranged from a rectangular depression
53 feet long, 10 feet wide and 10 feet deep, to a left shoe
print shaped trench some 65 feet long, 20 feet wide and
10 to 12 feet deep. In fact, the trench was found to be
approximately 60 feet long, 8 or so feet wide and only 6
feet deep. Estimates of the number of barrels in the trench
ranged from 30 to 150; approximately 90 were found. Per-
haps most difficult was that while there was some informa-
tion on the range of wastes generated by NEPACCO,
which included innocuous materials as well as flammable
solvents and dioxin, no one knew their concentrations or
quantities.
TEAM APPROACH
To develop a plan to solve this problem safely, Syntex
realized early in the project that the expertise of a num-
ber of disciplines was necessary. Needed was expert tox-
icological advice to assess what risks existed at each stage
of the operation and a determination of when protective
clothing was necessary and what types were advisable to
wear. Needed also was a determination how best to as-
sure that no materials would escape into the environment
if some untoward event occurred during excavation. State
of the art analytical capability was crucial for quick,
accurate assessments of the materials found in the trench.
Particularly important was a construction crew that could
perform under conditions made difficult by the presence of
extensive safety equipment and could operate their ma-
chinery with such delicacy that drums which had thin
sides due to corrosion would not rupture.
Finally needed were sophisticated legal and public rela-
tions input to help develop the plan and maintain the
firm's good relations with the EPA, MDNR and the local
community.
Therefore, Syntex gathered together a multi-disciplinary
team composed of an engineer, equipment operators,
skilled laborers, a lexicologist, specialists in environmen-
tal health and safety, analytical chemists, lawyers, public
relations experts and executives in order to devise a plan
that could safely tackle the Denney farm site disposal
trench.
PLAN DEVELOPMENT
After months of hard work, Ithis team developed a plan
whose goal was to minimize and protect against possible
dangers to on-site workers, the nearby community and the
environment while simultaneously maintaining the mutual
326
-------
CASE HISTORIES 327
objectives of the parties involved: to remove and dispose
of hazardous wastes located at the Denney farm site.
The Denney farm site Final Plan did exactly that. It
established a series of procedures to remove all materials:
liquids, flowable sludges, chemical solids, containers, vis-
ibly contaminated soils and other soil from the Denney
farm site disposal trench to provide immediate protective
containment of the hazardous wastes excavated and to de-
toxify those wastes with higher concentrations of dioxin
and to provide an environmentally safe way simultan-
eously to contain and detoxify any low level contaminated
soil from the trench.
Safety Program
However, prior to beginning actual work the Final Plan
required every individual who would work at the Denney
farm site trench to undergo special medical examinations
and to complete a specially designed training program. A
detailed health and safety review was conducted by occu-
pational physicians, toxicologists and hygienists; it focused
on determining if any worker had a particular sensi-
tivity or susceptibility to dioxin or other toxic chemicals
believed to be in the disposal trench. Particular atten-
tion was paid to routes of exposure via the respiratory
tract, skin and mucous membranes. Any worker with such
a sensitivity or susceptibility was not permitted to work on
the project. All workers passing the medical exams then
underwent extensive training at the Verona plant prior to
actually beginning work at the disposal trench. During this
training they were instructed on the tasks they were to per-
form at the trench and practiced safety and emergency
procedures.
The prospective workers also learned how to use the
various types of required protective clothing. The most
stringent protective clothing was Category I equipment
consisting of complete respiratory, skin and mucous mem-
brane protection.
Category I protection was used by employees work-
ing in the disposal trench while the waste from the trench
was being removed. These workers used a supplied air
continuous flow system for maximum working efficiency
consistent with the primary objective of full assurance of
respiratory safety. They were also protected by a full body
guardian chemical suit (akin to a "moonsuit"), a white cap
safety helmet completely covering the head and neck, and
neoprene safety boots. This equipment afforded com-
plete body protection and allowed the mobility necessary
to remove safely the drums and contaminated soil.
Category II equipment, used only outside of the dis-
posal trench, had less stringent protective requirements. It
consisted of a chemical protection suit, safety hat, gloves,
boots and an air purifying respirator.
Category III equipment was required for the relatively
minor operations. Category HI standard safety equipment
and clothing included a safety hat, disposable Tyvek suit,
coveralls and safety boots.
After those workers who would be participating in the
excavation and monitoring activities at the Denney farm
site underwent medical examinations, were specially train-
ed for the safe handling of dioxin, learned how to use
their safety equipment, and were schooled in emergency
procedures, construction at the disposal trench was ready
to begin.
SITE CLEANUP
The first order of business was to expand the 8-foot-
tall chain link fence already surrounding the disposal
trench to accommodate the additional structures neces-
sary to implement the Final Plan. This fence served to
prevent any individuals or animals from accidentally wan-
dering onto the site. A guard was also stationed on the
site to dissuade the curious and add an additional meas-
ure of safety and security to the area.
Next, a Temporary Work Site Protection Structure
(the TWSPS) 100-feet-long and 48-feet-wide was erected
over the entire disposal trench. The purpose of this struc-
ture was to provide a cover that would prevent rainfall
from entering the trench site, reduce the effect of wind
on the handling of contaminated soil and provide a safe
storage area for containment drums.
Once the TWSPS was erected over the disposal trench a
Temporary On-Site Drum Storage Area (TOSDSA) was
constructed to provide a diked, weather protected and se-
cure storage location for all drummed materials and over-
packs. The TOSDSA was constructed with reinforced con-
crete with an emphasis on structural integrity and leak
proofing.
Two Microbiological Degradation Basins (MDBs) were
then constructed in order to safeguard and treat contam-
inated soil excavated from the Disposal Trench.
Finally, before excavation began, ancillary but impor-
tant additional equipment was placed at the Denney farm
site: an air compressor to supply air to the workers who
would actually be in the trench, electrical wires, pipe lines
for water, fuel, nitrogen, a forklift, a generator and a back
hoe.
An anemometer was installed to determine wind speed.
If the wind velocity exceeded a certain level, the project
would be shut down to avoid the possibility of escaping
particles. Further, air sampling devices were placed about
the site and their filters were assayed regularly for dioxin;
none was ever found in any of these samples.
Excavation
The excavation began by removing the cap and the first
layer of earth. Then, workers carefully began digging
around the barrels in the trench with shovels. Both opened
and unopened barrels were uncovered. Once a barrel was
found, its contents and any material that had leaked out
were extracted by a special vacuum system and trans-
ferred into clean barrels. The contaminated barrels, once
emptied, were then carefully lifted from the disposal
trench and placed in overpacks. Once the new barrels and
overpacks reached their capacity, they were sealed, labeled
and taken to the TOSDSA.
After the barrels were removed, all the visibly contam-
inated soil was removed and placed in 55 gallon drums and
-------
328 CASE HISTORIES
moved to the TOSDSA. The remaining soil was then
placed in the MDBs. This soil will be tested and, if found
contaminated, subjected to treatment.
Using this method of excavation, removal, and storage
the disposal trench has been emptied—90 barrels were
found in it although some of the barrels were broken or
disintegrated so an exact count is impossible. The dis-
posal trench has been sampled and the samples are being
assayed to determine if the soil is still contaminated.
DIOXIN DEGRADATION
It is expected that the most contaminated materials will
be amenable to photolysis, which involves breaking down
the nuclear structure of dioxin using light energy, usually
in a specific wavelength range. Photolysis studies have
shown that dioxins may be photolytically degraded by
natural sunlight. A photolysis treatment plant provides the
photochemical reaction necessary to destroy the dioxin
compound by using an ultraviolet source of light in an en-
closed and controlled setting.
A photolysis treatment system was developed to treat
approximately 4300 gallons of an oily dioxin-contaminated
waste NEPACCO had abandoned in a tank at the Verona
plant. The first step was to make certain nothing leaked
from the storage tank. Six years of work preceded a de-
cision to go forward with development and use of the
photolysis process primarily because of its potential for
success and the degree of safety it offered. It is a low pres-
sure, low temperature, and fully closed system providing
almost no opportunity of human exposure to the material
to be detoxified as well as virtually no chance of toxic re-
lease to the surrounding environment.
The system was successful. During the summer of 1980
Syntex extracted virtually all of the NEPACCO waste
from the tank and destroyed 99% of the dioxin. This sys-
tem will be used to detoxify a portion of the materials at
the Denney farm site.
Some materials found at the Denney farm site can also
be disposed of through landfills and other EPA approved
methods. Other material cannot be so easily discarded.
Syntex has explored a variety of different technologies to
dispose of these substances such as microbiological degra-
dation which is the molecular degradation of an organic
substance resulting from the complex actions of living
organisms. To assist this biological process, various or-
ganic nutrients are to be added to treat the contaminated
soil in the MDBs. Syntex and the EPA will be closely mon-
itoring the dioxin levels in the MDBs to determine the
success of microbiological degradation.
CONCLUSION
While the project is not fully complete, the hazardous
wastes from the Denney farm site trench are excavated
and secured, with progress being made toward their ulti-
mate disposal. The success of this project is the product of
a team composed of a solid core of individuals with high-
ly developed expertise. It is also the result of the close
cooperation and mutual respect among Syntex, the EPA,
the MDNR and the local communities, without which this
project would have not been possible.
-------
A HAZARDOUS WASTE DISPOSAL PROBLEM
vs
A SYSTEMATIC APPROACH FOR
IMPOSING ORDER ONTO CHAOS
THOMAS O. DAHL
U.S. Environmental Protection Agency
National Enforcement Investigations Center
Denver, Colorado
INTRODUCTION
Occidental Chemical Company, a subsidiary of Hooker
Chemical Corporation and its parent, Occidental Pe-
troleum Corporation, owns and operates an agricultural
fertilizers and pesticides plant in the San Joaquin Valley at
Lathrop, California. This plant came into national
prominence in 1977 following discovery that a substantial
proportion of the pesticide workers were sterile. Pre-
sumably, the cause of this sterility was from exposure to
a nematocide, Dibromochloropropane (DBCP) which
was both formulated (1957 to July 1977) and manu-
factured (1963 to 1971 and 1974 to 1976) at the plant. Sub-
sequent to this discovery, the State of California banned
the sale and use of DBCP on August 12, 1977. This
initial action was expanded into a nationwide mandatory
suspension of the production and use of DBCP except in
certain limited areas.
The combination of the sterility issue and certain
Company documents indicating potential adverse envir-
onmental effects caused the U.S. Environmental Pro-
tection Agency (EPA) and State of California to investi-
gate the Company's operations. On December 18, 1979,
the U.S. and State of California filed suit in U.S. District
Court against Occidental. On February 6, 1981, following
over a year of technical studies and protracted legal
maneuvering, a consent decree was lodged with the Court
which established the framework for additional necessary
studies and implementation of remedies.
The activities which led ultimately to the successful
settlement of the case were characterized by a series of
investigations which first sought to understand the po-
tential problems, then defined the actual problems, and
finally designed necessary remedies. The first of these was
a multimedia environmental audit of the Company's
operations.
AUDIT OF OPERATIONS
From September 18 to 28, 1979, investigators from the
U.S. Government and State of California reviewed
Company records and conducted interviews of employees
at the Lathrop facility. The subjects of these inquiries
were: past and present waste-generating processes and
the disposal practices employed; in additional environ-
mental data were collected (air, water, soil, ground-
water). Attempts were made to gain information from
the beginning of operations (1953) to present. Informa-
tion on this time span was imperative because past dis-
posal practices could still be causing adverse effects in
the soil and groundwater systems.
The audit of Company operations in Lathrop, Cali-
fornia determined that past production and disposal prac-
tices relating to liquid and solid wastes were inadequate,
resulting in contamination of the groundwater and soil
in the vicinity of the plant. Major findings of the audit
included:
•Wet phosphoric acid operations, commencing in 1953,
resulted in the generation of considerable amounts of
gypsum (CaSo6) and losses of fluoride. The gypsum was
stored onsite during drying on permeable soil presumably
resulting in percolation of sulfate ions to the local
groundwater which was known to be between 2 and 7.6 m
of ground surface.
•Ammonia, ammonium sulfate and ammonium phos-
phate/sulfate operations resulted in discharge of am-
monium ions to unlined ponds, presumably resulting in
percolation to groundwater.
•Sulfuric acid plant operations resulted in onsite burial of
vanadium pentoxide catalyst causing, at a minimum,
soil contamination.
•Pesticide formulating and manufacturing operations,
some commencing as early as 1957, were characterized
by use of over 100 active ingredients. Liquid waste dis-
posal practices until at least 1976 included use of un-
lined ditches and ponds in permeable soil areas. Further-
more, until at least 1970 some pesticide waste solids
and concentrated liquids were disposed of onsite in shal-
low trenches. Presumably these trenches were either ex-
posed directly to the groundwater table or very close to
it. A review of Company records for selected years in-
dicated small percentage losses of pesticides. However,
although these were small in terms of percentages, they
represented large quantities when considering millions of
kilograms of pesticides were formulated and/or manu-
factured.
•A review of past environmental data available through
Company records and State of California files provided
strong indications of environmental contamination at-
tributable to the facility. Sulfate concentrations in one of
the Company wells increased from 28 mg/Vin 1960 to
over 2,000 mg/
-------
330 CASE HISTORIES
had been detected near the facility in soil, shallow
groundwater and process, irrigation and potable wells.
All these compounds were known to have been handled
at the facility. Of these, 12 are known animal carcino-
gens.
•An aerial overflight of the facility conducted as part of
the audit, using false color infrared imagery, indicated
dark and discolored areas to the west of previous dis-
posal areas. Vegetation appeared to be stressed; this pat-
tern could have been caused by leachate migrating from
the disposal site.
PLAN OF STUDY
Based on the audit findings and State of California
records indicating probably significant, albeit not fully
defined, environmental contamination, a plan of study
was formulated to address the problem. This plan of
study, developed jointly by the Company, the State of
California, and the USEPA National Enforcement In-
vestigations Center, presented a comprehensive approach
for systematically determining whether certain inorganic,
organic and radiological substances were present in the
groundwater and soils in the vicinity of the Company
facility and whether or not these substances were attri-
butable to the Company's operations. The study would
determine whether, and if so, which mitigation and re-
medial measures were necessary to prevent an imminent
and substantial endangerment to public health and envir-
onment. Programs for groundwater analysis, soil analy-
sis, private and public well monitoring and groundwater
modeling were designed. The plan of study was divided
into three phases:
Phase I—That phase of the groundwater' and soil
analysis program necessary to determine whether any or-
ganic, inorganic or radiological constituents were present
in groundwater or soils in the vicinity of the Company's
plant. This was to include, but not be limited to, the
following:
•Define the vertical and horizontal zone of the afore-
mentioned constituents within the groundwater and soils
•Establish the direction of movement of the subsurface
plumes and their rates of movement
•Develop proposed mitigation measures if any are de-
termined to be necessary to protect public health and
the environment against imminent and substantial en-
dangerment from contamination of groundwaters and
soil resulting from the Company's operations
Phase II—That phase of the groundwater and soil
analysis program necessary to:
•Evaluate the overall effectiveness of any mitigation
measures, if any were determined to be necessary
•Define any additional remedial measures needed to be in
compliance with waste discharge requirements and pro-
tection of public health and the environment from an
imminent and substantial endangerment
Phase III—That phase of the groundwater and soil
analysis program conducted after the implementation of
mitigation measures, if any were determined to be neces-
sary, that would evaluate ongoing compliance with the
waste discharge requirements and protection of public
health and the environment.
Groundwater Analysis Program
A total of 14 monitoring sites were selected surround-
ing the facility (Figure 1). Eight of these locations were
selected to reflect presumed upgradient and downgradi-
ent groundwater conditions, the areal confines of the fa-
cility and locations of past disposal sites. Six additional
sites were selected to reflect conditions radiating out sev-
eral hundred meters from the plant boundary locations.
Sampling at these sites would enable determination of
whether and to what degree contaminants would attenu-
ate in the directions of assumed groundwater movement.
Three permanent wells were to be constructed at each
site. The intent was to locate wells at depths of greatest
contamination, background conditions below the leachate
plume and approximately midway in the plume. It was
recognized that additional wells may become necessary if
the geohydrology of the area created several plumes
rather than one. In fact, there are now 55 monitoring wells
installed at the site (Figure 2).
One of the principal dilemmas faced by the formulators
of the plan of study was what parameter(s) would be
used to define the depths of the desired wells. As noted
previously, numerous compounds, both inorganic and
organic, had been handled at the facility, presumably with
varying attenuation and migration rates. Of prime'con-
cern were many of the organics with known adverse
health effects. After considerable discussion, it was de-
termined that existing state of the art information could
not afford an exact determination of the best paramet-
er (s). Intentions were to initially examine groundwater
conditions down to approximately 46 m or background
conditions, whichever was greatest. Because this would
mean numerous samples, the issues of expense and time
to analyze indicator parameter(s) had to be addressed.
Sulfate was finally chosen as the primary indicator be-
cause (1) considerable quantities had been introduced to
the groundwater by the facility from the gypsum ponds,
(2) sulfate was believed to be as mobile as any of the or-
ganics and (3) it could be monitored rapidly and inex-
pensively. To augment this indicator, additional indicator
analyses were performed onsite for pH, conductivity,
Redox potential and ammonia nitrogen.
In order to delineate areal stratigraphy, particularly at
the well sites, samples were to be collected at representa-
tive depths during drilling. Once all stratigraphy data were
collected from the 14 sites, three representative bore holes
to 61 m were to be drilled in the study area to collect
undisturbed samples for representative permeability and
porosity measurements.
Drilling techniques employed to accomplish desired
goals included both a dual tube reverse air rotary rig and a
conventional rotary rig. With the dual tube rig, air flows
down the outer annulus of the drill rod and carries cut-
tings and groundwater up the inner tube to ground sur-
face and exits via a cyclone. Since air flows are on the
order of 30 m/sec, an observer at ground surface sees es-
-------
CASE HISTORIES 333
Figure 1.
Study Area Showing Permanent Wells Sampled During
Phase I and Interim Phase II Study.
(Scale: 1" = 1,100')
sentially an instantaneous representation of cuttings and
groundwater. This procedure enabled a continuous logging
of stratigraphy by an experienced geologist and also af-
forded indicator groundwater sampling at approximately
3 m intervals. Once a depth of desired sampling was
reached, advancement of the drill rod was halted but the
air was left on. This afforded a representative water
sample since the rapid air flow purged the drill rod of
any contents. At certain times, when in clay formations
and some sandy material, plugging occurred and it was
necessary to add a "water mist" to the airstream to lubri-
cate the drill bit. Representative groundwater samples
were then obtained by stopping the mist approximately
1 m before the desired depth, advancing the drill rod and
allowing ample time for air purging of any mist water.
Once desired groundwater and stratigraphy samples
were collected at a site, conductivity, sulfate, ammonia
nitrogen and stratigraphy data were plotted against depth.
Three depths were then selected for permanent wells. Al-
though the depths were ostensibly to represent the pre-
viously mentioned criteria of highest sulfate concentra-
tion, background sulfate and mid-concentration, reality
often dictated a more detailed determination. For ex-
ample, in some instances, there were no sharply defined
curves; consequently, a combination of the data plots, the
existing stratigraphy and considerations of adjacent site
data had to be used to select depths of permanent wells.
Once depths were selected, 30 cm bore holes were
drilled using a conventional rotary rig for installation of
permanent wells of 15 cm diameter steel casing. Because
a major concern was precluding the introduction of any
contaminants into the wells, the rotary drill contractor
was instructed to only use water as a drilling fluid, if pos-
sible, rather than drilling mud. This proved successful at
all 42 wells installed at the 14 sites (three per site). At the
bottom of each casing was a concrete plug followed by a
1.5 m stainless steel screen to afford adequate withdrawal
from the water-bearing formation. All wells were gravel
-------
334 CASE HISTORIES
Figure 2.
Phase II. Permanent Well Sampling Locations
(Scale: 1" = 1,100')
packed to above the screen and then cement grouted to
the surface to isolate the screened section. To assure no
contaminants would be introduced at the surface, the cas-
ing extended to 0.6 m above ground surface and was
fitted with a metal cap which could be locked. Wells were
developed/cleaned by pumping at least 7,600 with an
airlift and/or a submersible pump. Purged water was col-
lected in tanks and returned to the Company's onsite
hazardous waste tanks for subsequent disposal.
Groundwater characterization sampling was conducted
with a submersible pump and an air-inflatable packer.
The pump/packer assembly was lowered to a point where
the packer was just above the top of the well screen. The
packer was then inflated via an air line to the surface to
isolate the well casing from the screen down. The pump
was activated and a volume of water equal to at least 5
volumes of casing below this screen was purged to the sur-
face via a Teflon* tube and discharged to hazardous waste
tanks. Representative samples could then be collected for
desired analyses (Table I). The selection of these parame-
ters was based on compounds handled at the facility in
greatest quantities, based on audit findings and com-
pounds previously found in groundwater and soil mirnpta
in the area. The analyses selected included 29 pesticide*,
12 metals and other inorganics and a radiological assay.
Once sampling was completed at a given well, the pump/
packer/Teflon-tube assembly was moved to the next well
and used again. It was assumed that the purging volumei
would be sufficient to carry away any contaminants left
on the Teflon from the previous well. To verify thto,
quality assurance sampling was periodically conducted by
purging the Teflon with deionized distilled water. Samphi
were collected of the deionized distilled water as wcU it
the water discharged from the Teflon at the end of purginf.
•Registered Trade name.
-------
CASE HISTORIES 335
Table I.
Parameters Chosen for Groundwater Characterization
Organics
DBCP
ETDB (EDB)
CE-BHC
LindaneCY-BHC)
/3-BHC
A-BHC
Aldrin
Toxaphene
Thiodan I (Endosulfan)
Chlordane
DDT
Perthane
Methoxychlor
Ethyl Parathion
Methyl Parathion
Malathion
Diazinon
Disyston (Disulfoton)
DBF
Dimethoate
Trithion
Dibrom
Methyl Trithion
Dioxathion
Fenthion
Phosdrin (Mevinphos)
Ethion
Sevin
DNBP
Inorganics and Metals
Nitrate Fluoride
Sulfate Copper
pH Zinc
Conductivity and Temperature Chromium
Vanadium Arsenic
Nickel
Other
Oxidation Reduction (Redox) Potential
Radiological Assay (gross alpha scan, radium scan, uranium scan,
thorium scan and gross beta scan)
Soil Analysis Program
The audit of the Company's past practices revealed the
existence of a number of past disposal areas, including
unlined ponds, conveyance ditches and burial sites. Al-
though the locations of some of the sites were obvious
since they were above ground and clearly visible, others
such as alleged pesticide burial trenches in the so-called
"Boneyard" area were not. No Company person inter-
viewed could identify the number or exact location of the
burial areas. In approximately 1970, use of the Boneyard
area had been discontinued and subsequent filling and
grading had rendered visible identification impossible.
The only document available which appeared to show
the locations of some of the trenches was a blueprint
submitted to the State of California in 1970. Its accuracy
was presumed by the Company to be questionable.
In order to determine whether previously disposed ma-
terials needed to be escavated and transported to State of
California approved hazardous waste disposal sites, a
sampling program had to be designed for the known areas
such as past disposal ponds and ditches. This included
core sampling to approximately 3 to 5 m below grade.
Sampling locations were selected representing the geo-
graphical bounds of the disposal sites, the discharge
points and the lowest points (in ponds) (Figures 3 and 4).
The procedure used was to collect discrete samples at the
surface, 0.6 m, 1.5 m and 3 m. The surface sample (0-10
cm) was collected by shovel and placed into a solvent-
rinsed glass or other appropriately prepared jars for sub-
sequent analysis (Table II). The remaining samples were
collected using a hollow stem auger drill rig and split-
spoon samplers. The sampling procedure included drilling
down to the desired depth and then advancing a split-
spoon sampler with the rig to collect the desired soil sample.
Samplers were opened and contents transferred onto a
sheet of Teflon to preclude cross contamination. Samples,
once split by quartering or mixing and aliquoting, were
tranferred to appropriately prepared glass jars for trans-
port to labs for analysis. The Teflon sheet was rinsed
with deionized and distilled water and a 1:1 solution of
hexane and acetone between samples. The split-spoon
sampler was cleaned in a similar manner between samples
to preclude cross contamination.
Table II.
Parameters Chosen for Soil Sample Characterization
Organics
EDB
DBCP
oe-BHC
Lindane(T-BHC)
/3-BHC
A-BHC
Aldrin
Toxaphene
Thiodan I (Endosulfan)
Chlordane
DDT
Perthane
Methoxychlor
Ethyl Parathion
Methyl Parathion
Vanadium
Nickel
Copper
Metals
Other
Malathion
Diazinon
Disyston (Disulfoton)
DEF
Dimethoate
Trithion
Dibrom
Methyl Trithion
Dioxathion
Fenthion
Phosdrin (Mevinphos)
Ethion
Sevin
DNBP
Zinc
Chromium
Arsenic
Radiological Assay (gross alpha scan, radium scan, uranium scan,
thorium scan and gross beta scan)
PH
Relocating past disposal trenches/pits presented a most
interesting technical problem. As noted previously, the
only written evidence of possible locations was a 1970 site
map. A reconnaissance of the area indicated that some of
the map reference points could still be located. After con-
siderable thought and discussion the following approach
was adopted:
•Assume the site map was at least generally accurate
•Locate and physically stake out the disposal sites from
site map locations
•Attempt to locate disposal sites by cutting exploratory
trenches at right angles to the assumed directional lie of
the trenches. This included using a backhoe and cutting
-------
336 CASE HISTORIES
Figures.
Study Area Showing Soil Coring Locations Sampled During the
Phase I Study Excluding the Excavation Program.
(Scale: 1" = 460')
to the water table which was generally 1.8 to 2.4 m be-
low the ground surface (Figure 5)
•Once disposal trenches are intersected, rotate the explora-
tory trenches 90° and commence in directions of the dis-
posal trench. This procedure was employed to determine
the length of the trench as well as to assess, by visual
observation of trench contents, the need for immediate
total trench excavation versus sampling followed by pos-
sible excavation
•Once initial excavation and sampling activities are com-
pleted, collect soil core samples with depth adjacent to the
trenches for parameters previously referenced (Table II)
to ascertain whether migration of contaminants requir-
ing further excavation has occurred.
The approach detailed above led to the discovery of
four major disposal trenches and a number of pockets of
disposal pits. The discovery of one trench and how it wi*
subsequently addressed is particularly noteworthy. The ex-
ploratory trench intersected the disposal trench and effortl
began to continue exploratory trenches in the direction of
the disposal trench (north-south). Visual observation of
the contents (e.g., remnants of drums, bags and acid tank
liners), however, led the investigators to believe they had
located the major chemical trench referenced in internal
-------
CASE HISTORIES 337
Company records and interviews during the audit. This
led to a decision to mine out the entire trench with the
backhoe rather than proceed with exploratory trenches.
The approach taken was to first remove the approxi-
mately 0.6 to 0.9 m of overburden to one side of the trench
and then excavate from that point down to the water table.
Any metal drum remnants, pesticide produce containers
or other probable hazardous waste containers, were re-
moved by hand to hazardous waste bins or overpack
drums and hauled away to California-approved Class I
hazardous waste disposal sites. Soil and other trench con-
tents were excavated by backhoe and placed on sheets of
black plastic for subsequent sampling and decisions re-
garding disposal. The location of where the material came
from in the original trench was catalogued for future
reference.
While proceeding in this manner, thousands of '/2 and
1 pint glass pesticide bottles, some broken and empty of
contents, some partly full, or full were discovered. The
labels of these bottles indicated such products as 40%
DDT, 50% Chlordane, 2,4-D, 2,4,5-T and Dieldrin.
Ominously apparent was the fact that many containers
were still below the water table.
Because of the inherently hazardous nature of the dis-
posed material, a decision was made to excavate below
the water table and remove the material with sealed
trucks to Class I disposal sites. Considering the character-
istics of the disposal area and discussions concerning past
recollections of the trench operation, it was decided that
excavation 0.9 m below the water table should remove any
containers etc. originally deposited. The approach for
accomplishing this included use of a "mud wave." This
•8E V f*
Figure 4.
Study Area Showing Soil Coring Locations During
Excavation Program.
(Scale: I" = 135')
-------
338 CASE HISTORIES
90 worm
limi «• MO MITI MI1»UL
Mill «*H«I*V
OIMOMl 1MMH 0" W1
»0*IN« LOC1TIOMI
1IMIL SUHIMI UMIMTMtIM
VAMDIIM PUTOllOC miltl «
H*VI HIM Cifaxe vf
®AmN»iu*Tf locnic* or w*ti IMU
WT ni •HKN en NOT fwtwifll <
Figures.
Trench Location Plan
included starting from one end of the trench with a bull-
dozer, cutting down 0.9 m and moving the excavated
material a short distance forward in a wave. Clean fill
was immediately brought in to replace the excavated ma-
terial and the process was repeated. Water, pushed for-
ward by the wave, was pumped to a truck from a crude
sump carved out of the downstream end of the trench and
taken to the Company's onsite hazardous waste disposal
tanks. This procedure was continued until the trench was
entirely excavated and filled with clean fill. Soil exposed
at the bottom of the cut by the bulldozer blade while
advancing the wave had a clean appearance, supporting
the belief that the procedure selected would escavate all
previously deposited containers and solid materials. Eval-
uation of potential migration of contaminants through
the soil was accomplished by the previously described
core sampling program and the groundwater monitoring
wells.
Worker safety during excavation was of primary im-
portance to the investigators. Extensive physical examina-
tions were performed on the workers before and after
the project. While working at or near the trenches, workers
-------
CASE HISTORIES 339
were required to wear hard hats, safety glasses, fresh
respirators with multiple purpose cartridges (including
organic vapors), rubber boots with steel toes, rubber
gloves and cloth coveralls which were changed at least
daily. Cloth coveralls were selected over a more imper-
meable suit since actual physical contact, other than with
the rubber gloves, with excavated materials was minimal.
Since summertime temperatures in the Lathrop area
often exceed 38 °C, the cloth coveralls also posed less risk
to the worker from the standpoint of heat exhaustion.
For those workers who were collecting samples or other-
wise having greatest contact with materials, impermeable
aprons were provided.
Private/Public Well Monitoring Program
As part of the effort to ascertain the extent of ground-
water contamination as well as to assure protection of
the public health, both private and public wells in the
plant vicinity were sampled (Figure 6). Parameter coverage
included the Table I list, excluding pH, Redox potential,
conductivity and temperature.
Wells sampled included all known drinking water wells
within approximately 1.6 km of the facility. In addition,
the Company offered a one-time sampling to any of its
employees who were using private wells, however remote
from the facility. Procedures employed included first
conducting a well inventory survey followed by gaining
consent to sample. Sampling of approximately 28 wells
was conducted twice monthly for the first 3 months and
then monthly for 2 months. By that time at least four
complete sets of-data were available for each well sampled.
This number of complete samplings was judges necessary
by the State of California to support decisions regarding
possible closures of wells. Sampling methods included
collecting the samples in prescribed containers as close to
the well head as possible. Water lines were flushed for a
period of time judged sufficient to gather a representa-
tive sample.
Groundwater Modeling
Although the previously described monitoring of
groundwater wells and examination of soil samples was
essential to document the extent of contamination, it
was also recognized that a groundwater modeling effort
was necessary, particularly in designing and assessing
long term remedies. The purposes of the Phase I model-
ing efforts were to:
•Develop a preliminary two-dimensional model of the
study area
•Gain an understanding of the relative importance of the
factors that influence groundwater movement in the area
•Make a preliminary assessment of any existing and future
pollutant transport paths
•Define future groundwater modeling efforts for subse-
quent phases
Inputs to the first phase modeling effort included using
data gathered during the groundwater, soil and private
well monitoring activities as well as independent data-
gathering efforts. From the ongoing activities data was
gathered relating to stratigraphy, piezometric head
measurements from the permanent wells, and soil char-
acteristics (i.e. porosity, permeability, grain size, analy-
ses). In addition, data were gathered or estimated for
existing land use, well pumpage and construction charac-
teristics in the area, return flows from major users in the
area, precipitation recharge, boundary conditions and
anticipated soil-chemical-water relationships. Pump tests
were also performed using a combination of existing pro-
duction wells, permanent groundwater monitoring wells
and additional piezometers as needed.
Once the data were synthesized into the preliminary
model, several sensitivity studies were conducted to de-
termine the importance of:
•Vertical depths of the flow field
•Boundary conditions
•Recharge sources
•Pumping patterns (extent and seasonal patterns)
•Aquifer characteristics
Since an integral objective of the plan of study was to
determine necessary remedial measures, several model
simulations were conducted to forecast future pollutant
transport pathways resulting from various remedial al-
ternatives.
RESULTS
Implementation of the Plan of Study substantiated
initial concerns regarding the seriousness of the problem.
Considerable pesticide-contaminated soils were found in
a former disposal area, the pesticide pond bottom and
the ditch which conveyed pesticides to the pond. Based on
soil core sampling data, decisions were made regarding
the degree of need for offsite disposal at California
Class I or II-1 sites. In all, approximately 1100 m3 of ma-
terial were excavated and hauled to Class I sites and 2400
m3 to II-1 sites. Excavated areas were backfilled with clean
soil to within 1 m of grade. A 0.3 m layer of clay (106 cm/
sec permeability) was deposited above this, and then 0.6 m
of clean soil was deposited. The final grade was sloped to
accelerate runoff from the areas.
Groundwater sampling demonstrated the presence of
numerous pesticides in wells adjacent to past disposal sites.
DBCP was detected in numerous onsite wells, reaching
concentrations as high as 1.2 mg/5 in one well 61 m deep.
Sampling of domestic wells confirmed the presence of
DBPC in wells to the north by the Oxy facilities. Concen-
trations were considerably less than those found in the
test wells, (approximately 4 mg/p). Occidental, upon re-
view of these data, agreed to provide a new permanent
source of water to the affected and potentially affected
residents.
The data collected from the test and domestic wells were
also used to select an ongoing monitoring program for
February, June and October of subsequent years, as well
as locations for additional wells to further refine definition
of the areal and vertical extent of the groundwater con-
tamination. Two of the wells completed to over 91 m
yielded no detectable DBCP. Another well nest con-
structed west of the Libby Owens Ford (LOF) plant also
showed no detectable concentrations of DBCP. Ground-
-------
340 CASE HISTORIES
water modeling efforts confirmed the migration of
ground water to the west, consistent with assumed re-
gional travel, but also indicated a strong vector toward
the northeast, a result of heavy groundwater pumpage.
These physical studies and modeling results were used by
Occidental to design a proposed groundwater purge sys-
tem to cease further migration of contaminants offsite.
Purged groundwater from five extraction wells will be
treated onsite in an activated carbon treatment system.
Pilot plant studies have indicated removals of pesticides to
below detection limits. The Company has proposed in-
jecting treated wastewaters into a deep aquifer (more than
100 m) which already contains high dissolved solids. The
governmental agencies are currently reviewing this pro-
posal and once a remedial system is finally approved, it
must be installed within 180 days.
The extensive work referenced in this report was ac-
complished over an approximately two year period. In ex-
cess of 50,000 analyses of groundwater and soil were con-
ducted. It would be misleading to say that the work is fin-
ished or that it progressed without any differences of opin-
ion as to necessary studies and interpretation of results.
However, considering the state of the art nature of the
work and the extreme difficulty in working in the shadow
of an ongoing court case, the cooperation was remarkable.
The author believes this project served as an excellent ex-
ample of how industry and government can put their col-
lective expertise to good use in solving complex technical
problems.
0 - 0.6 MILES
BB -2.2 MILES—»
C -0.2
»0fc - 0.8
Figure 6.
Study Area Showing Domestic Wells Sampled
During Phase I Study.
(Scale: 1" = 1,100')
-------
IMPLICATIONS OF THE
CHEMICAL CONTROL CORP. INCIDENT
ADAM M. FINKEL
RICHARD S. GOLOB
Hazardous Materials Intelligence Report
World Information Systems
Cambridge, Massachusetts
INTRODUCTION
On April 21, 1980, a massive hazardous waste explosion
and fire took place at the Chemical Control Corp. above-
ground storage site in Elizabeth, New Jersey, consum-
ing at least 20,000 of the more than 45,000 drums of
hazardous waste present at the site. An unusual combina-
tion of wind and temperature conditions and the results
of a 12-month pre-fire cleanup, prevented the occurrence
of an unprecedented public health disaster.
As of September 1981, the post-fire cleanup had cost
more than $27 million, and the final site mitigation had
not yet been completed. At least four major civil and crim-
inal suits have been filed as a result of this incident, and
private damage claims total over $500 million.
The Chemical Control incident provides an important
case study of the problems surrounding hazardous waste
management and uncontrolled waste sites. No acute inci-
dent in the U.S. involving hazardous wastes has incurred
a greater response cost and no single acute incident in the
U.S. has involved a larger quantity of hazardous waste or
a more varied collection of waste.
In this paper the authors will focus on the events sur-
rounding this incident, and their implications for future
waste-site responses and for federal and state policies reg-
ulating these responses. The paper divides the incident in-
to five distinct phases:
(1) the operations of Chemical Control before the New
Jersey Department of Environmental Protection
(DEP) took any organized action at the site
(2) The cleanup that occurred during the 12 months be-
fore the fire
(3) The fire itself
(4) The intensive response actions during the year after
the fire
(5) The work remaining at the site and the ongoing costs
and litigation associated with it.
PRE-RECEIVERSHIP
This phase began in the mid-1970s, when Chemical
Control began accepting hazardous waste from area gen-
erators; it ended in January 1979, when DEP obtained
a court restraining order and, for the first time in history,
took control of a private company. According to the Over-
sight and Investigations Subcommittee of the House of
Representatives Interstate and Foreign Commerce Com-
mittee, DEP solid waste officials knew about the potential
problems at the Chemical Control storage facility at least
by the spring of 1978, at which time the site contained
about 25,000 drums.
DEP and the New Jersey Division of Criminal Jus-
tice (DCJ) both agree that their respective agencies did
not communicate adequately during the pre-receivership
phase because of understaffing. In addition, a DCJ mem-
orandum in 1978 claims that DEP investigators were not
adequately trained and, although aware of the problems
at Chemical Control, did not have the expertise to assess
their severity. Consequently, according to the memoran-
dum, the investigators treated the site as a landfill and
applied landfill guidelines to its regulation. Other state
officials have said that, if DEP regulators had made ade-
quate checks of the state-required manifest forms, they
would have recognized that Chemical Control, licensed as
an incinerator facility only, was accepting large quanti-
ties of physically non-incinerable wastes.
Since this occurred, New Jersey has developed the 11-
state Interagency Hazardous Waste Strike Force, a model
program for other enforcement and regulatory personnel
around the country. In addition, New Jersey's status as
the leading hazardous waste generating state has made its
regulatory agencies the leaders among the 50 states in deal-
ing with these problems.
The unregulated existence of Chemical Control during
its early history does not appear to be solely a function
of the times, though. Testimony given at an Oversight
and Investigations Subcommittee hearing on December 16,
1980 claimed that, even now with the implementation of
a national manifest system under the Resource Conserva-
tion and Recovery Act (RCRA), operators can still main-
tain sites that pose similar threats as the Chemical Con-
trol facility. The principal witness at the hearing was Har-
old Kauffman, a former official of Duane Marine Corp.
in Perth Amboy, New Jersey, who had turned state's wit-
ness after being charged with complicity in illegal haz-
ardous waste management.
During his testimony, Kauffman recommended that the
U.S. EPA and state regulatory agencies change the focus
of their oversight of the hazardous waste industry. Ac-
cording to Kauffman, the state and federal governments
emphasize the routine filings of applications, and at pres-
ent, nearly anyone can pay a fee and become a licensed
341
-------
342 CASE HISTORIES
hazardous waste hauler or begin the siting process for a
landfill. John Albert, who allegedly assumed ownership
of Chemical Control in 1977 at gunpoint, claims that he
received a DEP hazardous waste hauler's license without
asking for one. Albert's other company, Jersey Sanita-
tion, recently received a renewal of its license, even
though its principals had been convicted of felonies.
Kauffman said that, because no agency currently moni-
tors the background and safety history of applicants for
hauling and disposing operations, incidents such as Chem-
ical Control will continue to take place.
Kauffman testified that he did not believe that a com-
pany with holdings in the garbage and sanitation indus-
tries should be allowed to enter the hazardous waste in-
dustry, for several reasons:
(1) operators of garbage trucks can easily mix quantities
of toxic wastes with the municipal garbage and
(2) garbage haulers usually have established relationships
with municipal dumps, which are less stringently regu-
lated than RCRA hazardous waste landfills.
Consequently, these haulers can pass on significant quan-
tities of hazardous wastes to the municipal dumps. A
subcommittee spokesman noted that Kauffman's com-
ments were based on the activities of a limited number of
small companies operating in metropolitan New York, and
do not necessarily apply to the major firms that deal with
both solid and hazardous waste.
Leonard Tinnan, corporate development director for
the nation's largest waste management firm, Chemical
Waste Management Inc. of Long Beach, California, com-
mented that the large national companies with both solid
and hazardous waste operations had largely succeeded in
maintaining separate corporate structures for their differ-
ent functions. Tinnan said that companies such as Chem-
ical Waste Management and Browning-Ferris Industries
have gradually separated the organization of solid and haz-
ardous waste operations due to the different backgrounds
needed to manage each operation, as well as due to the
public fears about the practices that occur among smaller
companies.
In addition, at the subcommittee hearing, Congressman
Albert Gore (Democrat-Tennessee) said that, when a
hauler has a tacit agreement with a disposal facility, the
hauler and facility owner could potentially collude to falsi-
fy the national RCRA manifests so that, instead of trans-
porting the wastes out of state to another authorized
facility, the hauler could mix the waste with household
garbage and dispose of it at a municipal landfill. Alleg-
edly, through such practices, hazardous wastes which were
supposed to have been incinerated at the Chemical Con-
trol facility ended up at the Kin-Buc municipal landfill
in Edison, N.J., currently one of the nation's largest un-
controlled hazardous waste sites at 220 acres.
Although one U.S. regulatory agency—the Interstate
Commerce Commission (ICC)—is currently attempting to
assert its authority to make background checks on haz-
ardous waste haulers, the ICC is receiving considerable
opposition from the regulated community. Trade associa-
tions, including the Spill Control Association of America
and the Hazardous Waste Services Association, agree that
a federal agency should take steps to curtail midnight
dumping, but believe that ICC regulation of hazardous
waste haulers would duplicate existing EPA and U.S.
Department of Transportation regulations and would too
closely control the entry of legitimate firms into the haz-
ardous waste industry on a geographical or a market-
share basis.
