Proceedings of the National Conference on Hazardous Wastes and Hazardous Materials, March 16-18, 1987, Washington, DC

-------
groups of data. The samples were analyzed for volatile organics.
The results of the analyses are presented in Table 2.

Table 2
Sample Data Set Summary of Compounds Delected ABN and VOA
Analyiei In PPB
Simple HP,
Volatile Organic Analysis

1,1,2.2-Tet rachloroathyleno

Dichloromethane

1,1,1-Trlchloroethana

l,l,-DLchloro«than«
240

10
HA2

HA
Di-N-Butylphthlate
Bis(2-Ethyleh«iyl)
Phthalat*
22,000
1. The concentration in the blank is greater than Vi the method detection limit and is greater than
V^ the concentration delected in the umple.
2. Analyses not conducted.

Obviously, there is little or no correlation between the ana-
lytical data and the Held screening data. This discrepancy may
be the result of numerous factors effecting the field screening
data. The remainder of the discussion in the paper will address
the shortcomings of field screening and possible explanation for
the discrepancies between instruments utilized for the screening
investigation and the analytical results.

DISCUSSION
Clearly, there are some problems concerning the acquisition
and interpretation of data from soil pore gas investigations.
Several questions that should be addressed when considering
whether a soil pore gas investigation is an appropriate tool to
use while assessing a particular site include:

• Which instruments should be used for field screening?
• What is the most appropriate technique to obtain a represen-
tative soil pore gas sample?
• What impacts do variations in weather, soil and other physi-
cal conditions have on the quality of the data?
• How can the data from a soil pore gas investigation be veri-
fied?

First, which instrument should be used for analyzing soil pore
gas samples? Ideally, the investigator should know what com-
pounds are likely to be contaminating a site or migrating from a
facility. If these compounds are clearly defined, the instrument(s)
should be selected based on its ability to analyze the target com-
pound^). If methane or hydrogen sulfide is likely to be pres-
ent in substantial concentrations, the device which is selected
must be able to either distinguish these compounds from the con-
taminants of interest or selectively eliminate the interference
caused by their presence.
Clearly, the most desireable instruments to be used in a soil
pore gas study are those which have the ability to quantitatively,
as well as qualitatively, define those contaminants which may be
present. Instruments, such as the Century 128 OVA, operated in
a gas chromatograph mode or the Photovac Model 10S10 port-
able gas chromatograph have the ability to identify the com-
pounds and quantities in which they are present. Obviously,
these instruments are not as common as the field screening in-
struments utilized for this study. However, the quality of the data
would be greatly enhanced through their usage.
What method is most appropriate to collect a representative
soil gas sample? The key word in this question is "appropriate."
There are several commercially available soil pore gas sampler.
which have been recently developed. This equipment is readily
obtainable and may represent the most appropriate samplers for
use in a particular study. However, site conditions vary tremen-
dously and no one piece of equipment or method may be suitable
in all instances.
The method used in this study probably resulted in represen-
tative soil pore gas samples. However, correlation of those re-
sults to soil or groundwater analytical results is difficult. An ex-
planation may be the existence of thin horizons in which con-
taminants may be present in higher concentrations relative to the
surrounding soils. These layers may significantly contribute to the
concentration of contaminants in soil pore gas samples, whereat,
a composite sample collected from a borehole may contain lit-
tle or none of this material. Therefore, correlation of the data it
difficult, if not impossible.
The most appropriate method is the one which works best. The
method must sample the soil pore gas at a designated depth or
depth range and have the ability to seal the sample intake from
the zones above and/or below. In most cases, the critical seal is
the one at the ground surface which prevents dilution of the soil
pore gas sample by the troposphere. Sampling devices may be
constructed from monitoring well materials and be permanently
or temporarily installed in the vadose zone.
What impacts do variations in weather, soil and other physi-
cal conditions have on the quality of data? Soil pore gas investi-
gations are conducted such that each sampling point is compared
relative to the adjacent sample points. Therefore, assuming they
are constant during the investigation, the impact of such physi-
cal parameters is insignificant. In addition, the effects of tempera-
ture are minimized by the continuity of the ground's tempera-
ture.
Frost and snow cover may cause the concentrations of contam-
inants in soil pore gas to increase. Winter conditions act as a
broad seal capping the site and allowing the soil pore gas to equi-
librate. When the frost or snow is punctured, the resulting con-
taminant concentrations may be increased above the time when
such cover is not present. However, anomalies representing po-
tentially contaminated groundwater and/or soil will still be pres-
ent.
Changes in physical conditions during investigations require
careful scrutiny. The date, time, soil and weather condition
should always be recorded when samples are collected. The in-
vestigation should be organized into blocks of sample locations,
especially when the investigation will require more than one day.
Finally, if investigation is expected to take more than one day,
sample points from the previous day may be sampled to deter-
mine fluctuations resulting from weather related changes.
How can soil pore gas data be verified? Verification of soil
pore gas data may be attained through a selective soil/ground-
water sampling program and/or a quantitative and qualitative
assessment of the contaminants present in a soil port gas sam-
ple. The physical, chemical and environmental factors affecting
the soil pore gas must be properly understood to obtain reliable
results. For example, if a portion of a site is underlain:by a layer
of richly organic peat, the concentrations of methane and/or
hydrogen sulfide may be elevated in the immediate vicinity indi-
cating the presence of a contaminant source. Analysis of a soil
pore gas sample may indicate high levels of methane. Boring in-
formation may define the aerial extent of the peat layer which
corresponds to the contaminant levels delineated in the soil pore
gas study. Therefore, knowledge of the stratigraphy and chem-
ical content of the pore gas resulted in the proper characteria-
202 MONITORING & SAMPLING

-------
tion of this portion of the site.
The result of a soil pore gas study should always be verified
with laboratory results. Studies designed to delineate soil and/or
groundwater contamination should be verified with soil and/or
groundwater samples, respectively. The samples should be se-
lected from the entire range of values, not only those locations
in which contamination is expected. The analytical results must
then be compared with field screening values to determine the
ability of the investigation to detect contamination on a site-spe-
cific basis.
CONCLUSION
Soil pore gas studies are a powerful technique to cost-effec-
tively evaluate a site or facility for the presence of groundwater
and/or soil contamination. The worth of an investigation is de-
pendent on the ability of obtain representative soil pore gas sam-
ples which may be qualitatively and quantitatively analyzed for
contaminants present.
The investigations must be performed with knowledge of local
physical, chemical and environmental factors which may effect
the acquisition or interpretation of the data. The method selected
to collect representative soil pore gas samples must be appropriate
for the site-specific conditions. The instruments used to analyze
soil pore gas samples must be sensitive to target parameters of
concern, while having the ability to distinguish contaminants for
naturally occurring compounds.
The data resulting from soil pore gas investigations must be
verifiable through a laboratory analytical program. The soil gas
data from the full range of values must be compared and corre-
lated to the analytical results.
Soil pore gas investigations are cost-effective and easy to per-
form. However, significant care must be used when interpreting
the results.

REFERENCES
1. Marrin, D.L. and Thompson, G.M., "Gaseous Behavior of TCE
Overlying a Contaminated Aquifer," Groundwater, 25, 1987, 21-27.
2. Spittler, T.M., Fitch, L and Clifford, S., "A New Method for De-
tection of Organic Vapors in the Zone," Proc. of the Characteriza-
tion and Monitoring of the Vadose Zone Conference, National Water
Well Association, Denver, CO, 1985.
3. Lappala, E.G. and Thompson, G.M., "Detection of Groundwater
Contamination by Shallow Soil Gas Sampling in the Vadose Zone,"
Proc. of the Characterization and Monitoring of the Vadose Zone
Conference, National Water Well Association, Denver, CO, 1983.
MONITORING & SAMPLING 203

-------
                               Screening For  Characterization of
                            PCB-Containing  Soils and Sediments
                                                   Bruce A. Fowler
                                                  Joseph T. Bennett
                                                   E.C.  Jordan  Co.
                                                   Portland, Maine
ABSTRACT
  Transformer recycling operations  at  a salvage  yard have
resulted in soils  and sediments becoming contaminated with
PCBs.  PCB compounds found in site soils and sediments were
identified as Aroclors 1242, 1248, 1254 and 1260. A gas chroma-
tograph with  electron capture detector (GC/ECD)  method to
screen  for the presence of PCBs was utilized to determine the
distribution of PCBs in  soils and sediments  at the site.  The
method was used in conjunction with confirmatory PCB analysis
using procedures  of the U.S. EPA Contracts  Laboratory Pro-
gram (CLP).
  The PCB screening procedure (screen) was applied  to over 300
samples of surface and subsurface soil and sediment  collected at
the site. Calibration of the screen was accomplished by submitting
samples that were split in the field for CLP analysis of PCBs. The
results indicate that for samples with PCB concentrations of less
than 100 ppm, the accuracy of the screen is essentially equivalent
to the CLP procedure. For samples with PCB concentrations ex-
ceeding 100 ppm, the screen underestimated the PCB concentra-
tion  by an average of 60%.
  The rapid turnaround of the screening procedure  allowed for
continual re-evaluation of the sampling program during the field
investigation. The screen's low cost allowed for the collection of a
greater number of samples, thus providing for a more accurate
site characterization. As a result, the use of the screen provided a
focused, expedient and cost-effective means of identifying the
distribution of PCBs at the site.

INTRODUCTION
  Defining the area! and vertical distribution and concentration
of PCBs in site soils and sediments was a primary objective of a
remedial investigation (Rl) conducted at  a Region 1 Superfund
site.  The  site  was  an abandoned salvage and scrap  yard where
transformer disassembly operations had occurred. The operations
consisted  of the disassembly of transformers and salvage of their
components.  Some spillage of transformer oils, one  component
of which was PCB, occurred during the disassembly process. Sur-
face  water runoff transported oils and oil-containing soils to on-
site lagoons and adjacent wetlands.
  Preliminary sampling  by the U.S.  EPA Region  1 Field  In-
vestigation Team in the early 1980s revealed high concentrations
of PCBs in the site soils and lagoon sediments.  PCB  compounds
found at the site were identified as Aroclors 1242, 1248, 1254 and
1260. In 1985 the site was placed on the NPL. A potentially re-
sponsible party (PRP) agreed to conduct a remedial investigation
and feasibility study (RI/FS).
  Although the baseline chemical data at this site was generated
using CLP analyses, the PRP sought an  analytical method that
would supplement the data base without significantly increasing
the cost of the overall analytical program. A screening technique
adopted from Spinier1 was selected as a means of generating ac-
curate site data for further characterization of the on-site soils
and sediments, at a low cost.
  The objectives for using this PCB screen at the site were to:

• Obtain rapid  sample analysis turnaround,  enabling real-time
  field sampling decisions that would optimize the length and
  cost of the field program
• Accurately characterize the site distribution of PCBs, particu-
  larly in the low concentration  range (i.e.,  0 to SO ppm), by
  analyzing a large number of samples
• Optimize the selection of samples for off-site laboratory analy-
  ses by CLP methods
  This paper presents  a brief review  of the project design, a
description of the screen and CLP procedures, a discussion of the
methods used to compare the procedures, a qualitative and quan-
titative evaluation of the results  and conclusions regarding the
overall effectiveness of the PCB screen in achieving the program
objectives.

PROJECT DESIGN
  The potential use of the GC/ECD method as a screening  tech-
nique for PCBs in a soil matrix has  been demonstrated  on a
bench-scale basis by Spit tier.' The method is capable of quantify-
ing PCB concentrations in excess of 1 ppm.  For this study, aD
sediment and soil samples collected were analyzed for PCB using
an adaptation of that screening technique. Sample analysis took
place at a nearby laboratory.
  The screen was calibrated for identification and quantification
of the four Aroclors found on-site. As calibration of the screen,
selected sediment and soil samples were split in the field and sub-
mitted  for extraction and analysis using the CLP procedure.'
ANALYTICAL PROCEDURES
  Both  procedures identify and quantify the concentration of
PCB through GC/ECD analysis of sample extracts. To identify
PCBs in the sample extracts, peak retention times and relative
peak  areas of  the sample chromatogram  were compared to
reference  chromatograms  of  the appropriate PCB Aroclors.
Quantitation of PCBs is based on a calibration curve constructed
from analysis of a reagent blank and calibration standards at a
minimum of three concentrations. The primary difference in time
and cost between the two procedures is incurred in the extraction
procedure. These procedures are summarized below:
 204     MONITORING & SAMPLING

-------
Screening Extraction
The screen extraction procedure consists of placing a 2.0 g ali-
quot of sample in a 10 ml glass test tube and adding 0.5 ml water,
2.0 ml methanol and 2.5 ml hexane. The test tube is capped and,
for 30 sec is either shaken vigorously or mixed on a vortex mixer.
After the mixture is allowed to settle for 5-10 min, the mixing step
is repeated.
The aqueous and organic phases are then separated by settling
or centrifuging. The hexane layer of the extract is transferred to a
test tube containing 3.0 ml of concentrated sulfuric acid. This
mixture is capped and mixed three times for 10-15 sec by shaking
or by using a vortex mixer, with a 5-10 min settling period be-
tween mixing. The acid-cleaned hexane layer is transferred to a
precleaned vial for subsequent GC/ECD analysis.
Using commonly available laboratory equipment, one analyst
can extract 10 to 20 samples in less than 2 hr. The cost of an
analysis, including GC/ECD analysis and data reduction is be-
tween $50 and $100. For this particular study, the extractions and
analyses were performed in the laboratory; however, the pro-
cedure could be performed in an on-site field trailer.

CLP Extraction
This procedure2 uses a more complex solvent extraction process
performed with the aid of ultrasonic agitation. A sample totaling
30 g is extracted with 300 ml of acetone and methylene choride
(1:1 v/v). The extract is concentrated and solvent exchanged to
hexane. The hexane extract, after cleanup (if necessary), is used
for subsequent GC/ECD analysis.
The time required for this procedure is approximately 1 hr per
sample. Because of the requirement for relatively long extraction
times, large volumes of solvent and expensive laboratory equip-
ment, the analytical cost per sample is over $300. In addition,
sample turnaround times for CLP analyses often exceed a month.

EVALUATION METHODS
To test the accuracy (ability to duplicate the CLP result) and
precision (ability to obtain reproducible results) of the PCB
screen, three evaluations were undertaken.

Precision and Accuracy of Procedures
For this evaluation, site soils were dried, homogenized and split
into replicates in the laboratory before analysis by both the screen
and CLP procedures. This comparison evaluated the accuracy
and precision of the screen under controlled conditions.

Soil Matrix Variability
This evaluation compared the results of duplicate samples that
were split in the field and analyzed by the screen. The evaluation
of field-split duplicate samples is important in determining the
degree of variability introduced by the soil matrices. The results
of duplicate analyses of both field and laboratory samples were
subjected to the one-way analysis of variance to compute the
within-sample variability.

Screen Calibration Analysis
The final evaluation determined the accuracy of the PCB screen
by comparing the results for samples that were split in the field
and submitted for both screen and CLP analysis. These results
were compared quantitatively using linear regression analysis and
frequency histograms.
For each of these comparisons, data for PCB Aroclors 1254
and 1260 were compared on a dry sample weight basis. To ac-
complish this, screen analyses (which were performed on undried
samples) were converted to dry weight concentrations using the
measured moisture content of the sample. Aroclors 1242 and 1248
were not used in the comparisons as they occurred only very infre-
quently at the site, and therefore represented a very small sample
population.

RESULTS

Precision and Accuracy of Procedures
A comparison of the CLP and screen procedures for PCBs in
subsurface soils is presented in Table 1. The soil samples utilized
for this comparison were air-dried and gently ground with a glass
rod prior to analysis.
The data indicate that for separate aliquots of a homogeneous
soil sample, a significant difference does exist between the PCB
concentrations measured by the two procedures. The t-statistic
values for the replicate measurements for samples A and B (see
Table 1) are 0.3 and 0.2 respectively. A t-statistic of greater than
2.4 indicates that the sample results of the two procedures were
statistically different (at a probability level of 95%).

Table 1
Comparison of PCB Analytical Techniques
11.6 7.8 7.B 9.1 ± 2.2

10.0 9.8 B.2 7.i 9.0 8.9 ± I .1
Soil Matrix Variability
An evaluation of analytical variability in soils was performed
through analysis of split duplicate samples collected during both
the surface and subsurface sampling programs. Table 2 reports
the precision based on coefficient of variation (CV) estimated for
the PCB screen for subsampling conducted in the laboratory and
for duplicate samples collected in the field. The reported precision
values show that the within-laboratory precision of 12-15% in-
creases to 27-159% for duplicate field samples, with the subsur-
face sample group exhibiting the greatest degree of variability.

Table 2
Precision of PCB Screen Analysis

Type of
Sample
Surface Soil
Subsurface Soil
Surface Soil
Subsurface Soil
Sediment

Location of
Sampling Splitting
Laboratory
Laboratory
Field
Field
Field

Number of
Samples
4
2
12
12
1
Coefficient of
Variation1 (I)
15
12
27
159
10
= Standard deviation divided by sample mean multiplied by 100X.
Computation of overall standard deviation and overall sample mean
was by one-way analysis of variance.

Screen Calibration Analysis
To calibrate the screen, split samples from 77 field locations
were analyzed by the CLP procedure. In contrast to the com-
MONITORING & SAMPLING 205

-------
parison of the procedures performed in the laboratory (Table 1),
this comparison incorporates differences between samples split in
the field.
AroclOf* 123«*lZflO (pom)
I7T1 CLP 1X^3 itr,./,
Figure 1
Frequency Distribution of PCB Concentrations in Soils and Sediments

Illustrated in Fig. 1 are the frequency distributions of measured
PCB Aroclors 1254 and 1260 concentrations for the screen and
CLP determinations. The histogram illustrates that the CLP and
screen procedures frequency distributions are similar and do not
show a constant bias of one procedure versus the other. The
histogram, however, does not address the accuracy of the screen
procedure, i.e., its ability to arrive at the same result on the same
split sample.
Linear regression analyses of CLP versus screening results were
conducted to more quantitatively address the accuracy and pre-
cision of the PCB screen. The results, presented in Table 3, have
been divided into three sections on the basis of the respective
sampling media (i.e., surface soils, subsurface soils and sedi-
ments). An additional statistical analyses was conducted to
evaluate the screen's performance for all media.
Overall the results of the data set indicate that, on the average,
the screen measurements are 40% of the CLP determinations.
Closer examination of the data suggests that the screen more ac-
curately reflects the CLP determination in the 0 to 100 ppm range
(regression slope of 1.05).
Table 3
Linear Regression Analysis of Screen Versus CLP PCB Results

Saapla
Surfaca Soil
Sublurfaca Soil
Sadlaent
All Saiaplaa
All Sanplei (<100 ppn)
All S.mple. (>100 ppn)

Saaplaa
32
)6
7
77
64
1)
•agraaalon
Slop.'
0.41
0.1)
0.40
0.40
1.03
0.3»

Coa((lclanc(R)
0.98
0.46
O.S)
0.95
0.7}
0.94
A scatter diagram of screen versus CLP results for the entire
data set is shown in Fig. 2; an inset of that plot on the 0 to 100
ppm scale is shown in Fig. 3. The figures suggest that the tendency
for PCB concentrations to be underestimated by the screen it
most pronounced in samples with PCB concentrations above ap-
proximately 100 ppm. This observation is substantiated by the
regression analyses (Table 3). For PCB concentrations lest than
and greater than 100 ppm, the screen measurements (as a percen-
tage of the CLP determinations), average 105% and 39%, respec-
tively.

1
I
8
;j
ft g
ll
|
i
#

01 -
0 • -
07 -

09 -
O.J •
O.a •

03 -
02 -

0
0

0
a c

0,Jo°° °
otVr, n „ .
0 02 0.« 0.9 Of
O.' trecm ! 214* 1290 (PP-)

Figure 2
Comparison of Screen and CLP Measurements of PCBs
in Soils and Sediments
100 -r-
K> -

90 -

70 -

90 -

50 -

«0 -

20
40 «D

O.P AroclOT I2S«»I290 (ppm)
1 • Standard llnaar ragraaflon aquation y • «ut + b wharf y • tcraan
raault, • • ilopa, x • CLP raiult, and b • y-lntarcapt.
Figure 3
Comparison of Screen and CLP Measurements of PCBs in Soils and
Sediments Where PCB Concentrations Do Not Exceed 100 ppm
DISCUSSION
Qualitatively, PCB screen data exhibit a frequency distribution
that appears similar to that of the CLP data (Fig. 1), with no con-
sistent bias of one procedure compared to the other. Further, the
screen data exhibit a distribution consistent with known on-site
commercial activities, geologic features and weathering and
redistribution processes of site soils and sediments. The horizon-
tal distribution of PCBs in soils exhibits well-defined zones of
high concentration that consistently decrease with increasiBS
206 MONITORING & SAMPLING

-------
distance from the source areas. CLP data support these observa-
tions.
In order to further support the applicability and accuracy of the
screen, quantitative method performance information is also re-
quired. The results of the laboratory replicate samples analyzed
by both the screen and CLP procedures (Table 1), illustrate the
close agreement of the two procedures for soils with PCB concen-
trations of 5 to 10 ppm. Sample precision of 12 to 15% for the
screen analysis falls within the 5 to 20% range calculated for the
CLP analysis for the same samples. Similar performance data
have been reported for U.S. EPA approved procedures for PCB
measurements of liquid and solid matrices.3-4
The evaluation of replicate analyses for the PCB screen in-
dicates that the field split samples (Table 2) exhibit greater
variability than the laboratory split samples (Table 1). This dif-
ference can be attributed to the heterogeneous distribution of
PCBs in the soil. The subsurface soils, for example, typically con-
sisted of a dense, stiff, clayey silt which often exhibited extensive
desiccation fracturing. PCB-containing oils travelling in the sub-
surface are likely to follow preferential pathways (e.g., fractures)
yielding greater PCB concentrations along the fracture surface.
As a result, subsurface soils at the site are susceptible to large
changes in concentration over very short distances (i.e., 1 cm or
less). These physical observations of the subsurface soils are con-
sistent with the high variability found within the subsurface soil
duplicates (CV = 159%; Table 2). The low correlation coefficient
(R = 0.46; Table 3) obtained from the regression of screen versus
CLP results is also consistent with the nature of heterogeneous
soil. In contrast, the surface soils and sediments at the site have
undergone extensive weathering and redistribution processes
which have resulted in a much more homogeneous matrix. As ex-
pected, they exhibit only slightly greater variation in duplicate
analysis than the laboratory split samples.
The linear relationship between the two analytical procedures
was also analyzed. For those samples with concentrations less
than 100 ppm, the screening result approaches that of the CLP
value (regression slope = 1.05; Table 3). For samples containing
PCBs in excess of 100 ppm, the screen tends to underestimate the
concentration of PCB relative to the CLP determination. One ex-
planation for this apparent underestimation is the heterogeneous
distribution of PCBs in a highly contaminated sample and how it
affects the analytical methods. Samples with high PCB concentra-
tions are likely to contain "hot spots" (i.e., very high PCB con-
centrations), which the 2 g aliquot necessary for the screen is less
likely to include than the 30 g aliquot used for CLP analysis. A se-
cond possible explanation is that the screen does not extract PCBs
as efficiently from samples with elevated PCB content. More
laboratory data are necessary to test both hypotheses.

CONCLUSIONS
The analytical results indicate that for samples with PCB con-
centrations less than 100 ppm, the sampling and analysis precision
and accuracy of the PCB screen are essentially equivalent to the
CLP procedure. At concentrations less than 10 ppm the precision
of the screen was within that of the CLP procedures; and similar
to performance data reported by the U.S. EPA for PCB
measurements of liquid and solid matrices.3'4
The accuracy of the PCB screen at low PCB concentration was
instrumental in accurately delineating the fringe areas surroun-
ding known site sources. While the screen tended to underesti-
mate PCB concentration in the highly contaminated areas, it still
provided preliminary indications of soils containing very high
PCB concentrations. This information was useful in selecting
samples for more accurate (and precise) CLP analysis.
Use of the PCB screen should be reviewed carefully prior to im-
plementation. The following considerations should be noted:
• The screen is generally performed on a "wet" soil matrix;
therefore, differences in sample moisture content should be
considered when evaluating PCB sample content.
• All split samples should be completely homogenized in the
field to minimize variability within split-duplicate samples.
• The screen must be calibrated and the results must be quanti-
fied for all Aroclors thought to be present at the site.
In summary, the performance findings of the PCB screen, com-
bined with the minimal extraction equipment requirements and
rapid turnaround time justify its use—at PCB sites—in sup-
plementing the CLP data base. Continual data updates, in con-
junction with the collection of a large number of selectively placed
samples, can result in the accurate, cost-effective quantification
of PCB-containing soils.
ACKNOWLEDGEMENTS
The authors thank D. Twomey, S. Turner and B. Wallin for
technical assistance with the PCB screen analyses, and G.
Mikeska and D. Poor (all of B.C. Jordan Co.) for continued sup-
port and assistance with data reduction and evaluation. They also
thank CompuChem Laboratories for technical assistance with the
CLP analyses.

REFERENCES
1. Spittler, T.M., "Field Measurement of Polychlorinated Biphenyls in
Soil and Sediment Using a Portable Gas Chromatograph," Environ-
mental Sampling for Hazardous Wastes, American Chemical Society,
1984, 37-42.
2. U.S. EPA, "Statement of Work for Organic Analysis, Exhibit D:
Analytical Methods," U.S. EPA Contract Laboratory Program,
7/85 Revision, 1985, K1-D134.
3. U.S. EPA, "Method 608—Organochlorine Pesticides and PCBs,"
40 CFR Part 136 Federal Register, 49, 1984, 43321-43336.
4. U.S. EPA, "Method 8080—Organochlorine Pesticides and PCBs,"
EPA SW-846 Test Methods for Evaluating Solid Waste, 2nd Ed.,
8080/1-8080/17, U.S. EPA, Washington, DC, 1982.
MONITORING & SAMPLING 207

-------
Performance of the Toxicity Characteristic
Leaching Procedure

Lynn R. Newcomer
Wilson Laboratories
Salina, Kansas
W. Burton Blackburn
S-Cubed
LaJolla, California
Todd A. Kimmell
U.S. Environmental Protection Agency
Washington, D.C.
ABSTRACT
The U.S. EPA recently has promulgated the use of a leaching
procedure (TCLP) to be used as a criterion for disposal of hazar-
dous wastes under the Land Disposal Restrictions Rule. The
TCLP also has been included in the proposed rule for Identifica-
tion and Listing of Hazardous Wastes.
Summarized in this paper are results of eight different studies
designed to evaluate the TCLP method. Study results indicate
that the TCLP procedure can be applied consistently by a diverse
group of laboratories.

INTRODUCTION
RCRA directs the U.S. EPA to identify and regulate wastes
which pose a hazard to human health and the environment. In
order to meet this mandate, the U.S. EPA listed a number of
wastes as hazardous and identified four hazardous waste
characteristics. One of these, the Extraction Procedure (EP) Tox-
icity Characteristic, addresses wastes which exhibit leaching
potential into groundwater supplies if improperly managed.
In 1984 Congress amended RCRA directing the U.S. EPA to
make changes in the existing Extraction Procedure to predict
more accurately the leaching potentials of hazardous wastes and
expand its application to a greater number of toxic constituents.
In response, the U.S. EPA developed and proposed the Toxicity
Characteristic Leaching Procedure (TCLP). The second genera-
tion leaching procedure, TCLP, was published as a draft protocol
on Dec. 20, 1985; it was officially published in the Federal
Register (Vol. 51, No. 9, Jan. 14, 1986) as part of the proposed
Land Disposal Restrictions Rule and again in the Federal Register
(Vol. 51, No. 114, June 13, 1986) as part of the proposed rule for
Identification and Listing of Hazardous Waste. The final land
disposal restrictions rule was published in the Federal Register
(Vol. 51, No. 216, Nov. 7, 1986).
The TCLP protocol calls for an 18-hr extraction of a waste
sample with either an acetic acid or sodium acetate solution
followed by the determination of metals, pesticides and semi-
volatile and volatile organic compounds in the leachate. Metals,
pesticides and semi-volatile organic compounds are extracted us-
ing a bottle or jar similar to the EP procedure. For the extraction
of volatile organic compounds (VOCs), a new device known as a
Zero Headspace Extractor (ZHE) is used.
To date, there have been various single and multi-laboratory
studies conducted to evaluate the TCLP procedure. Twenty-three
government, industry, research and commercial laboratories par-
ticipated in various phases of the evaluation studies. At least 15
different types of wastes were used including metal manufactur-
ing, power plant, refinery, electroplating, textile and Publicly
Owned Treatment Works (POTWs) wastes.
This paper combines and condenses data from these various
studies and discusses the method (TCLP) performance, i.e., com-
parison to the EP procedure, ruggedness, precision and reproduc-
ibility. Data presented herein represent a fairly comprehensive set
of data based on the analyses of the various wastes.
In order to condense available data, only results of analytes
detected in greater than 90% of the samples are reported.
Outliers, as determined by the Dixon Test1 at the 95<7* confidence
level, also are excluded. In several of the studies, the Dixon Test
had not been applied to the study results. Where raw data were
available, the Dixon test was applied by the author.
Most of the original studies conducted under U.S. EPA con-
tract or by private industry as referenced in this paper are
available for public review in the U.S. EPA RCRA Docket
(S-212) at the U.S. EPA, Washington, D.C.
TCLP VERSUS EP
Two studies comparing the TCLP and EP extractions are sum-
marized. The sponsoring industries selected 13 different wastes
for extraction and analysis. Only results reported by more than
one laboratory and on analytes which were detectable are in-
cluded here.

Inter-Industry Collaborative Study
The Inter-Industry Collaborative Study' was sponsored by six
trade associations with the goal of evaluating the TCLP on wastes
of interest to the organizations. Only one waste, waste 16, is
reported since it was the only waste analyzed by all six par-
ticipating laboratories. This waste consisted of a composite of
baghouse dust from a steelmaking operation and sinter waste
from a lead smelting facility. The results of this study are sum-
marized in Table 1. For the five metals for which data are given,
there does not appear to be a large difference in the amount of ex-
tractable analytes. Although the RSDEP for lead is significantly
greater than the lead RSD-rcLp, the other four metals indicate
similar extraction repeatability. Since the raw data for this study
were not available, it is not known whether outliers were dis-
carded. The study concluded that although the TCLP and EP are
not precise methods, they are similar in precision.
208
MONITORING & SAMPLING

-------
                          Table 1
       Inter-Industry Collaborative Study Summary Results
          of EP and TCLP Metal Extracts of Waste #6
Pirueter
Arsenic
B«riu«
Cldaiim
Chroniua
lead
Avenge
XEP
0.048
5.9
0.19
0.090
293
XTCLP
0.039
14. 5
0.17
0.069
256
SEP
0.024
3.77
0.061
0.065
471
STCLP
0.021
10.5
0.067
0.048
241

ZRSDjp
50
64
32
72
161
76
ZRSDrCLP
44
72
40
69
94
64
      Units - «g/L
      X • Mean values for 11-12 extractions
      S • Standard Deviation
      ZRSD - Relative Standard Deviation
Electric Power Research Institute
  A second study comparing TCLP and EP procedures was spon-
sored by the Electric Power Research Institute (EPRI).3 Seven dif-
ferent utility wastes (including fly ashes, bottom ashes and flue
gas sludges) were extracted in duplicate by three laboratories. The
six extracts of each waste were divided equally among the three
laboratories and analyzed in quadruplicate.  A summary of the
results is reported in Table 2.

                          Table 2
EPRI Results of EP and TCLP Metal Extracts  of Seven Utility Wastes
Me tall
Anenic, W-2
W-7
Biriim, V-l
W-3
W-7
Cidiiui, W-l
W-2
W-5
W-6
W-7
Chroniuffl, W-l
W-2
W-3
W-4
W-5
W-7
Lead, W-2
Selenium, W-7
Zinc, W-l
W-2
W-3
W-5
W-6
W-7
Average
Without
Zinc. w-2
Average
XEP
ND
0.051
0.406
0.430
0.177
0.013
0.224
0.033
0.005
0.006
0,427
0.016
0.008
ND
0.030
ND
ND
0.061
0.171
5.36
0.092
1.48
0.174
0.151
0.3
XTCLP
0.317
0.149
0.327
0.819
0.446
0.016
0.233
0.028
0.004
0.006
0.470
0.921
0.010
0.004
0.042
0.059
0.181
0.135
0.238
5.37
0.164
1.6
0.306
0.234
0.50

SEP

0.034
0.178
0.219
0.060
0.004
0.038
0.133
0.002
0.001
0.077
0.023
0.006

0.026


0.024
0.142
0.3
0.129
0.12
0.052
0.073
0.08
STCLP
0.228
0.021
0.147
0.336
0.138
0.004
0.100
0.005
0.002
0.002
0.070
0.405
0.005
0.002
0.036
0.018
0.122
0.036
0.171
0.38
0.174
0.22
0.278
0.140
0.13
tRSDgp

67
44
51
34
31
17
25
48
18
18.
1451
75

85


39
83
6,
140
8
30
48
51
iRSDrcLp
72
14
45
41
31
25
43
17
38
32
15
44
46
55
86
30
67
27
72
7,
1062
14
91
60
45


0.17
0.29

Note:  Dnit
     W-l
     W-2
     W-3
     W-4
     W-5
     W-6
     W-7
      -mg/L
      Alkaline fly ash
      Acidic fly ash
      Alkaline bottom ash
      Neutral botton ash
      Forced oxidized flue gas desulfurization sludge
      Flue gas desulfurization sludge
      Neutral fly ash
Appro
      "cuLLal lly avn
    " 50Z of the results were not detected

  2 One lab reported results 10X lower than the other two labs
   Analyses of one lab's extracts were approximately 10X greater
   than those of the other two labs
(*i"ately 40 analyses were perforned on each waste type.
  During the period of public comment, one of the frequently
mentioned concerns of the TCLP was  the possible aggressive
leaching potential of the acetate extraction medium. The average
total metal extractables appear to be somewhat higher for the
                                                                TCLP extraction (0 .5 mg/1 vs 0.39 mg/1 or 0.29 mg/1 vs 0.17
                                                                mg/1) if zinc, waste W-2, results are not included; the relatively
                                                                large  zinc values tend to  buffer significant differences  of the
                                                                smaller values. Eighty percent of the results in Table 2 showed a
                                                                ratio of XxCLP:XEP of 0.8-2.0 and 15% fell within the 2.0 to 3.0
                                                                range (disregarding the data pairs containing "ND"). Although
                                                                these  ratios tend to support a conclusion that the TCLP is more
                                                                aggressive, all extractable results reported are well below hazar-
                                                                dous  waste threshold levels. If the TCLP is more aggressive, it
                                                                would appear to be waste and metal-specific.
                                                                  The EPRI study concluded that the TCLP appears to provide
                                                                better extraction reproducibility but that inter-laboratory varia-
                                                                tions  are a significant factor in TCLP analyses.
                                                                  For the majority of wastes analyzed  in these  two studies, the
                                                                TCLP and EP are similar in extraction efficiency and method
                                                                precision. Where differences exist, the TCLP tends  to provide
                                                                better precision, is  easier to perform and produces a more ag-
                                                                gressive leaching medium.

                                                                RUGGEDNESS
                                                                  Two ruggedness studies have been performed to determine the
                                                                effect of various perturbations on specific elements of the TCLP
                                                                protocol. Ruggedness testing determines the sensitivity of small
                                                                procedural variations which might be  expected  to occur  during
                                                                routine laboratory application.  Method  variations  which  are
                                                                observed to significantly affect analytical results need to be
                                                                carefully controlled.
                                                                  Both studies followed the partial factorial design described by
                                                                Youden.  In this design, seven conditions were slightly altered and
                                                                eight extractions were performed. This design provided sufficient
                                                                information to identify those areas of the method which were af-
                                                                fected by procedural variations.

                                                                Metals
                                                                  A study by ENSECO4 investigated the following conditions for
                                                                metals results on two wastes:
                                                                   1) Liquid/Solid ratio
                                                                   2) Extraction time
                                                                   3) Headspace
                                                                   4) Buffer #2 acidity
                                                                   5) Acid-washed filters
                                                                   6) Filter type

                                                                   7) Bottle type
                                                                                            19:1 vs21:l
                                                                                            16 hrs vs 18 hrs
                                                                                            20% vs 60%
                                                                                            190 meq vs 210 meq
                                                                                            yes vs no
                                                                                            0.7 /tm glass fiber vs 0.45
                                                                                            vs polycarbonate
                                                                                            borosilicate vs flint glass
  Of the seven method variations examined, acidity of the extrac-
tion fluid had the greatest impact on the results. Four of 13 metals
from an API separator sludge/electroplating waste (API/EW)
mixture and two of three metals from an ammonia lime still bot-
tom waste were extracted at higher levels by the more acidic buf-
fer. Because of the sensitivity to pH changes, the method requires
that the extraction fluids be prepared so that the final pH is within
±0.05 units as specified.
  Although not directly apparent from this study, other studies
referenced have indicated that flint glass, non acid-washed filters
or other filter materials may also affect results. Therefore, glass
fiber filters (acid washed for metals analyses) are required, and
borosilicate glass is recommended.
                                                                Volatile Organic Compounds
                                                                  A separate  ruggedness  study was performed  by ERCO/
                                                                ENSECO5 to investigate  method variations on volatile organic
                                                                compounds (VOCs) in API/EW and ammonia lime still bottom
                                                                wastes. The following parameters were tested:
                                                                                            MONITORING & SAMPLING    209

-------
                                     19:1 vs21:l
                                     0°7o vs 5°/o
                                     60 meq vs 80 meq
                                     ADM vs Millipore
                                     Syringe vs Tedlar bag
                                     yes vs no
                                     0  lb/in.2 vs 20 lb/in.2
1) Liquid/Solid ratio
2) Headspace
3) Buffer (Cl  acidity
4) ZHE device
5) Method of storing extract
6) Aliquotting
7) Pressure behind piston
  The only parameter having a significant effect on  the results
was the choice of the extraction device. The original Zero Head-
space Extractors (ZHEs) manufactured by Millipore  had prob-
lems with leaking valves and with piston movement which resulted
in loss of volatile compounds.  Millipore has corrected the valve
and piston problems so that both the Millipore and ADM ZHEs
should perform similarly. No problems were  reported with the
retrofitted Millipore  ZHEs used in the  S-Cubed  Collaborative
Study referenced later in this paper.
  The ERCO/ENSECO study reported that the ZHE was "ade-
quately rugged with respect to the parameters investigated."

PRECISION
  Substantial data have been  generated from TCLP precision
(reproducibility) studies.  Both  single- and  multi-laboratory
studies to assess method precision have been sponsored by private
industry and  by the U.S.  EPA. Precision  results from recent
studies are encouraging. The general consensus from these studies
is that the precision of the TCLP is comparable to or exceeds that
of the EP procedure and that method precision is adequate. One
of the more significant contributions to poor precision appears to
be related to sample homogeneity and inter-laboratory variation
which is not surprising due to the nature of waste materials.

Metals
  The largest source of TCLP precision data comes from the
metals analyses of TCLP extracts. Twenty-three laboratories per-
formed extractions and  analyses  on 15 different  wastes. The
results from  these  analyses are presented  in  the  tables below.
Tables 3 and 4 contain single- and multi-laboratory results, re-
spectively.  Results included in Tables 1 and 2 also provide pre-
cision information. The general range for mean percent RSDs in
Tables 1, 2, 3 and 4 is 22-74%. Although not necessarily an in-
dication of a precise method, the range is not unreasonable con-
sidering the waste types and the relatively low levels of metals
determined.

                           Table3
     Single-Laboratory TCLP MeUb, Precision (William*,  el aL)
                        Table 4
Multi-Laboratory TCLP Meiali, PrtcUlon (Biackbira ud Show)
Uaete
Aaauale Lia*
Still lottou





API/EW
Mixture




Extraction
UuU

tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
Natal

Sariuai

Chroflluai

Zinc

l.rlu.

Ckroailuai

Zinc

I

0.283
O.JJ2
0.04*
0.0*2
0.170
0.257
0.69*
1.01
0.125
14.4
11*
347
»

0.01)
0.015
0.021
0.012
0.12*
0.071
0.020
0.01
0.0*2
1.6
7.5
14.2
USD

4.6
4.6
47
1*
71
2«
2.S
0.6
4*
25
5.5
4.1
                                          XUD Unga - 0.6-71
                                          Mean USD  - 22
Vaete
AaMoaia Lla»
(till lottou




AFI/EV
Mixture




Foeell fu«l
t\J Ask




Extraction
ruid
tl
H
tl
ti
ti
ti
ti
ti
ti
ti
ti
ti
ti
ti
ti
ti
ti
ti
(total
Caoalu.

CkroaUeai

Lead

CaiUlUB

Ckroaloa

Lead

CaaaUum

CkroaUuai

Lead

X
0.053
0.02)
0.015
0.0012
0.00 JO
0.0012
0.0046
0.0005
0.0561
0.105
0.0011
0.012*
O.OM
0.0*1
0.017
0.070
0.0017
0.0457
1
.011
.017
.0014
.0017
.0027
.002S
.0021
.0004
.0227
.011
.00)1
.011*
.06*
.0*7
.014
.040
.0074
.00*)
SUB
M
H
tl
111
M
(7
tl
77
40
17
100
110
M
72
15
57
15
11
                                                                                                               ItSO laa*e '
                                                                                                               HU0 ZKD  '
                                                  17-111
                                                  74
                                                                     Iota:  X - Mean reeults from 4-12 different laboratories
                                                                           Onite • a)f/L
                                                                           Extraction FUld tl - pal 4.*
                                                                                         «2 • p. 2.*
                                                                      Less than 5V» of these wastes would be regulated as hazardoot
                                                                    wastes under 40 CFR, Pan 261, based on the data presented in
                                                                    these tables. The results indicate that a single analysis of a waste
                                                                    may not be adequate for waste characterization and identification

                                                                                              Tables
                                                                               Static-Laboratory Seaai-Votettka. Predakm
Uaete
Aaaunia
Llaa Still
lottou






















AFI/EV
Mixture






Coa*alpkenol

Napktkalaoa

2-MetkrlMBktkalane

Extract ioa
N«U

tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
tl
11
I

1*000
1*400
2000
1MO
7*40
74*0
121
M7
1*20
M27
2*0
271
117
117
701
Ml
151
15*
2*1
2*1
11.
H.
25.
2*.
40.
If.
11.
41.
115
145
2*5
200
1

22M
*2*
2*7
S2.I
DM
200
u.
45.
411
17*
44.
U.
22.
7.
I*.
20.1
17.*
2.1
22.7
7.*
t.l*
l.SS
1.1
l.«
11.5
1.71
».15
i.«
2t.4
24.S
tl.2
ll.t
U,

ll.(
4.1
14.*
2.1
U.4
2.7
l*.t
14.*
10.5
*.«
15.5
7.1
12.1
l.»
12.7
1.0
11.7
1.1
».*
1.)
1M
4.5
7.1
7.1
11.0
*.!
21.)
U.t
15.1
11.0
21.1
».S
                                                                                                              XKSD Inta - 1-U
                                                                                                              Maa> ItSO  - 12
Note:  All extraction! vere perforead in triplicate.
      Extraction Fluid fl - pH 4.*
                    tl • pH 2.9
      Unite - aig/L
                                                                   Mote:  Unite • u|/l
                                                                         Extractions were perforaad  In triplicate
                                                                         All reeulce nra at laaat IX tka detection limit
                                                                         Extraction Fluid *1 - pH 4.*
                                                                                       «2 - pH 2.9
210    MONITORING & SAMPLING

-------
requirements,  a comment reinforced by the Inter-Industry Col-
laborative Study and in the Background Document.6

Semi-Volatile Organic  Compounds
  Semi-volatile organic compounds have not been studied as ex-
tensively as the metals. There are, however, two studies (a single-
laboratory evaluation4  and  a  multi-laboratory collaborative
study') which provide data from several different waste  types.
Results of these studies are summarized in Tables 5 and 6.

                          Table 6
           Multi-Laboratory Semi-Volatiles, Precision
                                                  Table 7
                                       Single-Laboratory VOCs, Precision

Vaste
imwnia Lime
Still Bottoms (A)
API/W
Mixture (B)
Poiiil Fuel
Ply Ash (C)

Compound
BNAs

BNAs

BNAs

Extraction
Fluid
«
*2
*1
tl
*1
*2

X
10043
10376
1624
2074
750
739

S
7680
6552
675
1463
175
342

ZRSD
76.5
63.1
41.6
70.5
23.4
46.3
                                                 Mean ZRSD - 54
Note:  Units - pg/L
     X • Mean results  from 3-10 labs
     Extraction Fluid  #1 - pH 4.9
                   /2 - pH 2.9
                                ZRDS Range for Individual Compounds
                                     A,  *1
                                     A,  *2
                                     B,  tl
                                     B,  *2
                                     C,  *1
                                     C,  #2
 0-113
28-108
20-156
49-128
36-143
61-164
  Single-laboratory precision was excellent with greater than 90%
 of the results exhibiting an RSD less than 25%. More than 85% of
 all individual compounds in the multi-laboratory study fell in the
 RSD range of 20-120%. The actual ranges are included in Table 6.
  The single-laboratory  evaluation4  reported somewhat  better
 reproducibility for semi-volatiles  with the more acidic extraction
 fluid, #2. However, the multi-laboratory study7 did not confirm
 the same relationship between extraction fluid acidity and pre-
 cision.
  Both studies concluded that the TCLP provides adequate pre-
 cision. It was also determined that the high acetate content of the
 extraction fluid did not present problems (i.e., column degrada-
 tion of the gas chromatograph) for the analytical conditions used.

 Volatile  Organic Compounds
  Both single- and multi-laboratory data are available  for VOC
 precision. Three studies using a  combination of three different
 waste types are summarized.
  A single-laboratory study conducted by S-Cubed" involved the
 use of two waste types plus a  VOC-free sand/water mixture
 blank. In order to assure that volatile compounds were present to
 extract, each waste was separated into two fractions and spiked
 with a mixture of VOCs. One fraction was spiked at mg/1 concen-
 tration and the other at /tg/1 concentrations. The spikes were add-
 ed to the system at two different times and locations in the extrac-
 tion process. Group 1 compounds (Table 7) were spiked into the
 sample immediately after placing the waste in the ZHE (both
 ADM and Millipore extractors were used). The initial waste liquid
 was removed by the ZHE, as specified in the TCLP protocol, and
 collected for analysis. Group 2 compounds were added with the
 extraction fluid as  the fluid  was being pumped into the  ZHE. At
 the end of the 18-hr extraction procedure, the extraction fluid was
 collected from the ZHE and combined with the liquid collected
 from the first solid/liquid separation. Results of the analyses are
 presented in Table 7. To reduce the amount of data, compound
 groups were combined for  the API/EW and sand/water blank
 wastes.
Waste

Ammonia
Lime Still
Bo t tons





































API/EW
Mixture




Sand,
Spiked
Blank






Compound
Group 1 Compounds
Acrylonitrile

Carbon disulfide

2-Butanone

Benzene

Toluene

Chlorobenzene

Group 2 Compounds
1 , 1-Dichloroethylene

Chloroform

1 , 2— Dichloroe thane

1,1, 1-Trichloroe thane

Carbon tetrachloride
3
Trichloroethylene

1,1, 2-Tr ichloroe thane

Tetrachloroethylene

1,1,1, 2-Tetrachloroe thane
3
1,1,2, 2-Tetrachloroe thane

Surrogate!
d4-l ,4-Dichlqroethane

Bromof lurobenzene

d.-Toluene

Group 1 Compounds

Group 2 Compounds

Surrogates

Group 1 Compounds


Group 2 Compounds


Surrogates


Spike Level

200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm

200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm

200 ppb
1 ppm
200 ppb
1 ppm
200 ppb
1 ppm
1 ppm'
5 ppm*
1 ppm
5 ppm1
1 ppm
5 ppm
20 ppb}
200 ppb1
1 Ppm,
20 ppb;
200 ppb.
1 ppm1
20 ppb
200 ppb
1 ppm
X

69
62
47
20
75
66
93
46
87
47
74
52

80
89
90
94
93
95
SI
88
74
86
136
151
86
88
63
69
88
96
16
15

96
95
97
99
101
101
51
56
38
35
101
102
62
54
92
91
93
34
95
102
99
S

5.3
3.6
3.3
3.2
6.0
6.6
3.0
8.3
4.0
7.1
2.2
5.7

4.3
4.2
2.9
8.0
3.0
7.7
3.3
3.2
7.0
4.8
6.9
13
1.6
4.1
3.7
5.0
1
0
4.6
8.4

3.4
5.6
5.8
1.2
3.9
4.1
KD
ND
MD
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
IRSD

7.6
5.8
7.1
16
8.1
10
3.3
18
4.7
15
3.0
11

5.4
.7
3.2
8.5
3.2
8.1
4.1
3.6
9.5
5.6
5.1
8.6
1.9
4.7
5.9
7.3
12
11
29
56

3.5
5.9
6.0
1.2
3.9
4.1
7.6
9.0
8.5
7.6
3.0
2.5
8.2
10
7.2
1.5
7.3
20
2.0
4.6
3.6
                                                                   Mean ZRSD  - 8
                                                                   ZRDS Range - 3-20
                       Note:  X • Mean Z recovery for five replicates
                            . Compounds spiked into the waste
                            - Compounds spiked into the extraction fluid
                            , Recoveries indicate  possible transformations due to dehydrohalogenation
                              Surrogate and Blank  data not included
                            ND - Not determined

                         Although the data presented in Table 7 are reported in percent
                       recoveries, the reproducibility of percent recoveries is an indicator
                       of precision. The precision results  (RSDs) are  excellent. Only
                       tetrachlorethane  and trichloroethylene  showed  atypical  re-
                       coveries. The report suggested that dehydrohalogenation may
                       have occurred converting tetrachloroethane to trichloroethylene
                       and HCI. Since both Group 1 and 2  RSDs fall in  the same range,
                       it is unlikely that the spiking scheme contributed significantly to
                       the excellent precision.
                                                                                            MONITORING & SAMPLING    211

-------
  A second study directed by Oak Ridge National Laboratory*
evaluated the performance of ZHEs with two wastes. Ten extrac-
tions were  performed on fortified  wastes that were spiked with
halocarbons and aromatics in a scheme similar to that described
in the  S-Cubed study. The liquids from the  initial liquid/solid
separations and the final extracts were analyzed  individually.
Four of each set of 10 extractions were analyzed by Oak Ridge
National Laboratory while the other six of each set were analyzed
by another laboratory. The results  are summarized  in Table 8.

                          Table 8
           Multi-Laboratory (Two Labs) VOCs, Precision
                            Table 9
            Multi-Laboratory (11 Labs) VOCs, Prcdclon
U«lt«
Ammonia Lime
Still Bottoma










API/EV
Hixture




Compound
Bentene

Bromodichlorome thane
Bromoform
Carbon tetrechlorlde
Chlorobencene
Chloroform

Hethylene chloride
Tetrechloroethjrlene
Toluene

Bensene

Toluene

Hethylene chloride

Liquid
F ee liquid
L achate
r ee liquid
F ee liquid
T ee liquid
F ee liquid
P ee liquid
Leechete
Free liquid
Free liquid
Free liquid
Leachete
Free liquid
Leechete
Free liquid
Leechete
Pree liquid
Leechete
X
36. J
90.3
24.7
30.1
3*
)}
32
17.2
235
59.8
35.2
65.2
74
99
01
44
14
52
t
11.2
47.7
16.3
11.4
47.1
25. S
85.2
11.2
136
31.1
11.6
22.3
117
262
71.5
210
545
424
ItSD
10.7
52. »
66.2
61.2
34.6
19.4
19.7
6). 2
57.7
52.1
32.9
14.2
50.0
S7.4
14.4
47.4
106
93.7
                                            Neen (USD - 53
                                            ZU>S l»|e - 19-106
      Unite - MI/L
      X • Keen reeult* fro* 9-10 direction!
  In general, precision results are similar to those reported in the
study involving semi-volatile organic compounds (Table 6).
Although both wastes were spiked with a mixture of compounds,
the  oily  characteristics  of  the AP1/EW  waste apparently
prevented partitioning of the volatiles into the extraction fluid.
Many of the analytes either were not detected or showed a wide
range of concentrations. This can be seen from  the larger RSDs
for several of the compounds, methylene chloride and benzene.
Although the variations may be  greater for  oily wastes  which
reduce the effectiveness and applicability of the  method, the
variations are not out of line when compared  to other  multi-
laboratory results. It does, however, indicate that not all wastes
may be effectively characterized by the TCLP method.
   A third study, coordinated   by S-Cubed (Blackburn  and
Show), was a collaborative study in which 23 laboratories partici-
pated. Only 11 laboratories participated in the ZHE testing. Two
wastes, API/EW and mine tailings wastes, were fortified with a
mixture of VOCs. All participating laboratories  had  prior ex-
perience in using the ZHEs (ADM and Millipore). A summary of
the results is included in Table 9.
  Precision results for VOCs tend to occur over a considerable
range. However, the  range and mean RSD compare very closely
to the same collaborative study metals results (Table 4). Black-
burn and Show concluded that at the 95% level  of significance:
• Recoveries among laboratories were statistically similar
• Recoveries did not  vary significantly between the two sample
  types
• Each laboratory showed the same pattern of recovery for each
  of the two samples

  One can, therefore, conclude that the two samples contributed
equally to the leachable VOCs and that the RSD range does not
u»ct
Mliu Tilling!


















Aaanala Lima
Still Bottom*

















Compound
Vinyl chloride
Hetkyleoe chloride
Carboa dliulflde
1,1-Dickloroechaaa
1,1-Dlchloroathaa*
Chloroform
l,2-Dlchloroetb»ae
2-BvtaOO«e
1,1,1-Trlchloroe thane
Cerbaa tecrechloride
Trlcklorottheoe
1 , 1 , 2-Trtchloro«th«n«
Beaaeae
1 , 1 . 2. 2-Tetrachloroethana
Toluene
Chlorobenaene
ItkylbenieM
TrickloroflaorametlUM
Acrjrlooltrile
»lnyl chlorite
NatbyleBa chlorid«
Carbon diaalfide
1,1-Dlckloroatkena
1 , 1-DUkloroethaoe
Chloroform
1,2-Olchloroothane
2-Batanoae
1,1, 1-Tri Chi era* thane
Carbon tetrechloriae
Trlchloroe theme
1.1,2-Trichloroaehana
Banxene
1 , 1 , 2,2-Tetrachloroethame
Toltteae
Chlorotooaeme
Ethrlbemaeo*
Trichloroflaorome thane
Acrylomitrlle
I
*.36
12.1
3.57
21.
31.
44.
47.
41.
20.
12.
2*.
1*.
17.
J*.
».
15.
4.27
1.12
7«.7
5.00
14.1
1.17
52.1
52.1
M.7
41.1
59.0
51.*
7.10
57.1
6.7
61.3
3.16
W.O
71. 6
1.70
4.05
2*.*
1
t.M
11.1
2.11
27.7
25.4
29.2
13.*
It.*
20.»
t.2
21.2
10.*
21.7
25.*
11.2
lt.1
2.M
4.40
110.S
4.71
11.1
2.07
l».i
25.*
21.4
11.5
3*.*
40.*
«.l
M.2
4.7
2*.*
2.1
li.S
12.0
2.2
4.1
34.«
tin
100
M
51
127
11
U
n
15
WO
M
M
M
7*
n
M
M
M
115
1*4
*t
>2
tl
75
41
M
71
(7
7*
It
CO
70
44
**
27
17
51
11*
111
                                            IMS
                                                     n
                                                     • 17-14*
note:  Oaite - ot/L
                          Table 10
            TCLP Sammary of Twelve POTW Wastes
Parameter

Metale (1 SP Hetale)

Seai-Volatilee

Voletllee

Analytee Pound in at
Leaet 501 of the Waetee
Barium
Cadaiuai
Tolu ne
2-Bu anooe
Ithy bemade
Neth lleobutylketone
Kyi. e
p-Cr eol
Phen 1
Laboratorr

SPA
POTW
EPA
POTW
EPA
POTW
(ante of Analyta
Concentrations (•£/!.)
0.2-1.*
0.02-0.21
0.0008-0.95
0.0165-2.2
0.0006-0.0085
0.0048-0.064
0.0045-0.05
0.17-1.5
0.0008-0.19
Z of Aulrtaa Deuctt*
TCLP V
27 11
16 at
2 U
2 «A
4 U
10 IA
Propoied Utalatory
ThraahoU to/I)
100
5.0
14.4
7.2
m.
m.
ML
10
14.4
 Not*:  NA • Not antl/Bcd
       ML - Ho Unit* h«Vi b*«n proposed
212     MONITORING & SAMPLING

-------
preclude acquiring consistent results.

TCLP AND MUNICIPAL POTW SLUDGES
Two studies have been conducted to determine the impact of
the TCLP on Publicly Owned Treatment Works (POTWs)
wastes. S-Cubed10 and the U.S. EPA's Office of Water" have
coordinated testing of 12 POTW sludges. These wastes were
analyzed as split samples by S-Cubed, the U.S. EPA contract
laboratory and by the POTW laboratory or by a POTW con-
tracted laboratory. S-Cubed determined both EP and TCLP con-
centrations of the wastes.
Statistically, it is difficult to evaluate the data in terms of RSDs.
However, some general observations are in order. Table 10
presents a summary of these studies.
Results reported from POTW wastes indicate that analytes of
interest, TCLP constituents, are present in relatively low concen-
trations. Even though the results do vary, indications are that the
TCLP method can adequately and reproducibly be used for
municipal wastes. The only compound approaching the threshold
limit was 2-Butanone found in one waste. An earlier U.S. EPA12
study of six municipal sludges found that benzene and chloro-
form also approached the threshold limit for two sludges. It ap-
pears that the ZHE/VOC analysis is the most critical application
of the TCLP to POTW wastes.
The results of the POTW study are encouraging in light of the
previous studies. Some of the POTW tests were performed by
POTW laboratories or other commercial laboratories which did
not have experience with TCLP equipment. Quality control
guidelines, while included in the method, were not otherwise
specified. In other words, the results may be more typical of
laboratory variations among all types of laboratories. It could,
however, be argued that since the POTW sludge had relatively
low levels of constituents of concern, that a true test of labora-
tories' abilities to perform the test had not been measured.
The POTW study results do not appear to differ greatly from
the method performance studies.

CONCLUSION
From strictly an analytical application, the TCLP provides a
workable method which performs equally well or better than the
EP procedure. With practice and good laboratory skills, the ex-
traction procedure can become routine. As stated earlier, dif-
ficulties in achieving sample homogeneity and in collecting
representative samples from waste sources may imply that a single
TCLP analysis of a waste could provide misleading information.
The TCLP was designed to provide reproducible extraction
results, not to determine total constituents of a waste. It is likely
that some waste constituents may give highly variable results
which are matrix dependent and unpredictable. Distribution coef-
ficients (K,js) or sorption coefficients of compounds will vary with
the sample matrix. In general, TCLP performance appears to
satisfy regulatory needs.

ACKNOWLEDGEMENT
Todd A. Kimmell, former U.S. EPA Project Director for the
Development of the TCLP procedure is currently employed by
NUS Corporation, Gaithersburg, Maryland. Gail A. Hansen is
the current Project Officer at the U.S. EPA.

REFERENCES
1. Youden, W.J. and Steiner, E.H., Statistical Manual of the Associa-
tion of Official Analytical Chemists, AOAC, 1975.
2. "Inter-Industry Collaborative Study of the Toxicity Characteristic
Leaching Procedure, Addendum to Compilation of Phase IA and
Phase II Data," Docket Number F-86-TC-FFFFF, Sept. 15, 1986.
3. Mason, B.J. and Carlile, D.W., "Round-Robin Evaluation for Se-
lected Elements and Anionic Species from TCLP and EP Extrac-
tions," Pre-publication Version of EPRI Report No. EA-4740,
Draft Report, Apr. 25, 1986.
4. Williams, L.R., Francis, C.W., Maskarinec, M.P., Taylor, D.R. and
Rothman, N., "Single-Laboratory Evaluation of Mobility Procedure
for Solid Waste," EMSL, ORNL, S-Cubed, ENSECO.
5. Henry, B., "Evaluation of the ZHE TCLP Protocol," Final Re-
port, Contract No. 68-01-7075, ERCO/ENSECO, June 11, 1986.
6. "Background Document, RCRA, Hazardous and Solid Waste
Amendments of 1984, Land Disposal Restrictions Rule, Solvents and
Dioxins," U.S. EPA, Nov. 7, 1986.
7. Blackburn, W.B. and Show, I., "Collaborative Study of the Toxicity
Characteristics Leaching Procedure (TCLP)," Draft Final Report,
Contract No. 68-03-1958, S-Cubed, Nov. 1986.
8. Taylor, D.R. and Shurtleff, A.B., "Precision Evaluation of the
Toxicity Characteristic Leaching Procedure (TCLP) for Volatile
Contaminants," Final Report, Contract No. 68-01-7266, S-Cubed,
July 2, 1986.
9. Maskarinec, M.P. and Francis, C.W., "Precision Analyses for the
Zero-Headspace Extractor," Draft Interim Report, Contract No.
DE-AC05-840 R211400, Oak Ridge National Laboratory, Jan. 15,
1986.
10. Taylor, C.L., Blackburn, W.B. and Swanson, G.R., "Analytical
Data Report for POTW Sludge Testing," Contract No. 68-01-7266,
S-Cubed, Oct. 1986.
11. Walker, J., "Cooperative Testing of Municipal Sewerage Sludges by
the Toxicity Characteristics Leaching Procedure and Compositional
Analysis," Draft, Residuals Management Branch, U.S. EPA.
12. Walker, J., "Report on Six POTW Sludges Tested for Composi-
tional and TCLP Analysis," Memorandum, U.S. EPA, July 11,
1986.
MONITORING & SAMPLING 213

-------
                           Sampling  Strategies  for  Site  Evaluation
                                                     Jeffrey C. Myers
                                                       Rex  C. Bryan
                                                         Estox, Inc.
                                                    Golden, Colorado
INTRODUCTION
  Sampling for site assessment or remediation is a complex issue
combining many problems. To compound the situation, each site
is unique. This diversity demands detailed analysis to meet each
site's individual needs. Such efforts can consume time and money
very quickly unless they are carefully controlled.
  Also, the very nature of sampling is cause for concern. Very
small amounts of sample material are taken from areas or vol-
umes several orders of magnitude greater  than the sample. The
scientist must aspire to make each sample representative  of the
area in question. This generally is not an easy task. In addition,
proper QA/QC must be utilized so that poor procedure does not
invalidate a properly chosen, representative sample.
  To avoid the pitfalls of "sampling error," which take many
forms, the sampling strategy must be carefully outlined before-
hand. Questions such as objective, resources of time and money,
site knowledge and data adequacy must be addressed.
  Despite these significant problems, modern  tools and tech-
niques provide us with efficient and objective solutions to many
sampling problems. Field portable microcomputers simplify rou-
tine tasks such as data entry, data base management,  statistical
analysis, mapping, graphics and data communication. Geostatis-
tical techniques that incorporate economic parameters  maximize
the sampling dollar and yield objective, quantifiable results.


SAMPLE SYSTEM DESIGN
  Using  these resources,  the modern scientist  can address the
more specific questions involved in  sample design. Sample loca-
tions subject to time, monetary, geologic and cultural constraints,
generally are selected using some form of pattern. These patterns
take many  forms, including rectangular grids,  staggered grids,
random stratified grids, radial designs, random placement and
combinations thereof.  Each pattern  has  its  own  benefits and
flaws, depending upon the site characteristics.
GEOSTATISTICAL THEORY
  Geostatistical  theory  brings new opportunities to sampling
strategy with the analysis of precision, single as opposed to multi-
stage sampling and economics. Since nearby samples generally are
correlated, more information can  be extracted  from the same
number of samples. This spatial correlation is site-specific and is
quantified through the variogram,  thus allowing the scientist to
custom tailor his approach to a  given site. "Jack-knife" tech-
niques assure the use of a proper correlation model or variogram.
  Spatial correlation analysis is crucial to the determination of
optimal sampling density. As more samples are taken, the greater
is the precision of the analysis. The precision can be graphed in-
teractively by computer along with the sampling cost to determine
the optimum number of samples in a marginal benefit analysis.
This computer analysis also can be linked to the maximum toler-
able variance approach which balances the sampling costs, the ex-
pected cost of false difference (alpha times cleanup costs) and the
expected cost of false no-difference (beta times exposure costs).
  The misclassification ellipse is another powerful tool which
assists in making decisions. The ellipse, defined by the samples
and the estimation error, quantifies the amount of material above
a given threshold that will be left after remediation. The sensitiv-
ity to additional sampling and different threshold  can be easily
calculated.
KRIGING
  Kriging, a best linear unbiased estimator (BLUE), is an invalu-
able tool at both the precision and mapping stages. Many types of
kriging are available for specialized site characterization. Krigrng
can map not only the contaminant values but also the error asso-
ciated with the estimate. These quantities can be merged in the
max-max approach to help find sources.
  Using kriging to  follow the action line provides added benefits
to using the misclassification ellipse. By focusing on the threshold
value, the sampling value is maximized within the ellipse at the
most critical values and provides a greater return on the sampling
dollar. Within the action zone, exhaustion theory can place sta-
tistical confidence on the probability of missing zones above the
threshold.
CONCLUSIONS
  The combination of geostatistical and computer tools also re-
sults in additional benefits. Using these methods allows the scien-
tist to quickly and easily produce high quality maps and other
graphic displays. Good visual representation of analytical results
is essential at all stages of the project. Perhaps most important,
however, is the clear, concise display of final results to those who
are not  intimately involved in the project but whose decisions
will affect future action at the site.
  Sampling strategies for hazardous waste sites can no longer be
taken lightly. Serious consequences can result from poor or inade-
quate sampling programs. Fortunately, new tools and techniques
give the scientist  the power to make more accurate, objective, de-
fensible and  cost-effective  decisions in sampling analysis and
design.
 214    MONITORING & SAMPLING

-------
                     RCRA  and Its Implications for CERCLA

                                              Stephen M. Smith
                               Office of Emergency and Remedial Response
                                  U.S. Environmental Protection  Agency
                                              Washington,  D.C.
INTRODUCTION
 The following material is taken from a series of slides presented
at a conference seminar on the topic of RCRA/CERCLA rela-
tionships. The author, in his presentation, provided significant
detail for each of the following topical considerations. Inasmuch
as all sessions have been taped, those readers who wish to avail
themselves of the complete oral presentation may purchase the
tape either at the conference or through HMCRI: (301) 587-9390.

PURPOSE OF THE BRIEFING
D  Define applicable or relevant and appropriate requirements
  (ARARs)
s Review potential RCRA ARARs
* Describe applicability of  RCRA requirements  to  CERCLA
  actions
• Identify issues
APPLICABLE OR RELEVANT AND
APPROPRIATE REQUIREMENTS
• SARA §121 requires CERCLA cleanups to at least attain fed-
  eral or state ARARs
• Applicable: The  standard or requirement is designed to apply
  to problems identical to those encountered at the  CERCLA
  site
• Relevant and Appropriate: The standard or requirement is de-
  signed to apply  to problems sufficiently similar to those en-
  countered at the  CERCLA site that use of the standard makes
  good environmental sense
  -Land disposal restrictions
  -Corrective action
  -Closure
WHAT CERCLA ACTIONS
CONSTITUTE DISPOSAL?
• Disposal
  -Consolidation elsewhere*
  -Treat and replace
  -Consolidate in same area or unit*
• Not disposal
  -Cap waste in place
"•Including relocating spilled material
                        Figure 1
          What CERCLA Actions Constitute Disposal?
BASIC REQUIREMENTS FOR RCRA
APPLICABILITY
• Does site contain RCRA hazardous waste? and
• Did facility receive waste after Nov. 19, 1980; or
• Does remedy constitute treatment, storage or disposal?
POTENTIAL RCRA ARARs
• Location standards
• Special waste-specific requirements
• Action-specific standards for treatment, storage or disposal


RCRA ARARs FOR DISPOSAL
• Action-specific RCRA requirements affecting disposal
  •Minimum technology requirements
Disposal?
Nr>

                                          RCRA
                                        Dkpos.nl Rep
                                        Not Applicable
1
Minimum
Technology

I
Lund
Dis|ios:il

1 1
Corrective Q
Action n
Closure
                         Figure 2
                Applicable RCRA Requirements
                                                                                RCRA SITE REMEDIATION    215

-------
HOW MINIMUM TECHNOLOGY REQUIREMENTS
MIGHT APPLY TO CERCLA
• Double liner and leachate collection system applicable to:
  -New landfill or surface impoundment
  -Replacement or lateral expansion of landfill or surface im-
   poundment
• Applicable to CERCLA when remedy includes construction of
  new landfill or surface impoundment on-site
                                           Is it
                                        Disposal?
      Disposal?
             Yes
     Is It A New,
   Replacement or
   Expanding Unit?
             Yes
      Minimum
     Technology
       Applies
                          No
No
              RCRA
          • Disposal Regs
          Noi Applicable
 Minimum
Technology
    N/A
                                               Yes
                                           Is it
                                     Banned Waste?
                                                                         Yes
                                                                     Is it
                                                                  Placement?
                                               Yes
                                      Land Disposal
                                       Restrictions
                                          Apply
                                                No
    RCRA
 Disposal Regs
Not Applicable
                                                          Figure 5
                                                   Land Disposal Restrictions
                         Figure 3
              Minimum Technology Requirements
        Consolidate
        eltcwher*
            • Minimum Tech.
            Requirement!
                                                                                     Figure 6
                                                                            Applicable RCRA Requirements
                         Figure 4
                Applicable RCRA Requirements
HOW LAND DISPOSAL RESTRICTIONS
MIGHT APPLY TO CERCLA
• Land  disposal  restrictions  prohibit  placement  of  specified
  wastes in or on  the land without treatment (BOAT)
• Applicable to CERCLA when placement of banned waste oc-
  curs
• Statutory  exemption for two  years for soil  and debris from
  CERCLA response actions and RCRA corrective actions
  -Developing rule in next two years  for soil and debris
• Regulatory variance for solvents and dioxin
                                  HOW CORRECTIVE ACTION
                                  REQUIREMENTS MIGHT APPLY TO CERCLA
                                  • Pre-HSWA
                                    -Applies to units receiving waste after July 1982
                                    -Covers releases to groundwater
                                    -Prescribes groundwater protection and cleanup program
                                    -Sets groundwater protection standard
                                    -SARA sets special requirements for ACLs
                                  • HSWA
                                    -Applies to solid waste management units
                                    -Covers releases to all media
                                    -Regulatory requirements are being developed
                                  • Applicable to CERCLA  when treatment, storage or disposal
                                    occurs or occurred
216    RCRA SITE REMEDIATION

-------
Disposal?
No

RCRA
• Disposal Regs
Not Applicable
                  Was Waste
                Received After
                    7/82?
       Prc-HSWA
     Corrective Action
         Applies
        Are There
     Releases to Media
     Other Than GW?
                             No
    HSWA
Corrective Action
    Applies
    Yes
Disposal?
No

RCRA
• Disposal Regs
Not Applicable
                                                                                       1*'
                                                        How Do You
                                                           Plan to
                                                           Close?
Landfill Closure
Closure by
 Removal
                                                                   Figure 9
                                                                    Closure
                         Figure 7
                     Corrective Action
   • DDAT
    Treatment
    IW
   •HSWA
    Corrective
    Action
                         Figure 8
               Applicable RCRA Requirements
HOW CLOSURE REQUIREMENTS MIGHT
APPLY TO CERCLA
• Closure with waste in place (landfill closure) requires:
  -Stabilize waste
  -Cap
  -Post-closure care
• Closure by removal requires:
  -Remove or decontaminate all hazardous waste and constitu-
  ents, including groundwater
  -Drinkable leachate standard
  -No post-closure care
* Hybrid closure (draft proposed rule) requires:
  -Majority of contamination removed
  •Caps and post-closure monitoring based on site-specific con-
  ditions
* Applicable  to CERCLA when treatment, storage or disposal
  occurs or occurred
                                                                  Trent and
                                                                  replace

                                                                 • BOAT
                                                                  Treatment

                                                                  I9S7
                                                                 • Pfe-HSWA
                                                                  Corrective Action
                                                                                           • Closure wnite
                                                                                            in pbce
                                                                                                            Consolidate In
                                                                                                            snme area or
                                                                                                            unit
                                      I9S7
                                     1 Pre-HSWA
                                      Corrective
                                      Action

                                     • ClaMirc w.-isic
                                      ill plilVti
                                                                                                                         Cup
                                                                  Figure 10
                                                         Applicable RCRA Requirements



(
Minimum Ttfknalofy

MIA
Replace!*
EA pa Klin
r«t
Minimum
Technototy
Appliei'

New
em Or
1 Unit?
iMi

Minimum
Teehnotef,

LnulDOfaal
till
Donned
Waittt
1 la
till
PUcemenl?
1 to
Lind Ditntjul
Reilriclioni
Applv
Di.c<

~^TyL
Ytl
_
RCRA
Oil pout Rtft
Noi Applicable

Cerrttliv* Mian Cloiurt
— \N/A]
Wane Received I Mow Do You 1
After V82? 1 Han To Cteie? 1
r«l I MI 11
jj
r
Pre-MSWA
Corrcclive
Action Applies

1
Are Th
Release
Media C
ThanC

lere
s 10
WT
tSSri* ^ndfi" Cloiure Dy
Action Appliei Cloiure Removal
r-«

                                                                   Figure 11
                                                              Disposal Under RCRA
                                        ISSUES RAISED BY RCRA
                                        APPLICABILITY TO CERCLA
                                        • Inconsistencies between CERCLA response actions and RCRA
                                          regulations not designed specifically for  response actions (or
                                          RCRA corrective action)
                                          -Need for hybrid closure regulations
                                        • Consistency between current CERCLA response actions and
                                          RCRA corrective actions
                                          -Need for consistent remedies for similar problems
                                                                                           RCRA SITE REMEDIATION     217

-------
 Consistency between future  RCRA corrective action regula-       RCRA APPLICABILITY TO CERCLA
 tions and CERCLA selection of remedy process                   What Is being done to address Che Issues?
                                                              • Addressing integration of land  disposal restrictions in options
 -Need for consistent  approaches and  standards for similar         for selection of remedy
  activities                                                     • Participating in RCRA  corrective action work group
                                                              • Preparing manual on CERCLA compliance with other laws
  „                   ,     .        ...  „,,„„..           • Preparing guidance on implementing the land disposal restric-
1  Consideration of cost of remedy as required  by CERCLA             tjons un(jer CERCLA
                                                              • Preparing rulemaking on implementing the land disposal re-
  -Need to consider when implementing land disposal restric-         strictions for  soil,  debris and other complex  wastes from
  tions and when developing corrective action regulations              CERCLA sites
218    RCRA SITE REMEDIATION

-------
                    Development  of  Engineering Design  Strategies
                            For Remediation  and  Retrofitting of
                                   Existing Waste  Disposal Sites
                                                  Thomas J. O'Brien
                                               Benjamin G.  Siebecker
                                         Wehran Engineering Corporation
                                               Middletown, New York
INTRODUCTION
  The determination of cost-effective engineering design strate-
gies for retrofitting, remediation and the design of state-of-the-
art disposal facilities is predicated on understanding the site's
hydrogeologic setting. The purpose of our seminar is two-fold.
First, the methods, techniques and data analyses for the proper
identification of the hydrogeologic condition of waste  disposal
sites will be previewed. Second, engineering theory, concepts and
remediation/retrofitting engineering design strategies will be ex-
plored.
  The seminar will emphasize the  dynamic  interrelationship of
hydrogeology  and engineering  design strategies by using slides
as visual aids. The seminar will culminate with a discussion and
slide show of an actual case history.

HYDROGEOLOGIC STUDIES FOR
DISPOSAL FACILITIES
  The scope of hydrogeologic studies for disposal facilities is
largely dependent upon whether the proposed facility will be a
new site, the expansion of an existing  site or the remediation or
retrofitting of an operating or closed facility.
  For new disposal facilities, study focuses on the hydrogeologic
setting and  the defining of the "virgin" hydrogeologic condi-
tions. It is essential to indicate anticipated ranges for the hydro-
geologic parameters so that conservative engineering assumptions
can be utilized for the successful state-of-the-art design.
  For an expansion of disposal facilities, remediation or retro-
fitting of an existing site, much data already may be available
on the site-specific hydrogeologic setting. In this case, the study
likely will focus on the performance of the existing facility. In-
itially, groundwater quality data would be reviewed for existing
monitor wells, and recommendations would be made regarding
the  effectiveness  of  the  monitoring  program.  Typically,  the
hydrogeologic setting is altered by the placement of waste, and a
groundwater mound has been created. The determination of the
altered groundwater flow regime and the evaluation of the exist-
ing leachate collection  system if present,  would be of  primary
concern.

Background Information
  All hydrogeologic investigations should first review the avail-
able references in the literature and local area information, if
available. These  "paper studies"  do not  eliminate the need for
site-specific information; however, they do provide the basis for
the regional impact from a proposed or existing facility.

Hydrogeologic Field Studies
  Hydrogeologic field studies should concentrate first on  relative-
ly inexpensive field techniques and gradually develop into full-
scale boring programs. Typically, a newly prepared base  topo-
graphic plan is requested. It is vital for the success of the project
to have a good quality base plan, especially at waste disposal sites
where features and elevations change constantly.
  As mentioned, site-specific exploration is essential for the de-
sign of disposal facilities. Field investigations typically employ the
following studies or combination of studies:
  (1) Test pit investigation
  (2) Geophysical programs
  (3) Piezometer/monitoring  well  installations/surface   water
     staff gauge installations
  (4) Well development
  (5) Permeability testing

Hydrogeologic Report
  A hydrogeologic report should be prepared that is sufficiently
detailed to provide the basis for engineering design strategies. The
report should include:

  (1) A description of the site and its physical setting
  (2) A discussion of the objectives and procedures of the hydro-
     geologic investigation
  (3) A description of the regional hydrogeologic setting
  (4) A detailed discussion of the site geology as determined by
     the  investigation, including soil and bedrock types and
     depths and the continuity of geologic units
  (5) A detailed description of the  site-specific hydrogeologic
     conditions, including: the results of the in situ permeabil-
     ity testing; the presence and significance of aquifers/aqui-
     tards; groundwater flow directions and velocities (both hor-
     izontal and vertical); and groundwater quality, including
     the probable extent of the contamination plume, if present
  (6) A discussion of the significant conclusions reached during
     the course of the investigation; in particular, a preliminary
     assessment of potential remedial or  mitigative measures
     would be provided, if warranted, by the presence of signif-
     icant contamination as would recommendations for further
     investigation, if necessary.

  In support of the text, a number of maps, cross-sections and
tables should be prepared. These graphic aids generally would in-
clude a location plan, a detailed site map, a generalized geologic
column, generalized  geologic cross-sections based on exploratory
test pits, borings and existing wells and a water table contour
map. Results of other data, including permeability testing,  would
be presented in tabular form as would the groundwater quality
data.
                                                                                        RCRA SITE REMEDIATION    219

-------
ENGINEERING METHODS AND
TECHNOLOGIES FOR SITE REMEDIATION
  The answers an engineer requires from the hydrogeologic in-
vestigation are generally two-fold. First,  the identification, ex-
tent and severity of the groundwater contamination problem for
a given site are essential. Second, the hydrogeologic conditions,
both beneath and  around the  site, are critical from the stand-
point of engineering design alternative. The identification of the
groundwater contamination problem will define the current and
impending impacts on surrounding environs, identify  potential
receptors of contamination and most likely dictate the urgency
for remediation. The hydrogeologic site conditions will help iden-
tify possible remedial options geared to the solutions of the con-
tamination problem. The following discussion assumes that the
first question has been answered and a certain priority for remed-
iation has been established.

Leachate Generation
  Leachate is  generated in three ways: (1) percolation of pre-
cipitation into the waste whereby it becomes contaminated water;
(2)  surface flow or runoff; and  (3) groundwater  entering the
waste and subsequently becoming contaminated. Once the leach-
ate has been generated, it leaves the landfill as contaminated
groundwater or as leachate seepage, thus polluting surface waters
and  groundwaters. The most common scenario encountered is
leachate  generated by precipitation percolating into the waste
with off-site leachate migration via the groundwater system.
  Leachate generation by infiltrating precipitation  is computed
by models. Briefly, one subtracts runoff and evapotranspiration
from the total amount of precipitation, leaving a resultant value
of percolation. Other factors  considered  include temperature,
depth of soil, type of soil cover, topography and vegetation. Two
common water balance methods are: (1)  an adaptation of Thorn-
waites model  by Fenn, et al.\ and (2) the  HELP (Hydraulic
Evaluation of Landfill  Performance) model developed by the
Army Corps of Engineers. The Thornwaite/Fenn model is useful
for predicting long-term average leachate generation, while the
HELP model is good  for evaluating past conditions based upon
actual meteorological data.
  Leachate migration in the groundwater is  best quantified by
examining the hydrogeologic variables of  permeability, gradient
and  aerial  extent.  Unless  other substantial leachate generation
and migration pathways exist, the groundwater flows and perco-
lation quantities should balance. This quantification  of flow is an
important tool for evaluating leachate conditions at a site and de-
termining appropriate remedial actions and potential costs.
Remedial Actions
  The single, most effective, remedial action to a waste site is to
eliminate precipitation from percolating through the waste. Con-
trol of percolation is accomplished by the placement of an "im-
pervious cap." This prevents infiltration of precipitation into the
waste. Typical landfill  caps  are: PVC or HOPE flexible mem-
brane liners;  or a thick layer of clay or some composite of the
"synthetic membrane"  and "natural" cap. It is desirable for cap
material to have a minimum of 1   x  10 -'  cm/sec permeability
(i.e.,  "impermeable"). Should site water balances  reveal  that
contaminated groundwater must be further controlled because
the site will remain active, leachate is being generated by ground-
water inflow  or the contamination levels are very hazardous, a
cutoff barrier around the site should be considered. A cutoff bar-
rier usually not only contains  leachate, but also excludes clean
off-site  groundwater from entering the leachate collection sys-
tem. This collection system normally is installed between the cut-
off barrier and the waste, below the groundwater table. Cutoff
walls usually  are constructed of bentonite clay slurry and select
backfill, cement bentonite or compacted native clay. Cutoff bar-
riers can be utilized only when appropriate hydrogeologic con-
ditions are defined.

Design of New Waste Disposal Sites
  New waste disposal sites are designed to prevent groundwater
contamination from the onset. This is accomplished by using a
base liner and leachate collection underdrain system. Base liner
systems vary, and selecting the appropriate site liner may have lit-
tle to do with liner competence and cost-effectiveness, but maybe
primarily a political decision based solely on regulations.
  Additionally, since leachate will be collected,  treatment avail-
ability and costs must be considered. This potential treatment cost
adds emphasis to minimizing leachate generation from infiltrat-
ing precipitation, and so forth. Strict operational control of waste
placement and prompt  final covering with a cap as soon as die
waste disposal area has  reached final grade, can serve to substan-
tially reduce leachate generation.

CONCLUSIONS
  Siting a new landfill or expanding an existing one is a difficult
process. One thing that must be recognized is that  finding the
perfect landfill site may be impossible; therefore all new landfills
or retrofits must identify the hydrogeologic system and be engi-
neered to collect  leachate (i.e., "not leak").  This groundwater
and surface water degradation do not occur and state-of-the-art
waste disposal is realized.
 220    RCRA SITE REMEDIATION

-------
RCRA Corrective Action at a Large Chemical
Manufacturing Facility; Process and Results

George A. Furst, Ph.D.
U.S. Environmental Protection Agency
Boston, Massachusetts
ABSTRACT
Corrective action addressing all releases from solid waste man-
agement units (SWMUs) to the environment from a facility is a
major new requirement of RCRA, as amended by the Hazardous
and Solid Waste Amendments of 1984 (HSWA). This process can
be initiated either through administrative orders or through per-
mits issued by the U.S. EPA or the authorized states. Presently, a
guidance document is being prepared by U.S. EPA headquarters
for use by facilities conducting investigations of releases from
SWMUs. The investigation described in this paper is being used as a
case study in the above U.S. EPA guidance document to illustrate
this process, called the RCRA Facility Investigation (RFI).
This investigation was initiated in the summer of 1981 by the
owner/operator at the largest chemical manufacturing facility in
New England. In the fall of 1983 the U.S. EPA issued a unilateral
order to the facility under §3013 of RCRA; the order was subse-
quently modified to a CERCLA §106 consent order.
Simultaneously, the state issued an equivalent consent order to
the facility under the Massachusetts General Law 21 E (the State
Superfund Law). The final report, including corrective action
recommendations, is expected to be complete in the spring of
1987.
A phased approach incorporating distinct review steps is a
hallmark of this investigation. At the end of each phase, time is
allocated to review the project direction and to ensure that no
critical environmental media have been overlooked. Project
momentum is maintained by involving all parties in field deci-
sions.
During the investigation many methods have been used to
characterize the site geology, hydrogeology and contaminant pro-
files in groundwater, surface water and soil. A brief listing of this
process is presented with attention given to those techniques that
have been innovative or particularly useful in defining locations
of old liquid waste lagoons, buried containerized waste and con-
taminant groundwater plumes.
In the process of the investigation there have been continuous
meetings between the facility and the regulatory agencies. Con-
tractors have been used effectively by both the U.S. EPA and the
facility to supply technical expertise and to conduct field in-
vestigation activities. Specific questions that arose and were
resolved involved: the use of drilling muds during installation of
monitoring wells, techniques for field screening for volatiles,
location of test pits for containerized waste, media to be sampled
and the use of oversight sampling by the regulatory agencies.
Clear project goals and a commitment to the RFI by the facility
have helped keep the project on schedule. Cooperation between
the Massachusetts DEQE and the U.S. EPA have allowed effi-
cient utilization of respective resources.
The next step in the process will be the issuance of a Corrective
Action Permit containing the proposed plans for remediation of
the facility. This HSWA permit, in conjunction with the facility
Part B permit application, if approved, would constitute one of
the first full operating permits at a major chemical company to be
issued under RCRA in Region 1, U.S. EPA since the enactment
of HSWA.