PRE-FIRE CLEANUP
This phase extended from April 1979, when DEP hired
Coastal Services Inc. to perform preliminary cleanup
work, until April 1980, when the Chemical Control stor-
age facility caught fire. When Coastal Services began
work, the facility contained a maximum of 60,000 drums
and laboratory packs of chemicals, including polychlori-
nated biphenyls (PCBs), pesticides, cyanides, compressed
gases, organic peroxides, high explosives including TNT,
picric acid and nitroglycerin, bottled hospital and lab-
oratory specimens and radioactive wastes.
During the one-year pre-fire period, Coastal Services
spent approximately $1.5 million and supervised the re-
moval of about 10,000 drums and laboratory packs con-
taining some of the most dangerous wastes. These wastes
were disposed of at the CECOS International, Inc., chem-
ical landfill near Niagara Falls, New York. The major-
ity of the high explosives and compressed gases were
detonated under controlled conditions at a nearby ord-
nance works.
Although the preliminary cleanup probably helped to
avert a major public health disaster, the fire did take
place and caused major damage. As a result, the ques-
tion remains as to whether the pre-fire cleanup was con-
ducted in an expedient manner and with the proper re-
moval priorities.
Before discussing the contributions of Coastal Services
to the pre-fire effort, it must be noted that several ob-
servers have said that, at the same time as Coastal was
removing wastes, DEP was allowing more wastes to ac-
cumulate at the site. A Congressional staff member said
that a computer analysis of aerial photographs indicated
an influx of several thousand drums after September
1977. According to a former Coastal employee, DEP
realized at the early stages of the pre-fire cleanup that, as
a result of the increasing national attention given to the
site, Chemical Control was virtually guaranteed federal
and state cleanup funds. According to this argument,
DEP decided to consolidate the wastes present at several
smaller waste sites in northern New Jersey and store them
at one central location—Chemical Control—where they
would be cleaned up along with other wastes at the site.
In contrast, Tom Dalton, former vice-president of
Coastal Services, said that the drums which were trans-
ported to the Chemical Control site during the pre-
fire cleanup represented the contents of 14 trailers aban-
doned by the former Chemical Control operators at vari-
ous places throughout the city of Elizabeth. Dalton said
that the contents of these trailers contained a maximum
of 1000 drums, and that reports of massive drum ac-
cumulation resulted from a misreading of the aerial
-------
CASE HISTORIES 343
photographs, which actually showed not an increase in
the total number of drums but rather an increase in the
surface area covered by the drums due to the transfer of
drums stacked six and seven high to piles of drums stacked
two and three high. Nevertheless, the Chemical Control
operators used the allegations of drum accumulation from
outside sources in their trial defense in an attempt to les-
sen their proportionate liability for the waste consumed
in the fire.
The success of the pre-fire cleanup depended largely on
the success of the DEP effort to identify the original gen-
erators of the Chemical Control waste and encourage
them to reclaim their drums. In practice, th is effort
took the form of threatening those generators who could
be identified from the drum labels with legal action in
the massive liability surrounding the Chemical Control
site if they did not come forward and reclaim their
wastes. During the pre-fire phase, DEP identified several
major generators who voluntarily removed almost 7900
drums from the site.
Gregory Heath, technical representative for Peabody
Clean Industry, formerly Coastal Services, said that, dur-
ing the pre^fire cleanup, DEP had refused Coastal's sug-
gestion that it transport the overpacks and drums de-
signated for generator reclamation to an off-site ware-
house for temporary storage. Heath also said that, as a
result, some of the drums which had been bulked and
overpacked for reclamation were consumed in the fire.
While DEP was operating under severe financial con-
straints, as evidenced by the depletion of the $20 million
New Jersey Spill Compensation Fund and the subsequent
decision to temporarily raise the chemical and petroleum
feedstock taxes that supply this fund, several observers
commented that speed rather than economy should have
been the primary objective during the pre-fire phase.
Superfund
The same observers expressed concern that the current
EPA administration is encouraging slow, deliberate and
economical action at waste sites. The Superfund In-
terim Removal Guidance, which was sent in early August
to the federal on-scene coordinators (OSCs), mandates
that the OSCs not initiate any long-term remedial action
at a site, unless it is among the 400 waste sites on the
national priority list and unless a significant effort has
failed to locate a responsible party capable of providing
the necessary cleanup funds.
Under the guidance document, OSCs have authority to
begin so-called "immediate removal actions," or
"planned removal actions," in order to stabilize the site
and prepare it for long-term remedial work. OSCs must,
however, receive approval from EPA headquarters be-
fore authorizing a removal action, and EPA has indi-
cated that it intends to "establish a high threshold for
undertaking such actions."
In addition, the interim Superfund guidance, which
will remain in effect until the final National Contingency
Plan is promulgated sometime in late 1982, contains six
case histories designed to provide the OSCs with models
of typical response actions. At least two of these cases
criticize earlier EPA actions at specific waste sites as
overly ambitious and expensive. According to the interim
guidance, the case history of the Kin-Buc landfill demon-
strates EPA's new policy that removal actions should not
be initiated at waste sites with continuous chronic re-
leases of hazardous wastes, unless those releases directly
threaten human populations.
In practice, EPA policy now prohibits the EPA-funded
use of groundwater or leachate treatment systems unless
alternative approaches, including the "no-response op-
tion," have been considered; in effect, the interim guid-
ance discourages the massive contaminant removal and
groundwater treatment actions that EPA and DEP used
at the Chemical Control site. Furthermore, in the case
history of the PCB storage site at Sharpstown, Maryland,
the new EPA interim guidance suggests that drums should
be removed to an off-site location only when the drums
are actually leaking, and that deteriorating storage con-
tainers should only be staged or overpacked. A senior
official at O.K. Materials Co. of Findlay, Ohio, the
major post-fire contractor, said that this "reduction in
the formerly aggressive posture of EPA" may result in
greatly increased costs when cleanup is finally initiated
and possibly in an increased incidence of catastrophic
events at waste sites where the federal government has
adopted the "wait and see" approach.
Many observers believe that the "wait and see" policy
is inappropriate when evaluating the need for security at
an uncontrolled site. The Elizabeth Fire Department, the
New Jersey DCJ and the Federal Bureau of Investigation
have not yet completed their official arson investigation
into the Chemical Control incident, but since the fire
occurred during the cool springtime, it seems unlikely that
its origins were totally spontaneous. Following the Chem-
ical Control incident, at least one other significant ha-
zardous waste site fire has occurred; on July 10, 1981,
the fire at the General Disposal Co. waste site in Santa
Fe Springs, California, consumed over 10,000 drums of
waste and resulted in cleanup expenditures of up to $2.7
million. The probable cause of that fire was arson; a local
resident allegedly tossed a Molotov cocktail onto the
site with the intention of starting a small fire to drama-
tize the need for security and eventual cleanup at the site.
Fire
The actual emergency and subsequent response were
characterized by a combination of luck and effective
action. Two physical factors decreased the severity of the
fire itself. First, the large number of drums and their
high packing density created an intensely hot fire, over
2000 °F at its center. As a result, the fire itself acted as a
natural incinerator to degrade the hazardous wastes. DEP
analyses during the fire showed high concentration of ben-
zene, up to 10,000 ppm in the smoke plume, substanti-
ating the theory that many of the substituted aromatic
compounds were degraded into simpler molecules during
the fire. Second, repeated wind shifts and a high ceiling
kept the smoke plume from settling over a densely popu-
lated area near the site.
-------
344 CASE HISTORIES
The firefighters, under the supervision of Elizabeth
Fire Director Joseph Sullivan, were particularly effective
at combatting the acute chemical emergency. More than
a year before the fire, Sullivan had instituted a special
site-specific training program in anticipation of problems
at the Chemical Control site. Firefighters learned about
the location of access routes, the topographic conditions
at the site, and the availability of water supplies and
utility outlets.
Sullivan said that the training program was instru-
mental in increasing the speed of the initial response to
the blaze. He said that more than 250 firefighters worked
on the response, which began at 2254 LT on April 21,
and that they successfully contained the fire by 0914 LT
the following morning. The land-based firefighting crews
uses more than 7.2 million gallons of water to control
the fire and two fire boats stationed in the nearby Eliza-
beth River used several million gallons more.
Heath of Peabody Clean Industry said, however, that
DEP should have ordered the firefighters to let the fire
burn and to apply water only to control the spread of
the fire to adjoining properties. He said that this ap-
proach would have created less contaminated water and
degraded more material, resulting in a cheaper and less
dangerous post-fire cleanup. Alden McLellan, DEP as-
sistant commissioner, agreed that the "free burn" was a
viable option, but cautioned that temperatures at the
outer edges of the fire might not have been high enough
to degrade the most persistent toxic compounds.
Acting Elizabeth Fire Chief Charles Swody said in
September 1981 that his department had learned several
lessons as a result of the fire. Swody emphasized the need
for fire companies to maintain an adequate supply of ex-
tra breathing masks because, during the Chemical Control
fire, many off-duty personnel and firefighters from
neighboring communities provided services and needed
protective clothing. Swody also said that, to reduce the
amount of worker exposure, his department might have
placed greater reliance on remotely-operated equipment
and relieved its workers on a more frequent basis. He
claimed, however, that no fire department in the country
has the budget to afford state-of-the-art protective cloth-
ing for each responder at a site incident such as Chemi-
cal Control. One firefighter is suing the Elizabeth fire de-
partment for pulmonary disabilities allegedly contracted
during the response.
POST-FIRE CLEANUP
Introduction
This phase began on April 24, 1980, when DEP con-
tracted with O.H. Materials Co. to secure the site against
the immediate threat of contaminated surface water en-
tering the Elizabeth River. Also involved was the removal
of hazardous waste drums and contaminated soil from
the site, as well as the treatment of contaminated ground-
water.
The post-fire cleanup phase lasted until April 1981,
when the New Jersey Spill Compensation Fund and the
Federal Water Pollution Control Act Section 311 (k)
funds were withdrawn and O.H. Materials removed its
groundwater treatment system. During this one-year
period, more than 350 million pounds of contaminated
debris and soil were removed to the CECOS landfill in
Niagara Falls for disposal, more than 252,000 gallons of
incinerable liquid wastes were taken to the Rollins En-
vironmental Services incinerator in Logan Township, New
Jersey, for destruction, and about' 36,450 crushed empty
drums and 3770 intact drums were removed to secure
landfill disposal.
Three basic questions regarding this phase of the re-
sponse deserve attention: 1) Was the approximately $25
million spent during the year justifiable in terms of the
results obtained? 2) Was the level of personnel safety
during the post-fire response efficient for an action in-
volving such a large amount and complex mixture of
waste? and 3) What new techniques and instrumentation
used at Chemical Control could prove valuable at future
responses?
Cost
During the post-fire cleanup, O.H. Materials used a
team of 71 workers to remove 25,000 barrels, at a total
cost of more than $17 million. In contrast, in a similar
one-year period, Coastal Services used a team of 14 to
remove 10,000 of the most dangerous drums from the site,
at a total cost of $1.5 million.
A DEP official commented that, in the days following
the fire, DEP did not have time to negotiate the costs of
the response because of intense public pressure to initiate
an immediate and complete cleanup action. DEP selected
O.H. Materials as the cleanup contractor not only be-
cause it had the requisite capabilities but also because it
offered the most immediate initial response. The DEP
official also commented that, in developing its response
program, O.H. did not give adequate attention to cost
considerations, encouraged by the relatively limitless fed-
eral funding for this project.
The DEP official further said that, in the future, state
agencies should consider establishing standing contract
relationships with cleanup firms and that such contracts
should have clearly delineated cost parameters. He cau-
tioned, however, that such relationships tend to form
exclusive ties with one company and generate charges of
favoritism when incidents with a high potential for profit
and recognition occur.
In response, Robert Graziano, vice president of O.H.
Materials, said that O.H. Materials took a meticulous,
technology-intensive approach in securing and cleaning
up the Chemical Control site. He said that O.H. com-
pleted more than 90% of the tasks outlined in the original
post-fire contract, even though problems with on-site
PCBs and deeply buried wastes were encountered during
the work. In addition, Graziano said that, prior to the
fire, Coastal Services did not have to deal with the prob-
lems that O.H. Materials faced with debris, damaged
drums and unlabeled piles of drums.
As an example of the legitimate cost differences in vari-
ous incident responses, Graziano contrasted the Chemical
Control incident and response with that at the Seymour,
-------
CASE HISTORIES 345
Indiana waste site. Although Seymour response crews
were faced with a similar number of drums as that at
the Chemical Control site, they were able to finance the
response to date at approximately 20% of the cost, pri-
marily, according to Graziano, because: 1) the drums were
segregated before the state of the cleanup, 2) there was
no fire to cause a deterioration in site and drum condi-
tions and 3) the remote nature of the Seymour site did
not require as extensive community safety precautions.
Worker Safety
The National Institute of Occupational Safety and
Health (NIOSH) has not yet completed its investigation
into the various cases of worker exposure to hazardous
substances during the post-fire cleanup at Chemical
Control. James Melius, chief of the NIOSH Hazard
Evaluation Section, reported that, in July 1981, NIOSH
completed a follow-up medical survey among the ap-
proximately 390 firefighters and other response workers
at the initial incident. Of the 224 workers who had re-
ported no adverse pulmonary symptoms, such as wheez-
ing, excessive phlegm, and shortness of breath, immedi-
ately after the fire, 110 reported the onset of one or more
of these symptoms in the following year. Melius said this
incidence of delayed effects was serious enough that
NIOSH decided to conduct a detailed medical study of
the 390 workers to determine the extent of incident-
related pulmonary problems.
During the summer months, NIOSH had difficulty de-
termining whether certain transient worker illnesses were
due to cyanide poisoning, as suggested by the odors at
the site, or due to heat stress. Three cases of worker in-
capacitation occurred during the early morning hours,
however, when heat-stress would not be expected to be a
problem. Although NIOSH photoionization tests were
negative, NIOSH noted that the vapors could have been
acute for transient periods only and may not have been
present during the test period.
NIOSH did make several recommendations about work
periods at the Chemical Control response and at future
actions, including: 1) each thirty-minute work period in
full breathing apparatus should be followed by a thirty-
minute rest period, 2) workers should be acclimatized
to possible heat-stress problems upon entering the site
and 3) the dispensing of diuretic drinks, including coffee,
colas and alcohol, should be discouraged because of the
potential for dehydration.
Melius also reported that, as a result of the cases of
heat stress among the workers, NIOSH developed a
system for continuously monitoring both the environ-
mental conditions and the worker's heat stress in the
field. During August 1981, NIOSH tested the system at a
non-chemical site cleanup in Arizona, monitoring the
axillary temperature and heart rate of each worker.
A Coastal official said that Coastal Services walked
off the Chemical Control site because of the insistence
of a DEP official that Coastal workers wear full self-
contained breathing apparatus at the early stages of clean-
up. This observer disagreed because, at that point in the
cleanup, there was a 15-knot wind and a light rain falling,
conditions he said created little danger of acute concen-
trations of fumes. He said that the SCBA gear severely
impaired the ability of Coastal workers to navigate
through the rubble and wet drums during the early stages,
and that such limited mobility actually made conditions
less safe than if the workers had been wearing respira-
tors alone.
New Technology
Several new protocols and technologies, with potential
usefulness at other responses, have emerged from the
Chemical Control incident. These new techniques, which
were developed by O.H. Materials, include the following:
(1) A special barrel grappler that remotely handles
drums and moves them to a staging area, thereby
replacing the standard method of overpacking each
drum individually. The grappler consists of a
Caterpillar backhoe with a specially-designed arm
that can rotate through 360°. IT Corp. used this
device at the Santa Fe Springs post-fire cleanup.
(2) A proprietary compatibility sampling protocol,
which uses a mobile field laboratory to test both
for the toxicity and chemical compatibility of the
liquids contained in the drums. This sampling pro-
tocol allowed O.H. to identify those drums which
could be bulked to reduce the transportation volume
of the wastes.
(3) A specially-designed compatibility chamber for
liquid wastes, consisting of a 12,000 gal open-top
vessel fitted with a series of thermocouples. After
adding new materials to the chamber, O.H. per-
sonnel could precisely monitor the heat of reaction
from a safe distance to determine if the addition of
the materials should be slowed due to incompati-
bility.
(4) A health and safety protocol for moving personnel
on and off the site, consisting of several clearly
delineated "hot," "warm," and "cold" zones.
(5) A proprietary scheme for identifying and repackag-
ing the contents of the "lab packs." O.H. con-
structed a portable building on-site for this purpose
and retained a professional chemist to perform the
testing and repackaging of the laboratory chemi-
cals.
(6) The first field use of the O.H. Underground Re-
covery and Treatment System (URTS) at a ha-
zardous waste site. The URTS uses a pneumatic
recovery system to create a zone of depression,
thereby accelerating the movement of contaminated
leachate toward the recovery wells. O.H. was able
to use a variety of techniques in conjunction with
the URTS, including: 1) the high-temperature
steam-stripping of solvents in the contaminated
leachate, 2) a clarifier unit to precipitate out cer-
tain insoluble constituents from the groundwater
and 3) an activated carbon treatment unit for final
"polishing" of the leachate in preparation for dis-
charing it back onto the site.
In addition, the consultants at Roy F. Weston, Inc. in
West Chester, Pennsylvania, took advantage of the Chem-
-------
346 CASE HISTORIES
ical Control incident to make the first successful use of
ground penetrating radar (GPR) technology to locate
hazardous waste targets in saltwater. The U.S. Coast
Guard had expressed concern that the force of the initial
Chemical Control explosion had thrown chemical drums
into the Elizabeth River and that the water spray during
the subsequent firefighting operation had caused sections
of the river bank containing the drums to slump into the
river. Although previous GPR surveys in saltwater had
failed because the uneven salinity of the water attenu-
ated the return signal to an indeterminate degree, the
Elizabeth River had a relatively constant salinity and the
Weston researchers were able to use a combination of GPR
and fathometry readings to determine the exact degree of
attenuation at each point and calibrate the results ac-
cordingly.
Ongoing Action
After a cleanup program lasting two years and the ex-
penditure of more than $27 million, the Chemical Con-
trol site still remains a candidate for Superfund remedial
action. According to the former DEP on-scene coordi-
nator, the site has received 56 "points" on the Mitre
ranking model, placing it about 20th among the nation's
problem sites. DEP has ranked Chemical Control 13th
among the 300 New Jersey sites that it wants EPA to con-
sider for Superfund remedial actions. In early September
1981, the EPA Office of Emergency and Remedial Re-
sponse received a detailed request from DEP outlining
the remaining work necessary to clean up the Chemical
Control site, which DEP said still presented an uncon-
trolled leachate threat.
The DEP estimates that an additional $700,000 will be
needed to perform the following remedial actions:
(1) Removal and disposal of 1500 cubic yards of con-
taminated debris, including contaminated soil, ply-
wood and concrete
(2) Removal of sludge from the groundwater clarifier,
contaminated spent carbon from the URTS and 75
crushed drums
(3) Removal and detonation of 200 pressurized gas
cylinders of "uncertain" integrity
(4) Dismantling of a 10,000-gallon underground steel
tank found during the cleanup
(5) Removal and disposal of 60 drums containing PCB
wastes concentrated from settling tanks installed
during the cleanup. Some of these drums contain
up to 977,000 ppm PCBs, although the majority of
them contain less than 1000 ppm
(6) Removal and disposal of small quantities of miscel-
laneous material, including one drum of contami-
nated clothing, two drums of arsenic pentoxide, one
can of potassium cyanide and a 2000-gallon tank
truck left on-site by the former Chemical Control
operators.
(7) Additionally, the final URTS treatment of the con-
taminated groundwater had not been completed
when the funds were withdrawn.
At least four civil and criminal cases were filed as a re-
sult of the Chemical Control incident. In one recent ac-
tion, a federal grand jury in Newark, New Jersey, sen-
tenced John Albert and Eugene Cordon, the two men who
allegedly took over operation of Chemical Control in 1577,
to five and three year jail terms respectively for commit-
ting two counts each of mail fraud in connection with the
Chemical Control operation. These sentences represent the
first successful federal felony prosecutions against op-
erators of hazardous waste storage facilities. Since the il-
legal storage took place before the November 19, 1980
effective date of RCRA, the U.S. attorneys decided to
invoke the federal fraud laws.
The prosecutors based their indictment on evidence
gathered from the investigation of the acceptance by
Chemical Control of a single cylinder of highly unstable
perchloroisobutylene from a nearby generator, since the
unusual nature of this small shipment made the tracing
of bank checks and acceptance letters easy and verifi-
able. Albert and Conlon were charged with defrauding
DEP by promising to treat various wastes at the Chemi-
cal Control incinerator, when in fact the wastes were
transported to another unlicensed dump site. They were
also charged with defrauding various generators by prom-
ising to properly dispose of their wastes. The following
three suits are still pending:
(1) A criminal trial to bring nuisance charges against
Albert and Conlon for maintaining a hazardous
situation at Chemical Control. Since New Jersey
recently amended its criminal code in 1980 to in-
clude hazardous waste violations as felonies, the
two Chemical Control operators will only be
charged with misdemeanors.
(2) A civil suit filed by DEP to recover nearly $24 mil-
lion in cleanup costs from the Chemical Control
operators. This suit also seeks penalties under the
New Jersey Solid Waste Act amounting to more
than $25,000 per day; DCJ was pessimistic about
the ability of the insolvent Chemical Control
operators to provide even a small fraction of the
sum needed for significant reimbursement. The
federal government may also sue to recover its
Section 311 (k) fund expenditures.
(3) A suit against DEP filed by the Loizeaux Builders
Supply Co., owners of the property adjacent to the
Chemical Control facility. Loizeaux alleges that
DEP forced him out of business by taking over his
property for use as a drum staging area shortly after
the fire.
Perhaps most importantly, the Chemical Control inci-
dent contributed to public awareness of the need for en-
vironmentally sound disposal facilities. It also had a major
effect on the passage of New Jersey's new "Major
Hazardous Waste Facilities Siting Act," signed on Sep-
tember 10, 1981 by Governor Brendan Byrne. The Act
contains the following provisions encouraged by the
Chemical Control incident:
(1) Siting approval shall not be granted to an applicant
if any party to the application has been convicted
of a crime relating to improper hazardous waste
management during the previous ten years
-------
CASE HISTORIES 347
(2) All new major facilities shall be totally or partially
constructed above ground and be designed to allow
the extraction of 99.9% of all stored or disposed
waste
(3) Local inspectors who supply information on im-
proper facility management shall be given one-half
of any penalty collected in the ensuing litigation
(4) A facility must undergo a formal siting review pro-
cess if a capacity expansion of more than 50%
is planned.
Conclusion
Despite the controversies surrounding the regulation of
Chemical Control, the response to the chronic site prob-
lem and the costs and approaches involved in the post-
fire response, the final cleanup of the uncontrolled site
will represent the successful conclusion of a hazardous
waste project whose difficulty may remain unsurpassed
in future years. The cooperation between the cleanup con-
tractors and the federal and state regulators proves that,
given sufficient time, money and good fortune, the mis-
takes of the unregulated past can be efficiently and
safely corrected.
In the future, researchers, engineers and government
officials need to work closely to develop protocols, as
well as checks and balances, that can ensure positive re-
sults at site actions where these factors are not so favor-
able. As the Superfund cleanup program begins, the
private sector and government participants in uncontrolled
site actions should have ample opportunity to perfect the
knowledge gained from the Chemical Control incident.
REFERENCES
The authors have relied extensively on information that
appeared in the Hazardous Materials Intelligence Re-
port, an international weekly newsletter focusing on the
safe management of hazardous wastes and hazardous ma-
terials and published by World Information Systems,
Harvard Square Station, P.O. Box 535, Cambridge,
Massachusetts 02238. The Hazardous Materials Intelli-
gence Report began reporting on the Chemical Control
incident in its Vol. I, No. 5 issue on 13 June 1980 and,
since then, has provided regular coverage of the cleanup
operation, legal proceedings and other issues related to
the event.
-------
A FAST TRACK APPROACH TO MANAGEMENT AND
IMPACT ASSESSMENT (PART i)
GREGORY A. VANDERLAAN
U.S. Environmental Protection Agency
Region V
Chicago, Illinois
INTRODUCTION
The approach discussed here was developed during a
Federally funded emergency cleanup action under author-
ity of Section 311 of the Clean Water Act. The subject of
the effort and cause for action, the Seymour Recycling
facility, is quintessential of the many abandoned, haz-
ardous waste sites found throughout the country. It is a
large site covering an area of 13 acres with a diverse collec-
tion of hazardous compounds stored in a variety of con-
tainers.
BACKGROUND
Located in the Indiana community of Seymour, approx-
imately ninety miles south of Indianapolis, the Seymour
Recycling Center was established in 1971 solely to re-
cover methylene chloride used for industrial operations
at the Seymour Manufacturing Company. The recycling
center soon expanded its operation to include the reclama-
tion of other industrial chemicals.
In 1975, the Seymour Recycling Center incorporated
separately from the Seymour Manufacturing Company.
In 1978, when the Recycling Center was sold, owner es-
timates of on-site waste storage included approximately
42,000 55-gallon drums, 100 bulk storage vessels contain-
ing approximately 676,000 gallons of waste and seven
20,000 gallon rail cars.
Shortly after the change in owners, serious environ-
mental problems began to develop. The incinerator was
operating poorly and without an emission control sys-
tem. Micrometeorological conditions often resulted in
fumigation events affecting nearby residents. These occur-
rences prompted the affected people to complain to local
and State pollution control officials, demanding that
they intervene in the operation and force a cleanup to
attain compliance with the appropriate regulations.
Representatives of the recycling facility signed a con-
sent decree in June, 1978, with the State of Indiana re-
quiring them to shut down and dismantle the incinera-
tor and to reduce their drum inventory on a scheduled
basis. Fourteen months after the consent decree was
signed, however, the recycling center's drum inventory
actually increased by approximately 3,400. This was con-
firmed through aerial photographs evaluated by U.S.
EPA's Environmental Photographic Interpretation Center
in Warrenton, Virginia. According to recycling center
officials, the terms of the consent decree were not being
met for several reasons including labor problems, lack of
availability of secure landfills and lack of State approval to
place drums containing solids in non-secure landfills.
Several near catastrophic events developed at the re-
cycling center after agreement and implementation of the
consent decree. A major fire occurred in October, 1978,
when a vehicle struck a drum containing lithium aluminum
hydroxide in hexane, igniting it and several adjacent
drums. Also, several major releases occurred on-site in-
cluding an alkanol amine spill, a chromic/nitric acid spill
and an alleged hexachlorocyclopentadiene spill. Finally,
in the fall, 1979, after repeated failures on the part of
ownership to meet the consent decree schedule, the State
of Indiana brought the site to our attention for en-
forcement action.
AGENCY EFFORTS
The National Enforcement Investigation Center com-
pleted a field investigation01 designed primarily to identify
the materials on site and to establish the potential for any
toxic materials to adversely affect the local population,
Liquids and soils were sampled at locations within the re-
cycling center, the drainage ditch north of the site, nearby
residential and industrial wells and the city sanitary sewer
downstream from the recycling center. Analyses of the
samples discovered the presence of a number of com-
pounds in the area soils with documented adverse human
health effects (Table I.).
Table I.
Carcinogenic Compounds Found in Area Soils
Compound
Benzene
Chloroform
Methylene chloride
Toluene
Concentration 0*g/l)
5
20
200
2000
The soil in the area was found to be very porous, po-
tentially allowing infiltration of contaminants into the
local groundwater system. Although there were no data
documenting groundwater pollution, EPA believed the
348
-------
CASE HISTORIES 349
presence of such large quantities of hazardous substances
on the site and the nature of the soils in the area virtually
guaranteed shallow aquifer contamination.
While this initial field investigation was intended to de-
velop information on which to base litigation proceedings
under the Resource Conservation and Recovery Act of
1976, its findings made clear the need to seek other more
immediate alternatives to achieve a site cleanup. By the
time this investigation was completed in late 1979, the re-
cycling center's management was operating in total disre-
gard of the consent decree and had allowed the drum in-
ventory to approach 60,000.
As litigation proceedings went further and cases were
filed, the recycling center owners officially declared bank-
ruptcy in early 1980. Without a responsible party to fi-
nance cleanup, government action was now the only means
by which the site could be stabilized.
In March, 1980, an explosion occurred at the site as a
result of waste materials reacting after escaping from their
decaying containers. The State of Indiana requested assis-
tance from EPA Region V. A Federal On-Scene Coordina-
tor (OSC) from the Environmental Emergency Section
traveled to the site to determine if the explosion and leak-
ing containers constituted a discharge or threat of dis-
charge to navigable waters. It was necessary to make this
determination to establish Federal authority and gain ac-
cess to emergency cleanup funds under Section 311 of the
Clean Water Act.
EPA Responds
Shortly after arrival at the recycling center, the OSC
made an official determination that the abandoned site
represented a threat to public health due to potential sur-
face water contamination from site run-off and probably
ground water contamination from infiltration. The OSC
convened a Regional Response Team (RRT) meeting. All
members for the team agreed that a substantial threat of
run off to waters of the U.S. existed as long as the con-
ditions of the site were left unchanged. The U.S. Coast
Guard representative made it clear to the response team
membership that, as the 311 fund administrator, removal
of materials from the site or any activities not directly
associated with containment would be considered inappro-
priate and not eligible for funding.
The OSC hired a contractor under Coast Guard emer-
gency procurement authority. Contractor activities for the
next thirty days involved drum segregation and staging,
site grading and paving, earth work designed to direct
surface water runoff around the site and placement of a
runoff lagoon and treatment system to collect runoff from
the site itself and treat it through a carbon adsorption
system.
311 Action
At the conclusion of this effort which cost $866,000 and
after evaluation of the work completed, EPA realized that
an alternative approach to contractor procurement was
needed to insure a more cost effective use of Federal
funds. Placed in the perspective of implementing a site
cleanup, the funds expended did not produce significant
results. It was evident that extension of traditional 311 pro-
curement practices to address uncontrolled waste site
cleanup would mean large expenditures with little cost con-
trol capability. After a thorough review of the 311 effort,
EPA felt it would be possible to achieve a greater level of
effort at less cost by:
(1) removing the constraints on use of the 311 fund for
removal of materials from the site and
(2) employing an emergency procurement approach in-
corporating limited competition.
At this time, EPA's policy was to use Section 311 au-
thority wherever possible at hazardous waste sites
threatening surface waters and to undertake whatever was
necessary in effecting cleanup to mitigate the threat or dis-
charge. Coast Guard's 311 policy differed. Their main
concern was to conserve the 311 fund. They believed the
best way to do that was to limit emergency activities at
hazardous sites to containment only. The EPA and Coast
Guard collective experience at Seymour Recycling indi-
cated that the two policies on use of the 311 fund were
not only inconsistent but often in diametrical opposition.
Major Cleanup Needed
In July 1981, approximately three months after con-
clusion of the initial response, a special congressional ap-
propriation for Section 311 was enacted. An On-Scene Co-
ordinator made a site visit to assess the current conditions
and define the activity phases necessary to provide greater
site stabilization.
The OSC convened a RRT meeting and put forward a
phased plan for complete site cleanup beginning with the
bulk tanks, progressing through removal of drums, in-
vestigations for buried materials, removal of the contam-
inated soil mass and ground water recovery and treat-
ment. The members of the RRT believed that the relation-
ship between the site, groundwater and nearby surface
water could only be established by completing a ground-
water study. The Regional Response Team recommended
that while the groundwater study was underway, a per-
manent dike and security fence should be constructed.
Groundwater Monitoring
The OSC felt that the rapidly deteriorating conditions
at the site would not allow time for implementation of a
standard piezometric and analytic effort and that the
relationship would need to be established in a two to four
week period. A quick turnaround approach for identify-
ing the presence of organically contaminated groundwater
plumes entering nearby surface waters did exist and had
been applied successfully in determining rural waste water
treatment needs.(2) The OSC felt this approach could be
applied at the Seymour site, yielding results of sufficient
quality and clearly defining the relationship between the
site itself and nearby surface water (Table II). Details of
the groundwater study, the instrumentation used and re-
sults are discussed in detail in Part II of this paper.
-------
350 CASE HISTORIES
The results of the groundwater effort were available
within ten days. They confirmed that the~shallow ground-
water aquifer was contaminated and contaminants were
entering the nearby surface water system. The bulk stor-
age vessels and drums, all in a general state of disrepair
and decay, were in fact a source of an on-going discharge
to the East Fork of the White River.
Table n.
Information Derived from Groundwater Study
Rate and direction of groundwater flow
Relative level of groundwater contamination
Source of groundwater contamination
Positions of groundwater inflow into nearby stream
The instruments utilized during the study included
groundwater flowmeters and ultraviolet flourescence.
CONTRACTOR SELECTION
Having established the site as a source of contamina-
tion, the OSC requested funds for the removal of ma-
terials stored in the bulk vessels. Both the OSC and the
USCG Contracting Officer agreed that negotiated pro-
curement would provide the best insurance for obtain-
ing the "cost effective" approach. They also realized,
however, that the normal procurement process covers a
period of time far beyond that considered responsive to
this situation.
A "fast track" procurement process needed to be em-
ployed in order to begin removal activities as soon as
possible. The system ultimately developed was based on a
brief scope of work and evaluation criteria ranked in order
of priority (Table III).
Table m.
Fast Track Negotiated Procurement
Distribution of solicitations
Review and ranking
Cost negotiations
Selection
The evaluation criteria were distributed as part of
USCG's solicitation package. All contractor proposals re-
ceived were reviewed by a Technical Evaluation Team
consisting of representatives from EPA, the Indiana State
Board of Health and USGC. The proposals were ranked
based on point totals generated through review of the
information provided by the contractors in response to
the Coast Guard solicitation (Table IV).
Table IV.
Solicitation Requested Information.
Overall company approach
Disposal techniques
Safety precautions
Transportation specifics
Equipment to accomplish task
Pricing
After ranking, the proposals were grouped into three
divisions:
(1) acceptable,
(2) unacceptable but could be made acceptable with slight
change and
(3) unacceptable.
Negotiations were begun with those contractors grouped
in the second division. Technical and pricing modifica-
tions to their proposals were requested. These contrac-
tors were extremely responsive and modified their pro-
posals for resubmission in approximately one week. These
proposals were evaluated again (Table V) and a selection
was made.
Table V.
Major Criteria Contractor Selection
Most Environmentally Sound Approach
Most Reasonable Site Safety Program
Most Cost Effective Solution
This entire procurement process required only four and
one-half weeks to complete (Table VI) and realized a sig-
nificant savings of Federal resources. Final cost esti-
mates for acceptable proposals ranged from $350,000 to
$4,000,000. After being evaluated on technical and safety
terms, cost effectiveness became the pivotal factor. Re-
moval of the materials in the bulk storage vessels has been
completed. Costs for that effort went far beyond the con-
tractors estimate but remained less than the next most
cost effective alternative.
CONCLUSION
This procurement approach has also been used for
two other efforts at the Seymour Recycling Center. These
activities involved a detailed hydrogeological investigation
and the other involved removal of all liquids in 55-galIon
drums on the site. Cost estimates before negotiations were
$64,000 to $400,000 and $800,000 to $13,000,000 re-
spectively. The groundwater effort has been authorized
but drum cleanup has not because of the limited 311 fund
situation.
Table VI.
Time Frame for Negotiated Procurement
Element
Solicitation Preparation
Contractor Proposal
Proposal Evaluation
Proposed Revisions
Negotiations and Final Evaluation
Total Time to Contract Award
Time Span
(weeks)
1
2
0.5
1
1
4.5
-------
CASE HISTORIES 351
CONCLUSION
EPA Region V believes that the approaches used dur-
ing the 311 action at the Seymour recycling center pro-
vided valuable information on which to base key de-
cisions in a time frame consistent with the true meaning
of emergency response. Defining the role of groundwater
at an uncontrolled waste site can provide the OSC with in-
valuable information on which to base an immediate or
planned removal decision. The benefits of the negotiated
procurement process far outweighed the necessary re-
source commitment to effect them in terms of potential
gross savings versus the staff time needed to complete in-
tensive review and quick turnaround.
After experiencing the success of using these ap-
proaches, EPA Region V was particularly pleased to see
the development of a planned removal scenario as part
of the Superfund effort and are convinced that this ap-
proach, if managed in a timely manner, can achieve sig-
nificant public health protection quickly and cost effec-
tively.
REFERENCES
1. Forba, R.W., "Field Investigations of Hazardous
Waste Site Seymour Recycling Company" U.S.
EPA, Office of Enforcement, EPA-330/2-80-010,
March, 1980.
2. "Alternative Rural Wastewater Treatment Systems-
Case Study Number Two" U.S. EPA, Region 5,
draft Environmental Impact Statement, June, 1979.
A FAST TRACK APPROACH TO MANAGEMENT
AND IMPACT ASSESSMENT (PART ID
WILLIAM B. KERFOOT, Ph.D.
K-V Associates, Inc.
Falmouth, Massachusetts
INTRODUCTION
On August 15, 1980, K-V Associates, Inc. was con-
tracted by the U.S. Environmental Protection Agency
Region V under a Section 311 action for emergency service
to determine the direction of groundwater flow and loca-
tion of any subsurface plumes of material associated with
chemical waste materials improperly disposed of at the
Seymour Recycling Center. The analytical services in-
cluded the following tasks:
•Determine the direction of groundwater flow from the
Seymour Recycling site by obtaining flow data with a
Model 10 Dowser groundwater flow meter from roughly
20 locations of shallow augered excavations
•With the aid of existing PVC monitoring wells, deter-
mine the vertical profile of flow, if possible
•Clean, develop, and pump monitoring wells surrounding
the site
•Obtain at least 20 samples of groundwater in the direc-
tion of prevailing flow and along the stream shoreline
with a well-point sampler
•Analyze water samples in the field for conductance and
fluorescent aromatic hydrocarbon content
•Map the direction of groundwater flow and positions of
leachate from the site, indicating locations of inflow into
stream, if observed.
The Seymour Recycling Center is located on the per-
iphery of the Freeman Field Industrial Park southwest of
Seymour, Indiana. To the north of the Site a stable stream
flows directly east-west, then bends northward to enter
Heddy Run which discharges into the east fork of the
White River.
GROUNDWATER FLOW METER MEASUREMENT
Although several areas of ponded water were observed
at the surface near the site, subsurface profiling showed
the water to be perched on compacted clay-loam deposits.
Augering through the soil would frequently encounter
water at 8 to 9 foot depths below the surface. The inflows
occurred through limited veins of sand penetrating the
sandy-clay deposits. The water was under some pressure
and would enter rapidly, fluidizing the sandy loam de-
posits until eventually stabilizing at a depth of 6 to 7 ft be-
low grade, depending upon grade elevation.
Initially, a 3.5 in. diameter probe with the ground-
water flow meter (Model 10 Dowser, K-V Associates,
Inc.) was inserted into the sandy deposit at the base of
the hole. Attempts at measurement indicated no apparent
flow. Samples of the soil mixture were removed from the
hole and placed in a 4 in. diameter soil flume to measure
the capacity of the soil for flow. No measurable flow could
be detected through the mixture with a two-inch water
head even though the soil was a sandy consistency. A suf-
ficient content of clay or silt existed to effectively seal off
flow through the material.
-------
352 CASE HISTORIES
1000 2000 3000 4000 5000 6000 7000 FEET
CONTOUR INTERVAL 10 FEET
DATUM IS MEAN SEA LEVEL
OUAOIUNCU IOCAUM
Figure 1.
Location of Seymour Recycling Site Showing Stream Areas
and Discharge to the East Fork of White River
depth
O
4 drain
;j:::-;i
Grounc
Flowmet
aga pipa 1O
organic turtaca
aandy loam
sand • clay
*" ~- — — Piciomatr.c Haad— """
madium aand ' j /-\".;l
Corrfmad Aquitar .- . . .
water Well
er Probe Sam
tor meaiuramant
*. -'_• ..•-"'" ,
Point
pier
Figure 2.
Typical Soil Profile Adjacent to the Site
9
8-
7
«
« 6
a >•
i! -
u. g. 4
2-
1 •
0
Down the Well Probe
Seymour, Indiene
August 198O
10
Readout
Fin* slnd
ft. O.a void frwftofl
•lop* • O.37B
15
20
Figures.
Model 10 Dowser Calibration Curve for Water Bearing Strata
-------
CASE HISTORIES 353
Svymour. Indian
K-V Allocates.
Filmoulh, Man
Auguit 19BO
i • Sit* Plan
Inc.
•chui«tt*
Figure 4.
Direction and Rate of Groundwater Flow Near Seymour Site
To avoid blockage of measurement by intrusion of the
fine material, the hole was cased down to the porous
strata. The hole was augered to 10 ft and a 10 foot sec-
tion of 4 in. diameter drainage pipe installed. The gelled
sand which had fluidized into the hole was bucketed out,
allowing only washed sand from the aquifer to intrude up
the interior of the pipe. The 2 in. down-the-well probe was
then inserted into the sandy material at the base of the
casing, the casing pulled back above the probe to avoid
interference with capillary flow, and the probe oriented to-
wards north before measurement. After initial measure-
ment of flow, the probe was then rotated 180° and the
measurement repeated to provide a check on the first de-
termination. The results of field measurements taken from
cased holes surrounding the Seymour site are given in
Figure 4.
The Model 10 Dowser flow meter was calibrated for
flow rate against the Seymour sandy aquifer material ob-
tained from cased holes I and J placed in a 4 in. cylindri-
cal flow chamber. Using a peristaltic pump, water of
known volume was circulated through the chamber. With
an independent measurement of porosity of the soil, the
rate of movement in feet/day could be calculated. A cal-
ibration curve for the sandy water-bearing strata is plotted
in Figure 3.
GROUNDWATER SAMPLING
Samples of groundwater were obtained with a Model 12
miniature well-point sampler (K-V Associates, Inc.). Two
procedures were followed. With wells AA through J, the
well-point sampler was inserted into hand-augered, cased
holes (with 4 in. PVC draining pipe) and thrust through
the sandy bottom to 2 ft. below the casing.
With uncased holes or cornfield sampling, the well-
point was driven directly through the soil to approx-
imately four feet below standing water level (piezometric
head). A high ratio of striking force to crossectional area
of the shaft (5/8 in. diameter) enabled the well-point to
penetrate the sandy or clayey subsoil. Water was then
withdrawn with the sampler into a silt trap, washed out,
and refilled. A sample of the second pumping was taken
for analysis by a field fluorometer and later scanning by
a Perkin-Elmer Model 204 fluorescence spectrophoto-
meter.
The conductance of groundwater and surface water
samples was determined with a Model 1484-10 conductiv-
ity meter (Horizon Ecology Co.). Samples were obtained
for cross-calibration in the laboratory.
SURFACE WATER TRANSECT
On August 20, surface water samples were taken along
the stream course from west to east, progressing up-
stream. These samples were analyzed immediately for
fluorescent aromatic content. The stream flows through a
2 to 3 ft wide channel as it meanders through a 30 ft wide
man-made ditch between the cornfield and pastureland.
Grass covers the sides of the stream bank in unshaded
areas. The sandy stream bed is soft in numerous loca-
tions, apparently caused by upwelling waters from ground-
water recharge.
Two increases in fluorescent organics were observed,
one near Sample 2 at the opening of the storm drainage
ditch from the Seymour site and in the vicinity of sur-
face water Sample 7 taken near groundwater Sample R12.
FLUORESCENT ANALYSIS OF WATER SAMPLES
Fluorescence spectroscopy is well-known as an analyti-
cal tool for organic analysis because of its exceptional
sensitivity and selectivity.(1) The U.S. Coast Guard Re-
search and Development Center recently conducted a sur-
vey of the luminescence of hazardous materials/2' A large
number of hazardous organic compounds were identified
as possessing useful fluorescence. Of 113 materials sup-
plied to the Coast Guard study, 96 compounds were fluor-
escent at room temperature, although some admittedly
weak. The stronger fluorescent compounds include the
benzene and naphthalene derivatives common to petrol-
eum products, numerous pesticides and herbicides, com-
mon solvents such as toluene and xylene, dye products
such as aniline, and the polyaromatic hydrocarbons
(PAH's) such as anthacene, chrysene, and pyrene.
Fluorescence has recently found increasing use as a tool
in the analysis of complex mixtures of organics. Frank(3)
-------
354 CASE HISTORIES
•£ 300
a
u
x
UJ
Room Temperature Fluorescence
of 95 Toxic and Hazardous Materials
•- •<
250 300 350 400
Emission Wavelength (nm)
Figures.
Excitation-Emission Plotting of Fluorescence of 95 Toxic and
Hazardous Materials. Region of Sensitivity of Field Fluorometer
Indicated by Box
Room Temperature Fluorescence
of 95 Toxic and Hazardous Materials
250 30O 350 400
Emission Wavelength (nm)
Figure 6.
Zones of Sensitivity of Fluorometer Scans of Water Samples
used excitation of samples at 290 run and quantification of
emission at 329 nm as a method of characterizing crank-
case oils in situations not amenable to analysis by gas
chromatography. Quantification of number 4 oil directly
in water has been studied by Frank*4' for analysis of oil
contamination. More recently synchronous scanning,
accomplished by linking the excitation and emission mon-
ochromators together, has gained in popularity because of
the elimination of scatter interferences.(5>
Field water samples were analyzed by excitation at 254
nm and emission from 280 nm to 400 nm. Upon return to
the laboratory, the same samples were analyzed by three
different scanning methods: excitation at 290 nm and emis-
sion scan from 310 to 450 nm (a procedure similar to
Frank01 for crankcase oil), and synchronous scans (30 nm
and 50 nm separation). Results from the four methods are
superimposed over a two-dimensional plotting of the peak
locations of fluorescence of hazardous substances deter-
mined previously by the Coast Guard in Figure 6.
Results
The organic analyses yielded the following informa-
tion:
(1) Between cased holes BB and D a substantial con-
centration of fluorescent aromatic organics occurred.
All scanning procedures revealed a similar pattern
with the highest concentrations coinciding with holes
BB and D. The variety of scans also revealed two
fluorescence peaks dominating the elevated regions:
Wave Length (nm
Excitation/Emission
290/324
335/365
Class of Organic Compounds
phthalic acids, naphthalene
anthracenes, plasticizers, light oils
Previous gas chromatographic analyses of well water
samples had identified naphthalene, diethylphalate,
ethylbenzene, toluene and tetrachloroethylene exterior
to the storage site.(6)
(2) Similarly, groundwater samples downflow of the site
along a transect through the cornfield Rl through R17
revealed a rise above background from Samples R7
through R16.
(3) Surface water samples from the east-west creek indi-
cated inflow of fluorescent aromatic organics in the vi-
cinity of the drainage ditch confluence. The regions of
infiltration were strongest in the vicinity of surface
Samples 2 and 7 coinciding with the drainage ditch
and the central portion of the subsurface plume.
(4) During a heavy thunderstorm, a substantial inflow of
fluorescent aromatic organics was observed issuing
from the drainage ditch to the east-west creek. The in-
flow was readily observable as a discernible pulse en-
tering Heddy Run.
Based on these results, one can conclude that aromatic
organics from the Seymour site are entering the east-west
stream-Heddy Run confluence through two pathways:
(1) storm runoff through the western drainage ditch and
(2) groundwater infiltration into the east-west stream and
from a large subsurface plume containing materials
originating from a primary source at the site (as seen
in holes AA through D) and a secondary source as in-
filtration into the aquifer of storm drainage from con-
taminants in the western drainage ditch.
Storm Runoff
On August 19, a violent thunderstorm occurred at 4:30
p.m., forcing abandonment of groundwater sampling. In-
stead, the leachate detector was moved to the confluence
of the east-west stream and Heddy Run. Samples were re-
moved from the stream from 4:50 p.m. to 6:00 p.m. at 15
min. intervals. At 5:30 a large surge of fluorescent aro-
matic compounds was observed at the confluence. At 6:15
samples were obtained from the outflow of the west drain-
-------
CASE HISTORIES 355
5678 9 ^10 11 12
\ /' ,'
drainage ditch /.\
v :
Seymour Recycling
Seymour. Indiana • Site Plan
K-V Associates, Inc.
Falmouth, Massachusetts
August 1980
500
Figure?.
Sampling Locations and Position of Observed Plume
-------
356 CASE HISTORIES
Surface Water Samples - E-W Stream
Field Fluorometer Scanning Analysis
Relative Fluorescence
50 nm synchronous scan
leachate from storm drainage ditch
1 23 4 5 6 7 8 9 10 11
Figures.
Fluorescent Aromatic Concentrations Observed in Surface
Water of East-West Stream
Well-point Samples - Cornfield Transect
100-
90-
80-
70-
60-
S50-
U
• 40-
o
3
I 30-
20-
10-
0
I
Field Fluorometer Scanning Analysis
..
17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
R aeries
Figure 9.
Concentration of Fluorescent Aromatic Found in Groundwater
from Cornfield Transect Adjacent to Stream
age ditch from the Seymour site and immediately upstream
of the outflow along the east-west stream. Analysis of the
water samples confirmed that the bulk of the discharge was
issuing from the vicinity of the drainage ditch.
STRESSED VEGETATION
On edges of the cornfields, trees at various heights oc-
curred and could be seen to be under vegetative stress. The
position of various dead and stressed trees were plotted.
Most were in the vicinity of the western drainage ditch,
although a few trees occurred upstream along the east-
west stream. The corn plants in the region would not be
affected by groundwater contamination since their root
systems are shallow and rest in the sandy loam above the
aquifer. However, taller trees ( >20 ft) have roots penetrat-
ing into the lower strata which may contain the trans-
ported organics.
0
u
CB
O
CO
4)
O
!Z
0
ra
0
QC
20-
10-
0
h"H" series
• 1 • .1 .. 1
17 16 15 14 13 12 11 10 9 8 7 6 54 3 2 1
100-i
90-
80
70-
60-
50-
40-
30 •
20-
10
O
. 1
Illllll,
II....
Z YXWVUTSBBCCDDABC DE FGH I J
Station
Figure 10.
Relative Fluorescence of Groundwater Samples Observed with
50 nm Synchronous Scan Procedure
Table I.
GWF Directions & Relative Rates
Seymour, Indiana August, 1980
Hole
A
B
C
D
E
F
G
H»
I*
J
p..
AA
BB
CC
DD
R17
R14
R15
Compass
Direction
338°
340°
327°
310°
352°
347°
315°
u
u
344°
Unstable
320°
334°
346°
335°
311°
335°
23°
Rate
(Readout)
3.5
2.5
6.5
2
4.5
4
3.5
6.5
6
2
3.25
2.5
1.5
3.25
2.75
Velocity
(ft/day)'
1.3
1.0
2.5
.8
1.7
1.5
1.3
2.5
2.3
1.8
1.2
1.0
.6
1.3
1.0
•H and I were unstable in flow characteristics and appeared to be in-
fluenced by a possible leak in a water pipe or other localized source.
•*P was taken following rainfall and showed erratic flow (towards
83 °E).
•••0.375 x readout = ft/day.
u = Unstable.
-------
CASE HISTORIES 357
Relative Fluorescence
30 nm synchronous scan
20-
10-
17 16 15 14 13 12 11 1O 9 8 7 6 5 4.3 2 1
100
90
SO
7O
60-
50
4O-
3O-
20-
1O
O
•lllllL
ZYXWVUTSBBCCDDABCDEFGH I J
Station
Figure 11.
Relative Fluorescence of Groundwater Samples Observed with
30 nm Synchronous Scan Procedure
Conductance micromhos/ cm.
800
400-
17 16 15 14 13 12 11 10 9 8 7 6 3 4 3 2 1
2400
o
o
£2000
o 1500
U
1200
800-
400-
'ZYXWVUTSAABBCCDDABCDEFGHI J
Station
Figure 12.
Conductance of Groundwater Samples
STREAM INFILTRATION
The shallow (0.5-2.5 ft deep) east-west stream formed
hydraulic continuity with the shallow aquifer. The water
level of the stream on dry days was roughly 5.5 ft below
the grade of the cornfield. This placed the sandy bottom of
the stream at the top of the porous water-bearing strata.
A hole was augered near Station Two to the aquifer and
the difference in water head compared by inserting a
primed hose between the hole and the shallow stream. The
observed flow of water was rapidly towards the stream,
clearly indicating infiltration of groundwater up into the
stream.
To estimate the rate of inflow, Lee-type lysimeters were
installed in the stream bed to determine the rate of ground-
water infiltration in the vicinity of the plume intersection.
Site
1
2
3
4
Flow Rate
246 ml/10 min
145ml/15min
60 ml/30 min
72 ml/20 min
X =
Infiltration (ftVftVday)
0.67
0.26
0.05
0.10
0.27
The section under investigation averaged 2.6 ft in width.
With an estimated 0.27 ftVftVday infiltration along a
projected zone of length of 800 feet of stream bed inter-
cepted by the plume, the rate of discharge would be about
560 ftVday or approximately 4200 gal/day.
WELL DEVELOPING
On August 22, existing observation wells were pumped
for developing. The deep well occurring close to the drain-
age ditch south of the indicated position of well 9 was
well-developed with no fines apparent upon successive
pumping. Wells 1, 5, and 6 were found to penetrate to
10 ft below the top of the casing. Wells 1 and 5 were devel-
oped by drawdown and surging with 5 gal/mm centri-
fugal pump with hose to bottom of the casing. Well 6 was
drawn down to the bottom of the screen with little or
no apparent inflow. After two hours, only 2 in. of water
had filled the well screen. This well did not appear cap-
able of being developed. Well 5 had sustained slight dam-
age due to earth operations in the vicinity. A 20° bend
was apparent 2.5 ft below casing top (one ft below grade).
CONCLUSIONS
From August 16 through August 23, 1980, K-V Asso-
ciates, Inc. investigated the extent of subsurface con-
tamination of groundwater in the immediate vicinity of
the Seymour Recycling Facility in Seymour, Indiana.
Hand augering on the site revealed that water-bearing
strata existed at depths of 8-9 ft below grade.
Thirty shallow observation holes were excavated across
the site area and used to measure the direction and rate
of groundwater flow. After the major direction of ground-
water flow was determined, samples were taken of the
-------
358 CASE HISTORIES
groundwater across the axis of flow with a portable
well point sampler driven 14 ft below grade to determine
the location of any material moving through the aquifer.
Samples were analyzed for conductivity and petroleum-re-
lated aromatic organics (by UV fluorescence).
The following conclusions were reached:
•The mean groundwater flow direction was found to be
away from the town of Seymour and in a Northwester-
ly direction (approximately 316° NW) evidenced by direct
flow measurement and the location of a plume of or-
ganic leachate.
•A plume of petroleum-related aromatic hydrocarbons
was observed extending from the site in a Northwest-
erly direction discharging into surface waters at a con-
fluence of an east-west stream and Heddy Run South of
Hangman Crossing.
•The position of the subsurface plume coincided with a
zone containing numerous dead or dying trees, partic-
ularly larger trees with deeper root structures.
•The mean flow rate of the shallow aquifer was meas-
ured at 1.2 feet per day compared to a previous estimate
of 0.016 ft/day. The combined influence of siphoning of
flow to the nearby streams and restriction of lateral flow
strata by natural clay deposits probably contribute to the
more rapid flow than previously estimated.
•During storm period temporary changes in flow of the
streams (direction) were observed in the aquifer flow
when storm flooding occurred, further substantiating the
connection between both water bodies.
•Since the shallow aquifer is intercepted by the stream
bed of Heddy Run and in places by an east-west trib-
utary, it is possible that the plume may have stabilized
in position and is no longer moving laterally. Instead the
90-
80-
70-
0
£60'
0
»50-
0
30
20
10
Storm Runoff Sampling
50 nm Synchronous Scan
Heddy Run Confluent
Sources
\
III
1
Time
18 19 20 21 22 26
« o
«6 «D
-------
HAZARDOUS WASTE SITE INVESTIGATION
SYLVESTER SITE, NASHUA, NEW HAMPSHIRE
JOHN J. GUSHUE
GHR Engineering Corporation
New Bedford, Massachusetts
JOHNE.AYRES
Goldberg-Zoino and Associates
Newton Upper Falls, Massachusetts
ALVINJ. SNYDER
Environmental Resource Associates
Warwick, Rhode Island
INTRODUCTION
In this paper, the authors describe the results of an
engineering and hydrogeological investigation of the Syl-
vester Site undertaken by GHR Engineering Corporation
and its principal subcontractor, Goldberg-Zoino and Asso-
ciates (GZA). The investigation began in April 1980 and
was completed in July 1981 with submittal of the final
project report to the New Hampshire Water Supply and
Pollution Control Commission (NHWS & PCC) and the
U.S. Environmental Protection Agency (U.S. EPA.)(1)
The investigation of the Sylvester Site and adjoining
properties was designed to:
(1) Define the extent of soil, surface water and ground-
water contamination;
(2) Develop an engineering assessment of the hydrogeo-
logical and hydrological conditions governing contam-
inant movement from the site; and
(3) Evaluate alternative remedial actions on the basis of
technical feasibility, estimated costs and environ-
mental implications.
The evaluation of remedial actions included a labora-
tory-scale assessment of the feasibility of removing con-
taminants from groundwater in an on-site, above-ground
water treatment plant. The final project report, includ-
ing recommendations for remedial action, provided the
basis for the first Federal/State cooperative agreement
under Superfund, with $2.3 million being awarded to the
State of New Hampshire in August 1981 for design and
construction of remedial measures at the site.
SITE DESCRIPTION
The Sylvester Site is an approximately six-acre open
dump located behind the C & S Disposal Company gar-
age on Gilson Road in Nashua, New Hampshire. The site
is a former sand pit which was excavated in places to ele-
vations lower than seasonal high groundwater levels. Upon
completion of sand mining operations, the pit was grad-
ually filled during the 1970s with various types of refuse
and waste materials, consisting primarily of construction/
demolition debris, but also including large amounts of
chemical liquids and sludges. The waste materials extend
into the water table throughout much of the site. Sev-
eral methods were used to dispose of hazardous liquid and
sludge wastes at the site, including:
•Subsurface discharge of large volumes of liquid wastes
through 2-3 ft diameter pipe extending into the site from
the garage building
•Subsurface burial of drums of waste in at least two loca-
tions on the site
•Surface and subsurface disposal of sludge-like material
in several areas of the site
•Stockpiling of 55-gal drums of chemical wastes on the
surface of the site
In May and June 1980, all drums visible and access-
ible from the surface of the site were removed by Recra
Research, Inc. of Tonawanda, New York under subcon-
tract to GHR Engineering. A total of 1,314 drums were
transported from the site for final safe disposal in New
York and Ohio. A complete report on the drum removal
work was submitted to the NHWS & PCC in July 1980.(2)
The area surrounding the site is primarily residential
(Figure 1), although the land across Gilson Road to the
south is undeveloped swamp and woodland. Directly ad-
jacent to the east and north of the site are two large trailer
parks. The closest residences, in the Rodger's Trailer Park,
are within about 100 ft of the northeast perimeter of the
site. The Thompson property, which abuts the site to the
west, is a large, open parcel that is an active local source
of sand and gravel.
As indicated on Figure 1, contaminated groundwater
from the site flows northwesterly toward Lyle Reed Brook,
which is about 680 ft from the site and the Nashua River.
The contaminant plume from the site has already moved to
and beyond Lyle Reed Brook. Fortunately, the residents of
Jensen's Trailer Park are not dependent upon local
groundwater resources for drinking water, since the trailer
park is served by the municipal system.
Not so fortunate are residents along Route 111 in the
path of groundwater flow. Their individual drinking water
wells are in jeopardy, i.e., they will eventually be con-
359
-------
360 CASE HISTORIES
laminated as the plume continues to expand. Trace levels
of organic contamination have already been found in a
ground-water observation well located between Tumble-
brook Road and the Pennichuck Water Department well,
which is not presently in use. In addition to drinking water
supplies, a principal public health concern at the site is the
Lyle Reed Brook into which contaminants moving in
upper groundwater zones are being discharged, eventually
being carried in surface flows to the Nashua River.
FIELD EXPLORATION AND TESTING PROGRAM
Upon completion of drum removal work in mid-June
1980, an extensive field exploration and testing program
was carried out to develop the information needed to eval-
uate remedial options for the site. The complete explor-
ation and testing program was a multi-phased effort that
consisted of the following:
•Subsurface investigations of the disposal site by remote
sensing (magnetometer and partial electrical resistivity
surveys)
•Excavation of 34 test pits within the confines of the dis-
posal site, which included soil sampling and analyses,
groundwater sampler installations, and stratigraphic
logging.
•Execution of 15 test borings with soil sampling, rock
drilling and field analyses for organic chemical contam-
ination
•Installation of permanent groundwater sampling systems,
both conventional observation wells and gas drive
samplers, at 23 locations in the vicinity of the disposal
site (Figure 2)
•Installation of a 1.0 ft diameter well and execution of a
5-day aquifer pumping test.
The pumping test was conducted to:
(1) refine estimates of the hydraulic characteristics of the
aquifer;
B indicates private well in jeopardy
Rodger's
Trailer Park
Figure 1.
The Area Surrounding the Sylvester Site
(Not to Scale)
-------
CASE HISTORIES 361
(2) observe the impact of pumping on groundwater qual-
ity; and,
(3) collect a representative sample of the aquifer for use in
laboratory-scale testing of the effectiveness of several
alternative water treatment processes in renovating the
contaminated groundwater.
Groundwater quality in the vicinity of the Sylvester Site
has been monitored periodically since July 1979, beginning
with the installation of a ring of 10 groundwater ob-
servation wells around the site by the NHWS & PCC.
During the GHR/GZA site investigation, another 27
groundwater observation wells were added to the monitor-
ing network. In addition, 42 private drinking water wells
and 3 Pennichuck Water Department wells (1 production
well, 2 observation wells) have been tested by the U.S.
EPA as part of this site investigation. Surface water qual-
ity in Lyle Reed Brook and the Nashua River has been
monitored by the U.S. EPA on several occasions since
March 1980.
FINDINGS: CONTAMINATION
LEVELS AND DISTRIBUTION
Types and Amounts of Contaminants
The site and its associated groundwaters are highly con-
taminated with a large variety of toxic organic and inor-
ganic compounds. The groundwater contaminant plume
extends from the disposal site to the vicinity of the Penni-
chuck Water Department well between Tumblebrook and
Trout Roads.
The most contaminated groundwater zone extends from
the disposal site to the vicinity of wells #5-D and #M-2 on
the Thompson property; however, some mobile volatile
organic compounds have travelled as far as the Penni-
chuck well. Several of these compounds, including primar-
ily toluene and tetrahydrofuran (THF), have been con-
sistently detected in Lyle Reed Brook in the 5 to 10 mg/1
range since July 1980. The less mobile extractable organic
and inorganic compounds known to be present in the
groundwater contaminant plume have not been detected in
Lyle Reed Brook to date.
A tabular summary of the types and amounts of con-
taminants detected in groundwater and surface water
monitoring since July 1979 is presented in Table 1. The
data in this table provide an indication of the extent of
the adverse water quality impacts of the Sylvester Site and
a general perspective on observed contaminant migration.
However, the very dynamic temporal and spatial char-
acteristics of the contaminant plume and its movement
are oversimplified here. The ranges of the pollutant con-
centrations presented in Table I are the lowest and high-
est levels detected to date at sampling stations grouped as
Mows:
(1) Groundwater monitoring wells nearest to the down-
gradient perimeter of the Sylvester Site, including wells
#2A, #3, #4, #M-1 and #A-4
(2) Groundwater monitoring wells located between the site
and Lyle Reed Brook, including wells #A-2, #A-3,
#M-2 and #5-D. In this area, only well #A-3 has been
analyzed for extractable organics
The principal groundwater contaminants in terms of
concentrations are volatile organic solvent compounds
which, in total, have been found at levels as high as
approximately 1,800 mg/1. Concentrations of non-volatile
organic priority pollutants in groundwater have been in the
5 to 15 mg/1 range. No priority pollutant pesticides or
PCBs have been detected in groundwater. Arsenic levels
in groundwater are elevated (within the 0.20 to 1.2 mg/1
range) throughout much of the study area. Concentra-
tions of total metals in groundwater are within the 100 to
1,000 mg/1 range in the study area, with iron and man-
ganese being by far the major inorganic constituents.
Volatile organic contaminants in groundwater from the
Sylvester Site were first detected in Lyle Reed Brook in
March 1980 and have been steadily increasing since. Con-
taminant levels in the brook in September and December
1980 were in the 7 to 10 mg/1 range in the vicinity of Tum-
blebrook and Trout Brook Roads, and in the 2 to 3 mg/1
range at the Route 111 culvert. The decreasing concen-
trations along the brook may reflect the effects of dilu-
tion, aeration, volatilization and biodegradation.
No contaminants other than volatile organics have been
detected in the brook to date. The Nashua River has
been tested by the U.S. EPA at a point approximately
1,000 ft downstream of where the brook enters the river,
and no volatile contaminants present in the brook have
been detected in the river at that point.
Contaminant Transport Mechanisms
The disposal site itself is a depression carved into the
natural landscape as a result of sand mining operations.
The subsequent waste disposal activities have left a very
irregular surficial topography within the approximately 6-
acre area now enclosed within a chain link fence. The
ground surface immediately outside the fence consists of
steep embankments to higher ground to the west and
north, and relatively flat areas to the east and south. All
surface runoff in the immediate vicinity of the site flows
into the site; therefore, surface runoff is not a pollutant
transport mechanism from the site itself. Surface trans-
port of pollutants in the study area comes into play as con-
taminated groundwater is discharged to Lyle Reed Brook
and then flows in the brook to the Nashua River.
The characteristics of the groundwater flow regime in
the study area will ultimately determine the fate of pollu-
tants released into the aquifer underlying the site, and it
is apparent that the principal threats posed by the Syl-
vester Site are associated with groundwater. For this rea-
son, the hydrogeologic properties of the aquifer, along
with the characteristics of the flow regime, are of primary
importance.
Groundwater contours developed from water table mea-
surements made in December 1980 are shown in Figure
2. Assuming groundwater flow to be perpendicular to the
contours presented, the principal direction of flow is to the
northwest toward Lyle Reed Brook. Regional flow is also
in a northwesterly direction toward the Nashua River and
it is assumed that the Nashua River is the ultimate "sink"
for groundwater leaving the disposal site. A portion of the
flow reaches the river via Lyle Reed Brook and the bal-
-------
362 CASE HISTORIES
ance remains as groundwater, probably flowing in a path
roughly parallel to but beneath the brook.
Hydraulic gradients (taken as the slope of the water
table) across the site range from less than 0.2% to ap-
proximately 0.8%, but typically average 0.3 to 0.4%.
Anomalies in the gradient values as well as the overall
shape of the groundwater contours were consistently ob-
served in the vicinity of wells #A-2 and #M-2. Steepening
of the water table slope at this point is potentially indica-
tive of a localized decrease in aquifer transmissivity, poss-
ibly caused by a geological anomaly. Seasonal variations
of the water table elevations observed at the site during the
duration of the current study indicate that seasonal fluc-
tuations will not have a significant impact on the ground-
water flow regime or, correspondingly, on contaminant
migration.
Based on a combination of pumping test data, soil grain
size analyses and published information for the soil types
encountered, the hydraulic constants of the aquifer were
estimated. The estimated transmissivity of the fine to med-
ium sands underlying most of the site is 700 to 2,400 ft2/
day, representing a coefficient of horizontal permeability
of 20 to 70 ft/day (based on a saturated aquifer thick-
ness of 33 ft). For sand and gravel deposits encountered
north and west of the disposal area, transmissivity is
estimated at 3,300 to 6,600 ftVday; corresponding co-
efficients of permeability are from 100 to 200 ft/day. The
average linear or "transport" velocity of the groundwater
at the site was estimated to be 0.8 ft/day for the fine to
medium sands and approximately 1.6 ft/day in the areas
north and west of the disposal site, where coarser de-
posits would be expected to control groundwater move-
ment.
Based on the calculated transport velocities, travel times
were estimated between selected points for transport of
contaminated groundwater originating from the Sylvester
Site. A summary of transport times from the midpoint of
the disposal area to relevant points is tabulated below:
Location
Lyle Reed Brook
Pennichuck Well
Route 111
Nashua River
Nashua River (from Pennichuck
well)
Transport Time
3-5 years
4-6 years
5-7 years
7-9 years
2-5 years
4
1. XW to Nllhiu Rlv»r *.
Plum* Llrnlli ^j /
* " 1 I CIS Olipoul
/ Coapany Cerage
\l
Cltson Road * •lA-fi
— . L
^
\
\
i
/
a
O Multllwd Wdl
^ „ • W.llpolnl W«ll
+ Pinplng Will
C Property Ura
Figure 2.
Monitoring Wells, Groundwater Contours and
Contaminant Plume Limits at Sylvester Site
-------
CASE HISTORIES 363
Since the calculated travel times were based on a num-
ber of simplifying assumptions, they are considered to be
approximations only. Total volumes of contaminated
groundwater within various portions of the contaminant
plume were estimated from assumed porosity values and
observed saturated aquifer thickness. Considering the en-
tire plume from the disposal site to Lyle Reed Brook, the
total volume of affected groundwater is approximately
60 million gallons. Of this total, about 30 million gallons
lie within the immediate confines of the Sylvester Site as
delineated by the fenced area. The total estimated flow
of groundwater through the disposal site, considering a
cross-section across the width of the containment plume
at multilevel well #M-1, is approximately 31,500 gal/day
(with a range of from 13,000 to 65,000 gal/day).
Contaminant Distribution
At the Sylvester Site, the wide area over which waste
disposal occurred has caused the formation of a corres-
pondingly wide contaminant plume in the vicinity of the
source (see Figure 2). As the contamination migrates
northwest, the geometry of the water table contours tends
to counteract the spreading effects of hydrodynamic dis-
persion, thereby limiting the width of the plume. The re-
sult is a fairly sharply defined zone of contamination
whose southerly and easterly borders are confined to the
immediate area of the site. As the plume approaches Lyle
Reed Brook, a degree of spreading is inevitable as much
of the contaminated groundwater flows into the brook,
leaving the site as surface water. The balance of the plume
flows beneath and parallel to the brook.
Interpretations of contaminant distributions at the Syl-
vester Site based on most recent analytical data are shown
in Figures 3 and 4. Contours of pollutant concentra-
tion depicted on the figures are approximate, idealized
and subject to change with time, but the representations
adequately illustrate significant trends.
The observed distribution of volatile organic chemicals
(Figure 3) closely resembles theoretical modeling of the mi-
gration of a contaminant from a continuous point source
in a uniform flow field. The point source appears to be
centered in the vicinity of the subsurface leaching trench
found at the rear of the garage. From this point, lines
of equal concentration form elliptical lobes which extend
northwest in the direction of groundwater flow. Slight dis-
10' to Nashua River \
\
Figures.
Distribution of Total Volatile Organics in Groundwater
at Sylvester Site (in PPM, December 1980)
-------
364 CASE HISTORIES
tortions of the contours are observed in the central and
southern portions of disposal area and are probably the
results of localized dumping incidents.
Within the main lobe of the contaminant plume, con-
centrations of the major constituent, tetrahydrofuran
(THF) are often greater than 1,000 mg/1 For the other
principal volatile components of the waste, including
toluened, methylene chloride, and MIBK, corresponding
concentrations are generally one to two orders of magni-
tude lower. The distributions of these compounds and the
additional volatiles detected typically display more irreg-
ularity than the corresponding distribution of THF. Local-
ized "hot spots" or highly contaminated zones were ob-
served for various compounds in test pits and observa-
tion wells on the premises of the site, probably indicative
of "slugs" of pollutants percolating into the groundwater
from the unsaturated zone.
The distribution of metallic compounds (Figure 4) ob-
served in the vicinity of the Sylvester Site significantly
differs from the pattern displayed by the volatiles. Marked
lateral distortion of the theoretically elliptical concen-
tration contours indicates either a large source area or
multiple point sources. Highest concentrations of contam-
inants were observed in the central portions of the disposal
area, suggesting that the primary source of the metals is
not the liquid wastes from the leaching pit.
Although significant attenuation was observed with dis-
tance from the site, concentrations of total metals sig-
nificantly higher than background values were detected
up to 500 ft from the disposal area. Since metallic com-
pounds are typically less mobile in groundwater than vol-
atiles, this would suggest that the source of the metals is
significantly older than the source of the volatiles.
Summary of the Extent of Contamination
Before discussing the various remedial alternatives, sev-
eral important findings relative to the extent of the con-
tamination problem should be summarized. While signif-
icant adverse environmental effects, i.e., degradation of
surface water and groundwater quality, have already
occurred in the study area, the worst is yet to come.
The groundwater contaminant plume is moving from
the site toward the Nashua River at an approximate rate
of from 0.8 to 1.6 ft/day. The leading edge of the plume
extends to the vicinity of the unused Pennichuck Water
1. 900' to N.ihu. River
/ *
ThOBpson
Property
• lA-I
\
1
1
1
t CIS Dlipojal
Company Girage
Cllwn Raid I • •*-«
'\\
V
1
1
1
' 1
O Multilevel Well
^ „ • Wellpohil Well
^ Puieplng Well
t Properly Line
Figure 4.
Distribution of Total Metals in Groundwater at
Sylvester Site (in PPM, October 1980)
-------
CASE HISTORIES 365
Table I.
Summary of Types and Amounts of Contaminants
in Groundwater and Lyle Reed Brook
COMPOUNDS /PARAMETERS
A. Volatile Orsanics (ppb)
Vinyl Chloride*
Chloroethane*
Methylene Chloride*
1 , 1-Dichloroethane*
t-1 , 2-Dlchloroethylene*
Chloroform*
1 , 2-Dlchloroethane*
1,1. 1-Tr Ichloroechane*
Trlchloroethylene*
Tetrachloroethylene*
Benzene*
Toluene*
Ethyl Benzene*
Xylenes
Tetrahydrof uran (THF)
Methyl Echyl Ketone(MEK)
Methyl Isobutyl Ketone(MIBK)
Acetone
B, Extractable Organlcs (ppb)
2- Chlorophenol*
2 , 4-D imethylphenol*
2-Nitrophenol*
Pentachlorophenol*
Phenol*
o-Cresol
m-Cresol
Benzole Acid
1 , 2-Dichlorobenzene*
1 , 4-Dichlorobenzene*
Naphthalene*
Bls(2-ethylhexyl)Phthalate*
Dl-n-Butyl Phthalate*
Dlethyl Phthalate*
Dimethyl Phthalate*
Pesticides*
Polychlorlnated Blphenyls*
C. Inoraanlca (ppn)
Arsenic*
Barium*
Cadmium*
Chromium(total)
Lead*
Marcury*
S.l.nlum*
Sllv.r*
Coppar
Iron
Managanaaa
Nlckal
Zinc
D. Other Pararatttrt
pH
Specific Conductance (uMHOi)
Total Organic Carbon (ppm)
Chemical Oxyjen Dimand (ppn)
Total Phanollci (ppn)
LEVELS FOUND IS WELLS LEVELS FOUND IN WELLS LEVELS FOUND IN
NEAREST TO THE SITE"' BETWEEN SITE AND BROOKb' LYLE REED BROOKC '
ND to 950
ND to 320
ND to 47,500
ND to 210
ND to 5,000
ND to 1,600
ND to 890
ND to 1,400
ND to 2,700
ND to 1,600
ND to 2,800
< 10 to 100,000
ND to 2,700
SD to 1,000
ND to 4,300
44,000 to 1,200,000
ND to 7,000
ND to 33,250
ND to 48,000
ND to < 10
ND to 40
ND to 15
ND to < 10
ND to 2,700
ND to 1,137
ND to 864
ND to 320
ND to 88
ND to < 10
ND to 45
ND to < 10
ND to < 10
KD to 47
»D to UO
ND In this area
ND In this area
.04 to 1.7
< .1 to 2.5
* .005 to .01
.01 to .7
< .01 to .5
< .0002 to < .001
< .01 to < .01
.01 to .02
< .01 to .6
18. to 640.
2.8 to 115.
* .02 to .9
< .003 to 20.5
5,9 to 6.8
250, to 4,092,
12. to 7,600.
711. to 19,300,
.298 to 18. S
ND In this area
ND in this area
ND to 122,500
ND to 15
ND to 18,000
ND to 31,000
ND to 7
ND to 2,000
ND to 15,000
ND to 570
ND to 3,400
< 100 to 29,000
ND to 1,100
ND to 1,200
ND to 10,000
4,400 to 1,500.000
ND to b.OOO
ND to 13,000
ND to 36,000
— ND in OW A- 3—
ND in OW A- 3
— ND in OW A-3
— ND in OW A-3—
6,464
1,969
3,672
6,011
— ND In OW A-3
ND In OW A-3
ND In OW A-3
ND in OW A-3
ND in OW A-3
— ND In OW A-3
ND in OW A-3—
ND In OW A-3
ND in OW A-3
.006 to 1.05
.08 to .9
* .002 to «.005
.01 to .19
« .01 to I. 01
«,0002 to '.001
« .01 to < .01
.01 to .01
.04 to .6
16, to 580.
.2 to 120.
.2 to .9
,09 to 9.
5.9 to 6.8
228. to 9,517.
6. to 4,700.
— no data —
« .00} to 25.2
ND • not detected; NC - preeence »uapect«d, but not conflrmid; * - a priority
— ND In brook
ND to 11
ND in brook
ND to 81
ND in brook
ND to < 10
— ND In brook
ND in brook
ND in brook
ND In brook
ND to t 10
SD to 2,700
ND in brook
ND to 20
ND to 16
< 100 to 7,650
— -ND In brook
— NC in brook—
— NC In brook
ND in brook
ND in brook
ND In brook
ND in brook
ND in brook
ND in brook
ND in brook
ND In brook
ND In brook
ND in brook
ND in brook
ND in brook
ND In brook
ND In brook
ND in brook
ND In brook
ND In brook
-—no data
.06
- .005
.01
« .01
» .001
— no data —
« .01
* .1
10.
4.2
.1
.02
— no data —
— no data — -
— no data—
— no data —
— no data —
pollutant,
b> walU bitwxn iltt and brook aril »A-2. I/A-3, "M-2, and «-D,
c> Brook •ampllnn utatloni rapraiintid In table art: NHDB 04, 04A, 05, 05A, and 06. Inorganic! data ara
available only for atatlon NHDB 04A along thin portion of Lyli Hied Brook.
-------
366 CASE HISTORIES
Department well located between Tumblebrook and Trout
Brook Roads. Contaminants (in groundwater) that have
already moved to and beyond Lyle Reed Brook are, for all
intents and purposes, not recoverable. Thus, the risk of
organic contamination reaching private supply wells along
Route 111 is high regardless of what is done at the site.
The estimated travel time for groundwater to move from
the Pennichuck well to Route 111 is from 1 to 3 years
(travel time to the Nashua River is estimated at from 2 to
5 years). It is also possible that contaminated ground-
water is moving in bedrock joints. Thus, any bedrock
wells located downgradient of the disposal site are in
jeopardy.
The most contaminated zones extend from the disposal
site to within from 100 to 200 ft of Lyle Reed Brook.
If not prevented, these zones of highest groundwater con-
tamination will reach Lyle Reed Brook within a year; the
Pennichuck well in 1 to 4 years; Route 111 in 2 to 6 years;
and the Nashua River in 3 to 9 years. Furthermore, if
contaminants are not prevented from leaving the disposal
site, then groundwater contamination between the site and
the Nashua River will steadily increase for several years
and then remain at high levels for an indefinite period.
Since a major portion of the groundwater plume is dis-
charged to Lyle Reed Brook, pollutant concentrations in
the brook would also increase for several years before
stabilizing at high levels for an indefinite period. Of
course, any increase in the pollutant load in the brook will
be reflected in the Nashua River as well.
REMEDIAL ACTION ALTERNATIVES
During this investigation, several alternative measures
available for mitigating the environmental and public
health threats posed by the Sylvester Site were considered.
These included:
•No Action—natural flushing in combination with fre-
quent water quality monitoring
•In-place hydrologic isolation of contaminated soil and
groundwater zones
•Interception and treatment of contaminated groundwater
in the area between the site and Lyle Reed Brook
•Excavation of the disposal site for secure landfill dis-
posal either in an existing facility in upstate New York,1
or in a new facility that would have to be built in the
Nashua area for that purpose
•Combinations of the above
The various alternative remedial action measures are
briefly discussed below.
No Action (Natural Flushing)
Under certain conditions, an appropriate means of deal-
ing with an existing soil and groundwater contamination
problem is to take little or no direct preventative action,
thereby relying upon natural processes to dilute, disperse
and degrade the contaminants over an extended time
period. In terms of environmental protection, a no action
policy is the least desireable alternative due to the high
potential for adverse impacts on surface water and ground-
water. The key unknowns are the magnitude of subse-
quent stream and groundwater contamination; and the
time-span required before contaminant levels in water
would decrease to "acceptable" levels. For the Sylvester
Site, the "no action" option was considered to be unac-
ceptable due to the extremely high contamination levek
in groundwater, the mobility of the contaminant plume
and the potential for long-term adverse environmental and
public health conditions.
In-Place Hydrologic Isolation
Hydrologic isolation options for the Sylvester Site in-
clude the installation of a slurry cut-off wall around con-
taminated zones (possibly with bedrock grouting) and in-
stallation of an impermeable cover over the contaminated
area. A slurry cut-off wall is a vertical, essentially im-
permeable barrier installed to prevent groundwater inflow
and outflow from the contaminated zone. A slurry cut-
off wall installed to surround contaminated zones would
prevent clean groundwater moving above rock from be-
coming contaminated, and would prevent further leach-
ing of contaminated groundwater from the isolated area.
There are two viable options for the location of the
slurry trench cut-off wall around the site (see Figure 5).
The first option would isolate only the immediate area of
chemical dumping. The second option would encompass
all of the area within the plume from the site to the edge
of a bedrock valley which runs north-south roughly par-
allel to Lyle Reed Brook.
The second slurry wall option would isolate the bulk of
the contaminant plume, while the first would have no
remedial impact on the portion of the contaminated zone
which has already moved beyond the confines of the site.
In conjunction with the slurry cut-off wall, an imperme-
able surface cap consisting of either impervious soil (i.e.,
clay) or a synthetic liner would be required to prevent in-
filtration of precipitation through the enclosed waste.
Hydrologic isolation by a combination of a slurry wall
and injection grouting of the bedrock carries the contain-
ment alternative discussed above one step further. In addi-
tion to the cut-off wall within the overburden and the im-
permeable surface cap, fractured bedrock along the desig-
nated perimeter line would be sealed by the injection of
a bentonite/cement grout under high pressure. The en-
suing reduction of rock permeability would be accom-
panied by a decrease in contaminant migration. At the
Sylvester Site, bedrock has been cored in only one loca-
tion (boring M-l) and was found to be moderately frac-
tured.
The slurry wall/grout curtain alternatives would retard
essentially all movement of contaminants across the con-
tained area. Over a period of time, however, there is the
possibility of leakage through the cut-off walls. For this
reason, a monitoring system would be needed. An intrin-
sic problem with any hydrologic isolation system, is the
inability to alleviate contamination which has already
moved off-site. The limits of the containment area would
have to be expanded beyond practical proportions to iso-
late all contamination associated with the site.
-------
CASE HISTORIES 367
Groundwater Interception and Treatment
A preliminary assessment of the technical feasibility and
estimated costs of an above-ground treatment system for
renovating the contaminated groundwater underlying the
Sylvester Site was prepared through a laboratory-scale
treatability study. This study was conducted with assis-
tance from Environmental Resource Associates, Inc. The
purpose of the laboratory-scale treatment study was to de-
velop a "first-cut" assessment of the treatability of the
highly contaminated groundwater at the site. As such, the
study was confined to a small number of alternative water
treatment methods tested under a limited set of operating
conditions. Using an integrated batch testing approach,
the pollutant removal effectiveness of two alternative
treatment process trains were evaluated:
•Air stripping followed by neutralization/precipitation
followed by activated carbon adsorption
•Air stripping followed by neutralization/precipitation
followed by biological treatment
Contaminated groundwater for the treatability study
was obtained during the aquifer pumping test conducted
during the period January 28-February 2, 1981. The raw
groundwater sample was collected toward the end of the
pumping in order to base the treatability assessment on
feedwater that would be most representative of the effluent
produced under extended pumping conditions.
Using the information developed during the laboratory-
scale evaluations, preliminary design and cost estimates for
a full-scale, on-site treatment system were developed. The
conceptual treatment system proposed for the Sylvester
Site groundwater is illustrated in the block flow diagram
in Figure 6. Longer duration and more operationally flex-
ible pilot-scale testing will be needed prior to final treat-
ment system design. Among the key uncertainties to be
resolved in pilot-scale testing area the need for a metals
removal step and the optimum operating conditions for
the biological and carbon adsorption processes.
Total Removal Alternatives
For comparison with the remedial measures outlined
above, preliminary cost estimates were prepared for the
so-called total removal options involving excavation of the
disposal site for secure landfill disposal. Assuming an
average depth of removal of 10 ft over the 6-acre site,
1.900' to Nashua Rlv.r
OlM-2
• 12
Cllson Road
CCS Disposal
Company Garage
• lA-6
"^A
\
\
1
1
I
/
O Multilevel Wall
0 „ • Wellpoint Well
+ Pumping Well
t Property Line
Figure 5.