INTRODUCTION
The purposes of this paper are to present the RCRA Corrective
Action strategy process as it has evolved to date and to illustrate
the RFI stage of this process by utilizing a remedial investigation
that was initiated 3 yr ago under CERCLA authorities. Included
in the example of the RFI are lessons learned by the author work-
ing first as a contractor for the U.S. EPA and more recently as the
Region 1, U.S. EPA RCRA Facility Manager for the project.

THE CORRECTIVE ACTION PROCESS
The Hazardous and Solid Waste Amendments (HSWA) of
1984 greatly expanded the authorities of RCRA by requiring cor-
rective action for releases of hazardous wastes and constituents to
all media (i.e., groundwater, surface water, air and soil) at
facilities that treat, store or dispose of hazardous waste. Section
3004(u) of RCRA requires corrective action for releases from any
Solid Waste Management Unit (SWMU) at a storage, treatment
or disposal facility that is seeking or is otherwise subject to an
RCRA permit. Section 3008(h) of HSWA authorizes the U.S.
EPA to require corrective action or other necessary response
measures when information indicates that there is or has been a
release of hazardous waste or constituents from facilities that
have had, or should have had, interim status.'
These recently enacted RCRA corrective action authorities are
broad, and the universe of RCRA facilities to which they poten-
tially apply is diverse. Furthermore, corrective action re-
quirements apply to these facilities regardless of whether they are
continuing waste management operations or closing these opera-
tions.2 Current estimates indicate that nationally there may be as
many as 5,000 treatment, storage and disposal (TSD) facilities
that will require corrective action as a condition of permit is-
suance.
The stages of the RCRA corrective action process have
equivalents in CERCLA which are listed in Table 1. How closely
these programs will coincide is still under discussion within the
U.S. EPA. A primary objective of the U.S. EPA program will be
to maintain consistency with CERCLA and simultaneously meet
the specific objectives of RCRA, a program that has developed
through policy, guidance and regulation.3 The first step in the
RCRA corrective action process is the RCRA Facility Assessment
(RFA)4 which consists of an examination of information on waste
RCRA SITE REMEDIATION 221

-------
management practices at the facility, a review of past releases of
hazardous waste that may have occurred and the identification of
SWMUs with actual or potential releases. The RFA process also
includes a site visit to identify problems, and media sampling to
confirm, the absence or presence of releases from  the  above
SWMUs. The highest priority facilities  for initial RFAs  under
§3004(u) are land  disposal facilities  that have demonstrated
significant environmental problems.

                           Table 1
         RCRA Corrective Action Activities and Equivalent
          CERCLA Pre-remedlal and Remedial Activities
RCRA
                                  CERCLA
Facility Screening Process
RCRA Facility Assessment (RFA)
    (screening phase)
RCRA Facility Investigation (RFI)
Media Protection Standards
  (MPS)
Corrective Measures Study (CMS)
Permit—Proposed Permit
  Modification (PPM)
Order—Amendment of the Order
  (AO)
Corrective Measures Implementation
  (CMI)
Preliminary Assessment/Site
  Investigation (PA/SI)
    (pre-rcmedial phase)
Remedial Investigation (Rl)
Endangcrmcm Assessment
  (EA)
Feasibility Study (FS)

Record of Decision (ROD)

Record of Decision (ROD)

Remedial Action (RA)
  If a release has occurred or if there is a high probability of a
release at the facility, the company will be required to conduct an
RCRA Facility Investigation (RFI)' This process will either be in-
itiated by issuance of an enforcement order under authorities such
as RCRA §3008(h), §7003, §3013, CERCLA §106 or by HSWA
special permit conditions.'
  The purpose of the RFI is to determine the nature, extent and
rate of migration of contaminant releases occurring at the facility
and is similar  in  scope to the Remedial  Investigation in the
CERCLA program. The RFI must also provide sufficient data to
determine appropriate corrective measures or to document that
no  action is necessary. While the RFA generally is conducted by
the U.S. EPA, the RFI is  performed by the facility owner/op-
erator. As stated above, the RFI is initiated either by an order or
permit, and the U.S. EPA  oversees this activity. This paper will
focus on the RFI stage of the corrective action process.
  After completion of the RFI, the Media Protection Standards
(MPS) are assigned to each hazardous constituent which has been
released at the facility. Once these facility proposed concentration
limits are approved by the U.S. EPA, the MPS serves as a correc-
tive action goal for the facility to achieve by corrective measures.
The next stage  of  corrective action, the Corrective  Measures
Study (CMS), is performed by the owner/operator and consists of
a selection  of a number of appropriate corrective measures and
their submittal to the U.S. EPA for review. Also, at this stage, the
public has  an opportunity  to review and comment on the pro-
posed cleanup options and selection of final measures.  The U.S.
EPA will evaluate the owner/operator recommendations and ap-
prove or disapprove them. Public comments will  be considered
when making these decisions.
  After the remedy is selected by the U.S. EPA, the owner/op-
erator will take the appropriate measures needed to implement,
maintain and monitor the remedy chosen. This final Corrective
Measures Implementation (CMI) step is anticipated to be required
by  a Proposed  Permit Modification or an Amendment of the
Order  with  oversight  by the U.S. EPA.' Currently  one state,
Georgia, is authorized to operate the HSWA process and it is an-
ticipated that a limited number of states soon will be authorized
to operate the Corrective Action program described.

RFI CASE STUDY
  The facility conducting  this RCRA Facility Investigation  is
located  in Central Massachusetts. Fig. 1 shows an aerial view of
the facility. The 328-ac site is located in a river valley; it is bound-
ed by a major river on the northern boundary and a 100-ft escarp-
ment  along the  southern  boundary.  For over 80 yr, chemical
manufacturing companies have been located at the site; presently,
it is occupied by the largest chemical manufacturing facility in the
Region. A  prior company manufacturing cellulose shirt collars,
brushes, combs, piano keys  and nitrocellulose movie  film oc-
cupied the site from 1904 to 1938. At that time, the present owner
bought the site. As this company expanded in the 1940s, the com-
bined plants at this  location became one of the largest and most
diversified polymer  manufacturing sites in the United States.
  Over  the  years,  various  landfills  were utilized  to  dispose of
demolition  debris and  hazardous waste. Depressions  were filled
with liquids of off-specification products,  while burning pits were
utilized  to incinerate combustible materials on the western por-
tion of the property. The locations  of the facility buildings and
waste disposal areas (SWMUs) including dates of use are shown in
Fig. 1. Plant buildings were constructed on top of old landfills,
and from time to time  there were underground fires in this area.
  Initial studies, begun in 1981 to assess the extent of contamina-
tion, focused on groundwater associated with the waste disposal
areas. The  limited number of monitoring wells installed in 1982
showed  groundwater contamination by chlorobenzene, benzene,
toluene and xylene. In 1983, the U.S. EPA issued an RCRA §3013
Unilateral Order, and  the Massachusetts  Department  of En-
vironmental Quality Engineering (DEQE)  issued a 21E  Order
under a newly approved State Superfund authority. The Orders
required the submittal  of a Remedial Investigation Plan by the
facility.  In  the spring of 1984, the U.S. EPA's Unilateral Order
was amended to a CERCLA §106 Consent Order. At that time the
enforcement of the  order was overseen by U.S. EPA CERCLA
personnel because, prior to HSWA passage on Nov. 8, 1984, U.S.
EPA  RCRA personnel were oriented  toward compliance rather
than remediation.
  Meetings with the facility were held by the DEQE and the U.S.
EPA  during 1983  and  1984  to review the  draft  Remedial In-
vestigation  Plans (RIP) submitted by the facility. These meetings
were crucial as they set the tone for oversight of the project by the
regulatory agencies and initiated the methods of communication
utilized  throughout the project. In  order to ensure a common
understanding regarding areas of agreement and disagreement,
extensive notes were taken at these meetings and distributed, with
highlighted decisions,  to all  involved parties. For example, at
these  meetings it was agreed that a limited number of field deci-
sions, field chemical screening and QA/QC of data would be in-
itiated early in  the investigation.  The RIP was approved in
February 1984.
  The final plan was organized using a phased approach. Each
phase was a building block, based upon the results of previous ac-
tivities (i.e., each phase enlarged the information on geology, geo-
hydrology, geophysical characteristics and  the extent  of con-
tamination at the facility). The final  plan was as detailed as possi-
ble without being overly rigid. As information was gathered dur-
ing each phase, the locations of future monitoring wells and geo-
physical studies were modified. These field decisions, which in-
volved on-site  meetings of the company,  the U.S. EPA, DEQE
and/or their respective contractors, were a critical component of
the project, allowing it to move on schedule by expediting the col-
222     RCRA SITE REMEDIATION

-------
/ft// DRUMS FOUND IN AERIAL
PHOTOGRAPHS.
SCALE IN FEET
PREPARED BY
C.C.JOHNSON a MALHOTRA, P.C.

SOURCE i BLASLAND a BOUCK ENGINEERS, PC.
NOV. 1983
Figure 1
Facility Plan Location of SWMUs
lection of information. This phased approach also has been
recommended for the CERCLA RI/FS stage.8 Table 2 shows the
schedule of the project up to the recent submittal of the final RFI
report.

Phase I of Investigation
The Phase I tasks consisted of the collection and review of ex-
isting information regarding prior waste disposal practices. This
process included a review of all relevant company documents in-
cluding photographs of waste disposal practices and aerial photo-
graphs dating back to the early 1930s. All regional geohydrologic
information including drilling logs, surveys of groundwater users
in the area, site-specific geologic documents and information on
the tributary brook and Chicopee River also were reviewed. A site
visit, attended by all personnel involved, was conducted early in
the study. This site review process was important because it al-
lowed discussions about the scope of the project and the timing of
exploratory tasks and sampling events prior to the initiation of
field work. The visit was also important because it brought into
perspective distances and evidence of past land use and provided a
common reference for future discussions.
A 50-ft grid, encompassing the study area, was constructed us-
ing labelled stakes driven at grid intersections. This grid provided
a reference for locating seismic lines, magnetometry lines and
sampling locations. The grid also helped reference sampling loca-
tions over large open areas and permitted data to be computerized
for data manipulation."A resurvey of the site was completed at a
1-ft contour interval. This survey was compared with the original
1938 survey and provided extremely useful information about the
thickness of waste piles and extent of excavation and also deter-
mined undisturbed areas. This phase ended with an extensive
report summarizing the information collected to that point.
A meeting was held 2 wk after the report was issued and pro-
vided a forum for review and Phase II planning. Agreement on
the location of test borings for the next phase was accomplished
and areas requiring additional effort were identified. Again, ex-
tensive meeting notes were taken and distributed to all involved in
the study.

Phase II of Investigation
Phase II work began immediately following the Phase I review
meeting and in part consisted of field work involving 33 con-
tinuous split spoon borings with refusal at 10 ft into till. Field
chemical screening as described below was initiated in this stage.
Areas where the magnetometry scans in Phase I indicated high
concentrations of metal were scanned using earth penetrating
radar, a very useful method which indicated whether buried con-
tainerized waste or demolition debris was located in the waste
areas.

Field Screening During Investigation
During this second phase of the investigation, field screening
techniques were emphasized. This early use of field analysis
methods, as described below, allowed development of a more ac-
curate conceptual site model early in the project. This is not to say
that more rigorous analyses were not conducted at this stage. For
RCRA SITE REMEDIATION 223

-------
Table 2
Schedule of Completed Tasks of Remedial Investigation

Pre-Remedial Investigation Work
Initial Investigations 1981
Installation of Monitoring Wells MW-1 to
MW-12 July-Sept. 1982
Sampling of Monitoring Wells Fall 1983
Meetings with DEQE & EPA Summer 1983
DEQE & EPA Orders Fall 1983
Remedial Investigation Plan-Approval Feb. 1984

RI Phase I
Trip to O'Brian and Gere Analytical Lab April 1984
Walk Over of Site May 1984
Seismic Refraction May 1984
Magnetrometry (50 ft) May 1984
Sampling of Monitoring Wells May-June 1984
Phase I Report Meeting Aug. 1984

RI Phase II
Soil Boring with OVA & HN Screening Aug.-Nov. 1984
Earth Penetrating Radar Nov.-Dec. 1984
Installation of Monitoring Wells Dec. '84-May '85
Monitoring Wells were gamma logged April-May 1985
Sampling of Monitoring Wells May-July 1985
Phase II Meeting & Report Aug. 1985
Meeting to discuss Hydrogeology and further
project work Feb. 1986
Test Pitting Proposal June 1986
Sediment & Sampling Proposal July 1986
Test Pitting & Magnetrometry (10 ft) July 1986
Sediment & Surface Water Sampling Aug. 1986
Phase III Report & Meeting Nov. 1986
example, 30 full priority pollutant analyses of groundwater and
soil using stringent QA/QC standards were completed during this
stage of the project whereas over 900 samples were analyzed for
volatiles using the following screening technique. The samples
taken daily during exploratory split spoon soil borings were
sampled in a 40 ml VOA vial with teflon septum. Later in the day,
in the company laboratory, the samples were placed in a 100 F
water bath for at least an hour, and a headspace aliquant was in-
jected into a gas chromatograph to determine total volatiles in the
sample, if the volatile content was greater than 2 ppm, a chroma-
togram was made and compared with site-specific standards. Us-
ing this information and the geologic information, a site cross-
section, including both geological (glacial) units encountered and
the location of contamination could be plotted at peaks (with
numbers in ppm on diagram) associated with each soil boring as
shown in Fig. 2. This information was utilized later as a guide for
setting screen depths in the monitoring wells.
Screening results helped determine the width and thickness
(breadth) of contaminant releases emanating from the different
solid waste management units. Also, the low detection limit of the
technique for the soil allowed determination of the contaminant
plume edge. This technique was especially useful since early
groundwater and soil sampling data indicated that volatile con-
tamination in the groundwater was the main contamination en-
countered at the facility. Surface soil samples from different site
locations also were screened using the above headspace analysis
approach.

Installation of Monitoring Wells
The next stage of the investigation was the installation of 52
monitoring wells. This network consisted of one four-well, two
three-well and 19 two-well nested sets, each set used to determine
A'
/£.
~r-—>-"
LEGEND

,T UOUMmtrt* CUV
PREPARED BY
C.C. JOHNSON • MM.HOTM* , P.C.
•OUNCC- M.AH.AND • MUCK tMMCCRS.
NOV. l»«
Figure 2
Cross Section A-A'
Soil Borings Showing Volatile Screening Results
224 RCRA SITE REMEDIATION

-------
© GROUNDWATER MONITORING
WELL
atf SOLID WASTE
m^ MANAGEMENT UNITS SWMOl

//// DRUMS FOUND IN AERIAL
PHOTOGRAPHS.
SCALE X FEET
PREPARED BY
C.C. JOHNSON 8 MALHOTRA, P.C.

SOURCE i BLASLAND B BOUCK ENGINEERS, RC.
NOV. I9B5
Figure 3
Location of Groundwater Monitoring Wells
the vertical gradient at specific locations. The location of these
wells, and their relationship to SWMUs, is shown in Fig. 3.
The original proposal also included three bedrock wells. After
the installation of the first bedrock well, the water level was ex-
amined and found to be 14 ft higher than that in the adjacent well
screened immediately above the till-sand interface. Additionally,
the well had passed through 90 ft of lodgement till. This was a
case where the data guided us to a technical decision. Both the
strong vertical gradient and the thickness of till indicated that
there was little likelihood of contamination extending to bedrock.
Analytical data collected from the bedrock well supported this
contention.
A complete sampling by the facility of all 57 monitoring wells
was the final field task prior to submission of the Phase II interim
report in August 1985. The U.S. EPA also sampled 23 of the
monitoring wells to confirm the analytical results. The final
meeting was held in August after reviewing the draft Phase II in-
terim report. As this meeting, the facility and their consultant
presented the data collected to date with emphasis on the site hy-
drogeology, extent of contamination and results of the earth pen-
etrating radar survey of the magnetic anomaly locations.
The decisions made at this meeting are an example of making
focused decisions based on the information available at this stage
of the project. It was agreed that there would be sediment and
surface water sampling of locations where vertical flow gradients
and hydrologic flow lines indicated that contaminated ground-
water may be impacting these media. A formal request was to be
submitted by the U.S. EPA and DEQE with the understanding
that the facility would submit a work plan for this task. This sub-
mission of formal requests by the regulatory agencies and work
plans by the facility indicated that a change in the project had oc-
curred (evolved) as the work tasks in the Remedial Investigation
Plan were now less well defined than previous tasks.
In order to ensure agreement on the appropriate level of effort
and goals for the remaining tasks, it was agreed that all tasks not
listed in the RIP would be initiated by submission of work plans
by the facility for review by regulatory agencies. Quick review of
the formal requests by the facility and their consultants and
review of the work plans by the agency were critical at this stage.
The work plan included a time frame when the task would be ac-
complished. An example of this process is the surface water and
sediment sampling task and the confirmatory test pitting task as
described below.
The results of the magnetrometry and earth penetrating radar
survey indicated a high probability of buried containerized waste.
Re-examination of aerial photographs indicated stacked drums at
Solid Waste Disposal Area 1 (SWDA #1) on Fig. 1. Since DEQE
and the U.S. EPA were requesting the test pitting, the regulatory
agencies submitted to the facility a formal request for this task.
This meeting was also an opportunity to discuss in more depth
the remedial actions applicable to the site and what information
still must be gathered to support the various remedial options.
The final part of Phase II concentrated on completing the
focused tasks which were necessary to complete the investigation.
The summer of 1986 was spent conducting the test pitting and
sediment and surface water sampling. The final report will sum-
RCRA SITE REMEDIATION 225

-------
marize all information gathered, concentrating on the most recent
findings. This meeting will explore in detail the remedial options,
media protection standards, and all supporting data applicable to
remediate the facility.

Future Direction of Project
  The  final phase of the RF1,  Phase 111, will concentrate on the
remedial options for the study area. This will include the prepara-
tion of a corrective measures study and the setting of media pro-
tection standards to be met by corrective measures implementa-
tion.
  Related to the overall  goal of  the  facility to receive a  full
operating permit, the facility submitted the Part B application in
August of 1986. Massachusetts DEQE will review this permit for
completeness during 1987. Since the 1984 HSWA amendments re-
quire that a  federal  HSWA permit (Corrective Action  Permit)
also be issued to the facility to receive a full operating permit, the
work done to date will be integrated into special permit conditions
of the  HSWA permit. The study will be expanded to cover  other
areas of  the facility where hazardous waste presently is managed
and there has been or is a high probability of a release of hazar-
dous waste. These studies will be  integrated  into the above in-
vestigation and written  into the  special permit  conditions  as
discussed. When the decision on remediation is made by the  agen-
cy, with  public review, a Proposed Permit Modification will be
issued. This is followed by the final stage of corrective action, the
Corrective Measures  Implementation by the facility.

GENERAL COMMENTS ON OVERSIGHT
OF RFI  PROCESS
  Below are  listed several general considerations that have been
learned over the last 3 yr of overseeing this project and from other
projects  with which  the author  has been involved either  in an
oversight role or as the contractor with U.S. EPA oversight.
• One can order the facility to do  the investigation in a specific
  way that  may be  contrary  to their  technical  judgment, but
  there are consequences to this  approach. Sometimes it may be
  necessary to be unyielding but that posture has the potential to
  slow down the project in many ways. One critical way is that
  it may cause  communications to break down. Another conse-
  quence is that the project schedule may be adversely affected.
  An example of this is the drilling mud issue that is discussed by
  Gagnon.' This disagreement affected the project schedule.
• One should treat the facility personnel with dignity. When the
  majority of the dumping of wastes and waste burial occurred,
  it was not  known  to have an adverse impact on the  environ-
  ment and often no state or federal law was being broken. Treat-
  ing the facility personnel as adversaries can force them into
  that role. They,  too, want a clean environment, but  financial
  constraints mean they often need information to provide a
  justification to act. Clearly stated requests and a good working
  relationship help in this process.
• One should maintain constant communication. This sugges-
  tion cannot be over-emphasized. Often there are many parties
  involved in the RFI process. For example, the above example
  involved the facility, the facility's contractors, the state agency
  (DEQE), the U.S. EPA and the contractors for the U.S. EPA.
  Also, lawyers become involved during certain stages and  there
  are times when the public  is involved, too. These investiga-
  tions are complex and even when all parties are aware of the
  latest findings, conflicts may occur; through communications,
  the basis of consensus is present. In the above project,  close
  communication was consciously  made a high priority at an
  early stage. The notes of critical meetings were shared with all
  personnel.  This communication  process helped to minimize
  disagreements later as to what was agreed upon at the meet-
  ings, and it also built up a sense of trust.
• One should let the data guide the project direction. Often argu-
  ments  at  meetings occur due  to differences  in perception of
  what  the data  to be collected will indicate. Sometimes it is
  worth  waiting until the data are available to make a decision
  on the next step. This is a distinct advantage of the phased ap-
  proach as described above. As the base of information in-
  creases, the investigation can focus on those areas which the
  data indicate have been environmentally impacted and there-
  for must  be addressed in the next phase.  Time is not spent
  arguing over areas where data later indicate there has been no
  environmental impact. The advantage of this process is that
  the decisions are made based upon the available data.

CONCLUSIONS
  The RCRA Corrective Action  process will affect a large num-
ber of interim status facilities that are either closing or applying
for Part B permits. This paper focused on one portion of this pro-
cess, the RCRA Facility  Investigation. This example utilized the
phased approach for gathering information on rate and extent of
contamination at a large chemical manufacturing facility. The in-
formation gathered will be incorporated by the U.S.  EPA into the
HSWA portion of the Facility Operating Permit and will be com-
pleted through modifications to this permit."
  Corrective Action is a complex process that requires continual
communication as well as good  technical judgment.  It is an-
ticipated  that this RCRA program will evolve into a process that
will lead to safe handling of hazardous waste and a clean environ-
ment. Related to this goal is a need for early screening of all media
to define those media that must  be focused upon during the in-
vestigation so that  the U.S. EPA, the states and  industry can
work together for a cleaner environment.

DISCLAIMER
  The Corrective  Action approach described above is not final-
ized and  is based  upon current information  available to the
author.  The opinions or assertions  contained herein  are the
writer's and are not to be construed  as official or reflecting the
views of the U.S.  EPA.

ACKNOWLEDGEMENTS
  The author would like to express his appreciation to Patricia
Hynes. Her  inspiration still guides the project. Special thanks go
to the U.S.  EPA Region 1 Corrective Action Workgroup led by
Mary Jane O'Donnell and the following individuals  for their sug-
gestions,   conversations  and  guidance:  Gem  Falco,  Richard
Leighton, Craig Johnston, Gwen Porus and John Zannos. Figure
preparation  by Blasland and Bouch Engineers and C.C. Johnson
are  hereby acknowledged.
REFERENCES
 I. U.S. EPA,  "Nov. 8, 1984 HSWA Amendments", Final Codifica-
   tion Rule, Federal Register. 50, July 15, 1985. 28,702.
 2. Porter, J.W., "Development Plan for Technical and Procedural
   Regulations for Corrective Action at  Solid Waste Management
   Units (SWMUs)," Office of Solid Waste and Emergency Response,
   U.S. EPA memorandum, Oct. 23, 1986.
 3. Pagan, D., "Options Paper: Development of a Regulatory Program
   for Corrective Action under Section 3004(u)", U.S.  EPA Office of
   Solid Waste and Emergency Response,  Nov. 19, 1986.
 4. U.S. EPA,  "RCRA Facility Assessment Guidance," (Final), Sept.
   1986, Permits and State Program Division, Office of Solid Waste,
   Sept.  1986.
226     RCRA SITE REMEDIATION

-------
5. U.S. EPA, "RCRA Facility Investigation Guidance" (Draft), Waste         8. U.S. EPA, "Phased  RI/FS Concept Paper." Prepared for U.S.
  Management Division, Office of Solid Waste, Oct. 1986.                    EPA by Camp Dresser & McKee, Inc., June 17, 1986.
6. U.S. EPA,  "RCRA Corrective Action Process" (Draft), prepared         9. Gagnon, J.,  "The Remedial Investigation Process: An Industry
  by Region 1 Corrective Action Workgroup, Oct. 22, 1986.                   Prospective," Proc. National Conference on Hazardous Waste and
1. Porter, J.W., "National RCRA Corrective Action Strategy", Of-           Hazardous Materials,  Washington, DC, 1987.
  fice of Solid Waste and Emergency Response, U.S. EPA memoran-        10. U.S. EPA, "RCRA Corrective Action Plan",  (Interim Final), Of-
  dum, Oct. 14, 1986.                                                    fice of Waste Programs and Enforcement, Nov. 14, 1986.
                                                                                              RCRA SITE REMEDIATION     227

-------
                             The Remedial Investigation  Process:
                                        An  Industry Perspective

                                               James E. Gagnon, P.E.
                                            Monsanto Chemical Company
                                              Springfield,  Massachusetts
ABSTRACT
  The environmental assessment of a hazardous waste site can be
an extremely costly and time-consuming process, and it is only a
first step leading to a final closure or securing of the site. This re-
medial action or corrective measures phase also can be costly and
can take years to define and implement. An efficient and com-
prehensive study phase therefore is essential to  a cost-effective
and environmentally sound site closure.
  The author of this paper utilizes a 2-yr remedial investigation
at Monsanto Chemical Company's Indian Orchard Plant in Mass-
achusetts to evaluate the  environmental assessment  process. A
second major aspect of the paper is a discussion of the manner
by which decisions were  made during the  investigation. Some
positions adopted by  regulatory agencies  and Monsanto  were
productive and contributed to the investigation; other positions
were not as productive.

INTRODUCTION
  The Monsanto Company has  owned and operated a  plastics
and resins manufacturing facility in Springfield, Massachusetts
since  1938. Products manufactured at the site include: polysty-
rene,  formaldehyde, polyvinyl butyral, polyvinyl acetate, poly-
vinyl  alcohol and  various other surface coatings and adhesives.
Until  1972 Montanto disposed of much of its process and non-
process waste on-site in various landfills, open pit burning areas
and liquid disposal areas. Fig. 1 is a plot plan of the Indian
Orchard Plant on which the various waste disposal areas are iden-
tified.
  Since 1981 Monsanto has been conducting studies to assess the
impacts these former on-site disposal areas are having or are like-
ly to have on the public health and the environment. In March
1983 Monsanto published a preliminary site assessment report
that concluded that groundwater contamination at the site  was
relatively minor and was not widespread. The report also stated
that further study of the site was warranted.
  In the spring of 1983 both the U.S. EPA and the Massachu-
setts Department of Environmental Quality Engineering (DEQE),
after reviewing the preliminary site assessment report prepared by
Monsanto, requested that Monsanto conduct further study of the
site. This paper concentrates on the decision-making process be-
tween the U.S. EPA, DEQE and Monsanto. The paper will illus-
trate positions or approaches that Monsanto feels contributed to
or hindered the progress of the remedial investigation.

REMEDIAL INVESTIGATION PLAN
  Nothing contributes  more to the smooth execution of a cost-
effective and comprehensive remedial investigation than an hon-
est and  open  exchange of information between all parties. This
does not mean that there cannot be differences of opinion con-
cerning technical approaches or data interpretation. What is im-
portant is that each viewpoint be evaluated and debated openly
and that understanding be reached before moving to the next
issue.
  During the early planning stages of a project, this "give and
take" is essential. For example, the remedial investigation plan is
a document that will be used on a site investigation for perhaps
years, and the quality of this document can have an impact on the
entire investigation. All concerned panics need to be a part of the
development of the investigation plan since all parties will have
authorities and responsibilities detailed in the plan.
  The remedial investigation plan developed to guide the investi-
gations at the Monsanto Indian Orchard plant in Massachusetts
has weathered 2 yr of field work and has proven to be a major
positive element in the progress at the site. The plan was devel-
oped  cooperatively by  the U.S.  EPA, Massachusetts DEQE,
Monsanto  Chemical Company and consultants for both  Mon-
santo and the U.S. EPA.
  Some details of the development of the investigation plan are
worth noting since they  illustrate a method of operation that can
contribute much to the site investigation process. Also, the tone
set by all parties at this early stage in the process can be carried
through the investigation and thus can have a lasting influence on
subsequent phases of a project.


Preliminary Assessment
  As  mentioned previously, Monsanto conducted a preliminary
site assessment during 1982 which concentrated on the two solid
waste disposal areas and the liquid waste disposal area. Work
tasks included test borings with the subsequent installation of 12
monitoring wells,  surface water and groundwater sampling and
analyses, permeability analyses and groundwater flow analyses.
The conclusions and findings were consolidated into a final report
indicating no excessive groundwater contamination and no obvi-
ous surface water contamination. The report did conclude, how-
ever, that further monitoring was justified in certain areas. This
preliminary assessment was invaluable as it provided a useful re-
serve of site specific data to present  to the regulatory agencies as
well  as a background data base to plan future site assessment
activities.
  Monsanto transmitted all reports, evaluations and  analytical
results to the DEQE for review. DEQE then relayed all the in-
formation  to the Region 1 office of the U.S. EPA. Despite the
findings of the preliminary investigation and Monsanto's position
to pursue further  investigations as a corporate policy, the U.S.
EPA  and DEQE insisted that the additional assessment would
have to occur under officially designated program(s) that would
228    RCRA SITE REMEDIATION

-------
SITE PLAN
^

\^^
^<:
^L

^<>—
3 " .If
DD

^==^~===^::^^
/
LEGEND

WASTE DISPOSAL AREA
Figure 1
Monsanto Polymer Products Co. Indian Orchard Plant,
Springfield, Massachusetts
be initiated by a consent order. Therefore, in anticipation of the
receipt of the consent order(s), a meeting was held in mid-August
1983 to acquaint the U.S. EPA with the Monsanto plant site, the
work done at the Indian Orchard Plant to date, Monsanto's
future site work plans and to discuss the consent order to be given
to Monsanto. This began the precedent of establishing a formal
working relationship between the U.S. EPA, the DEQE and
Monsanto. All parties agreed that their goals would be to create a
program that would be satisfactory to all parties involved without
untimely delays and with a united effort.

Consent Order
As anticipated, the U.S. EPA issued a unilateral order in
November 1983 (authorized by RCRA (30:13), and the DEQE
issued a consent order (authorized by Mass. 21-E, the State
Superfund Law, enacted in 1983). After issuing the order, the
U-S. EPA agreed to replace the unilateral order with a consent
order that encompassed the same technical program. Through
the fall of 1983 and spring of 1984, portions of the consent orders
were negotiated. During these negotiations, the U.S. EPA,
DEQE and Monsanto continued to work together to establish
the detailed technical aspects of a program defined as the Remed-
ial Investigation Plan (RIP).
The RIP was to attain certain fundamental goals such as
(1) defining the nature and extent, if any, of environmental con-
tamination by hazardous wastes or hazardous substances from
past waste disposal practices and (2) formulating, implementing
and monitoring a practicable plan for appropriate remedial ac-
tion without dictating actual techniques to achieve these goals
with the exception of using EPA approved analytical methods
including QA/QC procedures and RCRA statistical analyses.
In December 1983 all parties met to discuss the Consent Agree-
ment and Order, addressing the disposal areas to be studied, field
investigation techniques to be utilized, previous Monsanto re-
ports, analytical techniques, indicator parameter analyses vs.
priority pollutant analyses and the program outlined by Mon-
santo. This meeting was a technical meeting providing all parties
with a forum to exchange information of acceptable and prac-
tical field techniques.

Remedial Investigation Plan
At this meeting, a schedule for the submittal and review of
the RIP was agreed upon by both parties. To reduce regulatory
review time for the draft RIP, standard protocols for site inves-
tigation activities were reviewed by the regulatory agencies prior
to the draft RIP submittal. The protocols included the follow-
RCRA SITE REMEDIATION 229

-------
ing:  drilling/sampling  methods, well installation method,  geo-
physical  techniques, analytical  procedures,  QA/QC  and  site
safety protocols.
  A draft RIP was transmitted to the U.S. EPA and DEQE in
January 1984. The RIP contained three phases: Phase 1  provided
planning, surveys and review; Phase 2 included soil borings, mon-
itoring well installation and  water quality analyses; and Phase
3 included more detailed site assessment programs.
  Following the submittal of the draft RIP,  the U.S. EPA and
DEQE proceeded with a timely  review such that a meeting was
held in late January to discuss the rationale behind the tasks and
techniques proposed in the RIP. The regulatory agencies were sat-
isfied with most of the RIP proposed; a few areas were discussed
and revised reflecting regulatory viewpoints and Monsanto's will-
ingness to compromise to achieve approvable results.
  One area of discussion included the  proposed soil screening
technique. Monsanto proposed using an HNU and appropriate
Draeger tubes in the field to determine gross levels of contamina-
tion with limited GC work in a laboratory. The U.S. EPA, how-
ever, has had experience using an OVA  (organic vapor analysis)
with a gas  chromatography (GC) attachment for  in-field  soil
screening. The U.S. EPA felt that the HNU and Draeger  tube
method  might not  be as comprehensive a screening technique
given the potential for a wide range of chemical parameters and,
as  a compromise,  Monsanto re-evaluated  the  proposed field
screening program. Still acting under the belief that the GC analy-
sis  belongs in a controlled  environment,  Monsanto  rewrote this
portion of the final RIP to include in-field screening with an
HNU followed by a controlled  OVA screening (run  on total
mode) at the end of each sampling day in the confines of a build-
ing  on-site. The U.S.  EPA  and DEQE were pleased  with  the
change  as  it reflected an incorporation of their logic in field
screening.
  Also, as a result of the  late January  meeting, certain  details
of the work tasks were left  as joint field  decisions between Mon-
santo and the agencies. For example,  the total network of seismic
refraction lines was left undefined dependent on the success of the
initial seismic refraction  lines over each  different disposal  area
to be surveyed. If the seismic refraction technique  could pene-
trate through the first layer, a mixture of sand and refuse, addi-
tional lines would not be done.
  Also, the exact number  and location  of groundwater moni-
toring wells was not specified  initially but rather was left to be de-
cided by agreement of  all three parties after a review of soil bor-
ing data.
  The development of the investigation plan was made easier and
more productive by the commitment of all parties to exchange
information, negotiate and incorporate  various  viewpoints into
the final document. The  benefits of using this approach were: a
timely review  of all detailed  submittals  to the regulatory agen-
cies; the ability to institute field decisions into  the process of con-
ducting investigation programs; and the development of a work-
ing relationship between Monsanto and the regulatory agencies.

FIELD DECISIONS
  One of the chief benefits of the cooperative development of
the Remedial  Investigation Plan  for  the Indian Orchard  Plant
was  the consensus that field decisions would  be an integral part
of the investigation. Adopting the concept of  field decisions into
the process allowed the investigation to move forward logically
and in measured phases with each work task being built upon in-
formation gathered from previous tasks. Thus, in most instances,
site-specific conditions were allowed to be a major factor in decis-
ions concerning the scope and techniques of the investigation.
  In general, this approach worked  quite well and  allowed for
progress with all parties satisfied with the scope, pace and costs of
the investigation. The issue of bedrock wells as an element of the
Indian Orchard Plant investigation is a good example to demon-
strate the decision-making process and how field decisions were
utilized during this investigation.
   Both the federal and state agencies felt that a minimum of three
wells screened in bedrock would be necessary  to determine the
quality and flow  direction of groundwater in the bedrock under
the site. Monsanto took  the position that if  site-specific data
could  document  that  the possibility  of migration  of contami-
nants into bedrock from the site overburden was extremely low,
then bedrock wells were not a necessary component  of the inves-
tigation.
   Extensive information on site geology was gathered from 50
soil borings. The data  indicated that a hard thick unit of glacial
till was present above bedrock under the site. Monsanto took the
position that due to the extent  (present in every deep borehole),
the thickness (up to 50 ft) and the permeability (5 x 10"' cm/sec)
of the  till,  it  was not likely that contaminants would migrate
through the glacial till into bedrock.
   The agencies' position remained unchanged, and thus they con-
tinued to insist that bedrock wells were necessary to  establish the
condition of groundwater in  bedrock. Monsanto, recognizing
that the agencies needed additional documentation to be assured
that bedrock was not being impacted by contamination above the
till layer,  agreed  to install one  bedrock well. One well was in-
stalled approximately 20 ft into bedrock. The groundwater level
in the rock well proved to be about 15 ft above the groundwater
level in an adjacent well screened in the overburden. The indica-
tion was then that a strong upward gradient  existed and that the
direction of groundwater flow was from bedrock up into the over-
burden. This finding coupled with the characteristics  of the till
unit  gave  the agencies  the supporting documentation that they
needed to be assured that bedrock groundwater would not be im-
pacted by the site. No additional wells screened in bedrock were
required.
   Decision-making of this nature in the field  helps keep an inves-
tigation moving forward with the support of all parties. Agency
personnel  were able to obtain sufficient information  to eliminate
bedrock from further study, and Monsanto  was able to demon-
strate a commitment to  the comprehensive study of the site as well
as cost-effectively assess bedrock impacts.
   Allowing  site investigation  decisions to be made  in the field
does not assure agreement on all issues. This approach can lead to
constant disagreement  between  the parties and can  hinder pro-
gress. Indeed, a strong case can  be made for  defining most if not
all of the investigation issues before any field work begins.  Under
such an approach, all parties know exactly what is expected and
little cause for disagreement in  the field remains. However, this
rigid approach does not allow one to make reasonable decisions
as more information becomes available. If managed well, an in-
vestigation that includes field decisions as an integral component
can progress more quickly and the study findings will be sup-
ported by all parties.
   During  the investigation at the Indian  Orchard Plant,  many
major decisions concerning the investigation were reached after
evaluating  site-specific data.  Not all decisions were  fully sup-
ported by  all parties, and one such decision will be discussed here
to illustrate potential problems with this approach.


Problem Encountered
   What is the best drilling method for the installation of ground-
water monitoring wells? This question is raised  during many site
investigations, and the  answer to this question can have a major
impact on the cost of an investigation as well as the quality of the
230    RCRA SITE REMEDIATION

-------
data generated.
Much discussion and some disagreement surrounded the issue
of drilling techniques for the installation of groundwater mon-
itoring wells at the Indian Orchard Plant. The intent here is not to
review the technical details of the issue but rather to present a
brief summary of positions and highlight the decision-making
process and results of the discision.
Monsanto originally had proposed using the hollow stem auger
technique for monitoring well installation. However, much site-
specific information had subsequently been gathered during com-
pletion of about 30 soil borings at the site, and Monsanto wished
to utilize this information in the selection of a drilling technique.
Monsanto proposed that borings for monitoring well installa-
tion be completed using a combination of mud-rotary and drive
and wash techniques. The concept was to advance the boring to
within 15 ft of the final depth using the mud rotary method and
then place a casing in the hole and flush with fresh water. The cas-
ing would then be driven to final depth for well installation. Mon-
santo felt that this was a sound technical proposal that would pro-
vide a cost-effective installation and a monitoring network that
would yield excellent data.
The agencies did not agree with Monsanto's assessment that the
proposed drilling method would provide valid groundwater qual-
ity data. The agencies took the position that mud rotary drilling
could lower the aquifer permeability, could promote cross-con-
tamination within different zones and could alter groundwater
quality. The agencies allowed no debate on this issue and related
that it was a policy position taken by Region 1 U.S. EPA that the
use of drilling mud for monitoring well installation would not
be allowed.
Monsanto felt that the position adopted by the agencies was
quite arbitrary and was unsupported by data and weakly sup-
ported by the literature. A detailed response to each agency objec-
tion was prepared, but this was only for the record since the U.S.
EPA would not consider the merits of Monsanto's proposal.
Monsanto chose not to challenge the agencies through appro-
priate legal routes but rather chose to complete monitoring well
installation with the casing drive and wash technique. The impact
of this decision was significant since the pace of well installation
was greatly slowed and, as a result, costs escalated. An an aside,
the hollow stem auger technique was not utilized since experience
at the site was poor with this method due to fine sand continually
running up into the augers.
In addition to increased time and cost, the approach adopted
by the agencies in the above example had a broader impact. The
open exchange and debate of technical issues that have character-
ized this investigation were lost during this incident. While it is
hard to identify exact impacts, it is certain that the project pace
and the level of cooperation among all parties were hampered for
a period of time.

CONCLUSION
Over the next several years the remedial investigation or site
assessment process will be conducted at many manufacturing
locations. These investigations will be carried out with extensive
U.S. EPA and/or state oversight. Experience at the Monsanto
Indian Orchard Plant has shown that the investigation process
moves forward efficiently and cost-effectively when there is a free
exchange of technical information and a willingness by all parties
to evaluate proposals based on technical merits. Also, incorporat-
ing field decisions into the process allows site specific data to be
used to modify the investigation plan as necessary to ensure a
comprehensive and cost-effective site assessment.
ACKNOWLEDGEMENTS
I wish to acknowledge the following individuals for their fine
work throughout this project and for their assistance with this
paper:
George W. Lee, Jr.,CPGS
Principal, Blasland & Bouck Engineers, P.C.
Syracuse, NY
Stephen J. Rossello
Senior Project Geologist
Blasland & Bouck Engineers, P.C.
Syracuse, NY
S.B. Gensky
Project Geologist
Blasland & Bouck Engineers, P.C.
Syracuse, NY
RCRA SITE REMEDIATION 231

-------
                         Radial  Wells and  Hazardous  Waste  Sites

                                                    Wade Dickinson
                                                 R. Wayne Dickinson
                                                  Petrolphysics,  Ltd.
                                              San  Francisco,  California
                                                     Peter A. Mote
                                             Harding Lawson Associates
                                                  Novato,  California
ABSTRACT
  Horizontal radial wells can now be remotely placed from a cen-
tral vertical well to intercept, inject or monitor fluids and gas.
This new technology has great potential in the areas of hydro-
carbon recovery, contaminated groundwater and toxic gas con-
trol, mineral leachate recovery and domestic water supply devel-
opment. This paper describes the methods by which horizontal
radial wells are installed and discusses potential applications in re-
mediation and monitoring of hazardous waste sites.
  Current conventional methods for groundwater control rely on
placement of numerous vertical drill holes and wells to character-
ize, monitor and clean  up hazardous waste sites. There are limi-
tations and inherent difficulties with vertical wells. These include
surface and subsurface access to target  zone, geology, hydro-
geology and geologic structure, aquifer permeability and cross-
contamination of aquifers. Horizontal radial wells provide viable
solutions to these limitations.

INTRODUCTION
  Horizontal or lateral radial wells  have been in limited use for
decades. Traditional lateral wells include hillside drains  and col-
lector radials from large diameter wells. However, these applica-
tions  are either limited by terrain  or require large  subsurface
access areas to install them downhole. In the oil industry, devi-
ated drilling from the vertical to the horizontal is available with a
large  radius of curvature (1800 to  3000 ft) or with a  medium
radius of curvature (20  to 40 ft). The critical limitation of current
large and medium radius deviated drilling is that only one radial
may be placed in each  vertical well and the horizontal wells can-
not be placed at a precise level. Such precise placement requires
curvature from vertical to horizontal via a small radius turn at a
known depth.
  The recent development and commercialization of a novel high
pressure, hydraulic jet drilling system by Petrolphysics and Bech-
tel now allows the placement of multiple horizontal radials at a
small, very sharp 12-in. radius of curvature from a central verti-
cal well. With this new system, horizontal radials can be placed at
any depth radiating from a 5.5  in. or larger diameter  vertical
casing. Several stacked layers of radials can be placed from the
same vertical well.
  Placing horizontal radials into a contaminated aquifer from a
conventional vertical well  can markedly enhance access to the
formation. Because many  aquifers  are near horizontal, placing
wells in the plane of the aquifer should increase the area of col-
lection/injection per well, as compared to vertical wells across
the aquifer. With that intrinsic geometric advantage, the volume
of formation that can be serviced by a well is markedly increased.
Hence fewer vertical wells are needed, and a more uniform ac-
cess to the contaminated aquifer is provided.
  This paper will first describe the short radius horizontal radial
drilling system and associated well completion methods. Applica-
tion of this horizontal technology  to hazardous waste site prob-
lems will then be discussed.
     CABLE
RESTRAINT TRUCK
                                                 RIG
  HIGH PRESSURE
  TUBING STRING
        FORMATION
  UNOERREAMEO
       ZONE
                                       1-1/4 INCH STEEL
                                       DRILL STRING
                                       CASING
                                       HIGH PRESSURE SEAL
"  T^U  .
                  DRILL
                  HEAD
 _   RADIAL BORE
 •   INTO FORMATION


    HYDRAULIC ERECTION
    CYLINDER
                            Figure 1
                      Overall Drilling System
232    RCRA SITE REMEDIATION

-------
HORIZONTAL RADIAL DRILLING SYSTEM
Over the past 6 yr, Bechtel Group and Petrolyphysics Ltd. have
been developing a completely new technology of placing and com-
pleting horizontal drain holes or radials. That technology was
developed primarily for enhanced recovery of both heavy and
light oil in both shallow and deep oil reservoirs.
The Petrolphysics/Bechtel radial system consists of a central
vertical well from which radiates a pattern of radials. Several
such patterns can be vertically spaced at different depths in a
single well. These radials extend into the formation for distances
of up to several hundred feet (Fig. 1). To place the radials, an
erectable whipstock is lowered downhole into a previously under-
reamed cavity. The whipstock is loaded with a long 1.25-in. diam-
eter continuous steel tube (50 to 200 ft long) which has a hydraul-
ic drill head welded to the leading end (or nose). High pressure
drilling water is pumped through the steel tubing. The resultant
hydrodynamic forces on the drill head pull the 1.25-in. tubing
through the whipstock making a 90°, 12-in. radius turn. The tube
then advances horizontally through the formation.
.CENTRAL VERTICAL WELL
RADIAL
FOUR RADIAL PLAN VIEW
\
TWELVE RADIAL PLAN VIEW
Figure 2
Radial Patterns
Several radials can be placed at specific horizons to gain access
to contaminated hydrogeologic units. Each radial is completed
as an in situ well by electrolytically perforating the tubing and by
hydraulically placing a gravel pack around the steel tube inside the
radial well bore. Clean water can be pumped into the formation
to develop hydraulic barriers to prevent contaminant migration.
Conversely, contaminated groundwater can be extracted from
the system.

FIELD RESULTS AND COST
OF RADIAL PLACEMENT
To date, over 27,000 ft of 1.25-in. diameter horizontal radials
have been placed during development and commercial production
operations. These radials have been placed in several parts of the
United States and in Canada at various depths ranging from near-
surface to 6,800 ft and at various lengths up to and beyond 200 ft.
As shown in Fig. 2, multiple radials can be placed at one or more
levels from a single vertical well. As many as 100 radials have been
placed in a single well at three different levels.
The drilling is done hydrodynamically, and the radial is pulled
into the formation by its own hydrodynamic force (Rabbit Force)
which always keeps it in tension. Hence, the radial naturally
tends to go straight in the desired inclination. In the laboratory,
a variety of sandstones, chalk, granites and limestone have been
drilled successfully and rapidly. In the field, a wide variety of
geologic conditions ranging from unconsolidated formations and
Austin chalk to beds of granite cobbles have been successfully
penetrated.
In the oil industry the total cost of radial placement without
gravel packing in deep oil reservoirs is about $135,000 for four
1.25-in. radials of 100- to 200-ft length without gravel packing.
This completed cost for 400 to 800 ft of horizontal radial in-
cludes the deep drilling rig, tubing strings, pumps and all services
including logging. The incremental cost of placing more than four
radials is small because the placement time per radial is very short.
At most hazardous waste and contaminated groundwater sites,
these radial placement costs should be substantially reduced for
shallow radial applications because equipment requirements are
less and rig time requirements are substantially less.

DRILLING SYSTEM DESIGN
Drilling and Drill String Propulsion
Formation drilling and drill string (radial) propulsion use the
same hydrodynamic force. This force concurrently pulls and
pushes the drill string forward and digs the bore hole. In effect,
the drill string makes its own hole and jumps into it, hence it is
called the Rabbit Force.
In drilling, the drill string begins vertically, proceeds around a
90° whipstock turn and then enters the formation horizontally.
Typical velocities range from 5 to 120 ft/min in unconsolidated
formations. The same fluid pressures that drive the system can
be regulated to control the direction of the drill string as it ad-
vances. The overall drilling system configuration is shown in
Fig. 1.

Pulling & Pushing Force
The 1.25 in. drill string in the central vertical well is contained
in a sealed chamber consisting of larger diameter drill pipe that
extends from the surface down to the whipstock. The drill string,
which is advanced inside the larger string, passes through a fixed
high pressure chevron seal in the top of the whipstock.
The high pressure drilling fluid, usually water at about 8,000 to
10,000 Ib/in.2, creates the Rabbit Force on the nose of the string.
This pulling force keeps the string in tension so that it tends to go
straight. The same fluid pressure exerts a pushing force on the
cross-section of the posterior of the drill string. This force helps
push it through the seal/whipstock system. For example, at
RCRA SITE REMEDIATION
233

-------
10,000 lb/in.a pressure in a 1-in. ID tube, the net pulling force is
about 8,000 Ib and the net pushing force, which is exerted on the
posterior of the 1.25-in. tube, is approximately 3,500 Ib. The
combined total push/pull forward force on a 1.25-in. tube, there-
fore, is about ll,5001b.

Whipstock
The whipstock (Fig. 3) is lowered on the end of a larger diam-
eter string and is erected downhole in an underreamed zone (Fig.
1). The whipstock contains a set of slides and wheels that enable
the 1.25-in. steel drill string to move continuously and smoothly
from the central vertical well, around the 90° whipstock turn,
through a straightener section and horizontally out into the
formation.
HIGH PRESSURE
TUBING STRING
ENTRY SECTION
{INITIATES 90* CURVE)
WHIPSTOCX BODY
EXIT SECTION
(STRAIGHTENER)
HYDRAULIC ERECTION
CYLINDER
Figure 3
Mark I Whipstock

Because the drill string is triaxally stressed during this process
(hoop stress plus axial stress plus bending stress), the deforma-
tion of the 1.25-in. string at a 12-in. radius is easily accomplished.
In effect, the 1.25-in. tube is in a transient plastic state as it passes
through the whipstock. Many radials can be placed at several dif-
ferent azimuths with the whipstock remaining at the same level
without moving it uphole. Many stacked layers of radials can also
be placed from the same central well.

Radial Restraint and Control System
The radial pitch and inclination can be controlled by regulating
the velocity of the drill string as it moves through the formation.
To provide velocity control, which is especially important as the
string progresses through the whipstock and enters the formation,
a cable restraint usually is applied. Restraint is accomplished us-
RECTIFIER

A/D
CONVERTER
xTOOL
ELECTRICAL SCHEMATIC

, suoe WIRE
FLEXIBLE VERTEBRAE CABLE SEPARATOR UNEAR VOUAOE
SKIN BACKBONE SPRINGS OfFSKXTW.
TRANSFORMER
VEKTI3RAE
FLEXIBLE SKM
CABLE BACKBONE
SUOEWHE
TOOL CROSS SECTION

Figure 4
Model V Radius of Curvature Tool

ing a calbe-restraint truck and a wellhead high-pressure grease
seal at the surface (Fig. 1). A connecting cable attaches to a re-
movable disconnect at the tail of the 1.25-in. drill string. This
disconnect is removed to permit the entry of wire line logging
tools. Instrumentation attached to the restraining cable provides
a real-time measurement of both extent and rate of drill string
movement.
An alternate second-generation restraint system, which also
keeps the radial from moving too fast, uses a hydraulic tail that is
constructed much like a double acting shock absorber that con-
trols the velocity by a closed-fluid system. The cable connection
to the 1.25-in. radial is eliminated using the hydraulic system.

Electrical Downhole Logging
During or following placement, the radial can be precisely lo-
cated in the subsurface radial bore hole in three dimensional
space. The highly flexible directional wireline logging system—the
Model V Radius of Curvature Tool—has been developed and
commercialized for this purpose.
The Model V Tool (Fig. 4) is built around a flexible backbone
with small movable slide wires placed at 90° to each other. When
the tool is bent, differential movements of these slide wires on the
inner and outer radius result in electrical signals that provide a
continuous reproducible indication of both vertical and horizon-
tal curvature of the drill string through which the tool is ad-
vanced. The tool is pumped down the 1-in. ID drill string to yield
the three dimensional data on the radial trajectory. The tool pro-
vides vertical and horizontal location; the length of the wire line
cable provides the length of trajectory. The tool also can function
under several thousand Ib/in.1 hydrostatic pressure.
Model V can locate and print out the location of the radial with
reproducible accuracy of better than 99%.

Spinning Jet Drill Heads
The Spinning Jet Drill Head (previously called the nose of the
radial) is a major component of the drilling system. The present
generation spinning jet (Fig. 5) is a major improvement over ex-
isting fluid jet drill heads and was designed specifically for this
drilling system. This drill head produces a thin conical shell of
234 RCRA SITE REMEDIATION

-------
water which is believed to create a toroidal slurry body of cuttings
shaped like an inflated inner tube. These cuttings appear to act as
a rotating body of abrasive slurry that can cut hard crystalline
and consolidated sedimentary rocks. In effect, the cuttings are
the cutter. The technology also is applicable to other downhole
processes such as underreaming.
UNCONSOLIDATED
FORMATION
PERIPHERAL JETS
PENETRATE CONICAL
Figure 5
Spinning Jet in Unconsolidated Formation
VERTICAL WELL
/SLOTTED LINER
TAIL PERFORATED NOSE
FILTER HORIZONTAL FILTER
WELL
PERFORATION
Figure 6
Radial Well Completion with Perforations
With this Spinning Jet drill head and water drilling fluid, it is
possible to cut a 4- to 24-in. diameter borehole in unconsolidated
formations at 5 to 120 ft/min. The borehole diameter is con-
trolled by the internal angle of the conical shell of water particles.
For harder materials analogous to Berea sandstones, the pene-
tration rate for a 4- to 6-in. diameter borehole with water only is
several inches per minute. For hard granite rocks, the penetra-
tion rate is 1 in./min or more. Laboratory results indicate that the
addition of abrasives to the water results in either a much faster
rate of penetration or it allows cutting at a lower operating pres-
sure.

RADIAL WELL COMPLETION METHODS
Once the radial tubing is in place in the formation and its loca-
tion is logged, the well can be completed by gravel packing the
well bore around the radial, electrolytically perforating the radial
and placing a flexible slotted liner. Radial well construction has
been demonstrated in the laboratory on 80-ft full-scale radials.
The gravel packing is hydraulically placed as a slurry in the
annular space between the radial bore and the steel radial tube.
The packing is completed by a bi-directional process that pro-
vides a 100% fill of consolidated gravel.
To allow fluid infiltration from the formation into the radial
once the gravel is in place, the 1.25-in. radial tube is perforated
in situ by an electrolytic perforator tool using a low concentration
(12%) of potassium chloride at low flow (10 to 20 gal/min). The
tool makes 120 perforations simultaneously using a conventional
truck mounted electric welder as a power supply. The perfora-
tions are round, sharp-edged orifices, the size of which can be
controlled from the surface while the perforating process is
underway.
For some applications a flexible slotted liner may be desirable.
If so, it can be hydraulically placed within the perforated radial
to keep any packing or formation sand from entering the well per-
forations. Permeability of the flexible slotted liner can be selected
to preclude any fines from the formation from being carried into
the radial by the fluids. Finally, wire brush filters are placed at
the ends of the radial to prevent the gravel pack from entering
the radial. The radial well completion system is shown in Fig. 6.
In summary the final well system consists of an array of hori-
zontal wells that can extend laterally to 200 ft or more from the
central vertical well. The location of each horizontal well bore is
defined from the electronic downhole logging. The 4-in. or larger
radial bores are gravel packed, and the 1.25-in.^steel radial tube
is perforated and lined to prevent migration of the gravel pack or
formation fines into the 1.25-in. radial tube. These radial tubes
are terminated in the underreamed zone of the vertical well. A
standard submersible pump and a vertical slotted liner can then
be placed in that central vertical well.

APPLICATION OF HORIZONTAL RADIAL
WELLS
The potential applications of the radial well system to the sub-
surface injection or extraction of fluids or gas are numerous.
Some broad categories where application could be very beneficial
include:
• Hazardous waste/contaminated groundwater control
• Water supply development
• Economic recovery of mineralized leachate
Situations where the horizontal radial system could be advan-
tageously applied include:
• Injection/extraction from dense population centers or from be-
neath civil structures
• Injection/extraction from geologic conditions that are adverse
to fluid access with vertical systems (e.g., fracture permeabil-
ity)
• Injection/extraction from thin aquifers (less than 200 ft)
Given today's emphasis on protecting the environment, the
most obvious potential applications include groundwater control
and in situ modification of the ground of hydraulic conditions.
The benefits of horizontal drilling arise primarily from the ability
to intercept groundwater and geologic structures that cannot be
efficiently intercepted by vertical systems. Contaminated ground-
water plumes are prime candidates for the application of the
radial well system.

Groundwater Control
The typical well-field extraction scheme incorporates an array
of vertical wells that intercepts and withdraws the contaminated
groundwater for treatment (Fig. 7A). The design of the well field
is dependent on site-specific parameters such as aquifer trans-
missivity, plume geometry, groundwater and contaminant char-
acteristics and surface access limitations. Such limitations include
steep topography, water bodies, wet lands, power or process
plants, office and industrial buildings, urban developments,
RCRA SITE REMEDIATION 235

-------
streets, highways and powerline corridors.
Vertical wells draw water in all directions within the cone of
depression (Fig. 7A); to assure that the entire contaminated zone
is influenced by the wells, the cones of depression must neces-
sarily overlap into areas of clean groundwater. Thus, excessive-
ly large volumes of water are generated for treatment.
A horizontal radial well system (Fig. 7B) can overcome many of
these typical problems. A central vertical well can be strategically
placed to deploy horizontal radial wells within the contaminant
plume area while avoiding the obstacles mentioned above and
attaining efficient aquifer coverage.
Extraction from less permeable aquifers using a vertical well
array requires a large number of closely spaced wells. The con-
struction, capital and operating costs typically are high with such
well fields because of the number of wells required and the long
operational period required for extraction. The alternative of us-
ing fewer horizontal collector radials to do the same job could
provide significant cost savings by requiring fewer wells which
would reduce the cost of construction, capital and especially
operation.

Subsurface Drains—Leachate Interception
The horizontal radial system can effecitvely control seepage
from beneath existing surface and shallow waste facilities (Fig.
8). The horizontal radials intercept leachate from these facilities
before it migrates and mixes with groundwater.
WATER
SURFACE
Figure 7A
Conventional Vertical Well Field
RADIAL WELL
The radial well system can be used also for monitoring by plac-
ing radials in the unsaturated zone beneath a disposal site. This
would provide access for leachate or vapor detection devices.
Such devices could be designed to monitor isolated sections of
perforations within the radial, which would provide contaminant
concentration gradient data along the radial.
Figure 8
Leachate Monitoring or Collection by Horizontal Radials

Vertical Well System in Fractured Media
Figure 7B
Well Field with Vertical Well and Horizontal Radials
Horizontal Radial System in Fractured Media
Figure 9
236 RCRA SITE REMEDIATION

-------
In disposal areas where the leachate migrates primarily along
fractures, a perimeter horizontal radial array could be designed to
efficiently intercept fractures of nearly every inclination and
orientation. Such a perimeter array then could be used to con-
tain or extract contaminated groundwater or monitor a storage
site.
Containment options include: (1) developing a positive pres-
sure gradient toward the storage site to inhibit contaminant
migration (this could be done using either gas or water); (2) grout-
ing to seal migration pathways; (3) freezing the formation to seal
fractures. The ability to pump migrating leachate from a system
that has efficiently intercepted the fractures has obvious benefits
over most vertical well systems. A horizontal monitoring array
also has obvious significant advantages over conventional verti-
cal monitoring well systems, since effective monitoring is depen-
dent on the interception of vertical fractures (Fig. 9). The effec-
tiveness of intercepting vertical fractures with horizontal radials
for improved oil production has been successfully demonstrated
commercially.

In Situ Modification
The horizontal radial system has the unique ability to penetrate
contaminated soil and groundwater in lateral planes regardless
of depth below ground surface. The radial can be used for a wide
variety of in situ treatment methods: (1) injection of acid-neu-
tralizing chemicals, absorbing clays or reactive chemicals, (2) in-
jection of neutralizing bioorganisms and nutrients and (3) physi-
cal modification by freezing or grouting. The latter two methods
are of particular interest.
Bioremediation is a developing method for in situ treatment of
certain petroleum products and volatile organic compounds.
These contaminants often occur as lateral lenses within the geo-
logic formation or float along the water table surface. The intro-
duction of microorganisms and nutrients along the lateral line
or radial array within the contaminated zone provides for the
efficient and thorough placement of the organisms and nutrients.
In situ modification by solidification/stabilization can be done
by binding the waste within a solid mass. Solidification can be
accomplished through grout injection (cement or polymer) and
through freezing by continuous circulation of a refrigerant or in-
jection of a cryogenic liquid.
In a lateral grout injection system, cement or a chemical grout
is injected under pressure through the system of radials that have
penetrated the fracture system. The fluid grout permeates the
fractures in the formation and seals potential pathways. In situ
freezing is accomplished by injecting cryogenic fluids (nitro-
gen, liquid nitrogen or liquid carbon dioxide) through the array
of horizontal radials.
In some emergencies (Fig. 7B), application of radials could be
placed remotely and rapidly to control the subsurface effects of
an accident at a nuclear or chemical plant. Generally, at such fa-
cilities, the foundation geology is well documented as the result of
permitting and licensing requirements, and the radial well system
could be quickly designed to stop or inhibit the spread of contam-
inants away from the accident site. The radial well system could
possibly have been applied at Chernobyl in the USSR. Grouting
beneath the failed reactor could possibly have been conducted
remotely without requiring personnel to be near or under the re-
actor.

CONCLUSIONS
Application of horizontal radial methods to waste problems
offers another subsurface dimension to contaminant collection,
monitoring and immobilization in both unsaturated and satur-
ated materials. Such radials can be placed around the periphery
and/or beneath a waste disposal site and provide a boundary
piping system to collect or monitor leakage or to immobilize or
contain contaminants. Radials offer the opportunity to grout
along horizontal injection holes or to use cryogenic fluids to
freeze a plume in place. In emergency situations radials could be
placed rapidly beneath a failed and leaking plant to grout migra-
tion pathways to reduce or eliminate potential leakage into the
groundwater. Many radials can be placed from a single vertical
well. They can be placed in one horizontal or lateral plane or in
several stacked layers. The total radial array installation 'can be
done very rapidly.

ACKNOWLEDGEMENT
The authors gratefully thank the Bechtel Group Inc., Bechtel
Investments Inc. and all of their colleagues for their support and
encouragement of this work.

REFERENCES
1. Pendleton, L.E. and Ramesh, A.B., "An Innovative Method of Drill-
ing Horizontal Boreholes," Heavy Oil and Oil Sands Technical Sym-
posium, Calgary, Alberta, Feb. 1985.
2. Dickinson, W. and Dickinson, R.W., "Horizontal Radial Drilling
System," SPE 13949 Proceedings Society of Petroleum Engineers
(SPE) 1985 California Regional Meeting, Bakersfield, CA, Mar. 1985.
3. Dickinson, W., Anderson, R.R. and Dickinson, R.W., "A Second-
Generation Horizontal Drilling System," IADC/SPE 14804 Proc. of
1986IADC/SPEDrilling Conference, Dallas, TX, Feb. 1986.
4. Simmons, R.N., "Turning the Corner," Oil and Gas Investor, June
1986.
5. Crosby, T.W. and Head, H.N., Bechtel National, Inc., and Dickin-
son, W. and Dickinson, R.W., "Management of Uncontrolled Haz-
ardous Waste Sites," SUPERFUND '86, Vancouver, B.C., Canada,
Dec. 1986.
RCRA SITE REMEDIATION 237

-------
                           Closure  of  a Hazardous  Waste Landfill
                       Incorporating  a  Leachate Collection  System

                                                  Thomas A. Kovacic
                                                  Keith C.  Mast, P.E.
                                            Woodward-Clyde  Consultants
                                                       Solon, Ohio
ABSTRACT
  An inactive industrial/hazardous waste landfill has been closed
in accordance with regulations set by the Illinois Environmental
Protection Agency. The landfill was placed on the state priority
list for sites requiring remedial action to bring them into compli-
ance with regulations. Leachate  seeps emerging from the land-
fill, lack of a well drained, low permeability cover and  the pres-
ence of a stream flowing through the property and eroding into
the slope of the landfill were major concerns to be addressed. The
erosion of the slope was greatly increasing the possibility for a re-
lease of waste materials directly into the stream.
  The  three distinct engineering/environmental components of
the closure: (1) erosion protection;  (2) leachate collection; and
(3)  low permeability cover were constructed in two phases under
separate contracts. The  incorporation of a  high  density  poly-
ethelene (HOPE), flexible membrane liner (FML) into the design
and construction of a metal bin retaining wall was completed in
phase one. This design provided  for the monitoring and,  if re-
quired, removal of leachate that  may continue to seep  from the
landfill. The retaining wall allowed for restructuring of  the slope
and stream channel, protecting them from further erosion. The
placement of a low permeability, compacted clay cover is present-
ly under construction as phase two to reduce surface water infil-
tration, thereby reducing the potential for additional leachate.

INTRODUCTION
  Due  to an evergrowing national concern of the possible envi-
ronmental hazards of landfills,  a hydrogeologic assessment of
an inactive industrial/hazardous waste landfill site was conducted
in 1983. Past records regarding possible contaminants  disposed
of and  their probable locations within the landfill were  reviewed
for  background reference. A series of 13 groundwater  monitor-
ing wells (including five groups of two nested wells each) was in-
stalled  near the perimeter of the site to determine groundwater
elevation, flow direction and possible contamination.
  During initial site visits, it was  noted that the west bank of the
landfill was being eroded and undercut in places by the Redmond
Branch of Cedar  Creek  which flows adjacent  to  and  approxi-
mately  1,000 ft along the western and northern slopes of the land-
fill (Fig. 1). The undercutting action had already resulted in slope
failures. A number of apparent  leachate seeps were  observed
along the western slope.  Continued  erosion and resulting slope
failures could result in the major release of waste materials from
one or  more of the impoundments directly into the stream.  Also,
drainage of the  landfill was inadequate.  The area  had  been left
nearly  flat with a soil and fine  gravel covering, increasing  the
potential for surface water infiltration.
  Subsequent design and  construction of the two phases of work
as described previously have proven to be at least partially effec-
tive. Since the construction of the clay cap is presently being com-
pleted, it will not be known for some time just how effectively
the cap will stop the production of leachate. However, the retain-
ing wall has stopped erosion of the landfill slope, and the leachate
collection system has performed as designed.
  Following construction, a post-closure monitoring plan will be
developed for the site, primarily to continue analysis for chem-
ical constituents in well,  stream and leachate water. Careful in-
spection of past concerns at the landfill will be monitored in post-
closure activities.

SITE DESCRIPTION AND HISTORY
  The landfilled area occupies approximately 6 ac in the middle
of a 9-ac site (Fig. 1). The average elevation of the surface of the
landfill is 606 ft above mean sea elevation. The Redmond Branch
of Cedar Creek enters the property from the south at approxi-
mately 589 ft elevation and flows along the entire west bank of
the landfill. In the northwest corner, the stream makes an approx-
imate  120 degree turn and flows east along the north bank of the
landfill. The stream makes an approximate 90 degree turn north
about midway along the  north property line and exits the site at
an elevation of S83 ft. The stream averages  10 ft in width and
flows  on a stream bed of limestone bedrock. Thick vegetation
occupied both  banks of  the  stream, including the west slope of
the  landfill. Little or no  vegetation grew over a majority of the
landfill area. A 21-in. diameter, public sanitary sewer line lies be-
neath  the western edge of the  landfill. Neighboring property to
the south and west consists of pastureland that drains toward the
site. Railroad property borders the site to the north and east.
  The  current stream channel  was rerouted by the railroad
authority in the 1920s. Waste disposal operations began at the site
in the early 1950s with the disposal of industrial wastes includ-
ing  construction debris,  foundry sand, ferrous sulfate crystals,
carbide sludge and spent  sulfuric acid.  These wastes were created
by degreasing and painting operations during the production of
steel wheels. Most  wastes were disposed  of as sludges. Small
amounts of asbestos and  mercury were also disposed at the land-
fill.
  Disposal operations at  the site were discontinued in 1972 after
iron-enriched, acidic leachate  was  found to be discharging into
the  stream along the northern boundary of the fill area. At this
time, a leachate collection system  was  installed along the north
bank of the site (Fig. 1). Three pump houses were constructed
to pump leachate from a  "French drain" through a series of PVC
lines to the sanitary sewer located on-site. This liquid is regularly
monitored and tested to assure compliance with permit limits.
  By 1983, perimeter groundwater monitoring wells had been in-
238    SITE MANAGEMENT & CLOSURE

-------
stalled. Chemical analyses of water obtained from the wells, up-
stream and downstream creek locations, seeps detected along the
west bank of the landfill and collection system discharges pro-
vided necessary data regarding water quality associated with the
site.
To date, no chlorinated compounds have been detected in the
groundwater. Heavy metals (priority pollutant and drinking
water) have not exceeded standards or guidelines for ingestion.
Several organic compounds were detected in a few monitoring
wells, but concentrations were below recommended or proposed
maximum contaminant levels. Some seeps had a pH as low as 3.5.
During the 1983 investigation, three separate seeps were noted
in the locations shown in Fig. 2. The seeps were located in moder-
ate to heavily eroded sections of the streambank. Redmond Creek
flow is not gaged, and no accurate flow rates were measured.
During low flow conditions, the creek flow varies from 6 in. to
approximately 24 in. in depth. During flood periods, however, it
is not uncommon for the stream to rise 6 to 8 ft and drop back
again in the matter of hours. Flood insurance data indicated the
stream could overflow the banks onto the landfill. The flood
waters erode the weakest, clayey silt sections of the slope resulting
in the slope failures.
At the conclusion of the 1983 investigation, it was realized
that immediate erosion protection was required. Future planning
included the design and construction of the clay cover for the
landfill and the possibility of a slurry wall being installed to stop
groundwater flow.
•Minhoto X-40
Figure 1
Landfill Location Plan
AREAS OF OBSERVED EROSION
LOCATION OF OBSERVED/
SAMPLED SEEPAOE
Figure 2
Streambank Erosion and Seep Location Diagram
DESIGN
Retaining Wall/Leachate Collection
Although the retaining wall and leachate collection system were
designed and constructed together, primary concern was focused
on the retaining wall. An economical design that could be quickly
implemented was required for at least two heavily eroded bank
sections along the stream. The observed areas of erosion and
seepage are shown in Fig. 2. The major area of concern was in
the vicinity of manhole (M.H.) X-39, where the eroding slope
was encroaching on the former carbide sludge/acid neutraliza-
tion impoundment. Another section of the stream bank near
M.H. X-40, because of the configuration of the stream, was be-
ing eroded at a higher rate than other sections of the bank.
It was desirable to limit the extension of the stabilization sys-
tem into the existing stream in order to reduce the amount of
stream disruption and dewatering that would be required to facil-
itate construction. The excavation into the existing side slopes of
the landfill were to be minimized in order to reduce the possibil-
ity of exposing the sludge and to avoid damaging the sanitary
sewer line. Engineering concerns for the retaining wall were typ-
ical; one had to design against lateral earth pressure in slipping-
shear and overturning. The wall also would have to conform to
the general shape of the slope and allow for the installation of a
leachate collection system on the back side.
Six retaining wall systems were considered to be technically
feasible; crib walls with mechanical shear resistors to resist slip-
ping appeared to be the most available and economical. The 10-ft
wide by 8-ft deep metal bin sections are constructed of galvan-
ized steel panels front and back, bolted to corner columns. Adja-
cent bins are bolted to common columns. Interior spacer panels
prevent bulging or crushing of the bin sections. The manufactur-
ers guidelines for panel gauge thickness and batter offset from
vertical were utilized. The retaining wall design and compon-
ents were standard; however, the foundation was not.
Normally, metal bin retaining walls are set on compacted soil
fill so that during construction and backfill the wall may settle as
a unit. Since the wall in this application would be constructed on
the limestone stream channel bottom, no settlement would occur.
In addition, to permanently set the bins in-place, a concrete foun-
dation was designed to encase the lower bin panels and columns.
The rigid bin foundation would join the wall to the underlying
bedrock through the use of steel shear resistor rods drilled and
grouted into the bedrock. The manufacturer gave assurance that
even though a rigid foundation would be used, the metal bin sec-
tions would perform adequately.
The purpose of the wall was to protect the heavily eroded sec-
tions of the streambank against further erosion. The areas shown
in Fig. 2 were heavily eroded, while other sections showed little
or no erosion. The erosion on the west bank of the stream in the
northwest corner of the site posed no environmental threat and
was of little concern at that time. A decision was made to pro-
tect only the most seriously eroded sections of the streambank.
Should the need arise for additional retaining wall sections, they
could easily be constructed at a later date and could be added
directly onto the existing wall sections. For this reason, it was de-
cided to build two separate wall sections as shown in Fig. 3.
Preliminary water quality analyses of the seeps along the stream
bank indicated that measures should be taken to substantially
reduce the flow of leachate into the stream. A leachate collec-
tion/monitoring system was designed for the northernmost wall
of the two wall sections. The southernmost wall was not designed
with a collection system since it is far removed and upgradient
from the worst impoundment area. Behind the north wall, the
collection system would greatly reduce the amount of water en-
tering the bank during periods of high water as well as contain
SITE MANAGEMENT & CLOSURE 239

-------
any leachate emerging from that section of the landfill.
  The main element in the collection system design was a 60-mil
thick, HDPE FML. The liner, reinforced with a geotextile fabric
to protect against punctures, would be placed loosely against the
back of the retaining wall. All overlapping seams  of the mem-
brane would be sealed with an extrusion welding process  that
creates one continuous unit out of the individual pieces. The FML
provides the barrier to keep stream water from entering the land-
fill and leachate from getting out.
  Between the membrane and the natural slope, washed gravel
would allow leachate to flow into the collection system. Near the
base of the wall a 6-in. perforated PVC pipe collects the leach-
ate and drains it to a newly installed manhole, also located behind
the new wall. From the manhole, leachate could be monitored
and, if required, removed. A compacted clay cover would be con-
structed over the leachate collection system and  the top of the
wall.

day Cover
  The landfill  cover design  was developed  following  Illinois
EPA (IEPA) guidelines. A minimum 2-ft thick compacted  clay
cover should top the entire landfilled  area.
 r	•»-»•—L=r-—  -_	'  "
 \  ^"v±r.n.—~s'
  \  ^V llMM. I-»       /
                         LMVtTMIHT »«T

                   -•OUTM HIT MUM W*U.
                           Figure 3
            Completed Stream Bank Erosion Protection

  The final surface cover must be graded to a minimum 2% slope
in any direction and should promote shallow rooted vegetation.
  To achieve these basic requirements, the approximate middle
of the landfill would be elevated 8 ft to create the 2°?t slope. Two
to three feet of clay, capable of attaining an in-place permeability
of 1 x 10"7 cm/sec or less, would be compacted over miscellan-
eous site soil fill and covered with topsoil. In constructing the clay
cover, two swales would be excavated to the east and south edges
of the landfill. These swales would collect and channel surface
water to the creek on the west and north sides of the site. Rock
channel protection would be placed over geotextile at the ends of
both swales to minimize erosion. Wood excelsior matting would
be placed in both swales after seeding.
  A concern in designing the clay cover was exposing sludge ma-
terial buried within the fill. Any sludge excavated would be placed
in the deepest fill area, near the center of the landfill, beneath the
clay cover.

CONSTRUCTION
Retaining Wall
  The contract to erect the retaining wall/leachate collection sys-
tem was awarded in September, 1985, and clearing work was be-
gun shortly thereafter. Construction was to be completed before
winter so that no problems with freezing weather would be en-
countered.  However, a major delay in manufacturing  delayed
steel delivery until mid-December.
  The metal bin crib wall sections were erected in four basic steps:
excavation; concrete leveling  slab and shear rod  installation;
structural steel; and concrete foundations. Fig. 4 shows construc-
tion early in the project.
  After diverting water away from the work area, the streambed
was excavated 3 ft or more as necessary to reach competent bed-
rock.  Excavation normally extended approximately  30 ft down-
stream at a time beginning at  the south end  of the north wall.
Pumps were used to pump water seeping into the work area back
to the stream. After  excavating to rock, shear rods  were drilled
into the rock and grouted in  place. A thin, concrete leveling slab
was then placed. Fresh concrete was covered with Visqueen and
Styrofoam as needed to protect against freezing. When dewater-
ing pumps were turned off at the end of each day, water would
quickly return to the work area and cover the freshly poured con-
crete, creating a natural insulating blanket. Each following morn-
ing, water and ice would be removed from the work area in prep-
aration to erect structural steel.
  Semi-erected steel panels were set in-place by crane and bolted
together to form the first bin. Each consecutive bin was then
erected in the same manner,  bolting into the columns previously
set. At least one panel would be bolted on each edge of the cube
structure to provide stability  while pouring the foundation. Nor-
mally, three bins would be erected during  each day. The 6 to 1
batter and  design  elevation of the wall was maintained with
threaded leveling rods located  on the base plate of each struc-
tural column. The columns could be slightly adjusted  by screwing
the rods up or down on the thin slab poured the previous day.
  After each  group of three bins was erected, concrete founda-
tions  were placed.  The foundations rigidly connect  all columns
and bottom steel bin panels directly to competent bedrock. Cora-
pressive strength tests were performed on concrete cylinders and
grout cubes throughout the project to assure minimum 3,000 Ib'
in' strength.
  When completed, wall height ranged from approximately 5 ft
at the end of each wall to a maximum of 20 ft above bedrock. A
typical wall cross-section is shown in Fig. 4.
  The south end of the  south bin wall ends approximately 35 ft
from  M.H. X-40, located a few feet inside the property line. The
bin wall could not be extended to the property line because of the
                           Figure 4
               Typical Steel Bin Erection Photograph
240    SITE"MANAGEMENT & CLOSURE

-------
configuration of the stream and the location of the manhole;
however, the manhole was already exposed due to erosion and re-
rmired protection.
  An erosion control revetment  mat was installed  in this area.
The material used was a double  layer synthetic  fabric joined to
create a pillow form work for placing grout. The area to be pro-
tected was graded as necessary, and upper and lower anchor
trenches were excavated. A geotextile was placed on the slope to
protect the slope from soil erosion. The mat was  then placed with
 upper and lower edges in the anchor trenches. The south end was
 wranoed around the exposed face of the manhole, and the north
 end of the mat butts against the bin wall. The section of mat lying
 in the upper anchor trench was pumped full of grout first to hold
 the mat in-place. Grouting continued until  the entire  mat had
   Wn erec    was completed, backfilling began. Backfill-
 ing internally within the bins proceeded evenly with external back-
 filling to prevent uneven loading. The construction of the leach-
 ate collection system was incorporated  in the backfilling of the
 north bin wall section.
             QEOMEMBRANE
                                             BIN WALL
                                            STEEL PLATE
                                              CLAMP
            EXTRUSION WELD
              OEOTEXTILE
             REINFORCEMENT
                             Figure 5
                   Detail of the Oeomembrane Joint
    Leachate Collection System
      To protect the main component of the leachate collection sys-
    tem, the FML, a reinforcing geotextile was placed on the back ot
    the bin wall. The geotextile provided protection against punctur-
    ing of the FML. The geotextile was hung loosely from the top ot
    the bin wall. Vertical joints were overlapped 3 ft. The FML (a
    Gundel HOPE) extends the entire length and height of the back-
    side of the north wall. The FML was placed in two sections. The
    bottom section extends from the base of the collection system, 12
     ft up the back of the wall where flat plate clamps secure it to the
     wall (Fig  5)  The bottom section extends horizontally outward
     2 ft from the base of the wall. The top section hangs from the top
     of the wall down overlapping the bottom section clamps. The top
     section folds over and into each bin section. At each end of the
     wall, the liner was folded back into the slope to create an envelope
     effect.
  The membrane was installed by personnel from the manufac-
turer  Separate sections of the liner were seamed with an extru-
sion welding process that made one continuous unit out of the
individual pieces. In approximately 240 ft of seaming tested, only
two pinholes were detected and subsequently repaired.
  The bottom of the leachate collection system was constructed
of a 1-ft clay layer, the FML and then an additional 1-ft clay lay-
er  In-place density testing was performed on all clay lifts .The
clay layers act as an anchor  for the bottom edge of the fabrics
and prevent the movement of liquids from between the mem-
brane and bedrock.                                         ,
   A geotextile was placed on top of the second clay layer, extend-
ing from the geomembrane up the existing slope. The purpose of
the geotextile is to prevent the migration of fine soil particles into
the collection system.                       .
   Perforated  PVC pipe (6-in. diameter) was ins ailed on top of
 the filter fabric at the base of the wall. The pipe slopes from both
 ends of the wall and empties into a collection manhoe construc-
 ted approximately 80 ft from the north end of the wall The man-
 hole is keyed into rock and rises above final topsoil  elevation
 approximately 1-ft away from the top of the wall The new man.
 hole is situated close to existing M.H. X-39, so that any leachate
 may be pumped into the sanitary sewer system, if appropriate.
 Final grade for the pipe was established by filling with  rounded,
 washed gravel. This gravel also was used to backfill the entire area
 behind the wall to within 3.5 ft of final design elevation.
    While gravel was being placed behind the wall, sand was being
  compacted within each bin section with vibratory compactors In-
  place  density of compacted sand lifts was tested and  recorded.
  The membrane and geotextile reinforcement were anchorsd inside
  the bins between two sand lifts. The geotextile filter fabric was
  folded over the top of the rounded gravel to the back of the wall
    A 3-ft clay cover was compacted over the entire system, extend-
  ing from a 5-ft wide bench cut into the top of the existing slope
  to the face of the wall. Topsoil was placed and graded over he
  clay layers so that surface  water  would drain to the face of the
  wall.  Drainage scuppers were installed in every other panel on the
  face of the wall. A typical  cross-section of the completed collec-
  tion system is shown in Fig. 6.                         .„„„„!
    The south wall was backfilled in a similar manner, but no col-
   lection system  was  constructed. All  topsoil was seeded and
   mulched.
                               Figure 6
                 Typical North Wall Cross-Section as Built


     Landfill Cover/Final Closure
       The landfill cover project is being finalized as this paper is be-
     ing written Final closure consists of a compacted clay cover on
     the landfill, graded to drain adequately and the placement of

                       SITE MANAGEMENT & CLOSURE    241

-------
 rock-channel protection in specified locations.
   Miscellaneous, existing site soil was graded to achieve design
 elevations prior to clay placement. Some fill material  was re-
 moved from the northwest corner of the property. This area was
 graded to a maximum 2 to 1 slope and protected from  erosion
 with rock channel rip-rap. Erosion in this area may have even-
 tually  reached neighboring properties. Rock-channel protection
 also was  placed between the two bin wall sections as  a  precau-
 tionary measure.
   Two drainage swales were  excavated; both swales originated
 near the entrance gate in the southeast corner of the  landfilled
 area. One swale was graded westward and will channel  surface
 water  runoff to the creek at a point between the two walls.  The
 second swale  runs north  along  the eastern border of  the filled
 area and  turns to the west near the property boundary.  It then
 drops  to  the west until it empties into the stream where it exits
 the property.
   In excavating to required grade in  the swale areas and in the
 northwest section of  the landfill, some waste materials were un-
 covered. These materials were buried in the center of the landfill,
 in the  miscellaneous fill compacted directly beneath the first  clay
 liner.
   Topographic maps of the site  were drawn prior to and  immed-
 iately following clay placement. These maps will prove clay thick-
 ness in all locations over the landfill surface. A final as-built top-
 ographic map will be drawn to verify topsoil thickness.
   Several clay borrow sources were investigated before one site
 was chosen. The clay is a uniform material with laboratory tested
 permeability of 1 x 10"7 cm/sec or less at reasonable compacted
 densities. A range of compacted densities and moisture contents
 was established for field control. Optimum moisture contents are
 in the  16.5 to 20.5% range.
   At the  time of this writing, approximately 10% of all clay has
 been placed. Clay layers were placed in 6 to 8 in. loose lifts  and
 compacted with sheepsfoot rollers.  In-place densities of clay  lifts
 are verified by nuclear density gauge testing at the rate of  one per
 10,000 ft1 of clay placed. Sand cone density tests are also  run for
 comparison testing.
   After final grading of the last clay lift, topsoil will be  placed.
 Seeding and mulching of the site will be in accordance'with state
 highway department specifications. Both swale areas and slopes
 at greater than 10% grade will be covered with a wood excelsior
mat to reduce erosion. The ends of both swales also have been
protected with rip-rap placed on geotextilc.