Slurry Wall and Cap Isolation Options for
Sylvester Site
-------
368 CASE HISTORIES
the costs to excavate, transport and dispose of the ma-
terial in an existing secure landfill facility in upstate New
York and to design and construct a new secure landfill
disposal facility in the Nashua area were estimated. The
conditions at the Sylvester Site are such that, even with
removal of the disposal site, an extensive groundwater
interception and treatment system would be needed as part
of a total removal plan.
RECOMMENDED REMEDIAL ACTION PLAN
The recommendation of the GHR/GZA Project Team
was that the final remedial action plan for the Sylvester
Site include both in-place hydrologic isolation of contam-
inated zones and interception of contaminated ground-
water for on-site treatment. Specifically, it was recom-
mended that the final plan consist of:
(1) Design and installation of slurry trench cut-off wall
Option 2 to completely surround approximately 12.5
acres of the contaminated zone, including the 6-acre
disposal site
(2) Design and installation of an impermeable top-seal
over the isolated zone to prevent infiltration of rain-
water
(3) Design and installation of a groundwater interception
and treatment system to renovate that portion of the
contaminated aquifer remaining outside the isolation
zone
(4) Installation of 6 to 8 new multilevel groundwater ob-
servation wells to monitor the effectiveness of the
slurry wall system and track the contaminant plume
in the region between Lyle Reed Brook and the Nashua
River
In addition to the design and implementation of the
slurry wall/treatment system it was recommended that:
(1) Lyle Reed Brook be posted and fenced to prevent
human exposure to the contaminated stream and
(2) city water be provided to those residences along Route
111 between Countryside Drive and Jensen's Trailer
Park who now rely on private wells for drinking water.
The severly degraded condition of the groundwater in
the area from the site to Lyle Reed Brook requires that
the total remedial action plan for the site include ground-
water interception for on-site treatment. The groundwater
interceptor wells would be located downgradient of the
selected slurry wall. Through strategic location of the
wells, with respect to both areal orientation and depth,
a minimal amount of contaminated groundwater flow will
WELL FIELD
Removal contaminated
groundwater from
plume and deliveri
it to above-ground
treatment system.
VAPOR RECOVERY
Reduces levels of
organic solvents In
spent air from
stripping column.
Activated carbon
used to recover
solvents in liquid
form.
AIR STRIPPER
Removes volatile
organic solvents
from groundwater.
Utlllies air strip-
ping column to
pass air and water
in countercurrent
flow over a packed
inert media.
pH ADJUSTMENT
METAL REMOVAL
Adjusts pH of
groundwater to
1.5. Removes
metals by preci-
pitation and
clarification.
BIOLOGICAL
Will degrade alcohols
by biological activity
and remove remain-
Ing volatile organlcs
by air stripping.
CARBON ADSORB
Removes high
molecular weight
organlcs and
color by adsorp-
tion on activated
carbon beds.
LEACHING TRENCH
Covered In-ground
leeching trenches
to return treated
groundwalor to
aquifer.
Figure 6.
Block Flow Diagram of Conceptual Treatment System for
Groundwater Renovation at Sylvester Site, Nashua,
New Hampshire
-------
CASE HISTORIES 369
pass through the interception system to downgradient
areas.
The principal variables with the interception and treat-
ment scheme are:
(1) the treatment processes to be used,
(2) the degree of pollutant removal desired and
(3) the length of time that the treatment plant would have
to be operated.
Based on aquifer volume calculations and interceptor
system flow rates, the treatment periods required to treat
given multiples of the initial contaminated aquifer volume
were estimated. This was done for the two slurry wall op-
tions and the additional case of interception and treat-
ment without hydrologic isolation. The results are pro-
vided on Table II. the effluent from the treatment plant
would be recirculated back through the aquifer by dis-
charging into a subsurface leaching trench to be located
outside the slurry wall.
Table II.
Estimated Treatment Periods Required for Various Design
Flow Rates and Hydrologic Isolation Options
Remedial
Option
Treatment With-
out Hydrologic
Isolation
Slurry
Wall
Option 1
Slurry
Wall
Option 2
Pump-
ing
Rate
(GPM)
25
50
100
25
50
100
25
50
100
Treatment Periods in Years Re-
quired for Various Aquifer
Volumes
1 3 5
4.3
2.2
1.1
2.1
1.0
0.5
1.0
0.5
0.25
12.9
6.6
3.3
6.3
3.0
1.5
3.0
1.5
0.75
21.5
11.0
5.5
10.5
5.0
2.5
5.0
2.5
1.25
Notes: 1. A 1-volume treatment period represents time required to
pump a volume equal to the initial contaminated aquifer
volume at given pumping rate.
2. Treatment periods represent times required to treat given
multiples of initial contaminated aquifer volumes. For treat-
ment without hydrologic isolation, the initial volume is 60
million gallons; for slurry wall option 1, the initial volume is
30 million gallons; and, for slurry wall option 2, the initial
volume is 15 million gallons.
The critical aspect of the slurry wall alternative is the
degree of contaminant isolation achieved. The major de-
sign consideration for the slurry wall will be to formu-
late a slurry mixture that will have not only the necessary
very low permeability, but also the ability to resist chem-
ical reaction and breakdown by the organic contaminants.
The installation of an impermeable surface over the iso-
lated area will prevent the build-up of liquids inside the
slurry wall, thereby preventing development of a hydraulic
head to promote leakage. Proper design and installation
of the isolation system should result in minimal leakage
through the slurry wall, and any leakage that does occur
would be diluted by large volumes of groundwater.
It is anticipated that any leakage of contaminants
through the bedrock underlying the recommended isola-
tion area will be minimal due to absence of a driving
hydraulic head. However, only an extensive program of
bedrock drilling and permeability testing would reveal the
quantities of flow (if any) from the area through bed-
rock units. Bedrock grouting was not recommended since
the available data did not justify the expenditures that
would be necessary.
Cost Effectiveness Analysis
To evaluate the cost-effectiveness of the recommended
remedial action plan, the methodology used by the U.S.
Environmental Protection Agency for municipal waste-
water treatment projects was employed. Using this stand-
ard approach, the various remedial action alternatives for
the Sylvester Site were evaluated in terms of their com-
parative present worth.
To provide a common basis for comparing the costs of
the various treatment system alternatives, it was assumed
that groundwater treatment would be provided for at least
as long as necessary to treat five multiples of the initial
contaminated aquifer volume (see Table II). This assump-
tion was necessary in order that each alternative slurry
wall/treatment system have an equivalent remedial bene-
fit (i.e., treat five initial aquifer volumes). This assump-
tion also establishes the approximate time period over
which the treatment plant would have to be operated under
each alternative, and thus sets the estimated number of
years for which operating and maintenance costs would be
incurred.
Using this methodology, the costs of the various al-
ternatives were estimated on a comparative present worth
basis; the results are presented in Table III. The present
worth cost comparisons indicate that the recommended re-
medial action plan (Item 5a.) is the most cost-effective al-
ternative.
By basing the comparative cost analysis on treating five
multiples of an initial contaminated aquifer volume, the
evaluation has ignored aquifer recharge through rainfall
infiltration. In actuality, infiltration outside of the slurry
wall and cap system will probably require a 50 gal/min
interception rate under Option 1. As shown in the above
tabulation, even if a 50 gal/min plant is combined with
slurry wall Option 1 and is operated for only 5 years, the
recommended Option 2 system still has a lower present
worth. With the Option 2 slurry wall and cap, a 25 gal/
min pumping rate will be sufficient to intercept the con-
taminant plume.
CURRENT STATUS
The recommended remedial plan was selected for imple-
mentation in late August 1981, the design phase of the
slurry wall and cap isolation system was being undertaken
and plans for pilot-scale treatment system testing were be-
ing formulated. Also, an interim (emergency) groundwater
interception and recirculation system to reduce contam-
inated groundwater discharges to Lyle Reed Brook was be-
ing designed. These design efforts are being conducted
-------
370 CASE HISTORIES
Table in.
Summary of Remedial Action Alternatives and Estimated
Costs, Sylvester Site, Nashua, New Hampshire
DESCRIPTION OF REMEDIAL ALTERNATIVES
ESTIMATED COSTS
Total removal of dump site for secure landfill
disposal out-of-state, and intercept and treat
aquifer for 20 years 9 50 gpm (AS-N/P-BIO-GAC)
Total removal of dump site for disposal in new
secure landfill in NH, and intercept and treat
aquifer for 20 years 9 50 gpm (AS-N/P-BIO-GAC)
Aquifer interception and treatment alone for
20 years 9 50 gpm (AS-N/P-BIO-GAC)
Slurry cut-off wall and clay cap Option 1 to
isolate approximately 7.2 acres, with
a. aquifer interception and treatment
for 10 years 9 25 gpm (AS-BIO-GAC)
b. aquifer interception and treatment
for 10 years 9 25 gpm (AS-N/P-BIO-GAC)
c. aquifer interception and treatment
for 5 years 9 50 gpm (AS-BIO-GAC)
d. aquifer interception and treatment
for 5 years 9 50 gpm (AS-N/P-BIO-GAC)
Slurry cut-off wall and clay cap Option 2 to
isolate approximately 12.5 acres, with
a. aquifer interception and treatment
for 5 years f 25 gpm (AS-BIO-GAC)
b. aquifer interception and treatment
for 5 years t 25 gpm (AS-N/P-BIO-GAC)
c. aquifer interception and treatment
for 3 years 9 50 gpm (AS-BIO-GAC)
d. aquifer interception and treatment
for 3 years 9 50 gpm (AS-N/P-BIO-GAC)
$ 19,000,000 to $ 28,500,000
$ 13,000,000 to $ 19,500,000
$ 9,000,000 to $ 13,500,000
$ 3,420,000 to $ 4,850,000
$ 5,170,000 to $ 7,500,000
$ 3,110,000 to $ 4,360,000
$ 4,690,000 to $ 6,740,000
$ 2,770,000 to $ 3,810,000
$ 3,850,000 to $ 5,430,000
$ 2,720,000 to S 3,670,000
$ 3,780,000 to $ 5,280,000
Notes: 1. Costs are comparative present worth values.
Present worth of estimated annual treatment plant operating and
1.
2.
3.
4.
5.
maintenance costs were computed using an interest rate of 7-3/8
percent as provided by U.S. EPA, Region I, Boston, MA.
Treatment plant assumed to operate for number of years necessary
to treat 5 multiples of the initial contaminated aquifor volume.
Salvage values assumed to be zero.
AS « air stripping; N/P • neutralization/precipitation for metals
removal; BIO = biological lagoons; GAC = granular activated carbon
adsorption.
by GHR Engineering Corporation, again in conjunction
with Goldberg-Zoino and Associates.
ACKNOWLEDGEMENTS
The successful investigation of the Sylvester Site re-
quired a high level of cooperation and commitment from
many individuals. While it would be inappropriate to iden-
tify all major contributors to this effort, it is appropriate
to acknowledge the guidance and cooperation received
from Mr. Carl Eidam of the U.S. EPA and Mr. Michael
Donahue of NH WS & PCC.
REFERENCES
1. GHR Engineering Corporation, "Hazardous Waste
Site Investigation, Sylvester Site, Gilson Road, Nashua,
New Hampshire," Final Report prepared for the New
Hampshire Water Supply and Pollution Control Com-
mission and the U.S. Environmental Protection Agen-
cy—Region I Laboratory, July 1981.
2. Recra Research, Inc., "The Characterization of the
Drums of Waste Material Located on the Gilson Road
Hazardous Waste Disposal Site in Nashua, New Hamp-
shire: Operational Summary," prepared for GHR En-
gineering Corporation, July 1980.
-------
CASE HISTORIES IN HANDLING UNKNOWN HAZARDOUS
MATERIALS AT DUMP SITE LOCATIONS
THOMAS F. DALTON
Hazchem Services Inc.
Garwood, New Jersey
INTRODUCTION
The author's purpose in writing this paper was to pre-
sent a series of case histories involving the cleanup of
hazardous dump sites where the type or degree of hazard
or some of the materials handled were unknown.
Each of the three case histories involved a different
set of circumstances requiring decisions and techniques to
determine as close as possible what was the nature of the
problem, how best to handle or control the material and
then, how to dispose of it.
CASE HISTORY #1
CHEMICAL CONTROL—NEW JERSEY
This location, at the beginning of the remedial action,
presented a formidable challenge to state and cleanup in-
dividuals, mainly due to the wide variety of unknown
chemicals and the extreme danger involved in handling
them.
Initially, there were almost 500,000 laboratory size con-
tainers and bottles, indiscriminately mixed together in
drums, cartons or stacked on tables. They ranged from
unknown chemicals to correctly identified materials.
Classification System
The breakdown of each container into a specific cate-
gory, not only for classification purposes, but also for
disposal, created a situation which called for broad group-
ing with many divisions. Into each division, were sub-
divisions, which would allow the grouping of like mater-
ials and make it easy to combine those subdivisions so
they could be disposed of at a secure landfill based upon
the limitations and groupings required by the disposal
company. For example, acids were classified as follows:
Class A. Inorganic
1. Fuming (liquid)
2. Non-fuming (liquid)
Class B. Organic Acids (liquid)
Class C. Crystalline
The disposal protocol called for taking.Class A items
and placing them into Group A (compatibility grouping
of a disposal company) for disposal of laboratory packs.
Class B items would be placed into Group D and so forth.
It would be possible for grouping into any disposal com-
pany's compatability grouping. Following the need for
classification was a necessity for record keeping, samp-
ling and coding.
Record Keeping
This is probably one of the most tedious requirements
at any dumpsite location. Record keeping at a major
cleanup site requires the services of at least a full-time
person and the cooperation of everyone.
All of the items that were located which were consid-
ered extremely hazardous (highly toxic, very unstable,
explosives or highly reactive) had to be noted in the log
book along with those items requested by the state or
federal agencies which were followed with a chain-of-
custody (evidentiary) protocol.
Coding of Drums
The coding of drums is not a problem when there are
only several hundred drums or batches of drums with the
same characteristics. When there are 50,000 to 60,000
drums with many different physical and chemical proper-
ties, then one has to establish a more sophisticated sys-
tem.
At Chemical Control, a six digit coding was worked
out with the following design. Starting inside the build-
ing with #1, then going in a clockwise position, Front #2,
Back #3, Right Side #4, Left Side #6. Any materials
from a source off the property would be #6. The next 5
numbers were the drum numbers for samples, identifica-
tion and statistical purposes. In addition, each disposal
location was coded with a letter beginning with A (for
SCA). This system resulted in a drum so marked that one
could determine exactly where it came from and where it
was going. For example: DRUM #201109A. This de-
signation meant that the drum came from the front of
the building, was drum #1109 and was sent to SCA for
disposal.
It is important to set up a clear, simple means of review-
ing drums for statistical study and also as a means of
tracking the drums.
All of the laboratory pack drums were packed with all
bottles or containers identified and grouped into classes.
Each drum had a list of contents which could be verified
by subsequent inspection of the disposal company. This
made for rapid evaluation by the disposal companies af-
ter the lists had been checked and verified. Once the
system was set up, the transportation, checking and dis-
posal preceded quite rapidly.
371
-------
372 CASE HISTORIES
Sampling
Sampling was a tedious and difficult project to set up
and follow due to the large numbers of drums and solid
materials that were sampled. Various sampling devices
were used, ranging from a 40 in. by 3/8 in. PVC tube
for drum samples to turkey basters made of plastic for
the semi-liquid materials. There were also composite
samples for use in a CG unit, which were collected in an
8 oz. bottle, then transferred to a 30 ml plastic container.
It was difficult to write a six digit number with a wax
crayon on a vial this small, so numbers had to be written
on a plastic bag in which the vial was placed. Solids and
liquids were collected in a variety of ways and coded with
the same numerical system used on the drums.
Unexpected Situations
There was always a surprise or two each week at the
site. Sometimes there were cyanide drums or acid drums
leaking, evolution of H2S gas near the building, drums
containing metallic sodium and sodium anhydride would
explode after a heavy rainstorm, malodorous leaks of
multi-colored oil, greases and plastics periodically found
their way out of the stacks of drums.
After a few of these episodes, it became a regular prac-
tice of the local Fire Department to find out if they could
use water or sand and soda ash on a fire and what type of
safety equipment should be utilized.
An explosion and fire occurred at the site about one year
after the work commenced (and has been described in an-
other paper in these Proceedings). Today it has been
cleaned up and no longer is a threat.
CASE HISTORY #2
(PICILLO FARM, RHODE ISLAND)
This situation was different from Chemical Control be-
cause most of the drums were buried or partially covered
in several large trenches filled with several feet of water.
The area was remote and access to the site itself was lim-
ited. Working with state officials to determine the best
plan to remove some drums that appeared to be sodium or
aluminum anhydride called for an initial delay of six to
eight weeks until the ground could dry up.
Sodium anhydride, when contacted with water, reacts
violently with a tremendous heat release. Hydrogen gas
is also released and a hot mixture containing caustic
soda remains. When water is dumped into sodium anhy-
dride, it is impossible to control the speed of reaction,
hence it reacts in an explosive-like manner. For this
reason, it was necessary to wait until the water had dis-
appeared and the weather improved. The hydrogen gas, if
not dissipated, could collect and explode.
The plan, worked out with the State Department of
Environmental Management, was to remove the drums
from the ground, sample and identify the contents,
over-drum, if necessary and store for future disposition.
The drums were removed mechanically after the ground
dried out and the surface soil was removed, if necessary,
sampling was done with a 40 in. thief, the sample placed
in plastic sample jars with screw caps. Analysis indicated
both used and unused sodium anhydride were in the
drums.
Temporary storage was in a box trailer with more
permanent storage at an old ammunition silo at the Na-
tional Airguard Air Station.
There were two options for disposal:
(1) React under controlled conditions to end products
which could be disposed of locally. Cost between
$350-$500/drum.
(2) Ship to an approved disposal site located in Nevada.
Cost of disposal and transportation approximately
$800-$l,000/drum. This method was ultimately
selected due to the limited number of drums and
disposal out-of-state.
Later on, as the monies became available, a more ex-
tensive cleanup at Picillo Farm was initiated by the state
of Rhode Island. Most of the drums were buried or parti-
ally buried in shallow cells scraped out by a bulldozer,
then dirt was pushed over the drums. Due to the geology
of the area, the cleanup had to be done quickly to prevent
both surface and groundwater contamination.
CASE HISTORY #3
(KIN BUC LANDFILL—NEW JERSEY)
This landfill situation differed from the other two be-
cause all of the drums and materials were completely
covered with soil to a height of 80 ft above grade. The lo-
catioan of the site was also a problem, since there were no
good access roads to the area in which the work had to be
performed.
Initially, the plan was to treat oily wastes which were
leaching out periodically and finding their way through
the marshlands to the Raritan River. The USCG and EPA
had monitored this situation and decided to attempt to
control the leachate which resembled an ongoing oil
spill.
Workers sampled the deposits which were sent for analy-
sis. In the meantime, since there was some old technical
data available, it was examined as a logical safety protocol.
GS/MS reports on the old samples showed 50 to 60 dif-
ferent organics. The atomic absorption reports indicated
the presence of heavy metals, especially those highly tosic
to humans and water supply systems: lead, cadmium,
copper, chromium, zinc, barium and lanadium.
The old reports also indicated small amounts of PCBs,
along with many benzene ring homologs. This information
indicated that the initial work should be done with full
face respirators, as well as protective suits, gloves, boots.
Later on, as the level of PCBs and other phenolic and
naphthenic compounds were quantified, the workers had
to change their safety protocol.
The first step was to remove debris, set up filter fences
and install both a disposable and a permanent oil spill
boom. Due to the location of some of these boom and fil-
ter barriers, which had to be monitored daily, small boats
were used to negotiate the marshlands. This was especially
true after a heavy rain when the leachate volume in-
creased.
-------
CASE HISTORIES 373
When the volume of leachate began to reach 2 to 3 gal/
min with a well defined black organic layer, skimming into
drums was done in an effort to remove more of the or-
ganic layer of leachate instead of allowing it to course its
way through the swampy marshlands into the Raritan
River. The work recently has been accelerated and modi-
fied in an attempt to:
(1) Concentrate the leachate
(2) Treat the leachate, in situ
(3) Consider long term cleanup with either carbon or
incineration
With a potential for a high flow rate of leachate, de-
pending on local hydraulic conditions, Kin Buc will be a
problem of immense proportions for years to come. Cre-
ating Kin Buc took many years, cleaning it up will take as
long, if not longer.
SUMMARY
The three case histories, presented, represent hazardous
dumpsites in which the materials were:
(1) Above ground
(2) Partially buried
(3) Completely buried
While it is an oversimplification to state that each and
every landfill has to be approached differently, there is
no doubt that working with materials that are above
ground is easier, although there is an inherent danger of
spills and percolation into groundwater.
Buried materials still present a strong problem of
aquifer contamination, however, this process may be
slowed down depending on the nature of the materials,
the condition of the containers or drums and the percola-
tion rate of the soil surrounding the material. Those
drums, partially buried, can exhibit a higher corrosion rate
due to the metal/ground interface which causes a higher
concentration of electrolytes to occur at the interface.
Of the three sites mentioned, the last one (Kin Buc) is
still ongoing with a great deal of work to do. The capping
of the dome-like dump reduced the percolation through
the infrastructure of the dump. The problem of a varying
effluent occurs with higher concentrations of organics
and soluble inorganics in the leachate. This produces a
paradoxical situation in that there is less leachate, but of
higher toxicity.
-------
CONTAMINATION CONTROL AT ROCKY MOUNTAIN
ARSENAL, DENVER, COLORADO
JOHN WARDELL, Ph.D.
MARGOT NIELSON
JUDITH WONG
U.S. Environmental Protection Agency
Region VIII
Denver, Colorado
INTRODUCTION
In response to requirements to contain or clean up
contamination at Rocky Mountain Arsenal (RMA), Den-
ver, Colorado, the Department of the Army (DA), began
a program to mitigate the contamination problem at that
installation in 1975. The approach initially taken by DA
was to contain contamination at the installation's boun-
daries.
In 1978, a pilot wastewater treatment system began
operations at the ridrthern arsenal boundary to remove
organic contamination from ground water. The pilot
system proved successful. It has been expanded to in-
tercept all contaminated groundwater crossing that boun-
dary. A similar system is also being designed to intercept
all contaminated water crossing the northwest boundary.
This paper describes the history and operation of these
containment systems.
HISTORY OF THE CONTAMINATION PROBLEM
Established in 1942, RMA occupies 17,000 acres of land
10 miles northeast of Denver, Colorado. The northern
and eastern borders are bounded primarily by agri-
cultural land. Agricultural lands, an industrial area, and
a housing development are located adjacent to the
southern boundary. Residential areas are located along
the installation's western boundary (Figure 1).
Within eight square miles north of the northern ar-
senal boundary, the land is primarily agricultural. Popu-
lation density is 10-15 people per square mile. Water is
supplied by approximately 90 wells. Thirty percent of
these wells use water from the aquifer affected by con-
tamination from the arsenal. Irrigation, stock, commer-
cial and domestic wells are affected.
Constructed to support World War II, RMA manu-
factured and assembled toxic chemical end items and
incendiary munitions. At the end of the war, RMA was
placed on standby status. A small number of activities
including demilitarization of obsolete hazardous and
toxic munitions occurred at the arsenal until the begin-
ning of the Korean War.
In 1946, parts of the installation were leased to private
industry for chemical manufacturing. The major lessee,
Shell Chemical Company (Shell), had used a large por-
tion of these original chemical manufacturing facilities
since 1952. That company has made additions to their
leased facilities for manufacturing pesticides.
The arsenal was selected as the site to construct a fa-
cility to manufacture GB, a highly toxic, non-persistent
nerve agent. This production facility was completed in
1953. Manufacturing of this chemical agent occurred un-
til 1957, but filling of the munitions continued until 1970.
RMA produced a biological anticrop agent from 1959-
1962, and emptied cyanogen chloride and phosgene bombs
between 1956-1969.
Unlined Lagoon Used
Industrial waste effluents resulting from past DA and
lessee manufacturing activities were initially discharged
into an unlined basin immediately north of the original
"plants area" called Basin A. Basin A received these in-
dustrial wastes until 1957. When the GB facility was
placed in operation, this basin was enlarged and new un-
lined basins were constructed northwest of Basin A
(Figure 2).
In the summer of 1954, some farmers northwest of the
arsenal complained that groundwater used for irrigation
had damaged their crops. Precipitation was much less
Figure 1.
Vicinity Map of RMA
374
-------
CASE HISTORIES 375
Cm OF DENVER
Figure 2.
Location of Features at RMA
than normal in 1954 and increased pumping from shal-
low irrigation wells occurred to support crop production.
In response to these complaints and subsequent claims
for damage, DA:
(1) Retained a firm of consulting engineers to investi-
gate the problem of possible groundwater contam-
ination described by the farmers.
(2) Requested the U.S. Geological Survey (USGS) to
study water quality on RMA and in adjacent farms.
(3) Contracted the University of Colorado to initiate
plant bioassay, geological and chemical studies to
determine the identification and source of any con-
taminants causing crop damage.
Some stock and irrigation wells were abandoned be-
cause of high salinity, and compensation was paid to a few
landowners for crop damage. Evidence suggested that
high salinity resulting from groundwater migration from
unlined basins on RMA was the cause of the problem.
New Storage Constructed
As a result of this investigating, Basin F was construct-
ed. Built to hold 240 million gallons, it covered 93 acres.
It was lined with an asphalt liner covered with one foot of
soil. Upon completion in 1957, all industrial wastes pro-
duced by DA or Shell at the arsenal were deposited in
Basin F. An attempt was made to transfer residues from
the unlined basins into Basin F (Figure 2).
Contaminants Detected
In 1974, organic compounds were detected crossing the
northern arsenal boundary. Diisopropylmethyl phospho-
nate (DIMP) and dicyclopentadiene (DCPD) were de-
tected in surface water draining from a man-made bog at
the north boundary of the installation. DIMP is a per-
sistent material produced in small quantities during pro-
duction of the GB nerve agent. It is not very toxic, but it
is an excellent indicator compound because it originates
from a single activity. DCPD is a chemical used by Shell
in production of some of its pesticides. Though not very
toxic, it has an extremely foul odor at extremely small
concentrations making potable water in which it is con-
tained unfit for consumption.
DA initiated several actions after detecting these com-
pounds. A previously existing groundwater monitoring
program was expanded. Tests for DIMP, DCPD, and
several other compounds were added. These latter com-
pounds were chemicals that could have been disposed of as
a result of manufacturing activities at the arsenal. A dike
was built to stop off-post discharge of surface water
from the bog. The toxicological impacts of DIMP and
DCPD were also evaluated.
In December 1974, the Colorado Department of Health
(CDH) detected DIMP at less than one /xg/1 in an off-site
well. The detection of both DIMP and DCPD in off-site
surface water and subsequent discovery of DIMP in off-
site groundwater led to issuance of administrative orders
by the CDH in April 1975. These orders required an im-
mediate stop to off-post surface and subsurface dis-
charge of these contaminants, preparation of a plan to
prevent their further discharge and implementation of a
water quality monitoring program to demonstrate compli-
ance with the first two requirements.
Subsequent to issuance of these administrative orders,
other organic compounds have been detected at the
northern boundary. The most significant of these com-
pounds is dibromochloropropane (DBCP), a soil fumi-
gant manufactured by Shell. DBCP has been reported to
cause male sterility and is a possible carcinogen. A drink-
ing water standard has not yet been established for this
chemical.
In May 1980, contamination of groundwater off-post
northwest of the arsenal (Figure 1) was detected in wells.
Potable water sources contained between 0.2-0.4 /ug/1
DBCP.
APPROXIMAl
OFF-SITE
CONTAMINATION
COMME RCE
CITY
GROUKDUATER FLOWS
CITY OF DELIVER
Figures.
Estimated Divide between Northwest and
North Groundwater Flows
-------
376 CASE HISTORIES
GEOLOGY AND HYDROLOGY
The bedrock consists of shale and claystone with lenses
of siltstone and sandstone. Alluvium and windblown de-
posits cover this bedrock. Silty clay, silt, sand and gravel
form these unconsolidated deposits of alluvium. In many
places, wind-deposited silts and sand cover this alluvium.
The alluvium ranges in thickness between 5-80 feet. The
alluvium contains the majority of the highly permeable
materials."'3'61
Based on previous geohydrologic studies, groundwater
flow is believed to be continuous through the shallow
RMA aquifer.<4|6) If the alluvium is dry, water flows in the
upper bedrock formation.<3|6) North of the original "plants
area," water flows in a northerly, northwesterly and
northeasterly direction. Water underneath Basin A flows
to either the northwest or north boundary of the ar-
senal(4'6) (Figure 3). Basin A received large quantities of
potentially hazardous contaminants.
Groundwater flow across the northern boundary occurs
in both the alluvium and sands in the bedrock. The aquifer
is defined as all sand and gravel, gravel and sand units
that are unconfined above an impermeable barrier. The
alluvium and sands are hydraulically connected and should
be considered to be one hydrogeologic unit. Groundwater
flow, however, occurs primarily in the alluvium because
permeability values in the coarser sand and gravel strata of
the alluvium is much greater than the permeability of the
silty and clayish sands.<5)
Permeabilities of sand, and sand and gravel units were
estimated to be 3,000 gpd/ft2 (400 ft/day) and 5,000 gpd/
ft2 (668 ft/day), respectively, based on aquifer tests. Units
containing appreciable quantities of silt had permeabilities
of an order of magnitude less. Units containing clay had
negligible permeability compared with permeability of
sand, and sand and gravel units.7
Aquifer thickness varied from zero on bedrock highs to
about 40 feet about 1 mile south of the northern arsenal
boundary.(3i8) Analysis of cross-sections indicated that
this aquifer is generally continuous.(3|8) The aquifer thick-
ness consists of permeable materials above and below the
water table. The saturated thickness was determined by
calculating the difference between the potentiometric and
bedrock surface. The potentiometric surface was initially
evaluated in 1978, before construction and operation of
the contamination containment system. It was generally
toward the north with an average gradient of 0.006 ft/ft.8
Subsequent measurements taken in 1979, after the contam-
ination containment system had begun operation, showed
similar flow characteristics. Gradients between 0.006 to
0.008 ft/ft were observed along the north boundary.0"
The clearest picture of contamination migration and
distribution affecting the north boundary can be provided
by considering the underlying aquifer as two separate
bodies of groundwater. The most significant body moves
beneath Basin F toward the north boundary in a north-
easterly direction. Contaminants leached from the surface
waste basins (e.g., Basin A) are carried by this subsurface
flow across the northern arsenal boundary. The other body
of groundwater flows northwesterly parallel and beneath
First Creek. It is generally free of contamination and has
much higher volume than the body of groundwater flow-
ing underneath Basin F. In the vicinity of the north boun-
dary, water table contours of both bodies become parallel
to that boundary. Contamination, therefore, is carried di-
rectly north across the boundary.'3'8'
IMPLEMENTATION OF CONTAMINATION
CONTROL PROCEDURES
Introduction
Contaminated groundwater has been detected crossing
both the northern and northwestern arsenal boundaries
(Figure 2). Contamination crossing the northwestern
boundary was detected in 1980. Consequently, contain-
ment strategies there are only presently being designed.
The focus of this discussion, therefore, is on containment
strategies undertaken along the northern boundary.
A two-step approach was implemented. A pilot contain-
ment system was installed to evaluate the feasibility of the
overall approach. Once satisfied that the pilot system
could be operated successfully, the containment system
would be extended to intercept and treat contaminated
groundwater crossing along the entire affected part of
the northern boundary.
Pilot Containment System
The pilot containment system was designed to demon-
strate the feasibility of containing off-post groundwater
contaminant migration across the northern arsenal
boundary. It was placed in operation in July 1978 and its
performance evaluated in the summer of 1979.(1)
The pilot containment system is composed of:
(1) Dewatering well subsystem
(2) Treatment plant
(3) Recharge well subsystem
(4) Impermeable barrier
(5) Monitoring well subsystem
For a flow schematic of the system and plant layout
see Figures 4 and 5 respectively.
TREATlffitn SYSTEM
Figure 4.
Schematic Diagram of Pilot Containment System
-------
CASE HISTORIES 377
X
X
A
S\
HITORINC IJELLSXN
HHH
TREATMENT FACILITY
DEUATERIHG WELLS
Premature filter plugging is indicated by a high pressure
drop alarm. Backwashing is done one filter at a time using
filtered water from the onstream filter for a pre-estab-
lished time period. The backwashed filter is automatically
returned to service when this operation is finished. The
particle-laden backwash water is collected in a sump
where the particles are allowed to settle out. The de-
canted water is reprocessed through the treatment plant.
Pressure is used to move water from the particle filters
to one of two columns packed with 20,000 Ibs of acti-
vated carbon (Figure 6). Only one column is used at a
time. Water flows down through the column and the
activated carbon packed into the column absorbs the or-
ganic contaminants. The treated water flows to the re-
charge wells by gravity for reinjection into the aquifer.
Figures.
Overview of Pilot Containment System (not to scale)
Flow of contaminated water across the arsenal boun-
dary is stopped by the impermeable barrier. Dewatering
wells remove water from the aquifer for treatment in the
treatment plant. Recharge wells inject water back into
the aquifer.
Dewatering Well Subsystem
This subsystem is composed of six 8-inch diameter wells
placed within 30-inch diameter gravel-packed holes. The
wells are screened through the entire alluvial aquifer thick-
ness. They are approximately 225 feet apart. Each well
has been provided with a flow control system and sub-
mersible pump. The pumping system was designed to keep
a constant head within each well. Sensors were used to
recycle water back into the well instead of to the treatment
system when the pumping water level fell below a prede-
termined level.
Treatment Plant Subsystem
The contaminated water removed by the dewatering
wells is discharged into a sump. This water is then pumped
through a filter to remove suspended solids. The water is
passed through an activated granular carbon column to
remove organic contaminants (e.g., DIMP, DCPD,
DBCP). The treated groundwater is discharged to the re-
charge well system and then back into the aquifer by grav-
ity drainage.
Contaminated water is pumped through the treatment
system at a controlled rate (design flow is 10,000 gal/hr).
A flow control system regulates flow to the treatment
system from the sump on the basis of water level in the
sump. Contaminated water is filtered to remove particles
before being passed through the packed carbon column.
Two filters are used. Each filter contains four feet of
filter media. This media is a blend of graded coal and sand.
The filters are operated in parallel.
Backwashing is required periodically because the filters
become plugged with particulates. Backwashing can be
done manually or automatically on a preset time interval.
Figure 6.
Treatment Facility Prefilters (left) and Adsorbers (right)
Spent carbon must periodically be replaced with fresh
activated carbon which is delivered to the treatment fa-
cility in specially designed trailers. Transfer of the carbon
from this trailer to the column is accomplished by filling
the trailer with treated water to slurry the fresh carbon.
The treated water had been stored in the empty column
prior to arrival of the trailer. Spent carbon is transferred
to and from the trailer as a slurry by pressurizing the
column or the truck with compressed air. The slurry
water is drained into the sump and then processed through
the treatment system.
A major consideration in deciding to use carbon ad-
sorption columns was the ease of operation and small
manpower requirements. Downtime of the treatment fa-
cility was less than 1% during FY-79. Only simplistic
problems typical with system start-up were encountered.
Flow recorders did not operate properly, pump seals
needed replacement and a faulty solenoid valve required
repair.
Breakthrough
To determine when the carbon in a column requires
replacement, the concentration of DIMP in water being
-------
378 CASE HISTORIES
discharged from a column is monitored. When the DIMP
concentration approaches 50/ig/l, new carbon is ordered.
This level was selected so that the column could still be
used until the trailer containing fresh carbon arrives.
DIMP was selected as the indicator compound because
previous testing had shown that it was the compound
that "breaks through" the carbon before any of the other
organic compounds. To monitor DIMP passing through
the column, samples of influent and effluent were col-
lected daily. An example of a typical DIMP "break-
through" curve is given in Figure 7.
Pilot bench studies had suggested that the carbon needed
would be 1.10 Ib carbon/1000 gal of wastewater. Be-
tween July 1978 and June 1979, two columns were
changed. The ratios were 1.90 and 1.29 lb/1000 gal of
wastewater, respectively. The higher than expected car-
bon usage rates were attributed to bed siphoning and
subsequent formation of air pockets in columns packed
with carbon. To overcome this problem a siphon break
was installed.'"
Recharge Well Subsystem
The recharge well subsystem is composed of twelve 18-
inch diameter wells approximately 100 feet apart. They are
installed within 36-inch diameter gravel-packed holes.
These wells are screened along the full thickness of the
aquifer. Treated water is injected continuously by gravity
unless the level of water in the well rises above a pre-de-
termined level. At that time, a float control valve closes,
halting flow into the well until the water level drops
enough for flow to resume.
Impermeable Barrier
A 1500 foot impermeable bentonite barrier anchored
approximately 2 ft into bedrock was installed between the
dewatering and recharge wells. It was installed to physi-
cally prevent flow of the groundwater through the aquifer,
isolating upgradient and downgradient flow at the treat-
ment system site. This barrier prevents mixing of contam-
inated and treated groundwater.
Monitoring Well Subsystem
Observation wells were installed both upgradient and
downgradient of the pilot containment system. They were
small diameter PVC casing screened within the alluvial
aquifer. Water levels and chemical quality of the ground-
water are monitored periodically at each of these wells.
Evaluation of System Performance
In June 1979, the performance of the pilot containment
system was evaluated to determine its ability to remove
organic contaminants from the groundwater. If the sys-
tem operated successfully, it would be extended to inter-
cept and treat contaminated groundwater crossing along
the northern arsenal boundary/"
The results of the evaluation were:
(1) The carbon column removed organic contaminants
from the groundwater. The contaminants could not
be detected using gas chromatography. Eventually,
contaminants were detected because the adsorptive
= 500 -
REPLACED CARBON
REPLACED CARBOI
10 10 30 40 50 60 9&
WEEKS
Figure 7.
Graph of DIMP Breakthrough Curve During
Pilot System Operation
capacity of the activated carbon became over-
whelmed. Therefore, successful operation requires
periodic replacement of the activated carbon.
(2) Downgradient flow of groundwater was essentially
unchanged. This latter consideration is as impor-
tant as the first requirement in a water-scarce area.
Expansion of the North Boundary
Pilot Containment System
Construction of the expanded containment system was
begun in early 1981. It is scheduled to begin operations in
August of 1981. It consists of 42 additional dewatering re-
charge wells, and monitoring wells, increased treatment
capacity, and extension of the bentonite barrier 4500 feet
to the east and 1500 feet to the west of the present pilot
containment system barrier. The design of this addition
was based upon the performance of the original pilot
system.<2)
Northwest Boundary Containment System
As described earlier, DBCP contamination was detected
off site at the northwest arsenal boundary. A containment
system is being designed to address that problem. Present-
ly, the situation is similar to the north boundary system
with one exception. Preliminary calculations indicated that
the dewatering wells alone could effectively remove con-
taminated groundwater from the aquifer. Two rows of de-
watering wells are planned approximately 800 feet apart.
Therefore, an impermeable barrier was not included in
design of the system.
SUMMARY AND CONCLUSIONS
DA has implemented a containment strategy to prevent
off-site migration of contaminants at RMA. Initially, a
pilot containment system was constructed along the north
boundary to test the feasibility of this containment ap-
-------
CASE HISTORIES 379
proach. The pilot containment system demonstrated that
the approach:
(1) Removed organic contaminants of concern to below
detectable units from the groundwater
(2) Did not adversely affect the flow and distribution of
the groundwater downgradient from the system
As a result, an extension was designed and is being con-
structed to treat contaminated groundwater along this
boundary. The concept is also being implemented to con-
tain contaminants in groundwater crossing the north-
west boundary.
Though this approach works, there are important ques-
tions remaining to be answered:
(1) The length of time required to run the system must
be determined. Containment could cost more than
clean up in certain situations because of the costs
associated in operating a containment system for
many years. The maintenance associated with op-
erating a system for many years must also be con-
sidered.
(2) The impact on land use options caused by use of a
containment strategy needs to be addressed. In the
arsenal's example, sale of land may be precluded
unless the new owners accept responsibility for op-
eration of the system. Use of contaminated land is
also restricted during the time frame that contamin-
ants continue to migrate from that land.
REFERENCES
1. D'Appolonia Consulting Engineers, Inc., "Evaluation
of North Boundary Pilot Containment System, RMA,
Denver, Colorado," Project Number RM79-389,
July 1979.
2. D'Appolonia Consulting Engineers, Inc., "Conceptual
Design of the North Boundary Containment System,
RMA, Denver, Colorado (Expansion)," Project Num-
ber RM79-480, October 1979.
3. Kolmer, J.R. and Anderson, G.A., "Part 1— Pilot
Containment Operation, Final Environmental Impact
Statement, Installation Restoration of Rocky Moun-
tain Arsenal," Department of the Army, Office of the
Project Manager of Chemical Demilitarization of In-
stallation Restoration, July 1977.
4. Konikow, L.F., "Hydrogeologic Maps of the Alluvial
Aquifer In and Adjacent to the Rocky Mountain Ar-
senal, Colorado," U.S. Geological Survey Open File
Report 74-342, 1 sheet, 1976.
5. May, J.H., Tompson, D.W., Law, P.K. and Wahl,
R.E., "Hydrogeologic Assessment of Denver Sands
Along North Boundary of Rocky Mountain Arsenal,"
U.S. Army Engineer Waterways Experiment Station,
January 1980.
S.G., "Digital Model Study of Diisopro-
pylmethylphosphonate (DIMP) Groundwater Con-
tamination, Rocky Mountain Arsenal near Denver,
Colorado, Progress Report— Phase I," U.S. Geologi-
cal Survey, Denver, Colorado, June 1977.
7. Vispi, M.A., "Report of Findings, Rocky Mountain
Arsenal Pumping Test," U.S. Army Engineer Water-
ways Experiment Station, 1978.
8. Zebell, R.A., "Basin F to the North Boundary, Volume
I: Geotechnical Definition, Rocky Mountain Arsenal,
Denver, Colorado," U.S. Army Engineer Waterways
Experiment Station, 1979.
-------
THE EFFECTIVE USE OF RESOURCE RECOVERY IN THE
CLEANUP OF UNCONTROLLED HAZARDOUS WASTE
SITES—BASED ON THE CALIFORNIA EXPERIENCE
WILLIAM QUAN
California Department of Health Services
Berkeley, California
INTRODUCTION
One of the most important environmental concerns to-
day is the location and management of uncontrolled haz-
ardous waste sites. Normally, an uncontrolled site is iden-
tified through anonymous tips and/or governmental sur-
veillance activities. Once the location of such a site be-
comes widely known there is much pressure exerted by
the news media and the public on governmental agen-
cies to remove the hazardous waste from the site as
quickly as possible.