CONCLUSION
  The site has gone through several high water periods since the
construction  of the retaining wall. The wall sections and revet-
ment mat have performed admirably in protecting against ero-
sion. At some of the wall columns along the face of the wall, small
sink holes have been noted. The holes have been attributed to the
sand within the bins filtering through gaps in the wall panels at-
tached to the columns and not to settlement. The holes have not
increased in size since they were first noted, and it is believed that
they will  not.  Each hole has since been filled with compacted clay.
  Leachate seeps still are being noted along the foundation of
the south wall, but this seepage was expected until completion of
the clay cover. The leachate has been repeatedly sampled and no
environmental threat exists.
  Approximately 3 ft of liquid were observed in the north wall
leachate  collection system, but were not observed until construc-
tion of the clay cover began. Analyses of the liquid show some
leachate  constituents, greatly diluted. Additional water has en-
tered the collection system. Since the leachate level does not read-
ily fluctuate with the stream elevation, it is believed to be entering
through the landfill surface (runoff water trapped during excava-
tion for clay cover) and should diminish when the cover is com-
pleted.
  Metered volumes of leachate pumped from the original collec-
tion system installed along  the north slope  in  1972  will be ob-
served in the  future. A reduction in volume  is anticipated. Also,
leachate  seeps along the south bin wall are expected to dissipate
and stop after the clay cover is completed.
  In the future, if leachate seeps continue, a collection system
could be  installed behind the south wall. If necessary, a slurry wall
could be constructed along the south property boundary to cut-
off groundwater passing through the site. It is anticipated that
these measures will not be required.
  A post-closure monitoring plan for the site presently is being
developed for the entire site. The plan will include sampling and
analyzing all monitoring wells, stream water and liquids trapped
in the collection systems or from seeps. This monitoring plan will
continue for several years.
242    SITE MANAGEMENT & CLOSURE

-------
Effective Closure of Surface Impoundments
Requires a Multi-Phase Plan

Albert K. Langley, Jr.
Georgia Department of Natural Resources
Environmental Protection Division
Atlanta, Georgia
ABSTRACT
Closure of hazardous waste surface impoundments is technical-
ly complex and expensive. Choice of a closure option must care-
fully consider costs and future liability while meeting complex
regulatory requirements. A closure which utilizes a combination
of treatment and removal followed by closure-in-place has proven
to be the best option for many facilities. This paper discusses the
decision process and provides case histories for several Georgia
facilities.

INTRODUCTION
The enactment of the Hazardous and Solid Waste Amend-
ments of 1984 (HSWA) substantially altered RCRA, particular-
ly concerning the management of hazardous wastes in surface
impoundments. The amendments establish strict minimum tech-
nology standards and compliance dates for hazardous waste sur-
face impoundments. Currently operating intrim status surface
impoundments must either retrofit with a double-liner and leach-
ate collection system or discontinue treating, storing or dispos-
ing of hazardous waste by Nov. 8, 1988. Further, interim status
facilities were required to certify compliance with groundwater
monitoring and financial responsibility requirements and submit
Part B hazardous waste operational permit applications by Nov.
8, 1985 or lose interim status i(theregulations; governing these
impoundments are found at 40 CFR 260-266 and 270).
The part 265 financial responsibility requirements for a haz-
ardous waste surface impoundment include obtaining non-sud-
den pollution liability insurance. This coverage currently is not
available to many industries. In Georgia, 84% of all surface im-
poundments lost interim status on Nov. 8, 1985 for failure to
obtain the necessary insurance. When a hazardous waste surface
impoundment loses interim status, the owner/operator must file a
closure plan within 15 days. The plan must then be implemented
as soon as the required public comment period has expired. Thus
most interim status surface impoundments in Georgia currently
are undergoing or recently have completed closure.
With almost no exception, the surface impoundments which
retained interim status will close prior to the Nov. 8, 1988 dead-
line since retrofitting surface impoundments with double-liners
and leachate collection systems is in almost all cases impractical.
The enactment of the HSWA eliminated the management of haz-
ardous waste in surface impoundments as a viable option for most
Industries. The very stringent minimum technology standards,
the unavailability of the required insurance and the consequences
of contaminating groundwater combine to eliminate any eco-
nomic incentive to manage hazardous wastes in on-site surface
impoundments.
Closure of a hazardous waste surface impoundment is tech-
nically complex and expensive. Careful planning and choice of
closure methodology can result in environmentally sound and
cost-effective closures. This paper provides a description of the
decision-making process faced when choosing a closure option
and relates cases histories of several Georgia facilities.

Alternate Treatment System
Prior to closing a surface impoundment, an industry must
develop some other method of treating the waste being discharged
to the surface impoundment. Table 1 provides a breakdown of
the interim status surface impoundments operating in Georgia
prior to the enactment of the HSWA. Currently, only 10.5 % are
actively receiving hazardous waste. The remainder are at some
point in the closure process. The most common process changes
which facilities have initiated to eliminate the use of surface im-
poundments involve either installing a pretreatment system using
tanks and discharging the effluent to a Publicly Owned Treat-
ment Works (POTW) or replacing the existing surface impound-
ment with tanks and obtaining an NPDES discharge permit.
Neither option is inexpensive, yet this is the first step which must
be accomplished prior to initiating closure.
Table 1
Active Hazardous Waste Surface Impoundments In Georgia
Industry Type

Wood Preservation
Metal Plating
Drug Manufacture
Chlor-alkali Plant
Wastewater Treatment
(POTW)
Wire Manufacture
Solvent Reclamation
Paint Pigment
Manufacture
Magnetic Tape
Manufacture
Pipe Manufacture

Organic Chemical
Manufacture
Pesticide Manufacture
Aircraft Manufacture
Number of
Impoundments

15
9
1
4
12
5
1
1
Size
(ha)

0.1-5.0
0.1-1.5
0.2
34(total)
0.1
0.3

0.1
Haste

KD01 Sludge
Plating Sludge
Solvents
Mercury,
corrosives
Oirome
lead and
Solvents
40(Total) Corrosives
0.1
0.1

0.6
0.1
0.1

Solvents
Spent
Pickle
Liquor
Corrosives
Pesticides
Wastewater
sludge from
the
chemical
conversion
coating of
aluminum
Totals
28
86.6
SITE MANAGEMENT & CLOSURE 243

-------
IMPOUNDMENT CLOSURE
  Once a facility has initiated process changes such that the sur-
face impoundment is no longer necessary, a major decision point
is reached. The owner/operator must decide to either clean-close
the impoundment or to close with waste or waste residues left in
place and provide post-closure care as for a landfill.
  Clean-closure is always the most desirable option. However, it
is an option not readily available in most cases. The hazardous
waste regulations are extremely specific in detailing what must be
accomplished  to clean-close a surface  impoundment (265.228).
To effect clean-closure, all standing liquids, waste and waste resi-
dues, any liner and all underlying and surrounding contaminated
soil must be removed. Since current  analytical techniques usually
can detect  waste components in the ug/1, complying  with  the
standard of total removal is extremely difficult. Further,  many
surface impoundments  have  demonstrated  contamination of
groundwater,  making the possibility of clean-closure even more
remote. The special case of a surface impoundment which only
received characteristic (i.e., ignitable, corrosive, reactive or  EP
toxic) waste is an exception. In this case, the owner/operator only
must demonstrate that no waste or waste residue continues to be
characteristically hazardous for  the  impoundment to be clean-
closed. In Georgia,  8% of the active  hazardous waste surface
impoundments managed characteristic waste only. Most of these
have effected clean-closure.

Closure Options
  The majority of surface impoundments in  Georgia received
listed hazardous waste. These facilities must meet the more strin-
gent closure  performance standard of removing  all waste and
waste residue to analytically non-detectable or background levels.
In most cases, this simply is not technically or economically feas-
ible, particularly since 92% of the listed waste surface impound-
ments  have demonstrated  groundwater contamination and  re-
moval of all the contaminated groundwater would be required.
This situation has  resulted  in nearly all of the facilities propos-
ing some manner  of closure-in-place  followed by post-closure
care and corrective action for groundwater contamination.

Closure-ln-Place
  Once a facility has accepted the reality of closure-in-place, sev-
eral other decisions must be made. A facility which closes a sur-
face impoundment with waste or waste residue in place must pro-
vide a minimum of 30  yr  of post-closure care, must receive a
hazardous waste disposal permit, must provide corrective action
for releases of hazardous waste or constituents and must provide
financial assurance for these activities. The initial  response of
most companies is to leave all waste in place, cap the impound-
ment and then initiate corrective action. This appears, on the sur-
face, to be the most inexpensive option, thus preserving resources
which  may be used during the post-closure care and corrective
action  period. However, when closely considered, this approach
has major disadvantages.
  The corrective action provisions of the hazardous waste regula-
tions (264.100) are very stringent. A facility must operate a cor-
rective action  program which will return  the regulated unit to
compliance with the established groundwater protection  stand-
ard. It is not sufficient  for the program to maintain the status
quo. If the  corrective action plan does not prove  effective, it
must be modified. If the surface impoundment is closed with the
wastes left in place yet continues to release hazardous  waste or
constituents to the groundwater, then a probable first modifica-
tion to the corrective action plan is to require the wastes  to be
exhumed. Leaving the wastes in place results in the maximum lia-
bility to the facility. The probability is fairly  high that at some
future date the wastes will have to be removed or otherwise
treated to alleviate groundwater contamination.
  At the other end of the spectrum, the facility owner/operator
may choose to remove all wastes and waste residue for off-site
disposal. Since this still leaves  contaminated groundwater in
place, the impoundment still must undergo post-closure care and
corrective action. However, since all contamination has  been re-
moved, the facility's liability is minimized and the effectiveness of
the groundwater corrective action  program is maximized. The
obvious disadvantage is  the extremely high cost of off-site dis-
posal of large volumes of hazardous waste  and contaminated
soil.
  Faced with the above  choices, several Georgia facilities  have
opted for closure activities which combine both approaches in a
cost-effective manner. The actual hazardous waste and the most
severely contaminated soil are removed to a disposal facility or
treated in place. Residual contaminated soil is treated to the ex-
tent possible and then stabilized to immobilize the remaining haz-
ardous constituents. The impoundment is then carefully back-
filled in controlled lifts, and an impermeable cap is established.
The closed impoundment then must undergo post-closure  care
and, in  conjunction with the  issuance of a hazardous waste dis-
posal permit, initiate a corrective action program.
  Although the above approach is much more expensive than
simply leaving the waste in place, it is far less expensive than re-
moving  all  contamination. Long-term liabilities at  the  site are
greatly reduced since the most severely contaminated material has
been treated or removed. The effectiveness of the groundwater
corrective action plan is maintained since care was taken during
closure to immobilize to the extent possible the residual contam-
ination which was left in place. Any further release of hazardous
constituents should  be eliminated or greatly reduced, thus mak-
ing  the establishment of an effective groundwater corrective ac-
tion program much easier.
  To effectively manage a complex closure as described above re-
quires close coordination between the owner/operator and the
appropriate regulatory agency. Costs must be closely audited to
determine at what point  removal or treatment  should stop  and
stabilization should begin. However, as the following case histor-
ies demonstrate, such an  approach to closure can be both effec-
tive and cost-efficient.

CASE HISTORIES
  Many of the active hazardous waste surface impoundments in
Georgia are used to treat  or store wastewater from wood-preser-
vation using creosote and/or  pentachlorophenol. This results in
the  generation of the listed hazardous waste K001.  This sludge
contains a wide variety of organic compounds  (Table 2). Since
most of the wood-treating plants have been operating for over 20
yr, a large amount of sludge has built up in the impoundments.
The estimated sludge quantity for Georgia wood preservers aver-
ages 2500 yd'/facility. Additionally, extremely large quantities of
contaminated  soil are present, usually 2-4 times as much as the
actual sludge.
  With very few exceptions, these facilities have extensively con-
taminated the groundwater, making clean-closure virtually im-
possible. The  facility must decide on the type of closure which
will be performed based on both cost and on potential for future
liability.
  K001  sludge is quite amenable to biodegradation.  However,
biodegrading  to  non-detectable levels is  not  cost-effective.
Neither is it cost-effective to treat large quantities of only slightly
contaminated  soil. The sheer quantity  of soil  involved creates
massive logistical problems and, since the soil remains K001 waste
until delisted, there is little to be gained by treating this material.
244    SITE MANAGEMENT & CLOSURE

-------
Faced with these facts, Facility A (Table 3) embarked on a
multi-phase closure designed to minimize future liability while
controlling costs. The facility initiated closure by installing with-
in the impoundment a lined treating cell and sufficient aerators to
effectively suspend the K001 sludge in the water column. Nutrient
and pH levels are monitored regularly, and nutrients are added
as necessary. The pond is inoculated regularly with conditioned
bacteria to help maintain a maximum bacterial biomass. This en-
hanced biodegradation system allows the facility to effectively
treat approximately 50 yd3 of sludge/wk. Sludge treatment results
in the generation of very few residual solids. Upon completion
of the sludge treatment phase, approximately 15 yd3 of solids
will remain to be shipped off-site.
Treatment of contaminated soils will begin following comple-
tion of the sludge treatment. Over 90% of the treated soil will re-
main as solids following treatment. This treated material must be
moved to a partitioned section of the impoundment to allow for
treatment of new charges of soil. The goal of the treatment pro-
cess is to continue treatment until the remaining soils have K001
constituent levels no higher than indicated in Table 2. This will
result in the treatment of approximately 3800 yd3 of contam-
inated soil. Upon completion of the soil treatment, the impound-
ment will be capped and the required post-closure care will begin.
The facility owner/operator has determined that treatment of
soils to Table 2 levels followed by post-closure care should min-
imize his potential future liabilities as well as prevent any future
releases of hazardous constituents to the groundwater. This
closure option results in a significantly higher cost than closure-
in-place (Fig. 1) but is dramatically less expensive than removal
of all contamination.
Closure at Facility A will require approximately 18 mo to com-
plete. The process is almost one-third complete, is currently on
schedule and is within the estimated costs.
Facility B is also a wood-treating plant conducting a closure
very similar to that at Facility A. The impoundment at Facility B
has a larger surface area but contains significantly less waste than
Facility A (Table 3). Closure here also is proceeding on sched-
ule and within projected costs (Fig. 1). Upon initiation of closure
at Facility B, it was discovered that the entire impoundment is
Table 2
K001 Sludge Parameters and Treatment Levels Proposed by Facility A
Parameter
Pentachlorophenol
2-chlorophenoI
Phenol
2,4-dimethylphenol
2,4,6-trichlorophenol
p-chloro-m-cresol
Tetrachlorophenol
2,4-dinitrophenol
Naphthalene
Acenaphthene
Phenanthrene
Anthracene
Fluoranthene
Chrysene
Benz (a) anthracene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo (a) pyrene
Indeno(l,2,3,-cd)pyrene
Dibenzo (a,h) anthracene
Carbazole
Proposed Levels (ppm)

10
20
20
20
20
20
10
20
20
20
20
20
20
20
20
20
20
20
20
20
20
underlain by a shallow clay layer through which contamination
apparently has not penetrated. Thus, Facility B may well be able
to effectively treat and remove all contamination and thus per-
form a clean-closure.
In contrast to Facilities A and B, Facility C operated a syn-
thetically lined surface impoundment for fewer than 5 yr. The
impoundment managed a wastewater high in chrome content
(Table 3). Routine inspections of the impoundment liner docu-
mented tears, and the facility decided to close the impoundment
before any groundwater contamination could occur.
Table 3
Case History Facilities

Impoundment Volumes (Cubic meters)
Facility Process Size (ha) Age (years) Lined Waste Soil
A
B
C
D
Wood
Preservation
Wood
Preservation
Magnetic Tape
Manufacture
Pipe
Manufacture
0.7
2.8
0.3
0.1
23
12
5
7
No
No
Yes
Yes
4000
2300
600
200
6800
4600
50
400
1m' = 1.3yd'
$x105
30r
25
2Q
15
10.
5
ol _J
•-CLOSE-IN-PLACE

^-TREAT A CAP

[""I - REMOVE
A BCD

FACILITY

Figure 1
Closure cost estimates from four hazardous waste surface impound-
ments in Georgia. Included are comparative estimates for closure-in-
place, closure by treatment/removal followed by closure-in-place and
closure by removal.

Since almost no waste or waste constituents had migrated to the
soil and no groundwater contamination had occurred, Facility C
opted for the most expensive closure option, total removal of all
waste and waste residues (Fig. 1). This recently completed closure
SITE MANAGEMENT & CLOSURE 245

-------
demonstrates that clean-closure is a possibility, although a costly
one. The circumstances at Facility C are  uncommon since the
impoundment was synthetically lined, was operated  for only a
short time  and was used  to manage only a characteristic haz-
ardous waste (providing significant contamination of surrounding
soils and water did not occur).
  Facility D also operated a small synthetically lined impound-
ment for the management  of reactive wastes. Closure  of this im-
poundment involves the in situ treatment of the waste for a 90-
day  period followed by excavation of all  remaining  hazardous
waste. Again, since only a characteristic waste  is involved, the
facility most  likely will effect a clean-closure. However, if con-
tamination is more extensive than initially determined,  closure-in-
place may be required. As  Fig.  1 indicates,  there is a striking dif-
ference in costs between closure-in-place and removal of all the
waste. The treatment followed  by the removal option once again
is intermediary in cost.

REGULATORY CONSIDERATIONS
  Accomplishing a complex closure of a  surface impoundment is
technically and  logistically difficult.  However, the  regulatory
quagmire which can be encountered can dramatically increase the
difficulty. The regulations adopted under  RCRA  are extremely
complex. In  some places  they are excrutiatingly precise,  yet in
other areas they are distressingly obtuse. This problem with the
Federal regulations is further  complicated by state regulations
which may be more stringent than the Federal regulations or,  if
not  more stringent, are simply different. A  close and effective
working relationship between the owner/operator and  the appro-
priate regulatory agency must be created to effectively manage a
complex closure. Clear and direct lines of communication must
be opened  and maintained so  that there are no surprises  which
could endanger the completion of the closure.
  Several regulatory milestones must be  passed before and after
a hazardous waste surface impoundment is certified closed. The
first is the filing of an  acceptable closure  plan with the regulatory
agency.
  The RCRA regulations at 265.113 specify that all closure activ-
ities must be completed within 180 days  of initiation of closure.
Obviously, the types of closures discussed  in this article can re-
quire substantially  longer than  180 days to complete. A variance
to allow closure to encompass more than  180 days may  be granted
provided that the owner/operator documents the necessity of a
longer closure period. The owner/operator must provide precise
and compelling reasons for a longer than 180 day closure period.
Again, open lines of communication between the owner/operator
and  the regulatory agency  are necessary. The regulatory person-
nel  should  be familiar with the type of  closure that the facility
plans to conduct and the potential time schedule. Thus, when the
plan is submitted, the agency already should be prepared to con-
sider a longer time schedule than is common.
  Few closure plans are ever approved without modifications
requested by the regulatory agency. If open lines of communica-
tion are maintained, the agency and  the owner/operator should
be able to quickly and smoothly make any required changes in
the closure plan. However,  if any innovative or unusual treat-
ment or closure activities are proposed, the burden of proof will
rest on the  owner/operator. Regulatory agencies by their very
nature are conservative. The owner/operator must be prepared to
demonstrate with hard data the effectiveness of any proposed
closure techniques.
  Once a closure  plan  acceptable to the regulatory agency is
developed, the plan must be placed on public notice as required
by 265.112.  A major provision of the  HSWA is an expanded
public  participation program. The public may comment on the
technical adequacy of the closure plan and request modifications.
If the owner/operator and the regulatory agency have completed
adequate preparation and review, there should be no technical
objections to the closure plan from the public. In Georgia none of
the land disposal facility closure plans which have been placed
on public notice have required modification based on public com-
ments.
  Upon completion of an interim status closure plan which leaves
waste or waste residues in place,  an  interim status post-closure
care plan must be implemented. This  plan is to be conducted un-
til issuance of the actual hazardous waste permit specifying post-
closure care. As a practical matter in  Georgia, we attempt to co-
ordinate issuance of the post-closure care permit and completion
of the actual closure activities in a manner that does not require
an interim status post-closure care period. This goal necessitates
completion of the permit decision 45 days prior to the comple-
tion of closure. Once again, a clear  line of communication be-
tween the affected industry and the regulatory agency is impera-
tive. Given a good relationship, the transition  from an operating
surface impoundment  to a permitted, closed impoundment
undergoing post-closure care can be smooth.

CONCLUSIONS
  Cost-effective and environmentally sound closure of hazardous
waste surface impoundments are complex and require careful
planning and decision-making. Close  coordination  with  the
appropriate regulatory agency is imperative. The rules governing
these  closures are  complex  and  require very specific actions.
Thorough planning and  communication with the  regulatory
agency can result in final closures which meet the regulatory re-
quirements, minimize future liabilities are are cost-efficient.
246     SITE .MANATEMENT & CLOSURE

-------
A Successful Hazardous Waste Landfill Siting-
Maryland's Experience
Thomas D. McKewen
Recovery Associates
Annapolis, Maryland
Anne C. Sloan
Department of Natural Resources
Annapolis, Maryland
ABSTRACT
Maryland took an activist approach to hazardous waste man-
agement by supplementing its regulatory program with a pro-
gram designed both to seek out and possibly develop sites for
new treatment, storage and disposal facilities and to ensure that
suitable sites could not be vetoed by local governments.
A four-level screening and evaluation process and subsequent
site selection are described along with accommodations reached
with the host community to allow the project to move forward
without pre-emption of local decisions. Some conclusions devel-
oped in this process are offered to others who may confront a
need to site a hazardous waste management facility.

INTRODUCTION
To our knowledge it is the only new, commercially available
hazardous waste landfill to be sited since RCRA's passage in
1976. It received the first "Part B" permit issued by the U.S.
EPA to any hazardous waste landfill in the country. It was oper-
ating a mere 27 mo after the need for it was formally enunciated
and less than 2 yr after a site was selected. It was developed with-
out citizen litigation.
"It" is the Hawkins Point Hazardous Waste Landfill in Balti-
more, Maryland. Siting and development of this facility arose
from circumstances in some respects unique to Maryland but in
other ways typical of those found in most industrial states. The
decision-making process leading to development of the project,
while complex, was more orderly than the very short time sched-
ule (and some authors who have written about it) might suggest.
This paper describes the process in some detail and attempts to
draw some lessons which would be helpful to other states facing
the need to site new hazardous waste management facilities.

BACKGROUND
Early in 1976 the Maryland General Assembly enacted legisla-
tion aimed at controlling hazardous industrial waste. The legisla-
tion, prefiguring requirements of the Federal Resource Conserva-
tion and Recovery Act, was regulatory in nature and included
features such as manifesting and record keeping for waste gener-
ation and shipment and permitting of waste management facili-
ties. Prior to that time Maryland (like other states) had largely
ignored industrial solid waste, leaving sole responsibility for man-
aging it with the private sector. Partly because industrial waste
disposal was not considered the responsibility of public agencies,
unlike residential and commercial solid waste, relatively little was
known about the solid wastes produced by Maryland industries.
By the end of the decade, what was known, or at least felt, by
many industrial operators and state officials was that a crisis of
off-site hazardous waste management capacity was looming.
Some of the state's major hazardous waste generators urged an
active role for the state in assuming that commercial hazardous
waste facilities remained available locally.
During the fall of 1979 legislation was developed with the dual
purpose of clarifying the authority of the Maryland Environ-
mental Service (MES) to offer hazardous waste management serv-
ices and of providing a mechanism to overcome perceived local
government veto power over the siting of new hazardous waste
management facilities. The legislation was developed by staff
from several state agencies with input from the Maryland Asso-
ciation of Counties, which supported its passage, and the Mary-
land Chamber of Commerce. The Governor's office played a
leadership role.
A note on the state's organization may be necessary here.
Under Maryland's cabinet system, since 1980, principal environ-
mental regulatory authority has been vested in the Department
of Health and Mental Hygiene. The Maryland Environmental
Service is an agency of the Department of Natural Resources.
MES was established in 1970 to provide water supply and waste
management services for both liquid and solid wastes to munici-
palities or private interests anywhere in the state. In this capac-
ity MES may plan, design, build and/or operate facilities. It has
no regulatory role. It presently operates many water supply and
waste management systems which are regulated by the Health
Department.
Impetus to passage of the siting legislation was provided by
the prospect of closure of the state's only hazardous waste land-
fill. Broad interest in the problem of hazardous waste was pro-
vided, at the same time, by the discovery of an abandoned tank
farm containing PCB-laced materials leaking on the banks of an
Eastern Shore river. The difficulty of finding storage and disposal
services for these wastes, which required immediate removal,
weighed on officials in numerous state agencies. It also attracted
substantial media coverage.
Given the political power of the counties in Maryland—the
basic units of local government—and the well-established tra-
ditions of local autonomy in matters related to land use, the
Hazardous Waste Facilities Siting Program Act passed with re-
markable ease in the 1980 session of the General Assembly. The
SITE MANAGEMENT & CLOSURE 247

-------
Act provided for establishment  of an eight-member indepen-
dent Siting Board with power to issue a certificate of public neces-
sity for new hazardous waste facilities, in effect overcoming zon-
ing or other local land use restrictions.  The Maryland Environ-
mental Service, which at that time staffed the Board, was to con-
duct a study of types and quantities of wastes generated in Mary-
land (MES  is no longer staff to the Board,  and the Board now
conducts  the type and quantity  estimate independently). MES
also was to develop an inventory of potential sites for new haz-
ardous waste management facilities. The authority of MES to
develop and operate hazardous waste management facilities was
established almost parenthetically.
   An  appropriation  of  $300,000 was  made  to carry out  the
"needs" study and compile the site inventory; two staff mem-
bers of MES were assigned full-time to the two projects,  with
other staff made available as needed. Since a deadline of July I,
1981 was specified in  the legislation for completion of the inven-
tory, it was necessary to carry out the two studies simultaneous-
ly, rather than  first establish a specific facility  need and then in-
itiate the  siting activity. About $150,000 was budgeted for each
study.

THE SITING PROCESS
   One provision of the siting legislation required MES to estab-
lish guidelines  for placing sites on the inventory. Formulation
of these guidelines became the important first step in the siting
process, which got underway with consultant  selection in the early
fall of 1980. A series  of in-house workshops involving personnel
from  several state agencies and four  county planning depart-
ments preceded two Saturday workshops to which members of
numerous interest groups (e.g., farm bureaus, watermen's asso-
ciation, League of Women Voters and environmental groups)
were invited. In all of these workshops, structured presentations
and discussions were designed  to  elicit participants'  thoughts
about what characteristics made sites acceptable for hazardous
waste  treatment or disposal and what characteristics made them
less desirable or flatly unacceptable.
   The guidelines which  resulted from  this  process involved a
four-level screening and evaluation procedure designed to yield
perhaps a dozen sites considered worth further consideration.
The first and third levels in the process provided for eliminating
geographic areas based on the presence of undesirable features,
the second level was intended  to highlight desirable characteris-
tics and the fourth to evaluate specific candidate sites. All screen-
ing was to be based on existing published and file data, rather
than extensive field work.
   The draft guidelines were submitted to local governing bodies
for comment and were made available for comment by interested
members of the public as well. Open-house style public meetings
were held in two different locations in the state early in the win-
ter of 1981.  At the open houses, both material being developed in
the needs  study and the draft guidelines were available  for ques-
tion and comment, together with the responsible MES  staff and
two consultant  teams. There was  very little public interest in the
process at this time—one major metropolitan newspaper even re-
fused a specific request to publicize the open  meetings through
a news-feature story. Comment on the draft  guidelines  was min-
imal; in March, the final guidelines were published and screen-
ing was initiated.
   Early work on establishing the types and  quantities of Mary-
land generated  waste, the "needs"  study, was suggesting by
March that the type of facility needed most immediately in Mary-
land would  be a landfill, rather than some sort of treatment or
storage facility. In Maryland and its immediate neighbors there
did not appear to be an existing or anticipated shortfall of capac-
ity to treat or incinerate wastes for which these were the prefer-
able management options. The continued operation of existing
hazardous waste  landfills and their capacities appeared  more
dubious.
   Thus the siting  study was restricted to searching for a landfill
site using the more stringent of the criteria developed in the guide-
lines, which made a distinction between above ground and land
emplacement  facilities. A site  of about ISO acres was deemed
necessary to  ensure capacity to meet about 10-years' land dis-
posal need. This need was significantly driven by the fact the
state's largest generator requiring off-site land disposal was Allied
Corporation, which produced about 100,000 tons/yr of spent ore
from its chrome chemical works in Baltimore.
   The first level siting criteria was designed to eliminate  from
consideration large areas clearly unsuitable on geologic or other
environmental grounds. Application of criteria on some ten char-
acteristics eliminated about one-half the land area of the state, in-
cluding nine entire counties. Four Level II characteristics, in-
tended to be positive features, contributed very little to the pro-
cess. At this point, a substantial number of large (several square
miles) areas remained as search areas, constituting perhaps one-
third of the state's land area.
   Screening through Level II was  very coarse; data used  were
mapped  at a  scale of  1.0 in. to  4 mi,  so only major environ-
mental features or geological formations mappable at this scale
were considered. Obviously, many  "excluding" characteristics
were not mapped at this scale, chief among them was existing
development.  An additional 15 excluding criteria were provided
in the third level of screening, including most local environmen-
tal, social and land use characteristics.
   In order to check the accuracy of work on Levels I and II, and
also to gather data from local sources to go on with Level in, a
series of review sessions was undertaken with local government
staff during the late spring of 1981. Reviews were scheduled in
each of the 14 counties and Baltimore City which still remained
in the search area.  Open public review of the work was not sched-
uled for this midway point.
   At the scheduled review in Cecil County, County Commission-
ers and a reporter for a local paper appeared along with the ex-
pected county planning and public  works people. In this partic-
ular county, there were six areas remaining after Level II screen-
ing, ranging in size up to about 20 mi1 in size. Despite their large
size, the potential  areas were promptly labeled "sites" in the pa-
pers and immediately became the subject of considerable public
furor, in part fueled by a previously unsuccessful candidate for
public office.  The culmination was a march on the State House
during the summer of  1981. Demands were made that  these
"sites" be removed from consideration, and MES personnel were
lambasted in public meetings for their incompetence in not  con-
sidering such obvious facts as the existence of a small town in the
middle of one of the "sites." (Remember, existing development
was to be excluded in Level III. The town appeared as only a dot
on the 1 in. to 4 mi maps used up to this point.) Interestingly, the
one major mapping error of the whole Level  I-Level II effort
occurred in this county. This error was noticed by neither the
local people at the review meeting nor by the incensed public; it
was found by project staff.
  As a result of rumors and newspaper coverage arising from the
Cecil County  review meeting, MES staff decided  to reproduce
and distribute reduced black-and-white versions of the county
maps which were the output of the first two phases of screening.
Two more counties also saw angry protests during the summer;
like Cecil, these counties were quite remote from  the center of
hazardous waste generation and thus less likely to receive serious
consideration  in the final evaluation. MES personnel attended
248    SITE MANAGEMENT & CLOSURE

-------
several meetings in  all three  counties during the summer; the
Governor's office cancelled the plans for public review of the
Level III screening which was called for by the guidelines for fear
public anxiety levels would be too high.
  Review of Level III results with local government staff also was
stopped. This meant that selection of candidate sites for evalua-
tion in Level IV did not benefit from any local input, as had been
intended. A better set of candidate sites might have been selected
with local assistance.
  The Level III screening results were  not what had been antici-
pated when the siting process was established. Instead of a rela-
tively few large areas from which to carve out a dozen or so candi-
date sites, there were over 100 small areas, almost site-sized in
themselves, scattered over 11  counties and Baltimore City, This
situation  resulted largely from the way in which the "already
developed areas" factor was  mapped. During Level III  screen-
ing, ribbons of very low-density residential and commercial devel-
opment were found, and excluded, along most county roads
throughout the state—to the embarrassment of some local plan-
ners, there were very few areas in which this development pattern
was not found. Some kind of density  criterion might have been
a more useful one to employ to rule out areas where a hazardous
waste landfill  would adversely impact real neighborhoods  or
major commercial activities—areas  too large to be purchased as
a part of site acquisition.
   Given these  Level III results, public review would have been
difficult and, indeed, contentious. However, the published siting
process included in the guidelines had called for public review at
this point; cancelling a series of regional meetings scheduled for
September tarnished the credibility  of  the planning process. The
ill effects probably would have been more damaging if the whole
 study had not reached a conclusion relatively soon thereafter with
 selection of a site in an area where social and environmental im-
pacts were relatively low.
   Even before public uproar over  the partial screening results,
 project staff at MES and members of the Hazardous Waste Facil-
 ities Siting Board had come to realize the potential of an open-
 ended site inventory, as called for in the Siting Act, for causing
 long-term public anxiety. Once sites had been publicly identified
 as potentially acceptable for hazardous waste management facili-
 ties, what would happen to them and  perhaps more importantly
 to other property nearby? There was no authority or funding to
 acquire multiple sites to  assure their  availability if need arose.
 There was no funding even to perform detailed on-site investiga-
 tions to prove the acceptability or unacceptability of one or more
 of the candidate sites.
   During the spring of 1981,  the Siting Board was also growing
 concerned that no applications for additional landfill capacity ap-
 peared to be forthcoming from private waste management firms
 and that the existing hazardous waste landfill, nearing termina-
 tion of its permit and expiration of its zoning approval, would
 very soon cease operation. In May therefore, in accordance with
 a provision of the siting law, the Siting Board directed MES to
 select a site and develop it. The Board also believed that a state-
 developed facility would provide a yardstick against which to
 judge subsequent private applications. The Board's directive pro-
 vided the desired closure to the site inventory process (although
 getting a candidate site "off the list" remained a concern of some
 counties, once the report identifying the  14 candidates was re-
 leased, even after site selection was announced).
   Of the  14 candidate sites given Level IV evaluation, most were
 in the Baltimore metropolitan area; only  two real outlyers were
 included, in order to see what kinds of social and  environmen-
 tal impacts would be associated with a  site selected on the "keep-
 it-away-from-people" philosophy. The candidates were rated as
positive-neutral-negative on 26 characteristics. It was not possible
for a site to be rated positive on all the Level IV factors, since cer-
tain desirable features, such as availability of public water supply
and distance from residential and commercial land uses, would
not normally be found together.

SITE SELECTION
  Since one of the positive attributes identified in Level II was ex-
isting use of an area for waste disposal, it was not surprising that
one of the  candidate sites selected for evaluation  in Level IV
was the Hawkins Point area at the tip of Marley Neck in South
Baltimore. Two privately owned industrial waste landfills were lo-
cated there  along with a landfill for  Allied's spent chrome ore
operated since 1980 by MES on land owned by the Maryland Port
Administration.  The commercial landfill at Hawkins  Point had
been closed prior to reaching its ultimate capacity; it was believed
to contain hazardous wastes of unknown description and was
built on—or in—what was denoted on old maps as Chemical
Lake.  The adjacent  captive landfill  took nonhazardous  waste
from a nearby chemical plant. (At one time a plan had been devel-
oped, and then abandoned, to combine these two landfills into a
single  operation; a  considerable amount of disposal capacity
would  have  remained if this joint project were to be developed.)
Leachate from both of these landfills passed under a railroad spur
in two culverts and combined in a small creek crossing the Port
Administration property on its way to Thorns Cove on Baltimore
Harbor.
  Across the road from the two private landfills was a  large tract
of undeveloped  land owned by  the adjacent chemical plant. It
proved to be less suitable for hazardous waste disposal than the
remainder of  the area because its native clay had been largely
stripped. This part of Hawkins Point is now being developed as a
sanitary landfill by Baltimore City.
  The  whole Hawkins Point area is bounded on the south by the
Baltimore Beltway as it approaches the outer harbor crossing on
Key Bridge. On the  other side of the Beltway was a  small resi-
dential  community consisting of 22 houses and a church devel-
oped by a close-knit group of blacks in the early decades of the
20th century.  The whole  area is largely isolated from the re-
mainder of  Baltimore City by Curtis Bay and Curtis Creek, is
crossed by a drawbridge and historically was part of Anne Arun-
del County. The Anne Arundel County portion of Marley Neck
has fairly sparse residential development, primarily on  the water-
front, and large tracts of land slated for eventual industrial devel-
opment.
  As industrial development had occurred on Hawkins Point, the
residents of the  community had increasingly been assaulted en-
vironmentally. Their undesirable neighbors  included chemical
plants, a derelict park and an auto and truck graveyard. They also
had been largely overlooked by both the city and state. Their list
of grievances was long and legitimate. MES had become  familiar
with them through meetings and conversations incident to opera-
tion of the chrome ore landfill.
  Of the sites identified by the siting study, Hawkins Point of-
fered considerable advantages:

• It was in Baltimore City, the center of hazardous waste gener-
  ation in the state, thus  addressing  the equity  issue  as well as
  minimizing waste transportation with its attendant dangers and
  cost.
• Despite the presence of the small residential community, the en-
  tire area was heavily industrial in character, a character we be-
  lieved more compatible with  hazardous  waste management
  than other land uses. Additional industrial development was
  contemplated in local land use plans and the alternative to use
                                                                                     SITE MANAGEMENT & CLOSURE    249

-------
  of the site for hazardous waste management was some form of
  industrial development.
• By expediting construction of the rest of an interchange with
  the Beltway, virtually direct access from the landfill to a major
  highway could be provided, thus avoiding both city streets and
  suburban or country roads.
• The area was served by Baltimore water supply,  drawn from a
  distant surface water source, a backup precaution in the event
  the landfill were to leak and the leachate collection/treatment
  system fail.
• Because of the amount of subsurface data already available as
  a result of pre-existing permit applications, there was more con-
  fidence in the geologic situation than for any other site.
• Development of the 290-acre  Hawkins Point area in a unified
  way offered the potential to clean up an existing groundwater
  contamination problem, perhaps by building a treatment plant
  for leachate from the entire area.
• A part of the site was already  in state ownership  and was oper-
  ating as a hazardous waste landfill. Operation here could be ex-
  panded as the first phase of  a phased development. In hind-
  sight this is more crucial than it was at the time, when no one
  realized that delay in the U.S.  EPA's regulations  would make it
  impossible to develop a totally new site until much later.

  In the fall of 1981 discussions were opened with Baltimore
and state officials  regarding the development  of a commercially
available hazardous waste landfill at Hawkins Point. Bargaining
with the city resulted in a number of agreements to  allow the pro-
ject to move forward:

• The only portion of the site which  would be used for a haz-
  ardous waste landfill was the land already owned by the Port
  Administration.
• The site would be used for only 5 yr, rather than  the 10 yr orig-
  inally desired.
• The project would pay the city a $5 royalty for every ton of dis-
  posed waste.
• The state would provide a major restoration and redevelop-
  ment of Fort Armistead Park.
• No out-of-state wastes would be accepted.

  Separate discussions were undertaken by MES with the Haw-
kins Point Improvement Association regarding the residents' con-
cerns about expanded use of the Port Administration property.
Residents expressed a number of druthers, which  MES tried to
accommodate:

• That residents be considered for employment at the site
• That heavy equipment and debris left by  highway construction
  crews be removed
• That suspected hazardous dumpsites nearby be investigated
• That groundwater and garden soil in  the neighborhood be
  tested for contamination from existing waste disposal facilities
  in the neighborhood
• That the air be monitored  for pollution from nearby industrial
  sources and that something be done about damaging fall-out
  from these sources
• That a separate royalty payment  (set at $l/ton of waste re-
  ceived) be made to the residents, to do with as they wished

  MES was moving to  meet these requests when announcement
of the site's selection was made in October of 1981. Some pro-
gress was made on most  of these  issues during the  ensuing
months, until residents' attention became totally focused on the
possibility of having the city or  state buy their homes. This pos-
sibility was first raised by a  local member of the General Assem-
bly early in the winter of 1981-1982; the homes have, in fact, been
purchased and the population has been relocated.
SITE DEVELOPMENT
  In December 1981 the Board of Public Works, comprised of
the Governor, Treasurer and Comptroller of the Treasury, pro-
vided initial funding from the Resource Recovery Loan Act of
1974 to design and construct the new general hazardous waste cell
at Hawkins Point. It was intended that project revenues would
repay this initial outlay with the exception of funds used for re-
medial action on the Port Administration property.
  During the winter of 1981-1982 project development began in
earnest, with simultaneous  subsurface investigation, design and
environmental assessment. MES and Governor's office staff and
consultants met regularly with the Hawkins Point Improvement
Association throughout this process, arranging presentations and
site visits as needed. One point which MES discussed early with
residents was the question of whether MES or a private contrac-
tor should  operate the facility once it  was built and  accepting
waste. Residents expressed a desire that state employees operate
the facility, in the belief they would be more responsive to local
needs than  would employees and management of profit-seeking
private enterprise.  This  preference figured in the decision that
MES would be the operator.
  Since Baltimore City gave the necessary  land use approvals to
the project, it was not necessary for the Hazardous Waste Facil-
ities Siting Board to rule on  whether to give the facility a certif-
icate of public necessity. A  formal public meeting was held to
conclude the siting study in January of 1982, and a public hearing
on the facility permit was conducted by the  Health Department in
May of 1982. A suit by Browning Ferris Industries held up issu-
ance of the permit until Nov. 30,1982.
  Actual development of the Hawkins Point Landfill was not un-
eventful. Some subsurface conditions proved more troublesome
than anticipated; U.S. EPA regulations regarding expansion of
interim status facilities changed during the design phase of the
project; there were contractual constraints on use of some of the
Port Administration's property; design  standards for hazardous
waste landfills were evolving and  becoming more stringent, and
MES  considered itself morally, if not strictly  speaking legally,
obligated to incorporate higher standards in the Hawkins Point
design; the major  user of the facility.  Allied  Corporation, in-
sisted on keeping its portion of the facility totally separate, in
management and operation as well as physical configuration,
from  the general hazardous  waste portion. All of these factors
contributed to a higher-than-anticipated price tag for disposal at
the facility, which opened in July 1983.
  The general hazardous waste cell was designed to take about 1
yr hazardous waste generated in Maryland for which landfill was
the only feasible disposal method, an amount estimated in 1981 to
be about 40,000 tons. Liquids had long been excluded from land-
fills in Maryland and were  not  anticipated  at Hawkins Point;
other organic wastes were not planned for either. As it happened,
the  facility was not capable of handling drummed waste as initial-
ly constructed—this may have prevented some Maryland genera-
tors from using it. For less understood reasons, other wastes did
not find their way to Hawkins Point either. Cost, at nearly $100/
ton, was certainly a factor in the facility's non-use. Delay in open-
ing  Hawkins Point also may have played a role by causing some
generators  to enter contracts with out-of-state disposal firms.
After a year of losing money the general hazardous waste cell was
closed. The portion of the site dedicated to disposal of Allied's
chrome waste continued  operating until the chrome works was
closed.

CONCLUSION
  Despite its short operational life we consider the siting of the
Hawkins Point hazardous waste landfill a successful one, in part
250     SITE-MANAGEMENT & CLOSURE

-------
because no extraordinary measures were required to overcome
opposition. Although some observers considered it disappoint-
ing that the siting did not have to be resolved by the Hazardous
Waste Facilities Siting Board, we would suggest that the ability to
work out a solution with the host community, rather than appeal
to a pre-emptive state authority, is a measure of the success of
the effort.
  We learned, or relearned, a lot from the siting  and project
development effort, which we would summarize as follows for
other public agencies facing a similar endeavor:
• The need for a facility of a particular type should be established
  prior to initiating a siting study and this need should be widely
  recognized.
• As indicated earlier in this paper, there has to be closure to a
   siting effort—an open-ended inventory has too high a price in
   public uncertainty and anxiety to be worth any potential bene-
   fits. Additionally the entire process must be clearly laid out and
   understood by all participants and observers. Flexibility to
   change in response to public needs or changed circumstances
   should be included and publicly acknowledged.
 •  One "favorable" siting criterion, presence of an existing waste
   management operation, proved to be a drawback in  actually
developing a site  under real-world regulatory conditions in
which the greatest imperative seems to be to prove you didn't
cause degradation.
• While a deadline for siting may help to keep studies on track,
  more time and staff than were available to MES are necessary
  to carry out public education and response  activities with the
  desired flexibility.  Even 16 to 18 months following consultant
  selection  will seem to many people to be rushing things, pos-
  sibly "cramming something down someone's throat."
• Most of the public is used  to reacting to  decisions already
  made, rather than  helping to establish the direction of a study
  —e.g., the siting guidelines—or monitoring  progress and sug-
  gesting mid-course corrections. There must be ways to involve
  more people at these points in a project's development, but we
  did not find them.
• A careful and well-documented siting  study can substantially
  reduce the  time  required  for  post-decision  environmental
  assessment, although it will not meet public demands for an en-
  vironmental impact statement.
• Need for a facility, according to public policy dictates, and a
  market for that facility are  not synonymous. Public subsidy
  may be necessary to carry out public policy preferences.
                                                                                      SITE MANAGEMENT & CLOSURE    251

-------
                   Permitting  for  Hazardous Waste Incinerators-
                              Preparation of  the Trial Burn Plan

                                              Robert P. Newman, P.E.
                                                 Arthur B. Nunn, III
                                       Scott  Environmental Technology, Inc.
                                          Environmental  Services Division
                                            Plumsteadville, Pennsylvania
ABSTRACT
  This paper  summarizes information and technical considera-
tions important to the development of a trial burn permit appli-
cation for a hazardous waste incinerator.  The intent is to sum-
marize the most  important elements  of the application and to
identify areas which are potentially  troublesome.  The specific
regulatory requirements and operational stipulations also are re-
viewed.
  The paper addresses not only the sampling and analytical por-
tion of a trial burn plan, but also requirements for describing the
incinerator and control equipment design, waste  characteriza-
tion, process monitoring and the establishment of operating con-
ditions.  The overall objective is to provide a usable framework to
assist environmental managers and plant  personnel responsible
for the permitting of hazardous waste incinerators.

INTRODUCTION
  The U.S. EPA has established regulations under RCRA per-
taining directly to hazardous waste incinerators. The regulations
require  that new and existing hazardous waste incinerators de-
stroy hazardous organic compounds and comply with specific lev-
els of paniculate  and chloride emissions. To demonstrate compli-
ance with Federal regulation, owners and operators of such in-
cinerators are required to perform a trial burn test. Prior to actual
testing,  a comprehensive trial  burn plan is submitted to the regu-
latory agency for review. The trial burn plan is a  critical com-
ponent of the overall permitting program. The plan defines  the
design  of the  incinerator and proposes  operating conditions,
waste characterization, test burn sampling and analysis proced-
ures and process monitoring requirements.

REGULATORY  BACKGROUND
  RCRA established a national regulatory program to control the
disposal of hazardous wastes. Specific standards for incinerator
performance was promulgated on Jan. 23, 1981, and amended
on June 24, 1982. The specific standards provide that the prin-
cipal organic hazardous constituents (POHCs) designated in each
waste feed must  be destroyed and/or removed  to an efficiency
of 99.99%, that particulate  emissions  must not  exceed  180
mg/dsm1 corrected  to 7% oxygen in the stack gas and that gas-
eous hydrogen chloride (HC1) emissions must be reduced either
to 1.8 kg/hr or a removal efficiency of 99%. The regulations also
specify a number of requirements for  waste analysis and inciner-
ator operation, monitoring and inspection. Finally, they establish
the procedures by which permits to hazardous waste incinerators
will be granted.
  For permitting purposes, incinerators are classified as being
new or in  "interim  status." Incinerators designated as being in
interim status must  have been  in existence on Nov.  19, 1980.
Provided the proper permit applications have been filed (RCRA
Part A), interim status incinerators are allowed to operate prior
to the issuance of the final RCRA permit. For such existing facil-
ities, the regulatory agencies decide when the trial  burn plan and
final RCRA Part B application are to be submitted.
  If an existing facility in interim  status is authorized to burn
hazardous waste, the facility technically needs no prior approval
to conduct a trial burn. However,  without prior  approval of a
trial burn plan, it is not known if trial burn data will be sufficient
to meet regulatory requirements.  It is  strongly recommended,
therefore,  that trial  burn plans always be prepared for interim
status incinerators.
  Because RCRA allows existing incinerators to operate under in-
terim status while awaiting the Agency's decision concerning per-
mit issuance, these facilities are not subject to the operating re-
strictions which complicate the permitting process  for new incin-
erators. Prior to construction of a  new incinerator, owners and
operators of the new units who will conduct a trial burn are re-
quired to submit a trial burn plan with the permit application.
The application will  be processed through all of the required ad-
ministrative procedures including preparation of a draft permit
and opportunity  for public comment and hearings. After com-
pletion of this process, a permit that establishes all of the con-
ditions needed to comply with all applicable standards will  be
issued. This permit will be the "finally effective RCRA permit"
required for construction of the incinerator.
  The permit will be structured to provide for four phases of the
operation. Operating conditions will be specified for each phase.
The initial phase begins immediately following completion of con-
struction.  During this  phase,  the  unit may be operated for
"shake-down" purposes in order to identify possible mechanical
difficulties and to ensure that the  unit has reached operational
readiness and has achieved steady-state operating conditions prior
to conducting the trial burn.  This phase of the permit is limited
to 720 hr of operation using hazardous waste feed (one additional
period of up to 720  hr may be allowed for cause). Note that this
does not limit burning of non-hazardous wastes or fuel.
  After timely and satisfactory completion  of all shake-down
operations, the second phase of the permit begins. This phase
consists solely of the period  allotted to conduct  the trial bum.
Following completion of the trial burn, a period of several weeks
to several months will be necessary to complete and submit the
trial burn results. During this period, which represents the third
operational phase of the permit, the facility may continue to oper-
ate under specified operating conditions.
  Detailed review of the trial burn results will show either that the
incinerator is capable of complying  with the performance stand-
ards when operating within  the trial burn conditions or that com-
 252     INCINERATION

-------
pliance was not attained during the trial burn and a second test is
necessary. If compliance was shown, the permit may be modified
to set, as the final operating  requirements,  those conditions
demonstrated during the trial burn. If compliance has not been
shown and an additional trial burn is necessary, the permit must
be modified to allow an additional trial burn. When  all permit
modifications are complete, the facility begins its fourth and final
operating phase which continues throughout the duration of the
final permit.

TRIAL BURN PLAN
  The purpose of the trial burn plan is to provide a clear delin-
eation of: the incinerator  system;  the nature of the hazardous
waste; sampling and analytical techniques; and proposed operat-
ing conditions for the  incinerator  and control  equipment. The
trial burn plan is the most critical part of the RCRA application
package. The plan not only identifies how the incinerator will be
tested, but also specifies future operating conditions which will
become part of the eventual RCRA permit.
  The trial burn plan basically consists of six distinct compon-
ents (Table 1). Each component will be addressed in the follow-
ing sections.

INCINERATOR AND CONTROL SYSTEM
DESCRIPTION
  The more detailed the description of the physical incinerator/
control equipment system, the less time will be spent in supply-
ing supplemental information to regulatory agencies. The items
critical to the discussion in this section include the following:
 • Type of incinerator (rotary, fluidized bed, etc.)
 • Manufacturer's name, model numbers of principal compon-
  ents
                           Table 1
              Major Components of Trial Burn Plan

  1 Incinerator and control system description
  1 Waste characterization and POHC designation
  1 Provisions for sampling and monitoring of incinerator process
  1 Test schedule
  1 Test burn protocol
   QA/QC
  1 Auxiliary fuel system
   Capacities of prime movers (including fan curves showing operating
   ranges for static pressures, RPM, horsepower, volumetric flow rate)
  1 Dimensions of incinerator showing all feed locations
  1 Nozzle and burner design
  1 Description of automatic waste feed cutoff system
   Construction materials
   Complete description of pollution control equipment
  1 Location of temperature, pressure, flow rate, and other process mon-
   itoring systems
  1 Stack gas monitoring systems (including calibration procedures)
   Description of source and amount of waste(s) that will be incinerated
   during the trial burn
   If wastes blended, description of process to ensure uniform mixing
 WASTE CHARACTERIZATION AND
 POHC DESIGNATION
   It is necessary in the burn plan to describe as completely as pos-
 sible all waste streams that are combusted in the incinerator. At a
minimum, the information listed in Table 2 should be presented
for each waste. The purpose of this information is to provide the
reviewer with a full understanding of the physical and chemical
properties of each waste.

                           Table 2
                     Waste Analysis Data
• Heating Valve
• Viscosity
• Chlorine Content
• Ash Content
• Appendix VIII Constituents/concentrations


  After the wastes have been fully  defined and all of the Appen-
dix VIII constituents in each waste  have been identified, it will be
necessary to select several of these constituents as Principal Or-
ganic Hazardous Compounds (POHCs) to demonstrate the incin-
erator DRE. Some of the items to be considered in making this
selection are presented in Table 3. In general, it is common prac-
tice to select the compound which is present in the highest  con-
centration and  the compound which ranks highest on  the  U.S.
EPA incinerability hierarchy. It is desirable to select POHCs with
different types of molecular structures. In  selecting POHCs it also
is advisable to carefully consider any long range plans which may
impact the types or composition of the wastes which are to be
burned. If it is anticipated that any changes will occur which will
result in inclusion  of any Appendix VIII  constituents which are
higher on the incinerability hierarchy than any currently existing
compounds, it will be necessary to spike  the existing waste  with
these compounds.  The same would be true for any anticipated
increases in compound concentrations. The idea is to ensure that
it will not be necessary to conduct a second trial burn at some
future date  because of changes in waste characterization  that
could have been anticipated and addressed in the initial trial burn.
  Another very important  factor to consider when  selecting
POHCs is the ability to adequately sample and analyze the com-
pound. Table 4 presents  a list of compounds that present  sam-
pling and analytical problems.  These  compounds  should be
avoided if at all possible when selecting POHCs. Table 5 presents
a list of some of the common products of incomplete combus-
tion that may be formed from several compounds during the com-
bustion process. This list should be considered carefully when
selecting POHCs to ensure that the  selected POHC is not likely to
be present in the incinerator exhaust as a product of incomplete
combustion of some other compound.
                           Table3
                    POHC Selection Criteria

  Most Critical Waste Stream(s)
  Highest Concentrations
  Lowest Incinerability
  Different Molecular Structures
  Long-Range Plans
PROVISIONS FOR SAMPLING AND
MONITORING OF INCINERATOR PROCESS
  The trial burn plan should define very clearly and completely
the sampling and analysis and process monitoring procedures that
will be used during the trial burn program. The plan should in-
clude descriptions of the stack  testing  procedures and process
monitoring procedures. It is strongly recommended that this sec-
                                                                                                         INCINERATION     253

-------
                           Table 4
                       Difficult POHCs
                                                          Table 6
                                                   Slack Test Methodologies
Compound
Acetonitritle
Benzene
Chloroform
Creosote
Dibutyl Phthalate
Dichloromethane
Dimethylhydrazine
1,4-Dioxane
Diphenylamine
Formaldehyde
Hydrazine
Maleic anhydride
Methyl ethyl ketone
Phosgene
Phthalic anhydride
Pyridine
Toluene
Toluene diisocyanate
               Problems
           Poor recovery
          Contamination
          Contamination
         Complex mixture
                Reactive
          Contamination
                Unstable
           Poor recovery
           Poor recovery
           Water soluble
                Unstable
                Unstable
           Poor recovery
             Highly toxic
                Reactive
           Water soluble
          Contamination
                Reactive
                           Table 5
              Products of Incomplete Combustion of
                   Several Common Chemicals
Compound

Carbon Tetrachloride


Chloroform

Chlorobenzene
Dichlorobenzene

Trichlorobenzene


Toluene
      Possible Products of
   Incomplete Combustion

Perchloroethylene
Perchloroethane
Chloroform
Perchloroethylene
Carbon Tetrachloride
Benzene
Chlorobenzene
Benzene
Dichlorobenzene
Chlorobenzene
Benzene
Benzene
 tion of the trial burn plan be prepared with the assistance of a
 qualified and experienced source testing organization.
   The most commonly  used stack  testing  procedures are pre-
 sented in Table 6.  Full descriptions of these procedures can be
 found in the references presented at the end of this paper. It is
 important  to tell the Agency what procedures will be used  for
 sampling and analysis of the stack gas and to present the rationale
 for selection of each procedure.
   Samples that must be collected and analyzed  along with  the
 stack gas include a sample of the wastes which are fed to the incin-
 erator, any auxiliary fuel, scrubber discharge, scrubber make-up
 and ash. At a minimum, the parameters presented in Table 7 must
 be analyzed for each of these samples. This section of the trial
 burn plan must describe what samples will be collected,  how these
 samples will be collected, what parameters will be analyzed and
 what analytical procedures will be used.
   The third area to be addressed in this section of the burn plan is
 how the operation  of the incinerator and all  associated control
 equipment will be monitored during the test program.  Examples
 of the types of process data that must be recorded are presented in
Table 8. The way in which each parameter will be monitored must
be specified. When in-line process monitors are to be used, data
High boiling point POHCs
Low boiling point POHCs
HCL
Paniculate
CO
NOX
C02, 02
Modified Method 5
VOST
Modified Method 5
Method 5
NDIR (Method 10)
Method 7
Method 3
                           Table 7
                  Parameters for Waste Analysis
Sample
Waste Feed
                               Auxiliary Fuel

                               Scrubber Feed Water
                               Scrubber Discharge

                               Ash
Parameters
POHCs
Viscosity
Density
Heating Value
Ash Content
Chlorine Content
POHCs
Heating Value
POHCs
POHCs
EP-Toxicity (Metals)
POHCs
EP-Toxicity (Metals)
must be presented on the make and model number of each moni-
tor and the calibration procedures and frequency of those mon-
itors.

TEST SCHEDULE
  It is  necessary to include a proposed schedule in the trial burn
plan. It is, of course, difficult, if not impossible,  to firmly iden-
tify the test dates at the time of the  burn plan  preparation. The
actual test date will be  contingent upon several factors including
the length of time for regulatory agency review and approval of
the plan, process and  operating schedules, weather conditions
and test team availability. An attempt should be made, however,
to identify a tentative time frame for the testing.
  A schedule of activities for the trial burn period also should be
included. An example of such a schedule is presented below:

 • Day 1        Arrive on-site
                Begin equipment setup
 • Day 2        Complete setup
                Conduct preliminary measurements
 • Day 3        Conduct Test 1
 • Day 4        Conduct Test 2
 • Day5        Conduct Test 3
                Pack Equipment/Leave Site
 • Day 8        Samples arrive at laboratory
 • Day 55       Sample analysis complete
 • Day 95       Submit test report
   This schedule may have to be revised to include actual dates
 after a firm test date is established.
                               TEST BURN PROTOCOL
                                 The test burn protocol is the section in the trial burn plan in
                               which the operating conditions of the incinerator and control
                               system during the trial burn are selected and discussed. This sec-
 254    INCINERATION

-------
Table 8
Process Date Requirements
Incinerator Temperature
Waste Feed Rate(s)
Combustion Air Flow Rate
Auxiliary Fuel Feed Rate
Scrubber Pressure Drop
Incinerator Pressure
Scrubber Make-up Water Feed Rate
Scrubber Discharge Rate

tion of the plan is extremely important because the conditions at
which the incinerator is operated during the trial burn will become
operating restrictions which will be a part of the final Part B per-
mit. As a result, these conditions must be very carefully selected.
Some of the specific items that must be addressed include the
following:
• Incinerator Temperature
• Waste Feed Rate
• POHC Concentration in Waste
• Combustion Air Flow Rate
• Auxiliary Fuel Feed Rate
• Waste Ash Content
• Waste Chlorine Content
• Scrubber Pressure Drop
It should be the intent of the owner/operator to select con-
ditions which provide a maximum degree of flexibility for future
incinerator operations. This flexibility can be achieved by test-
ing under worst case conditions; i.e., maximum expected waste
feed rate, maximum expected POHC concentration, maximum
expected waste ash content, maximum expected waste chlorine
content and minimum expected incinerator temperature. In order
to fully demonstrate the incinerator DRE under the variety of
conditions desired by the owner/operator, it may be necessary to
conduct more than one series of tests during the trial burn pro-
gram.

QUALITY ASSURANCE/QUALITY CONTROL
Quality assurance and quality control (QA/QC) activities must
be an integral part of a trial burn program from the very begin-
ning of the effort through to the end. A complete QA/QC plan
must be included as part of the trial burn plan. A complete QA/
QC plan consists of the following sections.
• Project Description
• Project Organization and Responsibility
• QA Objectives
• Sampling Procedures
• Sample Custody
• Calibration Procedures and Frequency
• Analytical Procedures
• Data Reduction, Validation and Reporting
• Internal Quality Control Checks
• Performance and System Audits
• Preventive Maintenance
• Specific Routine Procedures Used to Assess Data Precision,
Accuracy and Completeness
• Corrective Action
• Quality Assurance Reports to Management
This plan must be prepared in great detail because it will be
used as a way to measure the quality of the data generated dur-
ing the trial burn and, as a result, the acceptability of the test re-
port.

CONCLUSION
The preparation of a comprehensive trial burn plan is a critical
part of the overall permitting process for hazardous waste incin-
erators. The plan must detail the nature of the actual trial burn
test as well as technical considerations of the process and waste
streams. A well-developed plan will minimize the time required
to permit hazardous waste incinerators and preclude regulatory
non-compliance actions in addition to overly restrictive permit
conditions.

REFERENCES
1. Title 40, Code of Federal Regulations.
2. Title 40, Code of Federal Regulations, Part 60, Appendix A.
3. U.S. EPA, Office of Solid Waste, Washington, DC, "Test Methods
for Evaluating Solid Waste-Physical/Chemical Methods," SW-846.
4. U.S. EPA, Office of Solid Waste, Washington, DC, "Guidance
Manual for Hazardous Waste Incinerator Permits," SW-966.
5. U.S. EPA, Industrial Environmental Research Laboratory, Research
Triangle Park, NC, "Sampling and Analysis Methods for Hazardous
Waste Combustion," EPA-600/8-84-002.
6. U.S. EPA, Air and Energy Engineering Research Laboratory, Re-
search Triangle Park, NC, "Modified Method 5 Train and Source
Assessment Sampling System Operator's Manual," EPA-600/8-85-
003.
7. U.S. EPA, Industrial Environmental Research Laboratory, Research
Triangle Park, NC, "Protocol for the Collection and Analysis of
Volatile POHCs Using VOST," EPA-600/8-84-007.
INCINERATION 255

-------
                  Planning and  Design for  On-Site  Incineration at
                             Two Illinois  Hazardous Waste Sites

                                                   James F. Frank
                                        Division of Land  Pollution Control
                                    Illinois Environmental Protection  Agency
                                                 Springfield, Illinois
ABSTRACT
  This paper discusses  the planning,  design and procurement
phases of a transportable incinerator at one state and one NPL-
listed site.  Both sites have heavily contaminated soils. The NPL
site, LaSalle Electric Utilities  (LEU),  is a bankrupt capacitor
manufacturer. The state site, Beardstown Lauder Salvage Yard, is
a defunct  metal salvage yard that salvaged copper from trans-
formers.
  The LEU site has contaminated residential  lawns. The lawn
soils will be excavated and hauled to the main plant property for
incineration utilizing a  transportable incinerator erected at  the
site. The second phase of this cleanup may involve a transportable
incinerator to  detoxify PCB-contaminated soils  on the plant
property. Approximately 25,500 yd3 of contaminated soils will be
excavated and incinerated; lawns will be backfilled with clean soil
and sodded. The home interiors will be cleaned.
  This paper describes the pre-qualification and procurement
process as well as  a step-by-step chronology  of how to apply
transportable incineration technology on two different sites. Cost
information is presented. The Illinois Environmental  Protection
Agency's plans for future incineration projects  are discussed. In-
sight to community relations approaches is provided.

BEARDSTOWN PCB SITE
  The Beardstown  PCB-contaminated site was a metal salvage
yard occupying approximately 5 ac located 3 mi east of the Illinois
River at Beardstown, Illinois, in Cass County. The land originally
was purchased in the 1950s when Mr.  Lauder began his salvage
operation. The original  plot of land was purchased as 5 ac,  but
the boundaries were expanded to include portions of adjoining
properties.
  The site accepted all  sorts of  salvage  materials from general
scrap to junk automobiles. PCB transformers and  capacitors
were accepted for the copper which would be reclaimed. The oil
from the transformers in most  cases was dumped on the ground
to drain so that the copper could be removed easily.
  The Illinois EPA (IEPA) became interested in the site when in-
itial results of  soil sampling  collected  in the vicinity  of  the
transformers and capacitors indicated levels of PCB contamina-
tion as high as 12%. Due to the nature of the surficial soils and
the proximity of a trailer court to the northwest, the agency de-
cided to perform an immediate removal of all debris.
  The initial Phase I cleanup of the site included the removal of
approximately 4125 yd3 of clean scrap metals,  320 transformer
carcasses,  96 capacitors  and 120 yd3 of contaminated soils. The
second phase of the cleanup identified 4900 yd3 of contaminated
soils at levels exceeding 5 ppm PCBs, the established cleanup ob-
jective. From data collected during this phase, it was determined
that soils are contaminated to a depth of 5 ft with the majority of
contamination occurring in the top 2 ft. The third phase of the
cleanup will use the TTDU to incinerate  the 4900 yd3 of con-
taminated soils and treat to a level of 5 ppm total PCBs or less.
The treated  material then will be placed  back  on the site and
vegetated. No cap will be placed over the treated soil.

Geologic Setting
  The surficial  soils consist of Tine to medium grained sand with
little clay matrix. This sandy material extends to a depth of ap-
proximately  100 ft and is a major aquifer in the area due to the
high water yield. Most residential wells in the vicinity are sand-
point wells which are driven down to an average depth of 40 to 50
ft.
  Monitor wells indicate groundwater  flow toward the trailer
court to the northwest. These wells and a few select wells within
the trailer court have not indicated groundwater contamination.

Procurement Process for Beardstown Site
  In September 1985, IEPA put out a request for proposal which
read in part as follows:

   Request for Statement of Qualifications for Cleanup
Services Using  a Transportable Thermal Destruction Unit

  It is  the intention of the  Illinois Environmental Protec-
tion Agency, Division of Land Pollution Control ("IEPA/
DLPC" or "Agency"), to utilize Transportable Thermal
Destruction Unit (TTDU) technology at a hazardous waste
site that is now on the State Remedial Action Priority List
(SRAPL) or National Priority List (NPL).
   IEPA/DLPC is seeking qualified firms who will  agree to
own and operate a TTDU in conjunction with their hazar-
dous waste cleanup services on an Illinois hazardous waste
site to  be designated by IEPA/DLPC. Typical  contamin-
ants and wastes found at the site  will  be  PCBs at levels
about  50  ppm,  volatile  organic  solvents,  nonvolatile
organics, halogenated organics, pesticides, oils and other
petroleum products. These contaminants will be in several
different forms, including liquid  wastes, in  soils  or in
various objects (for example, 55-gallon metal drums).
   IEPA/DLPC desires proposing firms to provide the fol-
lowing services:

1. Own and operate a TTDU at the specified cleanup site.
2. Design and build a feed system that will handle  contam-
   inated soil,  liquids, sludges and various sized metal or
   fiber drums.
3. Perform all acts necessary for a hazardous waste clean-
256    INCINERATION

-------
up, including excavating, staging, processing and feed-
ing waste material to the TTDU.
4. Remove treated wastes from the TTDU for appropriate
management and/or disposal.
5. Test air emissions and obtain necessary permits.
6. Provide all necessary equipment and appurtenances to
make system functional.
7. Provide all necessary personnel to operate all systems.
The purpose of the Request for Statement of Qualifications
(RFQ) was to select from the universe of prospective bidders
those bidders who could best demonstrate to the agency that they
possessed the financial strength and resources, state-of-the-art
technology, management strength, relevant experience and satis-
factory plan for financing the program's capital expenditures to
successfully implement all of the services previously described.
The result of the RFQ and subsequent technical review of pro-
posals was that a contract in the amount of $2.4 million was
awarded to Roy F. Weston, Inc. of West Chester, Pennsylvania.
Weston submitted the low bid in this procurement action which
provided for a turn-key remediation service at the Beardstown site
with a unit cost of $246.00/ton. The current schedule for the
Beardstown projects is as follows:
Beardstown Schedule
Site Preparation
Equipment Mobilization
Equipment Preparation
Trial Burn
Full Scale Operations
Remediation Complete
Demobilization
March 1987
April 1987
May-June 1987
June 1987
August-September 1987
October 1987
November 1987
LA SALLE ELECTRIC UTILITIES SITE
The LaSalle Electrical Utilities (LEU) National Priorities List
(NPL) site is located in west-central LaSalle County, north-
central Illinois. The 1980 census data showed the City of LaSalle
to have a population of 10,347.
The bedrock in the area consists primarily of shale, sandstone,
dolomite and limestone. The upper bedrock is a highly weathered
shale found at a depth of approximately 20 to 25 ft. Overlying the
bedrock are approximately 10 ft of glacial till. Over the till is an
interbedded unit of sand, silt and clay.
There are four major hydrogeologic aquifers which occur in
this area of Illinois. The Mt. Simon-Elmhurst aquifer, the deepest
of the four, is not utilized in the LaSalle area due to its extreme
depth and its high mineral content. The next aquifer is the
Ironton-Galesville which serves the three public water supply
wells in the nearby community of Peru, Illinois. These wells are
approximately 2700 ft deep. The shallow dolomite and the sand
and gravel aquifers in the area service many domestic and public
wells.

Site History
LEU is a former manufacturer of electrical equipment. Opera-
tions at the plan began prior to World War II, and in the late
1940s the plant began utilizing PCBs in the production of capaci-
tors. This manufacturing practice continued until October 1978.
In May 1981, the company ceased operations at the LaSalle
plant. The LaSalle facility has been abandoned since that time.
Information on the waste management practices of the com-
pany both on and off the property is limited. Undocumented re-
ports allege that PCB-contaminated waste oils were regularly ap-
plied as a dust suppressant both on and off the property as late at
1969.
Beginning in September 1975, numerous government agencies
(including the U.S. EPA, the IEPA and OSHA) conducted vari-
ous inspections and issued numerous complaints and orders to the
LEU company as a result of its manufacturing and handling prac-
tices both past and present.
On Sept. 19, 1983, LEU petitioned for relief under Chapter 11
of the Bankruptcy Act. On June 26, 1986, the Bankruptcy Court
entered an order approving the company's planned liquidation
under Chapter 11.

Current Site Status
The information gathered in the various investigations and
sampling efforts was used as part of the site's remedial investiga-
tion (RI) which was conducted by IEPA. A draft copy of the RI
report was prepared and submitted by the IEPA to the U.S. EPA
in January 1986.
The contaminants of concern in the LEU off-site area are
PCBs. No other materials above normal background levels have
been detected in this area.
Based on the sampling data, the extent of contamination for
the off-site area was determined. The areas of contamination in-
clude the shoulders of St. Vincent Avenue for about 0.2 mi north
and approximately 1.2 mi south of the LEU property, the residen-
tial area directly east of the site, the small commercial area south
of the property and one residence north of the LEU property.
Concentrations of PCBs in the composite soil samples from
these areas range from less than 0.20 ppm to as high as 2600 ppm.
(The lower limit is the analytical detection limit.) Additional grab
samples from the most heavily contaminated residential yard
revealed a hot spot containing up to 5800 ppm of PCBs. Concen-
trations typically average about 75 to 125 ppm in most yards in
the area. The depths of contamination range from 0 to 12 in. in
most areas, to as much as 5 ft at a few heavily contaminated loca-
tions. The total volume of soil that is contaminated above the 5
ppm level is approximately 28,690 yd3. Of this amount, 6,450 yds
are along the shoulders of St. Vincent Avenue; the remainder is
on the nearby residential and commercial property.
There are 27 property owners who have contamination in their
yards above the 5 ppm level and who will be directly impacted by
the phased remedial alternative which has been chosen to alleviate
the contamination.

Alternative Screening and Evaluation
Five alternatives were examined in detail. In conformance with
the NCP and the FS guidance specifications, the alternatives were
put into five general compliance categories as follows:
• The "no action" alternative
• Alternatives that attain applicable and relevant federal and
state public health or environmental standards, guidance and/
or advisories
• Alternatives that exceed applicable and relevant federal and
state public health or environmental standards, guidance and/
or advisories
• Alternatives that meet CERCLA criteria for preventing or
minimizing the threat to human health and the environment
but which may not attain all relevant or applicable standards
• Alternatives that treat or dispose of the hazardous substances
at an off-site facility
With the exception of the "no action" alternative, all the alter-
natives would involve soil removal from the residential site area.
The soil removal process would be the same for each of these four
alternatives. It would include the excavation of soil over the ap-
propriate area and depth in order to achieve the specified cleanup
level. Table 1 presents the costs for the five options.
A record of decision for the LEU site involving the residential
soils removal was signed on Aug. 29, 1986 by Valadas Adamkus,
INCINERATION 257

-------
Regional Administrator for Region 5.  The IEPA is proceeding
with an advances state match funding arrangement under cooper-
ative agreement to the U.S. EPA to perform a state-led design.
The major work  elements  in time frame  for  the design are as
follows:

• The  design is being  performed by Ecology  and Environment
  over a 17-wk period  at a  cost of approximately $500,000.
• At the conclusion of the  design process, the state  will proceed
  with a formal advertisement for competitive bidding to pro-
  cure incineration services. Although the federal procurement
  process is somewhat different  from  that required by Illinois
  State Statute there are many similarities including the use of a
  performance base specification rather than a detailed design for
  the incinerator and pre-processing appurtenances. The state has
  selected this option  as opposed to a detailed incinerator de-
  sign  since we believe it is faster and more economical to allow
  many different thermal process systems to compete in the pro-
  curement process including various incinerator designs.

                           Table 1
          Cost of Remedial  Action Options for LEU Site
Alternative
                              Total Present Worth Cost
                              (Millions of 1986 Dollars)
                 5 ppm    10 ppm  25 ppm   50 ppm

No Action          0.0      0.0     0.0      0.0
Off-Site Landfill    19.5     13.4     8.6      7.4
Off-Site
  Incineration     176.4    120.5    77.0     66.6
On-Site
  Incineration      29.6     20.3    13.0     11.2
On-Site Storage      1.3      1.0     0.8      0.8

  The approximate schedule for the LaSalle Electric Utilities pro-
ject is shown in Table  2.
                           Table 2
         LaSalle Electric Utilities Remedial Design Schedule
Task
Project Initiation
Remedial Design Administrative
  Documents (QAPP, SSP, etc.)
Completion of Construction Plans
  Specifications and Bidding Documents
                               Completion Date

                               January 1987

                               February 1987

                               May 1987
Initiative Procurement for Remedial Action    June 1987
                       Remedial Action
Receive Sealed Bid Proposals                July 1987
Award Contract and Initiate Remedial
  Action                                August 1987

LIMITED WAIVER OF RCRA PERMITS
IN ILLINOIS SECTION 118(i) OF CERCLA
  Section 118(i)  of CERCLA became effective as a matter of law
on Oct. 17,  1986 with the enactment  of the  1986 Superfund
Amendments Reauthorization Act.  Section  118(i) provides as
follows:

1.  Limited Waivers in  State of Illinois
   1) Mobile Incinerators—In the case of remedial actions spe-
      cifically involving mobile incinerator units in the State of
      Illinois, if such  remedial  actions are undertaken by the
      State under the  authority of a State  Superfund law or
      equipment authority, the State may, with the approval of
      the  Administrator, waive  any permit requirement  under
      Subtitle C of the Solid Waste Disposal Act which would be
      otherwise  applicable to such action to the extent that the
      following conditions are met:
      (A) No Transfer—The  incinerator does  not involve the
          transfer of a hazardous substance or pollutant or con-
          taminant  from the facility at which the  release  or
          threatened release occurs to an off-site facility.
      (B) Remedial Action—The remedial action provides each
          of the following:
          (i)  Changes in  the character or composition of the
              hazardous substance or pollutant or contaminant
              concerned so that it no longer presents a risk to
              public health.
          (ii)  Protection against  accidental  emissions during
              operation.
          (iii) Protection of public health considering the multi-
              medial impacts of the treatment process.
      (C) Public Participation—The  State provides procedures
          for public participation  regarding the  response action
          which are at least equivalent to the level of public par-
          ticipation procedures applicable under CERCLA and
          under the Solid Waste Disposal Act.
   2)  Effect of Waiver—The waiver of any permit requirement
      under this subsection shall not be construed to  waive any
      standard or level of control which—

      (A) is applicable to any hazardous substance or pollutant
          or contaminant involved in the remedial action;  and
      (B) would otherwise be contained  in the permit.
          Such waiver of any permit requirement under Subtitle
          C of the Solid Waste Disposal Act shall only apply to
          the extent that the facility or remedial action involves
          the on-site treatment with a mobile incineration unit of
          waste present at such site. The waiver shall not apply
          to any other regulated or potentially regulated activity,
          including the use of the mobile incineration unit for ac
          tions not authorized by  the State.
   3)  Expiration of Authority—The authority of this subjection
      shall terminate at  the  end  of  3  years, unless  the  State
      demonstrates,  to the satisfaction  of the  Administrator,
      that the operation of mobile incinerators in the State has
      sufficiently protected public health and the environment
      and is  consistent with  the criteria required for a permit
      under Subtitle  C of the Solid Waste Disposal Act.

  The genesis of Section 118(i) occurred in the fall of 1985 as the
IEPA monitored the progress of Superfund reauthorization and
as IEPA initiated the procurement of a mobile incinerator unit to
incinerate hazardous wastes on-site at abandoned waste dumps in
the State of Illinois. During that time, two things became clear to
IEPA.  First, the  language,  which eventually became  Section
121(e) of the Act, that exempted the U.S. EPA on-site remedial
actions  from various federal and state permit requirements, in-
cluding  Subtitle C of the Solid Waste Disposal Act, would not be
expanded to  encompass similar exemptions where the remedial
action was undertaken by  the state at non-CERCLA sites. Se-
cond, unless an exemption  from RCRA permitting requirements
was obtained for mobile incinerator units, the long-term goals of
the state in encouraging ultimate destruction of hazardous waste
through  on-site treatment by incinerator could not be met. In-
stead, the shell game of toxic disposal would continue  as before.
  Because of these concerns, the IEPA requested the  assistance
of the Office of Governor James R. Thompson. In turn, the prob-
lems were explained to the office of Rep.  Edward Madigan. Rep.
Madigan, after having been convinced of the merits of the IBPA's
case, proposed that language be included in the law eliminating
 258
INCINERATION

-------
this inequity between federal and state cleanup actions. As a
result of these efforts, Section 121(m) ("Permits for On-site
Cleanup under State Authority") was included in the proposed
amendments as adopted by the House of Representatives on Dec.
10, 1985 (HI 1638). The Senate bill did not include a counterpart
to Section 121 (m). As the House and Senate conferees discussed
the proposed amendments during the spring and summer of 1986,
Rep. Madigan continued to play an active role in urging the con-
ferees to include language which would allow the state and the
U.S. EPA to reach "a mutual agreement to waive a permit for
the use of mobile incinerators." This would be done only after
assuring that the substantive requirements of the permit are met
and that the public health is protected.
The concept behind this limited waiver is to provide an oppor-
tunity for a state to demonstrate in a model program how mobile
incineration systems may be utilized in a safe and efficient manner
to detoxify hazardous waste at non-NPL sites. Presuming that
this test is exemplary, it is hoped that this privilege may be extend-
ed to other states which meet certain minimum requirements as
deemed appropriate by the U.S. EPA. Illinois is evaluating
several sites for use in this program at the present time. We intend
to submit the necessary petition to the administrator of the U.S.
EPA during January 1987.

COMMUNITY RELATIONS AT MOBILE
INCINERATION SITES
An integral part of any hazardous waste remediation program
is community relations. This is particularly true when incineration
is being considered as a remedial alternative. The IEPA has been
involved in three community relations projects involving the use
of mobile incineration.
The only common element that characterizes each of these sites
is a hazardous waste site remediation plan recommending the use
of a mobile incinerator.
At two of the three sites, a community relations program was
being implemented during the course of a remedial investigation.
Both of the communities (Beardstown and LaSalle) were kept ful-
ly informed of the remedial investigation activities and test results
as they were received. Agency staff visited each affected house-
hold in LaSalle to discuss the results of the soil analyses, to pro-
pose the soil removal option and to discuss the use of incineration
as the remedial alternative.
The third community is a state immediate removal site, and
community relations activities started concurrently with the
remedial investigation.
A consistent low-key approach was used in each community
when the use of a mobile incinerator initially was discussed.
Various methods of corrective action were described; information
was distributed; concerns answered; and mobile incineration was
identified as the agency's preferred remedial option. The agency
staff did not come on strongly, although they personally were
very enthusiastic about the proposal. Each community recognized
that the agency was listening to its comments and concerns.
There are several common elements that characterize the
citizens' feelings about the sites and the proposed technology. At
each site, the agency was called upon to explain why the waste
could not be "hauled away." The citizens reluctantly accepted
the concept that it is not feasible to transport large volumes of
contaminated soil long distances and dispose of them in landfill fa-
cilities. In addition, it is against state and federal policy to do so.
Citizens had extensive questions regarding the safety and
reliability of the technology. Citizens seem to place a degree of
trust in the lEPA's assertion that the technology and resulting
remediation were as safe as possible given a bad situation.
Citizens demanded a guarantee that the remediation would be of
a limited duration and that once the site was remediated the in-
cinerator would be relocated. They were particularly concerned
about the incinerator being left on-site and used to handle waste
imported from other sites.
On balance, IEPA has been quite pleased with the response to
its community relations programs involving on-site incineration.
This situation stands in stark contrast to the strong objections
voiced by citizens and political leaders to new permanently
located hazardous waste facilities including permanent incinera-
tion facilities.
CONCLUSION
The IEPA will have one active incinerator site operational and
another in the procurement phase in 1987. These two projects are
among the first large-scale remediation projects involving
thousands of tons of soil to be conducted in the United States. Il-
linois intends to pursue future mobile incineration projects at
NPL sites under Section 118(i) of CERCLA.
IEPA is proud of its accomplishments in promoting the use of
mobile incineration as an alternative treatment technology whose
time has come. The agency has been willing to match its convic-
tion with substantial amounts of state monies. The agency does
not mind being the "guinea pig" if it serves the public purpose of
furthering a promising technology. The agency's goal is to have
this technology provided by many different companies in a com-
petitive market place situation. This will allow the State of Illinois
to have its choice among qualified providers so that it is able to
landfill less hazardous waste and incinerate more hazardous waste
as it strives to achieve more permanent environmental solutions in
the state's hazardous waste management program.
INCINERATION 259

-------
                   Full-Scale Rotary Kiln  Incinerator Field  Trial:
                            Phase  I,  Verification Trial Burn  on
                    Dioxin/Herbicide  Orange Contaminated  Soil

                                            Major Terry L. Stoddart
                                                  Jeffrey  J.  Short
                          Headquarters Air Force Engineering and Services Center
                          Environics  Division, Environmental  Engineering Branch
                                        Tyndall Air Force Base, Florida
ABSTRACT
  The U.S. Air Force Engineering and Services Center is con-
ducting a series of field trials employing a variety of currently
available remedial action technologies that may be capable of re-
storing Herbicide Orange/Dioxin-contaminated sites. As part of
this continuing research program, the U.S. Air Force is conduct-
ing a full-scale field trial at the Naval Construction Battalion
Center, Gulf port, Mississippi. This test will employ a full scale
rotary kiln incinerator capable of processing up to 6 tons/hr of
dioxin-contaminated soils.
  The results of five verification trial burns reported here indi-
cate that the incinerator is capable of removing dioxin and Herbi-
cide Orange from the soil matrix to concentrations not detectable
at 10 ug/kg. Future plans call for  the unit to thermally treat
11,000 tons of contaminated soil to support reliability, main-
tainability and cost-effectiveness studies.