Very often the most expeditious way of cleaning up a
site is removal of the waste for disposal at a permitted haz-
ardous waste disposal site; this is not necessarily the most
prudent way of solving the problem. Furthermore, Cal-
ifornia and the rest of the country are now experiencing
increasing difficulty in siting new hazardous waste dis-
posal sites. Concurrently, permitted sites are facing in-
creasing public pressure to close. For example, even after
obtaining 185 permits the five-year-old Earthline haz-
ardous waste disposal site in Wilsonville, Illinois was
closed as a public nuisance."1
Therefore, if the limited number of currently operating
hazardous waste disposal sites are filled, the states will be
faced with frequent incidents of illegal dumping of haz-
ardous wastes and/or firms going out of business because
there is no facility which can accept their waste.
In addition, until recently, only surface and ground-
water protection were generally considered in the location
of a hazardous waste disposal site, but now there are also
concerns about air pollution and public health effects due
to emissions of hazardous waste from disposal sites. Re-
cently, several papers'2'3t 4) have addressed the problems of
air quality deterioration and the control of air emissions
from hazardous waste sites.
According to a 1980 EPA contracted study(5) of nine un-
controlled hazardous waste sites in the United States, the
remedial action technologies usually employed were on-
site containment and surface/groundwater monitoring,
and waste removal for landfill burial or incineration at
permitted facilities. Only once was resource recovery em-
ployed as part of a cleanup program. In the long run, even
though it usually takes a longer period of time for a site to
be cleaned up through resource recovery, it is probably, in
many instances, the soundest solution. Sometimes it may
take a joint cooperative effort of everybody concerned for
resource recovery to work. Furthermore, landfill disposal
and/or on-site containment measures may be only tem-
porary solutions.
Four California case studies of uncontrolled hazardous
waste sites are presented. For the purposes of this paper,
an uncontrolled or abandoned hazardous waste site is de-
fined as having one or more of the following charac-
teristrics:
(1) is no longer used or is inactive and is no longer main-
tained,
(2) is a location of unpermitted or improper waste dis-
posal
(3) has no known owner, and
(4) containment and monitoring techniques are inade-
quate.
Hazardous waste is defined in the California Admin-
istrative Code,(6) as any "waste material or mixture of
wastes which is toxic, corrosive, flammable, an irritant, a
strong sensitizer or which generate pressures through de-
composition..."
Due to requests for confidentiality or to pending litiga-
tions, the names and locations of some of the following
studied sites have been deleted from this paper.
CASE A: CONTRA COSTA COUNTY, CALIFORNIA
The site was formerly used by its previous owner for the
fabrication of steel. One of the operations was a galvan-
izing mill, which was used to coat steel pieces with zinc
from a zinc acid solution so as to protect the steel pieces
from corrosion. Apparently the spent galvanizing solu-
tions were ponded from 1960 until 1975 when the plant
closed. The pond was built in existing marshes where soils
vary from clay to sandy silt. In 1977, the property was
sold. The responsibility for the maintenance of the pond
was then transferred to the new owner.
During 1977, the pond overflowed. Subsequently, dikes
were constructed to contain future overflows. In the same
year residents registered complaints about the pond and
emissions from the pond. During 1978, the runoff rate
from the nearby railroad company's fill was periodically
exceeded which led to the pond overflowing once more and
into San Pablo Bay.
Over the years, wastes continued to seep from the pond
and some seepage had even reached the Bay although mon-
380
-------
CASE HISTORIES 381
Table I.
Summary of Analytical Results for
Contra Costa Steel Plant Lagoon
Parameter
pH
Sulfate
Cadmium
HexavaLent Chromium
Trivalent Chromium
Total Chromium
Lead
Zinc
Total Solids
Units
Units
Mg/Kg(l)
Kg/Kg (1)
Mg/Kg(l)
Mg/KgU)
Mj>/Kg(l)
ME/KR (i)
Hg/Kg(l)
Z (2)
Composite Pond
Water Sample
1.9
15,400 Mg/£
0.62 Mg/U
0.02 Mg/i
3.15 Mg/J.
3.15 Mg/i
0.10 Mg/fc
2,530 Mg/1
—
Pond Bottom
Surface Sedijnent
1.5
51,000
0.85
<0.2
18.6
18.6
190
5,570
51.6
CORE SAWLES
Station 1
0-4" 11-14"
2,1
15,000
0.97
<0,2
37.4
37.4
214
2,910
66.9
2.7
2,980
0.68
r0.2
18.5
18.5
23.9
2,050
83.4
Station 2
0-4" 9-13"
2.4
2,650
< 1.00
<0.2
14.1
14.1
191
416
71.8
3.2
1,260
1.60
<0.2
31.3
31.3
28.1
413
80.5
Station 3
0-4" 13-17"
1.8
109,000
0.88
<0.2
34.2
"34.2
26.3
9,650
57.6
1.9
56,000
4.55
<0.2
45.5
45.5
81.9
15,700
56.9
California
TTLC *
*2 (corrosive)
—
10
50
250
--
50
200
~
EPA
EP Toxic Ity**
i2
—
1.0
—
--
5.0
5.0
—
—
(1) Mg/Kg expressed as Wet Weight
(2) Oven Dry (103°C) Weight Basis
* TTLC's (Total Threshold Limit Concentration) are from the California Assessment Manual for Hazardous Wastes, p. 66-67.
** These values are not equivalent to California TTLC but are the same as STLC (Soluble Threshold Limit Concentration) .
© Sampling Station
Date of Sampling 8/13/80
Figure 1.
Core Sampling Points at Pond
itoring of the ground water, which is approximately three
feet below ground, indicated no contamination.
Remedial Actions
The remedial actions employed were basically contain-
ment measures, for example, diking. During these years
there were also proposals to treat the pond wastes for dis-
charge into the sanitary sewage system and San Pablo Bay,
disposal of the waste at a Class I hazardous waste disposal
site(7) or to permit the liquid to evaporate from the pond so
the pond bottoms and adjacent contaminated soils could
be excavated and hauled to a Class I site for disposal.
In 1980, the owners authorized a commercial laboratory
to sample and analyze the pond water and adjacent soils.
This was done to ascertain the alternatives for deposition
of the pond's waste water and the contaminated soils. As
the analytical report (Table I) indicates, three core samples
(the core samples were collected at the locations indicated
in Figure 1), a bottom surface sediment sample, and a
composite water sample were collected and analyzed. Bas-
ically, the results indicate that all samples are considered
hazardous by the California Department of Health
Services per the California Assessment Manual for Haz-
ardous Wastes® or CAM.
Subsequently, the Department's California Waste
Exchange (CWE) became involved. The CWE realized
from the analytical results that the primary chemical com-
pound, zinc sulfate, dissolved in the waste water is used
in the agricultural industry.(9'10)
Sulfur, assimilated as the sulfate ion, is considered a
secondary plant nutrient for protein synthesis and sulfur
deficiencies exist in California soils at the eastern foothills
-------
382 CASE HISTORIES
in the San Joaquin Valley, the central coastal range and
the Sacramento Valley. Zinc, a micro-nutrient, needed for
terminal plant growth, is assimilated as zinc ions. As with
sulfur, zinc is often found deficient in California soils.
A nearby firm, Veale Tract Farms, when contacted by
the CWE, expressed an interest in the three million gallons
of pond water for use as a micronutrient on its marginal
soils. Veale Tract Farms proposed to dilute the pond water
to an appropriate level before application to their soil.
Thus far, of the two proposals received by the owners, one
for hauling and disposal at a Class I site and the other for
hauling and application at Vale Tract Farms, the latter was
less costly by $700,000.
As for the pond bottom surface sediment and the con-
taminated soil, they may also be beneficially used by roto-
tilling at a 1:100 ratio with native soils, so as to reduce the
lead content to an acceptable level. In this way all of the
waste on-site could be recycled.
At this point, the most economic identified remedial ac-
tion technology is still resource recovery. Finally, the Cal-
ifornia Department of Health Services has estimated that
cleanup of the site through disposal at a Class I site can
cost as much as $29,000,000.
CASE B: CONCORD, CALIFORNIA
Prior to 1964 the site was owned and used by a salvage
company operator. In 1964, the site was sold but the sell-
er retained one half of the mineral rights to the eastern
portion of the site where the salvage operation had been
located. Apparently, the operator had used the site over
the years to dispose of 60 to 70 tons of material from the
nearby Allied Chemical's Union Collier Plant. The piles
of material disposed are believed to have been spent coke
filters used in sulfuric acid production (see Figure 2).
In 1972 the U.S. Navy acquired the above site as part of
a two mile buffer zone around a munition's area. During
1980, the site earned to the attention of the California De-
California Department of Health Services
Table n.
Summary of Analytical Data for Solid Waste Piles at
Allied Chemicals' Union Collier Plant
Mg/kg
Parameter
PH
Ag
As
Ba
Bi
Cd
Co
Cr
Cu
Fe
Hg
Mn
Mo
Ni
Pb
Sb
Se
Sn
Sr
Tl
V
Zn
Br
Rb
Y
La
Ge
Th
Te
Ti
Coke Pile
fl
3.1
26 ±14
Coke Pile
12
2.2
89 ±19
251 ±44
31±8
3.800
58 ±20
14±6
16±8
1,600
972
244
36±24
77 ±10
14±4
33±6
37 ±10
19±18
552
48 ±8
1,670
23 ±16
4,120
7,030
23±6
48±8
44±8
Calif.
TTLC
50
1,000
10
150
3,500
200
50
1,000
10
50
240
200
1,550
113±44
Figure 2.
Note: The analytical error for metal determination is approximately ± 10%. unleu otherwtM
indicated
partment of Health Services when it was identified as a
site that had received industrial wastes in the U.S. Con-
gress* EckhardtSurvey.
Preliminary results (see Table II) on samples collected
and analyzed by the California Department of Health
Services indicated that the coke piles were acidic, pH 2.2
to 3.1, and contained high levels of metals: lead, 1600 to
4120 ppm; arsenic, 251 ppm; selenium, 972 to 7,030 ppm;
and tellerium, 552 to 1550 ppm. Since these concentrations
of heavy metals exceeded the Total Threshold Limit Con-
centrations in the Department's CAM for lead and arsenic,
the coke piles are designated as hazardous wastes.
The site is presently unsecured; there are no fences and
based on preliminary information there may be leaching
of heavy metals into the surrounding soil. In addition, the
piles of waste have been cited by the California Regional
Water Quality Control Board (CRWQCB) as a source of
contaminated runoff to tidal and navigable waters. How-
-------
CASE HISTORIES 383
ever, since the site is not adjacent to a residential area
there appears to be no imminent threat to human health.
Remedial Actions
During early 1981, the Department requested the present
owner, the U.S. Navy, to perform extensive sampling and
analysis of the coke piles and the adjacent soils to deter-
mine the average concentration of the contaminants in the
coke piles and the extent of soil contamination or leach-
ing. Additionally, the U.S. Navy was requested by the
CWE to analyze the coke piles for sulfur.
Two California oil refineries contacted by the CWE ex-
pressed interest in the coke piles as a low grade fuel pend-
ing results of the sulfur concentration analyses. The 60-70
tons of coke in the piles is to be blended by the refineries
with their coke so as to produce a low-grade fuel with ac-
ceptable heavy metals and sulfur concentrations, in this
context, the coke piles would be a very small fraction of
the final fuel.
CASE C: OAKLAND, CALIFORNIA
The site near the Oakland Coliseum was formerly used
by Pacific Oxygen, which manufactured acetylene. Acety-
lene is produced from a reaction between calcium carbide
arid water.0 1}
CaC
2H20
»Ca(OH)2 + 134kcal/mole.
The by-product or waste is hydrated lime, Ca(OH)2.
According to the local health agency, the Alameda
County Health Department, the site had been inoperative
since 1976 but prior to that Pacific Oxygen used the site to
dispose of their acetylene sludge, CA(OH)2 (See Figure
3). When the site closed down 1,000-2,000 tons of acety-
lene sludge had accumulated.
In August 1979 a "concerned citizen" contacted the De-
partment via the Alameda County Health Department to
report that children had ventured onto the unsecured site,
the chain fence had been knocked down, to play in the ace-
tylene sludge piles. The site also had an open sump 6 ft
x 8 ft x 3 ft deep which was three-quarters full of liquid.
Subsequently, samples of the acetylene sludge and sump
liquid were collected and analyzed by the California De-
partment of Health Services. The results (see Table III)
indicate that although the acetylene sludge contain rela-
tively low amounts of toxic heavy metals, the pH was
high, 12.3, indicating that the sludge is corrosive and there-
fore a hazardous waste under California regulation. The
pH of the sump liquid was 11.04; therefore it is a likely
irritant, which makes the liquid also a hazardous waste.
In April, 1980, when the City of Oakland obtained title
to the Pacific Oxygen site, the Department's CWE was
contacted for assistance in recycling the wastes that were
on the site. The CWE determined that the wastes may be
recycled in several ways:
•as a soil amendment,
•as an agent to precipitate heavy metals from waste waters
at a sewage treatment plant,
•to formulate a paving material for road construction.
Eventually a local paving company, Gallagher & Burk,
Inc., was located to recycle the wastes. The wastes would
be mixed with rock and gravel to improve the sand equiv-
alent of gravel for use as a road base.
Result
1,000-2,000 tons of lime sludge and liquids were recy-
cled at no cost to the City of Oakland. This was the single
largest waste transfer the CWE facilitated during 1980.
With the sludge and liquids removed, the City of Oakland
California Department of Health Services
Table III.
Summary of Analytical Results for Pacific Oxygen
Mg/kg
Parameter
pH
As
Fe
Mn
Ni
Sr
Zn
Ti
Y
Sump
liquid*
11.04
5
1,033
34
122
Acetylene Sludge**
Pilel
V below surface
12.3±0.1
823 ±81
44±13
186±18
21 ±7
Acetylene Sludge**
Pilel
1 Vz' below surface
12.3±0.1
26±5
139±19
20 ±16
38 ±8
29 + 7
Calif.
TTLC
72 + 52
10 + 5
50
200
200
•Results were reported on 9/12/79
"Results were reported on 7/17/80
Note: Only results for all metals above the detection limit of the X-ray fluorescence spectrometer are reported.
-------
384 CASE HISTORIES
is now proceeding with plans to develop the site into a park
and recreation area.
CASE D: NORTHERN CALIFORNIA
The owner had apparently been using the site for un-
authorized storage and treatment (Figures 4 and 5) because
an investigation by the Department of Health Services'
surveillance and enforcement staff in mid 1980 revealed
that the owner has never applied for a hazardous waste fa-
cility permit. The owner had been, for a period of time,
hauling in drums of ink waste, ink wash, glues, waste oils
and solvents for storage until he could find a recycler.
Otherwise, the waste was hauled to a Class I site.
Although the site was used to store flammable waste,
the Department of Health Services staff saw no fire ex-
tinguishers and safety showers. Because of these and other
deficiencies the State Fire Marshal believed that a fire at
the site would cause the closing of an adjacent freeway.
The above site was also unsecured and unpaved. From
time to time, there had been numerous spills of hazardous
wastes. In addition, glues, ink waste and ink wash wastes
were disposed into unpermitted plastic-lined pits. An
analysis of contaminated soil by the California Depart-
ment of Health Services in mid 1980 showed relatively high
levels of chromium, copper, lead and zinc. Finally, there
was no containment provisions for any spills which could
threaten the pollution of surface and ground waters.
Remedial Actions
In late 1980, the CRWQCB issued a Cleanup and Abate-
ment Order (CAO) to the site owner after inadequate
responses to informal requests for cleanup. The order re-
affirmed previous informal requests to have all removed
materials (soils) disposed at a Class I disposal site. The
second part of the Order required that containment berms
and impermeable surfacing be constructed for all tank
drum storage and handling areas.
Figure 4.
Figure 3.
Figures.
Subsequent inspection by this Department and other
governmental agencies in early 1981 indicated that the
above storage/treatment site posed a clear and present
threat to public health and the environment. Drums of
hazardous waste were seen deteriorating. Accordingly,
the Department urged the local authority to proceed with
the enforcement and cleanup of the site pursuant to the
California Health and Safety Code.m
To expedite the identification of the wastes, firms whose
hazardous waste were observed were contacted. Upon
learning of the situation the Department's CWE provided
names of recyclers who might be of help in the site clean-
up. Eventually, the CWE with the help of recycling con-
tractors, facilitated the recycling of 900 of the 2100 drums
of flammable waste. In addition, 607 drums were recycled.
However, 150 drums of fiberglass and foaming resins
-------
CASE HISTORIES 385
were not recycled primarily because the producer insisted
they were cleanup residues and therefore they be disposed
of at a Class I site as originally intended.
ROADBLOCKS AND REMEDIES
Once an abandoned or uncontrolled hazardous waste
site has been identified, an immediate qualitative assess-
ment of the problem site must be made. Specifically, in-
formation (i.e., the presence of toxic and/or flammable
vapors and toxic heavy metals) must be quickly gathered to
ascertain if there is imminent hazard to human health and/
or the environment. If an imminent hazard exists, then
governmental agencies working with the responsible own-
er, if there is one, should immediately construct contain-
ment measures or even temporary ones, if necessary. Re-
gardless whether or not an imminent hazard exists, the site
should be secured from public access.
Once the problem site becomes public knowledge, the in-
volved government agencies should educate the media and
concerned groups about the situation. If in the opinion
of the governmental agencies, resource recovery can be
used to clean up the site, the media and all concerned
groups should be so informed. The benefits of resource re-
covery over land disposal should be extolled. Education
and allowance of input from concerned groups by govern-
iment agencies will help to allay public distrust and anxiety.
After an uncontrolled site has been identified, to assess
whether an imminent hazard exists, sampling and analysis
should immediately commence and almost concurrently a
comprehensive program for sampling and analysis should
be prepared. At this point, the resource recovery staff
should be consulted for recommendations on the param-
eters the sampling and analytical programs should be
addressed.
Until recently, the Department's CWE usually became
aware of a problem site only after sampling and analysis
of the site had been performed. Any lack of communica-
tion will unnecessarily delay the recycling of the wastes be-
cause in many instances of the long turn-around time
(months) between sampling and the completion of the
chemical analyses and the lengthy time needed in many in-
stances to locate recyclers. The Resource Recovery staff
should be involved as soon as possible. The time to locate
recyclers may be significantly reduced if a Resource Re-
covery staff has access to a computer-based waste ex-
change. Instead of a quarterly catalog of wastes available,
which is now distributed by most clearinghouse waste ex-
changes, a daily update would be available from the com-
puter data base to all interested parties. Presently, the
CWE is planning to complement its program through the
assistance of a computer-based waste exchange.
In addition, instead of depending on a response to a cat-
alog or computer listing of available wastes, the Resource
Recovery staff should also actively engage in the matching
of the waste with potential recyclers. Finally, because
sometimes wastes are recycled over great distances (i.e.,
1000 miles) there should be attempts to locate recyclers on
a regional instead of a State basis.
REMEDIAL ACTIONS AND FLOWCHART
Any remedial action program for an uncontrolled haz-
ardous waste site should consider the possibility of clean-
up through resource recovery. The following proposed
schematic (Figure 6) for maximizing resource recovery of
hazardous wastes from an uncontrolled site incorporates
the above remedies to roadblocks. Just before efforts are
initiated to qualitatively identify imminent human health
and environmental hazards, the Resource Recovery staff
should become involved. This early inclusion of the re-
source recovery staff allows additional time for the loca-
tion of potential recyclers. It has been the experience of the
CWE that approximately 2 weeks is frequently necessary
to determine whether there is a potential for resource re-
covery.
The resource recovery potential is assured when there
are known and willing recyclers. However, this is not the
usual case and the potential for resource recovery, in many
instances depends on the vision which is basically a func-
tion of the knowledge and experience in analytical and
industrial chemistry of the Resource Recovery staff.
To maximize the resource recovery potential of a waste,
the literature should be reviewed for uses for the chem-
ical constituents in the waste. Reference texts such as the
Encyclopedia of Chemical Technology(l3) Condensed
Chemical Dictionary(l4) Chemical Sources(l5) Directory of
Chemical Producers in the C/.S.,(16) California Manufac-
turers Register^ etc., have been found to be useful by
the CWE. The first four books identify the uses for a par-
ticular chemical. The last book, which is updated annually,
lists companies by geographic region, industry, and SIC
number and therefore permits identification of potential
recyclers for the hazardous chemical waste.
Oftentimes, the waste cannot be recycled as is, but must
be treated by biological, chemical and/or physical tech-
niques to reduce or remove certain chemical constit-
uents before recycling is possible, economics notwithstand-
ing. If this is the case, then the feasibility of recycling of a
waste will also depend on the treatment costs. In the end a
waste is only recyclable when it is both technologically and
economically feasible.
CONCLUSION
California has proven through its recent experience that
resource recovery can be an effective tool for the cleanup
of uncontrolled hazardous waste sites. Resource recovery
means recovery of raw materials and energy. In addition,
land is saved for future generations. As the problems of
siting new hazardous waste disposal sites increases and re-
source recovery becomes more publicized remedial action
programs for uncontrolled hazardous waste disposal site
will consider resource recovery as a matter of course.
ACKNOWLEDGEMENTS
The author wishes to express special thanks to Ms. Alyce
Tom for the typing of this manuscript and to my colleag-
ues who provided me assistance and information for this
paper.
-------
386 CASE HISTORIES
Location of
Site
Educate and
Inform Concerned
Groups
Qualitative
Identification
of
Hazards
/ On-site Visit/
--—-^Historical Information/
/ Qualitative Sampling /
/ and Analytical Program/
Containment
of Hazards and
Securing of Site
_L
Identification
and Amount of Waste
Identification
of
Treatment
Technologies
Historical Information/
~~Y Comprehensive Sampling /
and Analytical Program/
•—/ Llterlature /
I Manufacturers of /
/ Treatment Technologlces /
^Identlfl
Recycler
titled Potential
Figure 6.
Flowchart for Resource Recovery of Wastes
From an Uncontrolled Site
BIBLIOGRAPHY
1. "A permitted disposal site is closed down", Chem-
ical Week, Vol. 129, No. 4, P. 62, July 22, 1981.
2. California Air Resources Board, "Determination of
the Air Quality at the Abandoned Waste Disposal
Site in Fullerton, California," Nov., 1980.
3. Thibodeaux, L.J., "Estimating The Air Emissions of
Chemicals from Hazardous Waste Landfills," /. of
Hazardous Materials, 4, 1981, 235-244.
4. Shen, T.T., "Control Techniques for Gas Emissions
from Hazardous Waste Landfills," JPACA, 31, 1981,
132-135.
5. Neely, N., Gillespie, D., Schauf, F., and Walsh, J.,
Remedial Actions at Hazardous Waste Sites: Survey
and Case Studies, EPA-430-9-81-005, January 1981.
6. California Administrative Code, Title 22, Social Se-
curity Division 4. Environmental Health, Chapter 30,
Minimum Standards for Management of Hazardous
and Extreme Wastes, Section 66088, Register 79,
No. 19, 1979.
7. Franks, A.L., California State Water Resources Con-
trol Board's Waste Discharge Requirements for Non-
sewerable Waste Disposal to Land, July 1980.
8. California Assessment Manual for Hazardous Wastes,
June 1981.
9. Soil Improvement Committee of the California Fer-
tilizer Association, Western Fertilizer Handbook, 6th
Edition, 1980.
10. Berg, G.L., 1981 Farm Chemicals Handbook.
11. Kirk-Othmer Encyclopedia of Chemical Technology,
Third Edition, /, 1978, 203-210.
12. California State Department of Health Services,
Health and Safety Code, Hazardous Waste Control
Law, Div. 20, Chapter 6.5.
13. Kirk-Othmer Encyclopedia of Chemical Technology,
Third Edition, 1978.
14. Hawley, G.G., The Condensed Chemical Dictionary,
9th Edition, 1977.
15. Directories Publishing Co., Inc., Chem. Sources—
U.S.A. 1980 Edition.
16. SRI, International, 1981 Directory of Chemical Pro-
ducers United States of America.
17. Times Mirror Press, 1981 California Manufacturers
Register, 34th Edition.
-------
THE GENERATOR'S LIABILITY FOR PAST
HAZARDOUS WASTE DISPOSAL PRACTICES
JOSEPH M. MANKO
AND
MARC E. GOLD
Wolf, Block, Schorr and Solis-Cohen
Philadelphia, Pennsylvania
INTRODUCTION
One need only pick up any newspaper to be reminded
that improper disposal sites were used by industry as haz-
ardous waste burial grounds over the past decades. Most
new federal, state and local legislation and regulations
deal prospectively with the disposal of hazardous waste.
They leave for the courts the difficult task of determining
who will bear the legal responsibility for cleaning up the
hidden horrors which threaten public health and natural
resources. In this paper, the authors review the many com-
plex technical and legal issues involved in assigning and
apportioning this liability and provide some practical
guidance to hazardous waste generators on future pre-
cautions.
In discussing the potential liability of generators for past
hazardous waste disposal practices, the authors are not
focusing on owners or operators of hazardous waste treat-
ment, storage or disposal facilities or even the haulers who
brought the waste to such facilities. In most cases, their
liability is determined on a statutory basis, under certain
circumstances going back to the Refuse Act of 1899'" and
more recently the Resource Conservation and Recovery
Act ("RCRA")(2) and the Comprehensive Environmental
Response, Compensation and Liability Act ("Super-
fund").(3) What the authors do address is the legal respon-
sibility of a typical hazardous waste generator who hired
someone to haul hazardous waste to an off-site disposal
facility.
BACKGROUND
The "Throw Away Mindset"
Just a few years ago, before the recent hazardous waste
catastrophes raised the public's consciousness level, the
industrial mindset with respect to hazardous waste disposal
was generally "out of sight, out of mind." Most industries
assigned their purchasing agents the task of arranging for
the disposal of industry's unwanted product: its hazardous
waste. Purchasing agents were, for the most part, un-
familiar with the nature of the wastes or the proper dis-
posal methods required. Faced with relatively insignificant
disposal costs, even for the most toxic wastes and relatively
limited technology or interest in recycling such wastes,
arrangements for the disposal of hazardous waste were
routinely made, with little consideration as to its ultimate
disposition. Rarely would a generator take the time to es-
tablish and apply transporter or disposer selection criteria
other than price. It was the atypical generator who even
considered examining the transporter's equipment, inquir-
ing about its reliability or reputation, or even determining
from regulatory authorities whether it had valid permits to
conduct its business or any enforcement proceedings pend-
ing against it.
Perhaps most important of all, many generators never
even inquired as to the transporters' intended disposal
location and few, if any, ever followed a waste shipment
to assure that it reached the proper destination. Other
than in those rare instances where a generator had ample
land adjacent to its plant on which to dispose of its
waste,(4) little care was taken to make certain that hazard-
ous wastes were disposed of properly.
The Love Canal Syndrome
The tragedies that beset the community of Niagara Falls,
New York, captured the nation's attention and have held
it for the past several years. It was impossible to escape
reading or avoid seeing on television stories about the
anachronistically labelled "Love Canal." There followed
soon on the heels of this disclosure the discovery of other
hazardous waste nightmares such as the Valley of the
Drums in Kentucky, the dumping of PCBs along the road-
side in North Carolina, and the frequent midnight dump-
ing of hazardous waste in the back woods of New England.
Today, every state has its own litany of hazardous waste
horrors—monuments to unbridled industrial waste dis-
posal. And as if these hazardous waste exposures to the
soil, surface and groundwaters from these incidents were
not enough, the fire and explosion of chemical wastes
stored in Elizabeth, New Jersey, lifted into the air for
miles around toxic fumes deadly enough to threaten
millions of people with immediate harm. The notion be-
came aware that there were ticking timebombs, on and
above the ground which made everyone vulnerable to
hazardous waste impacts from a variety of environmental
pathways.
Enter the Government
As is so often the case, government rushed into the
breach with new legislation and regulations. The narrow
focus of the Toxic Substance Control Act ("TSCA")(5)
with its ambitious regulatory program requiring preregis-
tration of new chemical substances and the lengthy de-
387
-------
388 LIABILITY, LEGAL & PUBLIC ISSUES
lays in implementing RCRA—the two major federal leg-
islative efforts in 1976—demonstrated the need for further
legal authority to specifically address past problems. Sim-
ilarly, as states prepared to accept delegation of various
federal regulatory programs, the need to adopt new state
hazardous waste laws and regulate the siting of new
hazardous waste facilities became evident.
Although hazardous waste legislation and the educa-
tional benefits of the implementation of such programs
will prove to be helpful in the future, there was still no
direct mechanism for providing the requisite funds to clean
up the hazardous waste disposal sites which continue to
contaminate the environment. The issue then shifted to
one of economics which was compounded in large measure
by the insolvency of most illegitimate transporters (i.e.,
the midnight dumper), the nonaccountability through in-
solvency or disappearance of the owners and operators of
improperly run sites, and an overlay of extremely complex
legal and scientific problems.
Congress finally addressed these issues, albeit less clear-
ly than anticipated, by the passage of Superfund which
was signed into law in December, 1980. The approach
adopted was to establish a $1.6 billion fund to be used by
the federal government to respond to the most serious
hazardous waste disposal site problems on a priority basis,
88% of which would be raised by a tax on the petroleum
and chemical industries starting April 1, 1981. Since even
this ambitious undertaking would be inadequate to clean
up all such sites, it was left for the legal system, through
the application of statutory and common law principles,
to develop standards to determine and apportion liability
among the various participants in the hazardous waste dis-
posal chain.
RESPONSIBILITY FOR PRIOR PRACTICES
Typically, a hazardous waste disposal site will be dis-
covered where the transporters involved and site owner
and/or operator are incapable of responding to environ-
mental problems, either because they are bankrupt or have
fled the jurisdiction. Combining this fact with the pub-
lic's jaundiced view of industries, as having amassed great
profits from their manufacturing activities while gener-
ating and improperly disposing of tons of hazardous
waste, quickly results in pressure to place the clean up
responsibility on the generator.(6) Support for this ap-
proach is founded on the philosophy that industry can
either internalize the clean up costs, adversely impacting
only the generator's shareholders, or externalize them
through higher costs of goods to the consumers. Industry
counters this view by suggesting that the public bear the
expense through general tax revenues, like some of the
present proposals to clean up the Three Mile Island nuclear
plant, since it was the public at large that created the de-
mand for and presumably benefitted from the products
manufactured.
As this debate on the liability issue continues, the public
has grown more distrustful of corporate officials and with
the slow pace at which governmental officials have been
able to attack these problems—both technologically and
fiscally—the public's confidence in their ability to rescue
them has also waned. At the same time, scientific methods
have developed the capability to determine down to one
part per trillion the presence of contaminants in the en-
vironment. It is in this setting that the battle lines have
been drawn pitting against each other teams of experts to
determine the extent of needed clean up and the question
of who should bear the legal and financial responsibility.
Such determinations will not come swiftly.
COMMON LAW LIABILITY
The General Rule
Where a generator directly disposes of hazardous waste,
rather than hiring an independent contractor to remove the
waste for proper disposal, his liability is more easily de-
termined under the various statutory and common law
theories. Thus, it will be difficult for a generator of haz-
ardous waste to escape liability where it can be shown that
he dumped the waste at a site which was known or which
should have been known to be improper. Since most gen-
erators did not dispose of their own wastes, this area will
not be discussed further; instead, the more common pat-
tern will be emphasized: the generators who hired inde-
pendent contractors to remove hazardous wastes from
their plants for off-site disposal.
The general common law rule is that an employer is
not liable for the acts of its independent contractor.*7*
This rule is based on traditional contract principles where
one is free to shift responsibility to another independent
actor who is competent to perform the task. However,
over the years the courts have carved out important ex-
ceptions to this rule resulting in significant limitations on
its applicability.
The Exceptions to the General Rule
NEGLIGENTLY HIRED CONTRACTOR. First, a gen-
erator will be held directly liable for the acts of its in-
dependent contractor which cause physical harm to third
persons if the generator has been negligent in, or fails to
exercise reasonable care over, the hiring of a skillful con-
tractor.(8) The applicability of this exception will depend
upon the generator's compliance with an evolving set of
standards dealing with the degree of investigation of the
transporter's competence, reputation and license status.
In making an assessment of the independent contractor-
transporter, some predictable questions arise:
(1) Did the hauler have a prior record of violations?
(2) Had there been adverse press about the transporter's
practices?
(3) Were the transporter's prices significantly lower than
his competition?
(4) Was the transporter's equipment in good operating
condition?
(5) Did the generator check the transporter's references?
In short, liability may be established upon this principle
where it can be shown that the generator knew or should
have known of deficiencies in the transporter's qualifica-
tions. Assuming that the generator is found to have con-
ducted a sufficient investigation of the independent con-
tractor selected to remove the hazardous waste or that the
-------
LIABILITY, LEGAL & PUBLIC ISSUES 389
generator can show that even had such an investigation
been conducted, it would not have disclosed any negative
information, the generator has only addressed the thres-
hold issue in attempting to escape liability.
STRICT LIABILITY. The second exception to the inde-
pendent contractor rule involves principles of strict liabil-
ity for ultrahazardous or abnormally dangerous activ-
ities/9' This theory is premised on the need to find that
the generation of hazardous waste is "abnormally danger-
ous" and involves such a high degree of risk that no mat-
ter who is hired to remove the hazardous waste and no
matter how extensive the investigation of the independent
contractor may have been, the generator should be held
liable for any harm caused by the wastes.
The traditional example of an ultrahazardous activity is
the manufacture of dynamite, which has been found by
the courts to be so abnormally dangerous that the manu-
facturer remains liable for harm caused by the dynamite's
explosive characteristics. Concluding that an activity or
material is ultrahazardous and thereby imposing a stand-
ard of strict liability merely makes a societal judgement
that while the conduct or material itself should not be
banned, it carries such a high degree of risk that the pro-
ducer must be held absolutely accountable for any result-
ing harm.
Much of the recent legislation has attempted to define
retroactively certain activities involving hazardous waste as
"ultrahazardous", or "abnormally dangerous."(10) This
raises a myriad of questions as to whether the waste be-
comes ultrahazardous at the time it is improperly disposed
of, or whether it retains an ultrahazardous character from
the point of its generation. This area is extremely complex
and raises such fundamental issues as the constitution-
ality of the retroactive application of statutory declara-
tions and the precise factual determinations necessary to
form the underpinnings of a plaintiff's case.
DIRECT LIABILITY FOR A PECULIAR RISK. If the
generator properly selected an independent contractor and
principles of strict liability do not apply, he may still be
directly liable for harm caused by the hazardous waste if
it posed a "peculiar risk" against which special precau-
tions should have been specified but were not.(ll) Here,
much will depend upon the nature of the waste and the
specific arrangement between the generator and the inde-
pendent contractor as to the recognized peculiar risks and
the concomitant precautions that were specified.
Information about the hazardous nature of the waste
and the precautions necessary to be taken in handling it
should have been communicated to the independent con-
tractor if the generator is to discharge its obligations in
this area. However, some of the many unanswered ques-
tions include determinations of whether the risk was
''peculiar" and the specified precautions were reasonable.
VICARIOUS LIABILITY. Even assuming the generator is
able to escape direct liability, he still faces the possibility
of;being found vicariously liable. For example, even if
special precautions were identified by the generator but the
independent contractor failed to implement them and
harm resulted, the generator may be held vicariously
Gable. (12>
Similarly, the generator may be vicariously liable for a
nuisance or trespass caused by the independent contractor
depending on the extent of the generator's knowledge or
the foreseeability of such results.(13) Until more courts
address these issues, the standards of common law liabil-
ity will still be evolving and the establishment of direct or
vicarious liability will turn on the particular set of facts.
Possible Defenses to Common Law Liability
Assuming that a court would find a generator liable
under any one or more of the common law rules described
above, there are defenses that may be available to avoid
generator liability.
1. Where negligence is an element of the cause of action
against the generator, contributory or comparative
negligence may be asserted as a defense. For example,
where the operator of the disposal facility that wrong-
fully accepted the waste sues the generator for the re-
sultant harm, the generator could point to the consent
of the operator or other acts in an attempt to shift the
liability.
2. Lack of causation linking the generator's waste to the
site or to the harm will most certainly be raised by gen-
erator-defendants. Here the generator could show that
although the waste was improperly disposed of it did not
create harm or even the likelihood of harm, that the
likelihood of harm was not foreseeable, that the waste
was not a substantive factor in causing the harm or that
the waste cannot even be traced back to the particular
generator.
3. The statute of limitations for the cause of action may
already have run. The key question here is when did the
cause of action against the generator arise: Was it when
the disposal took place, the subsequent discovery of the
waste site, or the subsequent injury? Often the type of
action brought will define the length of the applicable
statute of limitations and the court's answer to the
above questions will determine the number of years that
have elapsed prior to suit.
Finally, assuming that liability is established, there are
significant questions as to the measure of damages and
the way in which such damages are to be apportioned.
These issues may turn on the extent to which cleanup is
necessary and the value of the property before the injury
as compared with its present value. In addition, there
are extremely critical issues to resolve as to the apportion-
ment of liability among many defendants, methods of
contribution, and joint and several liability. Answers to
each of these questions may dictate litigation posture,
strategies and settlement possibilities.
FEDERAL LAW ENFORCEMENT
Prior to the adoption of Superfund, the federal gov-
ernment had proceeded against hazardous waste genera-
tors for past practices primarily under the emergency pro-
vision of RCRA, Section 7003,(14) which was amended in
1980 to require the government to prove only that an immi-
nent and substantial endangerment to public health and
welfare may exist.(15) To date, all but one court has held
-------
390 LIABILITY, LEGAL & PUBLIC ISSUES
this provision to be jurisdictional only and the courts have
required the. government to prove its entitlement to relief
under the substantive standards established by the federal
common law of nuisance."" Although the courts have not
yet held hazardous waste generators liable under this
RCRA provision for their disposal of hazardous waste
through independent contractors, there are likely to be
many cases filed by the federal government on this theory
as well as by affected landowners under similar common
law theories."7'
Superfund was heralded as the tool which the federal
government would use in the future to impose strict lia-
bility on hazardous waste generators for past disposal
practices. Conceptually this is analogous to the standard
of liability in the products liability area.
Congress relied on the standard of liability established in
Section 311 of the Clean Water Act"8' for establishing
strict liability. It is important to recognize that "hazardous
substances," the term used under Superfund, is very
broadly defined"" to include hazardous pollutants under
Section 112 of the Clean Air Act,'20' hazardous and toxic
substances under Sections 304, 307 and 311 of the Clean
Water Act/2" toxic substances under Section 7 of
TSCA,(22) and hazardous waste under Section 3001 of
RCRA.'23' In addition, Section 102 of Superfund'24'
authorizes the Administrator of EPA to list additional sub-
stances as hazardous not already included within this
broad definition. It is obviously all inclusive.
The first requirement imposed under Superfund was the
reporting of sites not registered under RCRA which had
been used as hazardous waste disposal facilities.'25' Based
on these reports, EPA will be able to identify and evaluate
thousands of potential sites for priority cleanup.
Included within the class of persons covered under the
liability section of Superfund are operators, owners, trans-
porters and generators.'2*1 The exposure is quite broad cov-
ering the costs of removal and remedial action incurred
by the government or other persons consistent with the
provisions of the National Contingency Plan.'27' These
costs would include not only site investigations, but
measures ranging up to and including the temporary or
permanent relocation of residents and the provision of al-
ternate water supplies, as well as reimbursement for dam-
age to natural resources under certain circumstances.
There is a significant requirement limiting the extent of site
cleanup under Superfund to actions that are cost-effective
which takes into account the limits of the fund and the
need to respond to other priority sites. However, that lia-
bility for personal injury claims was specifically deleted in
a compromise effort to pass Superfund.
The defenses available under Superfund are intentional-
ly very limited, covering releases of hazardous substances
caused solely by acts of God, war and a third party not
in a direct or indirect contractual relationship with the
generator. In order to come within this last defense, the
generator must show by a preponderance of the evidence
that he used due care, taking into consideration the nature
of the waste, and took precautions against foreseeable
acts of third parties and the consequences of such acts.
The true scope of the Superfund defenses will have to
await judicial review on a case by case basis before their
availability can be fully evaluated.
As indicated, Superfund would appear to foreclose a
generator from asserting many of the common law de-
fenses previously described; however, there is serious ques-
tion as to whether the retroactive application of this statute
will withstand constitutional challenge. Also, while the
liability provisions of Superfund purport to be immed-
iately effective, many of the statutory prerequisites, such
as revisions to the National Contingency Plan, the need
for state participation and the establishment of a priority
list, have not yet been implemented, so that a court may
be persuaded that asserting liability under Superfund may
be premature.
Also, President Carter's Executive Order08' delegating
much of the authority to various federal agencies, primar-
ily EPA, is coming under close scrutiny by the new ad-
ministration. On the other hand, the government will ar-
gue that Superfund was intended to be used as an immedi-
ate stopgap measure to clean up sites before they create
human disasters and that relief need not await the comple-
tion of the regulatory process or strict adherence to pro-
cedural preconditions. Obviously, this is another issue
which must await judicial interpretation.
STATE LAW ENFORCEMENT AND
RECENT DEVELOPMENTS
An example of the changing legal theories of liability
potentially applicable to hazardous waste generators is no
more evident than in Pennsylvania. Until September, 1980,
the Commonwealth of Pennsylvania relied primarily on
the Clean Streams Law'29' which regulates the pollution of
surface and groundwater essentially through a permit pro-
gram. This law provides for vigorous enforcement, abate-
ment actions and the assessment of penalties against,
among others, landowners of contaminated property.
Pennsylvania courts have held landowners to be strictly
liable for the cleanup of their property even if the damage
was caused by a prior owner.'30' In the context of generator
liability, however, unless it could be proven that the
generator "discharged" or "permitted the discharge" of
hazardous waste into the surface or groundwater, the
Clean Streams Law was not an effective means of impos-
ing liability. In instances where a generator did nothing
more than contract with a transporter for the disposal of
hazardous waste, courts would have to stretch the "per-
mit to discharge" language to establish liability.