INTRODUCTION
  Herbicide Orange (HO) was used  as a tactical military defol-
iant in Southeast Asia until 1971 when DOD suspended its use
because of suspected health risks.'  The formula was  a 50/50
mixture of me herbicides 2,4-dichlorophenoxyacetic acid (2,4-D)
and 2,4,5-tribhlorophenoxyacetic acid (2,4,5-T). Trace amounts
of  2,3,7,8-tetrachlorodibenzo-p-dioxin  (TCDD) contaminated
HO during the production of 2,4,5-T.  The average concentra-
tion of TCDD in HO was about 2 mg/1.1
  The barrels of herbicide were stored and shipped to the theater
from the Naval Construction Battalion Center (NCBC) in Gulf-
port, Mississippi. After its ban, 1,370,000 gal of herbicide from
Southeast Asia were  placed in storage at Johnston Island (JI).
Another 850,000 gal  were in storage at NCBC. The remaining
2,220,000 gal were destroyed in 1977 by incineration at sea (Oper-
ation PACER HO).
  Following Operation PACER HO, USAF OEHL initiated site
sampling at JI, NCBC and Eglin AFB, Florida—where HO was
application tested—to monitor migration and degradation of HO
and its associated dioxin. The JI and NCBC sites were contam-
inated with 2,4,5-T, 2,4-D and TCDD as a result of spills of the
HO during storage and handling. In 1980 the Secretary of the
Air Force directed that the HO sites be returned to full and bene-
ficial use. The Environics Division of the U.S. Air Force Engi-
neering and Services Center (AFESC/RDVW) was tasked to eval-
uate and develop management techniques to restore the sites and
reduce impacts of the contamination.
  The U.S. Air Force contracted with the Idaho National Engi-
neering Laboratory (Department of Energy), operated by EG&T
Idaho,  to  demonstrate in  the field innovative site restoration
technologies. As a part of this research program,  EG&G sub-
contracted with the ENSCO Corp. to conduct a full-scale demon-
stration of a modular, transportable, rotary-kiln incinerator at
NCBC. Pilot-scale technology demonstrations at NCBC include
the Huber Corporation Advanced Electric Reactor and the I.T.
Corporation Thermal Desorption/Ultraviolet process.
  This paper reviews the preliminary analytical results of the set-
up and trial burn of contaminated soil at NCBC with the ENSCO
transportable incinerator. The field trail, which includes a heavily
documented validation test burn, will treat approximately 11,000
tons of decontaminated soil. This full-scale field trial will allow
the U.S. Air Force to determine the efficacy, reliability and costs
associated  with  the use of an incinerator to reclaim contam-
inated sites.

Site Description
  NCBC is located northwest of Gulfport, Mississippi, approx-
imately 2 mi from the Gulf of Mexico. The elevation averages
30 ft above sea level. The HO storage area comprises nearly 18 ac
of flat sand and sandy loam type soils. The soils at the site were
mixed with Portland cement about 30 yr ago to provide a hard-
ened surface for heavy equipment operation and storage.
  Beginning  in April 1984, the U.S. Air Force  performed a
comprehensive site sampling effort to map contamination levels
at the HO sites.' This 2-yr study provided the necessary data to
support the demonstration of site restoration technologies. A re-
view of site photographs (not previously available) indicated that
there were other HO storage areas at NCBC not included in the
original sampling. An additional 5 ac was sampled in 1986 to
further determine the extent of HO contamination of "Area B."
  Contamination levels at the original 12-ac NCBC site ranged
from less than detectable at 0.1 ppb to over 500 ppb.1 Surface
concentrations averaged 14.3 ppb (based on 1639 samples). Pre-
liminary analysis of 500 Area B samples show that the contamina-
tion levels there ranged from non-detectable at 0.1  ppb to 350
ppb.  At these levels,  there are approximately 9,000 tons and
2,000 tons of contaminated soil that could be excavated for treat-
ment in areas A and B, respectively.

PROCESS DESCRIPTION
  The  ENSCO  incinerator (Mobile Waste Processor—MWP-
260    INCINERATION

-------
1
W«IM

UXUBMl
1*

W«l«f I
Hopp
\,
•r

W««uw*ur 1>
Fu«IJw»il*/jU — caJBu^n**
Ram or tcr«w U*4

Roury kiln
II

ConcaniraUd
• All solution
Figure 1
Schematic Flow Diagram of MWP-2000
2000) is designed and fabricated by the Pyrotech Division of the
Environmental Services Company (ENSCO) in White Bluff,
Tennessee.5 The MWP-2000 (Fig. 1) is a modular mobile incinera-
tor originally designed to destroy solid and liquid wastes con-
taminated with PCBs. The system achieved six nines (99.9999%)
destruction removal efficiency tests for PCBs and dioxin surro-
gates under TSCA and RCRA requirements.
The MWP-2000 is capable of treating approximately 100 tons/
day of dioxin-contaminated soil. A brief description of the incin-
eration process to be used at NCBC follows.
The contaminated soil is fed into a weighing hopper with a
front-end loader. From the weighing hopper, the soil drops onto a
conveyer belt, which drops the soil into a feed hopper. From the
feed hopper, the soil is fed by a rotary auger into the rotary kiln
incinerator. A shredder is used to provide a uniform size feed-
stock.
The soil is heated to 1000-1800° F in the rotary kiln which will
burn or volatilize all combustibles. The treated soil then exits the
rotary kiln and falls into a water-sealed, treated soil quencher.
A chain-drag conveyor discharges the soil into a large roll-off bin.
The soil is held for analysis to ensure no residual contaminants
exist. The soil is returned to an excavated area when determined
to be clean.
Meanwhile, the off-gas from the rotary kiln is drawn into the
secondary combustion chamber (SCC), where it is subjected to
temperatures of 2000 to 2400 ° F for a minimum residence time
of 2.2 sec in an excess oxygen atmosphere. The SCC ensures com-
plete destruction of organic gases. Acidic brine produced at the
scrubber is passed through tandem activated carbon filters and
neutralized.
Gases from the SCC then pass into the waste heat boiler to
produce steam for use in the ejector scrubber. From the boiler,
the gases then pass into the quench sump, which reduces the
off-gas temperature for subsequent processing in the packed
tower.
The packed tower removes 99% of the acid gases from the com-
bustion air. In the packed tower, the gases flow upward through
the tower and are scrubbed by a countercurrent flow of water.
From the packed tower, the off-gas is drawn into the ejector
scrubber. That device, which operates similar to an ejection
pump, provides the prime motive force for moving the off-gases
and helps scrub particulates from the off-gas. Steam generated in
the waste heat boiler serves as the motive fluid. The clean off-gas
then is forced up the 35-ft stack.

System Description
The residence time of the waste in the system is varied from 30
to 60 min by altering the speed, the angle of the kiln or the num-
ber and location of internal refractory dams. The temperature of
the rotary kiln is high enough to remove TCDD from the soil
matrix by volatilization of all organics into the off-gas. Kiln
temperature is controlled by adjusting fuel flow, combustion air
and soil input. At NCBC, the system is modified for natural gas.
The single gas-fired burner produces a minimum 14,000,000
BTU/hr. Wet soil substantially affects kiln temperature and, con-
sequently, the quantity of throughput.
The off-gases from the kiln pass through dual cyclones to re-
move the fine particulates. The particulates settle out into the
treated soil receiving tank. This tank receives the ash from the
lower end of the rotary kiln. Treated soil is discharged into a
INCINERATION 261

-------
bellows-sealed breaching at the lower end of the kiln into the
tank. This tank is filled with water to a level above the discharge
opening of the breaching providing a water seal and preventing
escape of gases. A chain-drag conveyor removes the treated ash
from the receiving tank and transfers it to a storage bin. Large
(20 yd1) roll-off bins are used to hold the treated soil until analy-
sis verifies that it contains no contaminants.
The purpose of the SCC is to completely oxidize the contam-
inants in the off-gas to simple combustion products: water, car-
bon dioxide and hydrochloric acid. The SCC uses a 24,000,000
BTU vortex burner designed to produce a short, turbulent flame.
Additionally, combustion air is introduced into the SCC to en-
sure maximum turbulence and optimal burning of the TCDD.
The gases exit the SCC and move to the waste heat exchanger
(boiler) through a duct equipped with an emergency vent in case
of coolant waterless.
The purpose of the waste heat boiler is to produce 8400 Ib/hr
of high pressure steam to be used as the primary motive force
in the ejector scrubber. This boiler reduces the off-gas tempera-
ture from approximately 2200° F to 388 ° F.
The air pollution control train consists of a quench system, a
packed tower, an ejector scrubber and a 35-ft stack. This equip-
ment is designed to further cool the gases, remove 1500 Ib/hr of
HCL and remove particles larger than 3 /i in size.
The quench system conveys exit gases from the waste heat boil-
er through a 90-degree elbow to the quench sump. The quench
sump collects excess recirculation water for the entire air pollu-
tion control system and provides some additional residence time
to cool gases. The quench elbow contains spray nozzles which
produce a fine water mist to cool the gases from 600 to approx-
imately 153° F. The mist also absorbs HCL contained in the off-
gas and settled into the quench sump or is carried to the packed
tower. Calcium carbonate is added to the water in the quench
tank to neutralizek the resultant acid. The quench sump is fed by
a raw water line in case of an emergency low-water condition.
The packed tower is designed to remove HCL from the off-
gas. Assuming a maximum loading of 1600 Ib/hr, the packed
tower is capable of removing 99"% of the HCL leaving the quench
sump. The packed tower is a fiberglass-reinforced plastic tank
packed to a 6-ft depth with 2-in. diameter plastic shapes called
"Tellerettes." The gases flow upward through the tower and are
scrubbed by a downward flow of water recirculated from the
packed tower sump and ejector scrubber.
The ejector scrubber is designed to remove additional particu-
late and HCL from the gases before they are discharged. The
scrubber removes 99
-------
Sample Analysis
Analytical protocols employed were taken from U.S. EPA
SW846, Certified Laboratory Program (CLP) dioxin protocols
or by generally accepted laboratory procedures. All preservation
and storage protocols specified by the stated regulations/proto-
cols were followed without exception.

Table 2
Test 2 Operational Parameters
Table 5
Test 5 Operational Parameters
PARAMETER
DATE

START

END

FEED RATE (TONS/HH)

KILN OUTLET TEMP (DEGREES F)

SEC COHB OUTLET TEMP (DEGREES F)

STACK OUTLET CO %

STACK OUTLET C02 %

STACK OUTLET 02 »

COMBUSTION EFFICIENCY %

STACK PABTICULATE (GR/DSCF)
Table 3
Test 3 Operational Parameters
PARAMETER
DATE

START

END

FEED RATE (TONS/HR)

KILN OUTLET TEMP (DEGREES F)

SEC COMB OUTLET TEMP (DEGREES F)

STACK OUTLET CO %

STACK OUTLET C02 t

STACK OUTLET 02 J

COMBUSTION EFFICIENCY %

STACK PARTICULATE (GR/DSCF)
VALUE
7 DEC 86

09:15

11:00

3.64

1377

2159

0.0

9.0

5.5

100

0.018
VALUE
7 DEC 86

14:55

16:05

3.71

1552

2167

0.0

9.0

1.5

100

0.015
PARAMETER
VALUE
DATE

START

END

FEED RATE (TONS/HR)

KILN OUTLET TEMP (DEGREES F)

SEC COMB OUTLET TEMP (DEGREES F)

STACK OUTLET CO %

STACK OUTLET C02 %

STACK OUTLET 02 %

COMBUSTION EFFICIENCY %

STACK PARTICIPATE (GR/DSCF)
15 DEC 86

11:15

12:55

6.31

1355

2101

0.0

7.6

6.6

100

0.019
RESULTS AND DISCUSSION
During the five verification test burns, the ENSCO unit was
operated within the operational envelope dictated by the U.S.
EPA Region 4 R&D permit. The only serious operational prob-
lem encountered was wet soil that resulted from extremely heavy
rains. Future operations of the unit at the NCBC site will in-
corporate soil drying as a routine activity.
The preliminary results of laboratory analysis are presented in
Tables 6 through 10. Additional analyses on stack-gas samples
were not available at the time this report was prepared. The data
presented indicate that the ENSCO MWP-2000 is capable of re-
moving the principal hazardous constituents from the soil matrix
found at NCBC. These data also indicate that the treated soil
should be delistable under the conditions of RCRA. Additional
tests are being conducted to determine the concentration of heavy
metals in the feed stock and treated soil.
Following submittal of all laboratory test results to the various
regulatory agencies, the ENSCO system will undergo reliability
and maintainability testing. The test plan calls for continuous
operation of the unit for 90 days. During this period, the unit will
treat approximately 11,000 tons of Herbicide Orange/Dioxin
contaminated soil. Following completion of the test period, engi-
neering and cost reports will be prepared to document the reli-
ability, maintainability and cost-effectiveness of the ENSCO
system.
Table 4
Test 4 Operational Parameters
PARAMETER
DATE

START

END

FEED RATE (TONS/HR)

KILN OUTLET TEMP (DEGREES F)

SEC COMB OUTLET TEMP (DEGREES F)

STACK OUTLET CO %

STACK OUTLET C02 %

STACK OUTLET 02 %

COMBUSTION EFFICIENCY %

STACK PARTICULATE (GR/DSCF)
VALUE
15 DEC 86

09:20

10:30

5.22

1185

2113

0.0

8.0

6.0

100X

0.022
Table 6
Results of Chemical Analyses Test 1 (2.82 tons/hr) Concentration as
Jlg/kg(ppb)
ANALYTE
FEED STOCK
TREATED SOIL
2,4,D
2,1,5-T
2,1,5-TRICHLOROPHENOL
3,4-DICHLOROPHENOL
PHENOL
TCDD
PeCDD
HxCDD
HpCDD
OCDD
TCDF
PeCDF
HxCDF
HpCDF
OCDF
56,000
100,000
1,600
ND(330)
ND(660)
32.1
ND(0.15)
ND(0.058)
ND(0. 10)
0.70
0.15
ND(0.23)
ND(0.081)
ND(0.070)
ND(0.0011)
ND (20)
ND(2)
ND(1600)
ND(330)
ND(330)
HD(0.0015)
ND(0.0018)
HD(0.0051)
ND(0. 00015)
0.0024
ND(0. 00085)
ND(0. 00018)
HD(0. 00031)
ND(0.0011)
ND(0. 00024)
ND= NOT DETECTED AT THE INDICATED DETECTION LIMIT
INCINERATION 263
-------
Table 7
Results of Chemical Analyses Test 2 (3.64 lons/hr) Concentration as
ug/kg(ppb)
ANALYTE
FEED STOCK
TREATED SOIL
2,1,0
2,1,5-T
2,1, 5-TRICHLOROf HENOL
3,1-DICHLOROPHENOL
PHENOL
TCDD
PeCDD
HxCDD
HpCDD
OCOD
TCDF
PeCDF
HxCDF
HpCDF
OCDF
ND= NOT DETECTED AT THE
3,300,000
510,000
3,700
ND(330)
ND(660)
51.2
ND(0.28)
ND(O.IO)
ND(O.ll)
0.61
0.19
ND(0. V5)
ND(0.060)
ND(O.IO)
ND(0.066)
INDICATED DETECTION
HD(20)
ND(2)
NDO600)
ND(330)
ND(330)
ND(0.0015)
ND(0.0029)
ND(O.OOll)
0.00039
0.00137
0.0129
ND(0. 00069)
ND(0. 00057)
ND(0. 00062)
ND(0. 00012)
LIMIT
Table 8
Results of Chemical Analyses Test 3 (3.71 tons/hr) Concentration as
ug/kg(ppb)
ANALYTE
FEED STOCK
TREATED SOIL
2.4,0
2,4,5-T
2,4,5-TRICHLOROPHENOL
3,4-DICHLOROPHENOL
PHENOL
TCDD
PeCDD
HxCDD
HpCDD
OCDD
TCDF
PeCDF
HxCDF
HpCDF
OCDF
120,000
220,000
3,600
ND(330)
ND(330)
38.0
NDC0.28)
ND(0. 18)
ND(0.25)
0.72
0.58
ND(O.ll)
NDC0.019)
ND(0.052)
ND(0.067)
ND(20)
ND(2)
ND(1600)
NDC330)
NDC330)
ND(0. 00089)
ND(0. 00028)
ND(0.0022)
ND(0.0017)
0.0193
0.0160
ND(0. 00129)
ND(0. 00068)
ND(0. 00050)
ND(0. 00027)
Table 9
ReiulU of Chemical Analyses Tat 4 (5.22 toni/hr) Concentration u
pg/kg(ppb)
ANALYTE
FEED STOCK
TREATED SOIL
2.4.D
2.4,5-T
2,4,5-TRICHLOROPHENOL
3,4-DICHLOROPHENOL
PHENOL
TCDD
PeCDD
HxCDD
HpCDD
OCDD
TCDF
PeCDF
HxCDF
HpCDF
OCOF
23,000
117,000
8,800
NDC330)
ND<330)
15.8
ND(0.21)
ND(0.9D
ND(O.II)
0.80
0.66
NDC0.12)
ND(0.050)
NDfO.22)
ND(0.16)
ND(20)
ND(2)
ND(1600)
HD(330)
NDC330)
ND{0.0022)
ND(0. 00035)
HD(0. 00011)
0.00058
0.02Z7
0.0067
ND(0. 00046)
MD(0. 00065)
0.00065
ND( 0.00028)
ND= NOT DETECTED AT THE INDICATED DETECTION LIMIT

Table 10
Result* of Chemical Analyse* Test 5 (6.31 lons/hr) Concentration as
"f/kg
-------
The U.S. Environmental Protection Agency's
Guidelines for Risk Assessment

Peter W. Preuss, Ph.D.
Alan M. Ehrlich, Ph.D.
Kevin G. Garrahan, P.E.
Office of Health and Environmental Assessment
U.S. Environmental Protection Agency
Washington, D.C.
ABSTRACT
In recent years the U.S. EPA has moved toward a risk assess-
ment/risk reduction framework for making regulatory decisions.
The agency has taken a number of steps to assure the quality and
consistency of the risk assessment component of those decisions.
The first, and perhaps most important of these steps, is the
development of agency-wide risk assessment guidelines. Five
guidelines have been published: carcinogenicity, mutagenicity,
developmental toxicity, chemical mixtures and exposure. The
provisions of the five guidelines are discussed in the context of the
four components of risk assessment.
Other activities designed to assure quality and consistency in
risk assessment, reduce uncertainty in risk assessment, ensure a
more efficient information exchange about risk and risk assess-
ment and develop the appropriate oversight mechanisms also are
discussed. These include: additional guidelines, the Risk Assess-
ment Forum, risk assessment research, the Integrated Risk Infor-
mation System, the Hazard Assessment Notification System and
the Risk Assessment Council.

INTRODUCTION
One key factor used by the U.S. EPA in developing a pollution
control strategy is the evaluation of scientific information in ways
that will permit assessing: (1) the risk from an environmental in-
sult, or (2) the degree that the risk may be reduced under any par-
ticular control scenario. Specifically, the Office of Solid Waste is
increasing its use of risk-based decision-making in implementing
its program under RCRA and the Hazardous and Solid Waste
Amendments of 1984. This change to risk-based decision-making
is, perhaps, most noticeable in the establishment of Alternate
Concentration Limits (ACLs) as a groundwater protection stan-
dard. As a result of this and similar changes in all U.S. EPA of-
fices, risk assessment has become increasingly important to the
regulatory process. It also has become clear that the distinction
between risk assessment and other parts of the regulatory decision
process needs to be carefully and comprehensively defined. This
regulatory decision process was basically defined several years ago
by the National Academy of Sciences (NAS) and can include
legal, economic, political and social factors as a part of the
management of risks determined by way of the risk assessment
process1 (Fig. 1).
During the past decade, the U.S. EPA has moved vigorously to
a risk assessment/risk management/risk reduction framework for
making regulatory decisions. As a consequence, the assurance of
quality and the consistency of assessments have become issues to
which we are paying a great deal of attention.
RISK ASSESSMENT
LAtoNATOMV AND FIELD
OMEHVATIOMI OF AOVEftlE
HEALTH EMECTI AND IX-
.OIUHII TO PARTICULAR
HIOH TO LOW OOIE AMD
ANIMAL TO HUMAN
RISK MANAGEMENT
10UHCI M4k
Figure 1
Elements of Risk Assessment and Risk Management'

There are a number of steps that we have taken in help achieve
these goals of quality and consistency; perhaps the most impor-
tant among them is the development of agency-wide assessment
guidelines. The U.S. EPA had developed such guidelines in the
past—carcinogenicity in 1976 and 1980, systemic toxicants and
mutagenicity in 1980 and exposure assessment in 1983.2-3-4-5 In
January 1984, the U.S. EPA began intensive work on six new or
revised guidelines: carcinogenicity, mutagenicity, reproductive
toxicity (subdivided into individual guidelines for developmental
toxicity and male and female reproductive toxicity), systemic tox-
icants (e.g., target organ toxicants), chemical mixtures and ex-
posure assessment.6
The first stage for each guideline was the development of drafts
by agency-wide work groups of scientists. These drafts were cir-
culated to scientists from academia, other governmental agencies,
industry and public interest groups.
Using this procedure, five guidelines (carcinogenicity,
mutagenicity, developmental toxicity, chemical mixtures and ex-
posure) were proposed for public comment.7,8,9,10,11 After the
public comments were received, the U.S. EPA staff evaluated the
comments, suggested revisions and sent the proposed guidelines
and the evaluation of comments to special review panels of the
RISK ASSESSMENT 265
-------
Science Advisory Board (SAB). The review panels and the Ex-
ecutive Committee concurred on the guidelines subject to certain
revisions and have subsequently concurred on the revisions.12'13
The proposed risk assessment guidelines recently were published
in the Federal Register.1*-"^1'1*
The U.S. EPA's guidelines set forth internal agency procedures
that will:
• Promote consistency across U.S. EPA risk assessments by de-
veloping common approaches to risk assessment
• Promote the quality of the science underlying U.S. EPA risk
assessments [by use of a consensus approach (discussed below)
where appropriate]
• Clarify the U.S. EPA's approach to risk assessment by inform-
ing the public and the regulated community about the process
by which the U.S. EPA will evaluate scientific information

The guidelines are not regulations; in fact, they are intention-
ally flexible to encourage the use of all data and the appropriate
scientific methods and judgments. The guidelines can, however,
influence the regulatory process by:
• Making the U.S. EPA's risk assessments more consistent and
of higher technical quality
• Familiarizing risk assessors throughout the country with the
U.S. EPA's approach
• Making it possible for scientists to plan their experiments to
collect the information that U.S. EPA scientists would like to
have available when conducting a risk assessment
Finally, these guidelines are intended to be evolving documents.
They are being updated even now, as the science base relating to
risk assessment leads to new understanding of the effects of toxic
substances or to a reduction of the uncertainty inherent in the risk
assessment process.
General agreement on the need for risk assessment guidelines
does not exist. Some scientists have argued that articulation of
guidelines is inappropriate and that every situation should be
evaluated on a case-by-case basis. They believe that this case-by-
case approach is necessary because of the complexity of the scien-
tific issues and their concern that general rules cannot be easily
developed or followed. On the other hand, others prefer detailed
guidelines that take risk assessors through each step of the process
and spell out specific approaches or scientific conclusions. As in
most disagreements, a middle ground exists, that is, to develop a
general logic for the kinds of information needed and to articulate
appropriate methods for assessment and evaluation. In this ap-
proach, the guidelines are intended to be tools in the hands of
skilled scientists; they encourage the evaluation and use of all the
available information on a case-by-case basis.
COMPONENTS OF RISK ASSESSMENT AND
THEIR RELATIONSHIP TO THE GUIDELINES
In discussing risk assessment and risk management, the NAS
divided the process of risk assessment into four components.1
These are:
(1) Hazard Identification: the determination of whether a par-
ticular chemical is or is not causally linked to particular health
effects
(2) Dose-Response Assessment: the determination of the rela-
tion between the magnitude of exposure and the probability
of occurrence of the health effects in question
(3) Exposure Assessment: the determination of the extent of
human exposure before or after application of regulatory
controls
(4) Risk Characterization: the description of the nature and often
the magnitude of human risk, including attendant uncertainty
To the extent possible, the U.S. EPA's guidelines follow the
Academy's definitions. The following sections describe each com-
ponent in greater detail and show how the guidelines relate to
them.

HAZARD IDENTIFICATION
The hazard identification component of a risk assessment con-
sists of a review of relevant biological and chemical information
bearing on whether or not an agent may pose a specific hazard.
Sometimes, there is enough information available for the
qualitative evidence to be combined into a formal weight-of-
evidence determination.
For example, in the guidelines for carcinogen risk assessment,7'14
the following information is evaluated to the extent that it is
available:
• Physical/chemical properties and routes and patterns of ex-
posure
• Structure/activity relationships
• Metabolic and pharmacokinetic data
• Influence of other toxicologic effects
• Short-term tests
• Long-term animal studies
• Human studies
Once these data are reviewed, the animal and human data are
each divided into groups by degree of evidence:
• Sufficient evidence of carcinogenicity
• Limited evidence of carcinogenicity
• Inadequate evidence
• No data available
• No evidence of carcinogenicity
The animal and human evidence are then combined into a
weight-of-evidence classification scheme similar to the one
developed by the International Agency for Research on Cancer."
This scheme gives more weight to human evidence when it is
available. The scheme includes the following groups:
Group A — human carcinogen
Group B — probable human carcinogen
Group C — possible human carcinogen
Group D — not classifiable as to human carcinogenicity
Group E — evidence of noncarcinogenicity toward humans
To some degree, these are arbitrary divisions along a con-
tinuum; therefore, it is important to stress that categories should
not be overinterpreted. The attached matrix (Table 1) shows how
the human studies and long-term animal studies are combined to
derive the first approximation of the overall weight-of-evidence
classifications. Other types of evidence are then used to adjust the
first approximation upward or downward as appropriate.
In the case of mutagenicity risk assessment,8-15 the goal is to
assess the likelihood that a particular chemical agent induces
heritable changes in DNA and the likelihood that the chemical
will interact with human germ cells.
Evidence that an agent induces heritable mutations in human
beings could be derived from epidemiologic data indicating a
strong association between chemical exposure and heritable ef-
fects. It is difficult to obtain such data, however, because any par-
ticular mutation is a rare event, and only a small fraction of the
estimated thousands of human genes and conditions are currently
useful as markers in estimating mutation rates.
Therefore, in the absence of human epidemiologic data, it is
appropriate to rely on data from experimental animal systems so
266 RISK ASSESSMENT
-------
Table 1
Illustrative Categorization of Evidence Based on
Animal and Human Data"
1
1
1
1
1
1
1
1
1
1
1
1

HIMAM
EVIDENCE

SUFFICIENT

LIMITED

INADEQUATE

NO DATA

EVIDENCE OF
NO EFFECT

ANIMAL EVIDENCE
NO EVIDENCE ()f
SlIFFICIENT LIMITED INADEQUATE DATA NO EFFECT

A A A A A

HI Rl Bl Bl BI

B2 C n 0 D

B2 C 1) D F

82 C D D E

NOTE: The above assignments are presented for illustrative purposes. There may be nuances in
the classification of both animal and human data indicating that different categorizations than
those given in the table should be assigned. Furthermore, these assignments are tentative and may
be modified by ancillary evidence. In this regard all relevant information should be evaluated to
determine if the designation of the overall weight-of-evidence needs to be modified. Relevant fac-
tors to be included along with the tumor data from human and animal studies include structure-
activity relationships, short-term test findings, results of appropriate physiological, biochemical
and toxicological observations and comparative metabolism and pharmacokinetic studies. The
nature of these findings may cause an adjustment of the overall categorization of the weight-of-
evidence.
long as the limitations of using surrogate and model systems are
clearly stated. The universality of DNA and the interest in the
possible causal relationship between mutagenesis and cancer in-
duction are partly responsible for the development of a large
number of both in vitro and in vivo mutation tests which may be
used to evaluate the potential mutagenic activity of specific
agents. The practical implication is that the available data for any
set of chemicals are extremely variable, thus precluding a precise
scheme for classifying chemicals as potential human germ-cell
nutagens. What has evolved is a rank-ordered scheme of
categories of evidence bearing on potential human germ-cell
mutagenicity. The highest category is reserved for human
epidemiologic data, recognizing that no such data are currently
available. There are then five other categories (in descending
order) based on the premise that greater weight is placed on tests
conducted in germ cells than in somatic cells, on tests performed
in vivo rather than in vitro, in eukaryotes rather than prokaryotes
and in mammalian species rather than in submammalian species.
Additionally, there are categories for defining a nonmutagen and
for cases in which there is insufficient information to make a
qualitative decision.
The specific statements of the eight categories are:
(1) Positive data derived from human germ-cell mutagenicity
studies, when available, will constitute the highest level of
evidence for human mutagenicity.
(2) Valid positive results from studies on heritable mutational
events (of any kind) in mammalian germ cells.
(3) Valid positive results from mammalian germ-cell chromo-
some aberration studies that do not include an intergenera-
tion test.
(4) Sufficient evidence for a chemical's interaction with mam-
malian germ cells, together with valid positive mutagenicity
test results from two assay systems, at least one of which is
mammalian (in vitro or in vivo). The positive results may
both be for gene mutations or both for chromosome aberra-
tions; if one is for gene mutations and the other for chromo-
some aberrations, both must be from mammalian systems.
(5) Suggestive evidence for a chemical's interaction with mam-
malian germ cells together with valid positive mutagenicity
evidence from two assay systems as described under 4, above.
Alternatively, positive mutagenicity evidence of less strength
than defined under 4, above, when combined with sufficient
evidence for a chemical's interaction with mammalian germ
cells.
(6) Positive mutagenicity test results of less strength than de-
fined under 4, combined with suggestive evidence for a chem-
ical's interaction with mammalian germ cells.
(7) Although definitive proof of nonmutagenicity is not pos-
sible, a chemical could be classified operationally as a non-
mutagen for human germ cells if it gives valid negative test
results for all end points of concern.
(8) Inadequate evidence bearing on either mutagenicity or chem-
ical interaction with mammalian germ cells.
In the guidelines, developmental toxicity includes adverse ef-
fects on the developing organism that may result from exposure
prior to conception (in either parent), during prenatal develop-
ment or postnatally to the time of sexual maturation.9'16 The ma-
jor manifestations of development effects include death of the
developing organism, malformation, altered growth and func-
tional deficiency. (The term teratogenicity refers primarily to
malformations and is a subclass of developmental toxicity.)
Short-term and in vitro tests, which frequently are used to assess
risks from suspect carcinogens and mutagens, are not appropriate
approaches to assess developmental toxicity because the develop-
ing organism is such a complex system. Instead, bioassays and
human epidemiologic data are the primary sources of information
used. The primary biological assays involve treatment of animals
during organogenesis and evaluation of the offspring at term.
These types of evaluations also may be done as part of a multi-
generation study.
The kinds of evaluations that are made in the U.S. EPA's
hazard identification/weight-of-evidence determination include,
as with all such evaluations:
• Quality of the data
• Resolving power of the studies; that is, consideration of the
significance of the studies as a function of the number of ani-
mals or subjects
• Relevance of route and timing of exposure
• Appropriateness of dose selection
and, more specifically in the case of developmental toxicity, an
evaluation of the information for a series of end points that may
include:
• In the developing animal
-deaths
-structural abnormalities
-growth alterations
-functional deficiencies in the developing organism
• In the maternal animal
-fertility
-weight and weight gain
-clinical signs of toxicity
-specific target organ pathology and histopathology

In the case of chemical mixtures,10'17 the U.S. EPA conducts its
hazard identification by considering the weights-of-evidence for
the mixture's component chemicals. Occasionally, and especially
for complex mixtures, the evidence for a health hazard comes
directly from studies on the mixture itself. Information on the
mixture itself, however, must be reviewed carefully for evidence
of masking of one toxic end point by another. For example, when
RISK ASSESSMENT 267
-------
one component chemical is a suspect carcinogen but the data
show marked toxicity in major organs (e.g., liver or kidney) and
no indication of cancer, there is the possibility that other toxic ef-
fects may mask the evidence of carcinogenicity. The hazard iden-
tification then would suggest no cancer risk at any dose when, in
fact, at doses below the threshold for systemic toxicity, there
could be significant risk of cancer.
Exposure assessment usually is a separate step in the risk assess-
ment process; the exposure guidelines are discussed in a later sec-
tion of this document. For mixtures, however, the exposure in-
formation must be considered to determine the chance that
chemical interactions in the environment could produce new
chemicals, over time or during transport, with different types of
health hazards resulting. This concept is discussed more fully in
the next section.

DOSE-RESPONSE ASSESSMENT
Classically, there are two general approaches to dose-response
assessment depending on whether the health effects are threshold
or nonthreshold. For threshold effects, discussed later in this sec-
tion, the assessment estimates the point below which we do not
expect a significant adverse effect. For nonthreshold effects, an
attempt is made to extrapolate response data from doses in the ex-
perimental range to resonse estimates in the dose ranges typical of
most environmental exposures. The largest number of such dose-
response extrapolations have been performed in the field of car-
cinogen risk assessment; therefore, the cancer guidelines give the
most detailed guidance on dose-response assessment.7'14 This as-
sessment includes guidance on the kinds of evidence that should
be used in the dose-response evaluation, such as:

• If available, estimates based on human edidemiologic data are
preferred over estimates based on animal data.
• In the absence of appropriate human studies, data from animal
species that respond most like humans should be used.
• The biologically acceptable data set from long-term animal
studies showing the greatest sensitivity generally should be
given the greatest emphasis.
• Data from the exposure route of concern are preferred to data
from other exposure routes; if data from other exposure routes
are used, the considerations used in making route-to-route
extrapolations must be carefully described.
• When there are multiple tumor sites or multiple tumor types,
each showing significantly elevated tumor incidence, the total
estimate of carcinogenic risk is estimated by pooling (i.e.,
counting the number of animals having one or more of the
significant tumors).
• Benign tumors generally should be combined with malignant
tumors for risk estimates.

Another major consideration is the choice of the particular
mathematical model used for low-dose extrapolation. Different
extrapolation models may fit the observed data reasonably well
but may lead to large differences in the projected risk at low
doses. In keeping with the recommendations of the Office of
Science and Technology Policy,20 the U.S. EPA will review each
assessment with respect to the evidence on cancer mechanisms
and other biological or statistical evidence that indicates the
biological suitability of a particular extrapolation model. A ra-
tionale will be included to justify the use of the chosen model. In
the absence of adequate information to the contrary, the linear-
ized multistage procedure will be employed. The linearized
multistage procedure is recognized as leading to a plausible upper
limit to the risk that is consistent with some proposed mechanisms
of carcinogenesis.
Additional issues are species- and route-extrapolation of the
doses. Currently, the U.S. EPA adjusts animal doses by the ratio
of animal-to-human surface areas. The evidence in support of this
approach is not strong, and research is in progress to improve the
method. Route extrapolation is used when the only available data
are for a route other than the route of concern.
In the case of mutagenicity risk assessment,8'15 dose-response
assessments can presently only be performed using data on ger-
minal mutations induced in intact animals. The morphological-
specific locus and biochemical-specific locus assays can provide
data on the frequencies of recessive mutations, and data on
heritable chromosome damage can be obtained from the heritable
trans-location test. As in carcinogen risk assessment, the U.S.
EPA will strive to use the most appropriate extrapolation models
for risk analysis and will be guided by available data and mechan-
istic considerations in this selection. However, it is anticipated
that for tests involving germ cells of whole mammals, few dose
points will be available to define dose-response functions;
therefore, a linear extrapolation will be used. The U.S. EPA has
recognized that pioneering work in the field of molecular
dosimetry ultimately may lead to useful extrapolation models.
The other major approach to dose-response assessment is con-
cerned with effects which the U.S. EPA has referred to as
systemic toxicants or noncarcinogenic health effects (see below).
Although this particular area presently is not covered by guide-
lines (they are still being developed), it is appropriate here to
discuss the general approach. We usually calculate what we call a
Reference Dose (RfD), that is, the dose below which we do not
expect a significant risk of adverse effects. The reference dose is
related to the more familiar concept of the Acceptable Daily In-
take (ADI) but strives to remove the elements of risk management
from the process. At present, we are not sure at what point above
the RfD there will be a significant adverse health effect. The dose-
response evaluation is done in the following way. The literature is
examined to determine both the critical toxic effect (that is, the
adverse effect that first appears in the dose scale as the dose is in-
creased) and the highest dose at which the effect does not occur
(often called the highest No-Observed-Adverse-Effect-Level or
NOAEL). This NOAEL is divided by an uncertainty factor which
generally ranges from 10 to 1,000; the uncertainty factor is com-
posed of a series of factors, each representing a specific area of
uncertainty inherent in the data available.
The RfD calculation is a generic calculation for most toxicants
considered to have thresholds. In addition, there is much work
being carried out in an attempt to develop more quantitative ap-
proaches for dose-response assessment for reproductive and
developmental toxicants both within and outside of the U.S.
EPA.
The dose-response procedures described in the chemical mix-
tures guidelines are a bit different.10'17 In this case, guidance is
provided for combining several different types of information on
the mixture of concern as well as on the mixture components. If
dose-response data are available for the mixture itself, such data
are used, and other U.S. EPA risk assessment procedures would
apply to the mixture as a whole. If data are not available on the
specific mixture, it may be appropriate to infer information from
sufficiently similar mixtures. When neither is available, we sug-
gest using what is called dose or response addition, appropriately
modified if interactions between components (such as synergism)
can be quantified. For most threshold pollutants, when interac-
tions cannot be quantified and when the component chemicals are
lexicologically similar, strict dose addition is used. This means
dividing each estimated intake level by its RfD, and summing each
of these quotients to calculate a hazard index. When the hazard
index is much less than one, no significant risk is expected from
the mixture. When the hazard index is much greater than one, a
268
RISK ASSESSMENT
-------
significant risk might be expected. When the hazard index is near
one, each case needs to be considered individually.
For carcinogens and for dissimilar systemic toxicants that have
dose-response data, response addition is used so that, at typical
environmental levels, the excess risks for each of the component
chemicals are summed to reach an overall risk estimate. Again, in-
teractions need to be considered, and we must recognize the add-
ed uncertainty in the assessment. The U.S. EPA intends to in-
vestigate this and other problems involving mixtures that contain
carcinogens over the next few years.

EXPOSURE ASSESSMENT
From the titles of the various guidelines, it is clear that four of
the five relate to health effects; in those cases, which have been
presented previously, discussions of hazard identification and
dose-response assessment are appropriate. In contrast, one guide-
line discusses exposure assessment.
The Proposed Guidelines for Exposure Assessment11 and the
Guidelines for Estimating Exposures18 provide a procedural
framework to estimate the degree of human contact to a chemi-
cal. The major areas to be evaluated when estimating exposures
are:
• Source assessment—a characterization of the sources of con-
tamination
• Pathways and fate analysis—a description of how a contamin-
ant may transport from the source to the potentially exposed
population
• Estimation of environmental concentration—an estimate using
monitoring data and/or modeling of contamination levels
away from one source where the potentially exposed popula-
tion is located
• Population analysis—a description of the size, location and
habits of potentially exposed human and environmental re-
ceptors
• Integrated exposure analysis—the calculation of exposure
levels and an evaluation of uncertainty
An integrated exposure assessment quantifies the contact of an
exposed population to the substance under investigation via all
routes of exposure and all pathways from the sources to the ex-
posed individuals.
Generally, exposure estimates may be presented by expressing
the magnitude and duration of an individual event of exposure or
by expressing potential lifetime exposure. For example, evalua-
tions of acute or subacute effects, such as developmental effects,
would use the magnitude of exposure per event or several events
over a short period of time. On the other hand, assessments of
carcinogenic risk often consider the daily average exposure
calculated over a lifetime. The nature of the toxic effect being
evaluated in the risk assessment will determine the appropriate
length of exposure presented.
For most risk assessments involving chronic exposure, exposure
(mg/kg/day) is calculated as a dose averaged over the body
weight (kg) and lifetime (days):
Average Daily Lifetime Exposure
Total Dose

Body Weight x Lifetime

The total dose (mg) can be expanded as follows:
(1)
Total Contaminant Contact Exposure Adsorption (2)
Dose = Concentration x Rate x Duration x Fraction
The four parameters in Equation (2) are defined as follows:
• Contaminant concentration represents the concentration of
the contaminant in the medium (air, food, soil, etc.) contacting
the body; typical units are mass/volume or mass/mass.
• Contact rate is the rate at which the medium contacts the body
(through inhalation, ingestion or dermal contact); typical units
are mass/time or volume/time or, for dermal contact, volume/
surface area.
• Exposure duration is the length of time for contact with the
contaminant.
• Absorption fraction is the effective portion of total contam-
inant contacting and entering the body. Entering the body
means that the contaminant crosses one of the three exchange
membranes: alveolar membrane, gastrointestinal tract or skin.
The six factors given in Equations (1) and (2) must be known
(or estimated) in order to estimate exposure. Research is in pro-
gress to better define how to estimate each of these factors for
humans as well as test animals.

RISK CHARACTERIZATION
In the guidelines, the risk characterization step is a summing
discussion in which information is put together in a useful way.
This means that the risk characterization contains not only a risk
estimate for a specific exposure, but also a cogent summary of the
biological information, the assumptions used and their limita-
tions, and a discussion of uncertainties in the risk assessment,
both qualitative and quantitative.
In the case of the cancer, mutagenicity and chemical mixtures
guidelines,7'8-10'14'15'17 the risk characterization specifically consists
of the dose-response extrapolation information as well as the
associated weight-of-evidence determination from the scale or
table contained in the guidelines. For mixtures,10'17 the weight-of-
evidence covers three areas: health effects information, toxic in-
teractions and exposure estimates.
In the case of the exposure assessment guidelines, a specific
mathematical technique has been developed to assess uncer-
tainty.11-18 In this case, the probability distributions estimated
for the uncertainty around each component in the calcula-
tion are entered into a computer program, and the probability
distribution of the results of the exposure assessment can then be
calculated or estimated.

SYSTEMIC TOXICANTS
The last of the six original areas of guidelines development is
systemic toxicants. Guidelines have not yet been prepared because
it is difficult to reach consensus for such a broad area. The U.S.
EPA essentially includes, under the umbrella of systemic tox-
icants, chemicals causing a variety of health effects other than
cancer, mutagenicity and specific acute effects. This umbrella
clearly covers many end points and many different target organs
that could be considered for any one chemical. In the absence of
consensus procedures, each Program Office at the U.S. EPA
historically has approached the problem in different ways.21 Ex-
amples include evaluating a specific adverse health effect rather
than determining the critical health effect (the adverse health ef-
fect occurring at the lowest dose) or assessing risk for less than
lifetime exposure rather than determining a lifetime chronic Ac-
ceptable Daily Intake (ADI) or Reference Dose (RfD). In addi-
tion, Program Offices have used different approaches for dealing
with uncertainty. Some estimate a lifetime chronic RfD based on
uncertainty factors tied to the available information and then es-
tablish criteria based on the RfD. Some calculate a Margin of
Safety between the highest No-Observed-Adverse-Effect-Level of
the critical effect and the estimated exposure and then evaluate
RISK ASSESSMENT 269
-------
that margin specifically in terms of the chemical of interest and its
expected exposure pattern. Some offices estimate an appropriate
degree of protection on a case-by-case basis, using their best tech-
nical and scientific judgment. Finally, some programs have de-
veloped their own quantitative techniques to extrapolate or inter-
polate across data gaps; few of these have gained general accep-
tance within the U.S. EPA. The guidelines development effort
has been postponed while a U.S. EPA-wide effort is under way
(discussed below) to review all RfDs that have been recommended
throughout the U.S. EPA, eliminating inconsistencies in RfDs for
the same chemicals calculated by various Program Offices, up-
dating the data base as appropriate and then achieving consensus
on the entire list. Generic issues resolved in this RfD review pro-
cess will form the basis for the guidelines on systemic toxicants.

OTHER GUIDELINES PROJECTS
The Science Advisory Board (SAB) reviewed the proposed
guidelines and suggested the development of two additional
documents: (1) guidelines for making and using environmental
measurements in exposure assessments and (2) a technical support
document for the guidelines on chemical mixtures. Work on those
projects is under way. In addition, a document is being developed
to focus on areas in need of research in developmental toxicology.
Work is also continuing on two other guidelines: (1) the Assess-
ment of Risk to the Male Reproductive System and (2) the Assess-
ment of Risk to the Female Reproductive System. In addition,
U.S. EPA staff are working on guidelines for the assessment of
systemic toxicants and are planning guidelines for the assessment
of ecological risk and the appropriate use of metabolism and
pharmacokinetic data and models.

RISK ASSESSMENT FORUM
Risk assessment guidelines are only one of the tools used in
making decisions; the guidelines, therefore, are only a part of the
procedures for making the U.S. EPA's decisions more consistent
and reliable. Another mechanism is the Risk Assessment Forum.
For any one issue, there may be data gaps or insufficiently
developed risk assessment technology. For those risk assessment
issues that could have a significant impact on regulatory decisions
or for those in which there is a strong U.S. EPA-wide interest, it is
desirable to develop a consensus U.S. EPA-wide science policy.
In addition, there may be areas that the guidelines do not cover
now, but which need some immediate or short-term resolution.
Lastly, as scientific theories develop and change or as experimen-
tal techniques and risk assessment assumptions change, there
needs to be a way to augment or amend the U.S. EPA's risk
assessment policies. The U.S. EPA, therefore, decided to
establish a standing group of senior scientists who would meet
regularly to provide a "forum" for those kinds of discussions and
decisions.6'22 As stated in its charter, the Risk Assessment Forum
promotes consensus on risk assessment issues and the incorpora-
tion of that consensus into appropriate risk assessment guidance.
To fulfill this purpose, the Forum assembles risk assessment ex-
perts from throughout the U.S. EPA in a formal process to study
and report on these issues from a U.S. EPA-wide scientific
perspective.
Forum activities may include: developing scientific analyses,
risk assessment guidance and risk assessment methodology for use
in ongoing and prospective U.S. EPA actions; using scientific and
technical analysis to propose risk assessment positions; and
fostering consensus on these issues. Generally, the Forum focuses
on generic issues fundamental to the risk assessment process, the
analysis of data used in risk assessment and the development of
consensus approaches to risk assessment. Peer review and quality
assurance of completed risk assessments or review of non-
scientific risk management issues are not standard Forum func-
tions.
In addition to the Forum, the U.S. EPA has established the
Risk Assessment Council. According to its charter, the Council's
goal is to provide executive oversight of the development, review
and implementation of U.S. EPA risk assessment policy. To
fulfill this purpose, the Council brings together senior officials
from throughout the agency in a formal manner to consider these
matters from a U.S. EPA-wide management perspective.
Council activities may include: coordination of intra-agency
(both in the regions and at headquarters) and inter-agency risk
assessment activities; development of initiatives to improve the
U.S. EPA's risk assessment process; guidance on the interpreta-
tion of risk assessment information in the U.S. EPA's decision-
making process; and referral of risk assessment matters needing
resolution to appropriate U.S. EPA offices.

REDUCING UNCERTAINTY
One critical need in risk assessment is to reduce the uncertainty
of the estimates. We are undertaking several activities to do that.
First, we have been planning three workshops. One, a "Consen-
sus Workshop on the Relationship of Maternal and Developmen-
tal Toxicity," was held in May 1986, to address issues of inter-
pretation of data in the area of developmental toxicity when tox-
icity to the maternal animal may also be apparent.23 Another
workshop held in October 1986, was on the use of pharmaco-
kinetic models in risk assessment. The goal of this workshop was
to identify the basis on which these models are formulated and the
assumptions and data that are necessary for their use. This
workshop addressed the practical application of pharmacokinetic
principles and models to improve risk assessment. Finally, in
February 1987, a workshop was held to address cancer research
needs relating to promoters. The workshop was intended to help
the U.S. EPA identify the needed areas of risk assessment
research relating to the role of promoters in carcinogenicity and
to establish a list of priorities for that research. We anticipate that
these workshops on pharmacokinetics and on carcinogenicity will
be the first of a fairly extensive series of workshops in these areas.

OTHER ACTIVITIES
The U.S. EPA was one of the pioneers in developing and
adopting risk assessment methods. For the first 10 yr, this meant
developing the techniques, applying them and then attempting to
reach consensus about their appropriate use. That effort has now
culminated in the publication of the five guidelines and in the con-
tinued plans for guideline development and risk assessment
research.
It is now appropriate to consolidate that effort by ensuring a
more efficient information exchange about risk and risk assess-
ment and by developing other oversight mechanisms to ensure
consistency and high technical quality in the U.S. EPA's risk
assessments.
One area of concern is the quality assurance of the hazard and
risk evaluations developed within various Program Offices; for
example, the work of the RfD review group referred to earlier.
The review group has been working since early 1985 and the first
150-200 RfDs have been approved for inclusion in U.S. EPA in-
formation systems (see discussion below). A similar review pro-
cess is under way for carcinogen risk estimates.
Another area of concern is the development of appropriate in-
formation exchanges about risk assessment activities in the U.S.
EPA in order to identify the hazard assessment activities under
way within the U.S. EPA and prevent possible duplication, to in-
crease the awareness of ongoing activities of interest to various
Program Offices and to improve coordination of these activities.
270 RISK ASSESSMENT
-------
This has led to the development of an internal Hazard Assessment
Notification System in which Program and Regional Offices will
list all hazard assessments in a data base and report on work that
is under way or anticipated.
Another such U.S. EPA-wide activity is the Integrated Risk In-
formation System (IRIS). One of the many problems encountered
by risk assessors both inside and outside of the U.S. EPA is ob-
taining coherent information about existing U.S. EPA risk con-
clusions useful for formulating risk assessments. Information
about the results of carcinogen bioassays, dose-response calcula-
tions, NOAELs, RfDs and other parameters exists for a large
number of chemicals, but this information previously has not
been integrated into an easily accessible, centralized information
base.
Information in IRIS is organized in a readily accessible elec-
tronic mail system on a chemical-by-chemical basis. Each item of
information is reviewed for consistency and quality prior to entry
into the system. As a chemical-based system, IRIS collects infor-
mation for a compound and constructs a file in which all numbers
fit into a particular format. Information is provided by four con-
tinuing efforts, each of which periodically supplies updated
assessments and information to the central IRIS management
unit: One effort contributes reviewed RfDs; a second effort does
the same for cancer risk estimates; a third effort lists acute hazard
information that the U.S. EPA has recently published24; and the
fourth effort provides risk management numbers, U.S. EPA-wide
(Reportable Quantities, National Ambient Air Quality Standards,
Water Quality Criteria, Maximum Contaminant Levels, etc.). In-
formation from these four projects is merged to produce a file
consisting of a series of chemical-specific documents. The user is
able to call up a chemical by name and review all of the U.S. EPA
summary material pertinent to it.
Other computer-related projects include the development of
toxicity data bases. The furthest along is "Studies on Toxicity
Applicable to Risk Assessment." This data base is unique because
it contains toxicity data for each dose group and includes pro-
grams for calculating and presenting the data in human equivalent
terms by using the extrapolation models and time-weighted-
averaging methods discussed previously. The second is a mixtures
data base, currently containing summary information on over
1,200 studies on toxic interactions such as synergism and an-
tagonism. Both data bases are being prepared so that they can be
accessible to the public to further improve risk assessment
methodology.
Administrator Lee Thomas has recently committed the U.S.
EPA to integrating risk reduction into the fabric of decision-
making. As part of this effort, Program Offices will use the tools
of risk assessment, among them the guidelines, to implement pro-
grams and set priorities. Two examples within the Office of Solid
Waste are the setting of Alternate Concentration Limits to protect
against contamination of groundwater and the use of risk assess-
ment to set priorities for banning wastes from landfills.
Finally, as the U.S. EPA emphasizes control of waste sites
(both through RCRA/HSWA and Superfund) the emphasis on
risk assessment will be at the sites themselves. Because of this
decentralization, the U.S. EPA is developing and implementing
programs to train field personnal in risk assessment techniques,
provide them with the necessary technical assistance and assure
the quality and consistency of the risk assessments carried out at
those sites.

CONCLUSIONS
Risk assessment at the U.S. EPA has evolved from an art
developed by a small group of people primarily discussing cancer
to a general analytic and decision tool used by many people in
many programs across the U.S. EPA. Furthermore, many of the
U.S. EPA's statutes are now predicated on a risk reduction basis
which requires more and better health risk and exposure analyses.
Since the possibility of overlapping and conflicting analyses ex-
ists, a larger role is necessary for review of risk assessments in
general, for oversight of the process and for development of more
detailed guidelines. The structures for assuring this quality and
technical consistency are now evolving within the U.S. EPA.
Therefore, as risk assessment becomes more sophisticated, as
more risk assessments are performed and as the need for
assurance of quality and consistency increases, the U.S. EPA will
develop guidelines for more end points, add more detail to ex-
isting guidelines, strengthen the management procedures for
building scientific consensus, publicize those resolutions and
maintain the appropriate degree of oversight. The results of this
process will be the development of better risk assessments with
less overall uncertainty and, ultimately, better protection of
public health.

DISCLAIMER
The views expressed in this paper are those of the authors and
do not necessarily reflect the views or policies of the U.S. EPA.
REFERENCES
1. National Research Council, Risk Assessment in the Federal Gov-
ernment: Managing the Process, National Academy Press, Wash-
ington, DC, 1983.
2. U.S. EPA, "Interim Procedures and Guidelines for Health Risk and
Economic Impact Assessments of Suspect Carcinogens," Fed. Reg.
41, 1976,21402.
3. U.S. EPA, "Mutagenicity Risk Assessment: Proposed Guidelines,"
Fed. Reg. 45, 1980, 74984.
4. U.S. EPA, "Ambient Water Quality Criteria Documents: Notice of
Availability,"Fed. Reg. 45, 1980, 79317.
5. U.S. EPA, "Guidance for Performing Exposure Assessments," un-
published draft, available from Science Advisory Board, U.S. EPA,
Washington, DC, 1983.
6. U.S. EPA, "Risk Assessment and Management: Framework for
Decision Making," Report #EPA-600/9-85-002, available from
Office of Policy, Planning, and Evaluation, U.S. EPA, Wash-
ington, DC, 1984.
7. U.S. EPA, "Proposed Guidelines for Carcinogen Risk Assess-
ment," Fed. Reg. 49, 1984, 46294.
8. U.S. EPA, "Proposed Guidelines for Mutagenicity Risk Assess-
ment," Fed. Reg. 49, 1984,46314.
9. U.S. EPA, "Proposed Guidelines for the Health Assessment of
Suspect Development Toxicants," Fed. Reg. 49, 1984,46324.
10. U.S. EPA, "Proposed Guidelines for the Health Risk Assessment
of Chemical Mixtures," Fed. Reg. 50, 1985,1170.
11. U.S. EPA, "Proposed Guidelines for Exposure Assessment," Fed.
Reg. 49, 1984, 46304.
12. N. Nelson, letter to Lee M. Thomas, Administrator, U.S. EPA, June
19, 1985, available from Science Advisory Board, U.S. EPA,
Washington, DC.
13. N. Nelson, letter to Lee M. Thomas, Administrator, U.S. EPA,
Mar. 14, 1986, available from Science Advisory Board, U.S. EPA,
Washington, DC.
14. U.S. EPA, "Guidelines for Carcinogen Risk Assessment," Fed. Reg.
51, 1986, 33992.
15. U.S. EPA, "Guidelines for Mutagenicity Risk Assessment," Fed.
Reg. 51, 1986, 34006.
16. U.S. EPA, "Guidelines for the Health Assessment of Suspect
Developmental Toxicants," Fed. Reg. 51, 1986, 34028.
RISK ASSESSMENT 271
-------
17. U.S. EPA, "Guidelines for the Health Risk Assessment of Chem- 21
ical Mixtures," Fed. Reg. SI, 1986, 34014.
18. U.S. EPA, "Guidelines for Estimating Exposures," Fed. Keg. SI, 22
1986, 34042.
19. International Agency for Research on Cancer, IARC Monographs 23,
on the Evaluation of the Carcinogenic Risk of Chemicals to Humans,
Supplement 4, Lyon, France, 1982.
20. U.S. Office of Science and Technology Policy, "Chemical Carcino-
gens: A Review of the Science and its Associated Principles," 24
Fed. Reg. SO, 1985, 10372.
Anderson, E.L. and Ehrlich, A.M., "New Risk Assessment Initia-
tives in EPA," Toxicol. Ind. Health 1,1985, 7.
Goldstein, B., "Strengthening the Assessment of Risk," EPA
Journal JO. #10, Dec. 1984, 5.
Kimmel, G.L., Kimmel. C.A. and Francis, E., "Proceedings of
the Consensus Workshop on the Relationship of Maternal and
Developmental Toxicity, May 12-14, 1986," Fund. Appl. Toxicol.,
to be published.
U.S. EPA, "Chemical Emergency Preparedness Program: Interim
Guidance, Revision 1, 9223.0-1 A," Nov. 1985.
272
RISK ASSESSMENT
-------
Comparative Risk Assessment of
Sources of Groundwater Contamination

Hope Pillsbury
U.S. Environmental Protection Agency
Office of Policy, Planning and Evaluation
Washington, D.C.
ABSTRACT
A wide variety of sources can contaminate groundwater. These
potential sources of contamination range from municipal landfills
and underground storage tanks to road salt application. This
study provides a single analytical framework which allows one to
compare the contamination potential of these widely disparate
sources. The study estimates the potential human health risk
posed by six types of groundwater contamination sources; the
risk is estimated both with and without considering the effect
taste and odor might have on an individual's exposure and his
subsequent health risk. The analysis also shows how contamina-
tion potential varies by environmental setting.

INTRODUCTION
A wide variety of sources ranging from large sources such as
municipal landfills to smaller sources such as septic tanks and
underground storage tanks can contaminate groundwater. His-
torically, at both the Federal and the state level, the responsibil-
ity for addressing possible contamination from these sources has
been split between different departments or even different agen-
cies. Consequently, no single analytical framework has been
available to compare the contamination potential of these widely
disparate source types. In addition to the need for a comparative
model, there is a need to determine how contamination poten-
tial may vary by environmental setting.
This study compares potential risks to human health from dif-
ferent sources of groundwater contamination. The purpose of
the model developed during this study is to provide a planning
tool which will help managers at the U.S. EPA or at the state level
set priorities among these sources. The model can be used to de-
termine which sources are good candidates for further study and
what the effect of different environmental settings is on these re-
sults. The analysis is too general, however, to support regulatory
efforts or to be applied directly to any specific site.
There are many ways to approach a comparison of these dif-
ferent sources. This study focuses on potential risk to a person
drinking water contaminated by one typical, example source type.
It provides estimates over time of the pollution expected at one
well located a specified distance directly down-gradient from the
source. It provides estimates of neither the number of other wells
affected nor of the total population affected.
It assesses the impact on groundwater from these pollution
sources in two ways: (1) the model measures the lifetime risk to
an individual drinking from a well and (2) the model produces
results based on the assumption that exposure would stop after
the individual tastes or smells pollution in water because an in-
dividual probably would not drink water that tastes bad.
The analysis does not provide estimates of the extent or volume
of contamination expected over time from a source type. Work
is underway to estimate damage to the groundwater resource.
This paper presents both an overview of the methodology used
in the analysis and the results from the comparison of source
types in different environmental settings. Additional analyses, re-
sults and supporting documentation are provided in U.S. EPA's
report, "Draft Comparative Impact Analysis of Sources of
Groundwater Contamination.'"

METHODOLOGY
The methodology entailed first choosing and characterizing
source types and environmental settings (Fig. 1). Each combina-
tion of source type and environmental setting was then modeled
using the U.S. EPA Office of Solid Waste's liner location model
(LLM).2 When data identifying the toxic pollutants and quantities
released from a source are provided, this model tracks the trans-
SOURCE SELECTION
- Potential Risk
- Analytical Feasibility
- Data Availability
- Usefulness of Results
SOURCE CHARACTERIZATION
- Size
- Constituents
- Failure/Release
- Geographic Distribution
ENVIRONMENTAL SETTINGS
- Depth to Ground water
- Net Hydrologic Recharge
- Aquifer Configuration
- Groundwater Velocity
CONTAMINANT TRANSPORT MODELING
- Contaminant Release
- Unsaturated Zone Algorithm
- Saturated Zone Model
- Distance to Receptors
HEALTH RISK ESTIMATION
- Dose Estimation
- Nature of Effect
- Dose-Response Function
- Risk Algorithm
RISK ANALYSIS
• Source Impact on
Human Health Risks
- Environmental
Vulnerability
Figure 1
Method for Analysis
RISK ASSESSMENT 273
-------
port of the pollutants in groundwater and predicts potential
health risk to an individual drinking that groundwater. The mod-
el originally was developed for use with hazardous waste land-
fills and surface impoundments. For this study, release data for
other source types were collected and analyzed using the LLM.
The results then were summarized both by source type and by en-
vironmental setting.

Sources
The study analyzes six source types chosen on the basis of con-
tamination potential, ease of modeling and expected usefulness
to the U.S. EPA. Each source was characterized in terms of iden-
tity, quantity and timing of the constituents released.
These sources are: municipal or "sanitary" landfills, industrial
surface impoundments, underground storage tanks, agricultural
feedlots, septic tanks and road salt. For each major category
there are two to seven subcategories based on either the size of the
source, the amount of release or the chemical constituents which
would be released. For example, the subcategories include small
and large municipal landfills; covered and uncovered salt piles;
and underground storage tanks containing gasoline, diesel fuels
and certain chemicals.
Pollutants
Modelled
Source Supcateoory Descrlptlcn

1. Septic 1»nk Leach Fields

a. High Density Field >40 tanks/ml* Sodlin chloride, nitrate,
methylera) chloride

b. Medina Density Field 10 to 40 tanks/ Sodium chloride, nitrate,
mi3 methylene chloride

c. Low tensity Field >in tenks/ml2 Sodium chlorine, nitrate,
meUiylene chloride
2. Agricultural Fredlou
Sodium chloride, nitrate
a. Beef Clttle, >1,000 23.550 m3
Hud FMdlots hold I TO rood

b. Beef Cattle,
tapotfrtnent
5. Unnenrmund Storage Tanks

a. Gasoline Ktael tank
n. Diesel Fuel

c. Chemical Storage

6. Road Delclng Salts
Steel tank

-------
found at a source. To be conservative, the pollutants chosen for
each source type were those which were particularly toxic, mobile
and nondegradable.
The analysis was limited to toxic pollutants (Fig. 2). No con-
trols limited the movement of the pollutants from the sources.
The single exception was covered salt piles, which were assumed
to be covered only one-half of the time.
Materials were stored, used or disposed by the sources for a
period of 30 yr. Chemical constituents were released by the
sources during the 30-yr period only, except for municipal land-
fills which, because of their size and the nature of their releases,
were assumed to release contaminants for as long as 200 yr.
Underground storage tanks were assumed to start leaking at age
25 and release all contents by year 30.

Environmental Settings
Releases from the source types were calculated using 72 en-
vironmental settings. Each environmental setting is defined by
depth to groundwater, net recharge (Fig. 3) and flow field (Fig.
4).
The model has four possible depths to groundwater ranging
from 1.5 meters to 22.9 m. There are four net recharge values
ranging from 1 in. (2.5 cm) to 15 in. (38 cm). The model requires
that several other important hydrogeologic variables such as ca-
tion exchange capacity, fraction of organic carbon in soil and soil
porosity and permeability be held constant in the analysis.
There are nine flow fields which have horizontal groundwater
velocities ranging from 1 m/yr to 10,000 m/yr. Certain flow fields
are further differentiated because they contain multiple layers
(aquifers and aquitards).
The model is appropriate only for environments in which the
aquifer or aquitard layers modeled are both homogenous and
isotropic. Therefore, it is not suitable for use in areas with karst
or fractured bedrock.

Contaminant Transport Modeling
The groundwater portion of the LLM has two key compon-
ents. First, the unsaturated zone is modeled using a simple algo-
rithm, the McWhorter-Nelson Wetting Front Equation. This
equation provides a time delay for pollutants to travel through the
unsaturated zone.
The second component is a modified version of the Pricket-
Lonquist random walk model which is a two-dimensional model
in the vertical plane. The transport of pollutants in groundwater
is modeled over a 400-yr time horizon (as long as 200 yr of release
plus 200 yr of transport after the last release) at a well located
600 m hydraulically down-gradient of the pollution source. Other
well distances are possible but, for simplicity, this analysis is re-
stricted to the 600-m distance. The model does not account for
transverse dispersion (i.e., horizontal dispersion perpendicular to
the direction of flow).
Health Risk Estimation
Human health risks are estimated for one individual drinking
2 I/day of water for 70 yr from a well which has been contami-
nated by the source type. Risk is defined in terms of the likelihood
that the single individual will experience the health effect of con-
cern. Thus the incidence rate of disease attributable to the pollu-
tant being analyzed is the unit of measure. Health effects are all
chronic but range from hypertension to cancer. Where there are
several constituents that an individual will be exposed to, risks are
assumed to be additive. Information on the hazard of the indi-
vidual constituents is from the U.S. EPA's Cancer Assessment
Group and EPA's Office of Solid Waste.
Presentation of Results
Results are presented assuming that disease incidence of over 1
in 100,000 is "significant." An individual would have a 1 in
100,000, or 10~5, chance of developing the disease or diseases
associated with the contaminants if exposed for the period of time
defined in the model. Any risks less than 10~9 (one in a billion)
were treated as zero risks.
If the source causes significant risks in more than 40% of the
environmental settings, it is labeled a "frequently significant"
risk source. Risks were termed "seldom significant" if they oc-
curred in less than 40% of the environments. Risks are termed
"never significant" if they never reach or exceed the "signifi-
cance" threshold of 1/100,000.
The analysis can use a different threshold of risk, such as 10~6
or 10~4. Additional sensitivity analyses were performed at these
risk levels to determine the difference they would have on the
final results.

RESULTS
Groundwater Risks by Source
There are six source categories divided into 22 subcategories.
As shown in Fig. 5, nine source subcategories are classified "fre-
quently significant" (significant in 40% or more of the settings).
These sources include iron and steel and organic chemicals sur-
face impoundments which show high risks in every environmental
setting modeled.
Four source subcategories show risk in less than 40% of the
environmental settings: gasoline and chemical underground stor-
age tanks, pulp and paper surface impoundments and large beef
feedlots.
Nine source subcategories showed no significant risks. Of
these, four posed some risk, although at levels which are lower
than the "significance" thresholds: high, medium and low den-
Source Subcateqory
Percent of Settings,
With risk >1Q-5 3
With Taste/Odor
Threshold:
Percent of Settings
With Risk >10"5
Frequently Significant
Iron & Steel Surface Impoundment (SI) 100.0 100.0
Organic Chemicals SI 100.0 50.0
Acid Mine Drainage SI B7.5 87.5
Oil t, Gas Brine SI 87.S 0.0
Municipal Waste Water Trtmt. SI 87.5 87.5
Large Municipal Landfill 57.8 57.8
Small Municipal Landfill 57.8 57.8
Uncovered Salt Storage Pile 57.8 0.0
Covered Salt Storage Pile 40.6 0.0

Seldom Significant
Gasoline Underground Storage Tank (UST) 37.5 12.5
Chemical UST 25.0 0.0
Pulp and Paper SI 25.0 0.0
Large Beef Feedlot 12.0 0.0

Never Significant
High Density Septic Tanks 0.0 0.0
Medium Density Septic Tanks 0.0 0.0
Low Density Septic Tanks 0.0 0.0
Small Beef Feedlot 0.0 0.0
Turkey Feedlot 0.0 0.0
Diesel Fuel UST 0.0 0.0
Road Salt Applic. (10 tons/lane-mile) 0.0 0.0
Road Salt Applic. (1 ton/lane-mile) 0.0 0.0
Alkaline Mine Drainage SI 0.0 0.0

Notes
1. For all subcategories there are 64 environmental settings. Originally there were 72, but in all
eight of the settings with flow field A, the groundwater velocity was too slow to transport
contaminants 600 meters downgradient within 200 years.
2. Risk refers to risk of disease, which may vary in severity from hypertension to cancer.
3. Assumes an individual would drink contaminated water for 70 years, regardless of its taste
or odor.

Figure 5
Frequency Table—Settings in which Average 400-Year Risk for an
Individual Exceeds 10 ~5 '-2
RISK ASSESSMENT 275
-------
sity septic tanks, pulp and paper surface impoundments and die-
sel underground storage tanks. The other five subcategories did
not pose health risks in any of the environmental settings: small
beef feedlots, turkey feedlots, both types of road salt applica-
tion and alkaline mine drainage surface impoundments.
Sensitivity analysis caused a few changes in the results. Using a
threshold of 10~6 caused gasoline USTs and pulp and paper sur-
face impoundments to switch from the "seldom significant" cate-
gory to the "frequently significant" category. Using a threshold
of 10~4 caused municipal landfills to switch from the "frequent-
ly significant" category to the "seldom significant" category;
chemical USTs and large beef feedlots switched from the "seldom
significant" category to the "never significant" category.
In most cases, a few pollutants drive the risk results for each
source type. For example, there is some controversy as to whether
arsenic is truly a carcinogen. If it is assumed that arsenic is not
carcinogenic, the risk category for municipal landfills (using a
10"5 threshold) drops from the "frequently significant" category
to the "never significant" category.

Groundwater Risks by Source
(Taste and Odor Thresholds Considered)
The previous results assumed a person would drink water re-
gardless of its taste or odor. The second form of the results as-
sumes that if an individual tastes or smells chemicals in ground-
water, he or she will stop drinking the water. Exposure then
would occur only when contamination levels are lower than levels
detectable by humans.
The literature showed that most organics can be tasted or
smelled.' No taste or odor thresholds for any of the metals in the
analysis have been found (Fig. 6).
The results showed that risks were reduced to zero for oil and
gas brine surface impoundments, salt piles, chemical under-
ground storage tanks, pulp and paper surface impoundments
and large beef feedlots when the taste and odor threshold of the
organic chemicals or the salt was taken into account. Risk from
Contaminant
Taste or Odor
Threshold
Arsenic na
Bariun na
Benzene 0.024
Cadmium na
Chloroform 0.1
Chromiun na
Cyanide na
1,1-Dichloroethane na
Ethyl Benzene 0.64
Ethylene Dibronide na
Lead na
Mercury na
Methyl Ethyl Ketone 1.0
Methvlene Chloride 3.2
Nitrate na
Phenol 0.0001
Sodium Chloride 200.0
Tetrachloroethylene 0.3
Tetraethyl Lead na
Toluene 0.024
1,1,1-Trichloroethane 0.16
Trichloroethylene 0.5

na = not available.
Source: U.S. EPA, Office of Solid Waite, "RCRA Rijk-Cosl Analysis Model." 1984.

Figure 6
Taste and Odor Thresholds for Pollutants Modeled
(Concentrations given as mg/1)
gasoline underground storage tanks and organic chemical surface
impoundments was reduced only partially because some pollu-
tants released by those sources continued to cause some signifi-
cant risk at levels below the taste and odor threshold.
The analysis showed that groundwater monitoring is particu-
larly important down-gradient of sources whose contaminants
may not be detectable by taste and odor. This monitoring is im-
portant for municipal landfills and several surface impoundments
because the metals released by these sources would not be tasted
or smelled and the organics concentrations were below taste and
odor thresholds.

Interaction Between Sources and
Environmental Settings
The study showed that within any particular source type, risks
generally are sensitive to velocity and are not particularly sensi-
tive to net recharge or depth to groundwater. For many source
types (septic tanks, underground storage tanks, surface impound-
ments and feedlots), the hydraulic head of the liquid in the source
provided more of the driving force for pollutant release than did
net recharge. Exceptions were municipal landfills and road salt
storage piles for which additional net recharge caused a greater
mass of contaminants to be released over time.
Another finding was that degradeable constituents (organics,
etc.) and nondegradeable constituents (metals, etc.) behave dif-
ferently in groundwater traveling at a given speed. In slow mov-
ing groundwater, metals may be of more concern because they
are not degradeable and thus, although they disperse some, they
still contaminate the well. Organics may have degraded by the
time they reach the well.
In contrast, in fast moving groundwater, organics may be of
more concern because they have had less time to degrade. Metals,
on the other hand, could be fairly widely dispersed by the time
they reach the well.

Comparing Modeled Concentrations to
Monitoring Concentrations
Results of the study were tested by finding studies in the litera-
ture of wells located in similar environmental settings, at similar
distances from the sources and contaminated by the same source
types as those modeled. A comparison was made between the
levels of contamination found at wells in the studies with the con-
tamination levels estimated in this study. The number of appro-
priate studies from the literature was small, but the comparison
showed the modeling to be reasonably accurate for metals. In
several instances, the modeling estimates underestimated the con-
centrations from organics. There were not enough data in the lit-
erature to check underground storage tanks or industrial (non-
hazardous) surface impoundments.

CONCLUSIONS
Several conclusions result from this analysis. First, interac-
tions between sources and environments are complex and should
be taken into account when developing strategies to prevent or
control groundwater contamination. Plans and strategies should
allow for the variation in pollution potential caused by the en-
vironmental setting of the source type.
Monitoring strategy should take into account the interaction
between environmental settings and sources. Monitoring should
be more frequent in faster flow fields. Monitoring is particularly
important where the pollutants released are not detectable by
taste or odor.
Nine sources showed "frequent significant" risks, four showed
"seldom significant" risks and nine showed "never significant"
risks. Risk from a number of these sources was significantly re-
276 RISK ASSESSMENT
-------
duced when human detection through taste and odor was taken
into account. Organic chemical surface impoundments, oil and
gas brine surface impoundments, pulp and paper surface im-
poundments, salt storage piles, chemical and gasoline under-
ground storage tanks and large beef feedlots fell into this cate-
gory.
Some sources are affected more by differences in environmen-
tal setting than others. Iron and steel and organic chemical sur-
face impoundments caused significant risks in all of the environ-
mental settings. Municipal landfills, certain other industrial sur-
face impoundments and salt piles are among the source types
that potentially cause high risk in more than one-half of the en-
vironments. Risks from gasoline and chemical underground stor-
age tanks, pulp and paper surface impoundments and large beef
feedlots caused significant risks in less than one-half of the en-
vironments. Many source types never resulted in significant risk,
regardless of environment. The cause of variation in source risk
by environment can be due either to the interaction of the
source's pollutants with the environment (factors such as degra-
dation and mobility), or the interaction of the source type with
the environment (factors such as net recharge or mass of pollu-
tants available for release).
This study is most useful as a framework for considering how
risks from sources compare and how they are affected by en-
vironmental setting. Further studies should be performed to con-
firm and elaborate on these results.

ACKNOWLEDGEMENTS
The author wishes to thank Cliff Rothenstein of the U.S. EPA,
who developed the liner location model (the tool used in this
analysis) and Randy Freed of ICF, Incorporated, the lead con-
sultant for the analysis on which this paper is based. Thanks
also go to Jeanne Briskin of the U.S. EPA, who co-managed this
project in its early stages. This paper reflects the view of the
author and does not, in any way, reflect the views, opinion or
policy of the U.S. EPA.

REFERENCES
1. U.S. EPA, Office of Policy, Planning and Evaluation, "Draft Com-
parative Impact Analysis of Sources of Groundwater Contamina-
tion," Jan. 1987.
2. U.S. EPA, Office of Solid Waste, "Liner/Location Risk Analysis
Project, Draft Report," 1986.
3. U.S. EPA, Office of Solid Waste, "RCRA Risk-Cost Policy Model,"
1984.
RISK ASSESSMENT 277
-------
Application of Expert Systems for
Environmental Engineering Decisions
Donald R. Brenneman
NUS Corporation
Pittsburgh, Pennsylvania
INTRODUCTION
Expert systems are becoming increasingly popular and interest
in their use for management engineering and scientific purposes is
rapidly increasing. Expert systems are a class of computer-based
programs that guide an informed program user through the logic
of a problem, asking questions and performing calculations to
solve a problem. These systems typically employ a rule-based pro-
duction system with built-in classification rules and scoring pro-
cedures derived from the knowledge and experience of experts in
the field of interest. They are designed to make the knowledge
and skills of an expert available to non-experts for problem solv-
ing.
Several potential applications for expert systems in the field of
environmental engineering include the development of permitting
advisory systems and risk assessments and in the selection of
remedial actions as part of the remedial investigation/feasibility
study process. Additionally there is potential application in the
development of discipline-specific advisory modules to assist
users in selecting the most favorable course of action.
Expert systems can be used to increase our ability to utilize
available knowledge that is generally not immediately available.
These systems are built in by capturing and encoding the relevant
experience of experts in a particular field of interest.
Our interest in this subject first emerged during the early days
of Superfund under the REM/FIT Contract. The prevailing guid-
ance for rating feasible remediation alternatives at that time was
to use a trade-off matrix approach. A table was prepared to show
the rating of each alternative/evaluation criteria combination.
The matrix includes many of the following measures of effec-
tiveness:

Capital cost
Operation and maintenance
Level of cleanup achieveable
Time to cleanup
Feasibility
Implementability
Reliability
Ability to minimize adverse impacts during action
Ability to minimize off-site impacts because of action
Remoteness of activities
Usability of groundwater
Usability of surface water

Each of the alternatives was rated 1 through 5 for each
measure, with 1 being the worst score and 5 the best. The formula
for an overall cost-effectiveness rating was as follows:

(Capital Cost Rating + O&M Cost Rating) x Effectiveness Rating
Sum = Cost-Effectiveness Rating
Alternatives were then grouped into categories of effectiveness
—high, medium and low. Depending on the level of effectiveness
judged by the U.S. EPA and other applicable regulatory agencies,
the least costly alternative^) were identified.
The matrix approach is no longer used being instead replaced
by a dialectic approach. The matrix approach was somewhat dif-
ficult to use and in some situations, quite subjective. However,
the matrix approach did provide the inspirational insight into tak-
ing a more rigorous scientific approach to decision-making.

TYPES OF EXPERT SYSTEMS
Fuzzy Set Logic
Our first attempt at scientific decision-making had its roots in
nuclear power plant siting studies. We used fuzzy logic theory to
assist selected decision problems. The rigorous mathematics of
fuzzy logic was originated by Dr. Lofti Zadeh. Typical the
algorithm for our decision model has been similar to that used by
economics and management scientists.2 The method of weighting
criteria upon which alternative decisions are based is credited to
Sealy.1 For further information on this subject, refer to the work
of Zadeh,3-4 which provides the concepts of fuzzy sets and fuzzy
algorithms. A microcomputer program listing for fuzzy decision
making is provided in Table 1. This program can be compiled
with Microsoft's QuickBasic Compiler (Version 2.0).
To use the program the user first enters the number of alter-
natives that are to be evaluated. Then, the number of decision
criteria are entered. The program will then ask the user to enter
the relative importance of each criteria on a scale of 0 to 10. The
final requirement for user input is a paired comparison analysis.
For this particular program, the user must first decide which
criterion is more important from a series of paired criteria sets
that are presented on the screen. At this same time, the relative
magnitude of importance for the criterion selected must be
entered. The magnitude of importance is based on a scale of 1 to
10. After all inputs are completed, the program performs a matrix
analyses and then prints out the consistency of the paired matrix
comparison and the best choice alternative according to the data
that was entered. Functionally this exercise is equivalent to the
former matrix analysis technique with the exception that some
measure of consistency for the evaluation criteria is presented.
Unfortunately, there are shortcomings to the fuzzy logic ap-
proach as presented. A major problem is that only an expert who
really "knows" the relative importance of the decision criteria
can effectively use it. This can be demonstrated by comparing the
consistency of paired matrix analysis when the same problem is
run by an expert and a novice. Considering that the overall objec-
tive was to develop a methodology with more universal applica-
tion, attention focused toward expert systems which utilize pro-
278 RISK ASSESSMENT
-------
..(or 0 to Quit): ";B
.(or 0 to Quit) ";A
Table 1
A Program for Fuzzy Decision-Making

DIM DC<20),E(20),D(20),R(20),C(20),F(20)
DIM H(20,20),A(20,30)
DIM CS(20),LS(20)
COLOR 11,0,0:CLS:KEY OFF
• ••**** MAIN I/O ROUTINE • • • • •
MAIN:
PRINT: INPUT "NO. OF ALTERNATIVES
IF B<1 THEN CLS:END
PRINT:PRINT "ENTER LABELS FOR ALTERNATIVES ...
FOR 1-1 TO B:INPUT L$(I):NEXT I
100 PRINT:INPUT "NO. OF CRITERIA FOR THE DECISION
IF A<1 THEN CLS:END
IF A > B+l THEN GOSUB SUB1
PRINT:PRINT "ENTER LABELS FOR CRITERIA ..."
FOR 1-1 TO A:A(I,I)-1.0:INPUT C$(I):NEXT I:PRINT
PRINT:PRINT "RATINGS OF THE ALTERNATIVES"
PRINT "****e«i«*****»»li««e»*«**«l*«e«*Se«****eeeel>e": PRINT
PRINT "IN EACH CASE ENTER A VALUE BETWEEN 0 AND 1;"
PRINT "WHERE A HIGHER VALUE MEANS ' BETTER'.": PRINT
PRINT: FOR 1-1 TO A: FOR J-l TO B
PRINT:PRINT "ENTER ";C$(I);" RATING FOR ",-LS(J)
INPUT W(I,J):NEXT J:NEXT I
PRINT: PRINT: PRINT
PRINT "RATINGS FOR CRITERIA"
PRINT "••eea«eeeee6eeeeeeeeeeeeeeeSee«eeeeeeee*ii0 GOTO 300:NEXT I
GOTO 400
300 FOR 1-1 TO A:C(I)-F(I) :NEXT I
GOTO 200
400 WX-HW
PRINT: PRINT "EIGENVALUE - ";WX
FOR 1-1 TO A:E(I)-F(I)«A:NEXT I
PRINT:PRINT "EIGENVECTOR ... ":PRINT
FOR 1-1 TO A:PRINT F(I)!:NEXT I
PRINT:PRINT:PRINT "ALPHA-VECTOR ... ":PRINT
FOR 1-1 TO A:PRINT E(I)!:NEXT I
MU-(WX-A)/(A-1)
Q-SQR(MU/2)
PRINT
PRINT:PRINT "CONSISTENCY OF THE PAIRED-COMPARISON MATRIX - ";Q
PRINT "(value should be < 1; if > 2, repeat the process)"
RETURN
>*•*.. PRINT RESULTS * * * • *
PRINTOUT:
FOR J-l TO B: DC(J)-9999.
FOR 1-1 TO A: W(I,J)-W(I,J)«E(I):IF W(I,J)MX THEN MX-DC(I)
CH-I:NEXT I:PRINT
PRINT L$(CH);" ! ": PRINT
PRINT "...is the best choice according to the data you have entered.
PRINT
PRINT " (press to continue or to quit) M
WHILE RR$ <> CHHS(13)
1000 RR$-INKEY$: IF RRS-"" THEN GOTO 1000
IF RR$-"Q« THEN CLS:END
WEND
GOTO MAIN
SUB1:
WHILE TS <> CHR$(13):PRINT
PRINT "The number o£ Criteria cannot be > than Alternatives + 1, "
PRINT "(try again ...press to continue...)"
2000 T$-INKEY$: IF T$-"" THEN GOTO 2000
WEND:CLS
RETURN 100

duction rules. Such systems are also called knowledge-based

systems.

Knowledge-Based Systems

Knowledge-based systems incorporate data about a subject or

group of subjects and the expert knowledge that represents how

that data is to be interpreted and used. In other words, knowl-
edge-based expert systems: (1) use specialized knowledge about a
particular problem area (e.g., hydrogeology or environmental
risk) rather than just general purpose knowledge that would apply
to all problems, (2) use symbolic (and often qualitative) reasoning
rather than just numerical calculations, and (3) perform at a level
of competence that is better than that of non-experts.
One of the most important ways in which expert systems differ
from traditional computer applications is in their use of heuristic
reasoning. Traditional applications are well understood and
therefore can employ algorithms that when followed lead to the
correct conclusion. For example, aquifer permeability can be
calculated according to a precise set of rules. A heuristic system
involves judgmental reasoning via trial and error and is most ap-
propriate for complex problems. Heuristic decision rules or in-
ference procedures generally provide a good, but not necessarily
optimum answer. This is because a heuristic rule is one which
when applied only moves us closer to the true solution.
A typical expert system (including knowledge-based systems)
will include a user interface for communicating with a human
user, a knowledge base of facts and rules related to the problem
and an inference engine or reasoning methods for utilizing the in-
formation in the knowledge base to solve problems.
In order to effectively use the specialized knowledge about
many different kinds of problems, artificial intelligence research-
ers have developed a number of "knowledge representation"
techniques. For this presentation, the technique for encoding
knowledge is through the use of production rules. Production
rules are particularly useful in building systems based on heuristic
methods.
Production rules are simple IF-THEN constructs that are used
to represent the empirical consequences of a given condition, or
the action that should be taken in a given situation. For example,
a simple hydrogeology expert system might have a rule like:
IF—it is necessary to describe the characteristics and properties
of water-bearing zones, and
IF—the aquifer hydraulic conductivity is less than or equal to
28 ftVday
THEN—consider using slug tests
For further understanding production rule systems, refer to Table
2, which is a template representation of a typical knowledge-
based system.