Prior to its amendment in 1980, the Pennsylvania Solid
Waste Management Act ("Act"),<31) proved quite inef-
fectual since there was no meaningful provision for direct
enforcement against hazardous waste generators and pen-
alties were limited to $300 per day. However, the 1980
amendments to this Act'32' provided a comprehensive
hazardous waste program including a manifest system,
permits, licenses for transporters (which exceed the
RCRA requirements for transporters) and a variety of ad-
ministrative, civil and criminal penalties, ranging up to
$500,000 and 20 years imprisonment. However, unless
the provision declaring certain hazardous waste activi-
-------
LIABILITY, LEGAL & PUBLIC ISSUES 391
ties "ultrahazardous" and thereby subject to strict liabil-
ity for any resulting harm (which on its face is retroactive)
includes the act of generating hazardous waste, generators
may still be able to escape direct liability for the prior off-
site disposal of hazardous wastes.(33)
Perhaps the trend of judicial decisions in this area can
be seen in New Jersey v. Ventron Corp. (see footnote 6)
which interpreted the New Jersey Spill Compensation and
Control Act(34) strictly and apportioned liability jointly
and severally against two parent corporations of wholly
owned subsidiaries. While the case is arguably limited to
its facts and is strictly an application of New Jersey sta-
tutory law, it reflects the judiciary's attitude on hazardous
waste disposal issues (in this case, mercury) that wherever
feasible, the defendant most financially able to effect a
remedy is likely to be responsible, since the potential ad-
verse public impact of improper hazardous waste disposal
is astronomical.
WHAT'S A GENERATOR TO DO?
SOME PRACTICE ADVICE
Once a generator has been sued, he obviously will have
to defend himself in court against the various theories of
liability claimed by the plaintiff, many of which have been
described in this paper. Open questions that will invariably
be litigated include the burden of proof necessary to es-
tablish causation, applicability of joint and several liabil-
ity, bases of apportionment, constitutional limits on retro-
active application of statutes and statutes of limitation.
Prior to suit, a generator can attempt to uncover po-
tential problem areas and take certain actions to be in the
best position in the event litigation is brought.
The generator can audit its hazardous waste disposal
records as far back as he can (since there may be no time
limitation) to determine whether its waste was hazardous.
This determination can be accomplished by consulting
with his engineers and chemists, reviewing raw material
records, etc. Next, the generator can review the names of
each of his transporters and disposers to see whether they
are any longer in business and whether any are now being
investigated or have already been prosecuted.
All contract forms for any waste disposal activities
should be reviewed to determine whether any "special
precautions" were identified. All personnel involved in
hazardous waste disposal activities should be interviewed
and the facts memorialized for future reference.(35)
Generators should review their insurance policies to see
whether or not there is any possible coverage should there
be claims arising out of hazardous waste activities. The
"sudden and accidental" exclusions in most policies have
now given way to more liberal coverage for environmental
impairment, but questions as to applicability for past prac-
tices are obviously quite important to any determination
of coverage. In addition, there is at least one reported
case where courts have interpreted insurance policies and
found cover age.(36)
The authors can do no more than note the problems
that they can foresee arising over the next few years in
the area of generator liability for hazardous waste prac-
tices. Very few "hard" answers can be given since very
few have yet been furnished; instead, most answers are
merely arguments for judges and juries, whose decisions
will become "the law." Then and only then will we know
what consequences will be meted out to generators of
hazardous waste for their past disposal practices.
REFERENCES AND FOOTNOTES
1. 33 U.S.C. §407.
2. 42 U.S.C. §6901, ef. seq.
3. 42 U.S.C. §9601, et.seq.
4. A generator with an on-site hazardous waste disposal
facility was still required to obtain the necessary per-
mits and approvals and to comply with existing oper-
ating requirements. By using plant property, the gener-
ator only avoided the risks of employing a transporter
for off-site disposal and the lack of control inherent in
such practices.
5. 15 U.S.C. §2601, et. seq.
6. For example, the court stated in New Jersey v. Ven-
tion, _____A.2d , 1 Chem. and Rad. Waste
Litigation Rptr. 348 (Super. Ct. Ch. Div., August 29,
1979):
As we become more sensitive to our environment
and more aware of the impact of pollution on our
environment, we must demand that the unchecked
development of products which release pollutants
into our environment be controlled. It does not of-
fend this court's sensitivities nor infringe upon a
manufacturing defendant's constitutional rights to
impose strict liability upon a defendant who, dur-
ing the course of a profit making venture discharges
into the environment a dangerous or hazardous
pollutant which results in damage or harm to the
public, notwithstanding an absence of intent or
negligence on the part of the defendant.
7. Restatement (Second) of Torts §409.
8. Restatement (Second) of Torts §411.
9. Restatement (Second) of Torts §427A. An "abnorm-
ally dangerous" activity is defined in Restatement
(Second) of Torts §520 to include the following factors:
•Existence of a high degree of risk of some harm to
the person, land or chattels of others;
•Likelihood that the harm that results from it will
be great;
•Inability to eliminate the risk by the exercise of
reasonable care;
•Extent to which the activity is not a matter of com-
mon usage;
•Inappropriateness of the activity to the place where
it is carried on; and
•Extent to which its value to the community is out-
weighed by its dangerous attributes.
10. See Pennsylvania Solid Waste Management Act, 35
P.S. §6018.401(b):
The storage, transportation, treatment and dis-
posal of hazardous waste are hereby declared to
-------
392 LIABILITY, LEGAL & PUBLIC ISSUES
be activities, which subject the person carrying on
those activities to liability for harm although he
has exercised utmost care to prevent harm, re-
gardless whether such activities were conducted
prior to the enactment hereof.
Also, in the legislative history of Super fund it is
stated that:
For the purposes of this act, Congress declares
that manufacture, use, transportation, treatment,
storage, disposal and release of hazardous sub-
stance are ultrahazardous activities. 126 Cong.
Rec. S 14972 (daily ed. November 24, 1980).
Note, however, that the Pennsylvania language
does not include "manufacturer" or "generator",
whereas Congress did include the former term in its
legislative intention.
11. Restatement (Second) of Torts §413.
12. Restatement (Second) of Torts §416.
13. Restatement (Second) of Torts §427B.
14. 42U.S.C. §6973.
15. The pre-1980 version of RCRA's imminent hazard sec-
tion required the government to prove that "the hand-
ling, storage, treatment, transportation or disposal of
any solid waste or hazardous waste is presenting an
imminent and substantial endangerment to health or
the environment. Now, all the government has to
prove is that the listed activities "may present" such
an endangerment.
16. See United States v. Midwest Solvent Recovery, Inc.,
484 F.Supp. 138 (N.D. Ind. 1980); United States v.
Solvents Recovery Service of New England, 496 F.
Supp. 1127 (D. Conn. 1980). Contra, United States v.
Diamond Shamrock, F.Supp (E.D.
Ohio 1981).
Further, in light of the holding in City of Milwaukee
v. Illinois, 451 U.S 68 L.Ed.2d 114 (Sup. Ct.
1981), that the federal commonal law of nuisance has
preempted in the water pollution area, it may also be
held to be inapplicable to waste disposal matters.
17. In Swell v. Petro Processors of Louisiana, Inc., 364
So. Rptr. 604 (Ct. App. La. 1978), generators who had
actual knowledge of improper waste disposal prac-
tices were held liable for damages to a neighboring
landowner's property.
18. 33U.S.C. §1321.
19. 42U.S.C. §9601.
20. 42U.S.C. §7412.
21. 33U.S.C. §§1314, 1317, 1321.
22. 15 U.S.C. §2607.
23. 42 U.S.C. §6921.
24. 42 U.S.C. §9602.
25. 42 U.S.C. §9603.
26. 42 U.S.C. §9607.
27. The National Contingency Plan was developed under
the Clean Water Act to describe the procedures and
criteria for responding to oil spills. 40 CFR Part 1510.
Under Section 105 of Superfund, 42 U.S.C. §9605, it
is to be significantly revised to incorporate pro-
cedures, priorities and methods of responding to re-
leases of hazardous substances.
28. 46 Fed. Reg. 9901 (January 19, 1981).
29. 35P.S. §691A et. seq.
30. National Wood Preservers Corp. v. Pennsylvania
Department of Environmental Resources, 414 A.2d
37 (Pa. 1980).
31. 35 P.S. §6001 et. seq.
32. 35 P.S. §6018.101 ef.*??.
33. Future off-site disposal activities by generators is
highly regulated in Pennsylvania and the Act specifi-
cally cuts off liability for civil and criminal penalties
provided a generator uses a licensed transporter and
the waste is received by an appropriate permitted stor-
age, treatment or disposal facility. 35 P.S. §§6018-
605, 6018-606.
34. 58 N.J.S.A. 10-23.11 et. seq.
35. It is important to consider the discoverability of any
written memoranda prior to their preparation so that
appropriate safeguards can be instituted.
36. Lansco Inc. v. New Jersey Department of Environ-
mental Protection, 138 N.J. Super. 275, 350 A.2d
520 (Ch. Div. 1975) aff'd. per curiam 145 N.J. Super.
433, 368 A.2d 363 (App. Div. 1975).
-------
RECOVERING DAMAGES TO NATURAL RESOURCES
UNDER CERCLA
EDWARD YANG, Ph.D.
AMY HORNE
Resources Program
Environmental Law Institute
Washington, D.C.
OSCAR ALBRECHT
Office of Research and Development
U.S. EPA, Cincinnati, Ohio
INTRODUCTION
Until recently, pollution damages to publicly owned
natural resources were seldom paid for, due to the dif-
ficulties of establishing "ownership", assessing physical
damage and deciding on monetary values of the resources.
Increased oil and hazardous substances spills with the last
two decades, together with improved knowledge of the
functions of natural resources, have led to the recogni-
tion that these functions warrant compensation for dam-
ages even though they are not explicitly traded in the
marketplace. The 1980 Comprehensive Environmental Re-
sponse, Compensation and Liability Act,(1) otherwise
known as the Superfund Act, establishes liability for dam-
ages or injury to, destruction of or loss of natural re-
sources resulting from hazardous substances releases, in-
cluding the reasonable costs of assessing such damages
(section 107 (a)). Further, in order to facilitate damage
recovery, the Superfund Act establishes a Hazardous Sub-
stances Response Fund which can be used by the govern-
ment to clean up spills and resource damage losses re-
cover. Consequently, the federal government has the un-
precedented opportunity to codify some damage assess-
ment techniques by promulgating regulations specifying
analysis procedures (section 301 (c) (1)).
The economic theory of externalities'2' holds that the
optimal compensation procedure will extract from pol-
luters an amount equivalent to the social value of the
damage. Otherwise, disregarding effects of income dis-
tribution, society will lose, due to misallocation of its re-
sources. In this paper, the authors briefly outline the eco-
nomic theory underlying resource damage assessment
and highlight some obstacles to establishing estimation
methods. Further, methods currently employed in legal
proceedings are reviewed and an alternative is considered
that can be applied under the Superfund Act. As such, the
paper's focus is on deriving a monetary measurement of
the damages rather than assessing physical damages, such
as identifying the pollutant, determining the contamina-
tion level and the effects caused by the pollutants.
PLACING MONETARY VALUE ON
DAMAGED NATURAL RESOURCE
According to economic theory, the appropriate valua-
tion of damages should be based on the demand that ex-
isted for the natural resources before they were injured.
The primary obstacle economists face is the non-market-
goods nature of open-access natural resources, such as
public beaches, state parks, open lakes or rivers. Because
these resources are not traded in the market place there is
no price from which monetary damages can be derived.
In such cases, economists resort to the concept of the
willingness of the users to pay for the natural resources.
A person's willingness to pay for a resource is the amount
of money he or she is willing to accept in place of using
the resource. When travel and equipment expenditures are
incurred, for example, by a user of a natural resource, the
willingness-to-pay figure must be reduced by these costs,
resulting in a "net" willingness-to-pay. For example, if a
sport fisherman is willing to pay $5.00 for catching a
trout, and the travel and equipment expenditures per
trout is $3.00, the net gain to him is $2.00 per trout.
The value of a natural resource, therefore, is the will-
ingness-to-pay net of the costs incurred to use the re-
source, aggregated for all of the users or potential users.
The value of the damage is the reduction of this aggre-
gated net willingness-to-pay that is caused by the incident.
This measure reflects society's welfare loss due to the
damage.
Economic techniques developed to measure net willing-
ness-to-pay(3) generally are expensive and time-consuming.
Moreover, many are designed for research rather than
damage recovery. The most widely used technique surveys
users of the resource and attempts to elicit their willing-
ness to pay. In order to obtain accurate results, the survey
must be a valid and sufficiently large sample of the user
population. Construction of such a sample often re-
quires a substantial effort.
Another economic technique relies on changes in prop-
erty values to estimate damage values. However, the
uses of this latter technique thus far have been mainly
limited to valuing the benefits of clean air.
TECHNIQUES USED IN LEGAL PROCEEDINGS
In the past, the considerable time and expense re-
quirements of the economic methods discussed above, lim-
ited their application in legal proceedings. Further,
courts lacked the technical expertise necessary to arrive at
a consistent framework to value the damages. In fact, only
recently have courts begun to hold the polluters liable for
393
-------
394 LIABILITY, LEGAL & PUBLIC ISSUES
damages to open-access natural resources, under the con-
cept of public trusteeship. Nevertheless, courts had to
develop methods that would quantify the damages so that
compensation could be made to the plaintiff. The follow-
ing subsections discuss these traditional assessment
methods.
Replacement Costs
In cases involving injury to a natural resource, replace-
ment costs are sometimes awarded by courts as a com-
ponent of restoration damages.<4) For example, a court
may order the defendant to pay for the cost of replacing
a grove of destroyed fruit trees with a comparable type
and number of seedlings. Similarly, a court may award a
plaintiff damages to cover the cost of restocking a pol-
luted stream or lake with fish and other aquatic life. In
these cases, the replacement cost is intended to com-
pensate for the loss of economic uses of natural resources
and restores the injured party to his or her pre-injury posi-
tion.
Replacement costs also have been used to determine the
value of non-economic uses of natural resources. In State
of Tennessee ex rel. Goodrich v. Riggen,™ the State sought
to recover the value of deer killed by negligent use of an
herbicide. At the trial, the State presented a number of
expert witnesses testifying as to value of the deer. Surveys
found that deer hunters spent $1,035-$ 1,727.47 for each
deer killed. Wildlife experts placed the recreational,
aesthetic, economic and ecological value of a deer at
$2,500-$5,000.(6> The plaintiffs sought $2,500 per deer
killed'7' based on the aesthetic, ecological and recreational
value of the deer. The court rejected the plaintiffs' as-
sessments of the value of the deer and instructed the jury
to consider instead:
"the average cost per deer, cost affected by the
survival ratio, restocking cost, the fact that deer re-
produce themselves, transportation costs for the deer
and economic benefits which you find the proof the
State of Tennessee may receive (sic), including that
from deer hunters; to reach an amount which would
fairly compensate the State of Tennessee for its al-
leged loss of deer and unborn fawns."...(8)
An important fact which perhaps influenced the court
is that in a different part of the State deer were so over-
populated that many would starve without control mea-
sures.'" The jury finally awarded the State of Tennessee
$1,119, based on testimony which asserted that the cost of
catching and transporting deer from the overpopulated
area was $100 per deer."01 The State lost on appeal.""
The replacement cost valuation technique was recently
employed by the plaintiffs in Commonwealth of Puerto
Rico vs. SS. Zoe Colocotroni.™ This case arose from an
incident in which an oil tanker dumped 1.5 million gallons
of crude oil in order to free itself after running aground
on a reef. The oil came ashore in Bahia Sucia, a Puerto
Rican lagoon which, prior to the spill was a "healthy,
functioning estuarial ecosystem" in which the angrove
forests functioned as breeding, feeding and nursery
grounds for fish, shrimp and benthic organisms as well
as being the primary food-producing agents of the organic
materials available to the aquatic food chain.03'
The oil spill resulted in the death of more than 92 million
marine animals in a 20 acre area for which no market
value, in the sense of lost market profits or revenues,
could be ascribed.04' In addition, the oil penetrated the
mangrove forests, resulting in a significant increase in
the mangrove mortality of the area.
More than five years after the incident, the mangrove
community showed a significant reduction in the number
of species of macrobiota, the number of molluscs and
crustaceans, and other marine organisms. The seagrass
community showed a significantly altered pattern of plant
biomass and the number of species and population density
of marine organisms.
The district court awarded the following damages: for
the more than 92 million marine organisms killed, the
defendants were to pay $5,526,583.20, a figures derived by
referring to the prices of marine organisms from biological
supply laboratories.
"The lowest possible replacement cost figure is
$.06 per animal... Accepting the lowest replacement
cost, and attaching damages only to the lost marine
animals in the West Mangrove area, we find the
damages caused by Defendants to amount to $5,526,-
583.20.""5'
For damages to the mangrove forests, the court awarded
damages for the 23 most affected spots in the West Man-
grove area. It found that the best means of reestablishing
these areas was by the intensive planting of mangrove and
restoration of this area to its condition before the oil spill.
Evidence showed that planting mangroves costs about
$16,500 per acre, and that the five year monitoring and
fertilization program would cost $36,000 per year or
$180,000 for the five years. The total damage of the pol-
lution to the West Mangrove equalled $559,000. The
court also awarded $78,108.80 in cleanup costs incurred
by the plaintiffs. The total award granted by the district
court was $6,164,192.09.
Contribution to Productivity
The contribution of a given resource to ecosystem pro-
ductivity had been used to place a value on the damaged
resource, on at least one occasion. In State of Florida v.
Bruce,(^ the administrative law judge assessed the value
of the detrital production of some felled mangrove trees.
The judge accepted the State's contention that detrital
value, the accumulated leaves, branches and seeds which
serve as an essential element in the estuarine food chain,
should be the basis for calculating damages. Its value is
defined from its contribution, through the food chain, to
the support of recreational and aesthetic fish species. The
court agreed that lost detrital value could be derived from
the size of the lost canopy cover, which could be found by
aggregating the measure of the cut mangrove tree diame-
ters. Thus, in this instance 2000 square feet of canopy was
lost. The value for the lost 2000 square feet was esti-
mated to be $2,760. This sum was derived by assuming a
$4000 per acre year value, which was multiplied by 15,
the number of years required for a mangrove seedling to
reach maturity.
-------
LIABILITY, LEGAL & PUBLIC ISSUES 395
In addition to showing that the lost detrital produc-
tion was a factor to be considered in calculating the value
of the damages, the State also argued that mangroves
served many other functions, such as water cleansing,
pollution control, nutrient uptake and assimilation,
shoreline stabilization, flood prevention, water conserva-
tion, oxygen production and sequestration of heavy metals
and other poisonous substances. However, compensa-
tion was not sought for the latter values.
Willingness-to-pay Survey
A technique for estimating the value of damages caused
by a pollution incident by surveying people's willingness-
to-pay for recreational uses of those resources was pre-
pared by plaintiffs for presentation to a court. Unfortun-
ately, no decision was reached in that case as to the
method's legal validity. In the Matter of the Complaint of
Steuart Transportation Company,(IT> a claim arose from a
spill of 250,000 gal of #6 fuel oil into the Chesapeake Bay.
As a result of the spill, thousands of waterfowl were
killed, 27 miles of wetlands were fouled, many inverte-
brates were killed, many oyster beds were extensively pol-
luted.08' Both the Commonwealth of Virginia and the
United States filed suit, the former seeking $731,500 in
damages to waterfowl and other costs, the latter seeking
$487,000 in cleanup costs and $1 million for loss of water-
~fowl.<19) The defendants eventually settled out of court for
$115,000.(20)
As part of its evidence as to the value of the damages
caused, the United States introduced a study which esti-
mated the lost hunting value resulting from the bird kill.<21)
An expert witness applied the results from a previously
completed study, which concerned the hunting expendi-
tures made by waterfowl hunters in the seven states lying
within the Pacific Flyway, to the facts in Steuart.
In that study, randomly selected hunters were sent a
questionnaire concerning household income after taxes,
number of seasons of experience in hunting waterfowl,
hunter's cost per season, hunter's seasonal bagged Water-
fowl harvest and the hunter's estimate of how much his
costs would have had to have risen above present costs
before he would have decided to totally abstain from wat-
erfowl hunting during the season. Using a statistical
formula, the marginal value of a bagged waterfowl for a
representative hunter was obtained.
Lost Business Opportunity
Damages to open-access natural resources can injure
business opportunity if business uses the resources as an
input, such as an oyster farm using clean open-water in
Skansi v. Humble Oil & Refining Company.m In this
case, the defendant argued that the plaintiff should not
receive more than his expected net profit on oysters de-
stroyed through water pollution. The court disagreed, say-
ing that the plaintiff had lost not only his net profits but
also all of the expenses of planting and cultivating, which
were included in the sales price.
In Carr v. United States,™ the court refined the lost
revenue valuation technique used in Skansi by limiting
the compensation to the net loss sustained by the plaintiff.
This figure was derived by multiplying the estimated fore-
gone yield in bushels by the average price per bushel for
that year and subtracting the costs of tonging and mar-
keting. Since the plaintiff had not expended money for
tonging and harvesting, the court was economically cor-
rect in eliminating those costs from the estimated lost
value. Consequently, the plaintiff received lost profits
plus those expenses that were already incurred.
Discussion
The above valuation methods, which were actually used
in legal proceedings, all attempt to measure the social
values of the damaged resources. Yet the methods are not
consistent: for the same damage, different values resulted,
depending upon which method was employed. These
methods also vary in their strengths and weaknesses. Re-
placement costs only address the supply side of the re-
sources, ignoring the level of demand for the resources
(i.e., should lost resources always be replaced or restored
at whatever cost?).
If only the supply side is considered, society might re-
place or restore resources whose costs exceed their social
value. The contribution to productivity method is more
useful when parts of ecosystem are damaged, disrupting
the aggregate function of the ecosystem. However, this
method's usefulness is limited if the functions or products
of the ecosystem are not well defined and not traded in
the market place.
The expenditure survey method has the strongest eco-
nomics basis, since it directly measures the users' willing-
ness-to-pay for the damaged resources. Its only weaknesses
are that: (1) costly surveys may be necessary and (2) the
accuracy of the results depends on the sincerity of the
people surveyed and on their interpretation of the ques-
tions. For example, in Steuart Transportation the answer
to the question of how much the costs of waterfowl hunt-
ing would have to rise above the present cost before the
hunter would decide to abstain from hunting may lead
to inaccurate results if the hunter is not constrained by
his actual income in answering.
The lost business opportunity method only applies when
a business using the open-access resources is injured by
damages to those resources. Hence, the method does
not address the value of the resources in public uses. In
fact, in the context of the Superfund Act, such private
damage recovery is excluded from the use of the Fund.
The applicability of the above methods under Super-
fund is limited by the potentially large number of spills
and types of natural resource damage. In order for the
Fund to process the claims expeditiously, the damage valu-
ation methods have to be, as Congress intended, simple
and require minimal administrative cost.
AN ALTERNATIVE SCHEME UNDER SUPERFUND
The damage valuation method under the Superfund
damage assessment procedure must strive for a balance
between conceptual validity and procedural simplicity. The
underlying concept is that a relationship exists between
the costs of conducting the damage valuation and the
-------
396 LIABILITY, LEGAL & PUBLIC ISSUES
benefits of accurately identifying the damage in
monetary terms. Because the social costs of miscalculating
the value of a unit of damage— the benefit of accurate
valuation— may be small in small releases,'24' the use of
expensive valuation methods can not always be justified.
Realizing this, Congress divided the damage assessment
regulations into two types:
"Such regulations shall specify (A) standard pro-
cedures for simplified assessment requiring minimal
field, observation, including establishing measures of
damages based on units of discharge or release or
units of affected areas and (B) alternative protocols
for conducting assessments in individual cases to de-
termine the type and extent of short and long-term
injury, destruction, or loss." (CERCLA) section 301
The Use of Value Tables for Small Releases
When the units of discharge or of affected resources are
small, Superfund can adopt value tables along the lines of
those used by the States of Virginia, Florida, Washington
and California.'23' In general, these tables set the values of
fish and game (except California, which includes other
natural resources, such as sand and flatworms) so that
damages can be recovered accordingly. In Virginia, the
court accepted the table as evidence of the value of the
lost fish in Commonwealth ex rel State Water Control
Board v. Weaver Mirror Co. No. 4722 (Franklin Co. Vir-
ginia Circuit Court, June 2).
Florida promulgated its table of values as regulations. <26)
Subsequently, the Florida Supreme Court accepted these
fish values in State Dept. of Pollution Control v. Interna-
tional Paper Company. It held that the values were proper
and relevant. (27)
The State of Washington published its Guidelines For
Evaluating Fish Kill Damages and Computing Fish Kill
Damage Claims (hereinafter Guidelines), which bases the
value of fish on its production cost at the state fish hatch-
eries. California used its table to assess compensation in
State of California v. Alice B. Copeland Vincilone,m
which arose after the defendant conducted leveling and
filling activities in "Big Hole," a complex of island,
water channels and land masses along the Colorado River.
Alaska also uses a table to assess pollution damage, but
with a different approach. Alaska's tables consist of a
schedule of penalties for oil discharges, which vary ac-
cording to: (1) whether the oil enters a freshwater, brack-
ish, or saltwater environment or public land (2) the toxi-
city, degradability and dispersal characteristics of the oil
and (3) the sensitivity and productivity of the receiving
environment.'291 This approach, although simpler than
the previous tables to administer, in that does not require
counting the damaged units, may not reflect the actual
value of the damage.
Although these value tables only cover a small por-
tion of the resources that can be damaged by hazardous
waste releases, they certainly can be expanded under
Superfund to include groundwater, wetland and other
natural resources. These tables can also be constructed on
a regional basis so for a given resource to reflect regional
difference in values. However, several weaknesses in the
existing value tables requires correction if they are to be
used for damage assessment under Superfund. The con-
struction of the values should be made more consistent
and be based on sound economics. Several examples will
help illustrate this need.
The fish values used in Virginia for example, are con-
structed from the average purchase price of the fish. This
is valid as long as the price is below the consumer's willing-
ness-to-pay for the fish. When the demand for the fish is
low the cost of replenishing the fish, including transporta-
tion and labor costs, may be an excessive compensation to
award.
This criticism also applies to parts of the Guidelines of
the State of Washington, which depends on replacement or
restoration costs. The Florida value table was derived
from a mixture of hatchery costs, expenditures and com-
mercial values. Expenditures of the recreational fishermen
cannot be used as the value, since they do not represent
willingness-to-pay for the fish. Again, commercial value
is also not accurate because the price includes other costs
of harvesting.
It is also questionable whether the price of the com-
mercial fish can be applied to species that are also in de-
mand for recreational fishing; there is no market where
open-access fish are allocated between commercial and
recreational fisherman.
The California table sums several types of values: re-
placement costs, use value and psychic value. This ap-
proach reflects the complexity of natural resources that
have multi-uses. However, the summation of the values
may result in double-accounting, since these values are
not always complementary. For example, the bird that a
hunter wants cannot be used for bird watcher.
Learning from the past experiences, value tables under
the Superfund can be constructed more consistently and
accurately. Consistency can be gained by employing the
best components of the state schemes and applying them,
to the extent suitable, to all small releases. Accuracy can
be improved by basing the techniques on more solid eco-
nomics. The economies of scale can allow careful construc-
tion of the values through willingness-to-pay surveys and
other techniques presently used in economics research. It
is likely that more than one technique will be used, due to
the different types of the resources (e.g., fish, wetland,
groundwater, lake and river).
When using surveys, the samples should be taken on a
regional basis, so that regional differences in uses will be
reflected in the values. Although simpler, value tables cor-
responding to units of discharge may be difficult to de-
fend, due to their theoretical weakness that the specific1
resource itself and its uses are not considered.
Use of Economic Techniques for Large Releases
When a hazardous waste release is large enough that
costs of conducting a case-by-case valuation can be justi-
fied, several economic techniques and approaches can be
used, including willingness-to-pay surveys, travel cost,
household production and revealed preferences. Although
these techniques are cumbersome, they can be simplified in
-------
LIABILITY, LEGAL & PUBLIC ISSUES 397
terms of procedure and understandability with minimal
loss of analytic rigor. Their chief advantage is that they
are all based on widely accepted economic principles.
This does not mean there is always consensus on how
the techniques are used. In fact, disputes do arise from dif-
ference in the empirical estimation procedures. It is ad-
visable, therefore, to determine which estimation pro-
cedures appear to have the support of most economists
practicing in this area.(30) The procedure should be adhered
to as much as possible to minimize variation in the
values resulting from different estimation procedures.
Another advantage of the case-by-case techniques is that
the values are generally derived from the sample of the
people who actually or potentially use the resources in
question. The resulting values, unlike those of value
tables, cannot be repudiated by the argument that the
values do not reflect the users of the specific damaged
resources.
The main objective to the case-by-case approach is its
high dollar and time costs. But with the status of rebut-
table assumption in court that is granted by the Superfund,
the techniques can be greatly simplified to reduce cost
and time. Elegance and rigor would have to be partially
foregone; the confidence interval of the estimates would
have to be increased. This often means small sample sizes
and short survey periods. Such sacrifices are worth the
benefits of bringing sound valuation techniques into the
effort to protect the nation's diminishing natural re-
sources.
FOOTNOTES
1. P.L. 96-510, 945 Stat. 2767 (Dec. 11, 1980)
2. Externality theory is concerned with divergences be-
tween private and social cost. For example, Firm A
emits smoke during its production process which soils
laundry cleaned by Firm B. The marginal social cost of
Firm A's activity is greater than the marginal private
cost and, consequently, the firm is producing more
than is socially optimal. According to A.C. Pigou, the
correct policy would be to levy a tax equal to the dif-
ference between marginal social and marginal private
costs. This tax would induce the firm to produce the
"right" socially efficient output—that at which price
equals marginal social cost.
3. See Freeman, A.M. Ill, "The Benefits of Environ-
mental Improvement" (Baltimore: Johns Hopkins
University Press, 1979) for a discussion of the tech-
niques.
4. See e g Commonwealth of Puerto Rico v. S.S. Zoe
< Colocotroni, 456, f. Supp. 1327 (1978)
5. Shelby Law No. 3 (1947)
i Assignments of Error and Brief of Appellant, at 19.
I. Michael J. Bean, "Law and Wildlife and Emerging
Body Wildlife Law," Environmental Law Reported,
Vol 7 March 1977, p. 50020.
8. Assignments of Error and Brief of Appellant, at 8.
9. John J. Tomanson, counsel for appellant, personal
communication, May 20, 1981.
10. Jim Allison, counsel for appellant, personal commun-
ication, May 21, 1981.
11. Assignments of Error and Brief of Appellant, at 28.
12. 628 F. 2d 652 (First Cir. 1980).
13. 456 F. Supp. 1327 (1978), at 1339.
14. Id. at 1344.
15. Id. at 1345.
16. DER Case No. 80-1481.
17. 495 Fed. Supp. 38 (D. Va. 1980).
18. James D. Range and Millicent A. Feller, "Con-
gressional Perspectives on the Need for Estimating En-
vironmental Damage from Oil and Hazardous Waste
Spills, "Proceedings of the 1979 U.S. Fish and Wild-
life Service Pollution Response Workshop," St.
Petersburg, Florida: U.S. Department of Interior,
September 1979.
19. General Accounting Office, Total Costs Resulting
from Two Major Oil Spills, Washington, D.C.; Gov-
ernment Printing Office, CED-77-71, June 1,1977.
20. "$115,000 Awarded in Oil Spill Suit," Chesapeake
Citizen Report, Vol. 3, no.2, March-April, 1981.
21. Gardner M. Brown, Jr. and Judd Hammack, "Com-
monwealth v. Steuart—Economic Valuation of
Waterfowl," (unpub. paper) May 5,1977.
22. 176 So. 2d 236 (La. Ct. Ap. 1965).
23. 136 F. Supp. 527 (D. Va. 1955).
24. Small releases can be defined on the basis of either
units of discharge or units of affected resources.
25. See Bruce C. Rashkow, "An Analysis of Alternatives
to the Traditional Approach to Damage Assessment,"
unpublished paper, Dept. of Justice, Washington,
D.C. 205-30.
26. In a case brought by commercial fishermen for dam-
ages resulting from the Santa Barbara oil spill, the
plaintiffs had difficulty establishing a base figure of
average commercial fish yield. And catch figures
normally vary widely and in addition are only sep-
arately reported, making them uncertain. Determining
lost profits with certainty was further dampened in
this case by additional damage the fish caused by un-
related flooding, Philip E. Sorenson, Environmental
Damage in Economics and Law: The case of the Santa
Barbara Oil Spill, Proceedings, Annual Meeting of
the Southern Economics Association, Atlanta,
Georgia, November 1976, pp. 115-119.
27. Fla. Stat. Ann. 403.141 (3) (1973) Doct. No. 15156,
California District Court, Riverside Superior Court.
28. Fla., 329 So. 2d 5 (1976).
29. As 46.03.758.
30. For example, the Department of Interior could es-
tablish a panel of economists expert in the area of
damage assessment.
-------
SITE CONTAMINATION AND LIABILITY AUDITS
IN THE ERA OF SUPERFUND
JOHN J. HOUSMAN, JR.
DAVID I. BRANDWEIN
DENNIS F. UNITES
TRC Environmental Consultants, Incorporated
Wethersfield, Connecticut
INTRODUCTION
Industrial managers responsible for corporate real es-
tate or environmental affairs are often not fully aware of
the legal and financial ramifications attending contami-
nated properties. With rapidly escalating government
efforts to locate and remedy abandoned or uncontrolled
hazardous waste sites, the need to inventory corporate
real estate, examine relative contamination potentials, de-
fine potential liability exposure and institute control
mechanisms is being recognized.
In this paper, the authors briefly outline the statutory
and common law initiatives and various liability path-
ways. However, the focus is on the presentation of a
structured multidisciplinary methodology, the Site Con-
tamination and Liability Audit, designed to identify site
contamination and liability problems and establish a
framework for their control. Examples are used to demon-
strate the need for such audits and to illustrate the audit
process.
The emphasis in the paper is placed on the property
screening aspects or early phases of the audit process.
The property screening phases are designed to obtain a
reasonable qualitative assessment of possible site contam-
ination problems with a minimum of effort and expense.
Site screening aspects are of particular value in an era of
increased corporate merger activity. Careful screening for
site contamination problems can prevent the acquisition
of severe liabilities.
FORCING STATUTORY AND COMMON LAW
INITIATIVES PREDATING SUPERFUND
Prior to enactment of the Comprehensive Environ-
mental Response, Compensation, and Liability Act of
1980 (Superfund), persons involved as defendants in lit-
igation related to inactive or abandoned contaminated
hazardous waste sites entered the courts under a variety
of statutory and common law provisions. The types of
actions which were initiated can be divided into the follow-
ing broad areas:
(1) Enforcement under Federal and State statutory pro-
visions
(2) Emergency powers provided by Federal and State
statutes
(3) Citizen actions taken pursuant to Federal and State
statutes
(4) Actions initiated by government pursuant to Fed-
eral and State common law
(5) Private suits brought by injured parties pursuant to
Federal and State laws
Enforcement Provisions under Federal
and State Environmental Laws
All Federal and State environmental laws have en-
forcement provisions making violation of specific sections
of the law either civil or criminal violations, depending
on the circumstances. Prior to Superfund, the most
widely used enforcement powers were those under RCRA,
the Clean Water Act, the Refuse Act and their state
counterparts.
The Clean Air Act, Safe Drinking Water Act, Toxic
Substances Control Act and their state equivalents have
seen more limited use. Since most of these statutory
provisions only address activities conducted after enact-
ment dates, effective use for abandoned or inactive waste
sites has been limited. Successful government actions
have most often occurred where the site has involved on-
going disposal activity.
Use of Statutory Emergency Powers
Most recent environmental statutes give Federal and
State officials broad authority if an environmental or hu-
man health hazard is imminent. However, in using these
powers, the government must first demonstrate that an
emergency exists. Federal and State courts have varied
widely in their interpretation of what constitutes an im-
minent hazard situation. Once this burden of proof is met,
the government has a wide range of remedies in equity
available under Section 7003 of RCRA, Section 504 of
the Clean Water Act, Section 303 of the Clean Air Act,
Section 1431 of the Safe Drinking Water Act, Section 7
of the Toxic Substances Control Act, various state equiva-
lents of these statutes and a number of State general
environmental policy laws. These remedies have included
actions such as:
(1) Order immediate restraint of an activity
(2) Require determination of the extent of contamina-
tion or environmental and health damage
(3) Require development of remedial plans
398
-------
LIABILITY, LEGAL & PUBLIC ISSUES 399
(4) Require implementation of containment or removal
measures
(5) Require installation of monitoring networks
(6) Order permanent restoration of the site
(7) Order installation of permanent alternative water
supplies
(8) Require posting of financial security through a
variety of mechanisms
Citizen Actions Pursuant to Federal and
State Statute
Each of the Federal environmental laws enacted in the
last decade and most comparable State laws contain so-
called "citizen suit" provisions. Generally, under such
provisions, any person or group of persons may commence
a civil suit against any party allegedly in violation of any
requirement of the act in question. These provisions also
generally provide for citizen suit against the administer-
ing regulatory agency for failure to perform an act or duty
under the law which is not discretionary with that agency.
Persons commencing such actions must first notify the
respective regulatory agencies and alleged violators of the
intention to bring action. No separate action may be
brought under these provisions if EPA or a state has be-
gun and is diligently pursuing the same action.
Government Common Law Actions
Federal and State agencies have used common law
principles in seeking relief against abandoned or uncon-
trolled hazardous waste sites. The common law theories
used have included nuisance, negligence, trespass, restitu-
tion and strict liability.
EPA has been particularly interested in imposing a
joint, several and strict liability standard on hazardous
waste generators, transporters and facility owners and
operators. For the most part, the courts have been sym-
pathetic to this approach. Typically, Federal and State
government attorneys have used the "shotgun" approach
in seeking relief under every possible avenue, citing com-
mon law theories along with the various statutory en-
vironmental authorities available.
Private Common Law Actions
Plaintiffs have brought suit under the common law
tort theories of negligence, nuisance, strict liability, tres-
pass and restitution. For the most part, these private
parties have found it difficult to muster the evidence
necessary to satisfy the strict burden of proof require-
ments of the common law tort theories. The govern-
ment, owing to their greater level of resources and clearer
statutory authority to enter property for the gathering of
evidence, has been more successful in obtaining judg-
ments pursuant to the common law tort theories.
INITIATIVES AVAILABLE UNDER SUPERFUND
The legal remedies available prior to the enactment of
the Comprehensive Environmental Response, Compen-
sation and Liability Act of 1980 (Superfund) pale as in-
significant before the awesome arsenal of authority now
available to government pursuant to the new Superfund
legislation. Essentially, the government now has a free
rein in its apparent power to address the problems pre-
sented by inactive, abandoned or uncontrolled hazardous
waste sites. Major sections of the new law of particular
importance to this discussion include:
(1) Section 103(a)—requires that the National Response
Center be notified of the release of any reportable
quantity of a hazardous substance.
(2) Section 103 (c)—requires that by June 9, 1981, any
person owning or operating, or who at the time of
disposal owned or operated, or who acting as an
transporter selected storage, treatment and disposal
facilities for hazardous substances, must have re-
ported to EPA the existence of such facilities,
hazardous substances received, and known, suspect-
ed or likely releases of hazardous substances from
known facilities (RCRA permitted or interim
status facilities are excluded).
(3) Section 104—grants the government authority to
respond to hazardous substance contamination in-
cidences or suspected incidences by directly employ-
ing investigation, remedial, or removal actions
(government discretion in these matters is implied).
(4) Section 107(a)—assigns strict financial liability for
response, remedial action, removal and destruction
or damage to natural resources to facility owners
and operators, past owners and operators, trans-
porters selecting facilities for disposal and persons
contracting for disposal at such facilities.
While clearly possessing the power to act, the govern-
ment will most likely begin exercising its new authority
slowly. A myriad of complicated plans, administrative
procedures and regulations required by the law remain to
be developed.
As the government acts, the regulated community and
potential defendants are not expected to sit back while
federal and state response teams descend upon them in-
curring large financial charges on their behalf. Structur-
ally, the law provides numerous opportunities for po-
tential defendants to question and challenge government
actions. However, before the proper questions can be
asked and meaningful challenges mounted, affected par-
ties must come to fully understand the scope and magni-
tude of their specific site contamination problems.
The Site Contamination and Liability Audit process
presented herein has been designed to provide industry
executives and middle level managers with a flexible out-
line of the actions needed to find and define their prob-
lems and the potential liabilities presented. After problem
definition and understanding, informed corporate re-
sponse can be fashioned.
SITE CONTAMINATION AND LIABILITY
AUDIT PROCESS
The Site Contamination and Liability Audit is a logical
stepwise procedure designed to find hazardous substance
site contamination problems, define these problems and
set the stage for informed corporate response. The audit
-------
400 LIABILITY, LEGAL & PUBLIC ISSUES
Table I.
Site Contamination and Liability Audit Phased Structure
Screening
Phases
Emergency
Action
Phase
Detailed Site
Investigation
and Remedial
Phases
Phase 1 Initial Property Inventory
Phase 2 Classification and Identification of Potential
Problem Properties
Phase 3 Preliminary Field Screening
Phase 4 Prioritization of Problem Properties
Phase 5 Immediate Emergency Stop Action Response
Phase 6 Detailed Site Field Investigation
Phase 7 Definition of Remedial Strategies, Risk and Financial
Liability Assessment and Remedial Cost Effectiveness
Phase 8 Selection of Preferred Remedial Strategy
Phase 9 Implementation of Remedial Action
Phase 10 Certification of Performance and Addressing Future
Potential Liability Issues
Table II.
Elements of Phase 1, Initial Property Inventory
Objective Gain understanding of the number, location and use of all properties under
corporate control (own, rent or lease) or planned property acquisitions.
Plant or
Facility Level
Plant or Fac. Mgr.
Plant Engineer
Environmental
Engineer
Info, and Data •Topographic map of site area (U.S.G.S. Quadrangle Map).
Req'd. for 'Site layout map showing property boundaries, acreage, building locations and
Initial adjacent property owners
Inventory •Period of time property under control
•Prior use of property, if known
•Activity summary (e.g., property is manufacturing location for chlorine gas)
•SIC (Standard Industrial Code) designation(s)
•Number of employees
•Number of production employees
•Summary of present environmental regulatory status of site (e.g., RCRA
Part A permit application filed for sludge storage lagoon, no present en-
forcement problems)
•Clearly indicate status of known contamination problems
All of the required information for this phase should be readily available to or
from the designated responsible individuals
If a large number of properties are involved, data processing may be re-
quired
Assignment of
Responsibility
Corp. Level
Director of
Envir. Affairs
Director of
Property Mgi
Environmental
Counsel
Division Level
Director of
Envir. Affairs
Director of
Property Mgt
Environmental
Counsel
Chief Engineer
Info, and
Data Sources
Equip, and
Services
Needed
Other
Considera-
tions
A questionnaire or data form may aid in obtaining uniform quality of re-
sponse from all locations
Table III.