EXPERT SYSTEMS APPLICATIONS
Considering the above, several knowledge-based expert systems
have been developed with unique objectives in mind. In one case,
a remedial technology expert system was constructed to inform
the user of the type of data that should be collected for each ap-
plicable technology. For example, if a user believes that a slurry
wall is a candidate remedial technology, the expert system would
advise that sufficient borings should be made to determine if there
is an approproate substrata into which the wall can be keyed. Fur-
ther, the user is advised to look for geotechnical contraindications
such as soil stability, earthquakes or adjacent streams. Function-
ally, this system can be called an advisory module that can be in-
corporated into a multi-objective system.
In a second case, a remedial alternative expert system was
developed to accomplish the same objective as the matrix ap-
proach described previously. However, in this case the system em-
bodied the knowledge of experts and the user did not need to be
an expert. Procedurally, this system builds upon the input of user-
defined site problems. Through a series of computer generated
questions and responses from the user, one or more conclusions
are made regarding the preferred alternative.
It is important to understand that these expert systems are real-
RISK ASSESSMENT 279
-------
ly never "complete" in their design. New rules and data must be
added to improve the systems and to refine the rules already in
place. In a sense, system maintenance will be perpetual. Just as
the human knowledge base grows and improves, so must a com-
puter knowledge-based expert system.
To demonstrate how a remedial alternative selection expert
system might be applied, it is informative to consider the analysis
of one operable unit for a hazardous waste site remediation pro-
ject. At this site, the groundwater was contaminated with volatile
hydrocarbons. Also, the contaminated aquifer was used as a
source of potable water. In this case the general response action
(GRA) was the treatment of contaminated aqueous and liquid
waste streams. The most applicable treatment category for this
GRA included physico/chemical treatment technologies. The in-
dividual technologies within this category were:

• Flow equalization
• Flocculation
• Sedimentation
• Activated carbon
• Adsorbents
• Ion exchange
• Reverse osmosis
• Liquid/liquid extraction
• Oil/water separation
• Steam distillation
• Air stripping
• Steam stripping
After a series of questions and answers the expert system ad-
vised that the following remedial alternatives were applicable to
this operable unit:

• Activated carbon
• Air stripping

The system also cautioned the user to consider treatment of the
air stripper off-gas based on the concentrations of hydrocarbons
in the feed stream.
While this example may appear to be straightforward and in-
tuitive, consider that: (1) the user was not an expert, (2) experts
were not consulted to determine all of the relevant issues, (3)
calculations were made to estimate the functional design require-
ments for the alternatives, (4) for the air stripper alternative, the
potential off-gas problem was identified, and (5) the time to reach
the correct conclusions took less than 1 hr. The only requirement
for the exercise was that the user had to be thoroughly familiar
with the site and be able to respond correctly to the questions.
Table 2
Knowledge Base Template
TITLES
This is were the titl« display or sign-on screen
for th« knowledge base la placed.
FACT DECLARATIONS
Certain data declaration will usually b« required. Thi*
requirement will depend upon the expert system •hell that
i« u*ed.
CONFIDENCE

This ii where the confidence handling toggle !• placed.

THRESHOLD
The threshold define* the minimum amount of confidence a fact
•uit have before It can fire a rule.
COALS
1. This Is the goal of the knowledge base
2. This it a competing goal
2.1 This la a subgoal
2.2 This is a competing subgoal
RULES
Here la where all of the rules of the knowledge
base are placed.

IF ... THEN
TEXT
To be most effective, the knowledge base program should allow for
text information that can be attached to each fact.
e.g.
confined aquifer...

What is the depth of the aquifer ?
EXPAND
Here is where explanatory Information is attached to a fact.
This information can explain to the user why the information
requested is required.
DISPLAY
Here is where displays unique to the knowledge base are placed.
These displays are available for viewing on command during a
session.
REFERENCES
CONCLUSIONS
Based on the experience gained, we conclude that expert system
technology will gain an increasing role in the field of hazardous
waste management. With expert systems, the knowledge of a
relatively small number of experts can be made available to many,
at a significant savings in cost and time. Additionally, the
remediation experience that has been gained can be used advan-
tageously in future remediation projects. An attendant benefit is
that the learning curve for new entrants into the hazardous waste
field can be significantly improved.
1. Sealy, T.L.. "A scaling method for priorities in hierarchical struc-
tures," J. Math. Psych.. 15, 1977. 234-281.
2. Yager, R.R., "Multiple objective decision-making using fuzzy stts,"
International J. Man-Machine Studies, 9, 1977, 375-382.
3. Zadch, L.A., "Fuzzy sets," Information and Control, 8, 196S, 338-
353.
4. Zadeh, L.A., "Fuzzy algorithms," Information and Control, 12,
1968, 94-102.
5. Zadch, L.A., "A fuzzy-set theoretic interpretation of linguistic
hedges," J. Cybernetics, 2, 1972, 4-34.
280 RISK ASSESSMENT
-------
Communicating Risk Assessment Findings
To the Public: Approaches that Don't Work
And One that Might
Robin Sandenburgh
Marion Cox
ICF Incorporated
Washington, D.C.
INTRODUCTION
During the last several years, the topic of "risk communica-
tion" has become an extremely popular and widely discussed
issue among environmental professionals, particularly among
federal and state agency personnel who often must try to convey
complex scientific information to a wide range of interested and
affected "publics." Many of these government decision-makers
are finding that their efforts to balance the benefits and costs of
risk decisions in environmental programs, and to credibly defend
their assumptions and conclusions regarding those decisions,
come under considerable scrutiny from all quarters, including
local officials, residents and private sector interests. These groups
increasingly take issue not only with the conclusions of their
analyses but with the very premises upon which these analyses are
based.
While this trend toward greater scrutiny of risk-related deci-
sions has been, for several decades, a part of public health and
consumer safety issues, these questions are arising with increased
frequency for other government agencies and programs. Indeed,
the scope of this heightened interest is having a significant effect
on the government's ability to make decisions in a timely and ef-
fective manner—and, in some cases, to make decisions at all. No-
where is this trend more evident than in the area of hazardous
waste management programs. Both private and public sector in-
terests, as well as the general public, are demonstrating increased
concern over how a wide range of hazardous waste management
decisions are made, and these same groups are increasingly
sophisticated in their efforts to affecting these decisions.
This "frontal attack" on government decisions has increased
dramatically since the early 1980s, when national attention was
focused on hazardous waste sites such as Love Canal, Times
Beach and similar incidents. In particular, government efforts to
communicate the relative risks' associated with hazardous waste
facilities and sites have been less than effective in meeting either
of two objectives: (1) allaying public and private sector concerns,
or (2) gaining public acceptance of proposed actions, based on the
findings of a site- or facility-specific risk assessment.
It is our belief, based on more than 15 yr of experience working
with federal and state government agencies implementing con-
troversial environmental and land use programs, that attempts to
communicate with affected publics about programs or actions
that involve risk often fall short of these objectives. Overall
strategies or approaches to risk communication have been, for the
most part, non-existent until such time as misunderstanding and
confusion occurs, at which point the over-riding approach is one
1. See W.F. Allman, "Staying Alive in the 20th Century" (Science 85,
October 1986) for a discussion of the public's perceptions of relative risk.
of "damage control." Where communication has occurred, it
typically takes place well into the project and focuses on the
results of findings of analyses that already have been concluded.
We believe communication efforts of this type are a result of
decision-makers' attempts to communicate with the public on
risk-related topics without first listening to, analyzing and under-
standing the sources of the public's specific concerns and ques-
tions.
In this paper, we discuss several common approaches to com-
municating risk-related information to the public, using risk
assessment findings as an example of the kind of information to
be conveyed. We then describe why these approaches may not
achieve the desired objectives. Finally, we outline an approach
that we believe is more likely to be effective in meeting the objec-
tives set forth above.

APPROACHES THAT FALL SHORT
In our experience government attempts to communicate risk-
related information to the public do not "fall short" for lack of
effort. Within the last several years, government agency staffs
responsible for making decisions related to environmental risks
have devoted considerable energy and resources to improving
their risk communication abilities. Rather, these approaches are
frequently less successful than desired because, in our view, they
do not address specific, fundamental and localized public con-
cerns. Instead, typical approaches to communication end up ad-
dressing government's perceptions of public concerns rather than
the public's actual concerns. Thus, at best these approaches suc-
ceed only partially and, in some instances, they do not succeed at
all.
Below we present three of the most common of these ap-
proaches. Each has elements that may be effective, some of which
are components of a preferable approach that we advocate in the
final section of this paper.

Overwhelm the Public with Information
A commonly used approach to conveying risk information to
the public is to present the information in exhaustive scientific
detail, under the assumption that public fears will be reassured by
the depth and breadth of scientific expertise being brought to bear
on the situation. A frequent corollary to this approach is to
"bring in the experts" to present the results of a risk assessment
or other scientific study.
At the root of this approach is a recognition of several impor-
tant factors that typically drive public concern. First, the public
all too often believes that it receives insufficient information
about either the government's decisions or its decision-making
process. Thus, any attempt by the government to respond to this
concern is a step in the right direction. Second, it is important to
RISK ASSESSMENT 281
-------
ensure the public that competent, credible personnel are working
to address their problem or respond to their concerns.
Why, then, does this approach fall short? In most cases, this
occurs not because the approach is fundamentally flawed; rather,
because it represents "too little, too late." Using the example of a
risk assessment conducted in the context of an RCRA permit for a
hazardous waste facility, it may not be sufficient to "inform" the
public of the method and results of a risk assessment that is
already completed. The assessment may focus on parameters or
media that are important to the toxicologist or other health pro-
fessionals who designed and conducted the study; however, these
findings may neglect concerns that are of primary importance to
the local community. Moreover, rather than inspiring confidence,
heavy reliance on scientific experts—especially in the context of a
public meeting or hearing—can appear to the public to be a
"snow job:" the "selling" of the findings of a risk assessment in
which they, members of the public, had no significant or mean-
ingful role.
In our experience, the result of this approach can be to reduce,
rather than maintain or enhance, the government's credibility in
the eyes of the public, a result that is exactly the opposite of what
is intended.

Finding the Right Words
Another frequently used approach to communications risk-
related concepts and findings to the public is to seek to
"translate" risk information in such a way that the public will
come to the same conclusions about the relative risk of a given
situation that the professionals who designed and conducted the
study have reached. Indeed, one important goal of the heightened
interest in risk communication in recent years seems to be to
discover the right words or analogies by which to convey the
government decision-makers' view of a particular risk to the in-
terested public. This approach seems to convey the belief, on the
part of both government decision-makers and risk assessment ex-
perts, that, "The public would accept what we're trying to do if
only we could find the right words to explain it to them."
Moreover, most attempts to date to "find the right words"
have involved comparative approaches: comparing the risk posed
by the environmental problem in question to another, more well-
known risk.J In this way, it is generally believed, the public will be
better able to understand the environmental risk in the context of
a risk that is more familiar to them. On the surface, this approach
is a logical one: comparative analysis is the means by which most
of us make a wide range of personal risk/benefit decisions. In
practice, however, comparative approaches have tended to fall
short of the desired objectives and, at times, to increase—rather
than reduce—public opposition to the government's findings or
decisions.
The reasons for the ineffectiveness of comparative approaches
have been discussed widely by others, and have focused on impor-
tant issues such as whether the risk is voluntarily or involuntarily
assumed, the extent to which risks are "equitably" distributed,
and the degree to which those assuming the risk have control over
managing the risk. In our view, however, an equally important
reason for the failure of the comparative approach is the extent to
which the bearers of the risk have been involved in the risk
analysis and assessment process prior to the presentation of the
findings of that process. Without the early involvement of the af-
2. See Allman, ibid., and Bradley R. Brockbank, "Contrasting Risk
Communication Tasks and Objectives for Superfund Sites, Underground
Storage Tanks, and Biotechnology Field Tests." Paper presented to the
Annual Conference of the Sociedty for Risk Analysis, Nov. 12, 1986.
fected community, attempts to communicate risk information by
comparative means are frequently in danger of being perceived by
members of the public as insensitive to their concerns and conde-
scending of their ability to comprehend risk-related concepts. In
addition, important assumptions made by technical experts are
not always compared to or evaluated against the local commun-
ity's assumptions at a point when such comparisons are most
valuable in ensuring community understanding and acceptance of
the risk assessment method and findings.

Fulfilling the Regulatory Requirements
For the foreseeable future, government agencies will continue
to have to respond to both the public's demands to receive con-
siderably more information about the risks associated with hazar-
dous waste facilities and expanded legislative provisions regarding
the public's involvement in hazardous waste decisions. It will be
essential, therefore, for government decision-makers to avoid
what has been, in the past, an approach to risk communication
that focused exclusively on fulfilling legislative "public participa-
tion" requirements rather than on meeting the public's need for
information. This approach can be characterized as "going
through the motions" of communicating with the public: holding
the requisite number of public meetings, delivering presentations
on risk and responding to citizens' questions.
Here again, this approach is a fundamentally sound one, on the
surface, providing the community with at least some access to
government decision-makers regarding risk management deci-
sions. In many cases this level and type of involvement in risk
management decisions will be sufficient to meet the public's needs
and expectations. Where the approach falls short, however, is in
meeting the informational needs and expectations of communities
which for a wide range of reasons desire an expanded or enhanced
role in the risk assessment and risk management process.
In addition, because this approach focuses on fulfilling explicit
public participation requirements, it often has the effect of
"decoupling" the public involvement component of the decision-
making process from the risk assessment component of that pro-
cess. The result is that two components proceed on schedules that
are more or less independent and that come together only, for ex-
ample, at the point when a public hearing is held on the findings
of the risk assessment. In other words, the risk communication
process is a sporadic, independent process rather than an integral
part of the overall risk management and decision-making pro-
cesses.
From the point of view of concerned citizens, this approach can
limit public involvement to a "review and comment" role on the
findings of studies or analyses that the public had little or no role
in designing. As a result, the public has very little personal invest-
ment in either the risk assessment process or in the results of that
process and thus may be more inclined to question the credibility
of that process or the veracity of its findings.

Summary
It should be noted that the three approaches described here do
not cover the full range of approaches to risk communication, nor
are they mutually exclusive. Rather, they represent three of the
most common methods that government agencies use in com-
municating with the public about risk. It must be reiterated that
these approaches are not inherently flawed; rather, they may not
be appropriate in all situations or sufficient for the full range of
interested publics.
An emerging theme, in some quarters, to ensure more effective
risk communication, is to focus almost entirely on techniques for
communication. It is our strong belief that sound and effective
communication techniques have long been available and are often
used by government agencies and decision-makers. Rather, it is
282 RISK ASSESSMENT
-------
the absence of effective overall strategies for communication that
we need to develop and highlight much earlier in the decision pro-
cess.'
In the final section of this paper we describe an approach to
risk communication that, we believe, may lead to more favorable
results from both the public's and agency's perspective.

AN ALTERNATIVE APPROACH
TO RISK COMMUNICATION
The approach we suggest below draws upon our experience,
research and constant evaluation of our work and of other risk
communication approaches that we observe—including the three
described above—and adds elements that we view as extremely ef-
fective in other public involvement and decision-making pro-
cesses. Our approach consists of five fundamental principles:

1. Involve the public early in the risk management process.
The earlier the agency can involve members of the affected
publics, the better the two-way communication process is likely to
be. This means soliciting public input on the overall risk manage-
ment process, including the design of the risk assessment or other
key studies, the schedule for conducting the process and the
degree to which members of the general public will be involved in
the actual decision-making process. The longer the process is
underway without communicating with the public, the greater the
likelihood that interested or concerned citizens will feel disenfran-
chised from and distrustful of that process.

2. Identify and incorporate the public's concerns, perceptions,
and assumptions about the potential risks.
An all-too-common situation in environmental programs is to
reach the final stages of a risk assessment and present the findings
of that assessment to the public, only to find out that many in-
terested citizens are concerned about pollutants, media or
pathways of exposure that were not even addressed in the assess-
ment. At that point, it is generally unimportant that there may
have been valid scientific reasons for not addressing these specific
concerns; from the citizens' point of view, a lot of time and
money has been spent and they still do not have the information
or the assurances that they would like. Thus, at the earliest possi-
ble time in designing a risk assessment, it is essential to engage
members of the public in a two-way discussion about the objec-
tives, assumptions, method and expected outcomes of the risk
assessment process and to ensure, to the extent practicable, that
their concerns are addressed within the context of that process.

3. Define the public's role in the decision-making process at the
beginning of that process.
Generally, a government agency defines early in a project the
extent to which citizens' input will be a factor in its decision-
making process. To the extent that the public's role has been
defined within the agency, that role should be explained explicitly
to the interested public at the beginning of the process. Raising
the public's expectations about its role, and then not fulfilling
those expectations, can be far more damaging to the
government's credibility than no effort at all. In the public's view,
a "redefinition" or an "unexpected clarification" of its role in

3. See Roger E. Kasperson, "Six Propositions on Public Participation
and Their Relevance for Risk Communication." Risk Analysis, Vol. 6,
No. 3, 1986, for an excellent discussion of key elements to such a com-
munication strategy.
the latter stages of the process smacks of betrayal and, perhaps
worse, an attempt to "railroad through" a decision that was
reached long ago.

4. Establish credibility within the community from the very
beginning.
For members of the interested public to trust and believe in a
risk management process, it is essential that they view that process
as legitimate and credible. One of the best ways to ensure the
credibility of the process is to establish early—and place a high
priority on—maintaining the personal integrity and credibility of
the government officials charged with implementing and manag-
ing that process. It has been our experience, in countless situa-
tions, that the personal credibility of an agency spokesperson can
go a long way in demonstrating the credibility of the entire pro-
cess and, conversely, that the lack of personal credibility can
damage an otherwise legitimate and effective process. Moreover,
the credibility of key individuals can be an important asset in
maintaining public confidence in the process in the event of unex-
pected events such as a delay in sample results or a change in the
analytical method.

5. Establish and maintain communication with trusted members
of the community.
In addition to establishing the credibility of key government
personnel, it is essential for key agency staff to identify and
establish effective communication with a member or members of
the local community who are both trusted and a credible source of
information. Such individuals can provide a vital link to the com-
munity, can provide access to key opinion leaders that an outsider
might never develop, and can help ensure that the public's con-
cerns are brought to the attention of government decision-makers
early in the process, when it is still possible to ensure that they are
addressed. "Outside" experts, no matter how knowledgeable or
personable, can never enjoy the same type of credibility with a
community that a member of that community can.
We believe, based on our experience and the experience of
others, that implementation of an approach consisting of these
five principles would go a long way toward establishing a risk
communication strategy that could be effective in a wide range of
environmental decision-making efforts.
CONCLUSIONS
We do not believe, as this paper suggests, that successful risk
communication requires new or as-yet undiscovered techniques.
The principles we have incorporated in our proposed approach
are not new and, in some cases, are not dissimilar to ones being
utilized in other approaches, including the three discussed above.
What has been lacking, in our view, is the systematic application,
testing and evaluation of an integrated approach to com-
municating with the public about risk issues related to hazardous
waste facilities and sites.
We hope, in the near future, that such an approach can be ap-
plied to a wide range of environmental decision-making efforts
and thus can be tested, evaluated and refined. Such efforts na-
tionwide can enhance, we believe, both government efforts to ad-
dress the legitimate interests and concerns of the publics affected
by risk management decisions and the overall quality of the deci-
sions that are made.
RISK ASSESSMENT 283
-------
Tank Testing Method for Detecting Leaks in DOD's
Large Underground Storage Tanks

L. Peter Boice
Office of the Deputy Assistant Secretary of Defense
(Environment)
Washington, D.C.
H. Kendall Wilcox, Ph.D.
Douglas E. Fiscus, P.E.
Midwest Research Institute
Kansas City, Missouri
ABSTRACT
The contamination of groundwater by leaking underground
storage tanks (USTs) is a major environmental concern. Local,
state and federal governments have sought to control contamina-
tion from these tanks with a variety of legislation.
The resulting UST regulations concern many tank owners. One
problem is that the leak detection methods these regulations spe-
cify generally have not proven adequate or appropriate for test-
ing large tanks, those larger than 20,000 gal, such as those owned
by the Department of Defense (DOD).
This concern prompted the DOD to contract with the Midwest
Research Institute to develop an acceptable strategy for the early
detection of leaks in all of DOD's very large USTs. This paper
summarizes the selection procedure that screened 37 leak detec-
tion methods and presents results of field tests that were con-
ducted on three test methods in August 1986 at two DOD installa-
tions.

INTRODUCTION
Underground storage tanks (USTs) are found practically every-
where in the United States. Best estimates suggest there are in
excess of 1 .4 million — in highly industrialized areas and on farms,
serving individuals, industries and government agencies. A recent
national survey by the U.S. EPA indicates as many as 35% of
these tanks may be leaking. '
Federal legislation to limit contamination from these tanks,
stated in Subtitle I to RCRA, is still evolving. In the absence of
comprehensive federal regulations, state and local governments
have adopted a wide variety of strategies for controlling USTs.
The resulting UST regulations are of great interest to the De-
partment of Defense (DOD), which has approximately 35,000
USTs nationwide. More than 3500 of these tanks are larger than
20,000 gal. These large tanks are frequently five to 200 times big-
ger than a "normal" underground tank, and include such dif-
ferent tank types as 50,000-gal jet fuel tanks found on military
flight lines and 50,000-bbl diesel tanks located at DOD tank
farms.
The large size and often complex configurations of DOD's
large tanks raise several important issues. DOD's large tanks are
significantly different from the 8 to 12,000-gal motor fuel tanks
typically found at retail gas stations. Yet, because regulations
generally have been written to control these smaller, but far more
numerous commercial tanks, they often are not applicable to
larger USTs. A similar problem limits most leak detection
methods, which have been developed primarily with these smaller
tanks in mind. These methods usually cannot meet the normal
tank tightness standard of 0.05 gal/hr when testing larger tanks.
These problems prompted the DOD to contract with the Mid-
west Research Institute (MRI) to develop an overall strategy for
the early detection of leaks in all its large USTs. MRI was asked
to review all existing UST regulations, identify and evaluate com-
mercially available leak detection options, conduct field tests at
several DOD installations and recommend an overall leak detec-
tion strategy. This paper summarizes findings of the initial selec-
tion process and presents the results of field tests conducted on
three leak detection methods at two DOD installations.

Table 1
Summary of Talk Leak Detection Methods
Mf KM n.0 UA* OfTlCTXW
O.AUOI BT1CK
COMMMC1ALMAM
All PttCIVOM *AM* TU1
AIMLA* TAM T10JMTV TI»T»M
AACO MTt UMMMAOWMO TAM UA*
ciNTinc niTma
ITHtv TAM IIMTOT
in CMIK UAA MT1CTOI
NIATM Pmonn tAM AMI LMI TUT
UAB LOUHM IT*!
MOOMTf TAM mi
PACI 1 L1AK OfTlCTOA
PALDJ LIA* DITICTO.
IAN* AUOfTM
TWO TWM LAII. MnWIMMAITIB mrfu
»0«VTXUI«TTaC TAJIK HIT MffTMOQt
ACOVlnCAL MOMTTOWM trfTIM
UvaOLD-MtftAIVt MUUM DfTtCTO*
•WITH AND DAWMN MUUM TWT
TAC KAMO UAK DtTICTOA
UlTKAAOMC UAK MUCTOa
VACU TICT
VAMAN LtAK OCT1CTOA
KHLUlln AMOLIAA.I
MHNCIFU INO COMMIHQAl MAMIi
COtllCTIOM tUUM
DTI MITMOO
OAOVNOWATIII AND IOIV IAMPUM
IMTIKIOTIAl MOMITOMNO Of DOUIU WALL T
OUtHVATION WILLJ
MWOT1 INPKA1HD IIM1MQ
•UNf ACI OlOmniCAL MITHOOt
U TUUI ICOLLICT1ON lUMPt
VAPOH WILL!
AMtmwCUUCMND
SOA nMHMTUM OHUIMIWinC*
SELECTION PROCESS
In preliminary work, MRI identified 37 leak detection methods
284 UNDERQROUND TANKS
-------
as candidates for intensive field evaluations2 (Table 1). These
may be divided into three basic categories: inventory monitoring,
tank integrity testing and leak effects monitoring. Inventory
methods alone generally are considered too insensitive to detect
most leaks in large USTs. Effects of leaks, monitoring methods,
although extremely useful, are designed to detect product or its
vapors after a leak, spill or overfill has occurred. By contrast,
tank integrity testing subjects the entire system—tank and pip-
ing—to an accurate, closely controlled analysis to determine if a
measurable amount of product is lost during a specified time
period. Consequently, MRI focused on evaluating the 23 tank in-
tegrity testing methods.
Several constraints were imposed on this evaluation. A tank in-
tegrity testing method had to:

• Be commercially available
• Be able to complete testing in one working day
• Not require overfilling the tank or placing it under pressure or
vacuum
• Have adequate precision and accuracy
One of the most important factors is commercial availability.
MRI's evaluation was not development-oriented. Hence, theoret-
ically appealing methods such as the two-tube laser interfero-
meter were excluded from further evaluation.
Second, it is desirable to complete testing within 12 hr to min-
imize site disruptions. Many of DOD's large USTs are located
on flight lines or other mission-essential activities and cannot be
removed from service for long periods.
Third, the tanks must be tested without overfilling. Many DOD
tanks have relatively complex systems. Substantial site prepara-
tion is required to isolate a tank from vapor recovery manifolds
and to remove overfill protection devices, pumps and gauging de-
vices. The complexity of the piping associated with most large
tanks would also involve considerable preparation and disrup-
tion.
In addition, these tanks cannot be placed under pressure or
vacuum. This normally is not a constraint for retail motor fuel
tanks, but large tanks at DOD facih'ties are not designed for these
conditions. Complications include vapor recovery systems, mul-
tiple manway entrances and automatic level sensing systems that
are difficult to seal. Bulk fuel tanks are usually concrete and have
joints between the tank wall and top that were not designed for
pressure or vacuum. At tank farms, 30 or more tanks are mani-
folded to a common vapor recovery line. Isolating individual
tanks would be a major construction project and would require
disconnecting and raising the complete vapor recovery manifold.
Finally, precision and accuracy are extremely important in de-
tecting a leak. A leak generally is defined as any product loss
greater than 0.05 gal/hr. The measurement sensitivity required to
detect such a leak increases as tank size increases. MRI has cal-
culated that only methods with a greater than 0.001-in. liquid
level measurement sensitivity can be considered for tanks of
20,000 gal or larger. Otherwise, the minimum leak rate that could
be detected in a reasonable period of time would be so large that
it would be meaningless.
Precision and accuracy are defined primarily by sensitivity of
tank liquid level measurement and ability to compensate for
temperature effects. Tank liquid level sensitivity is a direct func-
tion of surface area. The smaller the surface area of the product,
the better the sensitivity of the method. To minimize the surface
area, many test methods rely on overfilling the tank into the fill
pipe. Testing a partially filled tank requires a sensitivity at least
100 times greater than normal for 50,000-gal tanks and at least
10,000 times greater than normal for 50,000-bbl tanks. Thus,
the requirement to avoid overfilling has direct implications on
which tank testing methods can achieve the desired level of sensi-
tivity.
In addition to level sensitivity, temperature compensation can
significantly affect a test's precision and accuracy. In DOD's
bulk tanks, i.e., greater than 20,000 bbl, temperature changes of
less than 0.0001 °F produce volume changes of 0.05 gal. Yet,
accurate temperature measurement in the field is limited to
0.001 °F. Thus, even minute temperature changes in these tanks
could indicate a leak when none exists.
Tank product level and temperature precision and accuracy cri-
teria only apply to volumetric methods. Nonvolumetric tests are
independent of temperature changes and do not rely on detect-
ing liquid level changes. They also are qualitative, meaning they
are capable of detecting a leak but they cannot specify precise
leak rates.
In evaluating nonvolumetric methods, however, it was impor-
tant that any method selected for further study have the potential
for relating its measurements to a leak rate. This requirement
eliminated methods such as those relying on helium detection.
These methods can detect leaks that may not be of any environ-
mental consequence, because helium will escape through pipe fit-
tings that do not allow the passage of liquids or heavier gases.

DESCRIPTION OF TEST METHODS
SELECTED
Of 23 tank tightness methods reviewed, three were selected for
on-site evaluation—two volumetric (Leak Detection Systems
Tank Auditor and the Associated Environmental Systems [AES]
Precision Tester) and one nonvolumetric (Tracer Research Cor-
poration's [TRC] Rapid Leak Detector).3

TRC Rapid Leak Detector
The TRC tracer leak detection method involves mixing an inert
volatile tracer chemical with the fuel. If product leaks out of the
tank, the tracer chemical evaporates and diffuses into the soil gas
spaces in the backfill. Soil gas is evacuated from the backfill and
analyzed for the tracer chemical. If tracer is detected, a leak is in-
dicated.
Fig. 1 illustrates the fundamental TRC process. A vapor tracer,
which can be used to distinguish between product leaks and vapor
leaks, also is depicted. The vapor tracer is not mixed with the pro-
duct but is released into the air above the product. Leaks above
the product level are indicated by detection of both product tracer
and vapor tracer.
LEGEND *
A; PRODUCT TRACER
B VAPOR TRACER
| PRODUCT LEAK
NOTt
PIPE CAN BE USED FOR
TEST OF ADJACENT TANKS
IF SPACING BETWEEN
TANKS IB CLOSE ENOUGH
Figure 1
TRC Test Apparatus

The tracer analysis method is very fast and sensitive. Measure-
ments typically require 2 to 5 min, and detection limits for
UNDERGROUND TANKS 285
-------
tracer in the soil gas are in the low parts per trillion range.
Product leaks as low as 0.001 gal/hr can be detected using a tracer
concentration of 100 mg/1, or approximately 1 qt of tracer to
5,000 gal of product. This small quantity of additive has no im-
pact on the properties of the stored chemical. The analytical
method is a proprietary gas chromatograph technique developed
by TRC. The tracer chemical used at the test sites was Freon,
which has been approved for use at up to 100 ppm in JP-4 jet fuel
by the U.S. Air Force.
Calibrating a leak rate is done by adding a known quantity of
tracer to the backfill material around a tank and correlating this
simulated leak rate to the concentration of Freon measured in the
soil gas. This method is not as direct as volumetric leak tests, but
the TRC method can estimate the leak rate if one is found.
The TRC method is independent of temperature and baro-
metric pressure changes and can test total tank volume. The soil
gas evacuation pipes can be installed while the tank is in service.
Therefore, tank downtime is typically less than 8 hr.

Tank Auditor
The Tank Auditor method is based on changes in the buoyancy
of a probe partially immersed in product and suspended from a
weight-sensing transducer. The probe usually is a hollow vessel
only slightly heavier than the tank product. Its weight is adjusted
so that only a small force is required to keep it from sinking.
Buoyancy changes result from any factor that produces a change
in product volume. The test apparatus is shown in Fig. 2.
The unit is calibrated at the beginning of each test by adding a
known volume of product to the tank. The balance is connected
to a weight-sensing transducer so that changes in weight produce
an electronic signal which is recorded on a strip chart and by com-
puter. This method is very sensitive to small changes in force
since the transducer is measuring buoyancy as a result of product
level changes. Level changes as small as 0.005 in. can be detected,
allowing the method to be used to test partially filled tanks.

AES Precision Tester
The AES system shown in Fig. 3 is based on head pressure
changes that occur when tank level changes due to a leak. A sen-
sitive pressure transducer is mounted on a rigid support at the top
of the tank. Data are collected by a proprietary computerized
system which is capable of detecting changes of 1/12000 of the
full scale of the pressure sensor. Temperature also is monitored
at three levels using a temperature sensing system mounted on the
same support system. The system is calibrated by adding or re-
moving known volumes of product.
RADIO PMQUIMCT
r— WtlOMTUNMM
\ TUfttOUCU
Figure 3
AES Test Apparatus
SITE DESCRIPTION
A series of tests was conducted at two DOD installations to
evaluate the three tank testing methods. All three test methods
were evaluated at March Air Force Base (MAFB) near Riverside,
California, on two 50,000-gal tanks containing jet aircraft fuel
(JP-4). Additional testing was conducted using only the two volu-
metric methods at the San Pedro Defense Fuel Supply Point
(DFSP) located in San Pedro, California, on one 50,000-bbl
tank containing diesel fuel. Tests were performed on Aug. 11 and
12 at MAFB and Aug. 14 and IS, 1986, at San Pedro DFSP.
Details of each site are shown in Table 2.

Tibkl
Site Description
NO O* TAKml M f
WArt* TAItt. OttTAMCI
•now IUHACI
COV11IUAn«Ai
•*n ouuMm n n Mir
M»AMM NTOUUfr rrtTIM
tMMTCMAMD MVOAAMt
1 n A
KMJOIIAVtl OWVMMim rvfn
> MrtAM
*Wf SUCTOM
rwfluMI mow Two
•L«»MI TANK*
I MHJf AWAV
nvOVfMTITOAl
Figure 2
Tank Auditor Test Apparatus
At MAFB tank 16 was used for the AES and Tank Auditor
tests, and tanks 3 and 7 were used for the TRC test (Fig. 4). Dur-
ing the TRC test, tank 3 was used to mix the tracer chemical and
the fuel. The actual leak test was conducted on tank 7. Fig. 5
shows details of tank 16, which is fundamentally identical to
tanks 3 and 7.
At San Pedro DFSP, all bulk fuel tanks are essentially similar.
Tank 18 (Figs. 6 and 7) was tested. Manways and other openings
were sealed to the extent possible to prevent air currents that
286 UNDERGROUND TANKS
-------
might cause surface waves.

TEST CONDITIONS
To accurately evaluate the performances of the three test meth-
ods, test conditions had to be either controlled or closely moni-
tored. MRI equipment was used to perform:

• Calibration tests
• Simulated leak tests
• Measurement of environmental data (product temperature,
barometric pressure and ambient temperature)
Figure 6
Location of Tank 18 at San Pedro DFSP
FANCHO HTDMNT I
Figure 4
Overall Tank System Layout at March AFB

•"f | O '1
1 1

.
BUD

™ O ° ooo (^)

PLAN VIEW
lUVATIOMWnll
Figure 5
Tank Number 16 at March AFB
Calibration TEsts
Calibration was achieved by adding or removing known vol-
umes of product each day, using a portable transfer pump with
a volumetric meter. The Tank Auditor was calibrated only when
product was added because its computer program could only
accommodate added volumes for calibration. The AES method
used both removal and addition of product for calibration.

Leak Simulation Tests
Leaks were simulated using a variable speed peristaltic pump
apparatus. Rotameter readings were taken periodically to ensure
that the flow rate had not changed from the original setting. Dur-
ing all tests, pump operations were very stable.

MRI
TEMPERATURE
AND AES
TEST LOCATION

MRI
TEMPERATURE.
CALIBRATION,
SIMULATED
LEAK LOCATION

TANK AUDITOR
TEST LOCATION
Sl VENTS (2)
SOIL BJ

SFT
/ • 8 IN DIA GAUGE PORT *
-4- CONCRETE II 20 FT
1
(VARIES
4 FT TO 6 FT)
— 6 IN DIA DRAIN LINE

ELEVATION VIEW

Figure 7
Tank 18 at San Pedro DFSP
Environmental Data
Product level in an UST changes with temperature due to ther-
mal expansion or contraction of the tank and its product. Pro-
UNDERGROUND TANKS 287
-------
duct temperature can be modified by heat transfer from the sur-
rounding backfill material. If the tank product is in equilibrium
with the surrounding backfill, the major source of temperature
change will be from changes in surface temperature.
In theory, the product level in an LIST also can be affected by
barometric pressure. Although trapped vapor products can cause
changes in volume with changes in barometric pressure, vapor
products were not expected to be a major problem for these tests
because both volumetric methods were tested in a partially filled
condition; however, to record any ambient pressure and temper-
ature changes that might affect tank testing results during test-
ing, both barometric pressure and surface and subsurface temper-
atures were recorded.

TEST RESULTS
In this section, test results and a discussion of each method's
precision and accuracy are presented. Representatives of the
equipment suppliers for each method conducted their own testing
at the two DOD facilities. The major emphasis for both volu-
metric methods was to collect data on the behavior of the pro-
duct level in the tank so that temperature corrections could be
applied. Since neither method does this routinely, MRI collected
temperature data independently and applied appropriate temper-
ature correction factors to the level data.

TRC Rapid Leak Detector
Six probes were driven into the backfill to a depth of 6.5 ft. A
seventh probe penetrated to the bottom of the backfill to a depth
of 15 ft.
After the probes were installed, the air flow around the tank
was checked to verify that it was continuous at all locations on the
tank surface. The deep probe was used to check the bottom of the
backfill for the presence of water or product, because any por-
tion of the tank that is immersed in saturated backfill is not tested
by this method.
Next, background analyses were made on air samples from
each shallow air evacuation probe to verify that there was no
tracer present in the backfill prior to the start of the test.
Due to the large size of the tank being tested (50,000 gal) a
nearby tank was used to mix the tracer and fuel. Ten quarts of
Freon 11482 were first added to the mixing tank, which was full
at the time. The tank being tested was emptied to the lower limit
of the pumping system. Then, the JP-4 and tracer were mixed by
transfer. Only the product tracer was used in this test. A vapor
tracer would have been used only if a leak had been detected us-
ing the first tracer, to determine if the leak was above or below
the product level.
Some water or product was expected at the bottom of the tank
pit because all Air Force 50,000-gal JP-4 tanks are placed on con-
crete slabs in concrete cradles prior to backfilling. The concrete
floor acts as an impermeable layer on which water and product
accumulate. Approximately 5 in. of essentially clean water were
detected in the bottom of the tank pit. A very slight product film,
which typically results from a small amount of product spillage,
was discernible on the water surface. Since a leaking tank invari-
ably produces several inches to several feet of free product on top
of the water in the area immediately around the tank, the water
sample suggested the tank was not leaking.
The tank test with tracer confirmed that the tank was not leak-
ing. No tracer was detected in the backfill when soil gas was evac-
uated and analyzed.
After the tank leak test was completed, two sensitivity tests
were conducted by simulating leaks. In the first, performed by
TRC, JP-4 was dripped into the tank backfill through one probe
at a rate of 0.0016 gal/hr. The five remaining shallow probes were
evacuated and tested for presence of the tracer. The tracer peak
observed in the chromatogram was at least ten times greater than
the minimum detectable peak.
A second simulated leak was conducted by MRI. Tracer-lab-
eled fuel was released into the deep probe. The remaining probes
were evacuated until tracer was detected. The leak rate measured
and reported by TRC was 0.0004 gal/hr. The actual rate of the
simulated leak was 0.00032 gal/hr.
Test results indicated the tank was not leaking. The absence of
floating product on the water sample from the bottom of the tank
pit was positive evidence that there was no significant continual
leakage from the tanks in the area where the water sample was
collected. The simulated leak test verified there was no leakage
greater than 0.00032 gal/hr on portions of the tank above the
water saturated zone. It also demonstrated that the TRC method
could detect leaks less than 0.001 gal/hr, which is at least one
order of magnitude better than the measurement sensitivity re-
quired by UST regulations.

Temperature Corrections
Temperature is the most important single external factor that
must be accounted for in a volumetric test. Temperature data in-
dependent of the volumetric methods were collected on both test
tanks using six temperature sensors.
Three characteristics of the data arc noteworthy. First, the
periodicity of the variations is too large and regular to repre-
sent random fluctuations which might be due to unstable temper-
ature conditions.
Second, the variability of the upper sensor generally is much
greater than that of the remaining sensors. This variation might
have been caused by the addition or removal or product from
this region of the tank during calibration and leak simulation pro-
cedures.
Third, the temperature stratification observed for both tanks
was considerable. Results differ in that most of the temperature
differences for the San Pedro tank are in the upper third of the
tank. This difference probably is a result of heating the product's
surface and provides a strong indication that temperature
changes may not be uniform throughout the tank. This finding
suggests that a number of sensors should be used to test tanks of
this size.
The first two factors discussed above (periodicity of variations
and variability of the upper sensor) increase the uncertainty of
the temperature measurements.

Tank Auditor Results
Testing the Tank Auditor method produced mixed results.
Tank Auditor uses a buoyancy float to measure changes in liquid
level. The buoyancy measurements had good sensitivity and re-
sponded well to calibration volumes and simulated leaks; how-
ever, the high level of sensitivity proved to be a mixed blessing.
For example, interactive effects with tanks in use during testing
were observed. This interaction may have caused the phenom-
enon observed at MAFB during leak simulation testing, where
product level in the tank was observed to be increasing while pro-
duct temperature was decreasing.
Similarly, when compensations were made to the raw data to
account for product expansion or contraction caused by tempera-
ture changes, results were indeterminate. Uncertainties in buoy-
ancy probe and product temperature measurements resulted in
such a large data range that no useful conclusions could be
drawn.

AES Precision Test
The AES method produced the least satisfying test results. The
288 UNDERGROUND TANKS
-------
AES method uses pressure transducers to measure the mass of
liquid above the transducer. Level corrections for temperature
are made from a multipoint temperature probe. Field tests, how-
ever, proved that the AES pressure sensors were not sensitive
enough to be of practical value. Even large calibration volumes
of 1 to 5 gal in a 50,000-gal tank and 30 to 220 gal in a 50,000-
bbl tank could not be accurately detected. Simulated leaks rang-
ing from 0.31 to 3.50 gph were not reliably detected by this
method's current capabilities.

EVALUATION OF TEST RESULTS
The evaluation of the TRC method at March AFB demon-
strates that this method can reliably detect leaks above the soil
water saturated zone with sensitivity better than required by UST
regulations. These results are relatively clear because they are not
affected by tank size, thermal expansion or other changes within
the tank. The independence from temperature effects is particu-
larly important for large tanks. The presence or absence of tracer
outside the tank is the only variable of concern. Also, TRC was
the only method tested not impaired or compromised by a tank's
being only partially filled.
The TRC method is probably the safest and least disruptive of
the three methods evaluated. The tank and all components of the
system remain sealed during testing. Thus, the release of fuel
and vapors at ground level is reduced. Unlike the volumetric
methods, there is no need to isolate the tank prior to testing to
allow for thermal stabilization. Also, the TRC testing equipment
is not affected by ground vibrations, wind, noise, changing air
temperatures and the myriad of other factors that cause problems
for volumetric testing equipment.
The TRC method can be most economically applied to clusters
of interconnected tanks such as the March JP-4 tanks. A single
batch of tracer-labeled fuel can be transferred from tank to tank,
reducing the cost of the tracer chemical and the time and labor re-
quired to introduce and mix tracer with the product in each con-
secutive tank. An added advantage is that leaking pipes, valves
and fittings can be detected as the tracer-labeled product is trans-
ferred from tank to tank for testing. But, if a leaking tank is de-
tected or if other leaks are found, a different tracer must be used
for testing any remaining tanks. The release of the tracer pre-
cludes its further use at that site. Although the availability of
several different tracers mitigates this limitation, tune and cost
savings will be realized only if the system is tight.
The TRC method does have four significant disadvantages.
Most important is that it cannot be reliably used where the water
table is above the tank bottom. While Freon can be detected in
water samples taken from around the tank, TRC sensitivity de-
pends upon using gas samples. But, if no product is found on the
water table, there is no leak. This would negate the necessity for a
tank tightness test except to meet regulatory requirements.
Second, the TRC method may require angled or horizontal
probe installation for bulk storage tanks. The recommended
probe spacing is 20 ft between soil gas probes. Yet, the tank diam-
eter for 50,000-bbl and larger bulk tanks often exceeds 100 ft,
which could make it difficult to detect leaks occurring near the
center of the tank bottom. For this size tank, probes inserted at
an angle or horizontally from a probe pit would be required to
reach under the tank. This is not technically difficult to accom-
plish but would increase testing costs.
Third, the cost of Freon increases as tank size increases. In
practical terms, Freon is added at a rate of 10 qt (33 Ib) per
50,000 gal. To test a 50,000-gal tank required only $90 of Freon;
however, more than $3,000 of Freon would be needed to test a
50,000-bbl tank. As already noted, this cost can be reduced to the
extent that product can be transferred between tanks without en-
countering leaks.
TRC is actively working to increase their Freon detection
capability so that reduced Freon concentrations can be used,
thereby reducing the cost of Freon for large tanks. Some early in-
dications are that levels as low as 1 qt of Freon per 50,000 gal
may be possible. If this proves feasible, the TRC method would
only require approximately $300 of Freon to test a 50,000-bbl
tank.
Finally, the TRC calibration technique is not precise; however,
its high degree of sensitivity, being at least one order of magnitude
better than the regulatory requirement of 0.05 gal/hr, mitigates
this apparent flaw.
Conclusions regarding volumetric methods are less than en-
couraging for partially filled tanks greater than 20,000 gal. There
is considerable doubt as to the applicability of the classical tem-
perature-compensated level measurement approach to detecting
leaks in these tanks. Problem areas include:

• Requirement for very precise temperature measurements at
multiple locations in the tank. Temperature compensation pro-
cedures for these methods and most others need to be at least
an order of magnitude better than was displayed during this
evaluation.
• Effects of routine operations such as the receipt and dispen-
sing of fuel from nearby tanks
• Effects of vehicular traffic
• Vibrations caused by wind
• Level changes that are not readily explainable from the known
facts about the tank system
• Possible FM interference from nearby radio communications
• Need for better calibration procedures that minimize disrup-
tion to the tank
• Increased level sensitivity without attendant increases in noise
levels

Some of these problems could be eliminated, but major
changes in the hardware, operating procedures and data analysis
techniques still would be required.
CONCLUSIONS
Both volumetric methods have positive features which make
them attractive for potential future use on bulk tanks if needed
improvements in their methods are made. The Tank Auditor
equipment has a very sensitive level measurement system. If ade-
quate control of temperature effects can be gained, most remain-
ing problems can be minimized by minor testing changes and by
extending the test times.
The AES pressure sensing approach also has some strong fea-
tures. If a sensor capable of detecting level changes of 0.0001 in.
or smaller could be developed, this method would offer some ex-
tremely desirable characteristics for testing vertical tanks. For ex-
ample, a pressure sensor located at the bottom of the tank would
be self-compensating for temperature effects and would be insen-
sitive to surface ripples. Many of these limitations may be cor-
rectable in time. But, for now, neither volumetric method tested
can accurately detect leaks in DOD's large USTs.
The TRC leak detection method appears to be the best option
currently available to test DOD's large USTs. It is commercially
available, and it displayed a high degree of sensitivity during test-
ing at March AFB. It also is ideally suited to testing under par-
tially filled conditions, as its sensitivity is not related to the sur-
face area of the product. There is no reason to believe that the
method as it is now used could not be applied to the 50,000-bbl
tanks with only minor modifications.
UNDERGROUND TANKS 289
-------
REFERENCES Leak* in DOD's Underground Bulk Storage Tank*: Preliminary
1. Westai. Inc.. Midwest Research institute. Battelle Columbus Division. Strategy," Kansa. City. MO, June 1986, MR! Project Report
and Washington Consulting Group, "Underground Motor Fuel Stor- No- 8540-R(6).
age Tanks: A National Survey," U.S. EPA. Washington. DC, May 3 Mjdwejt Rwearch ^.^ ..^ Evaluation of Tank ^ ^^
I9S6- lion Methods: Draft Report," Kansas City, MO, Sept. 1986, MRI
2. Midwest Research Institute, "Recommended Strategy for Detection of Project Report No. 8540-R(97).
290 UNDERGROUND TANKS
-------
Safety Procedures for Testing and/or Removing
Underground Storage Tanks
Fred Halvorsen, Ph.D., P.E.
O.H. Materials Corp.
Findlay, Ohio
INTRODUCTION
During the past few years, the environmental hazards presented
by leaking underground storage tanks have been given increasing
attention by the regulatory community. Although final federal
regulations have not been issued, certain elements of a federal
regulatory program have been developed by the U.S. EPA in-
cluding tank notification (registration) and an interim standard
for underground tank installation. The proposed federal standard
will be issued shortly and must include requirements for tank
tightness testing, tank leakage monitoring systems, tank design re-
quirements (including possible secondary containment re-
quirements) and tank installation requirements. Many states have
also issued underground storage tank regulations—generally
along the lines of proposed federal regulations.
Historically, the basic rationale for placing storage tanks
underground was one of safety. Simply stated, tanks containing
flammables were made safe from fire, explosion and leakage by
burying them in holes in the ground. A secondary consideration
was the more effective use of space.
The only standards for underground tanks for many years were
those safety standards found in NFPA 30, Flammable and Com-
bustible Liquids Code. These standards were concerned with: en-
suring tanks were properly secured in the ground to prevent the
tanks from floating as a result of groundwater; control of ignition
sources; tank testing before being placed in service; tightness
testing; fire protection and identification; overfill protection; and
inventory records.
Unfortunately, for many years little consideration was given to
underground tank leakage. The lost product normally did not
constitute a safety hazard, and a few gallons of lost inventory did
not constitute an unacceptable economic loss to the tank
operator.
Times have changed, however, and underground storage tank
leakage now is considered a major environmental problem by the
public, the environmental community and insurance companies
responsible for the ultimate cleanup costs. Companies which test,
remove and install underground tanks have become a major in-
dustry.

A Problem
From the safety aspect, underground storage tank testing and
underground storage tank removal have not been given a great
deal o" attention, and from a health and safety aspect, con-
siderable hazards may exist during both tank testing and tank
removal. The purpose of this paper is to review the hazards
presented by these operations and outline safety guidelines which
should be applied in order to protect people, property and the en-
vironment. These guidelines were developed over the past 18 mo
and were used to ensure the safe testing of over 2,000 under-
ground storage tanks and safe removal of over 500 tanks.

BASIC HAZARD
The hazards presented by work around underground storage
tanks are primarily related to the product contained within the
tank. The vast majority of tanks contain combustible or flam-
mable petroleum fuels, but a significant portion of underground
tanks also contain products which may be toxic, corrosive or have
some other potential hazard.
While the products are safely contained within the tanks, there
is very little potential hazard. However, during both tank testing
and tank removal, the tank piping, tank barrier or boundary is
deliberately violated and product is placed in intimate contact
with workers in the vicinity of the tank. Additionally, if the tank
has leaked, soil adjacent to the tank may contain appreciable
quantities of the product, again potentially hazarding un-
protected workers in the area.

TANK TESTING HAZARDS
As just stated, the basic safety problems attendant with under-
ground storage tank testing usually are related to the product con-
tained within the tank. In the vast majority of cases, the product
is a petroleum fuel—most normally gasoline.
Most personnel connected with testing are "familiar" with
gasoline—normally to the extent of being cavalier and unap-
preciative of the hazards. Gasoline from a health standpoint is a
relatively toxic material from all modes of exposure. From a flam-
mability standpoint, gasoline is a volatile, readily ignited material
with a low flashpoint. If a less volatile petroleum product is pre-
sent, the hazards are appreciably reduced. If a chemical is pre-
sent, the toxic and/or flammability hazards may be altered.
The problem with tank testing is that the tank contents, which
are quite safe in the tank, must in some fashion be accessible to
persons testing the tank, who then may be exposed to the
chemical's attendant hazards. We need not dwell upon the actual
tank testing methods, but the most common tank tests are depen-
dent upon creating a head of product above the tank and measur-
ing the tank loss due to product leakage. In this testing process,
the product in the liquid and vapor phase may be present in toxic
and flammable concentrations in the area of the tank testing
operations.
Typically, underground storage tank testing is conducted in the
apron of a commercial gasoline station. The tank testing schedule
consists of first topping off the tank with fuel. While we advocate
that gasoline station operations be halted during testing, most sta-
tion owners attempt to remain open if at all possible during the
test procedure. This poses obvious difficulties with testing.
UNDERGROUND TANKS 291
-------
Basic Tank Testing Steps
Tank testing generally should follow these steps:

• Site assessment
• System preparation
• Tank testing
• Tank excavation

Site Assessment
A typical tank testing project may last less than one day, and
often the testing crew members are the first company represen-
tatives to visit the site. Because of these and other reasons, it is
not always practicable to develop a safety plan (which we ad-
vocate for all work) prior to the on-site arrival of the crew.
Development of a consensus safety plan should be the first
priority of the crew after arrival on the site. If the work site is an
industrial plan, the plant's safety officer should be contacted im-
mediately for a safety briefing (this is often a client requirement
before starting work). If the site is a service station or other com-
mercial business, the location of the nearest telephone should be
determined along with appropriate emergency telephone
numbers.
While doing the above, the crew also should make itself
familiar with general site conditions: ingress and egress, traffic
patterns, buildings, waterways, safety equipment, overhead
canopies, site gradient, hazardous material storage areas and fire-
fighting equipment. Any specific safety plan, such as an evacua-
tion route, will have to be conceived in the context of each site's
specific characteristics.
A detailed examination of the work area is the next step.
Overhead obstacles, tank orientation and access are only a few
considerations. Each tank should be identified by number, loca-
tion and product. Specific hazards such as power lines, noise and
traffic should be noted.
The next step is to develop work zones. At least two zones are
needed: the hot zone (or exclusion zone) and the support zone.
The hot zone is that area where hazardous or potentially hazar-
dous conditions exist; for example, where gasoline vapors would
present a fire hazard. The support zone is the area where ancillary
equipment is stored or where there is no need to enforce safety
rules related to eating or smoking. In certain instances, it is
necessary to designate a contamination reduction zone where per-
sonnel can wash up or change out of contaminated clothing.
The next step in developing the safety plan should be to
designate the parking area for the support vehicles, test equip-
ment and support equipment. Fire extinguishers, a charged
garden hose, eye wash stations and first aid kits should be placed
in a prominent, accessible area.

System Preparation
The next step is to prepare the tank for testing. In general, the
tank must be isolated from all other tanks for the test to be mean-
ingful. This step may require excavation to uncover piping
systems, rearrangement of piping and exposure of the tank man-
way in order to vent the ubiquitous air pocket which is present.
The pump may have to be removed from the tank. It is necessary
throughout this step to minimize exposure of personnel to the
product in the tank.

Testing Gasoline Tanks
The flammable property of gasoline presents the greatest
hazard to personnel when considering products normally tested.
Gasoline is an extremely flammable material, with high volatility
and low flashpoint. Gasoline vapors are heavier than air, and on
hot windless days, gasoline fumes may create a "vapor trail"
which may extend for 10-20 ft along the ground and act as a fuse
from an ignition source back to the gasoline. Personnel must
carefully consider the potential flammability hazard when testing
tanks which have contained or contain gasoline.
General safety considerations when working with gasoline and
other volatile, low flash, petroleum products include the follow-
ing admonitions:

• Clean up all liquid spills immediately with absorbent pads.
• Always keep all tank and piping openings closed unless there is
a need to have the opening open—use expandable rubber plugs,
if necessary, for closures.
• Always vent the tank when filling through the installed high
vent system; avoid allowing gasoline fumes to be vented into
sumps or at ground level.
• Never work in a flammable atmosphere—use the combustible
gas detector and "foam" or ventilate the sump if vapors in
excess of 20% of the LEL are found.
• Always be concerned about potential vapor ignition; wear
long-sleeved cotton clothing, safety shoes or boots, safety
glasses; have fire extinguishers readily available; have a charged
garden hose laid out.
• Isolate the work area; use hazard tape or orange safety cones to
define a work area at least 25 ft in all directions from tank
openings; post "No Smoking/Hazard" signs.
• De-energize and check ail electrical equipment including verifi-
cation that the equipment is secured; then tag the breaker
switches to prevent accidental closure of the breaker.
• Be aware of wind and weather; on cold, windy days, the vapor
hazard is much less than on hot, calm days; be aware that vapor
trails may move downwind or downhill.

Tank Testing
Specific safety rules include:

• Site setup
-The area should be cordoned off with yellow hazard tape
and/or orange hazard cones and posted with "No Smoking"
signs so that all potential ignition sources are kept at least
25 ft away from the test area.
-A listing of local emergency telephone numbers should be
carried by each employee and should be posted by the nearest
phone; make sure that the telephone is operational.
-Develop a work and safety approach at the beginning of each
project and review it with all on-site personnel.
-Daily safety meetings are mandatory.
-When working in high traffic areas, traffic cones or other
markers should be prominently placed to warn approaching
vehicles; all personnel should wear high-visibility clothing.
-Personnel should be familiar with the physical and chemical
properties of the materials in the tanks; a Material Safety
Data Sheet is required.
• Fire Protection
-A minimum of two, 20-lb ABC dry chemical fire extinguishers
must be easily accessible at all times during tank testing; it is
suggested that one be at or near the test area and one near the
fueling island.
-A charged garden hose shall be readily available at each site
as an emergency water supply; the system should be set on
fine spray.
-Fire blankets should be readily available at all job sites, es-
pecially when water is not available.
-Only nonsparking brass tools should be used when working
on fill pipes, and every effort must be made to eliminate vapors
rising out of a fill pipe.
-Every effort must be made to prevent product spillage; ab-
sorbents must be readily available in case of a spill. All spills
292 UNDERGROUND TANKS
-------
will be cleaned up immediately.
• Testing
-Any venting of tanks (such as occurs when filling or topping
off tanks) must be through the installed venting system on the
tank; do not allow tanks to vent at ground level or into sumps;
use rubber plugs if necessary.
-All tank openings should be kept closed except as needed for
access or testing.
-All pumps, dispensers and other electrical equipment at the
site shall be disconnected, tested to ensure that the electricity
is disconnected and the breaker switch tagged as being discon-
nected.
-Before working in any tank sump, a combustible gas monitor
shall be used to ensure that flammable vapors in the sump do
not exceed 20% of the LEL; if vapors exceed 20% of the LEL,
the sump should be "foamed" with an emulsifier and water
and/or ventilated using an explosion-proof mechanical ex-
hauster; no work should be performed until the flammable
vapors are less than 20% of the LEL.
• Electrical Safety
-Only heavy-duty, three-wire grounded extension cords should
be used; one should never have more than two pumps on one
circuit breaker.
i-Any time any electrical connection is made, male and female
plug ends should be elevated to eliminate sparking hazards
and prevent grounding.
-The extension cord on the pump must always be extended
fully to get the connection as far away from the test site and
fill pipe as possible; the minimum distance should be 25 ft.
-If it becomes necessary to remove a dispenser, no personnel
should attempt this due to the possibility of electrical spark;
the client's pump mechanic should be contacted.
-While doing any transferring, tanks and pumps must be
grounded.
-All lighting equipment is to have explosion-proof switches
(flashlights, satellites).
• Protective Clothing
-When testing tanks previously containing any chemicals or
petroleum product, safety goggles and hardhats with face-
shields must be worn while doing any plumbing or bleeding
of lines or flanges (remaining product may be floating on top
of water and will be the first thing to appear after air is bled
off).
-Hardhats, safety glasses, steel-toed shoes or boots and gloves
suitable for the product shall be used at all times when work-
ing.
-Long-sleeved cotton clothing and cotton trousers or cotton
coveralls shall be worn; short pants, short-sleeved shirts and
tennis shoes are specifically prohibited.
• General Safety
-No person may enter any confined space (tank, manway,
tank vault, etc.) without proper safety precautions (air moni-
toring, etc.) and a Confined Space Entry Permit.
-A "spotter" must be used when backing a truck into position.

UNDERGROUND STORAGE TANK
REMOVAL SAFETY
The hazards presented by removing an underground storage
tank again are related mainly to the product, although there are
obvious hazards created by excavation, handling large tanks and
cutting tanks for disposal. Basic safety considerations include
limiting exposure of tank contents to personnel, control of igni-
tion sources, use of proper excavation techniques and control of
flammable atmospheres in the tank.

TANK REMOVAL SAFETY PROCEDURES
Basic Tank Removal Steps
Tank removal normally should consist of the following steps:
• Site assessment
• Excavation
• Tank Removal
• Safety certification before tank rendering
• Tank rendering

Site Assessment
Site assessment is extremely important because this assessment
will determine the work which will be undertaken. In all cases, an
on-site review is suggested before work begins. If a site assessment
is not made before arrival, the site should be surveyed as the first
order of business upon arrival.
• Visually inspect the site to ensure that the work can be safely
done; special attention must be given to safe work surfaces
for equipment, the presence of overhead lines which may hin-
der equipment operation and local traffic which may be af-
fected.
• Call the local utility companies to locate telephone, power,
water and sewer lines which may be in the way of excavation;
ensure they are well marked before excavation.
• Locate the tank, together with piping, vents and manways.
• Sound the tank to verify that the tank contains the product
which was indicated; note liquid levels; check for the presence
of water and other contaminants.
• Sample the tank vapor space with an O2/LEL meter.
The site assessment is extremely important since the assessment
will establish the operational procedures to be followed.
The visual inspection will ensure that the work can be done
safely in the area from the standpoint of working surfaces for
heavy equipment and that obstructions and traffic will not hinder
the operation. Calling the utility companies and having lines
marked will ensure that you will not excavate in locations where
utilities may be present. Locating the tank and associated piping
will ensure that you are digging in the proper location. Sounding
the tank will determine that you are aware of tank contents and
quantity so you can plan cleaning and disposal requirements.
Sampling the tank for the presence of flammables will indicate
whether inerting or foaming may be necessary before commenc-
ing excavation.

Tank Excavation
Before excavation commences, the site supervisor should
review the site layout and markers to ensure that all underground
lines have been marked and that the excavation can be done safe-
ly.
The following general guidelines will be followed:
• Establish the boundaries of the exclusion zone so that unpro-
tected personnel will not accidentally come in contact with any
possible liquid splashes or vapors arising from the excavation.
• Work Surfaces: Ensure that all walking/working surfaces and
areas are in a safe condition. Can a firm footing for equipment
be established on the overburden? If not, the areas may need
to be stabilized. Is it possible for excavation or hoisting equip-
ment to contact overhead power lines? If so, these will need to
be de-energized prior to beginning operations. Note that for
tracked vehicles, the proper alignment with an excavation is
to have the tracks perpendicular to the excavation.
• Monitor: While the overburden is being removed, is free
UNDERGROUND TANKS 293
-------
product or saturated soil likely to be contacted? If so, has a
proper storage area for this spoil material been designated?
Does this area need to be lined and diked? If free product or
saturated soils are found, both toxic and flammable readings
shall be taken at the work site and perimeter.
• Slope or Shore: The excavation wall may have to be sloped or
shored to ensure that the walls do not collapse. Remember
that an excavation above shoulder level is considered a con-
fined space. There are definite OSHA standards for both ex-
cavations and confined space entry which must be addressed.
• Protective Equipment: If personnel enter the excavation, they
should use the proper protective equipment.
• Barricades: If open excavations are left unattended, proper
barricades and warning signals shall be set up to actively warn
all personnel of the open pit hazard. In certain areas, addi-
tional security measures may need to be instituted in the form
of fences or security guards.
The purpose of these precautions is to ensure that nothing
unexpected is contacted during digging, that heavy equipment
does not topple into the excavation due to wall failure or contact
energized utilities and that workers are afforded protection if they
enter the excavation.

Tank Removal
Before the tank is removed from the excavation, the following
precautions shall be taken:
• Remove Free Product: All products will be removed by pump-
ing (if possible).
• Inert or "Foam" Tank: If the tank atmosphere is flammable,
the tank may have to be inerted prior to removal. This process
can be done by introducing nitrogen or dry ice (CO^ until the
oxygen content is 8% of less, or the tank can be "foamed"
with firefighting or vapor suppression foam to suppress vapori-
zation, along with ventilation to below 10 LEL. Depending
on the situation and product, one method may be preferred
to the other; a description of each method is given under the
Safety Certification.
• Remove lines if possible; all lines will be removed by discon-
necting joints rather than cutting or burning; no hot work
will be performed without a hot work permit issued after LEL
testing.
These precautions are to ensure that a minimum amount of
product is present in the tank prior to removal, that the tank
vapor phase in inert or gas-free to prevent vapor ignition and that
the tank is moved immediately to the decontamination area for
cleaning.

Tank Decontamination
Decontamination, if required, is performed to remove residue
from the tank, so it can be disposed of as clean material. Before
tank decontamination, the following procedures and equipment
should be in place:

• The decontamination area will be marked as an exclusion zone;
proper personnel protective equipment, medical emergency
equipment, splash shower and eye wash should be available.
• Before opening the tank to permit entry for water blasting,
foaming or other cleaning methods are used, the tank will be
checked again for the presence of flammables and appropriate
actions will be taken to reduce flammable levels.
• A method for rinse water containment and proper disposal
procedures will be established.

Safety Certification Before Cutting Tanks
This procedure is used when tanks are to be cut up for disposal.
The preferred method of cutting (rendering) tanks is to use the
power shears attached to a boom vehicle. The less preferred
method is to use a cutting torch. A power metal chisel also may be
used. The primary hazard when cutting tanks using either shears,
cutting torch or chisel is the possibility of igniting flammable
vapors in the tank. When using the shears, the possibility of
catapulting metal pieces must be considered. Whenever tanks are
cut, all unnecessary personnal should be removed from the area.
The possibility of explosion of flammable vapors always should
be considered.
• Sample the tank atmosphere for flammables and oxygen; be-
fore a tank may be cut with either shears, cutting torch or
chisel, the tank atmosphere must be less than 10% of the LEL
or less than 8% O2. If the tank atmosphere shows flammables
in excess of 10% of the LEL, then some action must be taken
to reduce the flammable vapor concentration or reduce the
oxygen concentration to less than 8% O2.
• If the tank atmosphere is greater than 10% of the LEL, one of
these actions can be taken to make the tank safe for cutting:
-Ventilate the Tank: This procedure only will work with fairly
clean tanks with clean products; the tank will regas rapidly if
not a clean product. Readings should be taken at the location
of the tank exhaust to check for flammables; note that ex-
hausted vapors may be flammable.
-Clean and Ventilate the Tank. Use a cleaning method such as
butterworthing with hot water, pump out liquids and then
ventilate as above.
-Inert the Tank. Nitrogen from a liquid nitrogen tank or car-
bon dioxide from dry ice can be used to inert the atmosphere
in the tank to reduce the oxygen concentration below the level
necessary for combustion. Note that flammable vapors wiD
still be present and once the tank is cut or opened the inert gas
can be lost. The atmosphere must be diluted to less than 8%
O2 by volume (11% O2 is the theoretical lower limit with 8% a
safety feature) to be completely safe for normal petroleum
products. Oj/LEL meter must be used to verify the O2 con-
centration.
The quantity of inert gas which must be used depends on
how the gas is presented to the tank. In theory, the minimum
amount would be that necessary to lower the O2 concentra-
tion from 20.9% by volume to 8% by volume—about 1.25
tank volumes would be required. This assumes perfect mixing,
however. In practice, about 6 to 8 and perhaps as many as 10
tank volumes would be required, depending on how the
material is administered.
If dry ice is used, carbon dioxide will displace the air as it
sublimes. Carbon dioxide is a more effective inerting material
than nitrogen, but an 8% O2 concentration will still be re-
quired for safety purposes. Carbon dioxide at atmospheric
pressure gives off 8.7 ft 3 of gas (measured at 70 °F) for each
pound of dry ice at the sublimation point. Thus, for a 100-ft3
tank, 11.5 Ib of dry ice would be the amount which would
completely fill the tank by displacement.
There are several precautions which must be observed when
using dry ice. Solid carbon dioxide is extremely cold—109.3 °F.
Also, the gas produced will be absorbed into any water pres-
ent, thus effectively increasing the oxygen concentration.
In either case, when using either N2 or CO2, the major safety
consideration is to ensure that the tank atmosphere does not
exceed 8% before tank cutting commences.
-Foam the Tank: If product cannot be totally removed, a
method which has proven successful is to "foam" the sur-
face of the remaining liquid with a fire-fighting or vapor sup-
pression foam. This foam blend should be 3 or 4 in. thick and
will have the effect of suppressing vaporization of the vola-
294 UNDERGROUND TANKS
-------
tile material. The foam must be applied through a tank open-
ing and thus the tank must be opened. After laying the foam
blanket, the tank may have to be ventilated so that the vapor
flammable concentration is reduced to less than 10% of the
LEL on the combustible gas indicator.
In extreme cases, high expansion foam can be used to com-
pletely fill the tank. Obviously, no ventilation will be neces-
sary in this case.
The general rule of thumb for cutting tanks with a cutting torch
is that cutting slag should never be allowed to fall into free pro-
duct.
Naturally, when cutting tanks by any method, appropriate fire
protection should be available.

Other Situations
Under certain situations, the tank may need to be opened and
cleaned before removal. In this case, a reordering of steps is ap-
propriate. These are previously discussed:
Remove pumpable product
Test tank interior for flammables
If flammable, inert and purge
Retest atmosphere and repeat until safe
Open tank
Retest atmosphere for flammables and toxics
Clean as much as possible remotely with equipment
Allow properly protected personnel to enter for final cleaning
steps
SUMMARY
The industry has successfully tested and removed thousands of
tanks. Proper planning and hazard recognition have been keys to
this success. Before commencing any tank project, our policy is to
investigate the problem, develop an operational plan and
establish the various levels of environmental and personal protec-
tive controls and procedures. If in doubt about these procedures,
it may be necessary to consult a safety professional.
UNDERGROUND TANKS 295
-------
Underground Storage Tank Corrective
Action Technologies

Douglas C. Ammon
Hazardous Waste Engineering Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
S. Robert Cochran
PEI Associates, Inc.
Washington, D.C.
INTRODUCTION
An estimated 1.4 million underground tanks in the United
States are used to store liquid petroleum and chemical substances.
Government and industry sources estimate that in 1986 between
100,000 and 400,000 of these tanks and associated piping sys-
tems may leak, resulting in the contamination of subsurface soils,
migration of toxic or explosive vapors and contamination of
groundwater and surface water. Environmental and tank moni-
toring, tank replacement, product recovery and corrective ac-
tion technologies can reduce the potential of future releases or
remedy prior releases from underground storage tanks.
Industry sources estimate that the average cost of cleanup after
a tank leaks is approximately $70,000; however, should tank re-
moval and treatment of surrounding soils be required, costs may
approach or exceed $1 million. Corrective action costs often ex-
ceed $1 million if underground tank leakage has significant effect
on groundwater, surface water, drinking water supplies, reser-
voirs, sewers or utility trenches. The primary factors affecting
corrective action costs are: the magnitude of release; the toxicity
of the substance; the hydrogeologic setting; and the environ-
mental criteria, standards, levels or objectives applicable to the
cleanup.
Concern for the status of underground storage tanks has re-
sulted in the development of legislation regarding the registration
and monitoring of existing tanks, design and installation of new
tanks and corrective actions for releases. At the federal level, the
1984 Hazardous and Solid Waste Amendments added to RCRA
a new Subtitle I, "Regulation of Underground Storage Tanks."
Numerous states have passed laws requiring similar, often more
stringent, regulations.
Relative to Subtitle I, the decision-maker is influenced by sev-
eral factors in choosing the path of corrective action for a leak-
ing underground storage tank (UST). This paper provides some
guidance in establishing and maintaining scientific and technical
direction associated with response activities. The paper also in-
cludes general background information relative to UST construc-
tion techniques, leak detection methods and failure mechanisms.
Information is presented regarding transport pathways of re-
leased substances, techniques for evaluating the extent of release,
factors influencing risk to human health and the environment,
techniques for selecting initial corrective-action response technol-
ogies and detailed technical profiles of corrective-action tech-
nologies. Emphasis is on corrective actions associated with re-
leases from gasoline and petroleum USTs; however, comprehen-
sive profiles are provided for technologies used in responding to
chemical releases. A more detailed discussion of the topics pre-
sented in this paper is contained in a draft document entitled
"Underground Storage Tank Corrective Action Technologies,"
prepared by the U.S. EPA, HWERL.

UNDERGROUND STORAGE TANK PROFILE
Under RCRA Subtitle I, an underground storage tank is a tank
that stores "regulated substances" and that has at least 10% of
its volume below the surface of the ground, including piping con-
nected to the tank. Regulated substances include hazardous
chemical products regulated under CERCLA and petroleum pro-
ducts, including crude oil and refined products that are liquid
at standard conditions of temperature and pressure.
Millions of tanks have been installed underground to store
many materials. The configurations of these installations vary to
suit several constraints including geography of the site, material
stored, insurance underwriter requirements, government regula-
tions and the owner's operation. The installation most likely has
these basic components:
• Tank or tanks
• Anti-flotation anchorage
• Piping system
• Pumps
• Means for level gauging
• System for corrosion protection (if metal tanks)
The primary purposes for placing tanks underground are to
assure the safe storage of flammable, combustible and reactive
materials and to make optimum use of confined property sizes.
Many tanks were installed underground on the premise that they
would never develop leaks; thus, little consideration was given
to the consequences of any leakage that might develop.
Two broad categories of underground tank applications are
storage at gasoline stations and industrial/commercial installa-
tions. Underground tanks in gasoline stations are used to store
gasoline (leaded, unleaded and premium grades) and diesel fuel;
waste tank installations store a wide variety of materials includ-
ing solvents and various hydrocarbons. Industrial tanks may be
used to store either new materials or waste products. These in-
stallations usually have individual tanks for each stored material
and not multiple tanks connected by common suction piping, as
commonly found in gasoline stations. System piping in industrial
installations is not always located below grade but is run on over-
296 UNDERGROUND TANKS
-------
head racks, if they are available. In commercial installations such
as dry cleaning establishments, system piping generally will be
found below grade into the building.
Underground tanks can be found in various sizes, shapes and
construction materials. Metal tanks usually are welded and have
some kind of exterior coating or cathodic protection against cor-
rosion. Tanks fabricated from fiberglass, epoxy or other non-
metallic material may be found in newer installations; these gen-
erally do not require any coating for corrosion resistance. New-
er installations might also have tanks with double wall construc-
tion. The annular space between the walls is used in various ways
to detect product leaks from the inner tank wall or groundwater
leaks from the outer tank wall. Secondary containment systems,
generally a flexible membrane liner surrounding the tank(s),
are also found in newer installations.
Depth of tank burial varies, but the tank top is generally not
more than one tank diameter below finished grade. In some in-
stallations, tanks will be found in mounds, partially buried be-
low grade. Underground tanks usually are paved over to permit
traffic to travel above them. Manholes, caps or other hardware
are provided to cover and protect tanks appurtenances such as
the fill connections and gauge pipes. Some large tanks are equip-
ped with a manhole to allow access into the tank. Unless the water
table is well below the tank excavation, an anchor system is pres-
ent to prevent the tank from floating out of the ground.
Underground storage tanks generally include a method for
tracking inventory such as the use of a metering stick inserted in
the fill tube. More elaborate systems might employ bubbler type
pneumatic level sensors or use differential pressure instrument.

FAILURE MODES
Leaks can occur from storage tanks and associated piping and
pumping via leaks from corrosion, from system rupture due to
overloading, from external stresses or puncture and from faulty
construction/installation. In addition to other system failure,
overfilling is another source of tank system release.
The most common failure mode for underground tank systems
is corrosion of the tank or piping. Corrosion may be traced to
failure of the corrosion protection system through "holidays" in
the coating or taping system, depletion of the sacrificial anodes
and corrosion inside promoted by the stored product. Many leaks
due to corrosion will be encountered in older tanks that have no
corrosion protection at all.
Fiberglass-reinforced plastic and other nonmetalh'c tanks and
piping are inherently corrosion-resistant, but these materials have
lower structural strength than steel and require greater care dur-
ing installation. Reinforced ribs usually are incorporated to with-
stand both the internal stresses from the stored liquid and the ex-
ternal loads. Some resins used in tank construction may lose
structural strength when exposed to certain chemicals, while
others may dissolve, soften or become brittle in acidic saline soil
environments. Failure to observe the design limitations and to
follow the handling and installation requirements for tanks can
lead to tank failure.
A substantial percentage of underground storage system leaks
occur in piping systems. The leaks probably occur here because
the majority of piping systems utilize threaded joints, which are
vulnerable to corrosion from outside with inadequate corrosion
protection. Piping joints also may leak because of improper seal-
ing during installation or loosening due to vibrations, traffic load-
ings, temperature cycling, frost heave and backfill settlement.
Overfilling is another source of leaks or spills from tanks. Spills
of this type can be caused by human error, failure of shutoff
valves of the delivery source and failure of the tank level indica-
tor. These spills generally are small (less than a few gallons) but
can be much larger if the equipment is left unattended during tank
filling or if there is a failure of shutoff valves.

LEAK DETECTION
Leaks should be identified early and initial assessments should
be performed promptly to minimize the adverse effects of re-
leases from underground storage tanks. Leaks often are discov-
ered due to an odor in groundwater, in a confined space or near a
utility conduit. Unfortunately, discovery from odors or vapors
may occur well after the initiation of the leak. A successful tank
monitoring program should provide early detection, occur on a
regular basis and employ accurate leak detection systems. Once a
release is discovered, accurate characterization of the extent of re-
leases and the associated migration pathways is important.
The four classes of methods to detect leaks in underground
storage tanks are: (1) volumetric (quantitative) leak testing leak
rate measurement; (2) nonvolumetric (qualitative) leak testing;
(3) inventory controls; and (4) environmental effects monitoring
(outside detection). A state-of-the-art review leak detection is
given by Niaki and Broscious.'
Volumetric leak testing is based on detecting a change in tank
volume by measuring parameters such as liquid level, tempera-
ture, pressure and density. This category of leak detection in-
cludes methods that use an air bubbling system to monitor pres-
sure changes resulting from changes in product level in the tank.
Other methods detect level changes by using either a "J" tube
manometer, a laser beam and its reflection or a "dip-stick" type
device. Another approach to leak detection is to measure any
volume change by maintaining a constant level.
Nonvolumetric leak testing is used to determine the pressure by
qualitative methods, usually by using a second material other
than the product (tracer material). The tracer material, usually
helium, is generally used to pressurize a tank, and a leak is indi-
cated by either loss of pressure or the detection of the tracer gas
outside the tank by a sniffer mass spectrometer. Because of the
rapid diffusivity of helium, it can escape through a tank leak size
of 0.013 cm (0.005 inch). This type of testing can be used to de-
termine if there is a leak, while volumetric testing may be used to
determine the leak rate.
Inventory control is perhaps the simplest and most economical
leak detection method. Leaks are detected by keeping records of
tank inventories and noting any unexplained change in liquid lev-
els or amounts. Inventory monitoring can be performed by gauge
stick, by electronic level measurement and by weight monitor-
ing using pressure and density measurements. Although the prob-
lem of keeping records of tank inventory is complicated by the
fact that petroleum products and other chemicals are volatile
and losses due to evaporation are possible, the inventory method
should be used as a first and convenient method for gross leak
monitoring.
Environmental effects monitoring typically identifies leaks by
monitoring wells outside the storage tanks. Other environmental
monitoring techniques may be employed, such as volatile organic
carbon vapor analyzers, suction lysimeters and other similar tech-
niques. Environmental effects monitoring may not be able to dis-
tinguish which tank is leaking if multiple sources are present and
also may detect other sources and interferences. These methods
do, however, provide a direct measure of the environment and,
once installed, provide a method to frequently check for leaks.

CORRECTIVE ACTIONS
Corrective action responses for leaking underground storage
tanks can be categorized into two areas. Initial responses are
directed at controlling the immediate impacts resulting from new-
UNDERGROUND TANKS 297
-------
ly discovered or sudden releases. If an emergency situation exists
immediate responses may include notification of fire depart-
ments, evacuation, removal of ignition sources and venting of
flammable vapors. Permanent corrective measures are directed
at cleanup of the releases to some acceptable levels for protec-
tion of human health and the environment. A flow chart of cor-
rective action responses is presented in Fig. 1. The degree of de-
tail and amount of resources applied in these steps is a function
of site-specific conditions. Some of the most critical site-specific
conditions include the volume of release, the time frame within
which the release transpired, the site hydrogeologic conditions
and the proximity of environmentally sensitive communities and
human receptors.
Response options for leaking USTs include a large population
of control and treatment technologies; however, the process for
implementing these technologies can be narrowed to initial re-
sponses and corrective actions. Often it is practical to package a
set of these technologies within a corrective action to effectively
achieve the cleanup goals. Because of differing response objec-
tions, initial and permanent corrective actions often require dif-
ferent implementation strategies and varied application of similar
technologies.
Often the transition between initial response actions and im-
plementation of corrective actions is bridged by data assessment
and site investigation efforts. Assessment and investigation activ-
ities generally focus on determining the need for and extent of
additional corrective measures. In some cases, the extent of the
initial response action may simply be the completion of the site
assessment and continued environmental monitoring.
Figure 1
Corrective Action Response Process
During initial response actions, the focus is on alleviating the
immediate threat to public health and safety. Fig. 2 illustrates
the typical initial response process. The critical element of initial
response actions is the timing associated with implementation.
Effective implementation of initial response technologies com-
monly dictates that field deployment occur within hours of the
discovered release. It is essential that early discovery and contain-
ment of any suspected leak occur as soon as possible so that re-
covery procedures may be initiated and the influence of any re-
lease can be minimized. Normally, the first action taken is the
removal of any remaining product in the leaking UST. In addi-
tion, measures often are taken to minimize imminent and immed-
iate risk to human health and the environment. Fig. 3 illustrates
a list of options available to counter typical quick-response situa-
tions.
Response options listed in Fig. 3 are similar to options often
used in more comprehensive corrective action efforts. However,
the significant difference between the two is that typical response
efforts are generally of short duration and apply limited re-
sources. As shown in the initial response flow diagram, the im-
plementation of initial response efforts typically occurs within a
short time-frame (24 hr). Most initial responses use similar tech-
nologies such as tank repair/removal, soil excavation and free
product recovery, but the decision about which actions are ap-
propriate should be made on a site-by-site basis.
Figure 2
Initial Response Process
Figures
Quick Response Technologies

The selection of more permanent corrective actions requires a
higher level of analysis and involves a broader range of potential
technologies. Fig. 4 presents a listing of technologies that may be
used as permanent corrective actions for four classes of releases
profiles.
During permanent corrective action, the focus is on site char-
acterization and assessment to determine the need for further
action beyond source control, on calculating transport rates, on
assessing the hazard to the environment and potential human re-
ceptors, on collecting data for selecting and design of correc-
tive actions and on determining the effectiveness of corrective
actions. These activities may involve hydrogeologic, environ-
mental and atmospheric monitoring; chemical analysis, fate and
transport modeling; and analysis of corrective action effective-
ness.
When selecting corrective actions, the physiographic location
of the release should be considered. Often leaks associated with
gasoline stations are located within residential communities or
light commercial areas. As a result, corrective action planning
must take into consideration the following logistic and engineer-
ing factors:

• Limitations of the amount of land available for implementing
corrective actions
• Restricted property boundaries requiring easements to conduct
off-site activities
• Proximity to densely populated zones
298
UNDERGROUND TANKS
-------
fwn'-'^ fit lant lf"t aftf* I
L«ndf.!l,n3
Lmntftirmng
SoH vHhing
Dual ponp iyct*i

Surface nl/wawr

•pLfy^ftH tfrnji
Air dipping
Cwtxn cdKMpiion
ftdogicil tf««rn*nt
Shidg* d«wai*rjng

Gr.Ang
fttilflrHJM rf COfMITiMUd »•!.> •i
»n«t«« point at-u» ••tor •
lf»«im*nt o* paml-of-ut* water oupptH
FUplanrMM o( waltr *nd M««r bn>i
Cl*af«nQ/i*iia(B«an of water artf Mwar IIMI
ff
Figure 4
Corrective Action Technologies Relative to Release Volume and
Chemical Characteristics

• Complexities associated with subsurface cultural features (sew-
er, water and electrical utilities, basements, wells, subways and
tunnels)

Typical site problems are presented in Fig. 5 along with asso-
ciated categories of corrective action technologies. Once appro-
priate corrective action categories have been identified, the fol-
lowing technical, institutional, environmental and public health
factors may need to be evaluated:
• Performance including effectiveness, time to achieve given level
of responses and useful life
• Reliability including operation and maintenance requirements
• Bnplementability relative to site conditions
• Safety and health of nearby communities and workers

Oi»( pump •f*l«r>«
Flodtng /<«•!• pump*
Surlmef mllfwmlft ttpir
S>L»V ••Hi
Growing
Shirt pl*<
Sell Humming
eiofllmulftlen
Blalogle*! lr*mlm
P..op.uiioMl«oj(»l
Diltdvcd. •» HolMion
GxnUv m*4« W«rt«
Ion •>eh*nVM.iin id
On id Btiwtf reduction
* ••!« >uc«A«4
«« »nd «-tr hn*t
n e( ••)•' «nd wwi hr
Note: Technologies with bold print are likely to be used in
responding to UST releases at gasoline stations

Figure 5
Site Information Needs for Evaluation of Corrective Action Alternatives
• Public health and environment, including the ability to miti-
gate exposure concerns, effects on environmentally sensitive
areas and exceedance of environmental standards
• Costs, including capital, operational and maintenance
• Institutional, including regulatory compliance (local, state and
Federal regulations) and potential responsible parties
Each measure should be evaluated prior to committing to a cor-
rective action approach. The degree of evaluation is dependent on
the anticipated magnitude of response. The results of the decis-
ion-making and evaluation process should be documented for use
in later technical and institutional review.
BIBLIOGRAPHY
1. American Petroleum Institute, "Underground Spill Cleanup Man-
ual," API Publication 1628, API, Washington, DC, 1980.
2. American Petroleum Institute, "Laboratory Study on Solubilities of
Petroleum Hydrocarbons in Groundwater," API Publication 4395,
API, Washington, DC, 1985.
UNDERGROUND TANKS 299
-------
3. American Petroleum Institute, "Recommended Practice for Bulk
Liquid Stock Control at Retail Outlets. Publication 1621, API,
Washington, DC, 1977.
4. American Petroleum Institute, "Groundwater Monitoring and Sam-
ple Bias," Publication 4367, API, Washington, DC, 1983.
5. American Petroleum Institute, "Protecting Groundwater," API,
Washington, DC, 1985.
6. Camp Dresser, and McKee, Inc., "Fate and Transport of Sub-
stances Leaking From Underground Storage Tanks." Volume 1,
Technical Report. Interim report prepared for the U.S. EPA under
Contract No. 68-01-6939, 1986.
7. Montgomery, R.E., Remeta, D.P. and Gruenfeld, M., "Rapid On-
Site Methods of Chemical Analysis," in Contaminated Land, Smith
M.A., Ed., Plenum Publishing Corporation, New York, NY, 1985.
8. Niaki, S. and Broscious. J.A., "Underground Tank Leak Detection
Methods: A State-of-the-Art Review," EPA-600/2-86-001, 1986.
9. Shepherd, W.D.. "Practical Geohydrological Aspects of Ground-
water Contamination," Department of Environmental Affairs, Shell
Oil Company, Houston, TX, undated.
10. Bureau of National Affairs, "Toxic Contamination from California
Electronics Industry Proves Costly to Cleanup; North Carolina
Seeks to Avoid Problems," Environ. Reporter. 15, 1984, 972-977.
11. Camp, Dresser, and McKee, Inc., "Fate and Transport of Sub-
stances Leaking From Underground Storage Tanks, Volume 1,
Technical Report," Interim report prepared for the U.S. EPA under
Contract No. 68-01-6939,1986.
12. Camp, Dreiser, and McKee, Inc., Technical Summary, "Mobile
Treatment Technologies, Capacity and Capability of Alternatives to
Land Disposal for Superfund Wastes," Review draft prepared for
the U.S. EPA under Contract No. 68-01-7053, 1986.
13. Hansen, P., "L.U.S.T.: Leaking Underground Storage Tanks, In:
Proc. National Conference on Hazardous Wastes and Environmental
Emergencies, Cincinnati, Ohio," May 1985,66-67.
14. JRB Associates, "Methodology for Screening and Evaluation of
Remedial Responses," Draft report prepared for the U.S. EPA,
Cincinnati, OH, 1984.
15. Knox, R.C. et al., "State-of-the-Art Aquifer Restoration, Volume 1
and Volume2," EPA/600-2-84-182a and b, 1984.
16. Standen, A., Ed., Kirk-Oihmer Encyclopedia of Chemical Tech-
nology. 2nd ed., John Wiley & Sons, New York, NY, 1967.
17. U.S. EPA, Handbook of Remedial Action at Waste Disposal Sites
(Revised, EPA-625/6-85-006, U.S. EPA. Washington, DC, 1985.
18. Unterberg, W., et al.. "Manual for Preventing Spills of Hazardous
Substances at Fixed Facilities," Draft report prepared by Com-
bustion Engineering for the U.S. EPA under Contract No. 68-03-
3014, undated.
300 UNDERGROUND TANKS
-------
Compliance Strategy for Underground Tanks:
A Cost Assessment

Alan Lament, Ph.D.
Woodward-Clyde Consultants
Walnut Creek, California
James D. Hartley, P.E.
Woodward-Clyde Consultants
San Diego, California
ABSTRACT
This paper describes an economic model used to evaluate mon-
itoring and replacement strategies for existing underground stor-
age tanks. The model identifies the optimal strategy for an
individual tank taken in isolation and can be used to develop an
overall set of priorities and a strategy for a large inventory of
tanks.
The model makes use of fairly inexpensive information about
each tank and its site to estimate the total expected cost of
operation and leak remediation for a range of strategies, tak-
ing into account the probability of a leak and the expected remed-
iation costs if the tank does leak.
The use of the model to develop strategies for individual tanks
and a group of tanks is demonstrated.

INTRODUCTION
A number of states have promulgated regulations requiring
that all existing underground tanks be monitored for leaks and
that all new underground tanks be constructed with designs that
minimize the chance of a leak and which make monitoring leaks
far more effective. In addition, the new regulations, and the legal
environment in general, have made the tank owner responsible
for remediating the damage caused by any leaks that might occur
from his tanks. As a result, maintaining an underground tank,
particularly an existing tank, has become quite expensive and is
fraught with substantial financial risk.
This paper develops a probabilistic financial model of the vari-
ous options available to the owner of an existing tank. It iden-
tifies the basic policies which are generally most advantageous
based on the net present value of all the tank costs that the owner
might incur. The model will be most useful to the owner of a large
number of underground tanks, and the paper illustrates how the
results of the model can be used to develop an overall strategy
for managing a large inventory of tanks.

OPTIONS AVAILABLE TO THE TANK OWNER
The California state regulations, for example, allow the tank
owner to monitor his underground tanks using one of eight differ-
ent alternative monitoring systems. These alternative systems con-
sist of various combinations of tank testing, vadose zone mon-
itoring, groundwater monitoring, inventory reconciliation and
pipeline leak detection (Table 1). This paper analyzes six of the
options since the others often are practical only in special cir-
cumstances.
The monitoring alternatives are not the only alternatives avail-
able to the owner: the tank can be replaced immediately with a
new double walled tank, or the need for a tank can be eliminated
altogether. Each alternative should be considered.
The tank owner's options do not stop here. There is also the
question of when to remove the existing tank. It can be moni-
tored for a few years and then be replaced. Clearly, it is highly de-
sirable to remove the tank prior to the time it starts to leak. In
fact, the analysis below shows that the loss incurred by remov-
ing the tank early is far smaller than the risks incurred by leaving
the tank in beyond its anticipated life.
Table 1
California Tank Regulations Monitoring Alternatives
TECHNIQUES
MONfTOFUNG
SYSTEMS

CA-1

CA-2

CA-3

CA-4

CA-5

CA-6a

CA-6b

CA-7

CA-8
Tank Invent. GrndWal. Vadose Pipeln
Testing Reconc. Monit Uonil. Leak Detect.
.1

.2
1 tank gauging

2 tank gauging or inventory reconciliation
In this analysis, we assume that the tank owner sets a policy for
each tank. We will call this policy a "single tank" policy. The
single tank policy for a specific tank first states whether or not it is
to be replaced immediately (i.e., the existing tank will not be mon-
itored). If the tank is to be kept and monitored, the policy states
how it will be monitored. Finally, the policy sets a planned re-
placement year. If the tank leaks before the planned year, it is
replaced. If the tank does not leak before the planned replace-
ment year, it is replaced at that time. The objective of the eco-
nomic analysis is to determine the best policy for each tank.
For the multiple tank owner, a multiple tank policy is required.
Our work to date indicates that the best policy in most cases is
immediate replacement. This is not always financially or logis-
tically feasible for the owner of many tanks. A multiple tank pol-
icy specifies the order of tank replacement so that the most crit-
ical tanks are replaced first. This decision implies that many tanks
UNDERGROUND TANKS 301
-------
will be replaced later than implied in the optimal single tank pol-
icy. The model provides the information needed to do this.

APPROACH TO ECONOMIC MODELLING
The economic model estimates the expected present value of
the total cost of maintaining a tank in a given location for 20 yr
under each possible single tank policy. The costs estimated by the
model are not exact since they are based on a small amount of in-
expensively obtained information. They estimate the magnitude
of the expected costs for each tank, and they allow comparisons
between the tanks for planning purposes. More exact cost esti-
mates could be made using this approach, but the cost of making
them for a large population of tanks would be excessive.
The total cost is made up of the following component costs:

• Operating the existing tank until replacement
• Installing and operating a monitoring system on the existing
tank until replacement
• Installing and operating the new tank and its monitoring system
• Remediating any leaks

It is simpler to first discuss the economic analysis without con-
sidering leaks. This method removes the probabilistic issues in the
analysis and allows us to focus on some basic and important
issues that might otherwise be obscured. The probability of a leak
and its costs then can be added to the analysis relatively easily.

ECONOMIC ANALYSIS ASSUMING
NO LEAKS OCCUR
This part of the analysis focuses on the first three compon-
ent costs listed above. In the discussions below, we will compute
the present values of all of these costs for all the replacement years
between 0 and 20. This analysis assumes an interest rate of 10%.

Costs of Tanks Alone
The cost of operating the existing tank is simply the present
value of the stream of annual operating costs between now and
the replacement year. The cost of installing a new tank depends
on the size of the tank. In this analysis, we have assumed a linear
function of a fixed cost of $18,000 plus SO.SO/gal of capacity. In
California, there is a significant cost to closing an existing tank
and this is included in the fixed cost.
The cost of operating the new tank and its monitoring system
is the present value of the stream of annual operating costs. Note
that these costs start in the planned replacement year and must be
Fl Tnk Inst&OpCsl
70,000
Tank Vol.
(0*1)
••- 1000

•O- 5000

••- 10000

•D- 20000

•*- 40000
10,000
1 4 6 8 10 12 14 16 II 20
Plannwi Replacement Year
Figure 1
New Tank Installation and Operation Costs
(including cost of closing existing tank)
discounted back to the present.
Finally, the new tank will have an asset value at the horizon
year since it still may have useful life. This value is simply com-
puted by multiplying the proportion of its life remaining by its
installation cost and discounting back to the present.
The costs of these components varies with tank size and the
year of planned replacement (Fig. 1). This plot then shows the
costs of owning the tanks (both existing and replacement) for
20 yr as a function of the year in which the existing tank u re-
placed. The costs all decline as the replacement year approaches
20 yr. This result is due to the reduction in present value that is
achieved by delaying the one large expenditure for replacing die
tank. In the initial years, this decline is approximately equal to
the interest rate multiplied by the cost of replacement (assum-
ing the cost of operating the new tank is similar to the cost of
operating the existing tank). Due to this phenomenon, the sav-
ings achieved by delaying the replacement of a large tank are
greater than the savings achieved by delaying the replacement of
a small tank.

Cost of the Monitoring System on the
Existing Tank
The monitoring system on the existing tank has an immed-
iate capital cost and an annual operating cost. The costs used in
these analyses are based on our experience in California. The
total present values of the costs of the monitoring system on the
existing tank are shown in Fig. 2 as a function of the year that the
existing tank is replaced.
F2 WON SYS CST PLT
Man*. Syv
O- CA-3

»OM

O- CA-S

*- CA«i

«• CA-Sb
Figure 2
Monitoring System Installation and Operation Costs
(for existing tank)
Total Cost of Tank and Monitoring
System: Some Lessons
The total cost of the tank and monitoring system depends on
the size of the tank and the monitoring system chosen for the
existing tank. Fig. 3 through 6 show the costs for a small and a
large tank and for a relatively inexpensive and a relatively ex-
pensive monitoring system. These figures show the costs for the
installation and operation of the tank on the bottom and the cost
of the respective monitoring system on the top.
For the smaller tank with the less expensive monitoring sys-
tem (Ca-3), the present value of the cost increases if the tank is
replaced in 1 yr and gradually declines if the tank is replaced in
later years. However, even if the tank is not replaced until the
20th yr, the total cost is only about $12,000 less than immed-
iate tank replacement. The existing tank must be kept for 7 or 8 yr
302 UNDERGROUND TANKS
-------
F3 ins & opercsl5K tnk
40,000
Figure 3
Present Value of Installation and Operation
(5,000 gal tank with monitoring system CA-3)
F5ins & op cst 40k tnk
0 1 23 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Planned Replacement Year
cost of mon. syst. CA-3 • lank instal and oper. cost
Figure 5
Present Value of Installation and Operation
(40,000 gal tank with monitoring system CA-3)
F4 inst & opr cst 5k tnk
6 7 8 9 10 II 12 13 14 15 16 17 18 19 20
Planned Replacement Year
cost o( mon. syst. CA-6b
tank inst. and oper. cost
Figure 4
Present Value of Installation and Operation
(5,000 gal tank with monitoring system CA-6b)

before the total cost of the monitoring alternative is equal to the
cost of immediate replacement.
For the smaller tank with the more expensive monitoring sys-
tem, the total costs rise continually because the annual cost of
operating the monitoring system is greater than the saving
achieved by delaying tank replacement.
For the larger tank, we see somewhat the same pattern. The re-
sults favor keeping the existing tank longer because monitoring
costs are about the same, but the saving achieved by delaying re-
placement is much greater.
In summary, we see that it often is not economical to keep an
existing tank based on monitoring costs alone. This conclusion
is more valid for smaller tanks than larger tanks since there is a
smaller advantage to keeping the tank. In the next section, we
will add the expected costs of remediation. These costs will al-
ways tend to increase the cost of keeping a tank and will tend to
make immediate tank replacement more advantageous.

ECONOMIC ANALYSIS WITH THE POSSIBILITY
OFLEAKS
A tank will leak eventually if it is left in the ground. When it
does leak, it is likely to be quite expensive. The monitoring sys-
tems required on existing tanks will help to protect the owner by
alerting him when a leak does occur. However, these systems are
F6 ins & op cst 40k tnk
6 78 9 10 11 12 13 14 15 16 17 18 19 20
Planned Replacement Year
cost of moniL syst. CA-6b
tank instal. and oper. cost
Figure 6
Present Value of Installation and Operation
(40,000 gal tank with monitoring system CA-6b)

not complete protection. There are many possible reasons for
this: groundwater monitoring systems might be installed in the
wrong locations, vadose zone sensors might be allowed to de-
teriorate, inventory reconciliation and tank testing are prone to
random errors leading to false negatives. In any case, a certain
amount of material must leak out before the monitoring system
will alert the operator of a leak. In some situations, even this
amount of leakage may cause a significant remediation cost.
Therefore, an analysis such as this must include the cost of leaks.
Under our analysis of a management policy, it is assumed that
the tank is replaced in a specified year. Ideally, the tank will not
begin to leak before the replacement year. But there is some prob-
ability that the tank will begin to leak before then. To evaluate the
total expected cost of a policy, we must determine the probability
that the tank will leak prior to the planned replacement date and
the expected remediation costs that will be incurred if the leak
occurs. These two issues are discussed below.
Probability of a Leak
The age at which a tank begins to leak depends on many factors
including its material, construction, soil chemistry, soil satura-
tion and cathodic protection. For a given tank, one cannot know
ahead of time when it will leak, but one can estimate a probability
distribution over the time at which it will leak.
Several models of tank age to leak have been developed. In our
UNDERGROUND TANKS
303
-------
work, we have used a model developed by Warren Rogers Asso-
ciates' which is developed for single-walled steel tanks. This
model is useful for initial tank screening to identify those that
clearly are or are not risky. Another model was developed by
Pope-Reid Associates for the U.S. EPA.1 These are simple mod-
els based on limited data. There are more sophisticated proprie-
tary models available from private consultants covering a wider
range of tank types and are expected to provide better estimates.
Using any of these models, one can develop the probability dis-
tribution over the age at which the tank will begin to leak.

Expected Cost Given When a Leak Occurs
Once a leak occurs, the questions are: how much liquid will
escape before the monitoring system detects it, and how much will
it cost to clean up the spill? A monitoring system (ideally) uses
several leak detection techniques. Nearly all of the California
monitoring options are multiple technique systems. None of these
techniques is guaranteed to detect the leak. Further, each tech-
nique requires some amount of discharge must occur before that
technique can detect the leak.
Each monitoring system can be modelled by estimating the
probability that the leak will be first detected by each technique
in the system. Given the volumes associated with each technique,
we can derive the probability distribution over the volume that
will be leaked before detection.
For each volume leaked, there is a cost to remediate. These
costs depend on the material in the tank, depth to the water table,
potential beneficial uses of the water and the possibility of con-
taminating nearby utility vaults. For our work, we have developed
a simple computer model of remediation costs based on this sort
of site information. The model is useful to estimate the order of
magnitude of possible remediation costs as a function of volume
leaked and thereby measure the environmental sensitivity of the
site.
Once one has estimated the probability distribution over the
volume leaked and the costs as a function of the volume, one
can compute the expected cost of a leak for a given monitoring
system. For example, consider a system which includes vadose
zone and groundwater monitoring. The vadose zone system might
have an 80% chance of detecting the leak. If it detects, it will de-
tect after only a few gallons have been released. There is thus an
80"% chance that the cost of a leak will be quite small with this
monitoring system might have a 70% chance of detecting, but it
might require that 1,000 or 2,000 gal of material be leaked before
it does detect the leak. At this point, there is a 14% chance
[(1.0 - 0.8) x 0.7] that the leak will have been detected first by
the groundwater monitoring system with the expense of remed-
iating a 1,000 to 2,000 gal leak. This still leaves a 6% chance that
neither system detects and that the leak is detected through fumes
in a nearby sewer or by contamination in a downstream well.
This would lead to a much larger cost.

Total Expected Costs with Leaks
With this information, we can compute the expected cost of
remediation for each policy (consisting of a specific monitor-
ing system and a planned replacement year). This cost is com-
puted in the model by evaluating the probability that the tank will
begin to leak in each year up to the planned replacement year
and the probability that the tank will not begin to leak before
the planned replacement year. If the tank leaks before the re-
placement year, a cost of remediation is incurred. The expected
cost of the policy includes the expected cost of this remediation.
If the tank does not leak before replacement, no remediation cost
is incurred and only the normal operating and replacement costs
are incurred.
The total expected cost is approximately equal to the expected
cost of a leak at that location, with the monitoring system in
place, multiplied by the probability that the tank leaks before
the planned replacement year.
Fig. 7 shows the different expected remediation costs for sev-
eral different monitoring systems at a site with a very shallow
groundwater table (a few feet). This tank, Example Tank A, is
5 yr old and has a capacity of 8,000 gal of gasoline. The shallow
groundwater tends to make groundwater monitoring a relatively
effective monitoring technique and makes vadose zone devices
less sensitive. Fig. 8 shows a similar analysis for a site with a
deep groundwater table (150 ft), in this case, vadose zone devices
function well. With the deep water table, groundwater monitor-
ing is relatively ineffective because a large amount of material
must be leaked before detection occurs. This tank, Example Tank
B, is 4 yr old and has capacity of 1,000 gal of gasoline.
In each case, we see that the expected remediation cost is very
low if the planned replacement year is early. If replacement is de-
layed, the expected costs generally begin to rise substantially.
Note that there is a great deal of difference between monitoring
systems, however.
F7 Rem Csis S GndWal Pit
ktoiul Sftt.
•• CA-2

O- CJW

•• CA-4

O- CA-S

*" CA-6.

A- CA-6t>
10 12 14 16 18 20
Figure 7
Expected Remediation Costs
(Example Tank A, shallow groundwater site)

p
r 200.000
0
, 160.000
c
" 0
' » 120.000
1
V ,
• 80.000
1
u $
0 40.000 <
o
( 0 '
C

F8 Rem Csis D GndWal Pit

I
I
/" .'*'*"*
/ s
1 • _J»-B~II"IM>~B
f /'V1
m* a*S-5*/N 0.^-o-f>-o-o-o-o~o
2 4 6 8 10 12 14 16 18 20
Planned R0plac«m*nt YMT

Monl. Syct
•• CA-S
0 CA-3

m- CA-4
0- CA-S
A-CA-6.
A- CA-tt

Figure 8
Expected Remediation Costs
(Example Tank B, deep groundwater site)

Typical Results From (he Analysis
Fig. 9 and 10 show the total costs for the same two tanks. These
two examples have been chosen to illustrate several characteristics
of these analyses.
304
UNDERGROUND TANKS
-------

p
r
e
s
8 C
n o
1 s
a
1 '
u $
e
0
1

300,000 <
250,000 •
200,000 •

150,000
100,000

50,000
0

/'
/
/
'fa

jts'"

0 2

F9 Tot Exp Cst S Gndwat

/* Monit. Syst.
.• ^g.o-o-o-o-o-o-o-o-o-o-o
• o
£ .0 _n.n-n-n-Q-n-n-D-n-D-n-a

'•Sr^^

•- CA-2
0- CA-3
•• CA-4
a- CA-S
•A- CA-6a

A- CA-6b

4 6 8 10 12 14 16 18 20
Planned Replacement Year
Figure 9
Total Expected Costs
(Example Tank A, shallow groundwater site)

P
r
e
s
e C
t
s
v'
S
a
I '
:*

F10 Tot Exp Cst D Gndwat

300,000
250,000

200,000
150,000 •

100,000 •
50,000 :

0 •
I Monit. Syst.
/

/_
„•-•-•
,'
• ft A A rt A
• .X Ji**& **
/ X?-^A*^
> ' n-fl**^^ n-n-n-a-a-c-a-a
jffisSET'o*;* .11*1 ^f*J 1 A A i A t * i — t

•- CA-2

0- CA-3
•• CA-4
a- CA-S
*- CA-6a
^r CA-6b

0 2 4 6 8 10 12 14 16 18 20
Planned Replacement Year
Figure 10
Total Expected Costs
(Example Tank B, deep groundwater site)
First, it is apparent that some monitoring systems can be very
effective at controlling total expected costs if they are matched to
the site and if the site has relatively low sensitivity. Second, we
see two distinctly different patterns in these two analyses. In
Fig. 9, Example Tank A, we see that the best policy is immediate
replacement, regardless of which monitoring system is chosen.
Even if the best system is chosen, the total cost of keeping the
existing tank for & yr would be about $50,000 greater than sim-
ply replacing the tank immediately. And this is with a tank that is
only a few years old. To put this in perspective, it would cost
about $6,000/yr to keep this tank (during the next 8 yr) rather
than replace it now.
In Fig. 10, Example Tank B, we see a different pattern. Here
there is actually a slight cost savings if the tank is kept, provided
that we use monitoring system CA-5 (tank testing, inventory
reconciliation and pipeline leak detectors). This monitoring sys-
tem is very inexpensive and, although it is not a highly effective
system, it is still an economically advantageous system because
this site has a lower sensitivity (due to a very deep groundwater)
and a small leak would not cost that much to remediate. Further,
the tank is young and not likely to begin leaking for several years.
We should also note, though, that if the tank is kept beyond
about 9 yr with this monitoring system the expected costs begin to
rise significantly because after 9 yr the probability that the tank
will begin to leak starts to rise. The analysis shows that if the tank
is to be kept beyond 9 yr, it would be preferable to use either
monitoring system CA-3 or CA-6a, both of which use vadose
zone devices.
There is one other pattern that we have observed in some cases.
That is the cost of keeping the tank declines over the entire 20-yr
time horizon. In essence, the analysis indicates that one simply
put a good monitoring system on the tank and keep it until it
leaks. This is similar to the case shown in Fig. 10, but more ex-
treme. Again, it is due to a combination of low site sensitivity
and good monitoring conditions.

MANAGEMENT STRATEGIES FOR THE
MULTIPLE TANK OWNER
The analysis so far focuses on finding the optimal policy for
each individual tank. In our experience, the optimal policy for a
majority of the tanks is to replace them immediately with a
double-walled tank or eliminate the need for a tank altogether.
Because of the high capital costs involved, this may not be feas-
ible if an owner has a large number of tanks all needing immed-
iate replacement.
To develop a policy in this situation, the owner needs a set of
priorities to identify which tanks are to be replaced immediately
and which are to be kept for a few years and monitored. A set of
priorities can be developed by looking at the annual cost of delay-
ing replacement over the next, say, 5 yr. To identify this annual
cost, one must first identify the best monitoring option assuming
the tank is to be kept for 5 yr. Then take the total cost of using
that option and replacing the tank in 5 yr. Subtract out the total
cost of immediate replacement. That computation yields the total
additional expected cost that is incurred by keeping the tank for
5 yr. Finally, divide by five to get the average annual cost of
delay.
All the tanks in the inventory can be ranked according to an-
nual cost of delay. The optimal policy is to replace the tanks with
the highest annual cost of delay first. Fig. 11 shows such a rank-
ing for one project where the model was applied.
In Fig. 11, tanks are ranked by order of cost of delay which
is the order in which they should be replaced. The first eight tanks
have a positive cost of delay. That is, every year that their replace-
ment is delayed is expected to result in a net long-run cost to the
owner. Therefore, these tanks should be replaced immediately.
If this is not financially feasible, then the highest priority tanks
are those with the highest cost of delay.
The last three tanks in the figure have negative costs of delay.
That means that there is an expected net savings to the owner for
delaying their replacement. These tanks should be kept for a few
years before replacement in order to obtain the maximum eco-
nomic value from them.

CONCLUSIONS
This paper has presented a model of underground tank man-
agement economics that includes all the major costs associated
with owning an underground tank. It provides a practical ap-
proach to determining whether or not a tank should be kept, set-
ting replacement policies, choosing monitoring systems and man-
aging large numbers of tanks.
The analyses that we have done to date indicate that tanks fall
into three groups: those that should be replaced as soon as pos-
sible (any delay results in increased expected cost), those for
which there is an optimal replacement year sometime in the future
and those which can be economically monitored and maintained
for long periods of time. Of these, the largest group consists of
tanks that should be replaced as soon as possible. This group in-
cludes both new tanks and old tanks, the only difference being
the magnitude of the cost of delaying replacement. This group
UNDERGROUND TANKS
305
-------
FIICSTDEL REPL
-2.000
2.000 4.000 6.000
Annual Out of 0«lay. (
8.000
also tends to include smaller (up to 10,000 gal) rather than larger
tanks due to the economics of tank replacement.
REFERENCES

I. Petroleum Association for Conservation of the Canadian Environ-
ment, "Underground Tank Systems: Review of the Stale of the Art
and Guidelines," PACE Report No. 82-3. Ottawa, Canada, Feb.
1983.

2. Pope-Reid Associates, "Hazardous Waste Tank Failure Model: De-
scription of Methodology," report prepared for U.S. EPA, Office of
Solid Waste, Washington. DC, Jan. 1986.
Figure 11
Annual Costs of Delay of Replacement
(over first five years)
306 UNDERGROUND TANKS
-------
Minimization of Industrial Oil Wastes

Franco E. Godoy
Douglas L. Hazelwood
The Earth Technology Corporation
Alexandria, Virginia
ABSTRACT
Federal and state controls are expected to result in continued
increases in the cost of waste oil disposal. A significant portion of
the waste oil generated by industry could be eliminated through
management initiatives and technical improvements. Source
reduction, recycling and waste oil treatment are the most common
approaches for waste oil reduction. Relatively inexpensive off-
the-shelf equipment used in conjunction with routine oil analyses
can usually result in waste reductions of 50 to 90%. Savings
associated with comprehensive oil management programs typical-
ly result in a payback of investment costs with 1 to 3 yr.

INTRODUCTION
In 1980 used industrial oils accounted for 40% of the approx-
imately 480 million gal of used oil generated in the United States.

Table 1
Types of Industrial Oils
OIL TYPE
MAJOR CONSTITUENT
ADDITIVES
Hydraulic paraffinic base

Turbine Parrafinic base

Metalworking

Cutting t Grinding 1. Straight mineral
oil

2. Emulsions
Metal Forming

Quenching

Transformer

Refrigeration

Compressor

Natural Gas
Engine

Railway

Marine
3. Synthetic fluids

Straight mineral oil,
tallow, lard, palm oil

Paraffinic base

Straight mineral oils
paraffinic or naphthenic

Napthenic

Paraffinic base

Paraffinic base
Rust and
oxidation inhibitors

High temperature
oxidation inhibitors
Neat oil
Surface-active
emulsifing agent

Water

Wear reduction
additives

Oxidation Inhibitors
Naphthenic
None

Friction modifiers

Oxidation inhibitors

Rust inhibitors
Antifoam, Antiwear,
Detergent/Dispersant

Detergent/inhibitor
Package
Napthenic or paraffinic
base stack/high additive
Detergent/Dispersant
Antiwear, rust
inhibitors
detergent/dispersant
Most industrial waste oils have compositions similar to the virgin
oils but with impurities such as fine suspended metal particles,
dust, water and oxidation and decomposition products. Con-
taminants in used oils generally result from normal usage,
however, significant quantities may be introduced through poor
management practices.
Several states regulate waste oils as hazardous waste. In addi-
tion, federal regulations now being considered would mandate
more stringent controls over industrial waste oils throughout the
nation. These regulatory programs are having the impact of
significantly increasing the costs of managing waste oils. Many
facilities that had grown accustomed to being paid for their waste
oils are now faced with disposal and/or pickup fees for the same
materials.
This situation is expected to worsen in coming years as
regulatory controls increase while available disposal options are
reduced. Waste oil minimization represents the most attractive
approach to controlling costs and liabilities to reasonable levels.

CHARACTERISTICS OF INDUSTRIAL
WASTE OILS
Industrial generators use three major types of oil: straight
mineral oil, emulsified oils and synthetic oils. Table 1 presents a
summary of the major operations which generate waste oil, along
with information concerning make-up and additives.
The primary characteristics of an oil necessary to maintain ef-
fectiveness are:
• Viscosity sufficient to seal in the presence of internal system
pressures
• The ability to rapidly settle and separate insoluble solids
• Stability relative to oxidation to protect against rust and cor-
rosion
• Film strength sufficient to reduce friction and minimize wear
Even in a completely sealed system, forces combine to con-
taminate and otherwise compromise the critical operating proper-
ties of oils. The causes of oil deterioration and degradation are
highly system-specific. They commonly include bacteria, heat,
metal particles, oxidation, tramp oils and greases, solvents, dust
and water contamination. The impact of this degradation leads to
a change in the physical and chemical properties of the oil. The
critical properties often altered include:
• Flash point
• Corrosion rates
• Foaming
• Solids content
• Viscosity
• Gravity
WASTE MINIMIZATION 307
-------
• Water content
• Acid number

Table 2 presents specific causes of oil degradation and their im-
pacts.

Table 2
Causes and Impacts of Oil Degradation

CAUSE EFFECT
Excessive operating temperatures
Hater contamination
Metal particles (con wear
Dirt or other abrasive
contamination
Air entrainnent
Promotes oxidation
Increases viscosity
Causes sludging

pronotei oxidation
Promotes rusting
Forns oil-water emulsions
impairing lubrication

Promote fluid oxidation
Accelerates wear

Accelerates wear
Foaming and aeration
causes impairing lubricating
impaired properties
ESTABLISHING AN OIL
MANAGEMENT PROGRAM
Many industries have realized extensive benefits from the im-
plementation of waste oil minimization strategies. Recently, mini-
mization efforts have increased due to increased waste disposal
and raw material costs, federal legislation and liability exposure
associated with use and disposal of hazardous materials.
An initial condition necessary for a successful minimization
program involves management initiative. This initiative incor-
porates a perceptual shift from end-of-pipe disposal approach to
a more comprehensive, process-wide strategy of understanding oil
requirements and managing the oils to obtain a maximum life ex-
pectancy.
The first step to successfully minimizing waste oil is under-
standing the composition and performance requirements for the
oil and the reasons for its removal from service. Minimum accep-
table levels for critical oil properties, such as viscosity, water con-
tent and acid number are usually available from the manufac-
turers of the equipment in which the oil is used. These data should
be obtained and used as the benchmarks on which the minimiza-
tion program is based.
Once the required oil properties are known, the critical property
(or properties) that control the oil's useful life can be identified.
Inexpensive field tests can be performed for many oil properties
that will allow plotting of degradation and/or contaminant ac-
cumulation rates. These rates will determine the required oil pro-
cessing rates and allow proper sizing of oil management systems.
Often, the periodic careful examination of an oil while in service
will identify a previously unsuspected source of degradation that
can be readily corrected, such as a failed seal or a system hot spot.
Emulsified oils, such as metalworking fluids, arc usually
removed from service due to unacceptable levels of bacterial
growth resulting in rancidity, bad odors and skin irritations. By
evaluating system flow patterns and sources of contamination,
management changes or modifications to the fluid handling
system can often be identified that will significantly increase fluid
life. More than one machine shop has realized significant gains by
providing tobacco-chewing employees with spitoons and remov-
ing this common source of bacterial contamination! Removal of

308 WASTE MINIMIZATION
"dead spaces" in oil lines that promote bacterial growth also
helps to extend fluid life.
Several alternatives to conventional "drain and dump" ap-
proaches are available and should be considered:
• Source Reduction—the most attractive waste management op.
tion, focused on on-site changes which reduce or eliminate the
generation of waste oils. Source reduction can be divided into
three major areas: material substitution, technology modifica-
tions and managerial or procedural alterations.
• Recycling—to be considered after source reduction, has the
goal of recovering oil of a purity similar to the virgin oil for
reutilization. Recycling can be divided into three components:
direct reuse of the oil, removal of impurities/replenishment of
additives to obtain a reusable product and reclamation for a
separate use.
Implementation of an effective waste oil minimization program
will involve one or more of the above approaches. Unique oppor-
tunities for minimization will depend upon factors such as: the
type of application, specifications dictated by the application,
and the characteristics of the virgin and waste oil and their am-
menability to source reduction and recycle techniques.

SOURCE REDUCTION APPROACHES
Material Substitution
It may be possible to reduce waste generation rates by
substituting other materials for the oil currently employed. In-
dustrial oils function under a wide variety of conditions, acting as
lubricants, coolants, cleansers and sealants, or a combination of
these. There are several opportunities available for minimization
through material substitution, including: utilizing a higher grade
oil or substituting a synthetic oil or a synthetic petroleum-based
blend. Synthetic oils, which maintain essential characeristicssuch
as viscosity for a greater period of time, may be substituted for
petroleum-based stock. By maintaining their critical properties,
synthetic oils can reduce replacement requirements by a factor of
10. Alternately, petroleum-based stock may be blended with syn-
thetic oil to achieve comparable results.

Technology Modifications
Oil waste minimization may be realized through technological
modifications which decrease oil usage or extend its service life.
Opportunities include both process and equipment optimization
rather than operational changes. Reductions in oil waste genera-
tion through technological modifications may be achievable by
action such as: equipment modifications to reduce excessive
operating temperatures that promote oxidation and increase
viscosity or replacing equipment generating large quantities of oil
waste with more efficient equipment.

Operational and Managerial Changes
The procedural and organizational elements of an operation
have a great impact on potential waste generation. Waste oil
generation may be decreased through the following:
• Provide management initiatives to increase awareness of the
need and benefits of waste minimization
• Increase employee training to instill good operating practices
when working or servicing equipment to decrease waste oil
generation rates
• Expand programs involving preventive maintenance to avoid
waste generation through equipment failure, spills and leaks
• Increased segregation of waste oils to facilitate recycle and
reuse

These types of steps can easily be implemented and are general-
-------
ly very economical.
Maximum reductions in waste generation rates require im-
plementation of a comprehensive oil management program that
includes an oil inventory program and a quality control program.
Oil consumption tracking is essential to tailoring a recovery pro-
gram. Logs recording oil input and removal along with repairs
should be maintained for all equipment. Proper scheduling of oil
servicing will minimize waste generation. Periodic testing of the
oil to determine contaminant levels or critical properties will
facilitate oil waste minimization.

RECYCLING APPROACHES
There are several avenues open when considering implementa-
tion of a waste oil recycling program. Currently, three alternative
routes are available: waste oil recycling through on-site physical
reprocessing with some chemical replenishment, off-site re-
refining involving distillation and hydrofinishing and blending to
formulate either a specification or non-specification fuel.

Reprocessing
The feasibility for physical reprocessing must be determined on
a case-by-case basis. Properties such as viscosity are altered by
fluid or particulate contamination. There are several technologies
appropriate to most contaminant control situations:
• Gravity separation
• Filtration
• Distillation
• Centrifugation
• Magnetic separation
After physical reprocessing, additives depleted during use
and/or purification operations may be replenished.

Gravity Separation
Most industrial equipment provides for gravity separation of
oil and contaminants in a sump. Gravity settling is only effective
in removing undissolved water and large particular matter. Grav-
ity separation may be employed to segregate mixtures of the
following types:
• Liquids-liquids
• Liquids-solids
The principal mechanisms of gravity separation are sedimenta-
tion and coalescence. Sedimentation is defined as the removal of
suspended solid particles from a liquid stream by gravity settling.
The specific gravity of various solid contaminants allows them to
settle to the bottom of a settling chamber.
Coalescence is a method of separation of emulsified oils from
their water matrix . This operation involves passing waste fluids
through a series of coalescing plates to break the emulsion. Oil
droplets deposit on the plates and then separate by gravity.

Filtration
Filtration is the separation of solids from liquids achieved by
passing the liquid through a porous medium. The mechanism by
which contaminants are removed distinguishes the various filter
types:

• Mechanical filtration operates by passing the oil through a
medium which retains particles larger than the pore media.
Pleated paper, mesh screens and textile "socks" are examples
of common mechanical filters.
• Adsorbent filtration involves the collection of contaminant
molecules on the surface of a porous granular solid. Diatome-
ceous earth, fuller's earth and activated alumina are typical
filter media employed to retain dissolved contaminants such as
acids.
• Absorbent filtration occurs when free water is assimilated into
the fibers of the filter media. Although systems are not com-
monly designed to achieve contaminant removal by absorbent
filtration, its effects need to be considered when utilizing cellu-
lose filters, as the resulting swelling of media tends to reduce
void space and increase the pressure drop across the filter.
Other more exotic methods of filtration include membrane
separation processes such as ultrafiltration and reverse osmosis.
Reverse osmosis employs a semi-permeable membrane and a
pressure gradient of 200-600 lb/in.2. Ultrafiltration is used to
remove particles of 1 micron or smaller size, and filters are
typically porous membranes. Filtering pressures are approximate-
ly 100 lb/in.2.

Distillation
Distillation is an effective method for removing volatile con-
taminants such as dissolved water, acids and solvents from used
oils. Most oil distillation systems function by passing a thin film
of heated oil through a vacuum. The contaminants are removed
as a condensate stream for disposal. Temperatures of approx-
imately 75 °C and vacuums of 1 mm Hg are commonly employed.
Distillation is almost always preceded by mechanical filtration to
prevent system fouling.

Centrifugation
Centrifugation is similar to filtration except that centrifugal
force is used instead of a pressure gradient or gravity to separate
solids from a liquid. Centrifugation is also used to separate emul-
sions of two immiscible liquids. The centrifugal acceleration used
is many times that of gravity. Centrifugal force, commonly ex-
pressed as multiples of gravity, varies with the rotational speed
and with the radial distance from the center of rotation.

Magnetic Separation
A popular technique for clarifying metal-working fluids in-
volves magnetic separation. In this method iron-containing con-
taminants from cutting oils are separated by means of a magnetic
field.

Re-Refining
Waste oil re-refining is typically marketed on a custom recyc-
ling basis where a customer furnishes waste oil to be re-refined
and is returned oil blended to his specifications and needs. Re-
refining produces a high quality base stock suitable to compound-
ing and blending to both motor oil and industrial products. The
typical re-refining operation involves distillation and hydro-
finishing. The acid/clay re-refining process is no longer favored
due to costs and the difficulty of disposal of the acid/clay waste
products.
Table 3
Used Oil Fuel Specifications
CONSTITUENT/PROPERTY
ALLOWABLE LEVEL FOR BURNING
WITHOUT PERMIT
Arsenic
Cadmium
Cbromi uro
Lead
Total Halogens
Flash Point
5 ppm
2 ppm
10 ppm
100 ppm
1000 ppm
100 F maximum
WASTE MINIMIZATION 309
-------
Fuel Blending fuel oils to avoid exceeding source and air quality standards and
Waste oil, when properly treated, may be blended 10 formulate to minimise heat transfer surface fouling. In 1985, the U.S. EPA
a specification or non-specification fuel. Pretreatment to remove issued regulations to control the burning of used oils in non-
impurities such as metal-containing particulates and other industrial boilers. The allowable levels of contaminants are
materials is usually necessary. Waste oil is blended with cleaner presented in Table 3.
310 WASTE MINIMIZATION
-------
Developing a Minimization Strategy Effective
In Reducing Metal Hydroxide Sludge
Thomas F. Stanczyk
Recra Environmental, Inc.
Amherst, New York
ABSTRACT
With the influx of regulatory standards governing solid and
hazardous waste landfills, generators of metal hydroxide sludge
have been faced with rising disposal costs and stringent disposal
acceptance criteria dictating, in many cases, the need for chemical
stabilization and/or preferential pollutant removal.
Recent developments have been strategically applied to sludge
minimization programs. In most cases treatment performance
and the reduction of waste generation can be improved by imple-
menting process and control mechanisms effective in source
segregation and treatment, reagent substitution and sludge dry-
ing.
The components and trade-off considerations associated with
the development of a sludge minimization program are subject to
an array of industries concerned about waste generation, dis-
posal costs and risk reduction.

INTRODUCTION
Throughout the country waste hydroxide sludges comprise a
majority of the hazardous waste volume requiring ultimate land
disposal. Even though these sludges are generated from similar
plant operations and wastewater treatment systems, the resulting
waste sludges commonly vary in terms of characteristics and
hazard potential. This variability generally is dictated by feed-
stock usage, materials management practices and the perfor-
mance efficiency of treatment operations resulting in primary
sludge by-products. The commonality associated with metal hy-
droxide sludge is the mode of disposal.
Federal and state regulatory standards dictating land disposal
restrictions and the acceptance criteria dictating handling and
placement have had significant impacts on the design perfor-
mance standards of landfills. The acceptance criteria for land
disposal have, in many cases, required chemical stabilization and
preferential pollutant removal prior to direct placement and dis-
posal. These criteria will increase both pretreatment require-
ments and disposal costs.
Recognizing that many wastewater treatment systems were de-
signed to meet specified effluent discharge limitations, greater
emphasis recently has been placed on modifications to existing
systems to improve effluent quality and reduce both sludge vol-
ume and risk potential. In addition, with short- and long-term
planning, industries find that reassessing existing treatment prac-
tices can effectively improve metal reclamation, purification and
reuse.
The components associated with strategic planning are the sub-
ject of this paper. Where appropriate, the findings of applic-
able case studies have been incorporated.
PLANNING
The management of any industry invariably is motivated to
have control over plant policies and practices influencing pro-
duction and related costs. Metal hydroxide sludges generated as
by-products of mandated treatment systems are impacting com-
pany profit margins and the quantifiable concerns over future
liabilities. These concerns are being magnified by the realization
of new performance standards and landfill restrictions, thus plac-
ing greater emphasis on defining variables impacting pollutant
mobility and environmental risk.
The inadequacies and shortfalls of existing sludge manage-
ment technologies require re-evaluation of source segregation,
treatment and process controls governing feedstock usage and
handling. An effective strategy to minimize sludge generation
can be developed and implemented, but its success can only be
weighed after careful planning. Planning endeavors must estab-
lish as goals the need for detoxification, reclamation, purifica-
tion and reuse, volume reduction, chemical stabilization and safe
disposal. The necessary plan components include waste charac-
terization, degree of concern, treatment evaluation, disposal
assessment and economic/environmental trade-off considera-
tions.

WASTE EVALUATION AND ASSESSMENT
Waste characterizations should be the initial step in the ration-
ale development of a waste minimization strategy plan. Metal
hydroxide sludges should be characterized for chemical content
as well as properties influencing the method of handling, pack-
aging and disposal. Fig. 1 represents a logical sequence of re-
quirements encompassing both analytical testing and source re-
view.
The characterization program is an important element of the
strategy plan. As land disposal restrictions become increasingly
stringent, generators will be required to verify the absence/pres-
ence of hazardous substances. This issue is of prime concern for
metal hydroxide sludges which have the potential for concen-
trating organic as well as inorganic constituents. A typical range
of chemical content and physical properties for metal hydroxide
wastewater treatment sludge byproducts is presented in Table 1.
DEGREE OF CONCERN
Minimization strategies must address the reduction of waste
volume as well as the reduction and potential elimination of waste
constituent mobility and hazard potential. The prescribed dis-
posal acceptance criteria require hydroxide sludges to meet stand-
ards limiting free liquid generation, solubility, acid generation,
WASTE MINIMIZATION 311
-------
total volatile organic content, concentration of potentially reac-
tive constituents and teachable inorganic and organic content.
(VALUAHM R(OUltiM«U

o «C»A H4t*rd

blllly
0 DOT
o DUpo
(wf
I ••(.ordi
bU it/
utml •obllltr

Poi*«H*l

Itllf
nbiHii
Figure 1
Waste Evaluation and Data Assessment
Table 1
Range of Chemical Content and Physical Properties
Constituent Characteristic
Cadmium
Hexavalent chromium
Nickel
Cyanide (completed)
Cyanide (total)
Copper
Zinc
Aluminum
Arsenic
Barium
Mercury
Silver
Lead
Selenium
Tin
Chromium (total)
Titanium
Iron
Phenol
Volatile Organics
Fluoride
Ammonia
Water
Phosphates
Detergents
Alkalinity
T.O.C.
Sulfide
Specific gravity
pH
Cone. Range (%)
0-50
0-50
0-50
0-50

0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
0-50
1.0-1.5
5-14 units
Generators must be cognizant of disposal acceptance criteria
dictating pretreatment that may entail chemical stabilization,
physical solidification, chemical detoxification and/or preferen-
tial pollutant removal. When assessing degree of concern, the
sludge contents must be evaluated for factors influencing pollu-
tant mobility (i.e., excess alkalinity, in situ precipitation, oxida-
tion, alkaline hydrolysis and volatilization). In terms of chem-
ical compatibility, the chemical constituents within hydroxide
sludges should be assessed relative to potential waste: waste,
waste: liner and/or waste: leachate interaction. This assessment
will reduce the potential for creating a waste which will be more
difficult and costly to manage within the constraints of a land-
fill operation.

SOURCE SEGREGATION AND TREATMENT
Wastewater treatment traditionally has involved one or more
unit operations designed, alone or in combination, to remove
and/or reduce the concentration of toxic constituents prior to
discharge. The resulting by-products generally include a primary
sludge residue requiring further dewatcring and possibly stabil-
ization before the material can be managed directly.
Taking into account the potential chemical variability and
sludge complexity, it is beneficial to evaluate treatment alterna-
tives for wastewaters generated at the source prior to equaliza-
tion. By categorizing and segregating the wastewaters at the
source, it is conceivable that alternative treatment technologies
can be cost-effectively implemented. A typical treatment logic
chart employing the concepts of source segregation and treatment
of categorical waste receipts is depicted in Fig. 2.
It is important to note that the generator can successfully mod-
ify the properties and chemical content of the resulting metal
hydroxide sludge. The stated concepts have been utilized at indi-
vidual plant and commercial treatment facilities. Benefits recog-
nized from the concepts of source segregation include:
• Improved treatment performance and dewatering efficiency
• Improved options for feasible metal reclamation and reuse
• Preferential degradation/detoxification of hazardous constit-
uents using minimum reagent requirements
• Preferential removal of inorganic as well as organic constit-
uents resulting in sludge by-products having the potential for
delisting as a hazardous waste
• Reduction in sludge volume

CHEMICAL SUBSTITUTION
Wastewaters containing soluble metallic, nonmetallic and
organic constituents generally rely on chemical treatment to im-
prove effluent quality. Equalization followed by alkaline precip-
itation and sludge conditioning is a common sequence of unit
operations. The problem identified with most systems centers
around the desire to improve effluent quality with little regard
for the properties of the sludge by-product. Excessive reagent
usage, poor chemical immobilization and poor dewatering due to
organic interferences are among the problems encountered when
the chemistry of a treatment system and the sludge by-product
are ignored.
The concepts minimizing sludge generation should follow a
treatment logic exemplified in Fig. 3.
The key aspects of the process options depicted in Fig. 3 in-
clude:

• The ability to selectively remove one or more constituents us-
ing adsorption, membrane and/or ion exchange media prior to
treating wastewater with chemical reagents
• The ability to substitute reagents to remove and/or reduce the
concentration of soluble constituents and thus:
-reduce the volume of sludge by-product
-reduce the constraints dictating treatment performance effic-
iency
-optimize chemical immobilization through chemical interac-
tion and coprecipitation
-improve liquid-solid phase separation and dewatering effic-
iency
312 WASTE MINIMIZATION
-------
Product Process Material
Substitution Change Re-use
Segregate on the
In-Plant Basis of Chemistry,,
Generation Compatibility and
Physical Properties
Aqueous, Metal lies
Aqueous, Acids

Aqueous, Pickle
Liquors
Aqueous,
Ammonia-Based
Aqueous, Alkali
Solutions
Aqueous, Containing
Soluble Organlcs
Aqueous, Neutral
Solution
Aqueous, Oxldlzers
Aqueous, Requiring
Pretreatment (CN.S)
Metal Reclamation

Chemical Reduction
(Optional)
Ammonia Recovery,
Removal or Destruction
Liquid/Solid and
011 Separation
If Necessary
Volatile Organics
- stripping
- devolatilization
.Extractable Organics
- recovery
- select pollutant removal
Toxic Organics
- detoxification
- removal
Recovery or
Reuse Potential
Chemical Oxidation

Optimize Volume
Reduction Options
- phase separation
- dewatering
sludge drying
V
Sludge
Treatment
^Equalization Primary
' A Treatment
Assess Assess
Factors Factors
Influencing Influencing
Performance Reagent Usage,
Pollutant Removal
and By-Product
Waste
Generation

Figure 2
Wastewater Treatment, Viable Segregation and Treatment Schemes
SELECTIVE REMOVAL

ADSORPTION
ION EXCHANGE
MEMBRANE
Figure 3
Chemical Precipitation
SLUDGE TREATMENT
Treatment processes for residual metal hydroxide sludge re-
quire an understanding of sludge chemistry. Technologies which
should be included in the strategic planning and evaluation efforts
include:

• Chemical Stabilization/Solidification
• Mechanical Dewatering
• Extraction/Purification
• Drying/Volatile Stripping
The technology selected for implementation should effectively
reduce waste volume and residual chemical content mobility.
Each strategy should account for the considerations described
herein.
MECHANICAL DEWATERING
Dewatering metal hydroxide sludges has been employed by in-
dustries using one or more versions of the following filtration
options:
• Belt filter
• Vacuum
• Centrifugation
• Pressure
Each option can be applied to metal hydroxide sludges. The
performance efficiency generally will be dictated by the nature of
the waste, the degree of chemical conditioning, the frequency of
use, the in-house maintenance practices and the desired end-
products.
The solids generated with dewatering may range in total solid
content from a low of 15-20% by weight to a high of 50-60%
by weight. Organics in the form of soluble or emulsified oils
(i.e., machine and/or cutting fluids) could interfere with the per-
formance efficiency and would require chemical conditioning to
allow for proper particle size.
In some cases it may be beneficial to dewater high solid con-
tent residues, generated at the source, prior to treatment.
Fig. 4 is a cross-sectional view of a fiber cloth belt-type rotary
vacuum filter.
To optimize the performance of vacuum filters, one should
consider:

• Type and degree of chemical conditioning
• Specific resistance
• Cake thickness and weight
• Compressability
• Form time
• Organic content
WASTE MINIMIZATION 313
-------
EVAPORATION
MEAT FLUX
Figure 4
Cross-Sectional View of a Fiber Cloth Belt-Type Rotary Vacuum Filter
Figure 5
Evaporation Mechanisms
• Particle size
• Effluent and cake characteristics
• Cycle time

EXTRACTION/PURIFICATION
Recent applications of extraction technologies have been mod-
ified to allow for the direct treatment of sludges. The process
basically entails:
• Extraction with a solvent or aqueous media adjusted for pH
and maximum solubility of the constituents of concern
• Removal of constituents from the extraction media
• Recovery of the extraction fluid

This process sequence is being utilized to recover feedstock con-
stituents as well as reduce the concentration of hazardous con-
stituents preventing metal purification and/or direct land disposal.
Copper and nickel are two metals which commonly are recovered
by an extraction process.
With rising disposal costs industries are finding that metal pur-
ification and recovery techniques are economically viable. The
feasibility of reclaiming metals from sludge improves with source
segregation, treatment and feedstock control. Techniques war-
ranting consideration in the strategy plan include electrolytic
deposition, leaching/extraction, electrowinning, precipitation,
oxidation and ion exchange.

SLUDGE DRYING/VOLATILE STRIPPING
Unlike the removal of water using mechanical dewatering unit
operations, drying is a process whereby moisture entrained with-
in a sludge is evaporated through heating, producing a relatively
dry solid. This technology has been utilized successfully in indus-
try for over 50 yr and recently has become popular for residues
typically generated from mechanical dewatering. Dewatering
units generating filter cakes at a 25-35% by weight total solid
content can be further dewatered, producing dry solids display-
ing 85-95% total solids.
The application of drying technology to waste sludges should
be approached as a science. Each dryer will react differently, and
its performance will reflect the processing variances dictated by
material particle size, capillary voids, retention mixing action and
sludge makeup. In most applications, the drying step is preceded
by a mechanical dewatering system (i.e., vacuum or pressure).
Fig. 5 illustrates the evaporation mechanisms distinguishing dif-
fusion and capillary flow.
Water generally is present in hydroxide sludges on the surface
of conditioning additives, but it also can be mechanically bound
and chemically hydrated. Residual levels of solvents generally are
present in a similar manner. Therefore, the drying process must
effectively remove the liquid found in the interstices of sob'ds us-
ing diffusion and capillary action.
The drying process follows a logical sequence of steps entailing:

• Adding heat to the sludge optimum operating conditions
• Evaporating the moisture at a rate equating to the influent
moisture level
• Attaining a critical end point where the drying rate begins to
fall
The sludge characteristics during the drying cycle will undergo
changes varying from free boiling of liquid and plastic shearing
to a wet granular material and finally to a dry granular material.
Drying over continued periods of time can yield a free-flowing,
dry, powdery residue.
•IUDI tk O'-tf
Figure 6
Drying Operation with a Filter Press
314 WASTE MINIMIZATION
-------
Agitation is a critical component of systems on the market to-
day. The proper mix will have a significant impact on drying
rates. With sludge containing volatile impurities, the agitation
factor becomes increasingly important to enhance pollutant re-
moval and bring solid particles into contact with the heat-transfer
surface.
There are a number of drying units currently available and
applicable to both metal hydroxide sludges and sludges contain-
ing high volatile organics. Applications range from cored shaft/
blades for extended heat transfer to thin film, spray drying and
high speed shear relying on external radiant heat sources. In most
cases, the units are simplistic in terms of material loading, pro-
cessing, unloading and emission/dust control. Some of the units
discharge the resulting powder into bags minimizing employee
and work place exposure.
Fig. 6 is a conceptual illustration of a drying operation em-
ployed with a filter press.
It was previously mentioned that drying is a science that re-
quires the generation to fully evaluate the waste contents and
drying options. The technology can have wide applications yield-
ing a number of benefits if the chemistry of the sludge and the
factors influencing the mobility of water and/or volatile constit-
uents are well understood. Consider the following benefits achiev-
able with drying:
• Filter cakes can be reduced an additional 60-80% in volume
• The dried residues will display physical properties amenable
to direct disposal via landfill
• Purification techniques requiring metal reclamation and reuse
can be made feasible, thus minimizing treatment requirements
• The dried residues can be treated in a manner that will remove
volatile organic constituents to levels that will meet the U.S.
EPA's land disposal standards including the TCLP perfor-
mance standards
• The metal constituents found in the sludges are effectively
immobilized and amenable to potential delisting of hazardous
characteristics
• Disposal costs are significantly reduced by as much as 60-70%
• Pay-out analysis shows capital investment recovery within
relatively short periods of time
• The risks associated with large volumes of waste being shipped
off-site will be significantly reduced
• Off-spec charges and the potential need for solidification are
eliminated

SOLIDIFICATION/STABILIZATION
Metal hydroxide sludges can be effectively solidified by add-
ing one or more chemical reagents to cause a change in physical
strength and chemical mobility. A number of conventional tech-
nologies are applicable to metal hydroxide sludges. Engineers
must know how the reagents being applied interact with the
sludge contents and influence pollutant mobility in addition to
how they improve physical strength and load bearing capacity.
In many cases, pozzolanic treatment of metal hydroxide sludges
results in increased rates of metal and organic teachability. Waste
contaminants such as mercury, arsenic and phenol typically are
more mobile under alkaline conditions.
Solidification/stabilization is a science. The chemistry of the
sludge has to be well understood to ensure effective immobiliza-
tion. Organics, specifically low molecular weight solvents, will
interfere with hydration reactions occurring with alkaline reagent
treatment techniques. Residual solvents and oils can serve as car-
riers for other inorganic and organic constituents, substantially
increasing rates of leachability. While many sludges may appear
to be physically solidified, there has been no impact on chemical
immobilization.
It is also important to consider the operating variables of the
desired end product. Reagent selection and dosage can be im-
pacted by the amount of water, impurities in the sludge and time
allowed for curing. Solidifying a waste sludge after dewatering
can significantly reduce reagent requirements resulting in volume
reduction instead of volume increase. Extending the curing peri-
ods can reduce reagent requirements. Dewatering performed in
conjunction with drying can totally eliminate the need for addi-
tional stabilization.
Overall, stabilization technologies can prove beneficial. How-
ever, it is imperative that the chemistry of the sludge be reviewed
before selecting a process.

CONCLUSION
Industries have found practical solutions to the management
and minimization of sludge generation. Metal hydroxide sludges
can be effectively reduced by developing a waste strategy plan
that accounts for source generation, segregation, treatment, feed-
stock usage and material reclamation.
With greater emphasis being placed on the restrictions on di-
rect land disposal of hazardous wastes, industry must reassess
existing practices in a manner that will result in short- and long-
term options for metal reclamation/reuse, sludge reduction/de-
toxification and safe disposal within the constraints of pending
regulatory acceptance criteria.
Overall, there are a number of practical solutions to effectively
reduce disposal costs that are based on a clear understanding
of waste chemistry and its application to conventional engineer-
ing principals.
WASTE MINIMIZATION 315
-------
Small Generator Cooperative Effects
Economical Recycling
M.F. Malotke
Tencon Associates
Milford, Ohio
ABSTRACT
In this paper, the author discusses the method used to evaluate
wastes generated by a large group of Cincinnati automobile deal-
erships for whom individual recycling of spent solvents was not
economical. The analysis of the waste material, the previous dis-
position, the group strategy for collection, handling, processing
and reuse and the costs are described.

INTRODUCTION
Recycling companies in Southern Ohio are located near Day-
ton, over 50 mi from northern Cincinnati. Because of the dis-
tance, recyclers were reluctant to set up a "milk-run" collection
system for businesses such as auto dealerships which averaged
one to two drums of hazardous waste per month.
Through the Cincinnati Chamber of Commerce Solid Waste
Subcommittee, TENCON Associates was asked to put together a
program to help small companies meet RCRA's new "small gen-
erator" regulations and simultaneously promote recycling of
valuable resources. TENCON, in conjunction with the Greater
Cincinnati Auto Dealership Association, approached the deal-
ers and suggested the formation of a uniform laboratory/label-
ing/handling system that would reduce the overall cost of dis-
posal/recycling and would, at the same time, allow expeditious
handling of these small quantities of waste solvents throughout
the Cincinnati area.

ANALYSIS/WASTE PROFILE
The initial approach was an environmental audit of seven auto-
mobile dealerships selected at random to determine common
wastes. Samples were taken and analyzed. The results were tabu-
lated to determine what variations existed in the waste streams.
The surprising uniformity of results, as shown in the list below,
reflects a relatively uniform similarity of waste streams.
The dealers were discovered to have several recyclable waste
streams:
• Lacquer thinner with paint sludge
• Degreasing solvent (mineral spirits) with high lead above EP
Tox limitations
• Spent carburetor cleaner (methylene chloride)
• Waste gasoline
• Waste oil
A given dealership generation averaged:
• One to two drums of lacquer thinner/body shop/mo
• One drum degreaser/two mo/dealership
• One 5-gal can carburetor cleaner/6 mo/dealership
• One drum gasoline/6 mo/dealership
• 250 gal of waste oil/dealership/mo
The waste oil was established separately by a waste oil dealer
from the Dayton area. Since the quantity of oil was sufficient in
any given month to allow the collection firm to plan a weekly
routing in quadrants of the city, no "small quantity" handling
fees were necessary.

HAZARDOUS WASTE HANDLING
Data on the remaining drummed material, identified as haz-
ardous as a result of its lead content, were computerized for
each dealer. The initial laboratory work allowed these materials
to be identified as Hazardous Flammable Liquid N.O.S. This
determination minimized the labeling and manifesting variations.
Each dealer is given directions for labeling, dating and stor-
ing his wastes. His initial amount of waste is entered into the com-
puter file for his dealership. Subsequently, the computer system
generates a list of dealerships along a given route which have
not had a pick-up for more than 4 wk. A telephone call sheet is
established, and a clerk telephones the selected dealers approx-
imately 1 wk ahead of the recycler's scheduled visit.
Since the recycler's truck is usually two-thirds full from large
generators, the recycler is glad to make three to four additional
stops per trip to fill it up. Approximately monthly, they devote
one day to dealerships which are not on main routes and which
by that time have accumulated two or three drums each. This
scheduling has eliminated any pick-up or handling charge on the
dealership's part. Fees for the computer and clerical time are ap-
proximately S35 annually.

RECYCLING PROCESSES
Lacquer Thinner
The lacquer thinner is stripped/distilled for volatile materials
recovery; the sludge is concentrated. The thinner is evaluated for
quality and then is sold for several applications:
• Auto-body undercoating shops for use in thinning and gun
cleaning
• Small steel fabricators for use in priming (some color allow-
able)
Lacquer thinner which does not meet minimum quality specif-
ications is separated and blended into a fuel once the paint sludge
is removed.
Degreascr
The degreasing solvent, basically mineral spirits, is filtered
where feasible and reblended with new material to be sold as a
lower grade mineral spirit. Several of the dealerships purchase
this lower grade of mineral spirits for reuse in degreasing. By
blending at a 50/50 ratio, the material is kept "active" enough
to perform well while reducing costs.
316 WASTE MINIMIZATION
-------
The lead sludge, removed either through filtration or distilla-
tion, is added to the paint sludge for solidification and disposal.
Solvent which is too dirty is blended into the asphalt fuels pro-
gram.

Carburetor Cleaner
Carburetor cleaner, the other common solvent waste from
auto dealerships, is kept in open buckets for the entire useage
period, so that little solvent remains. This low percentage of sol-
vent and the small quantity generated per dealership (5 gal/6 mo)
indicate that the best alternative for this material at this time is
disposal. Although disposal costs are higher than recycle costs,
handling costs are eliminated since the recycler picks up the ma-
terial with the solvent drums. They consolidate the material and
re-manifest it to a chlorinated waste disposal service as full
drums.

Gasoline
The spent gasoline, usually contaminated with water or sugar,
is blended into the fuels program. In most cases, gasoline that is
sent from a dealership is of such poor quality that reuse in any
vehicle would be suspect, especially when the full fuel value can
be recovered in an asphalt fuels program.

PROGRAM COSTS
The program costs a dealership an average of $350/yr. Larger
dealerships average $550/yr. This cost includes a minimal $35
annual fee to cover the computer scheduling. Degreaser, on an
exchange basis, costs $5/drum over the non-bulk raw material
cost; lacquer thinner, with 8-10 in. of paint sludge, averages $60
per drum. Contaminated gasoline is averaging $35 per drum for
disposal.
This cost is offset by the money received from the waste oil,
which averages $150 to $250/yr. The total cost to the dealer then
is $100 to $200/yr. In addition, the dealerships have met all fed-
eral and state regulations with regard to U.S. EPA ID number,
have maximized the recycling of their waste materials and have
minimized any economic impact on themselves. An evaluation of
the cost to handle these waste materials without the association
indicates a potential increase of $600/dealer/yr. In addition, the
use of computer scheduling allows a yearly print-out of the wastes
handled, which can go directly on the annual generator reports.
CONCLUSION
At this time, the Chamber of Commerce, TENCON Asso-
ciates and the Cincinnati Auto Dealers Association feel that the
return from the "pooled" effort is well worth the cost. It is esti-
mated that some 88,000 gal of solvent/sludge material and some
265,000 gal of waste oil will be recycled annually. The challenge
now is to broaden the scheduling/laboratory assistance to other
groups.
WASTE MINIMIZATION 317
-------
Financial Analysis of Waste Management Alternatives
Richard W. Mac Lean
Corporate Environmental Programs
General Electric
Fairfield, Connecticut
ABSTRACT
Sound business decisions on waste program management can-
not be made without the consideration of long-term liability
issues. History has shown that today's inexpensive disposal op-
tion can be far more expensive in the long-term than the initially
more expensive route. It is difficult, however, to predict precisely
the liabilities which may be incurred in the future for specific
waste management practices. The problem is two-fold. First,
there is very little information available on which to evaluate the
technical and legal issues that impact future liability costs. Se-
cond, existing information has not been structured in an organ-
ized formal from which costs can be derived.
Recognizing this, General Electric has developed a workbook
and companion computer software program to help their divi-
sions estimate these future liability costs. The program presently
being implemented, while based on technology considerations, is
structured to reflect both corporate policy and sound en-
vironmental practice. Business managers can readily understand
and act on the output from the analysis which is presented. Thus,
environmental engineers are provided with a tool that allows them
to address in financial terms the liability and technical issues that
otherwise would be extremely difficult to quantify to manage-
ment.
INTRODUCTION
Business executives are faced almost daily with reminders of the
enormous cost, liability and public concerns for the generation
and disposal of waste. How do these individuals put their com-
pany on a sensible track to reduce waste? How can they do that is
right, both from an environmental and a cost-effectiveness stand-
point? Glowing policy statements on the merits of waste reduc-
tion may appear to enhance the company's image but do almost
nothing to attack the real issues. The middle managers who are in
the best position to take action do not need general policy state-
ments; they need the knowledge and tools to implement specific
programs. Agency hazardous waste regulations are sometimes
specific, but they provide little assistance for organizing a waste
reduction program. A policy based on "letter of the law" com-
pliance may lessen short-term liabilities but does very little to ad-
dress long-term issues or to establish cost-effective leadership.
Because of these difficulties some companies have established
"5-year plans" that have overall reduction targets for waste mini-
mization. Others have opted for waste end taxes as a policy tool
to force business components to reduce waste generation. For
some companies, policies such as these can be very effective,
especially for small firms or for large homogeneous firms, if they
have internal environmental technical resources. For large, highly
decentralized and diversified firms, these policy approaches are
extremely difficult to administer. For example, how do you main-
tain fairness between corporate divisions; differentiate among
treatment options and waste toxicity?; shift resources to the most
significant long-term needs?; and provide environmental
technical assistance at those plants where none may be available?

CORPORATE POLICY
Considering all of the above, Genera) Electric elected a funda-
mentally different approach. First, the policy had to be specific,
technically defensible and capable of being implemented at a
plant level. Second, the policy guidelines had to be written in
language that engineers (not necessarily environmental engineers)
and managers in highly diversified businesses could understand
and implement. Simply stated, the policy is this:

"Business components are to determine both short- and
long-term costs, including future liabilities, when
evaluating waste management plans. Those programs that
reduce all liabilities and are the best according to estab-
lished financial procedures are to be implemented."

The theory behind this policy is straightforward. If the total
costs are known, managers can implement better, more cost-
effective long-term waste reduction programs. However, from a
corporate responsibility standpoint, if long-term liabilities are
considered, the best, most environmentally sound decisions will
be made today, and the company and the community will win in
the long-term.
So much for theory-. How do you turn this into workable im-
plementation guidelines? General Electric's approach has been to
develop a financial analysis workbook and companion computer
software program which will allow plant personnel to determine
total waste management costs including future liability considera-
tions. The mechanics of this are discussed in the next sections of
this paper. It is important to mention that the techniques, white
they are based on a foundation of technology, do involve a signif-
icant amount of judgment in establishing the "weighing factors"
used in the calculations. Depending on the values selected, certain
waste management practices can turn out significantly better than
others. It is, thus, possible to establish sound environmental prac-
tices by means of a financial technique simply by establishing fac-
tors that drive long-term liability costs to unacceptable present
value costs for less desirable environmental practices. In General
Electric's program corporate establishes these factors and, in so
doing, establishes environmental policy.
Concurrent with this policy are two other activities that, in
total, define General Electric's corporate program for hazardous
waste reduction. First, beginning in 1987 all commercial treat-
ment, storage and disposal firms will be reviewed and qualified by
corporate to do business with the company. This review will con-
318
WASTE MINIMIZATION
-------
sist of a financial screening in addition to an on-site technical in-
spection conducted by outside consultants or by a team of in-
ternally trained inspectors drawn from divisions. The data obtain-
ed from these inspections, as will be shown later, will be used to
estimate the liabilities associated with these commercial firms.
Second, an internal audit program (called the "PULSE" self-
appraisal program) will be used to insure that divisions are com-
plying with internal policies and applicable local and federal
requirements for waste management.

TOTAL COSTS FOR WASTE MANAGEMENT
For the purposes of financial analysis, waste management costs
can be divided into two categories: (1) direct costs and (2) future
liabilities. Direct costs include the out-of-pocket costs routinely
associated with waste management and disposal. Examples of
direct costs include:
• Payments to vendors, including waste-end taxes
• Investment in waste management equipment
• Waste collection and transportation costs
• Raw material and labor costs associated with on-site waste
treatment, storage and disposal
• Production costs affected by waste management decisions
(e.g., raw material usage, energy or productivity penalty)
These costs usually are readily identifiable by standard finan-
cial and engineering techniques. In addition, there are numerous
reference handbooks on waste minimization and disposal that can
help engineers identify possible options.1'5 Waste reduction pro-
grams that go back further into the manufacturing operations
generally require the expertise of process engineers and are often
so site-specific that "universal" handbooks are of limited value.
Future commercial treatment and disposal costs and especially
long-term liability costs are much more difficult to quantify.6 For
this reason business decisions within industry have seldom includ-
ed more than current direct costs for waste management. There is,
in fact, almost no information in the literature that provides
guidance on how to estimate these future costs. In the case of off-
site disposal costs, factors such as changing regulations can signif-
icantly impact capacity availability and treatment options and, as
a result, disposal costs. Standard indices of inflation are of little
value to estimate future off-site disposal costs. In the case of
liability costs, what data do exist tend to report only cost informa-
tion for remedial action costs at specific disposal sites.7 Specifi-
cally, these liability costs can include:
• Response costs and natural resource damages incurred for
which the waste generator becomes liable under CERCLA
• Corrective action costs under RCRA at company-owned (on-
site) treatment, storage or disposal facilities
• Liabilities arising out of third-party lawsuits which seek com-
pensation for bodily injury and/or property damage

Direct Costs
The first step in analyzing the waste management program for a
specific manufacturing operation is to determine the present
"direct costs for the existing waste management practices. In order
to do this, the types of wastes that are generated and those that
result from intermediate treatment processes must be identified
and their quantities and constituents characterized.
The next step is to determine all the technically feasible alter-
natives to these current waste generation, treatment, storage and
disposal practices. Particular focus should be placed on process
changes that would eliminate the generation of waste entirely.
The present direct costs for these alternatives are then calculated.
Here again, there are numerous references that can assist manu-
facturing personnel in doing these steps, since this is standard
textbook environmental and process engineering.
At this stage in the analysis, it is worthwhile to assemble the in-
formation into an organized format that will lend itself to the
financial analysis that will be described later. One approach is il-
lustrated in Figs. 1, 2 and 3.
Current Wastes at point of generation, after subse-
quent management steps, when combined with other
wastes, and at "end of pipe."
Alternative Quantities or Types of Wastes that
would result from proposed waste management
changes such as source reduction or more broadly,
waste minimization.
Current Waste Management Steps including genera-
tion and any on-site storage, treatment or reclama-
tion which would result in the generation of a waste.
Boxes also represent any on-site disposal or reuse,
and any off-site management steps.
Alternative Waste Management Steps to those
designated by boxes, for proposed changes to gen-
eration and management based on quantity or type
of waste and on different technology approaches to
generation or management.

Figure 1
Symbols for the Waste Flow Diagrams

EXAMPLE FLOW DIAGRAM FOR ILECTROf LATTNG FACILITY WASTI MANAGEMENT:
T(
(lOOMOn)

IT;
DbdurMU
Auno^Kv*

T|
TTMlX«nl Mid
SMUT* Undltl

W, Ciuok lummun
W, Addk tUnwnun (chroM*-bMrlnt)
W4 CotlKUd Off-tuw (chn>M-tairki|)
W, Slud* (M ehromlua)
W, ftnUBr TraMtd Wuuwur (no chromkM)
Wt ScnibtMr Bio-Awn (ehromt-bMrlni)
W,, OtochutMtri* Purtlted Ouu
Figure 2
Current Waste Practices
Addle lUMmun (no chromium)
DUchiriubb TrttMd EIDutn
Sludgi (no chromium)

Non-huirdoui SubiUud fiudp
Figure 3
Alternative Waste Practices
WASTE MINIMIZATION 319
-------
These waste flow diagrams are especially useful in analyzing
complex processes since costs for each step must be identified.
Note that both on-site and off-site management practices are in-
cluded as well as all on-site generation practices. One must ac-
count for all costs because some treatment options such as recycl-
ing, may have positive cash flow.
Generally this is as far as most waste management reviews go:
the present cost for the existing program is compared against the
alternatives and the lowest present cost option wins. An
enlightened manager or an environmental engineer may "in their
heart" know that an alternative practice is better than current
practice, but the cost is greater. In the absence of a company
policy, lower costs options usually win. It is at this traditional
"end" point that our analysis begins.
Liability Costs
It is possible to view future liability costs as the price of an "in-
surance policy." An insurance company determines the price for
your life insurance policy (not knowing when or how you will die)
based on well-established actuarial tables. The General Electric
financial analysis workbook serves a similar function; it estimates
liability costs for specific types of facilities based on past history
and corporate policy. Just like the life insurance costs for a cer-
tain occupation, say a stuntman, are high, the liability costs of
certain types of disposal activities are high, even though the cur-
rent price may be low.
Actuarial tables on which insurance costs are determined are
very precise statistical tools since they are based on literally
millions of bits of information spanning decades of data.
Remedial costs, liability and regulatory issues on which one can
evaluate case histories are few and span only a decade or so. The
net result is that even the most complex, sophisticated technical
and legal analysis of a specific waste management program will
not guarantee that the cost estimates are an accurate predication
of what will happen in the future. With time, this situation will
improve, but for now a relatively straightforward analysis is all
that can be practically done. Remember also that the desired end
product of the analysis is the implementation of sound en-
vironmental policy as will be shown later.
From an overall standpoint, the two key determining factors in
estimating liability are the type of treatment, storage or disposal
activity and the type of waste involved. Furthermore, a financial
analysis must include two measurements: liability dollar amount
and the time at which this future expenditure is incurred. Both of
these measurements are dependent on the type of facility and the
nature of the waste handled. For the sake of simplicity, however,
these parameters are separate for financial analysis according to
the matrix presented in Table I. As illustrated it is assumed that
liabilities are incurred sooner for those wastes that are highly toxic
and mobile. In addition, it is assumed that the dollar amount
depends on key facility characteristics, specifically proximity to
local residents and water supplies, prior history of problems and
the treatment or disposal technology employed at the facility.
Table 1
Future Liability Bails
Factor
Waste
Facility
Parameter

• Toxicity
• Mobility
1 Technology
Population
Water Supply
1 Leak History
Input

Waste
Constituents
Site
Inspection
Program
Output
Time
Amount
Since most of the available information on liability cost is based
on remedial cleanup issues at land disposal facilities, the work-
book uses this as the base case. All other cases are extrapolated
from this "average landfill." Using this base case, the long-term
"excess" liability is estimated to be $350/ton. In other words, at
some time in the future a company will face an additional cost of
$350 for each ton of waste disposed in a typical landfill to pay for
the three liability issues listed previously. This dollar amount was
based on analysis of hazardous waste landfills meeting Part 264
requirements.8 The costs associated with inactive National
Priorities List (NPL) sites could be substantially higher, however,
this analysis is for facilities now in operation. Obviously some
landfills are better than others and some treatment and disposal
technologies present lower risk. In order to adjust the base case to
account for this, facility rating factors are determined next.
The rating factors shown in Fig. 3 are determined as follows:
• Population: Serves as an indicator for the number of people
who could consider themselves harmed by the release from the
facility. Estimates of injury claims, economic loss claims and
property claims depend on the affected population.
Score Description
1 low for a rural location
2 medium for an industrial location
3 high for an urban location
• Proximity to Water Supply: The workbook assumes that near-
by wells or surface water are sources for drinking water. Esti-
mates of natural resources damage, personal injury claims and
fluid removal and treatment costs depend on the proximity to
water supply. The workbook evaluates the proximity to water
supply as:
Score Description
1 low for a well more than 10 mi away or a
groundwater table more than 50 ft below
2 medium for a well between 1 and 10 mi away
or a groundwater table between 10 and 50 ft
below
3 high for a well less than 1 mi away or a
groundwater table less than 10 ft below
• History of Leaks: This measure serves as an indicator to de-
termine the probability of a leak. The total amount of future
liabilities is the expected value of these costs and equals the
amount incurred when the leaks occur multiplied by the proba-
bility of occurrence. The workbook calculates the probability
of leaks as:
Score Description
I low for no leak, spill or discharge
2 medium for any leak, spill or discharge that
has not harmed human health and/or the en-
vironment and for any potential leak, spill or
discharge
3 high for any leak, spill or discharge that has
harmed human health and/or the environ-
ment
The score for the base case or average landfill is 6 (i.e.,
2 + 2 + 2). Specific facilities (either company-owned or off-site)
are adjusted by totaling their score (from a low of 3 to a high of
9), dividing by 6 (the score of the base case) and multiplying by
the base case dollars per ton. In the case of General Electric, the
information required to determine facility rating factors is ob-
tained as part of the corporate audit program mentioned
previously. Since facilities are qualified by corporate for use by
320
WASTE MINIMIZATION
-------
divisions the resultant scores are typically low.
Even at this stage in the analysis the utility of the liability cost
estimates becomes obvious. For example, it is much easier to ex-
plain to a general manager that long-term liability costs will be cut
by X dollars per ton and this is the justification for switching to a
higher priced disposal facility that provides an environmentally
better service. Discussions involving geohydrology comparisons
between facilities are stimulating to technical people but are in-
comprehensible to most business managers.
The facility-adjusted dollar amount is next modified to account
for the type of technology employed. Table 2 lists examples of
multipliers that are used to either increase or decrease the liability.
These factors vary by more than 100-fold because technology is
probably one of the most significant determining parameters.
Note also that waste reduction is the best situation of all; no
waste, no waste liability!

Table 2
Weighing Factors Based on Technology

Technology Example
Multiplier
Treatment

Surface Impoundment 1.00
Chemical Treatment 0.01
Stabilization/Solidification 0.10
Incineration 0.01
Biological Treatment 0.01
Other Land-based Treatment 1.00
Other Tank-based Treatment 0.10
Surface Impoundment 1.0
Uaste Pile 1.0
Tank 0.1
Other Land-based Storage 1.0
Other Tank-based Storage 0.1
Disposal

Landfill 1.0
Surface Impoundment 1.0
Injection Well 2.0
Land Farm 0.5
Ocean 0.8
Other Off-stie Disposal 1.0
Other On-site Disposal 0.5
Consider two examples to illustrate how the above works.
A well-run incinerator with a rating score of 3 and a multiplier
of 0.01 would have a future liability cost of $1.75 per ton (3 + 6
X $350 x 0.01). On the other hand, a poorly operated deep well
injection unit that is inexpensive by today's market price could
have liability costs as high as $1,050 per ton—and one must
remember this amount is only for liability costs! One also must
consider direct costs. The liability costs can increase or decrease
depending on the specific weighing factors selected as a result of
company policy decisions and/or additional technical informa-
tion. Some companies may consider, for example, deep well injec-
tion safer than landfilling. They may change the 2.0 factor to,
say, 0.5.
The workbook assumes that these liability costs will be in-
curred. Thus, it is not a question of if, but when and how much.
It has been shown how to estimate how much. What follows is a
method to estimate when.
As mentioned previously, mobility and toxicity of the waste are
the parameters used to determine when. The mobility of the waste
measures the speed at which the primary waste constituent moves
through the ground. The higher the mobility, the faster the consti-
tuent moves and the sooner the corrective action and claims will
occur. The differences in the mobility of various chemicals re-
leased to similar hydrogeologic environment are due to dif-
ferences in their chemical properties. Chemical-specific retarda-
tion factors measure soil attenuation and, for organic chemicals,
are primarily a function of the partitioning characteristics of the
chemical in soils.
The toxicity of the constituent also effects the timing of future
claims. Some very mobile constituents might not trigger any cor-
rective action or claim because people do not perceive them as
dangerous. Based on experiments conducted with annuals, tox-
icologists establish dose-response relationships. These relation-
ships correlate the response or number of animals affected within
a sample population to the dose of hazardous constituent that was
ingested or inhaled by these animals.
For the sake of simplicity the toxicity and mobility of a waste
constituent are independently ranked as being high, medium or
low. The time span corresponding to low toxicity, low mobility
constituents was selected somewhat arbitrarily at 20 yr and for
highly toxic, highly mobile constituents at 4 yr. Examples are in-
cluded in Table 3. For mixtures, the most significant (i.e., in
terms of both concentration and toxicity) waste constituent is
used in the evaluation. Here again, in the absence of more
technical data, the rationale is some science, some crystal ball and
lots of policy.

Table 3
Number of Years for Liability Claims to Occur by Chemical
Hazardous
Constituent
Acetic Acid
Chlorine
Formaldehyde
Gasoline
Hydrochloric Acid
Mercury
Methyl Cellosolve
Methylene Chloride
Phenol
Potassium Cyanide
Sodium Hydroxide
Sulfuric Acid
Toluene
Mobility
High Medium Low
Low Number of Years
16
8
4
8
12
4
20
8
8
4
16
16
12
At this stage, we have determined both total dollar amount per
ton plus the time at which these costs will be incurred. A straight-
forward financial analysis flows from this point and is described
in the next section. It is, of course, possible to add or subtract
levels of complexity to or from the liability cost estimation de-
pending on the availability of additional information and the skill
and background of the workbook user. For example, one addi-
tional factor that is useful is to add a multiplier for those treat-
ment or disposal units that are almost exclusively used by a single
"deep pocket" company. In the event of future liability, the com-
pany may end up with a disproportional share of the cost of
cleanup.

Financial Analysis
Financial indices can be used to identify and rank order alter-
native waste generation and management practices which are
preferred to the current practice. The results depend on the rule a
company uses and whether real (i.e., start-year) dollars or
nominal (i.e., reported) dollars are applied. Example rules:
• Discounted Cashflow Rate of Return (DCRR)
Accept an alternative waste generation and management prac-
tice if the DCRR based on incremental cashflows is greater
WASTE MINIMIZATION
321
-------
than cost of money or discount rate
• Net Present Value (NPV)
Accept an alternative waste generation and management prac-
tice if its net present value based on the incremental cashflows
is positive
• Break-Even Point
Accept an alternative waste generation and management prac-
tice if the break-even point based on incremental cashflows
falls within X years of the project's start year
• Return on Investment (ROI)
Similar to the DCRR rule; i.e., accept an alternative waste
generation and management practice if the return on invest-
ment based on the incremental cashflows is greater than the
cost of money or discount rate

Finance staff and most engineers have access to computer soft-
ware programs that turn the drudgery associated with these
calculations into a simple exercise. Abbreviated example financial
input and output formats are illustrated in Tables 4, 5, 6 and 7 for
the alternative waste management case presented in Fig. 3. These
calculations were made using a standard Lotus 1-2-3* financial
software package that has been adapted to meet the specific needs
of the workbook. Table 7 demonstrates a preference for the alter-
native waste management program in Fig. 3.
Hodel Teir Stlrt:
Est1*lte In Stlrt Teir i
Table 4
Input Parameters
ASSUMPTIONS TABLE
1987
InfUtlon Rite (t! : 5.001
«o»lnil Cost of Honey (J) : 17.001
Investment Tn Credit Rite (») : 10.001
Fedenl !.« Rite (1) : 46.001
Oepreclitlon Schedule:

reir .
Percent:
Table

1
?51
5
Cashflows Associated with Alternative
CURRF.K7 PRACIlCt

28-Oct-St
Attl.Hr 1
Clpftll Costs
Operitlng Cipenses
Future UiDtlltles
Activity 2
Ciplul Costs
Opentlng Cxp«nsei
Future llibllltles
Activity 3
Ctpltil Costs
Operating Cipcnses
Future LUbllltles
Activity 4
Cipltil Costs
Operttlna. Cipenies
Future Uibltltles
Activity 5
CipUll Costs
Opentlng Cipenses
Future llibllltles
Activity (
Cipltil Costs
Opentlno. Expenses
Future Uibllltlei
TOTALS
Ctpltil Costs
Opentlng Expenses
Future Llibllltles

0
1986

100000
0
0

10000
0
0

5000
0
0

10000
0
0

5000
0
0

0
0
0

130000
0
0

i
1987

0
10000
0

0
1000
0

0
sooo
0

0
19000
0 .