Elements of Phase 2, Classification and Identification
of Potential Problem Properties
Objective
Assignment of
Responsibility
Info, and
Data Req'd.,
Aciion to be
Taken
Info, and
Data Sources
Equip, and
Service*
Needed
Oihef
Consideration*
Identify those properties where present or historic activity gives cause to
suspect possible site contamination
Corporate Level
Coordinator and
support staff
Division Level
Coordinator and
support staff
Plant or Facility Level
Plant or Facility Mgr and
and support staff
All Phase 1 data is reviewed by corporate and divisional coordinators. All
properties are classified: (t) known site contamination; (2) site contamina-
tion likely or suspect; (3) data for site not complete; and (4) site contamina-
tion unlikely.
•Obtain missing data for Category 3 sites
All data should be in hand as a result of Phae 1. Responsible individuals
provide missing data for Category 3 sites
If a large number of properties are involved, data processing may be re-
quired
A rating or scoring system ma> aid m the categorization of properties. All in-
dividuals involved should be counselled as to the importance of the project
Table IV.
Elements of Phase 3, Preliminary Field Screening
Objective Using simple and inexpensive field investigation techniques, confirm and
qualify site contamination problems
Assignment of Corporate and division coordinators assign a site investigation team to work
Responsibility directly with plant or facility personnel. The team should consist of a Hydro-
geologist, Environmental Scientist and an Environmental Engineer. The learn is
responsible for the design of each site screening program and for assuring that
the program is properly carried out. At least one team member should be on-
site during any field work
Site Research—Prior to field work the following data should be obtained and re-
Screening viewed:
Tasks 'Phase 1 and Phase 2 file
•Published and unpublished geologic-hydrogeologic reports and maps
•Expanded present and historic site activities report from plant or facility
personnel
•Copies of any RCRA notifications or permit applications
•Copies of any Superfund notifications
•Available site maps and surveys
•Present and historic aerial photographs of the site area
•Waste generation and disposition inventory from plant or facility personnel
(present and historic)
•Obtain any available environmental monitoring data
Field Program Design—Based on a review of the above data, a one or two day
site specific field "prospecting" program is designed. The following activities
might typically be included:
•Conduct structured interview of facility personnel followed by detailed
picture taking tour of site
•Water quality survey of standing water bodies, streams, drainage ditches,
seeps, monitor wells (if present) and surface impoundments
•Shallow soil sampling survey noting evidence of contamination (odor, dis-
coloration, etc.)
•Vegetation survey noting species present and evidence of death or stress
•Site access survey noting accessibility to general public (particularly children)
and means taken to control access
Info, and The responsible plant or facility personnel should be able to provide or
Data Sources obtain all of the research data required. Telephone assistance and advice can
be provided by the site investigation team
Equipment
Needed
Other
Considera-
tions
hand anger
post hole digger
shovel
surface water samplers
well bailers or pumps
sample containers
100 ft steel tape
35 mm camera
pH meter and probe
conductivity meter and probe
portable contaminant test
kits (HACH type)
reagents
distilled water
Enlisting the aid of responsible individuals most knowledgeable concerning
the site is strongly advised
program is flexible and can be adopted in whole or in
part as the needs in given situation demand.
Depending upon the goals of a particular audit, cor-
porate officials may require a simple problem potential
screening inventory of the properties under their control
and any planned acquisitions or a full scale site investi-
gation of the type, source and extent of contamination,
definition of possible remedial measures and costs, plan
for implementing remedial work, environmental per-
formance monitoring and liability control strategy.
Phases of the audit process which begins with general
property inventory steps and progresses through de-
tailed site specific investigation and remedial action
decision making are shown in Tables I to XI. The ele-
ments of each phase are described in detail, with at-
tention to objectives, assignment of responsibilities, in-
formation sources and data needs, equipment needs, out-
side services required and other considerations.
CASE HISTORIES—UNINVESTIGATED
LAND ACQUISITION
Asbestos Contamination
This case history clearly illustrates the need to conduct
careful screening audits of properties before purchase.
-------
LIABILITY, LEGAL & PUBLIC ISSUES 401
Late in 1978, a developer purchased an abandoned Con-
necticut gasket factory and surrounding acreage intending
to develop a shopping center on the parcel (see Figures
1 and 2). During the early stages of construction it was
discovered that the site was dotted with asbestos sludge
pits and asbestos gasket waste piles.
Concerned for the health of workers and area residents,
local health officials contacted State and Federal officials.
Teams of inspectors from the Connecticut Department of
Environmental Protection (DEP), U.S. Environmental
Protection Agency (EPA) and the Occupational Safety
and Health Administration (OSHA) descended on the
site. Construction was immediately halted.
After several days of negotiation, the developer signed
a DEP consent order. THe twenty compliance conditions
specified in the order involved everything from inform-
ing workers of the hazards associated with asbestos, to
the consolidation and proper on-site disposal of all as-
bestos wastes. Construction delays and compliance costs
added $900,000 to development costs. If a simple and in-
expensive site contamination screening audit had been
conducted, the developer would have avoided the prop-
erty and its problems.
Coal Tar Contamination
A Pennsylvania utility was notified by the U.S. EPA
that an oily discharge from one of its properties was
entering an important trout stream. The utility con-
tracted with TRC to investigate the source and nature of
the material and to delineate the extent of contamination.
A review of the land use in the site area showed that
coal gas had been manufactured on the site by the previ-
ous owner and that the oily substance seeping into the
stream was coal tar. Aerial photographs dating back to
1939 were used to identify suspect areas and help de-
sign the drilling and sampling program. The boundaries
of the contaminated area have been identified and the
feasibility of various containment and removal alterna-
tives is being studied.
Table V.
Elements of Phase 4, Prioritization of Problem Properties
Objective
Assignment of
Responsibility
Rating and
Prioritization
Tasks
Rating and
Prioritization
Info, and
Data Sources
Other
Considera-
tions
From field screening data develop preliminary estimate of the magnitude of
potential contamination and risk presented by each site and rate properties ac-
cordingly. Priority for site specific investigation and remedial action is estab-
lished
Corporate and divisional coordinators, site investigation team, and corporate
or outside counsel
Rating Potential for Health and Environmental Damage
Site rating methodologies have been developed by: 1) LeOrand;' ' 2) JRB
Associates;'2' and 3) the Mitre Corporation.'3' The methods consider po-
tential receptors, pathways for contaminant migration, contaminant character-
istics, and site engineering characteristics. These methods vary in specific data
requirements and emphasis. Preference should be given to the development
of a rating methodology synthesized from these approaches that is simple to
apply and requires only that data already in hand as a result of Phase 1
through 3 efforts. Using the method developed, each site is evaluated and given
a numerical rating.
Rating Liability/Risk Potential
Close approximations of the financial liabilities presented by a site cannot be
made until detailed site investigation work is completed and remedial action
option costing developed. However, a qualitative assessment of liability risk
potential can be made by answering the following questions:
•Is the property presently a source of off-site contamination?
•Has off-site environmental or property damage been documented?
•Are environmental regulatory enforcement or legal actions in progress or
pending?
•Has a "Superfund" notification been filed for the site?
•Does the site contain poorly maintained discrete accumulations of hazardous
wastes (i.e., drums, waste piles, sludge, etc.)?
•Is the property accessible to the general public?
•Are any private legal actions in progress or pending?
•Has environmental impairment liability insurance for the site ever been
denied?
Based on the answers to these questions, the potential financial liability is
rated High Risk, Moderate Risk or Low Risk. Three or more yes answers
should place a site in the high risk category.
Prioritization for Further Action
Sites with high health and environmental damage scores and a high risk lia-
bility rating would obviously be priority candidates for immediate emergency
stop action response and detailed site investigation
Phase 1 through 3, should have provided all the information and data needed
for Phase 4 prioritization. Follow up telephone conferences may be required
for clarifying data and answering liability risk questions
The value of legal counsel at this stage is of critical importance. The cor-
porate legal staff should be immediately briefed concerning all high score/high
risk sites
Table VI.
Elements of Phase 5, Immediate Emergency Stop Action Response
Objective Prior to the start of and during the detailed field investigation problems may
be uncovered that require immediate emergency remedial action. Rapidly con-
ceived and implemented remedial action is taken to mitigate imminent public
endangerment or potential catastrophic environmental release.
Assignment of Corporate and divisional coordinators, site investigation team, facility per-
Responsibility sonnel, corporate legal staff and outside contractors develop and implement
emergency remedial measures
Typical Stop
Action
Remedial
Measures
Equipment and
Services
Needed
Other
Considera-
tions
Actions taken will be site specific, responding to the special problems pre-
sented at each site. Typically, these measures might include:
General Response
•Immediately notify and brief all local emergency response officials
•Immediately provide site security—fencing, warning signs and 24-hour secur-
ity personnel
Drums and Tanks
•Identify contents and quantity, determine structural integrity, transfer con-
tents if necessary, label, number and properly stage
•As soon as practical, remove all containerized material for proper disposal
Pits, Ponds, Lagoons, Surface Impoundments and Waste Piles
•Identify contents and stored volumes, determine engineering specifics and
structural integrity of any liners or containment structures
•Take measures as necessary to eliminate or control contaminant release and
improve integrity of containment, such as:
1. cease discharge of waste to facility
2. remove waste for proper disposal
3. transfer material to secure temporary storage
4. install top liner
5. increase dike or berm elevation and
6. stabilize waste physically and chemically in situ by fixation and/or
solidification
Ground and Surface Water Contamination
Before any effective long-term program to eliminate or control ground and
surface water contamination can be designed and implemented, extensive
site investigations must be performed. However, the following stop gap
measures can be instituted to provide some degree of temporary control.
•Identify obvious pathways of contamination such as drainage ditches, seeps,
outflow pipes, etc. and seal them off
•Divert surface drainage away from contaminated areas
•Install combinations of cutoff walls, collection ditches and trenches to con-
centrate and collect strong leachates for treatment and proper disposal
To be determined on case specific basis
All emergency remedial work should be conducted by trained, experienced
and properly equipped personnel or outside contractors. Work must be
planned in detail and contracts carefully drafted. Contract provisions should
include proper insurance, attention to regulatory detail and adequate emer-
gency procedures and contingencies
-------
402 LIABILITY, LEGAL & PUBLIC ISSUES
Table VII.
Elements of Phase 6, Detailed Site Field Investigation
Object)', e A complete field investigation is designed and implemented to include samp-
ling and analysis of wastes and monitoring, sampling and analysis of area
soils, surface waters and groundwaters. The nature and extent of contamina-
tion, present impacts and potential future impacts are determined.
Assignment of This investigation will need to be conducted by outside contractors selected
Responsibility by the corporate, divisional and plant coordinators, together with the ap-
propriate legal staff
Detailed Preparation Tasks
Field •Assemble and review all data collected through previous phases
Investigation «Thc importance of site historical data cannot be overslressed. Prior to con-
Tasks ducting expensive field work, the record must be clear concerning property
chain of custody, hazardous materials use, and hazardous waste generation
and disposition. Techniques for obtaining missing information include:
searching land records; obtaining historic air photos and maps; reviewing old
industrial directories and commerce reports; interviewing local historians;
and interviewing employees of long standing, former employees and em-
ployees of past operations
•Prepare request for proposals for detailed site investigation
•Review proposals, interview finalists and select environmental consulting
contractor
Typical Field Tasks
•Complete a photogrammetric survey of the site area. Maps of a scale I" =
20' or 1" = 50' with a 1' or 2' contour interval are usually required
•Using the site maps and aerial photographs, prepare a detailed surface and
subsurface exploratory program plan. The plan considers what data is already
known and provides for the stepwise collection of the remaining data needed
to meet the Phase 6 objective. The program would typically consist of the
following:
I. Sampling and analysis of discrete waste accumulations, for which chemi-
cal analyses do not already exist
2. Based on waste characteristics and analyses, select contaminant paramet-
ers, contamination indicator parameters, and ground and surface water
quality parameters for subsequent monitoring, sampling and analysis work
3. Conduct surface water sampling and analysis program for any standing or
flowing water bodies immediately upgradient, on, or immediately down-
gradient of the site
Equipment and
Services
Needed
Other
Considera-
tions
4. Complete test pit and shallow soil boring program, logging and sampling
materials encountered
5. Install piezometers and/or groundwater monitoring equipment in selected
shallow borings and test pits which intersect the water table
6. Conduct deep drilling and boring program, collecting split spoon samples
at 5' intervals or change in Uthology
7. Install piezometers and/or groundwater monitoring equipment in selected
borings and drill holes
8. Perform pump tests on selected well installations to aid in defining site
hydrogeologic characteristics
9. Conduct geophysical survey (eletrical resistivity and/or seismic) to confirm
and augment subsurface data and
10. Define surrounding property use, present natural resource use and po-
tential future nature resource use
Data Reporting and Evaluation
•Even if a field investigation program is well designed and carried out, sur-
prise discoveries will require modification of the plan as work proceeds. This
is the norm rather than the exception
•At the conclusion of the investigation sufficient data will be in hand to de-
termine the extent and nature of contamination, present impacts and po-
tential future impacts. Typical outputs might include:
1. Results of waste, soil, surface water and groundwater sample analyses
2. Maps showing discrete waste accumulations, types and volumes
3. Maps presenting surface water quality data and flow information
4. Soil contamination maps and cross sections
5. Geologic (bedrock and surficial geology) maps and cross sections
6. Results from field and laboratory permeability tests, transmissivity cal-
culations
7. Hydrogeologic maps, cross sections and flow nets and
8. Groundwater contamination maps, cross sections and flow nets
To be determined on case specific basis
The contractor should be a full service environmental consulting firm with
site contamination investigation experience and staff expertise in the fields of
Geology, Hydrogeology, Chemistry, Environmental Engineering, Industrial
Engineering, Chemical Engineering, Hydrology and Environmental Science
Table VIII.
Elements of Phase 7, Definition of Remedial Strategies, Risk and
Financial Liability Assessment and Remedial Cost Effectiveness
Objective Several removal, containment and combination scenarios are developed to
remedy the environmental and related health problems presented by the site.
The costs of implementing each scenario arc calculated. Relative cost ef-
fectiveness is then determined for each case in terms of implementation dol-
lars paid out versus estimated financial liability risks remaining after imple-
mentation.
Assignment of It is recommended that the contractor on board for Phase 6 be retained for the
Responsibility Phase 7 efforts, working closely with the corporate, divisional and facility co-
ordinators and assigned legal staff.
Remedial Removal Scenario
Scenario *ln cases where discrete waste accumulations of known type and volume are
Development the sources or potential sources of significant environmental and/or health
damage, removal and proper disposal at permitted treatment and disposal
facilities is highly recommended
•Costs of removal and proper disposal are easily determined by obtaining
quotes from reputable disposal contractors
•This scenario assumes that any residual contamination left after removal is
insignificant and of no future consequence
Containment Sctnanos
•Containment strategies seek to control the release of hazardous contaminants
to the environment. Control is achieved by isolation, partial isolation,
contaminant collection, physical and chemical stabilization and alteration
of ground and surface water flow to prevent off site contaminant discharge
•In a given situation one or any combination of containment strategies
might be used
•Unit costs are determined using conceptual designs and standard engineer-
ing costing procedures
•Containment scenarios assume that long-term maintenance and monitoring
»ill be required to assure conunued performance of the containment goals
•Monitoring and maintenance costs for an assumed period (usually 20-30
years) are computed, adjusting for inflation, and added to scenario imple-
mentation costs
Combination Scenarios
•In a given situation removal and any combination of containment strategic!
might be used
Cost Costs
Effectiveness •Costs, as computed above, are taken as given for each particular remedial
scenario
•Cost effectiveness is determined by employing an event/consequences analysis
Eveni/Consequenct Analysis
•The event/consequence analysis is a semi-quantitative con effectiveness
assessment of the degree of risk control provided by implementing a re-
medial scenario of known cost
•Each remedial scenario is evaluated in relation to the answers to the follow-
ing questions:
I. How might the strategy fail?
2. What is the probability of failure?
3. Are obvious or planned fall back strategies available to detect and correct
failure situations?
4. What financial risk is exposed as a result of failure in terms of resource
restitution, replacement, awards for health and property damage, and en-
forcement penalties?
5. It the financial risk manageable; or should a remedial scenario requiring
larger initial investment, but posing substantially lower long-term financial
risk be employed?
Other It is recommended that an insurance consultant who is expert in enviroo-
Considera- mental liabilitiy matters be retained to review con effectiveness issues related
tions to financial liability risks
-------
LIABILITY, LEGAL & PUBLIC ISSUES 403
Table IX.
Elements of Phase 8, Selection of Preferred Remedial Strategy
Objective After careful examination of the pros and cons attending each remedial
scenario developed through Phase 7 efforts, a preferred strategy is selected for
implementation. The alternatives considered and the rationale for selection of
the preferred strategy are presented to the appropriate regulatory agencies for
review, comment, negotiation and final concurrence or approval
Assignment of Corporate, divisional and facility coordinators working with assigned legal
Responsibility staff and outside consultants
Info. Req'd. The data needed for presentation of the strategy selection process should be
in Data on hand as a result of Phase 6 and 7 tasks. This information will
include: detailed field investigation reports and maps; remedial scenario con-
ceptual designs and costing data; and cost effectiveness and financial liability
risk evaluations
A statutory and regulatory compliance brief should be prepared by the legal
staff to determine compliance issues and status related to implementation of
the preferred remedial strategy. Details concerning the need for any regulatory
permits and steps to be taken to maintain compliance throughout remedial ac-
tion implementation should be addressed
Other The importance of regulatory agency concurrence and approval of the re-
Considera- medial strategy cannot be overstressed. While in certain cases approvals are
lions not required, obtaining consent is still recommended. This reduces the poten-
tial for costly regulatory agency second guessing during and after remedial
work.
The legal staff must be convinced that any instrument (letter of approval, con-
sent agreement, compliance order, permit, etc.) issued by the appropriate
regulatory agencies is both understandable and conclusive.
Table X.
Elements of Phase 9, Implementation of Remedial Action
Objective The engineering, schedule, logistics and all contractual arrangements are
finalized. Any required permits are obtained. Remedial work is initiated and
completed.
Assignment of Corporate, divisional and facility coordinators work with assigned legal
Responsibility staff, outside consultants, clean up contractors, construction contractors and
disposal contractors.
Contract Most corporations have standard construction and service contract formats.
Considera- These formats often do not consider contract conditions needed for work re-
tions lated to the handling, transport and disposal of hazardous substances. Topic
areas to be considered when drafting site remedial work contracts are listed
below.
1. Definitions 20.
2. Overall Project Management 21.
3. Cleanup Operations Management 22.
4.Construction Management
i. Security 23.
6. Emergency Procedures 24.
7. Contingency Plans 25.
8. Personnel Safety and Training 26.
9. Safety Equipment 27
10. Utilities 28.
II. Materials Handling 29.
12. Transportation and Permits 30.
13. Treatment and Disposal Methods 31.
14. Treatment and Disposal Permits 32.
15. Record Keeping and Reporting 33.
16. Insurance 34,
17. Performance Bonds 35.
18. Indemnification 36
19. Subcontracting Authority 37
Inspection of Work
Sampling and Analysis Protocol
Services to be Performed,
Responsibilities
Facilities to be Provided
Materials to be Provided
Equipment to be Provided
Personnel to be Provided
Project Start and Completion Dates
Project Milestones
Compensation to Contractor
Method and Schedule of Payment
Rate Schedule
Amendment Clause
Payment of Taxes
Title and Disposition of Equipment
Release and Discharge
Termination
Force Majeure
Figure 1.
Fitzgerald Gasket Plant, Torrington, Connecticut, 1976
Figure 2.
Shopping Mall On Former Fitzgerald Gasket Property,
Torrington, Connecticut, 1980
Table XI.
Elements of Phase 10, Certification of Performance and
Addressing Future Potential Liability Issues
Objective Confirm that remedial work has been completed as planned and establish
mechanisms to control any future potential site liability
Assignment of Corporate, divisional and facility coordinators work with assigned legal staff,
Responsibility outside consultants and regulatory agencies
Certification *A licensed professional engineer certification of as-built plans and specifica-
of Performance tions for the remedial work is obtained
•The certification is reviewed by the appropriate regulatory agencies, in-
spections conducted and conclusive regulatory sign-off obtained via:
1. Letters of approval
2. Release of bonds
3 Statements of compliance with stipulated agreements
4. Statements of compliance with regulatory compliance order conditions and
5] Statements of compliance with specific regulations
Control of "Any required long-term monitoring and maintenance is initiated to detect
Remaining and control future liability
Liability «The availability and economic feasibility of obtaining environmental im-
pairment liability insurance, self insurance, or other forms of indemnification
is investigated
•Nonsudden environmental impairment insurance is now offered by the fol-
lowing firms:
1. Shand Morahan 5. Swett and Crawford Management
2. Alexander and Alexander 6. Travelers
3. American Intern'l Group 7. Aetna
4. Stewart-Smith 8. U.S. Insurance—Lloyds of London
Property "Property transactions involving hazardous substance site contamination
Transfer must consider the responsibilities of the seller, buyer, lessor and lessee with
relations to liability, site maintenance and site monitoring
•Responsibilities and liabilities can be shared or assumed by one party
•Contracts and sales agreements must be carefully worded by attorneys
working closely with technical experts
-------
404 LIABILITY, LEGAL & PUBLIC ISSUES
CASE HISTORY—INVESTIGATION AND
CLEANUP BEFORE PROPERTY PURCHASE
PCB Contamination
TRC performed a contamination evaluation for a
pharmaceutical manufacturer. The need for the study
arose when the potential purchases of the site required
that the present owner certify that the site was not con-
taminated. The TRC program included a number of
shallow borings to determine whether soil or unsaturated
zone contamination had occurred through spills. A test
well was drilled to determine the quality of site ground-
water. Because numerous industrial sources existed in the
area, routine analyses were keyed to an organic chemi-
cal produced by the previous owner. These routine analy-
ses showed that some product spills had occurred but
that the level of contamination was not significant.
Analyses for PCBs, however, showed that several spills
had occurred around the site. While levels did not ex-
ceed "regulated" amounts it was decided to delineate
and remove the contaminated soils to avoid future prob-
lems.
CONCLUSIONS
In response to a clear public mandate, statutory and
regulatory initiatives conceived to seek out, control and
assign liability for hazardous substance and hazardous
waste site.contamination problems are now in place. The
site contamination and liability audit process presented
herein has outlined methods for locating, investigating
and controlling hazardous site contamination problems.
The property screening phases are the most valuable
aspect of the audit process. Screening identifies hidden
problems before time can make them worse. More im-
portantly, however, is that a low cost investment in prop-
erty screening prior to purchase can prevent the acquisi-
tion of liabilities of vast dimension.
REFERENCES
1. LeGrand, H.E., "A Standardized System for Evaluat-
ing Waste-Disposal Sites," National Well Water As-
sociation Publication, 1980.
2. Kufs, C., et al., "Methodology for Rating the Hazard
Potential of Waste Disposal Sites," JRB Associates,
Inc., McLean, Virginia, May 1980.
3. Chang, S., Barrett, K., Haus, S., and Platt, A. "Site
Ranking Model for Determining Remedial Action
Priorities Among Uncontrolled Hazardous Substances
Facilities," Working Draft, The Mitre Corporation,
Metrek Division, McLean, Virginia, June 1981.
4. Anonymous, "Insuring Against Liability Under
RCRA," Chemical Week, August 19, 1981.
-------
INSTITUTIONAL LEARNING IN A BUREAUCRACY:
THE SUPERFUND COMMUNITY RELATIONS PROGRAM
STEVEN A. COHEN
Columbia University
New York, New York
THOMAS G. INGERSOLL
West Virginia University
Morgantown, West Virginia
JAMES R. JANIS
ICF, Incorporated
Washington, D.C.
INTRODUCTION
With the passage of the Comprehensive Environmental
Response, Compensation, and Liability Act in December
1980, the federal government acquired funding and au-
thority to clean up the multitude of abandoned hazardous
waste dumps threatening public health and the environ-
ment across the nation. Primary responsibility for ad-
ministering the Act's provisions was assigned to the U.S.
Environmental Protection Agency (EPA). EPA had, how-
ever, begun planning its program—the "Superfund"
program—some six months in advance, recognizing that
the widespread publicity given cases like Love Canal
made the need for such legislation clear.
In this paper, the authors analyze the planning and
development of one component of EPA's Superfund pro-
gram—community relations—as an illustration of a gen-
eral model for institutional learning in a bureaucracy.
The model is presented first in outline form. Then, in the
third section, the development of the Superfund com-
munity relations program is described to illustrate the
model. In the fourth section of this paper, the authors
analyze the structure of the community relations pro-
gram, in its present form, in order to relate the bureau-
cratic learning process to organizational aspects of the
performance of program functions. The adequacy of the
model to the learning process in other kinds of institu-
tions is commented upon in the conclusion.
The distinguishing feature of the development of the
Superfund community relations program, as a case of
institutional learning, is that it constituted more a re-
vision of an institutional memory than the acquisition of
new knowledge. EPA, as a ten-year-old institution, was
not only well-acquainted with the kinds of functions that
comprise community relations, but in fact, had a signifi-
cant bias toward their inclusion in programs aimed at
the protection and enhancement of the environment.
Thus, the learning that took place in the evolution of
Superfund community relations did not begin with a
tabula rasa; rather, it refined and redirected familiar func-
tions. But it did so in a political climate that was becom-
ing unreceptive to the value of those functions.
A Schema for Institutional Learning
The learning process illustrated by the Superfund
community relations program can be represented by
the following steps:
•The recognition of a problem to which the institution
must respond.
'Research on the nature of the problem and its implica-
tions for institutional organization and performance
•Development of policy based upon the research con-
ducted
•The development of means for disseminating the policy
•The acquisition of programmatic means for implement-
ing the policy
•Monitoring and evaluation of the program implemented
The learning process must not be equated with the re-
search step. Rather, it should be regarded as all those ac-
tions that lead to the incorporation of relatively novel
functions in the on-going activities of the institution. The
learning that takes place, in other words, is constituted
by the alteration of institutional behavior and the ac-
tions leading up to that alteration. Data-gathering, sub-
sequent analysis and the broadening understanding of
the staff of the institution are only steps in the process.
These steps may serve primarily as instruments useful in
changing policy rather than being valuable in their own
right as the acquisition of knowledge.
The alteration of institutional behavior is the result of
changes in three areas as a result of the learning process:
•The institution's self-conception.
•The institution's operations.
•The program in question and its results.
These factors require some explanation.
Corresponding to (or produced by) the institution's
memory is a conception of its role and significance
that is passed down through the staff. This self-concep-
tion may or may not accord with the conception that
others—e.g. the public or members of sister institutions—
have of the institution. In any event, the self-conception
will dictate, to some extent, the manner in which the in-
stitution's functions are performed.
405
-------
406 LIABILITY, LEGAL & PUBLIC ISSUES
On a more formal level, the institution's policy will be
codified in certain explicit standard operating procedures.
A learning process will alter policy, alter standard op-
erating procedures in turn and, ultimately, alter the ac-
tivities of the institution.
Finally, the program under whose auspices the learn-
ing process begins will be directly affected, even in its
structure and operations, by that process. These should be
consequences for the results obtained by the program,
that is, the success (or failure) in various aspects of pro-
gram activities.
The development of the Superfund Community Rela-
tions program illustrates this learning process, and the
factors changed through it, against the backdrop of a
prior institutional memory.
COMMUNITY RELATIONS POLICY
DEVELOPMENT
In April 1980 EPA established the Office of Analysis
and Program Development (OAPD) to provide pre-
implementation planning and legislative support for
the anticipated Superfund program. By mid-1980, OAPD
had established the theoretical parameters of the commun-
ity relations program. The office conceived of it as a co-
ordinated program of federal government interaction with
a local community where that community is the subject of
a federal action. Community Relations was considered to
include local media and press relations, public participa-
tion, local government and local interest group relations
and public information/education.
With the degree of public, congressional and media in-
terest in the pending Superfund legislation, OAPD recog-
nized that its planning process had to emphasize effec-
tive response. In emergency situations, an effective re-
sponse was considered one that ends an existent state of
emergency and permits a return to substantially pre-
emergency conditions. In proposed remedial action, an ef-
fective response was considered one that would permit
the Agency to maintain its credibility while eliminating
the environmental hazard. In either case, the Agency
recognized that its response must not only remove the
peril to the environment and to human health but also
remove the perceived threat to human and ecological well-
being as well.
With this understanding as background, OAPD pro-
posed that the primary goal of the community relations
program be to facilitate the implementation of cost-
effective solutions to hazardous waste problems by en-
suring adequate communication between government and
local communities during response actions. It was as-
sumed from the beginning that hazardous waste inci-
dents had the potential to generate extremely emotional
reactions in local communities. Thus, government ac-
tions in this extremely volatile political environment
would have to be explained with great thoroughness and
care if they were to be fully understood by the commun-
ity. In order to achieve the goal of ensuring adequate
communication, a number of secondary goals were ad-
vanced:
•Ensuring the local community a meaningful voice in
those implementation decisions that the community
considers most important
•Establishing a program of public information and media
relations that is appropriate to the degree of interest and
concern about the site
•Anticipate potential conflicts and attempt to avoid con-
flict whenever possible
•Establish standard operating procedures that ensured
extensive interaction between local government of-
ficials and federal/state officials.
OAPD generated an issue paper which outlined these
goals and the resources needed to reach them for a pro-
gram of the scope envisioned by Congress. It was antici-
pated that the Agency would be expected to have the or-
ganizational and technical capacity to address commun-
ity concerns. This would only be possible if, from the out-
set, community relations were declared a major program
element in Superfund. OAPD reasoned that waiting un-
til political pressures forced community involvement on
EPA would probably result in the loss of the opportunity
to manage community relations to facilitate rather than
impede cleanup actions.
Although the Superfund program was to have many
elements which would be new to EPA implementation,
the Agency was not without experience in emergency re-
sponse to hazardous waste incidents. Much valuable ex-
perience had been gained over almost a decade in the im-
plementation of Section 311 of the Clean Water Act. In
order to avoid serious blunders in the new program,
OAPD would have to do a lot of learning in a hurry. To
do so, it let a contract to ICF Incorporated to take a seri-
ous look at several hazardous waste emergency opera-
tions which had been conducted either under Section 311
or under state auspices in order to analyze the degree to
which community relations efforts had been effective. In
short, the purpose was to analyze what had been accom-
plished at these sites, what had been done well, where the
Agency had failed to provide for effective community/
government interactions and generate the lessons which
would be applicable to the greatly expanded scope of the
Superfund program.
The analysis was based on the assumption that with
knowledge of what had actually taken place in hazardous
waste emergencies across the country, a more realistic
program could be designed—one that was more under-
standing of, and responsive to, the needs of the public
and the capabilities of government. In addition, by look-
ing at the community relations techniques which had been
employed in the past in these situations, and seeing which
had worked well and which had not, EPA would learn
what techniques to use in the future or to avoid. The re-
sult, it was hoped, would be a program which would more
effectively—and efficiently—involve citizens in their gov-
ernment's response to these environmental problems and
inform them of actions underway in their community.
The methodology employed was to learn as much as
possible about the cases through newspaper accounts and
interviews with officials in Washington—the staff of local
congressmen and EPA personnel—and then to travel to
-------
LIABILITY, LEGAL & PUBLIC ISSUES 407
the site for at least a week of intensive interviewing. Of-
ficials in state environmental agencies and Regional
EPA offices were consulted at length. Interviews at the
site were conducted with citizens who were actively in-
volved in the problem, civic leaders, local government
authorities and environmentalists—as well as with people
living in the vicinity who had taken no interest or ex-
pressed no concern. The completed studies were en-
visioned as background material for the development of a
handbook that would assist EPA personnel responsible
for implementing the Superfund program.
By August 1980, it was clear that the researchers were
breaking new ground in the understanding of community
relations at the four sites under review, but it was equally
evident that a more complete picture of community re-
lations around the country was required. These initial
studies demonstrated a wide range of citizen concern with
and involvement in hazardous waste emergency opera-
tions; thus EPA decided to expand the study to include
sites in each of^the ten Regions. Site selection was based
on input from 311 emergency operations personnel
throughout the country, with the final selection geared
toward sites at which Superfund operations might be ini-
tiated. The cases analyzed in this second round brought
the total number of cases in the study to 21.
Even as the results of the expanded series of case
studies were being analyzed, EPA began moving into the
field with draft guidance to the Regions on the probable
extent of a community relations program under Super-
fund. The first draft of a guidance document was sent to
the Regions based on a preliminary analysis of the case
studies and the Superfund headquarters organization re-
quested that the Regions comment on the effectiveness of
the proposed guidelines. The draft manual suffered from
the fluidity of planning regarding the shape of the final
program. Although the legislation had now been passed,
the new Administration was not yet established and the
shape of the final program was in some doubt. By the end
of February, however, Regional comments had been re-
ceived and catalogued by headquarters and Michael B.
Cook, EPA's Deputy Assistant Administrator for Haz-
ardous Emergency Response, issued interim community
relations guidance for site cleanup.
The interim guidance document established the policy
by which Superfund implementation would be guided in
the community relations area. This policy requested that
the Regions adhere to the following principles:
•informing the local community about Agency actions
•empathizing with local concerns, learning about the lo-
cal community
•avoiding the generation of unrealistic expectations
•being open and forthright with information
•anticipating the formation of ad hoc citizen groups
•coordinating actions with local officials
•assigning community relations coordinators whenever
possible
•using a variety of participatory techniques
•considering the establishment of citizen advisory com-
mittees at sites and spills having a high degree of citizen
concern
•providing adequate training for Regional staff.
Operationally, each of these principles had been derived
from the 21 case studies which had been undertaken by
ICF and each was factored into the newly designed re-
medial and emergency response activity schedules for the
Superfund program.
The guidance document for community relations during
remedial response activity was divided into discrete re-
sponse phases and informed by a determination, on the
part of the EPA on-scene coordinator (OSC), of the de-
gree of citizen concern at a site, combined with the ex-
tent of environmental damage and the complexity of the
proposed technical solution. The determination of the de-
gree of citizen concern was to be based, at least in part, on
the results of interviews conducted by the OSC at the site
among local officials and concerned citizens.
The methodology which had been employed by ICF in
conducting the case studies had been found to be so ef-
fective in ferreting out the concerns of the local commun-
ity that the interim guidance incorporated that methodol-
ogy into the schema for Superfund response at all re-
medial actions. In addition, the value to the Agency of
the ICF case studies was exemplified in the requirement,
placed in the interim guidance, for the development of a
Community Relations Plan for each site at which remedi-
al action was to be taken. The plan was to be designed
and executed as an integral part of the remedial response,
not simply appended to the response as an afterthought.
During emergency response actions, the interim guid-
ance document noted the importance of keeping the com-
munity informed of the Agency's actions and the need for
the establishment of a close working relationship with
local officials. The same principles which mandated the
generation of a community relations plan during remedial
response prompted the inclusion of a similar plan in those
emergency situations which might be followed by remedial
action. As a result, the community relations component
of the Superfund program had become a significant ele-
ment of all responses which would be taken by the Agency.
By the summer of 1981, the National Contingency Plan
(NCP), first developed for the implementation of the
Clean Water Act, was ready for publication in the Fed-
eral Register. The newly revised NCP, as mandated by
CERCLA, was to contain the final guidance for the im-
plementation of the Superfund program. It would out-
line, in detail, the steps which could be taken at remedial
and removal operations under the authority of the Act.
Unlike earlier versions of the NCP, the newly issued
guidance fully incorporated the Community Relations
function into the operation of a response. Even the format
of the NCP points to this incorporation. The text of
the new NCP mandates the inclusion of all of the com-
munity relations techniques, including the methodology
developed for ICF's case studies, the use of the commun-
ity concern and degree of technical difficulty factors in
determining the level of response necessary for an ef-
fective community relations program, and the inclusion of
Community Relations Plans in all response actions.
-------
408 LIABILITY, LEGAL & PUBLIC ISSUES
Annex XI of the NCP, titled "Community Relations,"
specifies the mechanisms to be used, the design of the
community relations plans, and the nature of the commun-
ity relations program. The Annex states that:
"All community relations functions will be fully in-
tegrated within the operational units responding un-
der provisions of the Plan, and all aspects of response
operations must include provision for the execution of
community relations programs."
THE ROLE OF EPA HEADQUARTERS, REGIONS,
AND PRIVATE CONTRACTORS IN
IMPLEMENTING COMMUNITY RELATIONS
PROGRAMS
The basic thrust of all efforts in headquarters, Re-
gions and by contractors was to implement community
relations programs in the field. Superfund resources
would not be used for a nationwide education campaign,
but instead, would be directed to support response actions.
Headquarters
Headquarters' role included the lead on policy formula-
tion, reviewing community relations plans, training, pro-
gram evaluation and resource analysis. Headquarters pro-
vided informational materials on the Superfund program,
expert personnel for temporary community relations work
at critical sites, overall contract management and other
support services in the field.
Regions
Regions would take the lead on implementing commun-
ity relations programs, developing site-specific public in-
formation materials, supervising regional community re-
lations subcontractors and developing community rela-
tions plans for headquarters. The community relations
load unit was a matter of Regional discretion.
Contractors
Contractors would provide support services to head-
quarters and regional community relations program ac-
tivities. At headquarters, this would primarily be analytic
work in support of guidance and policy revisions. In the
Regions, contractor work would be staff and background
work to permit the Regions to stretch their own person-
nel to allow coverage at as many sites as possible. One
essential principle was to be maintained throughout the
program: contractors would never represent, or even ap-
pear to represent the Agency to the public, other govern-
ment officials, or the media (i.e., to anyone).
IMPLEMENTING COMMUNITY RELATIONS
Organizational Strategy: Headquarters
One of the four branches in the original design for the
Superfund DAA staff office was the "Community Gov-
ernmental Relations Branch." In the original proposal,
the Branch had three sections: (1) Community Relations,
(2) Public Information, and (3) Government Relations.
Typically, minimal staffing for a section would be three
professionals, one clerical person, and a supervisor.
When staffed, the Branch would have approximately 15
members. The original organizational proposal stated
that:
"The general functions of this Branch would be to:
(1) develop public participation/information policy
and guidance, (2) carry out a coordinated program of
public education and participation, (3) provide com-
munity relations support to on-scene coordinators at
critical abandoned waste sites, and (4) coordinate
the collection and dissemination of technical informa-
tion and public education materials."
The proposal was considered by the Superfund Na-
tional Program Manager. He decided not to establish a
distinct organizational unit to carry out this function. In-
stead, community relations policy development and co-
ordination was lodged in the Policy Analysis Branch of
the Deputy Assistant Administrator Staff office in the
Superfund headquarters organization and each operating
division was held responsible for ensuring the imple-
mentation of community relations policy within its own
functional area.
The rationale for this decision was that community re-
lations was to be an integral part of the Superfund pro-
gram: a normal, routine part of doing business. Just as
there would be no distinct organizational unit established
to monitor the health and safety of staff on-site, there
would also be no need to establish a distinct organiza-
tional unit to handle community relations.
This approach has certain intrinsic strengths and weak-
nesses. On the positive side, it helps ensure that commun-
ity relations could be a fully integrated component of
response. Those organizations making technical decisions
would also be responsible for ensuring that local com-
munities would be informed of and involved in response
actions. This type of organizational structure ensures that
community relations programs will not develop a "life of
their own," and will be given priority only in connection
with central technical functions. A more significant ad-
vantage of the integrated organization strategy is that a
difficult-to-explain, vulnerable function is given lower
visibility and protected from attack from OMB, Con-
gress and others.
In early 1981, when the Reagan Administration sub-
mitted the FY '82 budget, almost every public participa-
tion organizational unit in EPA was eliminated. It is pos-
sible that a distinct organizational unit for Superfund
Community Relations would have suffered a similar fate.
Nevertheless, a distinct organizational unit for com-
munity relations could bring significant benefits. Such an
arrangement allows expertise to be accumulated and then
concentrated in a single place. It also ensures that a super-
visory-level individual would be held responsible for de-
veloping and implementing the community relations pro-
-------
LIABILITY, LEGAL & PUBLIC ISSUES 409
grams. In the Superfund program, community relations
was a major responsibility of the headquarters policy
analysis branch chief, but was only one responsibility
among many.
The Superfund headquarters community relations or-
ganization was informal. At the staff level, Steve Cohen
was responsible for developing policy and ensuring its
overall coordination and implementation. Each operating
division in the Office of Emergency and Remedial Re-
sponse had a single individual responsible for ensuring the
implementation of community relations policy in Regional
office programs. These individuals met for a short time
as an "Informal Working Group," to work with Cohen
and develop a program to implement community rela-
tions policy. This informal group met several times be-
fore its collapse. Cohen found it more efficient to work
individually with each division's lead community rela-
tions person, rather than convene work group meetings.
Decisions reached by the group proved difficult to imple-
ment due to the fact that each community relations
"coordinator" had to clear all tasks with his or her
formal supervisor.
Regions
Regional office Superfund organizations evolved
throughout 1980 and 1981. During that time, community
relations was not organized according to any single pat-
tern. In mid-1980, headquarters asked each Regional
Administrator to name a Regional Superfund coordinator.
These coordinators, in most cases, came to head the Re-
gion's Superfund organizational unit. In late 1980, head-
quarters requested that each Superfund coordinator name
a Community Relations Coordinator. These community
relations coordinators were responsible for developing
community relations plans and ensuring the implementa-
tion of those plans. In some Regions, plan development
and implementation was assigned to the Public Affairs or
External Relations Office.
DEVELOPING AND IMPLEMENTING
COMMUNITY RELATIONS PLANS
The Interim Community Relations Guidance issued on
February 25, 1981, required the submission of commun-
ity relations plans. Because of the shortage of EPA per-
sonnel, the headquarters Superfund office planned to pro-
vide contract resources to the regions to assist in plan im-
plementation. Throughout the spring and early summer of
1980, the Hazardous Site Control Division (OERR) re-
quested, received and reviewed 30 community relations
plans. Although some of these plans were quite good,
overall:
•The majority of the plans did not contain a sufficient
quantity of the high quality information needed to make
an independent judgment regarding the adequacy of the
plans
•Very little attention was given to the technical complex-
ity of sites (as required in the Guidance)
•Most of the plans followed headquarters guidelines for
cost estimates, but it was difficult to determine whether
or not the level of community relations activity recom-
mended was desirable given the lack of background in-
formation
In order to develop consistent community relations plans,
it was decided to also provide contractor assistance to the
regions for developing these plans.