2 3
?« ISt

Practice
TEAR
7
199)

0
10000
0

0
1000
0

0
5000
0

0
19000
0

(Real

8
1994

0
10000
0

0
1000
0

0
1000
1475

0
1000
0

0
looo
0

0
5000
0

0
19000
1475

8 1
01 01

Dollars)

9
1995

0
10000
0

0
1000
0

0
5000
0

0
19000
0 ...

10
01

. zo
?oot

0
10000
0

0
1000
0

0
19000
0
Table 6
Incremental After-Tax Cashflows of Alternative Practice
Relative to Current Practice (Real Dollars)

net DU(,rence 8et»een Current ind AllernitUe >r«ttlc»l

REAL )
28-OC1-86
Cipltol Costs
Operitlno. Expenses x (1-FTR)
Future Ulbllltles I (1-FTR)
Net
CuxkulitUe
teir
0
1986
-40000
0
4000
0
0
-3(000
•36000

1
1987
0
4381
0
3780
0
8161
-27839 ...

7
1993
0
1308
0
3780
0
5088
6199

8
11*5.
0
0
0
3780
182M
Z»245

9
1995
0
0
0
3780
0
3780
32025 .

20
MM
0
0
0
3780
0
3780
.. 73605
Table 7
Value* of Financial Indices for Alternative Practice
Relative to Current Practke

FINAL OUTPUTS
28-Oct-86
Real :
Nomlnal :
Net Present Value 1n
Break-Even
Point
(rear)
1992
1991
1986 Dollars
Return
On Invest.
(I)
204
391
12457
KM
(1)
17.36
23.22
tioned previously, these can be significantly above the normal in-
dustrial price indices. Table 8 includes example increase. These in-
creases in market prices reflect changes that have occurred be-
tween 1985 and 1986. In the past 5 yr, there have been wide fluc-
tuations. For example, the market price for incineration went up
20<7o between 1984 and 1985.

Tables
Increases in Market Prices by Type of Commercial Treatment,
Storage and Disposal Facility
Increase
in Market
Type of TSDF Price (Percent)
Treatment

Surface Impoundment
Chemical Treatment
Stabilization/Solidif feat ion
Incineration
Biological Treatment
Other Land-based Treatment
Other Tank-based Treatment
Surface Impoundment
Waste Pile
Tank
Other Land-based Storage
Other Tank-based Storage
Landfill
Surface Impoundment
Injection Well
Land Farm
Ocean
Other Disposal
15
15
15
30
15
15
IS
20
20
IS
20
IS
20
20
25
IS
10
20
Other important considerations are the projected cost increases
for commercial storage, treatment and disposal services. As men-

322 WASTE MINIMIZATION
Cost increases for on-site waste management activities in many
cases can be less than those for commercial treatment, storage or
-------
disposal services (i.e., more in line with other site operational cost
increases). This situation may give the company an incentive to
construct its own waste management facilities. This is particularly
true if manufacturing process changes that eliminate waste
generation and avoid the regulatory permit process are im-
plemented. In addition, a company also may feel that it will exer-
cise greater control over its own facilities and further reduce the
likelihood of liabilities. After all, it has already assumed certain
risks because it generated the waste in the first place. Commercial
disposal firms would strongly disagree with this position and
point out that the company now will assume 100% of the liability
associated with future disposal issues. Furthermore, they would
point out that their facilities are larger and most cost effective. By
performing the financial analysis, at least a company will be in a
position to better understand the issues.
CONCLUSIONS
Future liability costs need to be considered when evaluating
hazardous waste management plans. Unless this is done, a com-
pany may select waste management practices that can later prove
to be both environmentally and economically inferior. Although
future liability costs cannot be predicted precisely, it is possible to
structure estimation techniques. General Electric's approach has
been described in this paper.
The specific methodology employed by General Electric is
described in a workbook and a computer software package de-
signed to guide users through the financial analysis steps. Other
methodologies are possible and they may be more suitable to the
needs of some companies.
The important consideration is that the system described can
act as a policy by which company environmental organizations
can guide business components toward better waste management
practices—and do this by utilizing a financial analysis technique.
This method offers several advantages over company policies
such as an end-tax on waste or a five-year waste reduction plan.
The primary benefits are that the method addresses the most
significant, long-term environmental issues and that the output of
the analysis is expressed in financial terms that upper manage-
ment can readily understand and utilize.

REFERENCES
1. Campbell, M.E. and Glen, W.M., Profit from Pollution Prevention:
A Guide to Industrial Waste Reduction and Recycling. Pollution
Probe Foundation, Litpak, B.C., Ed., Toronto, Ont., Canada, 1982.
2. Environmental Engineers' Handbook, 3 Volumes, Chilton Books,
Radner, PA, 1974.
3. Regan, R.W. and Craffey, P.W. "Assessment of the Hazardous
Waste Practices in the Paint and Allied Products Industries," Penn-
sylvania State University, Institute for Research on Land and Water
Resources, University Park, PA, Dec. 1984.
4. Monsanto Research Corporation, "Alternative Treatments of Organic
Solvents and Sludges from Metal Finishing Operations: Final Re-
port," Prepared for U.S. EPA, Dayton, OH, Sept. 1983.
5. Gupta, A., Johnson, E.R., Mindler, A.B. and Schlossell, R.H., "A
Central Metal Recovery Facility: A Preliminary Appraisal of Feasibil-
ity for the Treatment of Electroplating Job Shop Wastes," Prince-
ton University, Princeton, NJ, Oct. 1983.
6. U.S. Congress, Office of Technology Assessment, "Serious Reduc-
tion of Hazardous Waste: For Pollution Prevention and Industrial
Efficiency," OTA-ITE-317, U.S. Government Printing Office, Wash-
ington, DC, Sept. 1986.
7. U.S. EPA, "Remedial Action at Waste Disposal Sites (Revised),
EPA-625/6-85/006, Washington, DC.
8. ICF Technology, "A Comparison of the True Costs of Landfill Dis-
posal and Incineration of DOD Hazardous Wastes," Prepared for
Environmental Policy Directorate, Office of Secretary of Defense,
Defense Environmental Leadership Project Office, Washington, DC,
Sept. 21, 1984.
WASTE MINIMIZATION 323
-------
Waste Reduction in the Semiconductor Industry

Steve Pedersen
Semiconductor Industry Association
Cupertino, California
Mary Ann Keon
Kennedy/Jenks/Chilton
San Francisco, California
INTRODUCTION
The philosophy of industrial waste management has changed
rapidly in recent years due to regulatory and economic incentives
which have restricted and, in many cases, banned the discharge
of hazardous materials into the air, water and land. The increas-
ing cost and liability associated with waste management activities
have resulted in a crisis situation.
Regulators and industry recognize waste reduction as the most
economical and practical solution to the escalating hazardous
waste crisis. The emphasis, however, is still on waste manage-
ment, with over 99% of Federal and state spending devoted to
pollution control and less than 1% directed toward reducing
waste generation (U.S. Congress, Office of Technology Assess-
ment, "Serious Reduction of Hazardous Waste," OTA-ITE-317,
Sept. 1986). One major obstacle to achieving waste reduction is
the lack of industry-specific waste generation data. The develop-
ment and compilation of these data can accomplish the follow-
ing objectives.
• Facilitate technology transfer within industry
• Provide regulators with an adequate information base to realis-
tically determine what needs exist and what action to take in re-
sponse
• Provide industry with practical information to aid in response
to forthcoming regulations
To help meet these objectives, the Semiconductor Industry
Association has begun a study of waste generation and waste min-
imization techniques within the semiconductor industry.

BACKGROUND
The Semiconductor Industry Association (SIA), formed in
1977, currently represents 50 semiconductor manufacturers with
both domestic and foreign operations. In the past 10 yr, the SIA
has been involved in tax and trade negotiations, occupational and
health issues and, most recently, environmental concerns specific
to the semiconductor industry.
In 1984, the SIA initiated a survey of waste generation and
disposal practices within the semiconductor industry. Data on
the volumes of hazardous waste streams and methods of disposal
were compiled for 37 major semiconductor manufacturing facil-
ities located throughout the world. Nineteen of the 37 facilities,
which are part of 16 international firms, are located in California.
The relative volumes of the major waste streams produced by
the 37 facilities are presented in Table 1. As shown, the follow-
ing seven waste streams represent more than 98% of the total
volume of hazardous waste produced by the 37 typical facilities:
• Waste acids (excluding hydrofluoric)
• Solvent (non-halogenated)
• Waste hydrofluoric acid (and fluoride sludges)
• Metals-bearing liquids
• Solvent (halogenated)
• Stripper
• Photoresists
The 1984 survey was the first step in a long-range plan aimed
at the investigation and advancement of waste minimization pro-
grams within the industry.
In April 1986 the SIA applied for a grant under the California
Department of Health Services (DOHS) Waste Reduction Grant
Program. The grant, awarded in June 1986, funded: (1) the
collection and compilation of 1985 semiconductor data on pro-
cess waste generation and disposal practices, (2) the investiga-
tion of waste minimization techniques within the semiconductor
industry and (3) the preparation of a report summarizing the
findings.
The first stage of this project involved the development of a
waste generation survey form and a waste minimization question-
naire, which were distributed to members of the SIA with domes-
tic manufacturing operations. Copies of the survey form and
questionnaire are provided in Attachment A. A 379* response to
the survey form and 20% to the questionnaire were received.
As of this writing, the data are being compiled and analyzed
in preparation for a submittal to the DOHS and to the companies
which participated in the study. The final report, due in June
1987, will: (1) summarize the data from the waste generation
survey forms, (2) describe current waste disposal practices with-
in the United States and (3) present three detailed case studies of
successfully applied waste minimization techniques.
The remainder of this paper will provide the following:
• An overview of the process and waste streams specific to semi-
conductor manufacturing
• A discussion of several key issues encountered during the devel-
opment of this project
• Examples of waste minimization techniques currently applied
within the industry (as available)

SEMICONDUCTOR MANUFACTURING
Semiconductor Device Fabrication
Semiconductors are solid state electrical devices which perform
a variety of functions including information processing and dis-
play, power handling and the conversion between light energy
and electrical energy. The major semiconductor products are:
(1) silicon-based integrated circuits, (2) gallium arsenide wafers
for the production of light emitting diodes (LEDs), and (3) glass
wafer devices such as those used for liquid crystal display (LCDs).
The manufacturing process for all of these semiconductor pro-
324 WASTE MINIMIZATION
-------
Table 1
1984 Semiconductor Indistry Association Survey
Waste Stream Volumes
(16 companies—37 Facilities)
Volume
u.«te Stream (gallons)

W««t. Acid 2.897,749

Solvent
(Non-halogenatad) 1,587,214

W.It. Acid (HF) 1,124,623

Metals-bearing Liquid! 1,095,457

Solvent (Halogenated) 252,348

Stripper 199,863

Photoresist (+ & -) 170,632

Contaminated Solids 92,620

Photoresist (-) 45,377

Photoresist (*) 38,940

Fluoride Sludge 34,648

Vacuum Pump Oil 30,231

Lab Packs 13,728

Gallium Arsenide 11,495

HMDS 7,015

Waste Cyanide 5,184

TOTAL 7,609,619
Percent of Total
Waste Stream

31.1

20.9

14.8

14.4

3.3

2.6

2.4

1.2

0.6

0.5

0.4

0.2

0.2
approx. 100
ducts involves the use of organic solvents, acids and numerous
proprietary chemicals.
The process for the formation of silicon-based semiconductor
devices involves the fabrication of wafers from high-purity
polycrystalline silicon. Wafers currently manufactured range in
diameter from 5 to 8 in. and are approximately 0.2 in. thick.
The wafer undergoes five major processing steps which are uni-
versal to all silicon semiconductor devices: oxidation, photolitho-
graphy, etching, doping (diffusion and ion implantation) and
metallization. The semiconductor devices, commonly known as
chips, are fabricated on the wafer through a sequence of these
processes, which may be repeated several times.
Following fabrication, the wafers are sorted and die cut into
chips, which are then assembled and packaged. A 6-in wafer
could produce 800 semiconductor devices. The following is a
brief discussion of the major processing steps and typical chem-
ical usage:
Oxidation
Oxidation is the process of forming a thin film of silicon di-
oxide on the entire wafer. The wafers are heated in quartz con-
tainers hi the presence of oxygen or ultra pure water.

Photolithography
Photolithography is the procedure for forming extremely
accurate patterns on the wafer. Prior to photolithography (mask-
ing), the semiconductor wafer is cleaned using solvents, acids
and caustics. The wafer is coated with a thin layer of photore-
sist, containing an organic polymer, by spinning a small quantity
on the wafer. The wafer is then heated to bond the polymer.
Ultraviolet light shining through a mask containing a circuit
pattern determines where the polymer will be exposed. Some
polymers are normally insoluble to a developer solution but be-
come soluble after exposure to ultraviolet light (positive resist),
while others behave in a reverse manner (negative resist). The
undeveloped photoresist is dissolved and removed using a "strip-
per" containing acids and caustics.
The photolithography process produces much of the wastes
from the semiconductor manufacturing processes. Hexamethyl-
disilane (HMDS), which is used as an initial coating to increase
the adhesion of photoresist, is a waste produced from the photo-
lithography process. HMDS is typically in a solvent base of xylene
and/or Freon. Waste photoresist solvent and developers are also
produced from the photolithography.
For negative photoresist processes, a waste mixture of isoprene
rubber (the photoresist) and the developer (organic solvents in-
cluding xylene and other nonhalogenated hydrocarbons) is pro-
duced. Positive photoresist produces a waste stream of ortho-
diazo-ketone (the photoresist) and caustic (the developer).

Etching
Etching is used to dissolve away those places on the wafer
which are not covered by the photoresist. This process exposes
the silicon surface or substrate in preparation for doping with
impurities. The etching process primarily uses acids to remove
silicon dioxide and metals according to the patterns delineated by
the photoresist.
Waste sulfuric, hydrofluoric, hydrochloric, phosphoric, nitric
and chromic acids are produced as a result of etching the semi-
conductor devices. Waste lubricating oils from roughing pumps
used to evacuate the reaction chambers of the plasma etchers
are generated, too.

Doping
In doping (junction formation), diffusion or ion implanta-
tion techniques introduce impurities into select regions of the sub-
strate to form a boundary between conducting regions. In diffu-
sion, the wafers typically are stacked in a long, heated quartz
tube and exposed to gases containing impurities which diffuse
into the exposed part of the wafer. Ion implanters bombard the
wafer with ionized impurities. Junction formation produces
solid wastes containing arsenic, antimony, phosphorus, arsine,
diborane and waste pump oils.

Metallization
Metallization is the process of depositing complex patterns of
conductive material to interconnect the integrated circuits.
Metallization wastes significantly differ depending on the specific
plating process used. The waste streams from metallization in-
clude solutions of precious metals, heavy metals and acids.

Further Fabrication Operations
Interspersed with the processing steps are cleaning operations
using solvents and acids. The cleaning of quartz tubes and other
equipment and apparatus also generates waste hydrofluoric acid
and other acids, solvents and caustics.
Semiconductor chips are manufactured using highly sophisti-
cated mechanical/optical electronic instruments in special labor-
atories ("clean rooms") which are designed and operated to min-
imize contamination of the semiconductors. High capital invest-
ments are required to design and construct the clean rooms as well
as purchase and install the sophisticated instruments needed to
manufacture the semiconductor chips. Process equipment often
includes relatively mobile, batch unit operations.
However, the manufacturing processes are constantly changing
as a result of changes in the market and innovative semiconduc-
tor technologies and manufacturing processes. For example, the
need for greater accuracy in the photolithography processes to
manufacture the new semiconductor chips recently has caused
many of the semiconductor manufacturers to switch from nega-
WASTE MINIMIZATION 325
-------
live to positive photoresist processes. Consequently, the use of
alkaline and phenolic strippers has increased, and the use of
organic strippers has decreased. Thus, waste reduction and man-
agement processes must be flexible to respond to constantly
changing process needs.
In the semiconductor industry, the process engineers often try
to control physical and chemical processes at the molecular level.
Many of the physical and chemical factors affecting a manufac-
turing step are not always understood. Therefore, process steps,
including specific chemical usage, often are dictated by the exper-
iences of the fabrication managers, who may be reluctant to
change the steps (i.e., substitute chemicals or reduce doses).
In order for the process engineers and environmental engi-
neers to assess the applicability of waste minimization techniques,
the site-specific factors in the successful application of the tech-
nique must be defined. Making the process engineer aware of spe-
cific waste minimization techniques which are being successfully
applied to reduce the cost of manufacturing semiconductors
may be an effective way to expand the effectiveness of waste min-
imization programs.

Waste Disposal Practices
The wastes are collected by several techniques depending on the
design of the clean room, the electronic instrumentation and the
chemicals used. Many site-specific factors affect the systems for
collection of the waste from the clean rooms. The dilute acid
and alkaline aqueous wastes from rinsing operations typically
are piped outside of the building where they are neutralized and
discharged to a sanitary sewer. Hydrofluoric acid is aspirated to
a waste storage tank outside of the building or collected from the
process line in a container on a cart. The containers are then
transported to the waste storage area where the waste is trans-
ferred to larger storage containers or tanks.
Waste solvents, photoresists and strippers often are collected
in small containers. In a few instances, dedicated piping systems
are used to transport the waste photoresist and strippers to the
waste storage area.

ISSUES
Several key issues and difficulties have been encountered dur-
ing the development of this project. Because the project is still
being conducted, the following discussion will highlight our cur-
rent findings regarding these issues and provide possible solu-
tions that are being investigated. The issues which have been most
critical to the project's success are:
• Definition of waste minimization
• Basis for measuring waste reduction
• Units for reporting waste generation
• Incentives for waste minimization

Definition of Waste Minimization
For the purposes of this study, waste minimization was given
a broad definition that included waste treatment for volume and
hazard reduction as well as process modification, raw material
substitution, better housekeeping, on-site reuse, recycling and
reclamation. Because this project was the first of its kind to study
waste minimization within the semiconductor industry, this broad
definition was chosen to avoid restricting the scope of the inves-
tigation.
According to a recent Office of Technology Assessment (OTA)
study, defining waste minimization in a way that includes waste
management activities can divert attention away from waste re-
duction. They felt that if primacy is to be placed on waste reduc-
tion processes, then waste management activities that reduce the
amount of land-disposal waste should not be included in the
definition. Better and more efficient waste management prac-
tices can be dealt with as a separate issue.
Our findings indicate that while the semiconductor industry
recognizes waste reduction at the source (i.e., source reduction)
to be the ultimate goal, current practice still emphasizes cnd-of-
pipe management of hazardous waste. With a growing industry
as technically advanced and competitive as the semiconductor
industry, there is an extreme sensitivity to process changes which
could interfere with production. Therefore, we felt that a broader
definition of waste minimization that includes both the reduction
of waste generation at the source and waste management to re-
duce or detoxify land-disposal waste would be more useful, since
it is more reflective of current conditions in the industry.

Basis for Measuring Waste Reduction
A major dilemma in analyzing waste reduction is deciding how
such reduction should be measured. The basis for measuring
waste reduction over time must take into account product
changes, industrial activity changes and economic factors such as
product cost. Several possibilities include:
• Annual sales
• Units of production
• Wafer starts
• Chemical mass balance
Units of production are the only meaningful measure of waste
reduction because they are directly correlated to chemical usage
and are accurately determined by the semiconductor manufac-
turers. (For the semiconductor industry, units of production
translates into the number of chips produced on an annual basis).
However, the proprietary nature of production information
makes the industry sensitive to its release.
Using wafer starts or number of wafers produced annually as
a basis of measure would avoid proprietary conflicts. However,
wafer starts can be difficult to determine accurately and as a basis
for measuring waste reduction would not account for variations
in wafer size and chemical usage. Measuring waste reduction in
terms of annual sales introduces variables that are not directly
correlated to chemical usage, such as fluctuating market prices
and decreases in industrial activity.
Sufficient data are not available to draw any conclusions relat-
ing long-range sales figures with production activity. The last
option, measuring waste reduction by conducting mass balance
calculations at the plant level, would be extraordinarily difficult
and present some of the same confidentiality problems associated
with using production units.
One suggestion which may avoid the confidentiality issue and
still provide meaningful waste reduction data is to pool the waste
percentage data from each of the semiconductor companies and
average the values to give a measure of waste reduction on an
industry-wide basis. This option would give a true measure of
waste reduction and avoid proprietary restraints. The major dif-
ficulty with this approach is in the development of the present
waste reduction figures, which must be derived from produc-
tion-based waste generation data.

Units for Reporting Waste Generation
Responses to the 1985 survey indicate that individual com-
panies measure and report their waste generation in different
ways. Half of the responses to the survey form reported waste
generation figures in volumetric units, while the remaining half
were reported in mass units. Companies are being contacted in
an attempt to obtain appropriate conversion factors. Although
seemingly trivial, conversion factors have been difficult to obtain
are, in many cases, inaccurate. Future surveys will be modified to
326
WASTE MINIMIZATION
-------
request: (1) specific units for the waste generation figures and
(2) estimated conversion factors (density), if available.

Incentives for Waste Minimization
At the state and federal levels, there exist only indirect incen-
tives for waste reduction, such as the following:
• High cost of pollution control (i.e., landfilling, incineration)
• Current and future landfill bans enacted under the 1984 amend-
ments to RCRA
• Liability involved in the transportation and off-site disposal
of hazardous waste
In response to these indirect incentives, several companies have
developed programs to promote waste reduction at the plant
level. Examples include the following:
• Accounting methods which charge short- and long-term costs
of waste management to the fabrication laboratories and de-
partment responsible for the processes that generate the waste
• Employee programs which: (1) provide education and train-
ing in waste reduction techniques, and (2) set goals with pos-
itive inducement such as financial rewards or recognition for
achievement
• The communication of knowledge regarding waste reduction
techniques throughout the company by conducting regular in-
house seminars
These programs exist among the larger semiconductor firms
with the personnel and finances to support them. The competi-
tive nature of the industry has restricted inter-company commun-
ication about waste reduction methods and programs. As stated
earlier, one goal of this project is to transfer information and
provide a conduit for future communication between the mem-
bers of the SIA.
In studying the different incentive programs employed by in-
dustry, it is clear that the concept of waste minimization must
be sold at the top. Unless top level managers and executives are
committed to waste minimization within their company, it is un-
likely that any long-range programs will emerge. The environ-
mental managers and engineers, typically responsible for waste
minimization efforts have limited resources available; their major
efforts are directed at regulatory compliance and at "firefight-
ing" to resolve daily operating problems. An active waste min-
imization program must be a combined effort between upper
management, engineering and laboratory workers. Education is a
crucial step in the successful implementation of a waste reduc-
tion program. When quality control and production are critical,
education must overcome the resistance to change and innova-
tion so often present.
Each of the issues discussed above has raised more questions
and problems, rather than yielding solutions to the issue of waste
reduction. For example, the definition of waste minimization can
both emphasize the primacy of minimization at the source and
confuse the concept with waste management. Determining the
basis for measuring waste reduction involves serious conflicts
with the industry's need to protect proprietary information. The
variety of units used to report waste generation data underlines
the diversity present in the industry. Developing programs to en-
courage waste minimization should not be the sole responsibil-
ity of the environmental staff, as their time and resources are
too limited. Resolution of these issues and many more will only
be accomplished over time through a concerted effort by the regu-
latory, industrial and private sectors of society.
WASTE MINIMIZATION 327
-------
Department of Defense Hazardous Waste Minimization

Capt. Michael J. Carricato, USN
Andres Talts, P.E.
Joseph A. Kaminski, P.E.
Office of the Deputy Assistant Secretary of Defense
(Environment)
Washington, D.C.
Thomas E. Higgins, Ph.D., P.E.
CH2M Hill
Reston, Virginia
ABSTRACT
Hazardous waste minimization is essential to the Department
of Defense in order to reduce the serious liabilities associated with
its management and disposal.
DOD policy is to limit the generation of hazardous waste and
to recycle resources where practical. Any action that reduces the
need for disposal of hazardous waste is considered minimiza-
tion: hazardous materials control; material substitution; process
change; recycling; treatment, including thermal destruction.
Case examples of successful waste reduction efforts are pre-
sented in the paper.

INTRODUCTION
The Hazardous and Solid Waste Amendments (HSWA) of
1984 have made it more difficult and more expensive to dispose
of hazardous wastes. Some wastes are or will be banned from
land disposal entirely (e.g., liquid wastes) while other disposal
practices will be severely restricted. In addition, generators are
required to certify that they have instituted a waste minimiza-
tion program. The financial and legal incentives to reduce or en-
tirely eliminate the generation of hazardous wastes are becoming
more attractive.

DEPARTMENT OF DEFENSE
WASTE MINIMIZATION POLICY
Waste minimization is essential to reduce the serious liabil-
ities associated with the generation and subsequent handling
and disposal of hazardous waste. Department of Defense (DOD)
policy is to limit the generation of hazardous waste and to re-
cycle resources where practical.
Current law neither includes nor requires a precise definition.
The U.S. EPA's "working definition" of hazardous waste min-
imization includes anything that lessens the load on treatment,
storage and disposal facilities by reducing the quantity or toxicity
of hazardous waste.
Within DOD, minimization is any action that reduces the need
for disposal of hazardous waste. This can be accomplished by the
following methods:
• Hazardous material control
• Material substitution
• Process change
• Recycling
• Treatment, including thermal destruction
All DOD components are required to prepare hazardous waste
minimization plans. There are no absolute quotas or goals for re-
duction. However, a pragmatic goal is to treat all hazardous
waste for volume or toxicity reduction. As a management objec-
tive, all waste should be examined for this goal by 1992. The Used
Solvent Elimination (USE) Program, started in 1984, is an essen-
tial element of the DOD hazardous waste minimization program.
The Joint Logistics Commanders of the military services, who
are responsible for the major DOD industrial facilities related to
equipment and materials, have taken the lead in developing min-
imization programs for their commands. Their programs require
that they:
Prepare annual reports of hazardous waste generation
Implement hazardous material control programs
Assess and implement practical minimization technology
Assess and initiate R&D projects
Develop reduction goals and monitor progress
Establish minimization as an important consideration in acqui-
sition programs

SOURCES OF HAZARDOUS WASTE
GENERATION
The military services operate over 100 industrial facilities to re-
pair and recondition a wide variety of military equipment in-
cluding: aircraft; helicopters; ships; wheeled and tracked vehicles;
and other weapon systems and equipment. Common operations
performed at many of these industrial facilities include paint
stripping, solvent cleaning, metal plating and painting. Waste sol-
vents and toxic metal wastes from these metal finishing opera-
tions at the industrial facilities are the principal sources of haz-
ardous wastes produced by the Department of Defense.

CASE EXAMPLES OF WASTE
REDUCTION EFFORTS
The military services have for years made efforts to reduce
waste as part of their overall waste management programs. Sev-
eral individual efforts have been instituted, and innovative tech-
nologies have been developed by the services as part of their over-
all waste management programs.
The Environmental Policy Directorate of DOD investigated the
efforts of individual services and facilities to determine the suc-
cess of these efforts, to determine which technologies had been
successful, to encourage technology transfer of these Projects of
328
WASTE MINIMIZATION
-------
Excellence and to develop policy changes to identify and encour-
age the adoption of cost-effective methods for hazardous waste
reduction by the individual services and installations. Some ex-
amples of successful hazardous waste minimization projects
follow.

PAINT STRIPPING
Paint stripping is widely practiced in the preparation of equip-
ment and components for reconditioning and recoating. In a
typical paint stripping operation, sprays or baths containing
acidic methylene chloride, phenolic compounds or hot alkaline
solutions are employed to loosen and dissolve old paint. After the
paint softens, the resulting solvent-paint mixture is scraped or
brushed off.
The solvent-paint mixture usually is washed away with water,
resulting in the production of large quantities of wastewater con-
taminated with toxic metals, methylene chloride and phenols.
Treatments to remove metals includes hexavalent chromium and
reduction and metal hydroxide precipitation. Removal of the
methylene chloride and phenol usually involves air stripping
followed by activated carbon adsorption. Sludges and residues
from the treatment of paint stripping wastes are typically haz-
ardous due to the presence of metal pigments and/or toxic organ-
ics.
Several alternative paint stripping processes have been studies
at private and military industrial facilities. Among these are dry
media blasting, laser stripping, flash lamp stripping, CO2 pellet
blasting and cryogenics. The most promising of these techniques
is plastic medial blasting (PMB) paint removal, developed by
technicians in the Ogden Air Logistics Center at Hill Air Force
Base located near Ogden, Utah.

Plastic Medium Blasting
Conventional sand and glass bead blasting have been used to
remove paint and rust from metal surfaces. However, these ag-
gressive media can damage soft or delicate metal surfaces and
can produce a silicate dust which can cause silicosis, a respira-
tory ailment. Softer vegetable media, such as walnut shells and
rice hulls, have been successfully used in blasting. However, these
materials degrade rapidly, producing voluminous quantities of
dust and resulting in explosion and health hazards; they also
contribute to the generation of large quantities of hazardous
waste.
In PMB paint stripping, small angular plastic particles are air-
blasted at the painted surface, resulting in high local stresses
that dislodge the paint. The plastic material is harder than the
paint but softer than the underlying substrate. Since the particles
are plastic, they deform and return to their original shape rather
than shatter on impact like sand or vegetable media. By carefully
controlling the size or the beads and the conditions of the pro-
cess, the plastic media can be separated from the loosened paint
particles. As much as 95% of the plastic material can be recov-
ered and reused.
Generation of wet hazardous waste (solvents and paint sludge
in water) is eliminated. A small volume of dry waste is produced,
consisting of the decomposed plastic medium and paint chips.
Some samples of this dry waste have been determined to be haz-
ardous due to the stripping of chromate primers and incidental re-
moval of cadmium from plated parts.
The Ogden Air Logistics Center at Hill Air Force Base, Utah,
has been a lead military facility in the development and testing of
plastic media blasting technology. A full-scale PMB paint strip-
ping booth was constructed at Hill AFB. In recognition of this,
the PMB facility at Hill AFB was awarded the distinction of be-
ing a Project of Excellence for Hazardous Waste Reduction by
DOD, and workshops were held at the facility to encourage the
adoption of this technology at other facilities.

Workshops
The workshops served as a sounding board and brought out
such issues left to be resolves as dust generation, worker ex-
posure to metal contaminated dusts and potential damage to deli-
cate aircraft components. Further studies of the PMB process are
being performed, but the general consensus is that the problems
with the process are solvable.
Partly because of the workshops, but also because of the in-
formal networking of production personnel, the technology is
rapidly being tested and adopted at other military and private in-
dustrial facilities.

General Applicability
As part of this project, 26 industrial installations were surveyed
to determine the applicability of PMB paint stripping technol-
ogies. The installations surveyed included seven Army Depots, a
Marine Corps Logistics Base, five Naval Air Rework Facilities
(NARF), two Navy Shipyards, three Naval Ordnance and Wea-
pons Centers, a Naval Air Station, six Air Force Logistics Cen-
ters and a Tactical Air Command Air Force Base.
The general consensus was that PMB paint stripping had some
application at virtually all of the installations surveyed. Five of
the surveyed installations were performing PMB paint stripping
and had plans to expand their capabilities or had potential addi-
tional applications. An additional seven installations had plans to
add PMB paint stripping facilities.

Current Applications
Pensacola NARF has been designated the Navy's lead base for
the development of PMB paint stripping. Pensacola has success-
fully stripped paint from aircraft and helicopter parts using en-
closed glove boxes and open blasting in rooms. Pensacola is pur-
chasing two walk-in blast booths and blast cabinets. If testing
proves satisfactory, Pensacola plans to convert a wet stripping
hangar to accommodate dry medium stripping of whole aircraft.
Alameda NARF has been successfully paint stripping aircraft
and missile components in a walk-in blast booth converted from
walnut shell blasting to PMB. Alameda currently is constructing
a PMB hangar with a 120-ft by 120-ft by 50-ft interior clear space
capable of holding the P3 Orion aircraft.
Corpus Christi Army Depot has taken the lead in adapting
PMB paint stripping helicopters for the Army. Corpus Christi has
been operating PMB paint stripping facilities in converted offices
at the depot in which they have performed prototype testing,
developed methodology and conducted some production paint
stripping. Personnel at the facility have an active research pro-
gram and are in the process of purchasing a blast booth to accom-
modate paint stripping of whole helicopters.

RECYCLE OF SOLVENTS FROM
CLEANING AND PAINTING
Solvents are used at virtually every industrial facility. Metal
surfaces often are cleaned with solvents to remove accumulated
dirt, oils, greases and corrosion products. Trichloroethylene,
trichloroethane and perchloroethylene are used in vapor degreas-
ers, and mineral spirits such as Stoddard solvent and Varsol are
used in cold cleaning baths. Alcohols and Freon are commonly
used for metal preparation and precision cleaning of electronics
equipment. Volatile solvents such as toluene are commonly used
to thin solvent-based paints, and other solvents such as methyl
ethyl ketone (MEK) and xylene are used to clean up painting
equipment after use.
WASTE MINIMIZATION 329
-------
Disposal costs for waste solvents can exceed $100/drum and
are expected to increase substantially due to adoption of amend-
ments to RCRA banning the land disposal of liquid hazardous
wastes. When waste solvents are disposed of, fresh solvents must
be purchased. Recycling waste solvents can result in significant
savings due to the elimination of the costs and liabilities of waste
disposal and of virgin material.
There is great potential to reduce the quantities of solvents and
related by-products generated at DOD industrial facilities. It has
been estimated that DOD purchases and disposes of approxi-
mately 50,000 drums of cleaning solvent each year.

Batch Distillation
Solvents and other organic fluids are most frequently reclaimed
by simple batch distillation. Batch distillation systems typically
consist of a still pot, a heat source and a condenser. The waste
organic mixture is loaded into the still pot; heat is applied to the
contents, causing the mixture to boil; and organic vapors sep-
arate from the waste mixture and pass overhead to the conden-
ser. The clean solvent is then collected for reuse and still bottoms
are disposed as hazardous waste.
An atmospheric still can reclaim organic solvents that have
boiling points less than 325 °F. By adding vacuum, a distillation
unit can recover organic fluids which have atmospheric boiling
points up to 500°F while maintaining 300°F limit in the still's pot.
Since these simple stills cannot segregate components of similar
boiling points, it is imperative to segregate solvent wastes plan-
ned for recycle.
Waste solvents can be collected and transported to a central-
ized distillation facility for recovery, or they can be recycled at
the point of use. DOD facilities have been successful with both
approaches.

Central Facility
The main advantage of operating a large centralized facility is
that capital costs can be recovered quickly due to economies of
scale. A centralized facility can redistill large quantities of various
types of solvents. Since many different types of solvents are re-
cycled, great care must be taken with waste segregation and sam-
ple analysis. Another disadvantage of centralized reclamation is
that solvents must be transported to and from the point of use.
Localized facilities are sometimes preferable, however, because
the waste generator has total control over the recycling opera-
tion. Since only a few types of solvents are being redistilled at
the small facilities, laboratory analysis of waste solvents often is
not required. Labor intensive transportation and segregation
activities also are eliminated.
A centralized facility is dependent on a dedicated individual to
initiate and supervise operation of the system and an enthusiastic
staff dedicated solely to solvent collection, analysis, recycling
and distribution. Decentralized facilities require the conversion of
more personnel (foreman and operators) to adopt solvent recov-
ery as part of their routine.

Recycling Examples
Warner Robins Air Force Base, located in Macon, Georgia,
has operated a centralized batch, atmospheric still since August
1982. The organic fluid recovery system consists of a single stage
batch still, a water separator and an electrically powered steam
generator. The still, which can operate up to a temperature of
300 °F in a pot and can reclaim organic fluids at a rate of 55 gal/
hr is used to reclaim trichloroethane, Freon-113 and isopropanol,
with recovery rates of 70-99%.
From the initial startup through the end of 1984, it was esti-
mated that over $230,000 had been saved due to reduced need
for virgin material and reduced hazardous waste disposal costs.
It cost only $l3/drum to reclaim the used chemicals, whereas
disposal of the chemicals and repurchase of new materials would
have cost from $250 to $500/drum.
Robins Air Force Base has been able to successfully recycle
solvents in a large-scale operation because of careful waste seg-
regation, storage and transportation. Site managers are respon-
sible for segregation and labeling of waste drums at 30 different
collection areas. Before solvents are reclaimed, samples are ana-
lyzed to confirm the labeling. Samples are also analyzed after
distillation to ensure that they meet appropriate specifications.
Solvent recycling has been successful at Robins Air Force Base
because of a strong commitment from management to reduce the
quantities of waste solvents that must be disposed of. More im-
portantly, production personnel have cooperated with the re-
cycling team so that waste solvents can be segregated, labeled,
analyzed, transported and redistilled in an orderly and systematic
fashion. Proven distillation equipment was available and was
relatively easy to operate and maintain.
An example of successful decentralized solvent recycling is the
program at Norfolk Naval Shipyard in Norfolk, Virginia. The
paint shop foreman installed a $10,000 nonfractioning batch still
to recover waste solvents generated in the paint shop during
cleaning operations. This small still, which has a capacity of 2
gal/hr, is used to recover methyl isobutyl ketone, methyl ethyl
ketone, epoxy thinners and mineral spirits. Operators have the
option to operate the still with or without a vacuum system, de-
pending upon the volatility of the solvent to be recovered. After
a 15-min start-up period, the still runs without operator attention.
Over 80% of waste solvent is recovered at a cost of about 15C/
gal.
There is a potential for significant additional solvent recovery
at DOD industrial facilities. Potential impediments include the
manpower requirements needed to run a central facility and con-
cerns over meeting specifications required to maintain production
quality.

METAL PLATING
Plating is defined as the deposition of a thin layer of metal on
the surface of a base metal for the purpose of changing its prop-
erties. These modifications may be to improve the appearance
(decorative plating), to increase resistance to corrosion or to im-
prove engineering properties (hardness, wearability, solderability
or frictional characteristics) of the base metal.
The principal metals plated at military facilities are chromium
and cadmium. Chromium is used to rebuild worn parts or provide
corrosion resistance. Cadmium is used to provide corrosion pro-
tection, improve wear or erosion resistance or reduce friction.
The major discharges of hazardous waste from typical metal
plating facilities are: rinsewater contaminated by drag-out from
various cleaning and plating baths; cleanup of spills; disposal
of acid and alkaline cleaners and occasional plating bath dumps.
These wastes usually are treated in an industrial treatment facil-
ity by hydroxide precipitation, producing a sludge that is a listed
hazardous waste.

Waste Reduction Processes
The most common process modifications that have been im-
plemented at DOD plating shops to reduce generation of haz-
ardous wastes are: reduction of drag-out from processing baths,
reduction of rinsewater flows, improved rinsing efficiencies, re-
covery of metals from rinsewaters and material substitutions.
The Naval Civil Engineering Laboratory (NCEL), Port
Hueneme, California, implemented several process changes on a
chrome plating line at Pensacola Naval Air Rework Facility,
330 WASTE MINIMIZATION
-------
Florida. These process changes included installation of a recircu-
lating spray rinse for removal of drag-out from plated parts.
Clean rinsewater is available from a hand sprayer for final rinse
over the recirculating spray rinse tank. Following installation of
the spray rinse, rinsewater usage dropped from 350,000 gal/mo
to 1,200 gal/mo.
In addition to the rinse modifications, the plating temperature
was increased, resulting in increased evaporation from the plating
tank. Since this resulted in the evaporation rate from the plating
bath exceeding the rinsewater rate, rinsewater was returned to
the plating bath, recovering the chromium and eliminating the
discharge of waste. Additional modifications to racking and plat-
ing have resulted in increased plating rates and a reduction in the
number of plating baths required.
Since the recovery of drag-out could result in the buildup of
contaminants in the plating bath, an electrolytic bath purifica-
tion system was installed to continuously remove contaminating
cations from the bath and to convert trivalent chromium back to
its hexavalent form.
These modifications have resulted in a reduction in industrial
wastewater treatment costs of about $25,000/yr. In addition,
plating bath dumps have been eliminated, resulting in a signifi-
cant reduction in hazardous waste generation.

Zero Discharge
The "zero discharge" plating process at Pensacola was desig-
nated as a Project of Excellence for Hazardous Waste Reduction
by DOD, with technology transfer workshops held at the facility
similar to those for PMB paint stripping at Hill Air Force Base.
As part of this project, 22 industrial installations were surveyed
to determine the applicability of "zero discharge" chrome plat-
ing technologies. These included five Army Depots, a Marine
Corps Logistics Base, five Naval Air Rework Facilities, seven
Navy Shipyards, a Naval Ordnance Center and three Air Force
Logistics Centers.
The consensus was that the process was most applicable for
aircraft repair facilities where a large number of small, relatively
similar parts are rebuilt through hard chrome plating and machin-
ing. The process is less applicable to facilities that plate large,
highly variable parts.
All of the NARFs already had a variation of the process in
place or were planning to implement it. Three of the shipyards
had implemented the process but did not find it as applicable as
the NARFs. The Air Force ALCs did not adopt it because some
of the process changes conflicted with their specifications. The
Army depots had no plans to adopt the technology.
Cherry Point NARF independently developed a similar "zero
discharge" plating process. Rinsing has been simplified; parts are
first rinsed in a static rinse tank and then "hosed off" with fresh
water over the static rinse tank.

CONCLUSION
Individual installations have had some successes in developing
hazardous waste minimization technologies. Some technologies
have been developed by the services while others have been adop-
tions of civilian industrial techniques to military applications.
Workshops supported by the DOD acted as a catalyst to speed
up dissemination of information on successful waste minimiza-
tion technologies, although there is an informal network among
production personnel that promotes technology transfer among
facilities and between the services.
The DOD has established a policy of requiring the individual
services to establish hazardous waste minimization programs and
providing incentives to encourage the development and adoption
of cost-effective waste minimization technologies. Defense Envi-
ronmental Restoration Account monies have been committed to
support projects. Implementation is the responsibility of the in-
dividual services.
DOD has established a definition of waste minimization to in-
clude any action that reduces the need for disposal of hazardous
waste, including such methods as material substitution, process
change, recycling and waste treatment.
WASTE MINIMIZATION 331
-------
Reducing the Generation of Hazardous Waste:
Actions by Government and Industry
Joel S. Hirschhorn, Ph.D.
Kirsten U. Oldenburg
Office of Technology Assessment
United States Congress
Washington, D.C.
ABSTRACT
Moving from a concept that nearly everyone has endorsed and
on which some companies have focused attention to a major na-
tional commitment presents obstacles in both government and in-
dustry. Pollution prevention through waste reduction can be an-
other strategy for environmental protection goals. Pollution con-
trol or end-of-pipe programs can be complemented by waste re-
duction efforts that offer American industry a way to improve
profitability at a time when that is critical. Government action
does not have to mean the use of a traditional regulatory ap-
proach for waste reduction; there are a number of public policy
options that would stimulate attention to waste reduction by in-
dustry and offer it assistance, too.

INTRODUCTION
The best way to prevent future hazardous waste problems is to
prevent generation of the waste in the first place. As a means of
pollution prevention, reduction of the generation of hazardous
waste at its industrial source is the best and most certain way to
reduce risks to health and the environment. It is not the funda-
mental concept of waste reduction that is controversial; the issues
and differences of opinion concern its applicability, availability,
implementation by industry and tangible support by government.
OTA recently released its report, Serious Reduction of Haz-
ardous Waste; the 250-page report offers a comprehensive assess-
ment of the technical, economic and policy dimensions of haz-
ardous waste reduction. From its analyses, OTA concludes that
this approach to environmental protection is not an idealistic or
theoretical possibility but a practical, near-term alternative to
managing waste and controlling pollutants after they have been
generated. This has not always been the case. Circumstances have
changed because of the steady increase in the number of environ-
mental regulations and the steady increase in the costs of com-
plying with them. The greater awareness of regulatory noncom-
pliance and the growing recognition of the costs (e.g., Superfund)
and damages resulting from ineffective control of toxic waste and
other hazardous pollutants have also contributed to making waste
reduction more attractive. Prevention now makes a great deal of
environmental and economic sense, not as a replacement for
pollution control measures but as a complement to them.
The OTA report provides the information base and analytical

(The views expressed here arc those of the authors and nol necessarily
those of the Office of Technology Assessment.)
and policy framework for a national debate on waste reduction/
pollution prevention. The environmentalism of the 1960s can be
transformed by the economic sensibilities of the 1980s to create a
new environmental protection strategy. It is of particular signifi-
cance that after 10 yr of success in waste reduction, 3M can say
that the costs of pollution control "will almost always make the
total cost of this technology higher than the total cost of prevent-
ing pollution at the source." And DuPont says, "Reduced waste
will inevitably lead to lower cost for products, and thus, a higher
standard of living for all Americans." The pollution control cul-
ture that shapes how we think about environmental protection
was born in an atmosphere of crisis; now it can be joined by a new
waste reduction ethic.
There is an easy analogy between waste reduction and energy
conservation, but also a big difference. With energy conserva-
tion, there are multiple sources of energy, changing demand and
large market price fluctuations. But the costs of managing waste
and controlling pollution for a specific amount of waste or pollu-
tants can only continue to increase—and the economic driving
force for waste reduction/pollution prevention will not waver.
A better analogy is with preventative medicine—it has taken a
long time and sharply increasing health care costs to drive home
the point to many people that prevention is the most cost-effec-
tive strategy. The same applies to the environmental area.

CHANGE IS POSSIBLE
The most important word in the title of the OTA report is not
reduction, hazardous or waste—it is serious. For a decade, nearly
everyone said that waste reduction was the option of choice but
did little to make it a reality. The obstacles are not technical or
economic—they are attitudinal, behavioral and institutional. Old
ways have to be changed. But, as we all know, changing human
attitudes and behavior and government is no easy task. On the
other hand, our study has found solid evidence that such change
is possible.
First, pioneering U.S. companies have discovered how waste
reduction can offer substantial benefits quickly and without
spending a lot of money. 3M has saved about $300 million in the
last 10 yr and has used its previous success to spur its commitment
to continuing waste reduction. Companies may not always pub-
lish the technical and economic details of successful waste reduc-
tion efforts, but there is convincing evidence of the near term
practicality of waste reduction.
Second, some states and foreign countries have successfully
332 WASTE MINIMIZATION
-------
used non-regulatory approaches to help industry reduce waste.
We have examined state programs in great detail; although they
are not funded at high levels and are still new, there is clear evi-
dence that state technical assistance and information transfer pro-
grams help some companies, especially smaller ones, to reduce
waste. Some European countries that compete with us interna-
tionally are a decade ahead of us in waste reduction. Govern-
ments in Europe have spent substantial sums to help—but not
force—industry to reduce waste. The motivation for the joint
government-industry effort in Europe has been to serve their own
economic self-interests.

PRIMACY FOR WASTE REDUCTION
No one disputes the fact that the best environmental protection
comes from not producing waste and pollution in the first place.
However, many things clearly can stand in the way of widespread
waste reduction. Therefore, the OTA report emphasizes: (1) a
definition of waste reduction that is consistent with pollution pre-
vention—and the dangers of an ambiguous definition or one that
includes waste management; and (2) the case for the primacy of
waste reduction over pollution control. No technology to control
pollution or manage waste can reduce risks to health and environ-
ment as well as not produce some pollution itself.
It is not sensible to suggest that the pollution control regula-
tory system can be replaced. It is not realistic to think that we
can eliminate the generation of all pollutants and wastes. The Na-
tion will need adequate capacity to effectively treat the hazardous
waste that will continue to be generated and the enormous
amount of materials from site cleanups, although more and more
cleanups will be done on-site. But it is important to understand
that most people in industry are accustomed to thinking in terms
of end-of-pipe pollution control or the management of toxic
waste after it is created. Moreover, the pollution control regula-
tory system cannot by itself motivate—and does nothing to assist
—industry across all sectors and plants to expeditiously imple-
ment waste reduction to its technical and economic limits.
Our definition of waste reduction encompasses actions taken
by a waste generator to alter production practices to generate less
waste before the generator has to handle, transport and manage
it. Waste treatments such as incineration and off-site recycling
are preferred over land disposal but neither management method
is as effective as waste reduction. Approximately 10% of current
Superfund sites that require permanent cleanups were once waste
treatment or recycling facilities. We stress definition and the con-
cept of primacy because so much attention has gone to finding
alternatives to the land disposal of toxic waste that many people
have lost sight of waste reduction and, instead, focus on waste
treatment. While there is room for both approaches, the auto-
matic response from industry often is a traditional pollution con-
trol/waste management solution rather than an exploration of
waste reduction possibilities. The first is routine for environmen-
tal engineering or regulatory compliance departments; the second
must involve everyone involved in production, from shop work-
ers to engineers to R&D personnel to managers.
Waste reduction/pollution prevention offers something for
nearly everyone:

• Industry benefits from lower production and overhead costs
and reduced liabilities
• The public benefits from better environmental protection and
a more competitive industrial base
• The government benefits from having less to regulate and en-
force
Even with more environmental regulation—which is inevitable
—national environmental spending, instead of continuing to in-
crease, could decrease from the current $70 billion annual level if
waste reduction becomes widespread.
The chances are good that over the coming months a major
national interest in waste reduction/pollution prevention will
crystallize as the public understands that this is an immediate
option with environmental and economic benefits. A new posi-
tive, common sense approach to toxic waste and environmental
protection is available to the American public. The point is not to
blame anyone but to take advantage of the many available waste
reduction opportunities.

PRIVATE AND PUBLIC ACTIONS
The OTA report spells out what industry can do and what Con-
gress can think about doing to implement the statement of na-
tional policy in the 1984 amendments to RCRA: "The Congress
hereby declares it to be the national policy of the United States
that, wherever feasible, the generation of hazardous waste is to
be reduced or eliminated as expeditiously as possible."
The six major steps industry can take are:
• Conduct waste reduction audits, like energy use audits, to iden-
tify waste reduction opportunities.
• Revise accounting methods so that both short- and long-term
costs of managing wastes, including liabilities, are charged to
the departments and individuals responsible for the production
practices that generate the waste.
• Involve all employees in waste reduction planning and imple-
mentation. Waste reduction must not be seen as the sole
responsibility of environmental engineers or regulatory com-
pliance departments.
• Motivate employees and focus attention on waste reduction
by setting goals and rewarding employees' successful waste re-
duction efforts. Special education and training may be neces-
sary.
• Transfer knowledge throughout the company so that waste re-
ducing techniques implemented in one part of the company
can benefit all departments, divisions and plants. Company
newsletters and conferences may be necessary.
• Get technical assistance from outside sources, including state
programs, universities and professional consultants.
What can the Federal government do to spur more industrial
reduction of hazardous waste? Many possible options are dis-
cussed in the OTA report. However, the analysis does not sup-
port the use of a traditional regulatory or prescriptive approach.
Simply put, there are far too many industrial processes and site-
specific conditions for the government to establish technically
sensible standards or regulations.
The OTA analysis notes a number of reasons why staying with
the current system is not likely to bring about all the waste reduc-
tion that is technically and economically feasible. The current
system is based on voluntary industrial waste reduction without
any significant Federal assistance, while at the same time expect-
ing compliance with pollution control regulations. There are com-
peting demands for capital and technical resources in industry.
Voluntary waste reduction must compete with the legal require-
ments to comply with environmental regulations as well as other
economically attractive investments.
Moreover, there is no effective gathering of information on
waste reduction, no good definition of waste reduction and no
understanding of which wastes ought to be reduced. There are
important technical and sometimes economic obstacles to im-
plementing comprehensive waste reduction for many companies,
but there also are real environmental and economic benefits to
having widespread waste reduction take place sooner rather than
later. Relying on indirect incentives imposed by the current regu-
WASTE MINIMIZATION 333
-------
latory system and the ability of the marketplace to respond quick-
ly could mean that we will forego significant environmental and
economic benefits.
The OTA report offers Congress a middle course based on
Federal leadership and assistance. If the Federal goal is rapid and
comprehensive reduction of hazardous waste, then the following
six options offer a practical, near-term approach:
• Establish a grants program to fund a variety of activities that
support industrial waste reduction, such as technical assistance
and generic R&D, but do not fund specific waste reduction
efforts by individual companies.
• Enact new waste reduction legislation based on the multi-
media concept (i.e., deals with all wastes and pollutants in the
air, land and water), with expanded Federal reporting and plan-
ning requirements for industry.
• Establish reporting requirements on waste reduction for finan-
cial reports to the Securities and Exchange Commission.
• Create a new U.S. EPA Office of Waste Reduction with an
Assistant Administrator.
• Allow regulatory concessions (i.e., trade-off of certain limited
pollution control regulatory requirements for waste reduction
achievements).
• Create independent State Waste Reduction Boards to imple-
ment many of the new Federal initiatives.

These actions and perhaps others would be tantamount to
establishing a new waste reduction ethic for American society.
The result could be the elevation of waste reduction to a level
comparable to pollution control, giving America two ways to seek
more effective environmental protection.

REFERENCES
1. Office of Technology Assessment, U.S. Congress, Serious Reduc-
tion of Hazardous Waste: For Pollution Prevention and Industrial
Efficiency can be purchased from the Superintendent of Documents,
U.S. Government Printing Office, Washington, DC 20402, GPO
Stock No. 052-003-01048-8 for $12.
2. Summary booklets of the foregoing reports are available free from
OTA: Telephone (202) 224-8996.
334 WASTE MINIMIZATION
-------
Hazardous Waste Minimization by the U.S. Navy

Thadeus J. Zagrobelny, P.E.
Environmental Quality Division
Naval Facilities Engineering Command
Alexandria, Virginia
ABSTRACT
The Navy shore establishment generated over 240,000 tons of
hazardous waste in 1985 as a result of industrial operation to sup-
port Navy ships and aircraft. The Navy recognizes the need to
reduce the amount of hazardous wastes they generate to save
money now, to meet the U.S. EPA regulatory requirement to
minimize waste production and to avoid the long-term liability
associated with land disposal. The Navy program focuses on
proper hazardous material control, material substitution, pro-
cess changes, recycling and treatment. To support this effort, the
Navy is completing a comprehensive technology study to deter-
mine the exact wastes produced, the best current management
methods both within the Navy and in industry and research prior-
ities. A special Technology Transfer team is helping Naval in-
stallations take technology from the laboratory to the workplace.

BACKGROUND
In 1980 the Chief of Naval Operations (CNO) initiated a Navy
hazardous waste management program. The Navy goal was to en-
sure that Naval installations meet the requirements of the U.S.
EPA's national hazardous waste management program. CNO re-
quired all installations to develop hazardous waste management
plans that outlined where wastes were generated, who had dis-
posal responsibilities and how they would meet regulatory haz-
ardous waste disposal requirements. To ensure that any haz-
ardous wastes generated aboard ship were safely disposed, CNO
required shore activities to accept hazardous wastes from ships.
Certain installations were given the added duty to provide area-
wide services to take advantage of economies of scale and avoid
duplication of services.
Several factors have caused the Navy to expand the hazardous
waste management program to require hazardous waste minimiz-
ation:

• Disposal is expensive. Not only are the current disposal costs
significant, but also with increasing regulations, future costs
will increase markedly. Also, many land disposal facilities leak,
requiring costly correction. Even if the Navy contracts for
off-site land disposal of their work, it is liable for the wastes
and their effects forever. During the second half of fiscal year
1986, the Navy's disposal costs were up 60% and record dis-
posal costs are expected in 1987.
• Disposal is limited. With the 1984 Amendments to RCRA
tighter standards are reducing the number of disposal facilities.
The Navy's hazardous wastes must "compete" against other
hazardous wastes for limited national disposal capacity.
' The law requires it. RCRA now requires a generator to certify
on the hazardous waste manifest that he has acted to "mini-
mize the volume and toxicity of the waste generated to the
degree I have determined to be economically practical and that
I have selected the method of treatment, storage, or disposal
currently available to me which minimizes the present and fu-
ture threat to the environment."

HAZARDOUS WASTE GENERATIONS
As a part of the Navy's hazardous waste management pro-
gram, shore installations annually complete a Navy report that
details what waste materials they generate, treat, store and
dispose.
In 1985 Navy installations reported that they generated over
240,000 tons of hazardous waste. These figures include all wastes
that individual states may classify as hazardous, even if the U.S.
EPA does not regard them as hazardous wastes. This latter point
is significant, since some states require the management of waste
oil, including bilge water, as a hazardous waste. In addition, the
Navy generates over three times that amount of industrial waste-
water which requires treatment to avoid handling certain dilute
wastes as hazardous wastes. The total cost to the taxpayer for
Navy hazardous waste management, including internal handling,
treatment and contract disposal, was about $40M in 1985 (Table
1). With the 1984 Hazardous and Solid Waste Amendments to
RCRA increasing the costs of all aspects of hazardous waste man-
agement, the 1986 cost may exceed $50M and the 1987 cost prob-
ably will be more.
Table 1
1985 Navy Hazardous Waste Disposal Costs
Navy Disposal
DRMO Disposal for Navy
Navy Program Mgmt

Total
8.1M
9.8M
20.9M

$38.8M
Table 2
Top Ten Navy Installation Hazardous Waste Generators in 1985

1. PWC San Diego
2. NAS Jacksonville
3. NWIRP Bristol
4. NSY Long Beach
5. NWSEarle
6. NAVPRO Pomona
7. NASAlameda
8. NSY Portsmouth
9. NADC Warminster
10. NAS Corpus Christi
WASTE MINIMIZATION 335
-------
The major Navy generators of hazardous waste are Ship-
yards, Air Rework Facilities, Weapon Stations, GOCOs and
Public Works Centers (Table 2). These facilities are the large, in-
dustrial installations that are responsible for depot level mainten-
ance in ships, aircraft and weapons systems. The top 10 activities
generated 86% of the Navy's hazardous wastes. When looking
at who disposed the most waste (using the U.S. EPA's definition
of disposal to focus on land disposal), the top 10 installations
changed (Table 3). This change was a result of many of the top
generators operating their own treatment plants to reduce the
amount of hazardous waste that requires land disposal.
Table 3
Top Ten Navy Hazardous Waste Disposers In 1985

1. NSY Long Beach
2. NAVPRO Pomona
3. NASAlameda
4. NADC Warminster
5. NSY Mare Island
6. NSY Puget Sound
7. NWIRP Dallas
8. NAVBASE Norfolk
9. PWCPensacola
10. NOS Louisville
16% disposed
off-site
75% treated
on-site
2% treated
off-site ~ I

1% stored by
Navy
6% transferred
to DRMO
Figure 1
Navy Hazardous Waste Management in 1985

Table 4
Top Ten Categories of Hazardous Waste Generated In 1985

1. Petroleum wastes
2. HW, not otherwise specified
3. Cyanide cleaning baths (F009)
4. Miscellaneous Cyanide solutions
5. Corrosives (D002)
6. IWTP sludges (F006)
7. Cyanide plating baths (F007)
8. Sodium Nitrate
9. Paint
10. Ignitables(DOOl)
Figure 1 shows the fate of the hazardous wastes. The Navy
treated 75% of the wastes on-site in its own treatment plants.
Most of these wastes were dilute aqueous wastes contaminated
with oil, heavy metals or organics. The Navy directly contracted
for off-site disposal, normally burial, in a hazardous waste land-
fill, for 16% of its hazardous waste. The Navy sent 6% of its haz-
ardous wastes to the Defense Reutilization and Marketing Officej
operated by the Defense Logistics Agency (DLA). DLA gets rid
of the wastes by a variety of methods, including reuse, transfer,
donation, sales and disposal contracts.
The important waste streams are the 16% going to land dis-
posal and the portion of the wastes transferred to DLA that end
up in a landfill. This is the "bottom line" figure that the Navy'i
hazardous waste minimization program seeks to reduce, since thit
is the amount of hazardous waste that figures in the Navy's long-
term liability.
In their annual reports Navy installations do list the wastes that
they produce. The top 10 categories produced in 1985 are listed
in Table 4. One shortcoming of the data results from the Navy
using the same detail that regulatory agencies allow. For example,
the "N.O.S." label for not otherwise specified wastes is allowed,
but it does not help identify the specific waste or the exact pro-
cess that generates it. We addressed this problem through a de-
tailed technology assessment discussed later in this paper.

MINIMIZATION METHODS
There are a number of methods that the Navy uses to minimi??
hazardous waste generation. The techniques discussed in this sec-
tion are ranked in order of acceptability (Fig. 2). The first level of
minimization includes hazardous material control, material sub-
stitution and process change. If the process still produces a haz-
ardous waste after this step is taken, the next level of alterna-
tives includes recycling/recovery, treatment and thermal destruc-
tion. Eventually, a small residue may still require land disposal

Hazardous Material Control
Good hazardous waste control starts with good hazardous ma-
terial control. The Naval Air Rework Facility in Pensacola, Flor-
ida has a hazardous material control program that serves as an
example for the rest of the Navy. Only authorized users may ac-
quire hazardous materials and they may acquire only the specific
hazardous material that is needed for each job. The supply de-
partment issues hazardous materials in the minimum quantity
needed for the particular job. This procedure prevents hoarding
and the loss of good hazardous materials as their shelf life ex-
pires or workers are unaware that they have the hazardous ma-
terial on hand from a previous job and order more.
Bulk quantities are stored only in approved areas. Hazardous
materials have labels, and workers receive training in how to sale-
Materials Substitution
a
Process Changes
Recycle/Recover

J,
Treat to Non-HW

J.
Destruction

i
Disposal
Figure 2
Navy Hazardous Waste Minimization Flowchart
336 WASTE MINIMIZATION
-------
ly use hazardous materials and how to dispose of them as wastes.
The entire program is coordinated by a Hazardous Materials Con-
trol Committee. Finally, a base instruction documents the pro-
gram.
The Navy is developing a model program for other installa-
tions to use based on the Pensacola experience and strong haz-
ardous material control program at other Naval installations and
in private industry. In addition, the Navy is exploring developing
similar programs aboard ships to tighten the control of haz-
ardous materials and minimize the amount of material that turns
into wastes.

Material Substitution
The Navy sometimes can use alternate chemicals that are less
hazardous, cheaper or easier to recycle or treat. Material substi-
tution means using a substitute material without changing the
process. For example, most activities already use 1,1,1 trichloro-
ethane instead of trichloroethylene. More progress can be made
in this area. Major solvent generators are implementing programs
to recycle solvents. Another key part of the program is to use the
fewest number of solvents to allow for easier segregation. Final-
ly, the Navy is attempting to use the least toxic solvents available.

Process Change
There may be alternative processes that perform the job and
produce less or no hazardous waste. The Navy already is making
significant progress in two areas; aircraft paint stripping and hard
chrome plating. The Navy usually uses solvent stripping to re-
move paint from aircraft. A paint stripper, normally contain-
ing methyl chloride and/or phenols, is applied to the painted sur-
face and then washed off with water. The aqueous waste stream,
which can amount to 25,000 gal/aircraft, then goes to an indus-
trial wastewater treatment plant.
An alternative paint stripping process that shows great prom-
ise is plastic media stripping—"sandblasting" with a plastic grit
that is harder than the paint yet softer than the aircraft parts.
Another technology that the Navy is implementing is a zero dis-
charge hard chrome plating system. Instead of traditional plat-
ing following by treatment of the rinse water to reduce the heavy
metals concentration, the Navy is beginning to use a system which
uses a mist to rinse the parts, pipes the rinse water as makeup
water to the plating bath and then purifies the bath with a spec-
ial filter. This results in increased productivity, no remaining rinse
water to treat and no dumping of the plating bath.

Recycle/Recovery
Presently the Navy recycles and recovers some solvents and
oils and is planning to do more. The Used Solvent Elimination
(USE) program requires Navy installations to develop solvent
management programs. Consequently, many installations are in-
stalling solvent recovery stills that allow them to distill solvents
and reuse the recycled product, thus reducing both new material
and disposal costs. Many smaller installations are using commer-
cial firms that can provide a parts washer and clean (recycled)
solvent that is periodically removed to an off-site treatment loca-
tion.
Several large complexes collect and treat their waste oils and
oily waste waters. Recovered oil is downgraded for use in shore
boilers, saving significant amounts of virgin fuel.
Another popular recovery technique used by the Navy is the
waste exchange. One person's hazardous waste is donated to an-
other person, who then uses the waste as a feedstock for an-
other process.

Treatment
If it is not possible to change the process or recycle/recover
the waste, treatment is the next alternative. Treatment can in-
clude neutralization, solidification, volume reduction, biological
detoxification or other technologies.
Wastes, such as acids and bases, can be neutralized. The by-
product is often a nonhazardous salt and water.
Heavy metal sludges can be bound in a nonleachable matrix
and disposed of as solid waste. A filter press can squeeze water
out of a sludge, reducing its volume. Land treatment or extended
aeration wastewater treatment can use microbes to feed on the
hazardous wastes and detoxify them.

Thermal Destruction
Although the U.S. EPA calls thermal destruction a form of
treatment, it merits special mention. Normally, the Navy tries to
use the technologies described above before opting for thermal
destruction. Thermal destruction is expensive, complicated to
operate and places a significant administrative burden on the
user under the hazardous waste regulations. The Navy does not
have many thermal treatment systems now, but this may change.

PROGRAM MANAGEMENT AND
IMPLEMENTATION
It is one thing to list the hierarchy of management techniques;
it is quite another to actually do it in an organization as complex
as the Navy. The key is to make hazardous waste minimization
everyone's job. The Chief of Naval Operations is the overall pro-
gram sponsor for the Navy's hazardous waste minimization pro-
gram. In turn, his office has delegated specific responsibilities to
Echelon Two Commanders, who, in turn, require generating ac-
tivities to take action.
To support Naval installations with waste minimization tech-
nologies, the Chief of Naval Operations has asked the Naval Fa-
cilities Engineering Command to manage the overall minimiza-
tion program and provide field level support. The field level sup-
port is being provided through the Naval Civil Engineering Lab-
oratory (NCEL), the Naval Energy and Environmental Support
Activity (NEESA) and the Engineering Field Divisions.
NAVFAC is using centrally managed funds for hazardous
waste minimization program startup. This support includes fund-
ing for a comprehensive technology report discussed below, tech-
nology development, activity support and facility modifications/
construction.
NCEL has prepared a technology report of the Navy's cur-
rent hazardous waste generation and management practices, the
state of the art at other DOD activities and in industry and emer-
ging technologies that the Navy can apply. The technology report
identifies proven technologies for immediate Navy use and
recommends emerging technologies for further testing and eval-
uation by NCEL and other Navy laboratories.
The technology report also will be the basis for process mini-
mization factors that claimants will use to set minimization goals.
To transfer hazardous waste minimization techniques from the
laboratory to the field, NAVFAC formed a technology transfer
team at NEESA. The team will hold technology transfer work-
shops on successful technologies. NAVFAC's regional offices,
the Engineering Field Divisions, also will assist with implementing
projects demonstrated by the technology transfer team.

REDUCTION GOALS
The Chief of Naval Operations decided to require numerical
reduction goals for individual installations, based on the pro-
cesses that they use and technologies that are projected to be
applicable. The goals use 1985 generations as the baseline. The
NCEL technology assessment provides estimates on how much
each process can reduce the amount of hazardous waste gen-
WASTE MINIMIZATION 337
-------
crated or disposed. Each installation's hazardous waste produc-
tion is broken down by industrial process and the projected re-
ductions factors are applied to each process; the results are then
summed to provide an overall goal for the individual installa-
tion.

TECHNOLOGY ASSESSMENT REPORT
The technology report, which is more than a simple summary
of the annual reports, provides good insight into how the Navy
generates and manages hazardous waste. NCEL's work has been
documented in a two-volume "Initiation Decision Report" which
is a standard research and development report designed to assess
a technology, look at alternatives and propose a course of ac-
tion. The first volume includes an introduction, problem defi-
nition and a review of current hazardous waste management prac-
tices and technologies. The report looks at the industrial pro-
cesses that generate the most hazardous waste. The key chapter
discusses how Naval activities generate wastes and recommends
action that the Navy can take now as well as in the future as tech-
nologies evolve. The second volume will include technology pro-
jections, alternatives, recommendations for future actions, tech-
nology goals and capability goals.
NCEL visited installations that generate over 90% of the Navy
wastes; visits to Navy facilities were followed by visits to private
industry and regulatory agencies. The report of these visits identi-
fies not only the wastes the Navy generated, but also 33 processes
that generate hazardous wastes. The report then focuses on the 17
processes that generate 99% of the hazardous wastes produced by
Naval facilities and 92% of the reported costs (Table 5).
Each generating process is then subject to the hierarchy of man-
agement techniques (Fig. 2). If any hazardous wastes are pro-
jected to remain after the generating process is studied, the hier-
archy is applied to treating the hazardous waste. Similarly, if any
residue survives treatment, the hierarchy is applied to the ultimate
disposal step.
To help decide between alternatives, NCEL developed a tech-
nology assessment model. The model includes parameters for
logistics (such as procurement, operation, maintenance and
training), energy savings, volume reduction, the earliest date that
a technology will be ready, compliance with present and antici-
pated environmental laws and the degree of risk associated with
the technology. Each factor can be weighted to follow the user to
perform a sensitivity analysis. For example, "would an alterna-
tive still be chosen if logistics were a major problem?"
Table 5
Processes Generating Most Nivy Hazardous Wastes

I. IWTPs
2. Electroplating
3. Ordnance operations
4. Bilge water
5. Abrasive blasting
6. Painting
7. Demilitarization
8. Pipe flushing & cleaning
9. Boiler layup
10. Boiler cleaning
11. Fluids changeout
12. Solvent cleaning
13. Battery repair
14. Metal preparation
15. Bilge derusting
16. Chemical paint stripping
17. Torpedo operations
TECHNOLOGY TRANSFER
NAVFACENGCOM formed a technology transfer team for
hazardous waste minimization at the Navy Energy and Environ-
mental Support Activity (NEESA) in Port Hueneme, California,
The primary role of the technology transfer team is to transfer
new hazardous waste minimization technologies from the labora-
tory to the field. The technology transfer team's purpose if to
serve as an information center, review proposed minimisation
projects, demonstrate new technologies, transfer working tech-
nologies to the Navy and implement selected technologies.
The technology transfer team is not a library, but is a nucleus
of engineers and technicians whose goal is to reduce the Navy'i
hazardous waste generation. The technology report discussed
above lists the core technologies on which the team will focus.
Currently the technology transfer team has working knowledge
of hard chrome plating, plastic media blasting and solvent re-
cycling. The team is rapidly gaining knowledge about other min-
imization technologies. Since the Navy centrally manages iti
funds to start the program, NAVFAC also has asked the team
to analyze the proposed projects for technical feasibility.
As new technologies emerge the technology transfer team is
working with selected installations to demonstrate hazardous
waste minimization techniques. NEESA will have access to con-
tracts to adapt technologies for a site, install them provide activ-
ity training aid in trial operation, develop a list of lessons learned
and document the process. Once the team has demonstrated a
technology at a number of selected Naval activities, it will transfer
the working knowledge to Naval installations through a tech-
nology transfer process that will include publishing complete
user's packages, conducting seminars and workshops and recom-
mending changes to design criteria.
Hazardous waste minimization technologies generally will be
implemented in one of three ways: directly by activities, with
EFD assistance or with direct technology transfer team assis-
tance. The decision on which way to implement a technology will
depend on the complexity of the technology, degree of develop-
ment in the Navy, availability of "off the shelf" equipment and
funding constraints. Many activities could readily adopt some
technologies, such as solvent distillation, without EFD or tech-
nology transfer team help. Other technologies may be more com-
plex or we may want to quickly implement a technology through
one or several agents.

FUNDING
One key factor to the startup of a hazardous waste minimiza-
tion program is startup funding. There are several sources of
money available for program startup and continued operation.
First, NAVFAC has limited centrally managed funding available
as seed money for hazardous waste minimization. This funding
comes from "fenced" monies in an Environmental Restoration
account.
In fiscal year 1986, NAVFAC received $7.0M. It expects
S10.0M in fiscal years 1987, 1988 and 1989. This money is being
used for the technology assessment, technology development,
site-specific studies, equipment purchases and minor construc-
tion projects (construction that costs less than $200,000).
All requests for money are documented on a Pollution Control
Report that identifies the problem, proposes corrective actions
and gives a cost estimate. This money is intended to be seed
money only—there is no intention that this account will fund all
hazardous waste minimization.
Installation funding also is being used to implement the haz-
ardous waste minimization program. Many of the larger haz-
ardous waste generators, such as shipyards and air rework facil-
ities, are industrially funded, so they generate revenues from
338 WASTE MINIMIZATION
-------
"customers"—other components of the Navy that are respon-
sible for the operations of the ships or airplanes. These indus-
trially funded activities can use a portion of the funds that they
generate to pay for hazardous waste minimization programs.
If a construction project exceeds $200,000, it becomes a Mili-
tary Construction project, subject to a special set of approvals
that includes Congressional review. Since many of the Navy's in-
dustrial facilities date to the 1940s, the Navy must look at pro-
jects individually to determine if the projects should be classed
as hazardous waste minimization or plant modernization.
CONCLUSION
The Navy has established a framework for an effective haz-
ardous waste minimization program, including overall numerical
reduction goals, assessing current and developing technologies,
transferring the technology to the workplace and funding start-
up. Over the next few years, the Navy should significantly re-
duce the generation and amount of hazardous waste going to land
disposal.
WASTE MINIMIZATION 339
-------
Comprehensive Model for Hazardous Waste
Management Alternatives

William M. Sloan
Maryland Hazardous Waste Facilities Siting Board
Annapolis, Maryland
Jean Tilly
Stephen W. Bailey, P.E.
ICF Technology
Washington, D.C.
ABSTRACT
A computerized interactive data base for identifying most or all
the alternatives for managing a generator's hazardous waste has
been commissioned by the Maryland Hazardous Waste Facilities
Siting Board and developed by ICF Technology, Inc. The "Multi-
Option Model" provides practical information about waste re-
duction techniques, off-site recycling, commercial facilities, ap-
plicable waste-treatment technologies and cost of services.
The Board's knowledge of facilities planning and technical
assistance in Maryland and elsewhere was used to design the pro-
ject. For a number of reasons, many generators do not explore all
available alternatives. As a result generators do not use options
with lower cost and/or risk than their present methods of waste
handling/disposal. At the same time, the development of state-
government policies on new facilities suffers from data deficien-
cies that could be partly remedied by input from generators. The
input needed by the planner corresponds closely to the informa-
tion that the generator requires to make a reconnaissance of
alternatives. The technical information assembled in planning
contains much that could, with a little extra organizational effort,
be of practical use to the generator.
The Multi-Option Model is a computerized advisory system
that can be operated in the field on a portable computer. It pro-
vides information about management options "internal" and
"external" to the plant where waste is generated. Using waste-
type, waste-quantity and location data, the Model generates an
array of "external" management alternatives: (1) treatment and
disposal technologies applicable to the particular waste; (2) facil-
ities that have them; (3) engineered treatment costs; (4) recycling
possibilities; (5) brokers that handle the types of chemicals in-
volved; and (6) transportation cost to point of reuse, treatment
or disposal. Using information on product and unit processes,
the Model can assist "internal" management. The Model pro-
vides helpful hints to almost any generator wanting to reduce
waste. For some, the Model displays and prints process-specific
engineering. Technologies and engineered costs are applicable to
in-plant as well as to external waste treatment decisions.
The Model is "user-friendly" with respect to both computer
operation and data requirements. It is designed to require a min-
imum of sensitive information. The information that is entered
in the Model also would be used to indicate the realism of gov-
ernment waste-generation estimates and projections. The data are
not regulatory data and participation is not mandatory. Data can
be made confidential on request.
The Model will undergo in-house trial runs in January 1987,
and field testing and full application are scheduled for Spring
1987. There have been no insoluble problems in Model develop-
ment.

INTRODUCTION
Information from a variety of sources indicates a number of
differences between what most state governments consider best
or preferred industrial waste management and the actual practice
of industrial generators. Manifest statistics, for instance, have
shown that wastes which could be treated or incinerated were go-
ing to a landfill.
A tiny fraction, probably fewer than 5%, of Maryland's haz-
ardous waste generators list their "wastes" on an exchange at any
given time. Contact with waste-chemical brokers who promote
waste reuse indicates that the industry is immature and unable to
accommodate a large demand for services. Yet statistics of the
Northeast Industrial Waste Exchange show a one-in-five chance
of a successful exchange if the "waste" is advertised. And the
average exchange is worth $28,000.
Studies of in-plant waste management commissioned by the
Maryland Hazardous Waste Facilities Siting Board and staff con-
tacts over several years reveal many undeveloped opportunities
for internal at-source waste reduction. There clearly have been
many missed opportunities to improve waste management, at
least by government's rationale.
Government has its own problems assigning appropriate tech-
nologies. While working on generation estimates, Board staff
learned in a conversation that several thousand tons labelled
"lead" on the manifest were actually wooden pallets bearing
some lead-compound dust. The generator had listed the haz-
ardous constituent, but that failed to indicate a possible appro-
priate management technique. Classifications assigned to wastes
by the U.S. EPA often bore little relation to management op-
tions.
Some supplemental, independent, indicator of the realism of
statistically based estimates and projections was needed. Board
staff realized, when designing the third assessment of Maryland
treatment and disposal facility needs, that solutions to the two
problems had common elements.
Information developed in planning—which is characterized by
a broad view of alternatives—could be organized and presented
to industry as a practical tool to broaden the option a generator
could consider at small cost to the generator. A computerized
340 WASTE MINIMIZATION
-------
"expert system," called the Multi-Option Model, appeared work-
able.

Multiple-Alternative Planning and Data
Adequacy
In 1980 only a handful of government offices had addressed
the question of future commercial hazardous waste services.
Planning studies by New York State, the New England Reg-
ional Commission and the Delaware River Basin Commission
(including heavily industrial areas of Pennsylvania and New Jer-
sey) examined hazardous waste generation from the viewpoint
of treatment options. The U.S. EPA commissioned the "PHB"
report. There were few, if any, other studies. Estimates of gen-
eration were based on industry-base data, waste output estimated
from typical figures and some interviews.
In 1981 Maryland made the first planning estimates that used
regulatory data, using the manifests that were required to accom-
pany hazardous waste shipments beginning in 1978. Until super-
seded by the "U.S. EPA uniform manifest" in September 1984,
these forms contained a requirement for treatment-oriented
physical-chemical descriptions. Descriptions of this type appeared
on few, if any, other manifests. In sequential studies, data from
1980 (the first "reliable" year) and a study period of August 1981
through July 1982 were used in two assessments of generation
and facility needs. Generator interviews in connection with the
analysis of 1980 data were to detect influences and trends, not to
verify data.
Other states also began to analyze regulatory data as they be-
came available. Contact with generators was typical of these stud-
ies. These contacts generally confirmed that the type-codes pre-
scribed by the U.S. EPA were only vaguely indicative of man-
agement alternatives. The manifest and annual report data are
inherently limited, and direct contact provides a useful indepen-
dent check.
The objectives of state-level hazardous waste management
planning normally include defining needed facilities consistent
with best management practice. Since good practice begins with
at-source reduction but names landfill as a last resort, planning
embraces all the options.
Best management practice is widely accepted as beginning with
a four-level hierarchy of choices consisting of the following in
order of preference: (1) waste reduction; (2) recycling of low-
value byproducts; (3) treatment or incineration to reduce volume
or hazard by chemical change or decomposition; and (4) dis-
posal through deposition in a secure location. In general, the hier-
archy presents lower to higher-risk alternatives; in many in-
stances, choice of a higher-cost measure will mean less future ex-
posure. But lower actual cost is characteristic of reduction and
recycling. Economics and risk considerations will, in general,
lead the generator in the same direction.

Industry's Investigations Limited
Planning considerations appear to be broader than those of
most companies. Choosing a safe, legal, alternative, it seems,
can become complicated enough to discourage exploration of
measures that might be preferable in the long run. This appeared
especially true for smaller companies,—not merely the legally
defined "small-quantity generators," but any for which a high
level of effort in waste management is a significant diversion of
corporate resources.
While the Board had made management information devel-
oped in the cause of planning available to industry and public
since 1982, this release of data had been piecemeal. It became evi-
dent that structuring the filed information in order to uncover
and compare alternatives could materially assist many generators.
It also became evident that the input needed from generators
was the same information needed to define the State's facility
needs. The assistance features could dovetail neatly with data
acquisition. Computer technology could be applied effectively.

Waste-Reduction Advisory Funded by
U.S. EPA
The need for an introductory-level advisory on waste reduction
was indicated by studies commissioned by the Hazardous Waste
Facilities Siting Board in 1982. The project managers recom-
mended an introductory "walk-through" audit to correct obvious
problems and to scope further work. That is the basic purpose
of this proposed waste-reduction advisory. The U.S. EPA agreed
with this conclusion, and the waste-reduction component of the
project and the field interviews were funded through a grant
from Region III.
PURPOSE OF THE PROJECT
The main purpose of the Multi-Option Model is to help genera-
tors identify currently-available options—including, if the Model
is successful, some superior possibilities—in as all-embracing a
way as possible. "Superior" options might be less expensive in
immediate costs, or the options might involve a lower ultimate
risk and lower overall cost as true long-term "least-cost" op-
tions. "Superior" might also involve cost-risk trade-offs and
compromises within the firm, especially where contingent risks
were involved. The Model is to help identify options but not to
advise on corporate decisions.
Information is to be provided at an "introductory" or "re-
connaissance" level. "Introductory" information means a
description of capabilities, costs and results, with leads and refer-
ences for further screening. It is assumed that final choices will be
made on the basis of personal investigation; in fact, the Board
disclaims responsibility for decisions made from the model out-
put.
A main purpose of the Multi-Option Model from the viewpoint
of state government is to obtain the data needed to verify the
accuracy and realism of facility needs projections based on reg-
ulatory data. Since the assistance features and model mechanics
are the focus of this paper, however, this objective and the
"Types, Trends and Totals" Model to be used in planning new
units will not be developed further.
The Model is to be applied in the field following its develop-
ment. A target of 150 generator visits was set. This number was
judged by Board staff to be large enough to: (1) test the ap-
proach; (2) constitute a large enough proportion of Maryland
generators (10 to 20"%) that the technique would be visible;
(3) make up a reasonable sample to test estimate validity.
The planning aspects of the Model go beyond the verification
function mentioned above, however. The Model will illustrate
the waste-management implications of industrial location or pro-
cess-change decisions.
Identifying options will indirectly promote the four-level hier-
archy described in the introduction. The information available
through the Model will reduce the costs of exploring the options
and, thus, of making choices consistent with the hierarchy.
The Model holds promise as an educational tool. It could be
used by industry to demonstrate the range and consequences of a
decision to environmental or production personnel. It also could
be used to demonstrate options to citizens, government policy-
makers or students.
Future components could include directories of consultants and
specialties and compliance histories of facilities.
WASTE MINIMIZATION 341
-------
OVERVIEW OF THE MULTI-OPTION MODEL
The Model consists of two major components that apply "in-
ternally" to the plant at which the waste is generated and "ex-
ternally" to alternatives outside the plant of generation. Fig. 1 is a
sketch of the Model.

Measures Internal to the Plant
The internal component is concentrated heavily in source re-
duction, the most preferable approach. Source reduction is any
activity that reduces waste within a process. The source-reduc-
tion system is divided into: (1) a "general advisory" of helpful
hints applicable to almost any generator and (2) a number of ad-
visories (limited by budget and the experimental nature of the
work) with detailed help on particular industries or unit pro-
cesses. In-plant recycling is highly plant- and waste-specific and
appears in this version only incidentally with source reduction.
The waste treatment technologies of the "external" component
apply to in-plant treatment as well.

Measures Outside the Plant
Given the necessary data, the Model's expert system will pro-
vide guidance on:
• Applicable treatment technologies
• Facilities available
• Engineered costs
• Recycling opportunities
• Brokers that handle the particular waste
• Transportation costs
The Model was conceived as an innovative approach to acquir-
ing type of waste and quantity of waste data. In no instance
known to the authors is the generator given an immediate feed-
back on options as a result of providing data. The field applica-
tion of the Model will be designed to stress the immediate bene-
fits to the generator and the fact that the field representative will
have done his or her homework—including review of state
records applicable to the generator. The visit will not start cold.
This field work will be a strong boost to initial success. With a
positive history, the Model is expected to be accepted for use on
the generator's initiative.
Table 1
Waste Characterization Que» tlonnalre
5.

6.
Quantity (In 1 bs. or gallons or dru« per

Suspended Solids Concentration:
0-100 pp
100-1.000 «.
1.000-10.000 m
10,000-20.000 ff,
over 20,000 poi
pH:
0.00-6.00
6.00-6.00
8.00-M.OO
Heavy Metal Concentration:
Heavy Hetal

Cadali*
ChroBlu*
Lead
Hercury
Nickel
Silver
Zinc
Other
Cyanide Concentration (pp»):

Organic Constituents Concentration:
CO"C£HTRATIOR (pp»)
1.000-10.000 Over 10.
CONCENTRATIOR (w»)
0-1.000 1.000-10.000 0»er 10.000
Aroaallcs
Alcohols
Halogenated
Herbicides
Pesticides
Organic acids
Sol vents:
Halogenated
Non-ha1ogenate<5
Others

Heat Content (Btu/lb. or Btu/gallon):
Table!
Wute Exchange Lbtlng Form
MODEL INPUTS
The Model's external component queries the generator for:
(1) waste characteristics, (2) waste quantity and (3) location of
the actual generation. Given these, the Model can outline options
and calculate costs. Location obviously is simpler to obtain than
waste data, which are key and fairly complex. A principle of the
queries, however, is to keep input as simple as possible so the
work does not evolve into a data hunt.