The specific contractor assistance scheme which began
in August 1981, followed early in the Superfund pro-
gram provided the regions with staff assistance from
Architecture and Engineering (A&E) firms during tech-
nical work on-site, and from the Technical Assistance
Team (TAT) contractor. ICF Incorporated, the firm that
developed much of the community relations program
guidance material, was contracted to:
•Work with A&E firms to develop community relations
capability in these firms
•Develop community relations plans for regional office
approval
•Monitor contractor performance in the field
FACILITATING INSTITUTIONAL LEARNING
The entire effort at community relations policy de-
velopment and program implementation was an effort to
facilitate rapid institutional learning. The political dif-
ficulties that EPA found itself in at a large number of
hazardous waste sites around the country (e.g., Love
Canal, New York; Stringfellow, California; Memphis,
Tennessee; Valley of the Drums, Kentucky; and Jackson
Township, New Jersey) convinced Agency management
that new standard operating procedures had to be de-
veloped and instituted to forestall some of this opposi-
tion. It would be impossible to carry forth the mandate
of "cost-effective" remedial actions if citizen protest
caused delay and redesign at scores of waste sites.
The first stage in institutional learning is to develop
new institutional routines or standard operating pro-
cedures. The ambitious research program conducted to
develop community relations policy included the case
studies at 21 hazardous waste sites, interviews with per-
sonnel in every EPA Region, and considerable brain-
storming and analysis. Once the regional procedures were
developed and communicated to the Regions in the form
of the February 25 Guidance and the Community Rela-
tions Handbook, it was then necessary to provide a set of
incentives to secure program implementation.
The incentives provided fall under three categories:
(1) leadership, (2) training expertise and (3) staff-contract
resources.
Leadership
In a hierarchy such as EPA, consistent messages from
the top, and from headquarters to the Regions, are
necessary to facilitate institutional learning and organiza-
tional change. At the first headquarters meeting of the
Superfund Regional Coordinators and at most subse-
-------
410 LIABILITY, LEGAL & PUBLIC ISSUES
quent meetings, headquarters staff and the Superfund
National Program Manager stressed the importance of
Community Relations in Superfund. This message was
reinforced by a number of memos to the Regions.
Regional Superfund staff members were ready for and
in most cases welcomed this message. The field personnel
had first-hand knowledge of the nature of Superfund's
community relations problem, and were eager to take
steps to avoid these problems. Many Regional staff mem-
bers had already dealt with scared and angry citizens
and had been shouted at and harassed at public meet-
ings. The Superfund coordinators were fully aware of the
fact that Superfund and hazardous waste issues were un-
like anything else the Agency had ever faced. The vola-
tility of the issue was particularly obvious. This meant
that the Regions were receptive to this new message and
such receptivity made the task of institutional learning
considerably easier.
Training/Expertise
A training or "briefing" program was developed in
spring of 1980 and fielded in the fall of that year. The
purpose of the program was to brief Regional staff on
community relations issues and practices. In addition, a
HQ Public Affairs Assist Team was proposed to assist
the Regions in implementing Community Relations at
critical sites and will hopefully be established some day
as part of the Emergency Response Team in Edison,
New Jersey.
Finally, ICF Incorporated briefed the A&E and TAT
contractors on the lessons learned in the case studies. This
training program provided a means of disseminating new
organizational routines. It is, of course, too early to judge
the success of those efforts.
Contract Staff Resources
Even if the Regions had the disposition and the ex-
pertise to implement the Superfund community relations
without resources, all the good intentions in the world
will not result in program implementation. Accordingly,
headquarters attempted to provide contractor commun-
ity relations planning assistance to the field to pay for
and plan implementation. It was hoped that provision of
these resources would reinforce the importance of com-
munity relations and stimulate the Regions to allocate
some of their own resources to this function.
CONCLUSION
Change is a viable bureaucratic alternative. The in-
clusion of an effective Community Relations component
within the response capability of the Superfund program
demonstrates this viability. The entire program has been
affected by this inclusion, such that the knowledge gained
during program implementation can be applied to the
understanding of community relations effectiveness and
bureaucratic operations, as well. Had the community re-
lations function not been integrated into the operating
divisions of the program, this institutional learning
could not have occurred.
-------
THE COMMUNITY HAZARDOUS WASTE
COORDINATOR PROGRAM
BETSY GOGGIN
ANN RAPPAPORT
Department of Environmental Quality Engineering
Boston, Massachusetts
INTRODUCTION
All the resources required to solve the hazardous waste
problem cannot be found in any one level of government.
No one, not state government, not the federal govern-
ment, not industry, not any other group in society can
solve the hazardous waste problem alone. This premise is
the basis for the community hazardous waste coordinator
program.
The Massachusetts Department of Environmental Qual-
ity Engineering (DEQE) cannot do the whole job. The
Department's staff realized that in working toward the
resolution of the hazardous waste problem it must rely on
the interest and expertise of the people who are potentially
affected, that is, all the citizens of Massachusetts. It was
the recognition of this that led the Department to asking
the chief elected officials in Massachusetts' 351 cities and
towns to appoint community hazardous waste coordin-
ators. The problem of past improper waste handling at sev-
eral locations led to this request.
When the Division of Hazardous Waste was established
within the Department in January 1980, it faced two
immediate uncontrolled sites problems, a huge list of sites
potentially containing hazardous waste, limited staff funds
and the legacy of poor communications between state
government and communities during several hazardous
waste incidents in the past. Having a hazardous waste
coordinator in each community was seen as a step to-
ward the resolution of both of these.
The Division asked the community coordinators to as-
sist the Department in gathering the necessary historical
information on the potential hazardous waste sites and to
coordinate the activities of municipal officials and citi-
zens in this effort, because community officials and citi-
zens know their town better than anyone else and impor-
tant site-specific information is readily available to them.
The Division also asked the coordinators to serve as the
community contact persons for the Department in any
hazardous waste incident and to be responsible for en-
suring that the appropriate municipal agencies and citi-
zens are kept involved and informed.
Beyond the need for assistance with these particular un-
controlled sites-problems, the Department was convinced
that the coordinators also could provide a service to the
Massachusetts hazardous waste management program.
The coordinators could become advocates in the commun-
ities for responsible waste handling.
When the Department asked for the coordinator ap-
pointments, their role was described as coordinating local
groups and agencies in responding to hazardous waste
problems and serving as a lead community representa-
tive in all hazardous waste management activities. A pre-
vious knowledge of hazardous waste was not required for
the position but a strong interest and the ability to man-
age people and projects were strongly recommended. This
approach led to the large variety in the professional back-
grounds of the coordinators.
COORDINATOR RESPONSIBILITIES
The specific responsibilities of a coordinator vary de-
pending on the needs of the community and the interest
and expertise of the particular coordinator. The materials
sent to the chief elected officials and to the coordinators
themselves outlined the range of activities involved in pre-
venting, detecting and responding to hazardous waste
problems:
(1) Serving as primary community contact for the De-
partment and developing community contacts so that
they can relay important hazardous waste problems or
questions to the DEQE regional hazardous waste co-
ordinators.
(2) Learning which citizens/agencies/organizations within
the community are responsible for or involved in haz-
ardous waste activities:
•Fire Department issues licenses for storage of flam-
mable materials above and below ground and also has
primary responsibility for responding to hazardous
materials/waste spills
•Police Department has the ability to patrol the com-
munity around the clock and has the authority to ar-
rest individuals engaged in criminal activities
•Board of Health has broad powers to protect public
health, as well as specific directives including the ap-
proval of sites for hazardous waste treatment/dis-
posal facilities and receiving information from DEQE
regarding hazardous waste handled within commun-
ity boundaries
•Department of Public Works or Town Engineer
knows the locations of sewers, drain outfalls, etc.
411
-------
412 LIABILITY, LEGAL & PUBLIC ISSUES
•Conservation Commission has knowledge of wetlands
and other natural resources in the community
•Industrial Development Commission, Planning
Agency, and/or Chamber of Commerce know about
the types of industry currently in the community and
about which industries may be planning to come in-
to town—including hazardous materials users and
disposers. These agencies can also be very helpful in
encouraging the participation and cooperation of
local industries
• Water Department Superintendent knows the sources
of groundwater in the community, as well as sources
of drinking water, recharge zones, etc.
•Many citizens are knowledgeable in fields relating to
hazardous waste
(3) Organizing local resources:
•Contacting and developing local resources
•Forming a hazardous waste committee or task force
to assess the hazardous waste/materials situation in
the community and to choose activities to encourage
good hazardous waste management in the commun-
ity
•Working through the established authorities to en-
sure proper hazardous waste management in the com-
munity
(4) Educating citizens about:
•The threat to public health and environment caused
by hazardous wastes which are mismanaged
•How hazardous wastes can be sensibly handled by
using the available technologies for properly treat-
ing, recycling and disposing of hazardous wastes
(5) Preventing and detecting hazardous waste problems:
a. Assessing the hazardous waste/materials situation
in the community by identifying:
•All of the hazardous materials in the community
by categories, classes, volumes, locations and
type/age of containment. In addition to industrial
users, this survey would include facilities such as
hospitals, dry-cleaners, research laboratories and
gasoline stations
• What goes into the community landfill. Although
it is illegal, significant quantities of hazardous
waste, particularly that generated by small firms,
are deposited in landfills (with or without the
operator's approval)
•Old abandoned dump sites by checking old maps
and photographs and by talking to older commun-
ity residents
•Transportation routes for hazardous materials in-
cluding truck routes and railroad corridors. To
the extent possible, the frequency and timing of
hazardous materials shipments, as well as the
types of materials shipped, should be identified
• Where underground gasoline storage tanks are lo-
cated and beginning as a program to require own-
ers to test them for leakage
b. Mapping out local water supplies. Becoming famil-
iar with the network of aquifers, surface waters,
recharge zones, and wetlands in the community,
focusing on those that feed drinking water sources.
Determining where water supplies are with respect
to potential sources of contamination. Finding
maps which already exist.
c. Assisting the Department and the U.S. EPA in in-
vestigating suspected hazardous waste dump sites
by gathering the necessary historical information:
•A complete site-use history
*site uses over the years: industrial; residential;
waste disposal
"•owners of the site (past and present) property
maps, deed descriptions and restrictions, etc.
*chemicals manufactured or possibly dumped at
the site
•Types of waste suspected on site
*nature of industrial process that generated the
waste
•Evidence of pollution resulting from the site
•Location of nearest surface water, including
changes in surface water elevations (presence of
dams/diversions, culverting of streams) and
changes in direction of flow by natural or man-
made causes
•Direction of groundwater flow in relation to the
site now and in the past
•Location of nearest dwelling
•Location of nearest drinking water supply
•Type and permeability of the soil
d. If an uncontrolled hazardous waste site is found in
the community, participating in DEQE's and the
community's decision-making process to determine
what kind of action should be taken. The range of
choices at some sites includes:
•Removing the material for disposal at a secure
chemical landfill or treatment at an environmen-
tally sound facility (this is feasible only in a very
limited number of instances)
•Designing and constructing a secure containment
for the contamination at the site
•Bringing in an alternative source of water supply
or treating a water supply to remove contamina-
tion
The Department believes that making these decisions
must be a public process. Unless the potentially affected
parties and the community understand the options and
help decide on a particular solution, the particular situa-
tion will be extremely difficult to resolve.
e. Participating in the negotiations between the devel-
oper and the community, if a location for an en-
vironmentally sound hazardous waste treatment,
storage, or disposal facility is sound in the com-
munity.
(6) Ensuring that the community is prepared for a haz-
ardous waste/materials emergency:
•Learning about the resources available to respond to
hazardous waste/materials emergencies
•Working with the agencies responsible—fire dept.,
civil defense dept.—to ensure that the appropriate
people are prepared to respond effectively and ef-
ficiently
-------
LIABILITY, LEGAL & PUBLIC ISSUES 413
TRAINING
At this time, 83 percent or 290 of the 351 cities and
towns in Massachusetts have coordinators. A volunteer
steering committee of the coordinators has been formed;
the Division and the steering committee are currently plan-
ning a training program which will assist the coordina-
tors in performing the above activities.
The training program will provide the coordinators with
information on: enforcement and legal authorities, emer-
gency response procedures, how to detect and prevent
future problems and the siting of treatment, storage, and
disposal facilities. The training sessions will provide a
forum for the exchange of this information between pro-
fessionals involved or interested in different aspects of
the hazardous waste problem. The coordinators will be
able to use the information in their communities to en-
courage responsible hazardous waste management decis-
ions and avoid repeating past mistakes. Mistakes to be
avoided include: building a subdivision down-grad-
ient from the town dump, confusion during a hazardous
materials spill and enforcement actions which were not
well formulated because municipal officials, including
police officers, did not understand their jurisdictions.
PERFORMANCE
A number of communities already have a good under-
standing of sound hazardous waste management activ-
ities. For example, not long ago a developer applied to
a Massachusetts city to build a shopping center. The center
was to be located above a likely source of future water
supplies. The town agreed that the center could be built
but stipulated that a system to collect the rain water run-
off from the parking lot must be constructed and that no
small hazardous waste generators, including hair dressers
or dry cleaners, could be located in the center.
One of the most important resources for the coordina-
tors' training program are the coordinators themselves.
They are a diverse group of people having many areas of
expertise: health officers, conservation commission mem-
bers, firemen, policemen, lawyers, hydrogeologists and
engineers. Many of them are already developing their own
emergency contingency plans or are mapping water sup-
plies in relation to potential contamination sources.
One coordinator has given two lectures in the past sev-
eral months for other coordinators on how to map ground
water flow and potential sources of chemical contamina-
tion. He has also assisted one of Massachusetts' Water-
shed Associations in establishing a regional program for
community coordinators.
Another coordinator, an environmental engineer, has
done extensive emergency response planning for a large
chemical company in Massachusetts. He has participated
oh the coordinators steering committee and has provided
valuable emergency response planning information to the
rest of the group.
A third coordinator is a self-educated hazardous waste
expert. Since he was appointed- last fall, he has read as
much as possible and attended many seminars and lec-
tures. He has become an important resource for other co-
ordinators and a dedicated advocate for the development
of the training program.
During the past several months, many of the coordi-
nators have given the Department their ideas on how they
think the coordinator training program should be devel-
oped:
•legal authority within the municipality
•the general concepts—still a little confused between
"wastes" and materials
•specific suggestions as to organizing the municipal of-
ficials and departments in order to get better coopera-
tion
•how to inspire citizen responsibility for hazardous waste
management
•how neighboring towns can share information
•how to set up industrial-community seminars
•more about hydrogeology and its relationship to chem-
ical contamination
•possibly a listing of types of generators and the kinds of
waste they are likely to produce
•how to anticipate problems and deal with them when
they occur
•how to effectively handle emergency incidents—do's and
don'ts.
DEPARTMENT'S ROLE
The community hazardous waste coordinator program
has provided a number of challenges to the Department.
During the past eighteen months when the coordinator
program was conceived and organized, the Department
was undergoing a number of other major changes.
Changes involving organization, the delegation of respon-
sibilities and finding the resources to run the mandated
hazardous waste program. This was a difficult time to
attempt to develop a new program. Finding the resources
required for the training program has been one of the most
difficult tasks.
In the time of tightening budgets, it is often difficult to
fund the existing work load, let alone a new program. The
Department is using in-kind assistance and resources out-
side the agency to develop the program including local
colleges and universities, business and industry, environ-
mental education programs sponsored by civic, profession-
al, and educational organizations and other government
groups such as regional planning agencies and federal
agencies. Specifically, Tufts University, Boston Univer-
sity, the Monsanto Company, Massachusetts Cooperative
Extension Service, Nashua River Watershed Association,
Lincoln-Filene Center for Citizen Participation, Metropol-
itan Area Planning Council are among the organizations
that have been actively involved in the program. Essen-
tial, however, is the involvement of full time staff mem-
bers during the program's formative stages. A number of
staff people also are providing part time assistance in spe-
cific technical areas.
Funding a program using outside resources and in-kind
services creates the difficult problem of coordinating all
these activities and maintaining the spirit of cooperation
-------
414 LIABILITY, LEGAL & PUBLIC ISSUES
and enthusiasm. Whether this funding approach is viable
remains to be seen.
The unconventional nature of the community coordina-
tor program caused a certain amount of confusion within
the Department. It is certainly an atypical endeavor for a
regulatory agency. Although the program was a logical
step for the Department's key decision makers due to their
long-standing commitment to involving the public in agen-
cy decisions, it was not readily accepted by all the staff
members. Nearly a year of discussion and delays ensued
before the concept gained acceptance. Some staff mem-
bers held the traditional view that government officials
do their jobs well and that citizens need only be involved
minimally, if at all.
If the Department had anticipated the reluctance of the
staff to accept the coordinator program, some of the re-
sulting delays might have been avoided. The presenta-
tion of the program could have been altered to alleviate
some staff concerns; in particular, it was found that a
simple, direct statement of the objectives was the most
effective vehicle. The reasons were:
(1) To provide a solution to the serious communications
problems between DEQE and municipal governments
which occurred in several hazardous waste incidents
such as those in Acton, Bedford, Lunenburg, Nor-
wood and Woburn.
(2) To let the public and local officials know that DEQE
has extremely limited resources and cannot solve the
hazardous waste problem alone.
(3) To provide a person in each community who is both
concerned and knowledgeable about the problem of
hazardous waste to help avoid repeating past mistakes
such as:
•building a subdivision downgradient from the town
dump
•siting an industrial park on an aquifer or watershed
•panic/confusion during a hazardous materials spill
•enforcement actions which were not well formulated
because municipal officials, including police officers,
did not know the proper procedures
(4) To provide an advocate in each community for edu-
cating municipal officials, industry and the public
about:
•the variety of sound waste management options avail-
able
•the problem of continued emotional response to
hazardous wastes
•protecting groundwater resources
•promoting responsible waste management practices
within the community
•procedures for response to a hazardous waste/ma-
terials emergency
•enforcement activities at the Federal, State and
municipal levels
•detecting potential uncontrolled hazardous waste
sites, especially abandoned municipal dump sites
CONCLUSIONS
After reviewing the reasons why the Department asked
for the coordinators to be appointed, the staff agreed with
the intent of the program and became much more willing
to assist in its development. Perhaps an even more per-
suasive factor for the staff was the community coordina-
tors' demonstration of their interest and expertise.
Through their activities, it became apparent that local
citizens do know their communities best and that impor-
tant community specific information is readily available to
them.
In the long term, the benefits of having coordinators
should far outweigh the difficulties involved in getting the
program started. Today's limited budgets and the gen-
eral movement toward reducing the size of government
make the coordinator's program even more relevant than
when it was conceived.
The community coordinators are presently looking into
the possibility of forming a non-profit association. Such
an association would give the coordinators an independent
base outside of the Department and perhaps generate some
resources. Given the tenuous political support for environ-
mental issues in government, an association could provide
the coordinators with continuity they might otherwise
lack.
The coordinator program is a reflection of the axiom
that "hazardous waste is everybody's problem." As the
coordinators become more and more involved in prevent-
ing and resolving specific hazardous waste problems, they
will be a group of people actively taking responsibility
for problems created by mismanaged hazardous wastes.
The community hazardous waste coordinator program
is still in the experimental stage. The Department's staff
is learning an immense amount with each step that the
program takes and hopes that the coordinators are learn-
ing enough from the program to keep them actively in-
volved. The Department's staff encourages other agencies
to become involved in similar programs, are eager to learn
from the experiences of others, and are optimistic about
the benefits of this type of public participation.
-------
CITIZEN/GOVERNMENT INTERACTION AT
TOXIC WASTE SITES:
LESSONS FROM LOVE CANAL
L. GARDNER SHAW
and
LESTER W. MILBRATH
Environmental Studies Center
State University of New York at Buffalo
Buffalo, New York
INTRODUCTION
It is ironic, at least to environmentalists, that the effort
of William T. Love to develop hydroelectric power—a
form of energy conversion that is considered environ-
mentally benign—should eventually result in one of the
most traumatic environmental tragedies of the 20th cen-
tury. Love's grand scheme to develop a new industrial
city utilizing hydroelectric power projected the construc-
tion of a canal to carry water from the upper Niagara
River to the generation point. After nearly 1000 cubic
yards of the canal had been excavated in 1895, the pro-
ject was abandoned.
In the 1930's the abandoned trench had become a
dumpsite; beginning in the 1940's it was used by the
Hooker Chemical Company, the City of Niagara Falls,
and according to recent assertions, the U.S. Army. The
main user, however, was Hooker Chemical which for over
a decade dumped hundreds of tons of toxic chemicals,
most of them enclosed in 55-gallon drums, into the un-
finished canal. When the canal was full, Hooker claims
that it was sealed with a clay cap, an accepted procedure
for such dumps at that time.
During the late 1940's and early 1950's the City of
Niagara Falls expanded into the Love Canal area. Over
700 homes and a school were built in the vicinity of the
canal during the late 1950's.
Beginning in the late 1960's_and throughout the '70's,
the metal drums that had contained the chemicals de-
veloped holes due to corrosion and began to leak. Toxic
leachate rose to the surface, especially after heavy rains.
During the 1970's it was discovered that leachate had mi-
grated horizontally reaching the basements of nearby
homes and, even later, evidence was uncovered that the
leachate had traveled several blocks.
Chemicals and drums coming to the surface and the
appearance of numerous health symptoms, led a few local
residents to press government agencies to investigate the
problem and take remedial action. Early citizen requests
were ineffectual; local governments were not equipped
for the problems that citizens were asking them to deal
with.
In late 1976, the N.Y. Department of Environmental
Conservation (DEC), the City of Niagara Falls and U.S.
Environmental Protection Agency (EPA) began to in-
vestigate water quality problems in the area. Meanwhile,
the State Department of Health began the examination of
possible health effects associated with exposure to toxic
wastes. In April 1978, Commissioner of Health, Robert
D. Whalen, determined that a health hazard existed in
the Love Canal area and ordered the Niagara County
Health Department to cover exposed chemicals and install
a fence around the area. He also ordered health studies of
residents living adjacent to the canal.
In May 1978, EPA announced that air samples it had
taken from homes adjacent to the canal contained dan-
gerously high levels of toxic vapors. The State Health
Department studies began the following montluand_found_
high levels of toluene, chloroform, benzene and ehloro-
toluene.
On August 2, 1978, Commissioner Whalen issued an
order declaring a health emergency, recommending a
delay in opening the 99th Street School, and recommend-
ing the evacuation of pregnant women and children under
two years of age from the immediate canal vicinity.
Governor Carey visited the Love Canal area on August
7, and, at a public meeting, announced that the state
would purchase homes affected by the Love Canal chemi-
cals. Two days later it was confirmed that 239 homes in
two "rings" around the canal would be purchased. Resi-
dents within these first two rings were offered temporary
relocation at state expense, pending purchase of their
homes. Most of the residents affected by this decision
were evacuated to temporary housing.
During August, the State DEC had begun working with
the City of Niagara Falls and Conestoga-Rovers, the con-
sulting firm it had retained to review a site cleanup plan.
The safety and health of workers and nearby residents be-
came an integral and controversial aspect of the interim
construction plan. The scope of the safety plan included:
•Security and communication provisions
•Personal hygiene and workers' safety
•On-site monitoring and sampling
•Emergency evacuation provisions (this provision later
became quite controversial).
Also, during August, the State Department of Health
took blood samples from over 2,800 Love Canal area resi-
dents. With such a large number, there were long waiting
lines which led to complaints about the planning and or-
ganization of the health study. Following the sampling,
the residents eagerly, and later impatiently, awaited the
results. By August 1978, Love Canal had become a highly
415
-------
416 LIABILITY, LEGAL & PUBLIC ISSUES
publicized and politicized case marked by an extraordi-
narily high degree of citizen/government interaction.
ISSUES
By the end of August 1978, citizen activity in the Love
Canal area was directed toward a number of issues:
•How many homes should be purchased by the state and
which ones should they be?
•Which residents should be temporarily relocated, at
whose expense and on the basis of what criteria?
•What adjustments would be made in the property tax
liability of citizens whose homes had become less valu-
able and which, if any, homeowners would receive the
adjustment?
•What provision would be made for the safety of nearby
residents during remedial construction; who would ap-
prove of those provisions and who would supervise and
enforce them?
•What negative health effects were attributable to ex-
posure to toxic wastes at Love Canal; how would the
types and levels of toxic substances be determined; how
would health effects be determined and how would such
information be reported, and to whom?
ACTORS
Virtually all levels and branches of government that
might have had some possible connection with the Love
Canal case entered the picture at some point. The com-
plexity and frequency of governmental involvement fed a
sense of apprehension and urgency to the media, and
through them to the public. On the other side, the media
attention and the public clamor brought about greater
governmental involvement.
Government Agencies
The first agencies involved were the Regional Office of
the New York State Department of Environmental Con-
servation, the U.S. Environmental Protection Agency,
and the Engineering and Planning Offices of the City of
Niagara Falls. These agencies had become involved prior
to 1978 in an investigation of the causes of a water pollu-
tion problem noted in the Love Canal region. When the
source was traced to the Love Canal dumpsite, connec-
tions began to be made with the complaints of local resi-
dents of the surfacing of buried chemicals and the onset
of unusual health symptoms. At this point the State De-
partment of Health entered the stage.
Local governmental officials were, not surprisingly,
reluctant to act in a way that might attract unfavorable
attention to the area or to create animosity between the
people in the city and the chemical industry. The economy
of Niagara Falls relies on two industries: chemicals and
tourism—both of which could be injured by what was
beginning to transpire at Love Canal. Indeed, during all of
the early activity that followed the events of August
1978, the City of Niagara Falls was involved in negotia-
tions with Hooker Chemical Company over the latter's
prospective construction of a major office building in the
city.
Two days after the Health Commissioner's August 2
order was issued, New York Governor Carey created an
Interagency Task Force to deal with the problems as-
sociated with the Love Canal waste site. The Task Force
functioned at both the state and local level. The state level
Task Force, headquartered in Albany, consisted of the
heads of the departments of Health (DOH), Environ-
mental Conservation (DEC), Transportation (DOT),
Social Services, Insurance, Banking, Housing and Com-
munity Renewal, Equalization and Assessment, and
Disaster Preparedness (ODP). An on-scene "working
task force" had offices located at first in the 99th Street
School and later in homes, adjacent to the canal, that
had been purchased by the state. This local Task Force
consisted primarily of representatives of the Departments
of Transportation, Health, Environmental Conservation,
and Social Services. The on-site coordinator was Michael
Cuddy of the Department of Transportation. Throughout
this period most of the governmental activity at the Love
Canal was performed by agencies that were members of
the Task Force; hence, most citizen interaction was with
the Task Force or with the governor and/or his repre-
sentatives.
The State Department of Transportation was selected
as lead agency in the Task Force because of its experi-
ence with supervising construction activity and its back-
ground in dealing with property purchases and relocation
of families. There is a natural tendency for agencies com-
ing into a new situation to define a problem in terms
that are within their range of experience and compe-
tence; this experience also will shape the way that they
envision and pursue solutions to the problem they are con-
fronting. In this case, hardly any agency had experience
closely relevant to the problems that had to be tackled at
the Love Canal site. Experience with highway construc-
tion does not help very much in dealing with an appre-
hensive public fearful of explosions and toxic gas releases.
A background in negotiation with homeowners on the
purchase of property via eminent domain to build a high-
way does not necessarily equip an agency with a working
model to deal with the purchase of homes from owners
who are all too anxious to leave an area for fear of their
health.
The primary federal agencies to become involved were
EPA, FDAA (now the Federal Emergency Management
Agency, FEMA), and the Center for Disease Control;
other agencies also became involved (e.g., HUD and
HEW—now HHS). FDAA became an actor early when in
August 1978 Governor Carey began seeking federal
financial assistance; its successor, FEMA, ultimately was
a source of federal aid to both the city and the state. The
Center for Disease Control was involved in planning
health studies of residents and was responsible for setting
up and managing long range health studies.
To a great extent, the politicization of the Love Canal
case was as much a product of citizens' perceptions and
misperceptions, apprehensions and misapprehensions, as it
was of ineffective or inappropriate activity on the part of
governmental agencies. Numerous agencies were involved,
-------
LIABILITY, LEGAL & PUBLIC ISSUES 417
each perceiving the problem and their own role differ-
ently. Each tended to function in the manner for which
its experience had prepared it. Because so much learn-
ing had to take place by all parties, it is little wonder that
at critical junctures public passions rose because citizen's
perceptions of agency performance did not mesh well with
their expectations of what agencies should be doing.
Public appreciation of agency performance was highest
when citizens perceived agency personnel as being open,
straightforward and evenhanded. Agencies received low
marks from citizens, and citizen tension and anger grew
when agencies appeared to be concealing information,
soft-pedaling bad news or treating citizens in a con-
descending or inequitable fashion.
Citizens seemed particularly sensitive to continuity, or
lack of continuity, of agency personnel, often expressing
frustration with the removal of officials that they had
come to know and learned to work with. Many viewed it
as particularly salutory that Michael Cuddy was the co-
ordinator of the local Task Force through nearly all of
the period that the state government was involved.
Political Involvement
Governmental activity at that site was stimulated and
facilitated by several concerned legislators who attempted,
through constituent service and legislation to deal with
the various problems at the site. These included U.S. Con-
gressman John LaFalce, who represents that district and
State Legislators Senator Daly and Assemblymen Murphy
and Pillitera.
Citizen Activity
Much of the citizen activity in response to events at
the Love Canal coalesced around the Love Canal Home-
owners Association (LCHA). During the summer of 1978,
Lois Gibbs became more prominent in homeowners ac-
tivities, and she was ultimately elected president of the
Association. Her leadership was strengthened by being
invited to attend a meeting at the White House on Aug-
ust 9, that also included state officials, members of sev-
eral federal agencies and a representative of the President.
Mrs. Gibbs believes that state and federal officials made
a deliberate move at this time to recognize her leadership
because they felt that she was someone they could work
with.
In the first stages, LCHA members tended more to re-
act to governmental action rather than develop a strategy
to influence governmental officials. Shouting and react-
ing emotionally toward officials during public meetings
were modes of activity that proved to be effective in
gaining attention and sympathy in the short run, but
proved to be less productive techniques in the long run.
As time went on, the LCHA leadership learned to fit
strategies to situations much more effectively.
In addition to the Homeowners Association, several
groups emerged with smaller memberships and somewhat
different concerns. These included the Concerned Area
Residents, the LaSalle Development Renters Association,
the "93rd Street Group", and the Ecumenical Task Force.
The first two were composed primarily of residents of
government-subsidized housing adjacent to the canal.
Many of the former were senior citizens; many of the
latter were from racial minority groups, primarily black.
The Ecumenical Task Force is not really a citizens group;
rather it is a local arm of Church World Services, an in-
terdenominational cooperative relief and development
agency.
COMMUNICATIONS
Sources of Information
The public agencies used a full range of communica-
tion techniques. In many cases, the device was a press re-
lease or a press conference; several of the statements
from the governor's office and from the state Task Force
were of this variety. Public meetings also were utilized,
but with uneven success. Unfortunately, these meetings
tended to generate more heat than light. Many residents
felt that the high emotions that characterized these meet-
ings resulted because such meetings followed the style of
press releases which did not afford an opportunity for
questions and clarification.
In other instances, communication was more limited in
scope and more private. For example, many of the con-
sultations regarding the construction safety plan and the
tax abatement legislation appear to have been with only a
few members of the LCHA. Also, when the results of a
chromosome study were available in May 1980, a serious
effort was made (before the study was made public) to
communicate the results to the people who were tested.
The Task Force working at the Love Canal site util-
ized a weekly newsletter, and generally speaking, the meet-
ings of the Task Force included representatives of the
homeowners and renters associations and were often at-
tended by other interested citizens as well. The meetings
provided a regular source of information about activity
taking place on the local level. In addition to the news-
letter and word-of-mouth communications, the actions of
the Task Force also were reported regularly by the local
media.
Types of Information
The information that needed to be communicated to
the public could be classified as either: a) progress re-
ports, b) policy statements or c) technical data or reports.
Government agencies were most successful in communi-
cating on their progress. While communicating the ex-
istence of new policy positions did not prove difficult,
conveying the substance of policy did. Questions abound-
ed, for example, regarding the intent of the state in the
relocation of persons with health problems. It also was
unclear to many what role the FDAA intended to play in
early August 1978.
Most troublesome, however, was the communication
of information about health and environmental studies.
Obtaining results from them seemed agonizingly slow. On
the other hand, early release of a study from the State
Department of Health elicited citizen complaints about
-------
418 LIABILITY, LEGAL & PUBLIC ISSUES
lack of interpretive material or standards. One of the most
frustrating incidents for citizens was an academic and
scientific debate that ensued after the release of the
chromosome study in May 1980.
ROLES OF CITIZEN GROUPS
The events at Love Canal suggest several important
questions regarding the roles of citizen groups in such
settings: (1) What role should citizens play in the policy
process? (2) How should citizen involvement be timed?
(3) How can diverse citizen viewpoints be represented most
equitably and effectively? (4) How can damaging con-
flicts and controversies among groups be avoided or miti-
gated?
Most citizens are neither lexicologists, nor are they
public officials with experience in making or implement-
ing public policies; yet, citizens may become quite so-
phisticated about both technical and political matters re-
lated to toxic waste disposal. However, citizens are
resident-experts on their own communities and are best
equipped to read their own concerns and preferences.
Seeking, or accepting, citizens' suggestions on various as-
pects of a safety and evacuation plan could be a per-
fectly reasonable action. On the other hand, citizens
should not be expected to interpret complex scientific in-
formation couched in sophisticated terminology.
When citizens are given some formal role in public
policy making, it raises the question of the extent to
which different citizens' views should be represented in
that forum. A concern expressed frequently by members
of other citizen groups was that the Love Canal Home-
owners Association was afforded more attention and ex-
ercised greater influence on decisions, than citizens from
these other groups felt was appropriate. Yet, most of the
influence exerted by the LCHA was through informal
channels; the organization did not sit as a voting mem-
ber of any policy making bodies. Because citizen organi-
zations were not given a formal role in decision making
at Love Canal, it avoided some difficult problems that
formal representation might engender. The reaction of
citizen organizations to the perceived influence of LCHA
was more a reaction to what these groups perceived as
substantive inequities rather than procedural inequities.
There are two important exceptions to the previous
comments on formal representation. First, the LCHA was
afforded permanent representation at meetings of the
local Task Force very early in the latter's existence. Ob-
jections by renters groups that their interests were not ade-
quately being represented by the homeowners led quickly
to their inclusion as well.
Second, a Love Canal Revitalization Agency was con-
stituted with the intention of having three representa-
tives of the general community as well as representatives
of the Niagara Falls and nearby Wheatfield Municipal
governments. Disagreement arose over whether specific
citizens groups should be represented on the agency or, as
the local government officials felt, the three seats should
be occupied by representatives of the "community-at-
large." This position was partly based on the view that
the revitalization agency would be administering the pur-
chase of the last group of homes, as well as the reloca-
tion of renters, placing representatives of those
groups (homeowners and renters) in a conflict of inter-
est. The view of the municipal officials finally prevailed on
this issue, although one of the citizen members of the re-
vitalization agency, Bill Wagoner, remained associated
with several of the citizen groups.
EFFECTIVENESS OF CITIZEN/GOVERNMENT
INTERACTION
The effects of citizen/government interaction on policy
can be viewed from two perspectives: (1) the effect of that
interaction on policy makers and policy processes and
(2) the effect on policy decisions. With regard to the
former, citizen activity seemed to drive some agencies to
defensive, self-protective postures. This may have been
partly a function of the experience of the particular of-
ficials in dealing with the public. Some of the medical re-
searchers and other scientists involved in the events at Love
Canal were very reticent about talking with the public.
On the other hand, officials in the State Department of
Transportation (especially those active at the local level),
and local EPA officials were much more at ease.
The Love Canal case alerted the public to the dangers of
improper disposal of toxic waste. Citizen participation in
those events was instrumental in bringing out new di-
mensions of the problem. Through their persistence, citi-
zens at Love Canal interjected themselves into the formal
or informal decision making structure of many govern-
mental agencies. Citizens' organizations came to be re-
garded as parties that had to be consulted before many
types of decisions could be taken.
In addition to the general effect of shaping perceptions
of the dangers of improper disposal of toxic waste and of
general policies that must be pursued to avoid such prob-
lems, citizen participation at Love Canal influenced a
number of specific policy decisions. The most prominent
was the decision by Governor Carey to have the state pur-
chase homes in the Love Canal area; later there was a com-
mitment by the federal government to provide financial
aid for the relocation of the balance of the affected resi-
dents. It is doubtful that these decisions would have been
made this way without the citizens' vigorous and per-
sistent demands for help. The decision to purchase homes
in the outer ring is the most important achievement of
Love Canal citizen efforts.
State legislation providing property tax relief for Love
Canal homeowners, and federal legislation providing a
capacity for responding to toxic waste emergencies (e.g.,
the so-called Superfund) followed hearings in which Love
Canal residents participated. The occasions for the hear-
ings were themselves a response to pressure exerted by
citizens. In addition to the policies that were initiated,
numerous policies were revised or modified as a result of
citizen/government interaction. The safety and evacuation
plans and temporary relocation policies are illustrations
of this.
-------
LIABILITY, LEGAL & PUBLIC ISSUES 419
IMPLICATIONS
Many of the implications of the Love Canal case for
governmental response at future hazardous waste sites re-
late to the style in which agency officials interact with
the public. The importance of openness and straightfor-
wardness has already been mentioned. There were, how-
ever, several other techniques that had a bearing on the
nature of citizen/government interaction and which,
therefore, could be expected to influence the effectiveness
of hazardous waste response efforts elsewhere.
Citizens are apt to entertain somewhat naive concep-
tions of the role and capability of government, particu-
larly in the early stages of a critical situation. As citizens
gain experience and sophistication they are more likely to
differentiate the roles of various levels of government and
of agencies within a given level or recognize the con-
straints within which public officials must operate. Be-
fore this sophistication can be acquired, a series of
dashed hoped can reduce the credibility of some agencies,
and government in general, to the point where effective
interaction with the public becomes almost impossible.
Such problems might be avoided if officials encourage
limited, but realistic, expectations of agency capabilities.
An early problem the Task Force seems to have ex-
perienced was the release of information by lower echelon
members of the Task Force that contradicted other in-
formation being released. Reassignment of some staff
members to other duties may have helped alleviate that
problem. However, the reassignments contributed to the
problem of lack of continuity of personnel, cited previ-
ously. Release of inconsistent or contradictory informa-
tion not only reduces governmental credibility, but can
lead to anxiety heightening uncertainty among citizens as
well. On the other hand, rigid enforcement of a "party
line," or imposition of a "gag rule," on staff members
may have the effect of hiding agency activity from legiti-
mate public scrutiny or creating an impression of se-
cretiveness and lack of candor on the part of the agency.
This need not be an impossible dilemma; the dual goals
of coordination and candor seem attainable.
Promptness and orderliness in providing information is
important to agency credibility, to say nothing of its
benefits for relieving uncertainty among citizens. Es-
pecially when vital issues such as adverse health effects
are involved, citizens demand—and deserve—to know
where they stand. Information may not be available as
soon as the public wants it; therefore, it is important that
citizens be helped to understand from the beginning how
long it will take for such information to become avail-
able. Even the orderly and accurate release of informa-
tion may lead citizens to feel that they face serious risks;
disorderly communication can only exacerbate that prob-
lem.
Much of the confusion and anxiety generated by the
conflicting statements from government agencies and re-
searchers in May 1980 could have been avoided had EPA
officials been able to release chromosome study results in
a more orderly fashion—as had been their original in-
tention. The necessity of going to the public with test
results in order to pre-empt a press leak, precluded thor-
ough evaluation of the results of the study and created an
aura of importance around the study that was inappropri-
ate given its nature and scope.
When citizens are already divided by conflicting goals,
the potential for serious intergroup conflict can be in-
creased by a communication practice that appears to show
favoritism toward one group. There may be any number of
legitimate reasons why an agency may wish to encourage
or support a particular group but an appearance of pre-
ferential treatment may raise fears that other legitimate
interests are going to be overlooked. A tradeoff is in-
volved in that dealing with numerous groups may be
more cumbersome—and frustrating—than dealing with
one or two; yet, the diversity of interests may be
capable of effective expression only through several dis-
crete groups.
Evenhanded treatment of groups may become even
more of an issue where, as in the Love Canal area, cleav-
ages of interest regarding government policy toward the
toxic waste problem overlap other social cleavages in the
community—in this case, race and economic status. The
renters tended to be black, older, and of lower income
and they did not see their interests as identical to those
of the homeowners.
Perceived preferential treatment may also lead to a view
that agencies are pursuing a "divide and conquer" strat-
egy. The extent to which this perception was created by
malicious intent as opposed to bureaucratic ineptitude is
unclear. Perhaps the most frequently cited example of a
"divide and conquer" strategy being employed by gov-
ernment in the Love Canal case related to the issue of
tax abatement. An official in the state Task Force is re-
ported to have told residents on the west side of 93rd
Street that they were, in effect, "sold out" by the Home-
owners Association in discussions on the boundary lines
for tax abatement.
Although circumstances frequently dictate the use of
public meetings, such assemblages have widely recognized
drawbacks, especially when potentially emotional or
highly technical subjects are on the agenda. People who
are under great anxiety and stress will naturally press
their case in such public meetings and the meetings may
turn into chaos. The emotional concerns of citizens will
feed upon each other and almost any response by public
officials is likely to be perceived as inadequate.
In addition, such meetings are generally poor forums
for the display of technical or scientific virtuosity on one
hand and for the unembellished reporting of raw data on
the other. Data must be accompanied by interpretation,
although this should not be taken to suggest that citizens
are dullards. Many of the citizens active at Love Canal
became sophisticated in their understanding of the
language and concepts of toxic chemical wastes.
In some cases, an effective alternative to the large pub-
lic meeting is private meetings of agency officials with
the leaders of various citizen groups. At such "public
consultations," it is possible to provide an indepth pre-
sentation of information or explanation of policy, and to
share concerns about how the material might be received
by the public, in advance of a large meeting.
-------
420 LIABILITY, LEGAL & PUBLIC ISSUES
Another useful method for improving communication
between citizens and government is for the citizens to
acquire their own experts. In the Love Canal case one ex-
pert was funded by the state government and another
volunteered her services to the LCHA. Citizen-oriented
expert consultations have greater credibility than "gov-
ernment experts" and can significantly improve under-
standing by citizens.
Because there was no "model" for citizen/government
information at toxic waste emergency sites, decisions
about informing and involving people in policy making
tended to occur on an ad hoc basis at Love Canal. As is
typical when clear policies of equity have not been worked
out, it was in no small degree a case of the squeaking
wheel getting the grease. Unfortunately (from the stand-
point of citizen/government relations) the perception by
citizens that they needed to be vocal and contentious in
order to gain attention led to resentment and frustration
in many cases, and helped to create a pattern of adversari-
al relationships. While adversarial relationships may be
unavoidable in such situations, more careful attention to
the nature and timing of citizen involvement might help
to avoid their more counter-productive aspects.
-------