Waste Characteristics and Quantity
Table 1 shows the waste characterization query that will appear
on the screen. It consists of multiple-choice questions, making it
easy for generators to characterize their wastes. A detailed com-
positional analysis of wastes is not required. Generators need to
know approximate concentrations of key constituents, generally
within one order of magnitude. The questionnaire requests in-
formation about the minimum number of parameters needed to
match wastes to technologies. Some parameters do not apply to
certain wastes. Information about the heat content of a dilute cy-
anide waste, for instance, is irrelevant. Quantity, while not direct-
ly involved in technology-matching, is used in cost calculation.
The standard waste exchange listing (Table 2) also will appear
on the screen. It is the key to the exchange-broker directory,
"wanted" listings and an automatic exchange listing.
O «uu O P
a r*UITt O
w » li i ii O COMTVMXn Q VAMMU O OM f

• O •» O M. O t« O C , OK« O T
o m a NO
342
WASTE MINIMIZATION
-------
Table 3
Multi-Optional Model Treatment Technologies

Batch Still Distillation
Solvent Extraction
Liquid Injection Incineration
Rotary Kiln Incineration
Cyanide Destruction
Chromium Reduction
Chemical Precipitation
Carbon Adsorption
Steam Stripping
Dewatering
Biological Treatment
Neutralization
Chemical Stabilization
Land Treatment
Deep Well Injection

Generator Location
The Model asks for zip code. The Model then calculates the
transportation cost for treatment, disposal or exchange. The
waste reduction component affects both technology-matching
and transportation calculations by changing characteristics and
reducing volume.

Waste-Reduction Inputs
Since the waste-reduction component was funded after the ex-
ternal-option model was underway, the inputs are presently under
development. Some plant and process information will be re-
quired. Experience will show the trade-offs between data mini-
mization and quality of the expert-system response.

MODEL MECHANICS
The generator's data activates data bases in the Model that con-
tain information on waste management alternatives. The data
bases are:

• Management technologies and the waste types to which they
apply
• Characteristics of treatment residuals
• Costs of management technologies and transportation
• Directory of commercial facilities
• Waste exchange listings
• Directory of recycling brokers
• General waste reduction advisory
• Process-specific waste reduction advisories
The following paragraphs describe each data base.

Management Technologies by Waste Type
A feasible management practice is a combination of treatment,
storage, transportation and disposal technologies that apply to a
specific waste. Shipping a waste to another manufacturing plant
through the waste exchange is also a feasible management prac-
tice.
Table 3 lists the treatment technologies included in the Model.
This version of the Model includes technologies commonly used
in waste treatment. (Subsequent versions might include those that
are well developed but less commonly used industrial processes
or treatment technologies with limited application, such as re-
verse osmosis and electrolytic decomposition.)
Matching waste streams to technologies relies on the following:
' Physical form of the waste stream, e.g., solid, sludge or liquid;
aqueous/non-aqueous liquid content
Internal to plant: General waste-reduction advisory;
process-specific reduction advisory; treatment technologies and costs.
External to plant: treatment technologies and costs; comercial
facilities available; transportation costs; recycling opportunities.
Characteristics
Quantities
Location
Exchange Data
Product and
Process
Figure 1
Multi-Option Model Layout
Waste
Characteristics
Treatment
residuals

equations
Model
adjusts
costs
based on
distance
and
capacity.
Figure 2
Cost Procedures
• Chemical properties
• Sufficient identification of hazardous constituents to be sure
that the technology will have useful effect
• Overall composition and proportion of various constituents
Based on these considerations, the Model uses a set of treat-
ment criteria or rules to define the suitability of a waste stream for
each treatment technology and transportation mode. In assessing
the feasibility of the various technologies for particular waste
streams, it is important to consider the overall chemical com-
position as well as the hazardous constituents to be treated. Treat-
ment technologies are not always specific to individual constit-
uents, and a given technology may consequently be a poor choice
for treating a hazardous constituent that is a small fraction of
the overall waste composition. For example, steam stripping an
aqueous organic waste stream is effective only if it enables sig-
nificant separation of hazardous from non-hazardous constit-
uents or recovery of a portion of the waste stream.
Characteristics of Treatment Residuals
The application of a treatment technology modifies the char-
acteristics of the waste stream. The Model often uses "treatment
trains" or combinations of treatment technologies to produce a
residual suitable for transportation and disposal. The effluent
from the initial treatment step is the influent to the following
treatment steps. Solidification commonly produces one solid
waste residue for successive treatment steps. Other treatment
technologies, such as distillation or steam stripping, may produce
two separate streams that are both hazardous. The Model char-
acterizes all these residuals and determines feasible management
practices to handle these residuals.
WASTE MINIMIZATION 343
-------
Costs of Management Practices
The Model has grouped the various costs involved in using a
technology into three categories:
• Capital investment costs
• Operation and maintenance costs
• Closure and post-closure costs
The Model's cost values are engineering estimates for con-
structing and operating hazardous waste management facilities.
They do not necessarily represent prices as determined by facility
supply and demand for facility services. Cost procedures are
shown in Fig. 2.
Engineering estimates of cost are useful in a number of ways
even if they cannot predict the generator's actual cost on a given
day. First, engineered costs with a commercially acceptable re-
turn built in are an indication of what that option ought to cost,
what effect a restricted or glutted market is having and hence
what to watch out for in cost trends. For planning purposes, in
government, comparison of engineered and market costs is an in-
dication of the cost of restricted facility development.
Capital investment costs are for land, equipment (trucks,
tanks, etc.), installation, auxiliaries (instruments, controls) and
related facilities (laboratories). The Model's cost estimates are
based on information on the type and size of equipment at
standard facilities and are accurate to within + /- 30%. Opera-
tion and maintenance costs include raw materials, labor utilities
and equipment maintenance. Closure costs are incurred for the
decontamination of facility equipment and structures and for
certification and supervision, as required by RCRA regulations.
The costs of using a technology vary according to the charac-
teristics of the waste (e.g., quantity, fraction of non-water com-
ponents, fraction of suspended solids and specific gravity). In
addition, the Model takes into account economies of scale real-
ized as these technologies are used to handle increasing quanti-
ties of waste. Therefore, the Model's cost estimates are in the
form of equations that relate costs to waste stream characteristics
and facility capacity.
The Model combines these costs into a present value that allows
a direct comparison between waste management practices. This
present value calculation relies on several parameters (i.e., the
inflation rate, discount rate and tax rate) that can be either speci-
fied by the user or set at some default values.

Directory of Commercial Facilities
The Model contains a directory of commercial facilities in
Maryland and neighboring states (Pennsylvania, New Jersey,
Delaware, New York, Ohio and Virginia). The directory also lists
facilities in Alabama and Michigan because many wastes gener-
ated in Maryland are shipped to facilities located in these states.
For all these facilities the directory lists types of wastes ac-
cepted and services provided. For any waste generated in Mary-
land, the Model relies on the directory to list the names of the
facilities that can handle the waste. The Model then calculates
costs of waste management based on the capacity of the commer-
cial facility and on the distance between the generator and the
commercial facility.

Waste Exchange Listings
The Northeast Industrial Waste Exchange maintains both
printed bulletins and on-line listings of materials available and
materials wanted, which the Model will duplicate. Input of ex-
change listing data will provide an immediate matchup if there it
a "wanted" ad. The data also will be transmitted to the exchange
as an "available" listing. An exchange listing cannot, as with
other options, provide immediate response.

Directory of Recycling Brokers
The exchange-data input will also activate the directory of re-
cycling brokers. Names and addresses of recyclers, whose special-
ties are keyed to the waste-exchange classification, will be printed
out.

Waste-Reduction Advisory
This data base will present and categorize common source-re-
duction options applicable to most manufacturing plants. It will
present, in effect, a guide for a self audit for in-plant waste reduc-
tion and suggest direction for process-specific studies.

Process-Specific Waste Reduction Advisories
These may be oriented toward a particular manufacturing pro-
cess or, more likely, toward unit processes common to more than
one manufacturing area. This data base will contain characteris-
tics and limits on the manufacturing technology, potential mod-
ifications of operation, feedstock or equipment and possible sub-
stitute equipment.
CONCLUSIONS
The conclusions that can be drawn at this time (December
1986) are based on the work of developing the technical back-
up and logic of the Multi-Option Model. Thus far, the "user-
friendly" approach appears quite feasible. "User-friendly" refers
not only to the computer (it is pretty well accepted that interactive
programs can be made easy to use), but also to the data elements.
Project designers felt it important to convince the generator-man-
ager it was worthwhile to set at a keyboard for an hour. Field
testing will indicate the success of this approach. The field inter-
viewer will stress immediate benefits to the generator.
The main problem thus far in model development is micro-
computer memory. This appears soluble in this version. It could
recur, however, with new components.
The waste reduction work has not started in earnest yet.
Contact with the generator has barely been initiated. The re-
sponse has been "friendly caution." Protection of proprietary
information is an issue. Trade data can be kept confidential,
attorneys advise. The preferred approach is to avoid collecting
sensitive data.
By the time the paper is presented the Model will have had some
trial runs and more in-depth contact will have been made with the
generator community.
344 WASTE MINIMIZATION
-------
Characterization and Segregation of Waste Oils,
Solvents and Fuels at Naval Installations
Alan J. Kaufman
Black & Veatch, Inc.
Asheboro, North Carolina
Kendall M. Jacob, P.E.
Black & Veatch, Inc.
Greenville, South Carolina
ABSTRACT
Numerous site investigations at Naval installations across the
United States have been conducted by Black & Veatch, Inc. The
objective for the investigations was to provide the Navy with im-
proved waste management alternatives for oils and solvents. In-
itial inventory quantification and tabulation of the specific waste
volumes generated at each Naval base were followed by selections
of options to provide resource recovery, reclamation, recycling,
segregation and conservation of raw materials.

INTRODUCTION
The Department of Defense (DOD), like any industrial com-
pany, generates a variety of wastes. As in any complex industrial
sector, proper waste management results in lower disposal costs
and conservation of valuable materials. This paper details the
types of Naval activities that generate petroleum-based wastes,
types of materials typically generated, present typical waste
management practices and improved methods for characterizing
and segregating the materials.
TYPES OF FACILITIES
AND THEIR OPERATIONS

The generation of oils and solvents at any Naval facility
depends on the types of activities in which the base in engaged and
the size of each of its operations. The primary mission of the
Navy is to maintain, in a constant state of readiness, the defense-
oriented equipment for which it has responsibility. Defense-
oriented equipment can include such items as ships, boats,
airplanes, helicopters and automotive components. These types
of equipment are found at a variety of Naval activities. Some ac-
tivities have very specialized functions, while others have general
applications.
Some of the specialized activities include Naval Air Rework
Facilities (NARFs), Naval Air Stations (NASs) and Naval
Shipyards (NSYs). These activities typically include maintenance,
repair, rebuilding and fueling/defueling operations on aircraft
and ships. Activities having more diversified operations include
Public Works Centers (PWCs), Naval Stations (N A VST As) and
Shore Intermediate Maintenance Activities (SIMAs). These ac-
Table 1
Large Volume Oils Used at Various Naval Activities
HIL-L-2104 hIL-L-9080 KIL-L-2105 BRAND NAHE HIL-L-6081 HIL-L-17331 IUL-L-23699 HIL-L-46152 KIL-H-6083 HIL-H-83282 HIL-L-17672
GAS ENGINE DIESEL GEAR RISC AIRCRAFT SIEAH AIRCRAFT ENGINE HYDRAULIC FIRE LIGHT
ENGINE IURBINE TURBINE IU8BOSHAFI EQUIPMENT RESISTANT HYDRAULIC
LOCATION
NAVAL AMPHIBIOUS BASE (VA> 16000 13500

NAVAL AIJ REWORK FACILITY (VA) 11700

NAVAL AIR STATION (VA)

NAVAL STATION (VA) 4500 4800

NAVAL SHIPYARD (VA) 15000

MARINE CORPS AIR STATION (HI) 17800

NAVAL Alt STATION (HI) 3700

NAVAL STATION (HI)

NAVAL SHIPYARD (HI)

PUBLIC UORKS CENTER (HI) 5600

SHORE INTERMEDIATE MAINTENANCE (Hi) 2800

NAVAL AIR REUORK FACILITY (FL) 3000

NAVAL AIR STATION (FL)

CONSTRUCTION BATTALION (FL) 1900
2600
3300

47900
22800
3100
35500
3000
60900
4700

7800
4000
9000
11900

10900
3100
3800
135600
3G300
29600
3000
26000
5700
5200

1800
11500

2200
WASTE MINIMIZATION 345
-------
tivities typically include automotive and boat repair, maintenance
and rebuilding. Also, they provide support to the more special-
ized activities.

WASTE MATERIALS GENERATED
Three major types of petroleum-based wastes typically are en-
countered at any Naval base: oils, solvents and fuels. These
materials are classified as petroleum-based materials since they
are organic substances which contain carbon, hydrogen and
sometimes oxygen, nitrogen and one or more of the halogens.

Oils
Oils are formulated to provide improved viscosity, thermal
stability, antiwear, detergency, oxidation and corrosion inhibi-
tion. They also are used to maintain other mechanical properties
in machinery and engines. The Navy maintains a large number of
Military Specifications (MIL SPEC) for oils used in varying ap-
plications. These oils typically are used in automotive, hydraulic,
airplane, helicopter and ship applications. Some specific examples
of the MIL SPEC oils used at various Naval bases are provided in
Table I.
NARFs generate large volumes of a number of oils. These in-
clude oils for gas engines, aircraft turbines and hydraulic equip-
ment. NSYs generate large oil volumes from gas engines and me-
chanical systems. NASs generate their largest oil volumes from
aircraft applications. Smaller volumes are generated by PWCs,
SIMAs and SEABEES. Table 2 lists some of the actual sources of
oils generated at Naval facilities.

Table 2
Navy Sources of Used Oils
Storage and Distribution Areas
Power Plants
Deballasting Piers
Engine Repair Shops
Marine Terminals
Public Works Transportations
Shops
Oil Spill Cleanup Operations
Hobby Shops
Oil/Water Separators and
Grease Traps
Aircraft and Vehicle Washracks
Drydocks
Service Stations
Engine Test Cells
Machine Shops
Aircraft and Vehicle Maintenance
Shops
Aircraft Training Operations
Wastewater Treatment Plants
Ship Bilge Dewatering Operations
Oil SWOB Operations
DONUT Operations
Solvent*
Similar to the wide variety of oils, the Navy also uses a wide
variety of solvents. Typically, solvents are used for cleaning, de-
greasing, painting and equipment testing. Solvents are used for
these purposes due to their volatilities, evaporation rates and
good dissolving capabilities.
Two general classes of solvents exist; those that contain halo-
gens (chlorine or fluorine) and those that do not. The halogenated
solvents have high volatilities, rapid evaporation rates and low
flammabilities. These properties make them ideal for their use in
degreasing and high grade cleaning.
Nonhalogenated solvents are flammable and their volatilities
vary according to the specific solvent. Nonhalogenated solvents
including acetone, methyl ethyl ketone and ethyl acetate have
high volatilities; ethanol, toluene and xylene have intermediate
volatilities; and Stoddard Solvent has low volatility. These
solvents are used for degreasing, paint thinning and paint
removal. In some cases, these two types of solvents are found
mixed together or with other constituents to supplement their pro-
perties.
The process of solvent cleaning and degreasing serves to
eliminate unwanted oils, greases, carbon, dirt and corrosion pro-
ducts. The solvent removes the contaminant by solubilmng it
from the surface to be cleaned. Solvent cleaning usually is used to
prepare the cleaned item for subsequent operations such as repair-
ing and painting.
There are three types of solvent cleaning which typically are
performed at a Navy facility: cold cleaning, vapor degreasing and
precision cleaning. Cold cleaning is the most readily used type of
cleaning. The solvent is applied either by rag or brush or in a sol-
vent soak tank. The most commonly used solvent for cold clean-
ing is a mineral spirit solvent designated at P-D-680, Stoddard
Solvent or Varsol™ which is a petroleum distillate cut with a
wide boiling range making it one of the least expensive cleaning
solvents. Other solvents used for wipe cleaning include methanol,
ethanol, methyl ethyl ketone, ethyl acetate, toluene and naphtha.
Vapor degreasing entails the vaporization of a nonflammable
halogenated solvent to provide a cleaning medium. Steam or elec-
tricity is used to heat solvent in a tank to its boiling point. As the
solvent vapors rise in the tank, they are condensed by water or
refrigerated cooling coils. The item to be cleaned is suspended in
the solvent vapor. The washing action of the vapor and liquid on
the item provides the cleaning action. Several solvents are com-
monly used in vapor degreasers: 1,1,1-trichloro-ethane, perchlor-
oethylene and l,l,2-trichloro-l,2,2-trifluoroethane (Freon 113).
The Navy previously used trichloroethylene; however, the use of
this material was suspended due to its cancer-causing potential.
The halogenated solvents used for vapor degreasing are more
expensive than those described previously by a factor of between
two to four.
Precision cleaning is used where a high degree of cleanliness is
required. Applications requiring these high standards of contami-
nant removal include the cleaning of electronic components and
related precision instrumentation. Freon 113 almost always is used
in this application. It can range in cost from four to five times that
of the cold cleaning solvents.
A variety of solvents is used for paint removal and thinning
operations. Paint removal from large aircraft surfaces is accom-
plished by spray or dip tank application using a methylene
chloride (a chlorinated solvent) formulation. This solvent for-
mulation can contain acid, caustic and phenols or chromates in
addition to the solvent. Smaller scale paint removal for cleaning
and spot applications is done with nonhalogenated solvents such
as naphtha, toluene, xylene, ethyl acetate, acetone, methyl ethyl
ketone, methyl isobutyl ketone and various formulations of these
solvents.
These same solvents are used for paint thinning. The specific
solvent or formulation used is dependent on the type of paint in-
volved.
Table 3 provides a list of solvents, their uses and the Naval ac-
tivities where they are used.

Fuels
The fuels at Naval bases are used to provide energy sources to
power ships, boats, aircraft, portable equipment and automotive
vehicles. Ships, boats and portable equipment use Marine Diesel
Fuel (DFM); trucks use automotive diesel fuel; aircraft use
AVGAS, JP-4 or JP-5. Automobiles presently use unleaded
gasoline designated as MUS. These fuels have either high flash
points or low flash points. A fuel with a flash point below 140°F
has a low flash point. JP-4, AVGAS and MUS are all low flash
point fuels, while DFM, automotive diesel fuel and JP-5 are high
flash point fuels.
Most of the waste fuel at Naval installations originates from
spillage, defueling and cleaning of fuel tank operations. In some
cases, these fuels may be reused directly if they have not been con-
taminated. Table 4 provides a list of sources of waste fuels.
346 WASTE MINIMIZATION
-------
Table 3
Solvents, Their Uses and Locations Used
Solvent Name
(Fed Spec or Mil Spec)
Acetone (O-A-51)
2-Ethoxyethanol (TT-E-776,
TT-E-781)
f Carbon Remover (Mil-C-16173,
Mil-C-19853, P-C-11)
Dry Cleaning Solvent (P-D-680)
tEpoxy Stripper (Mil-R-81294)
Ethanol (O-E-760, TT-E-489)
Ethyl Acetate (TT-E-751)
Lacquer Thinner (TT-T-266,
A-A-857, Mil-T-19544, Mil-T-
60%, Mil-T-81772)
Methyl Ethyl Ketone (TT-M-261)
Naphtha, Aliphatic (TT-N-95)
Paint Thinner (TT-T-291)
Perchloroethylene
Petroleum, Ether(O-C-265)
Toluene (TT-T-548)
Toluene/Methyl Isobutyl
Ketone (Mil-T-19588)
1 1,1,1-Trichloroethane (O-T-620,
MU-T-81S33, Mil-T-81599)
t l,l,2-Trichloro-l,2,2-trifluoro-
ethane(Mil-C-81302)
Xylene (TT-X-915)
Navy
Application
Thinner-paint,
lacquer resin
Thinner-paint
Engine cleaner
Cleaner/degreaser
Epoxy paint
remover
Cleaner/degreaser
Cleaner/thinner-
paint
Cleaner/thinner-
paint
Cleaner/thinner-
paint
Cleaner/thinner-
paint
Cleaner/thinner-
paint
Dry cleaning,
vapor degreaser
Cleaner/degreaser
Cleaner/thinner-
paint
Cleaner/thinner-
paint, lacquer
Cleaner, vapor
degreaser
Precision
cleaner/
greaser
Cleaner/thinner-
paint, lacquer
Activity
NARF, NAS, NSC, NSY,
SUBASE
NARF, NSY
NARF, NAS, NAVSTA, NSY,
SIMA
NARF, NAS, NAVSTA, NSC,
NSY, NWS, PWC, SIMA,
SUBASE
NARF, NAS, NSY, PWC
NARF, NAS, NSY, NWS
NARF
NAB, NARF, NAVSTA,
PWC, SUBASE
NARF, NAS, NSY,
SUBASE
NARF, NAS, NAVSTA,
NSC, SIMA
NAB, NAS, NAVSTA, NSY,
NSC, NWS, PWC, SIMA
NAB, NAVSTA, NSY
NSC
NARF, NAS, NAVSTA,
NSC, NWS
NARF, NAS, NSY
NAB, NARF, NAS, NAVSTA,
NSY, PWC, SIMA
NARF, NAS, NSY,
SUBASE
NAS, PWC, SUBASE
Abbreviations:
NAB — Naval Amphibious Base
NARF — Naval Air Rework Facility
NAS - Naval Air Station
NAVSTA - Naval Station
NSC - Naval Supply Center
NSY - Navel Shipyard
NWS - Naval Weapons Station
PWC - Public Works Center
SIMA — Shore Intermediate Maintenance Activity
SUBASE — Submarine Base

t Halogenated Solvents
Table 4
Navy Sources of Used Fuels
Fuel Farms
Fuel Storage Areas
Fuel Depots
Aircraft & Ship Defueling Facilities
Power Plants
Piers
Engine Repair Shops
Fire-Fighting Schools
Marine Terminals
Public Works Transportation Shops
Aircraft & Vehicle Washracks
Drydocks
Service Stations
Engine Test Cells
Aircraft & Vehicle Maintenance Shops
Ship Bilge Dewatering Operations
CURRENT WASTE MANAGEMENT PRACTICES
The waste management practices currently employed at Naval
facilities have been determined by numerous field investigations
ty Black & Veatch, Inc. The majority of the site visits have con-
centrated on shoreside practices; however, some shipboard prac-
tices also have been examined. The purpose of each field survey
was to determine the following:

• The petroleum products and volumes of usage and disposal
from each base's tenant activities
• The details of how each product is managed
• The responsible organizations within the current petroleum
product management system
• The costs associated with the current management system
The above information was collected through interviews con-
ducted at various personnel levels ranging from upper manage-
ment to laborers. In addition, record searches were conducted to
obtain information not otherwise available or to verify the infor-
mation obtained during the interviews. The information was
tabulated and was comprised of the following:
• The building and shop where the specific materials were used
• Material name; brand names were entirely capitalized; manu-
facturers' names were included in capitals and parentheses
where appropriate
• The Military of Federal Specification (MIL-SPEC, FED-
SPEC) Numbers, the National Stock Number (NSN), corre-
sponding container size and major constituents of the mater-
ials where appropriate
• The manner in which each material was used as well as the
losses associated with the usage
• An average of the annual usage and disposal volumes obtained
as average weekly, monthly or annual volumes and equated to
annual volumes by the maximum and minimum values; activi-
ties that reported disposal volumes less than 20 gal/yr were
identified as insignificant.
• Disposal information with particular emphasis on the method
of disposal, materials combined, labeling of the disposal con-
tainer, location of waste receptacle, activity involved in col-
lection of materials, temporary storage or holding area and
AREA
Figure 1
Typical Used Oil, Solvent and Fuels Management
WASTE MINIMIZATION 347
-------
final disposal method; waste materials which were combined
are indicated by brackets; individual entries indicated the ma-
terial was segregated at the time of pickup.

After all the information had been tabulated, a flow chart was
devised to pictorially represent the management of petroleum
waste at the Naval base. Fig. 1 represents part of a typical flow
chart which could be expected at an activity.
Typically, no segregation is practiced when wastes are
generated from a shop-level function. Oils, halogenated solvents,
nonhalogenated solvents, fuels and aqueous wastes are all com-
bined in a drum or another vessel receiving wastes for disposal.
This mixing can prevent, in many cases, the reuse or recovery of
materials. This practice usually provides no alternatives other
than disposal by private contractor. In addition, the mixture of
various solvents, especially halogenated solvents and other
materials such as oils, can make an entire volume of waste a
hazardous waste even though the oils or other materials alone
may not normally be hazardous wastes. Instead of having the
ability to realize a monetary gain through the sale of oil, there is a
need to expend considerable amounts of money to have the wastes
properly disposed of in an environmentally safe manner.
In the past some of the wastes that were generated by various
activities were land disposed in surface impoundments and other-
wise land applied for grass control and road oiling. This practice
is now the exception rather than the rule, since many people are
now environmentally aware of the consequences of these types of
actions. In addition, the Chief of Naval Operations in April 1985,
issued a memorandum banning the landfilling of waste solvents.
This memorandum can be directly attributed to the Hazardous
and Solid Waste Act of 1984 which, in part, detailed the ban of
the landfilling of certain solvents.
Some other types of disposal included discharges to industrial
and municipal sewer systems, containerized storage for long
periods without regard to environmental regulations, treatment
and recovery of reusable commodities, use as boiler fuel for
energy recovery, donation to other government activities for
reuse, sale to private enterprises for monetary reimbursement, use
for fire-fighting training and combination with industrial trash.
Most wastes were accumulated in 55-gal drums, portable tanks
(bowsers) and underground storage tanks. In most cases, these
vessels were left either uncapped or unsecured, thus allowing
undesirable wastes to enter or unauthorized discharges to occur.
In the past, containers were not labeled, which prevented the
identification of the character or origin of the wastes. Drums were
allowed to remain exposed to weathering, causing rust accumula-
tion and eventual container deterioration.
More recently, due in part to stricter environmental regula-
tions, most of the labeling, lack of drum control and long waste
storage period problems are beginning to be addressed. The De-
fense Logistics Agency (DLA), through the local Defense Reutili-
zation and Marketing Offices, is requiring identification of drum
contents and their generators. Storage facilities at many Naval
bases are being upgraded or being newly constructed to comply
with regulatory requirements for hazardous waste storage. If
labeling requirements are not being met, Navy facilities are find-
ing many disposal firms will not transport their wastes. When
state environmental authorities inspect hazardous waste storage
facilities and inadequacies or improper disposal practices are un-
covered, fines are levied and corrective measures with specific
time frames are imposed.

IMPROVED CHARACTERIZATION AND
SEGREGATION OF WASTES
The Military Specification (MIL SPEC), Federal Specification
(FED SPEC) and National Stock Number (NSN) enable any
Naval activity to identify the constituents in the products it use*.
Useful information from each manufacturer or supplier of a com-
modity may need to be requested and kept on hand. This infor-
mation would not only enable the Navy to characterize its waste
product generation, but also comply with right-to-know regula-
tions and provide for the segregation of its wastes.
The characterization of waste probably will have to be con-
ducted by persons familiar with chemical constituents. These in-
dividuals would educate shop-level personnel and their manage-
ment concerning the safety precautions necessary to work with
the chemicals. Three of the most common precautions workers
need to be aware of are eye protection, skin protection and proper
ventilation. In addition, the knowledge workers gain in under-
standing the basic differences in chemicals along with a segrega-
tion program can help Naval activities reduce their generation of
wastes, reuse used materials and recover their economic value.
Black & Veatch has formulated some basic recommendations
for the Navy to improve its management of hazardous wastes.
The primary impetus the Navy has in reducing its generation of
hazardous wastes includes: reducing the costs for waste handling,
hauling and disposal; reducing the concern for future liability and
groundwater contamination; conservation of valuable com-
modities for reuse; and recovery of energy values.
One of the most efficient ways that the Navy can reduce its
hazardous waste generation is segregation of its wastes. Waste
materials should not be mixed together unless they are pan of an
approved management system and are compatible. The materials
should remain as uncontaminaled as possible by excluding water
or other foreign matter. The container designated to receive
specific wastes should be clearly labeled to identify the contents.
Labeling should include the material name, the generator and an
indication that it has been used. In many situations, color-coding
of different material drums will assist all the personnel involved in
handling the wastes.
Black & Veatch suggested one possible segregation scenario for
waste oils, solvents and fuels. The suggestion was to keep
petroleum oils, fuel and hydraulic fluids; synthetic oils and
hydraulic fluids, halogenated solvents and nonhalogenated
solvents segregated as completely as possible. Red, white and blue
RED
NONMALOOtNATCO
SOLVENTS
WHITE
HALOGENATED
SOLVENTS
PETROLEUM
OILS 'HI FLASH FUELS
GREEN
SYNTHETIC
OILS
PURPLE
Figure 2
Color Coding of Waste Drums
348 WASTE MINIMIZATION
-------
painted drums were to be reserved for nonhalogenated solvents,
halogenated solvents and petroleum oils, respectively. Freon was
to be accumulated in purple drums, while synthetic oils would be
stored in green drums. Fig. 2 provides an illustration of this
scenario.
Other suggestions were provided to the Navy to improve
management of wastes. Where high cost halogenated solvents
were lost due to evaporation, retention of the solvent could be
minimized by using covered containers, refrigerated vapor de-
greasing or substituting another less costly solvent where feasible.
Proper equipment maintenance and usage practices also were
established to retain these solvents. The color-coded containers
should be collected for on-site recycling or contract recycling off-
site by qualified vendors. Those used materials which are
amenable for use as boiler fuel should be burned for energy
recovery.
Oils, solvents and fuels should be containerized, and disposal
into sanitary sewers or trash dumpsters should be avoided; ab-
sorbents should be kept on hand to avoid accidental spills from
entering sanitary sewers. Foreign materials should be excluded
from accumulation drums; screened funnels can provide separa-
tion of large solids.
Since a large number of solvents used by the Navy have high
volatilities, drums should remain tightly sealed and should be
opened only when being emptied or filled. Spigots or drum pumps
usually can provide efficient solvent delivery. Drums also should
be closed tightly to prevent rain water from entering virgin or used
recyclable materials.

CONCLUSIONS
There are numerous opportunities for the Navy to improve its
management of petroleum-based wastes. It is no longer accept-
able to dispose of wastes indiscriminately in the most convenient
manner without regard for environmental concern. In the past,
not much thought was given to the economics of waste disposal
since it was relatively inexpensive; however, present environmen-
tal regulations have put high costs on disposal as well as penalties
for improper disposal practices. Expensive and larger volume
solvents can be recycled and reused repeatedly by distillation.
Some solvents and oils may be used to recover their energy con-
tent in lieu of disposal. The keys to successfully implementing im-
proved wastes management are segregation at the source of
generation and the willingness to change practices.
REFERENCES
1. Boubel, R.W., "Recovery, Reuse and Recycle of Solvents," Defense
Environmental Leadership Project, Washington, DC, Dec. 1985.
2. Higgins, T.E., "Industrial Processes to Reduce Generation of Haz-
ardous Wastes at DOD Facilities," Phase 2 Report, CH2M Hill for
the DOD Environmental Leadership Project, Washington, DC and
U.S. Army Corps of Engineers, Huntsville, AL, July 1985.
3. Kaufman, A.J., Jacob, K.M. and Smith, W.G., "Final Inventory
Report—Used Oil & Solvent Recycling Management Study," Black
& Veatch, Inc. for the Naval Energy and Environmental Support
Activity, Contract No. N62474-84-C-3387, Port Hueneme, CA, Mar.
1986.
4. Kaufman, A.J., Jacob, K.M. and Smith, W.G., "Draft Inventory
Report—Hazardous Waste Minimization and Used Oil and Solvent
Elimination Program," for the Naval Facilities Engineering Com-
mand, Southern Division, Contract No. N62467-85-D-0733, Oct.
1986.
5. Kaufman, A.J. and Spake, Y.E., Jr., "Final Report—Used Petroleum
Products Management Evaluation," for the Naval Facilities Engi-
neering Command, Atlantic Division, Contract No. NG2470-81-C-
3811, Aug. 1984.
6. Salvesen, R.H., "Used Oil and Solvent Recycling Guide—Final Re-
port," R.H. Salvesen Associates for the Naval Energy and Environ-
mental Support Activity, Contract No. N62474-84-C-3377, Port
Hueneme, CA, June 1985.
WASTE MINIMIZATION 349
-------
Faster Sample Throughput for
Environmental Analyses

James A. Poppiti, Ph.D.
Finnigan MAT
San Jose, California
ABSTRACT
Increased throughput in an environmental laboratory is a func-
tion of productivity and capacity. Increased productivity is
accomplished by changing laboratory operating parameters and
by automating equipment. Capacity increases usually are accom-
plished by adding equipment and personnel.
This paper examines the implementation of increased produc-
tivity and capacity in an environmental laboratory as they effect
overall sample throughput.

INTRODUCTION
It is generally accepted that the way to increase output of any
laboratory is to either increase productivity (i.e., analyze more
samples in a given time without significantly increasing resources)
or increase capacity. Increased productivity can be accomplished
by changing basic operations and by automating equipment. In-
creased capacity usually involves adding equipment and per-
sonnel to handle higher sample volumes. Increasing the capacity,
however, may also involve physically expanding or remodeling
the laboratory.
Sample flow through environmental laboratories involves a ser-
ies of sequential tasks. Samples must be received and logged in.
Chain-of-custody documentation usually is required along with
routing information listing what analyses are to be performed.
Once these tasks are accomplished, the samples can be homogen-
ized, aliquoted, extracted, digested or otherwise prepared for the
specific analytical technique that will be employed. After the
samples are prepared, they are analyzed and the analytics per-
formed. These results must then be validated to ensure data in-
tegrity and quality. Finally, a complete report is generated.
These tasks must be performed sequentially. That is, a sample
can be analyzed only after it has been prepared, and a report can
be prepared only after the analysis has been finished, the data val-
idated, etc. Sample preparation, analysis and validation usually
take longer than log-in and final reporting and are therefore crit-
ical to high throughput. While it is true that the overall through-
put is a function of all of the steps in the sequence and that in-
creased throughput in any one step does not automatically mean
that the overall throughput will increase, increased productivity
or capacity of any one step is a prerequisite to increased overall
throughput.
The major focus of the following discussion centers on in-
creasing throughput in sample analysis and data validation, since
both of these are crucial for increasing overall sample throughput.

INCREASING PRODUCTIVITY
Real increases in analysis and data validation speed can be
achieved by altering laboratory procedures and automation. A re-
cent example of increasing productivity by changing procedures
involves the analysis of volatile organic compounds.
The routine method for determining volatile compounds is a
purge and trap technique followed by packed column gas chrom-
atography. Recently, however, the U.S. EPA proposed using a
new capillary column method that shortens analysis time by ap-
proximately 30%. Table 1 provides the analysis times given in
the new method as well as a comparison of analysis times from
several other sources.

Tiblel
AnatjnJs Times for Volatile Compounds by Capillary ud
Picked Columns
Method
U.S. EPA Method 524.2
60 m VOCOL
30 m DB624
30 m DB5

U.S. EPA Method 524.1 (624)
1 mSPlOOO l^i

Dreisch and Munson
30 m SE54
1 mSPlOOO 1*»

Sievenson
30 m DB624
Tlmtdnla.)

28
25
19,

30
17.5
37
14
Although there are differences in columns, flow rates and
temperature programs used in Table 1, the analysis times for vol-
atiles on capillary columns are substantially lower than for packed
columns. We might also expect shorter analysis times if the car-
rier used were hydrogen instead of helium.
Another area that has received attention lately involves the use
of fast liquid chromatographic separations to achieve very short
analysis times for environmentally significant compounds. Phen-
ols, substituted acids and several other toxic compounds have
been analyzed using Thermospray LC/MS. The analysis time for
these compounds is usually less than 10 min and offers increased
specificity over the standard UV detector.
Improvements in detection systems can be used to increase
sample throughput. Many environmental laboratories now rou-
tinely use inductively coupled plasma (ICP) spectrometers in
place of atomic absorption (AA) spectrometers. The ICP's abil-
ity to simultaneously determine 10 to 20 elements has unquestion-
ably increased the productivity of many inorganic laboratories.
350
WASTE MINIMIZATION
-------
New Techniques
In a similar manner, newer techniques in mass spectrometry
such as MS/MS allow a chemist to determine several compounds
or classes of compounds simultaneously. For example, it is now
possible to program MS/MS instruments to look for specific ions
indicative of a given compound or compound class. If the ions
are detected, the mass spectrometer can automatically perform a
daughter ion experiment to verify the presence of the contami-
nant and then perform the quantitation if the ion is present. Such
experiments can be completed in 10% of the time of conventional
GC/MS methods.
When these newer techniques are used in conjunction with
automation systems, laboratory throughput is significantly in-
creased.
Automation
The introduction of automatic sampling devices has revolution-
ized the way laboratories function. Instruments that previously
required constant operator attention are now programed for un-
attended sample analysis.
To achieve high throughput, laboratories must quickly vali-
date data. Automated data transfer and quality assurance are
playing a significant role in increasing laboratory productivity.
To achieve high data validation speed, both industry and gov-
ernment are using automatic data transfer.
The U.S. EPA is now requiring that data be submitted on
floppy disks and is using an elementary expert system for prepay-
ment screening. Data submitted to the Agency under the new
Superfund program are provided in specified format on floppy
disks. The data are read and checked to determine whether all
contractual requirements have been met prior to payment. This
system is based on the premise that data meeting contract re-
quirements will, in fact, be valid.
Other groups within the U.S. EPA have recently employed ex-
pert systems to perform higher level checks. For example, sys-
tems that automatically compare results from several different
measurements for consistency are being used. Checking total
organic carbon against total constituent concentrations provides
a rapid method to determine whether a sample contains addi-
tional constituents. These checks are being used increasingly by
the U.S. EPA to ensure high quality data that can meet any scien-
tific or legal challenge.
Many laboratories are installing laboratory information man-
agement systems (LIMS) to meet new productivity goals. It is
clear that the future of environmental laboratories must include
automation at all levels and more efficient methods to provide
high quality data.

INCREASING CAPACITY
Although substantial increases in throughput can be achieved
by increasing productivity, there are cases where increasing pro-
ductivity alone will not meet a laboratory's goals for faster sam-
ple throughput. In these cases, laboratories must consider increas-
ing basic sample capacity.
Increased capacity is not a simple matter of adding equipment
and personnel. Spatial and environmental considerations must be
made prior to increasing capacity. Fortunately, modern instru-
mentation is much more efficient than previous models. New in-
struments now perform the same analyses as older models twice
their size and use less power.

CONCLUSION
High throughput environmental analyses require improved
productivity and capacity. Several approaches are available for
laboratories and each laboratory must decide which approach
makes best use of resources.
WASTE MINIMIZATION 351
-------
Hazardous Waste Reduction—Cornerstone of a
Waste Handling Strategy

Robert P. Bringer, Ph.D.
3M Company
St. Paul, Minnesota
INTRODUCTION
Waste has historically been regarded as something inherent in
industrial processes—undesirable, but inevitable. Waste as an un-
desirable by-product has usually been removed from the process
quickly and disposed of in the most expeditious and lowest cost
manner available. Periodic "waste reduction" programs have
been prompted principally by the need for cost savings, most easi-
ly achieved by reducing raw material usage. Raw materials have
generally been in plentiful supply and scarcity has seldom been
the reason for waste reduction programs.
During the last 10-15 years, however, some new factors have
appeared on the industrial scene which are causing corporations
to rethink their approach to waste and to reevaluate its impact on
their ability to survive. These new factors include:
• a complex environmental regulatory atmosphere which is rais-
ing the cost of waste disposal dramatically and taxing the gen-
eration of waste;
• a highly litigious society which can punish industry and others
for wastes disposed of even when within the law;
• a recognition that all resources, especially energy and raw ma-
terials, are not inexhaustible;
• a new emphasis on quality and the suspicion that high waste
generation and high quality are not compatible; and
• the advent of global competition which accentuates the need
to be a low-cost producer.
All these factors will continue to gain in importance over the
foreseeable future. That, in turn, will raise the tangible and in-
tangible costs of waste to levels that management cannot ignore.
The emphasis in waste management has most recently been on
proper disposal of hazardous waste. Because of the broader im-
plications of the new factors cited, waste management must
become broader in scope and, as such, a significant part of the
corporate quality effort.
Various companies have had waste management strategies or
programs for a number of years—and they are constantly evolv-
ing as new issues emerge. The status of the 3M program today is
probably typical of many large manufacturing corporations. We
all share the curse and the opportunities presented by volume and
variety in the wastes we generate, handle and dispose. However,
as a large manufacturing company, we also find ourselves pro-
cessing information on a large scale and servicing our products in
the field. So, we do have some of the same waste generating fac-
tors as the information processing and service sectors of the
business community. Small businesses may find similarities be-
tween what they do and what we do in individual and localized
cases. But small businesses do not usually have the opportunities
of scale working for them or internal staff expertise when han-
dling their waste and often need outside assistance.
Our overall waste handling strategy at 3M is a product of our
corporate environmental policy and some corporate desires. Our
policy was formulated by our Board of Directors over 10 years
ago, and it states that 3M will comply with governmental environ-
mental regulations and cooperate with the regulating agencies
(Table 1).

Tibk 1
3M Corporate Environmental Comemtion Policy
3M will continue to recognize and exercise its responsibility to:
• Solve its own environmental pollution and conservation problems
• Prevent pollution at the source wherever and whenever possible
• Develop products that will have a minimum effect on the environment
• Conserve natural resources through the use of reclamation and other
appropriate methods
• Assure that its facilities and products meet and sustain the regulations
of all federal, state and local environmental agencies
• Assist, wherever possible, governmental agencies and other official
organizations engaged in environmental activities

The policy also states that the company will prevent pollution
at the source, conserve resources, produce environmentally sound
products and, importantly, handle our own environmental prob-
lems.
Our underlying corporate desires include:
• Doing the environmentally positive thing. Many companies
have changed their commitments in the past several years from
what has to be done to what should be done environmentally;
often at the sacrifice of economic objectives.
• Reducing our current and future liabilities. This is more im-
portant today than ever before with the difficulty of obtaining
meaningful liability insurance and the litigation potential for
our wastes as they travel from "cradle to grave."
• Pursuing cost-effective solutions; we are a profit-making or-
ganization.
• Continually improving the quality of our products and pro-
cesses and lowering their potential for generating pollution
problems.
This combination of policy and desires has evolved into six
basic components which make up our corporate Waste Manage-
ment Program (Table 2).
The hard part of any program, of course, is carrying it out. At
3M that is complicated by the very nature of the company; for ex-
ample—

• large—$8 billion sales
352
WASTE MINIMIZATION
-------
• multinational—sell in 50 countries, manufacture in 20
• many smaller plants—113 registered EPA facilities in U.S.
alone
• multitiered regulatory contacts—federal, states, 113 counties
and cities in U.S.
• 45,000 products—25% less than 5 years old
• over 1,000,000 customers.
We produce a great number of different wastes in many dif-
ferent places subject to a variety of regulations, sometimes con-
flicting, and subject to the availability of suitable treatment and
disposal facilities. Dealing with waste at a company like 3M is a
complex and necessarily sophisticated business.
But, basically, we deal with two classes of waste—hazardous
and nonhazardous. Hazardous waste today is highly regulated
through RCRA and various state, county and municipal regula-
tions. And because of the nature of hazardous waste, its advanced
regulatory situation and its implications for corporations like 3M,
our program for managing hazardous waste is much farther along
than our program for managing nonhazardous wastes.

WASTE MANAGEMENT PROGRAM
The components of our Waste Management Program apply to
both hazardous and nonhazardous waste. And although today it
is more important and cost-effective to properly handle hazar-
dous waste than nonhazardous waste, nevertheless, there are
many parallels that can be drawn between the two cases. These
parallels can best be drawn in discussion of the individual com-
ponents of the Waste Management Program (see Table 2).

Table 2
3M Corporate Waste Management Program

1. Identify and quantify all waste streams.
2. Minimize waste generated at the source.
3. Maximize waste reclaim and reuse.
4. Apply most appropriate treatment and disposal technologies.
5. Maintain direct control of wastes to the maximum extent possible.
6. Comply with all regulatory requirements.
Identify and quantify all waste streams
This sound simple, but isn't, in a large, multiplant company
with complicated processes. At one of our larger and more
sophisticated facilities, no fewer than 999 separate waste sources
were identified. About two-thirds of these were hazardous, one-
third nonhazardous. With numbers like these, it becomes im-
perative to handle the information by computer.
Identification of waste streams is often easier than quantifica-
tion. Hazardous waste streams are more easily quantifiable
because they are usually part of the manufacturing process and
homogeneous in nature. Nonhazardous waste streams, on the
other hand, tend to be incidental to the process,
nonhomogeneous and difficult to quantify in simple weight or
volume terms. Obviously, we need better measurement systems
for nonhazardous waste than are now in place.

Minimize waste generated at the source and maximize
reclaim and reuse
These two components are the heart of 3M's Pollution Preven-
tion Pays or 3P program, which has been emphasizing waste
minimization at 3M for over 11 years. And the 3P program is the
cornerstone around which our waste management strategy is built.
There are four basic ways to prevent pollution at the source,
where a source is considered broadly as a product or a facility:
Develop nonpolluting products or reformulate existing products
This is done by substituting nonpolluting materials for ingredi-
ents that are pollutants.
A good example of this is our effort to substitute water-based
adhesives for those that contain solvents in the manufacture of
adhesive tapes. These solvents are pollutants because—untreated
—they contribute to the formation of ozone in the atmosphere.

Modify production operations
Examples are conversion to solventless coating processes or
changing from a batch to a continuous process in order to reduce
peak discharges and allow for water recycle.

Redesign equipment used in a manufacturing process
For example, at one of our plants, we converted a boiler so it
could burn solvent-laden exhaust from a coater, thereby
eliminating add-on pollution control equipment and saving
$270,000 per year in energy costs.

Recover and recycle waste products
As an example, at one of our plants, a film developing unit
generated wastewater containing 1,1,1-trichloroethane. A de-
canter system was installed to recover the developer for later
distillation and recycle. The decanter system installed cost was
$4,000 and produced a cost savings of over $20,000 for the first
four years of operation.
There are many examples of successful projects in all four basic
categories—with cost savings ranging from a few thousand to a
few million dollars. Our most prolific categories have been refor-
mulation and recycle. For other companies, the best opportunities
may lie in a different set of categories. But, the four categories are
the same, regardless of products or processes.

HOW THE 3P PROGRAM WORKS
First of all, it is a corporate-wide program fully endorsed by
management at all levels. Each of our 40 operating divisions and
30 OUS subsidiaries is encouraged to participate.
The 3P program is mainly directed at the company's 6,000
technical employees in manufacturing, engineering and labora-
tories. They are the ones most responsible for our products and
processes and thus the ones most able to work on pollution at its
source. Working in their own specialty areas—doing the work
they know best—they are asked to implement pollution preven-
tion concepts in their everyday activities. They become, in effect,
an extension of our Corporate Environmental Engineering staff.
When they believe they have made a worthwhile accomplish-
ment, they submit their effort to a 3P coordinating committee of
laboratory, engineering and manufacturing representatives for
review. Each effort is judged on the following criteria:
• It must result in an environmental benefit that can be quanti-
fied, such as the amount of pollution prevented.
• It must have a cost savings for the company.
• It must demonstrate technical achievement or innovative ap-
proach above and beyond standard engineering or manufac-
turing procedures.
If the effort meets with peer approval, then, and only then, is it
accepted as a 3P project. The employee receives a nice wood and
brass plaque. The presentation is made by a senior management
official in front of fellow employees, usually at a private
luncheon.
This recognition process is an important part of the overall pro-
gram. Other employees are encouraged—by management's
demonstration of support and commitment—to contribute 3P
projects. Employees also gain a sense of personally being able to
WASTE MINIMIZATION 353
-------
help the environment through their 3P projects. Top management
has always been highly supportive of the program. Annual
management reviews often take note of which product divisions
are doing the most—and the least—to produce 3P projects.
Another big part of the 3P program has always been communi-
cation—both inside and outside the company.
In addition to our recognition program, our internal communi-
cations have concentrated recently on sharing 3P success stories
with others in 3M—via one-page "Ideas" flyers. They combine
technical, economic and environmental information in easy-to-
read form with a little humor thrown into the title.
One "Ideas" flyer—entitled "Waste Stopper—Pumice on
Copper"—describes a project in which copper circuits are now
cleaned by scrubbing with pumice rather than by spraying with
corrosive acids. The result is a nonhazardous waste instead of
40,000 Ib/yr of hazardous waste—and a 2'/j-year payout of in-
vestment.
Another "Ideas" flyer—entitled "Pollution Abated—Cash
Crop Created"—describes a project in which waste ammonium
sulfate solution was transformed from a waste treatment problem
and cost to an agricultural fertilizer and a source of income. This
prevented 677 tons of water pollution annually and net invest-
ment was paid back in less than three years.
In addition to this kind of data on the "Ideas" sheets, we iden-
tify the 3M people responsible. This not only establishes an infor-
mation resource, but also represents another form of recognition
to one's peers.
We have always made information on our 3P program freely
available outside the company to anyone who requests it. We
believe in the 3P concept strongly. We also believe that by selling
the concept to others, we can help establish a leadership position
for industry in the environmental area.

Table 3
Pollution Prevention Pays Program
3M's 11-Year Results
Worldwide Savings (first year only)
Number of Projects
Countries Participating (subsidiary 3M
Companies)
Pollution Prevented Annually
—Air Pollutants
—Water Pollutants
—Wastewater
—Sludge/Solid Waste
$292 million
1,852

109,000 tons
13,100 tons
1,495 million gallons
274,000 tons
Let's now review the results of 3M's use of the 3P approach
(Table 3). Since the program's inception at 3M in 1975, 1,852 3P
projects have saved $292 million. These costs are for pollution
control facilities that did not have to be built; for reduced pollu-
tion control operations costs; for reduced manufacturing costs;
and for retained sales of products that might have been taken off
the market as environmentally unacceptable. Cost and sales sav-
ings are claimed for the first year only. Thus, the total savings are
stated very conservatively.
These cost savings have continued to grow over the 11-year
period with our OUS subsidiaries contributing 20% of the total in
that period.
From an environmental standpoint, the program has yielded
equally good results as shown in Table 3. In addition, the 3P pro-
gram's annual energy savings are estimated at the equivalent of
210,000 barrels of oil.
These 3P results from our company are an illustration of what
other companies also can do. The encouraging word today is that
many companies are, in fact, eliminating or reducing pollution at
the source.
Our 3P program has concentrated in the past on hazardous
waste reduction and reclamation because the economic and
regulatory incentives are usually much higher than with nonhazar-
dous waste. Much of the $292 million savings over 11 years has to
do with hazardous waste. This situation will change as landfills
close and economic incentives to minimize or utilize nonhazar-
dous waste grow.
A number of nonhazardous waste recycle programs are already
in effect at various 3M sites involving paper products, precious
and nonprecious metals, treated papers and plastics. These pro-
grams all involve multimillion pound quantities of materials.
They all save money and have positive environmental effects.

APPLY MOST APPROPRIATE TREATMENT
AND DISPOSAL TECHNOLOGIES
Essentially all of 3M's hazardous waste in the U.S. is sent to a
central hazardous waste incinerator where it is burned in a
regulated facility which has state-of-the-art air pollution control.
The volume of this waste is reduced by 95% to ash and sludge
residuals which are sent to regulated Landfills. The energy
generated by burning this waste is recovered as steam which is used
at the plant site—further saving the need to use fossil fuels.
Other examples of conversion of hazardous waste to energy in-
clude the use of recovered solvent and solvent-laden air as fuel in-
puts for industrial boilers. Converting nonhazardous waste to
energy is also an alternative for these types of materials. We are
just starting down that path as we study which plant sites can
utilize the energy created and also have an adequate supply of
nonhazardous waste. At a plant near Chicago, we are just com-
pleting a waste-to-energy incinerator for nonhazardous wastes
generated at that site. In London, Ontario, we recently completed
the public approval process for a nonhazardous waste incinerator
which will supply energy for our 3M Canada plant. The interesting
thing about this installation is that 80% of the nonhazardous waste
fuel will come from our industrial neighbors. In Alexandria, Min-
nesota, we are investing over $100,000 to allow our plant there to
utilize the steam generated at a public waste-to-energy facility.
Opportunities for converting nonhazardous waste to energy are
growing as landfills close and legislation dictates. However, in-
dustry and government must work closely to make sure that op-
timum choices between private and public funding are made.

MAINTAIN DIRECT CONTROL OVER OUR
WASTES TO REDUCE FUTURE LIABILITIES
This is the main reason for treating essentially all of our hazar-
dous wastes in our own facility. Ultimately, we would like to con-
vert the hazardous residues to delisted waste and bury them on
our own land. Then we would truly have "cradle to grave" con-
trol.
The use of in-house hazardous waste incineration is becoming
more prevalent in large U.S. corporations today. In addition to
the desire to reduce liabilities, there is also the growing need to
have an assured method of handling hazardous wastes in order to
ensure continued production of goods. This option is not general-
ly open to small and medium size companies, but the future may
see consortia formed for this very purpose.
There is a parallel with nonhazardous wastes. The landfills at
which nonhazardous wastes are buried today may become the
past disposal sites of tomorrow and require cleanup. Those who
have buried solid wastes will be implicated along with those who
buried hazardous wastes. Also, the definitions of what is hazardous
and what is not constantly change. The best approach to reduce
future liabilities is to reduce the volume through treatment and
354 WASTE MINIMIZATION
-------
bury residue in single purpose landfills—preferably your own.

COMPLY WITH ALL REGULATORY
REQUIREMENTS
The key to compliance is knowledge—identification and quan-
tification of waste streams, awareness of all laws and regulations
affecting your facilities and products and continuing review of the
environmental practices at your facilities. A good internal audit
system can help to ensure compliance and raise a company's con-
fidence level. A timely follow-up procedure is an essential ingredi-
ent in any audit program. Auditing and any other compliance
procedures above all require persistence.
Even full compliance, however, is no guarantee against liability
or litigation. It only guarantees peace with the regulating bodies.
One only has to have the costly experience of cleaning up past
disposal sites to know that living within today's rules guarantees
nothing about future liabilities. That is why maintaining direct
control is so important.
Hazardous wastes are tightly regulated. Nonhazardous wastes
are not—today. That is changing as people talk about co-disposal
of dry scrap and industrial solid waste. Household hazardous
waste is being examined. One thing we can be sure of—non-
hazardous waste will be more tightly regulated in the future.
SUMMARY
In summary then, it is important to have a waste management
plan today, no matter what kind of business you are in, large or
small. The opportunities to profit from proper management of
nonhazardous waste are growing as the old practices become
more expensive and create future unknown liabilities. And a great
deal can be learned from our management of hazardous wastes
and applied to the management of nonhazardous wastes. The
economic, liability and regulatory situations for hazardous and
nonhazardous wastes are on a path of convergence.
Finally, the best approach will always be the 3P approach.
Don't create it in the first place and you will never have to deal
with it.
WASTE MINIMIZATION 355
-------
Institutional Realities of Locating a Repository
For Low-Level Radium Waste in Colorado

H. Donald Ulrich
Joan van Munster, P.E.
CH2M Hill Central, Inc.
Denver, Colorado
John Brink
U.S. Environmental Protection Agency
Denver, Colorado
ABSTRACT
This paper discusses the techniques used, costs and institu-
tional realities faced in locating a repository in Colorado for the
disposal of 65,000 to 130,000 yd' of low-level radium waste cur-
rently situated in several areas of metropolitan Denver, Colorado.
In 1979, a U.S. EPA employee found a 1914 U.S. Bureau of
Mines document that referred to the National Radium Insti-
tute originally located in Denver. This initiated an investigation
that subsequently revealed 31 contaminated properties in various
parts of Denver. These properties were collectively included on
the National Priorities List of hazardous waste sites. CH2M
HILL, as the U.S. EPA's REM/FIT contractor in Region 8, was
directed to evaluate these sites and identify alternate methods of
cleanup.
One aspect of CH2M HILL's assignment was to identify a
final repository for the Denver radium materials. Under
CERCLA, the State of Colorado is required to assure the avail-
ability of a suitable repository for the disposal of waste from
Superfund sites. The intent of this study was to facilitate this pro-
cess by providing comparative information on a range of alterna-
tives. To determine relative feasibility, each site was evaluated
based on environmental, institutional and economic criteria.
The most technically acceptable option is co-location of the
Denver radium waste with the proposal uranium mill tailings fa-
cility to be located in Grand Junction, Colorado. However, the
feasibility of this alternative was diminished greatly due to insti-
tutional limitations resulting from the need to coordinate agency
schedules and to repermit the site. It also was determined that
identification of any number of dedicated single-use sites in east-
ern Colorado would be economically and environmentally attrac-
tive, with the relative impact between sites nearly indistinguish-
able. Unfortunately, the establishment of any new facility would
be faced with 5 yr of legal and permit-related battles, not to men-
tion anticipated public opposition.
As a result, no acceptable long-term site was available for im-
mediate disposal of the waste material, and it became necessary
to investigate alternatives for temporary storage. The U.S. EPA
presently is negotiating with Mentor Corporation, owner of one
of the contaminated properties, to use their site for temporary
storage of up to 40,000 yd' of contaminated material. After a
maximum of 8 yr the material must be removed from the Mentor
property. The study recommends continued negotiation with
DOE to ultimately co-locate the Denver Radium material at the
UMTRAP facility at Grand Junction.
INTRODUCTION
As pan of the Denver Radium Superfund Site Remedial Inves-
tigation/Feasibility Study activities, a focused feasibility study
(FFS) was undertaken to define options for the disposal of ma-
terial generated from the 31 contaminated properties. The objec-
tive of the study is to identify a cost-effective means of disposing
the contaminated material in an environmentally safe manner.

Focused Feasibility Study
The FFS addresses the selection of a disposal site for the con-
taminated waste material resulting from remedial actions at the
Denver Radium Site, including that material generated through
owner-sponsored cleanup activities. Section 104
-------
for at least 200 yr
• To limit radon gas releases to 20 pCi/mVsec (above back-
ground) or less
» To prevent increasing the concentration of toxic chemicals in
surface or groundwater

In order to place this material in perspective, a comparison of
the radioactivity of the Denver Radium contamination to that of
other radioactive materials is shown in Table 1. Notwithstand-
the relatively low-level radioactivity of the Denver Radium ma-
terial, the health hazards associated with it are worthy of consid-
eration due to the risks of lung cancer associated with life-long
exposures to radon and its decay progeny in buildings constructed
over radium contaminated soils.

Table 1
Comparison of Radioactive Materials
Materials from Denver Radium Sites

Colorado Plateau uranium ore
(typical of camotite)

Grand Junction uranium mill tailings

Typical environmental levels found
in soils

Typical granite rock

Typical levels found in brick

Typical levels found in concrete

Gypsum waste from phosphate production

Rose food (15% potassium)

Lualnous watch dial
Radium Concentration
b
5 to 1,000 ; avg. approx. 300

300 to 3,000 <0.1»-U ore)

2 to 7,590°; avg. approx. 720

0.1 to 2
1.4 to 2.6

0.9 to 2.0

30 to 40

3 to 300 million'1
rMeasured in picoCuries per gram (pCi/gm).
Host observations fall within this range, with isolated values up
cto 3,000 pCi/gm.
A typical watch dial will have 1 to 2 million picoCuries total
radioactivity. The paint used for watch dials ranges from 3 million
to 300 million pCi/gm and averages 25 million pCi/gm. Radium watch
dials are no longer made in the United States.

Sources: Nelson, Roger, 1985. Personal Communication, UMTRAP
Albuquerque Office, June 6, 1985.

National Council on Radiation Protection and Measurements,
1984, Exposures from the Uranium Series with Emphasis on Radon
and its Daughters. Recommendations issued March 15, 1984,
Bethesda, MD.

Toohey, Richard, 1985. Personal Communication, Argonne National
Laboratory, August 1985.
Volume of Denver Radium Material
The total volume of contaminated material found at the Den-
ver Radium properties is estimated between 115,000 and 231,000
yd'. From 65,000 to 130,000 yd3 eventually may be disposed
off-site by the time cleanup efforts are completed. It is also pos-
sible that from 39,000 to 78,000 yd3 could be left in place where
contamination does not exceed U.S. EPA standards.

EVALUATION METHODOLOGY AND
SITES INVESTIGATED
In this section of the paper, we describe the approach used to
screen and evaluate the candidate alternatives. A brief synopsis
of the alternative sites investigated and the basis for their inclu-
sion also is provided.
Evaluation Methodology: Screening
Criteria
Screening criteria progressed from very general to definitive as
the study progressed. A narrative on the evaluation methodology
follows:

Task 1: Potential Disposal Site Identification
The criteria for developing a shopping list of alternative repos-
itories was based on common sense. Project staff identified sites
that:

• Were in Colorado (excluding commercial sites)
• Were thought to be underlain by suitable geology
• Were already existing and used for the disposal of similar ma-
terials
• Were on the drawing boards or in the permitting phase (prin-
cipally UMTRAP sites)
• Were public properties with compatible land uses and near
Denver

Task 2: Site Evaluation and Screening
The screening criteria used for Task 2 are given in Table 2. The
intent of screening was to eliminate any obvious problems that
would absolutely prohibit the use of a candidate site for the Den-
ver Radium Waste. The evaluation was limited to an in-office re-
view of existing literature.

Table 2
Screening Criteria—Denver Radium Focused Feasibility Study

Engineering
• Capacity
• Compatibility of Wastes
• Reliability
• Transportation
• Construction
Environmental
• Land Use Compatibility
• Population Density
• Presence of Unique or Special Environment Features
• Oeology/Hydrogeology
• Safety
Implementability
• Public Acceptance
• Institutional Constraints
• Lead Time/Availability
• Permitting
• Potential for Temporary Storage
Cost
• Loading/Unloading
• Transportation
• Disposal Costs
• Operation and Maintenance
Sufficient information was generally available for existing or
imminent facilities. Conversely, other than general geologic base
maps, land ownership, transportation maps and census data,
essentially no environmental data were available for the new
single-use sites. Consequently, new single-use sites were screened
for geological suitability, property ownership, access to transpor-
tation and distance from Denver.
Data on the status of permits and availability of existing or
imminent sites were obtained from the involved state and federal
regulatory agencies. Rule of thumb estimates (i.e., 3 to 5 yr or
more) were used to evaluate the permitting schedule for new
single-use sites.
MINING & INDUSTRIAL WASTES 357
-------
Public acceptance was assessed based on what was known to be
unacceptable:
• Interstate transfer
• Intrastate transfer, i.e., crossing county lines
• Urban export to rural communities
• Transport from the eastern to western slope Colorado com-
munities
• Population density
• Precedent, i.e., recent resistance to similar projects
The possible safety concerns were limited to those associated
with moving materials to and disposing materials at the site. The
accident rates for truck transport within Colorado are based on
the Colorado Department of Highways report, "Large Truck
Accidents and Rates on the State Highways 1978-1982." The total
accident rate, based on average daily traffic counts and accident
statistics for specific routes, was calculated using the following
formula:

Total Accident Rate =
Total No. Ace x 10*

365 days x Avg Daily Traffic x Section Length
(1)
Accident rates were available for neither train transport nor
highway routes.
Order of magnitude cost estimates were based on units costs
developed from other related Superfund projects.

Task 3: Detailed Site A nalysis and Selection
Differences Between Task 2 and Task 3 Criteria. The evalua-
tion criteria for Task 3 are given in Table 3. In general, the cri-
teria are similar to those used in earlier steps, only a more rig-
orous analysis was conducted. In particular:
• Engineering: All of the existing or potential facilities were de-
scribed in detail to establish compatibility of operations and the
potential cost impact that would be realized through incorpo-
ration of the Denver Radium Waste. A prototypical single-
use site was designed based on UMTRAP design criteria. (A
single-use site is defined as a repository designed principally
for disposal of the Denver Radium material.) Railroad and
trucking companies were consulted for the design of the ma-
terials handling facilities. Again, this was done principally to
establish comparative project costs. Figure 1 depicts the hypo-
thetical plot plan for the single-use site.
• Environmental: The environmental assessment done at Task 3
was considerably more detailed than the assessment conducted
during Task 2. However, the review was based on existing per-
mit documents for the existing and impending projects. En-
vironmental work for the new single-use sites was based on site
visits, although no drilling was conducted to confirm subsur-
face geologic conditions. The safety evaluation remained un-
changed from Task 2.
• Availability: The status of availability for UMTRAP sites was
determined through a phone survey of DOE and State of Col-
orado regulators. Commercial sites were contacted directly.
The permitting schedule for a new single-use site was deter-
mined by meeting with representatives of the Colorado De-
partment of Health, Colorado Geological Survey, Natural
Resources, Local Affairs and the affected county governments.
• Cost: The cost estimates for Task 3 were based on quotes from
local trucking and railroad companies, quotes from commer-
cial operators, quantities to be moved and in-house estimating
guides. Development costs for single-use sites were modified
from the Task 2 estimates through a more detailed analysis
of access and materials handling. Estimates were prepared for
both the rail option which included siding, unloading facilities
and conveyors, and a trucking option which required only a
conveyor system. Estimates for permitting and engineering for
single-use sites were based on a percentage of construction
cost.

SITES INVESTIGATED
As presented in Table 4, 24 sites were identified and screened
using the criteria presented earlier. These 24 sites were reduced to
six candidates using the screening process described below.

Commercial Disposal Facilities
Five commercial facilities were evaluated through screening:
Lowry Landfill, owned by the City and County of Denver; a
new hazardous waste facility near Last Chance, Colorado, owned
by BFI; Handford Nuclear Reservation at Richland, Washing-
ton; Beatty Test Site in Nevada; and the proposed COMPACT
site in Colorado.
The Lowry Landfill site was eliminated due to institutional
constraints—its license prohibits disposal of radioactive waste,
and it is a designated Superfund site. (U.S. EPA's off-site dis-
posal policy severely restricts the use of sites that are not in full
compliance with environmental requirements.) Geologic con-
ditions at the site are also unfavorable. BFI's facility at Last
Chance is not available since the present permit application pro-
hibits disposal of radioactive waste.
Of the two existing commercial sites, Beatty, Nevada, was eval-
uated in favor of Hanford, Washington, because it is 400 miles
closer to Denver and poses fewer institutional problems since it
is presently the Rocky Mountain Low Level Radioactive Waste
Compact disposal site. Based on conversations with U.S. Ecol-
ogy, the operator of both the Hanford and Beatty facilities,
movement of the waste to Washington is likely to be met with
strong public resistance. Further, if regional low-level waste com-
Tiblc}
DtUlkd Evaluating Crilerta Focused Feasibility Study
Engineering
Proposed Design
Operations Plan
• Materials Handling
• Closure
• Monitoring
Environmental
Land Use
Topography/Surface Water
Subsurface Conditions
Cultural Resources
Vegetation and Wildlife
Public Safety
A variability
Status of Permits
Schedule
Institutional Constraints
Cost
Excavation and Loading
Truck Transport to Rail Siding
Rail or Truck (whichever was applicable to the site)
Site Development Cost (if applicable)
Disposal Fee (if applicable)
Operation and Maintenance
Engineering and Permitting (if applicable)
358 MINING & INDUSTRIAL WASTES
-------
pacts are ratified by Congress, the availability of the Hanford
facility for Denver Radium materials would eventually be re-
stricted since Colorado is not a member of the Northwest Com-
pact.
According to the terms of the Rocky Mountain Compact, a
new site must be established in a state other than Nevada by
1988; a proposed amendment would extend their deadline to
1989. Since the selection of a new Rocky Mountain Compact
site greatly influences this study, additional background informa-
tion is given below.
Colorado is part of the Rocky Mountain Compact formed
under the provisions of the Low Level Radioactive Waste Policy
Act of 1980. This act requires states to establish low-level waste
disposal facilities and allows them to form compacts to establish
regional disposal facilities. The four-member Rocky Mountain
compact has been ratified by the legislatures of Colorado,
Nevada, New Mexico and Wyoming and is awaiting ratification
by Congress. Since Colorado has the largest existing volume
and is the greatest producer of low-level waste among the Com-
pact members, it is expected that the site ultimately selected for
the Compact disposal facility will be located in Colorado.

Table 4
Sites Investigated and Screened
Commercial Facilities:
Lowry Landfill
Last Chance, Colorado
Hanford Nuclear Reservation, Richland, Washington
Beatty Test Site, Nevada
Colorado Compact Site
Uranium Mill Tailings Remedial Action Program (UMTRAP) Sites:
Clive, Utah
Durango, Colorado
Gunnison, Colorado
Grand Junction, Colorado
Naturita, Colorado
Public Property or Facilities near Denver:
Rocky Mountain Arsenal (RMA)
New airport expansion site
Existing Uranium Mills:
Cotter Corporation, Canon City, Colorado
Union Carbide, Uravan, Colorado
New Single-Use Sites:
Two sites near Fort Morgan, Colorado
Three sites near Last Chance, Colorado
Four sites near Limon, Colorado
One site near Manila
The State of Colorado currently is considering alternatives for
a Compact site, including co-location of the Compact site with a
DOE UMTRAP site. For the purposes of this study, it is assumed
that the new Rocky Mountain Compact site will be co-located
with the UMTRAP facility planned in the Grand Junction, Col-
orado, area. It is further assumed that the site could not be on-
line until late 1989 to early 1991 and that the facility would be
Privately owned and operated.
Of the three commercial facilities, only Beatty would be avail-
able immediately for receipt of the Denver Radium material,
Hanford probably never will be accessible and the new Rocky
Mountain Compact site will not be on-line for 4 to 5 ys. Disposal
fees at any of the three commercial facilities are anticipated to be
comparable and can be expected to average $400 to $500/yd3.
^BUI
£r
BURIED
MARKER
A
> STOCKMLC AHCA \
\ FOfi CXCAVATtD \
i MATfHIAL. IUrO*TtD }
> AM nmiAL, HAFHA? 3
UPCflAD.ENT
MONIfO^NG ^\
WEU ii
^
Figure 1
Single Use Conceptual Site Plan with Rail Access
(not to Scale)
UMTRAP Disposal Sites
The UMTRAP sites at Clive, Utah and Grand Junction, Col-
orado appeared to represent the most attractive and cost-effec-
tive solutions for disposal of the Denver Radium material. The
U.S. EPA and DOE have discussed the feasibility of disposing of
the Denver Radium material in the same facility as the uranium
mill tailings.
Inasmuch as the wastes are compatible and, in general, the
Denver Radium material is less radioactive, there should be no en-
vironmental or technical constraints to a program for combined
disposal.
Since the Clive facility was to be open for only about 18 mo
starting July 1, 1985, it was considered a logical option for dis-
posal until the opening of the Grand Junction facility. However,
the Clive facility was considered unuseable due to the unwilling-
ness of Utah officials to accept the Denver Radium material.
UMTRAP sites proposed at Durango, Gunnison, Rifle, May-
bell and Maturita, Colorado, were eliminated due to long haul
distances and timing considerations. It was concluded that a new
single-use site could be permitted in about the same time-frame
as those facilities (by 1990) but with less than one-half the haul
distance.
The Maybell and Rifle UMTRAP disposal facilities also may
offer a potential option for co-location. However, the Grand
Junction site was pursued as the most promising option because
it is among the first of the Colorado UMTRAP projects to move
to the construction stage.

Public Property or Facilities Near Denver
Both the Rocky Mountain Arsenal (RMA) and the new airport
site offered the advantage of a short hauling distance for disposal.
The RMA had to be eliminated due to army policy prohibiting
the disposal of hazardous waste on any of their installations and
the fact that the property is a proposed Superfund site. More-
over, the development of the new airport currently is being met
MINING & INDUSTRIAL WASTES 359
-------
with considerable opposition. This opposition to development
combined with unsuitable geologic conditions eliminated this
option.

Existing Uranium Mills
Uranium mills owned by the Cotter Corporation and Union
Carbide were eliminated at the Task 3 stage of the study due to in-
stitutional problems (both are U.S. EPA Superfund Enforce-
ment sites), incompatible schedules for completion and long haul-
ing distance. Moreover, disposal fees at these facilities were as-
sumed to be comparable to a commercial facility. Given these
considerations, there was no reason to investigate these options
further.
However, Union Carbide's millsite at Uravan re-emerged as a
potentially feasible disposal site in late 1986 when Colorado set-
tled its CERCLA natural resources damages claim against
UMETCO and its parent company. Union Carbide. The settle-
ment permits Colorado to use the Uravan site for Denver Radium
disposal at no cost to the state, provided that the disposal takes
place on lands owned by the defendants and that the Denver
Radium disposal takes place in an area that is separate from the
mill tailings that are being stabilized at the site.

Single-Use Sites
Single-use sites were identified through overlay mapping. The
information compiled included:
• Surficial and bedrock geology
• Prime farmland
• Sediment yield
• Transportation corridors
• Land use
• Land ownership
The analysis determined that a large number of technically suit-
able sites were available. For study purposes, 10 sites were iden-
tified for comparison. After screening, the 10 sites were pared to
four. Plates (photographs) 1 through 6 presented at the end of
this paper show the single-use sites evaluated.
The Fort Morgan area sites were eliminated from further con-
sideration because the area is projected to grow significantly
faster than other eastern Colorado areas.
The three sites near Last Chance, Colorado were eliminated
due to the absence of rail facilities and strong opposition. The
Last Chance area recently had been embroiled in the permitting of
a hazardous waste disposal facility being developed by BFI. Due
to this opposition, it had taken over 7 yr to complete permit-
ting.
Consequently, it was concluded that the most favorable area
for the facility would be near Limon, Colorado. Based on field
work, three sites that offered favorable transportation opportun-
ities were located in this area.
The site at Manila was included to show the trade-off be-
tween lower transport cost due to the proximity to Denver and
use of a relatively more conservative and costly liner design
needed to offset less suitable geologic conditions near Denver.
Finalist Sites
After screening, the following sites appeared to be the most
promising:
• Commercial Facilities
-Beatty, Nevada
-Rocky Mountain Compact Site in Colorado
• UMTRAP Facilities
-Grand Junction, Colorado
• New Single-Use Site
-Limon, Colorado
-Manila, Colorado
The locations of these sites are shown in Fig. 2.
Figure 2
Denver Radium Alternative Disposal Site Locations
DETAILED ALTERNATIVE DEVELOPMENT
AND EVALUATION
Alternative Development
Combining Alternatives is Most Feasible. At the conclusion of
this analysis, it became apparent that the small volume of material
to be initially removed may need to be disposed of separately
from the main volume of contaminated material due to the sched-
ule and the lack of an available repository. It is unlikely that any
acceptable long-term solution will be operable when it is neces-
sary to dispose of the initial portion of the material. On the other
hand, the options available for immediate disposal are not ap-
propriate for long-term disposal of large volumes due to loca-
tion cost, schedule for closure or politics. Thus, the best solutions
combine alternatives for immediate disposal with those suitable
for long-term, large-volume disposal.
All of the options use the commercial facility at Beatty,
Nevada, for the immediate disposal of 14,000 yd' of the Denver
Radium material. Although costly, this appeared to be the only
immediately available alternative that did not entail consolidation
and storage of wastes on-site until a permanent disposal option
becomes available.
360 MINING & INDUSTRIAL WASTES
-------
Detailed Alternative Evaluation
A brief description of the most promising combined alterna-
tives is presented below.
The options have been organized into three categories:
• Most feasible
• Feasible
• Less feasible
Relative feasibility between the categories is fairly apparent;
however, determining the most preferable option within a given
category is more difficult, with technical differences generally be-
ing negligible and subject to judgment factors.

Cost Not a Reliable Criterion. The cost estimates developed for
the individual and combined alternatives are sufficiently close so
that cost alone is not a reliable selection criterion when compar-
ing individual single-use sites or when evaluating the single-use
site concept against the UMTRAP facility at Grand Junction.
Cost estimates for the various combined alternatives are presented
in Table 5. Comparative costs for the alternative components
are given in Table 6.

Table 5
Summary of Alternative Costs8
(in Millions $)
Coajalnad jatarnatlra
Cowi-clal lit* at Matty. W/VtnW alta.
Cfind Junction, CO

CoBwrcUl lit* at Matty, MV/Blngla-Uaa Slta

. COMarclal lit* at Matty, HV/llnqla^Uae Slta

CoBMiclal alta at Matty, HV/Slngla-Uaa Sita

, CoaMfClal alta at Matty, HV/Blngla-Uaa Blta

. Conarelal ilta at Matty, HV/Mav Rocky Hount
Coaput Ilta
Mo. 11

Ko. 13
Diliilad co*t altlutaa ara provldad In tha raport
MIUMI • dlapoaal fa* of HS.BS/tt and nil/true
importation !• aatluti
t Situ la aaautwd to ba co-located with tha WTJWP Grand Junction fai
atMd on Intonation provldad by tha Stata ol Colorado, tha nav alta vlll ba prlvataly oparatad.
[«• lUlUr to thoaa at Baatty hava baan aaauMd.
Table 6
Comparative Project Component Costs ($/Ton)
Excavation
iltt Cattery Loading
Truck full Dlipaial
51 Ca Pei-Blli
Develop. & Engr.
S/Ton
t. ».» lock; Noua-
UlD Coapacl
Sit. (Co-
loutta,)'

1. DUMP Cltw,
It (Co-loeatloo)

>. IMUP Crnd
Aactloo
(Or location)

'• lltlli-Dit Slta:
""lla (Sit.
fc. 10)

'. !U|1|-UM SIM:
UK* (Sit.
Ma. U)
52

52
» 1.1 -
;
tb* "ockj Howiula C«IMCC SIM Mould DOC nqulra
"mlop-nt p.r toe.
!• ,1KU] far 220,000 ton* to bo coaMmtlvo.
/•INltDrjr Inclndot
co.tlr eollpi to coftpoao.
loot .alt*bl« goologlc coodlclou.
Drjr nclndot . wn co.tlr eollpi to coftpoao.M for loot .alt*« goologlc coodlclou.
j™61**" nil .tdlDK Md UBlMdlDC.
r~*" lot oitcnlDod whcthor tbc Duvar fadluo Htoriol. an vltblo tbc Jurl.dlctloo of the Rockr Houot.lc
^"lf«l lidlooetlv* Wfl*tt Compact. MoocttioloHt EPA It ovolutlng thl* option ID tho ovent tlut Colorado
fclll«pi • low-Uval wDBt« facility capable of baodllag tlw Duvar KadiuB Mtarlal.
Impact on the Natural Environment Not a Reliable Criterion.
While all the following alternatives would serve as acceptable re-
positories for the Denver Radium material, numerous other suit-
able disposal sites are potentially available. Some areas on the
western slope are potentially suitable but were not considered
for single-use sites in order to avoid mountain transport. How-
ever, other potential sites could be identified in the same eastern
Colorado study areas as the sites identified here. For example, 16
40-acre sites are available in a given section of land, and many
sections of land were found to be suitable from an environmental
and technical standpoint in the identified study areas. The sites
presented below are representative of the many options available.

Most Feasible Alternative

Commercial Facility atBeatty, Nevada/UMTRAP
Site at Grand Junction, Colorado
Of the alternatives evaluated, the only immediately available
permanent disposal option includes a combination of disposal of
the 14,000 yd3 of material to be removed initially at the commer-
cial facility at Beatty, Nevada, followed by long-term co-loca-
tion of the remaining 130,000 yd3 at the proposed UMTRAP fa-
cility at Grand Junction when the UMTRAP disposal facility is
established.
The advantages and disadvantages of this option are given be-
low:
Advantages
• This option currently offers the most expeditious solution, as
avoiding duplicate permitting could shorten the disposal sched-
ule for the Denver Radium wastes by about 3 yr.
• It is potentially the most cost-effective option, principally due
to minimal costs for permitting and engineering, and avoiding
high commercial disposal fees.
• It is the least institutionally complex, since the Beatty facility
presently serves as the low-level radioactive waste disposal site
for the Rocky Mountain Compact.
• There is reduced environmental impact when compared to a
new single-use site, as Beatty already exists and the UMTRAP
Grand Junction facility will be developed regardless of the co-
location of the Denver Radium material. This also will help
minimize the number of disposal sites.
• There is improved public safety, because the majority of the
haul can be provided by rail. All of the material (up to 130,000
yd3) disposed at Grand Junction can be hauled by rail to Grand
Junction. From Grand Junction, the material probably would
be trucked to the repositors. Rail transport to Beatty is pos-
sible. However, the shipment will be by rail only to Las Vegas
with trucking required for the remaining 100 miles to the site.
• Beatty can receive the Denver Radium waste material now.
• Since the Grand Junction site area is served by both rail and
highway, competition could result in lower hauling costs. How-
ever, as mentioned above, some truck hauling probably would
be required even with rail transport.

Disadvantages
• The co-location concept may delay receipt of BLM's approval
of the land withdrawal for DOE's proposed disposal site by as
much as 1 yr. This timetable may prove to have a sufficiently
severe impact on DOE's schedule to prohibit an acceptable
agreement with the U.S. EPA.
• Uncertainty exists regarding the opening of the Grand Junction
facility. If the site is not operating by late 1987 (the currently
projected opening date), this option becomes less attractive.
• There is potential public resistance to bringing waste materials
from Colorado to Nevada or from Denver to the western slope
MINING & INDUSTRIAL WASTES 361
-------
of Colorado. (However, public opposition is a disadvantage
common to all combined alternatives evaluated.)
• There will be high costs associated with the long haul distance
and disposal fees at Beatty. (However, this disadvantage is
common to all combined alternatives evaluated.)

Feasible Alternatives
If the alternative presented above cannot be implemented,
development of a single-use site dedicated for disposal of Denver
Radium material in eastern Colorado is the next most feasible
option.
In the discussion of the advantages and disadvantages of the
combined alternatives involving eastern Colorado single-use sites
below, only the attributes of the single-use site are addressed. For
a summary of the advantages and disadvantages of the Beatty,
Nevada, component of these options, the reader is referred to
the items related to the Beatty facility in the discussion of the
Beatty/Grand Junction UMTRAP option described above.

Commercial Facility at Beatty, Nevada/Single-
Use Site No. 12
Site No. 12 is approximately 8 mi east of Limon and 2 mi south
of Genoa, Colorado. The principal advantages and disadvan-
tages of this site include:
Advantages
• This option would present few, if any, land use conflicts due to
its remote location.
• The site is used for wheat farming and presents no valuable
habitat for wildlife nor does it include any native species of veg-
etation. No surface water is found on or adjacent to the site.
These factors result in little potential for adverse environmen-
tal impacts associated with site development.
• Access improvements would be minimal, as the site is served by
a paved road.
Disadvantages
• Permitting is expected to require a 5-yr lead time.
• The site could not be served economically by rail, thus increas-
ing the potential for vehicular accidents on 1-70.
• Waste material would have to be trucked through Limon. (This
exposure during transport has to be compared to any alterna-
tive in which the material must be hauled through the more
populous Denver Metropolitan area.)
• Access to the site would require exiting at Genoa, thus increas-
ing truck traffic and the associated noise and inconveniences.

Commercial Facility at Beatty. Nevada/
Single-Use Site No. 13
This option was identified as Site No. 13 and is located approx-
imately 13 '/2 mi southeast of Hugo, Colorado. The advantages
and disadvantages of this site include:
Advantages
• The site could be served by both rail and highway. This could
increase competition between haulers and reduce transporta-
tion cost.
• Similar to Site No. 12, this option is remote and would present
minimal land use conflicts.
• The immediate site vicinity appears to be over-grazed, and the
general absence of vegetation diminishes its value as a wild-
life habitat.
Disadvantages
• Permitting is expected to require a 5-yr lead time.
• Material would have to be hauled through Limon and Hugo,
Colorado. If rail transport were used, the traffic through
Limon and Hugo would increase by 2%, assuming a 3-yr re-
moval schedule.
• The site is within Yi mi of the Big Sandy Creek riparian corri-
dor. This area offers a shallow groundwater resource and prime
habitat for a variety of bird life and small and large mammals.

Less Feasible Alternatives
The alternatives presented below are technically feasible but
probably not implementable due to a combination of factors.

Commercial Facility at Beatty, Nevada/
Single-Use Site No. 10
Site No. 10 is located approximately 25 mi east of Denver,
Colorado. This site was evaluated to provide a comparison be-
tween transportation costs and the provision of a more conserva-
tive design to offset geologic conditions that are defined by the
State of Colorado as marginally suitable for the disposal of low-
level radioactive waste.
The relative advantages and disadvantages of this site include:
Advantages
• It has the shortest hauling distance (for either rail or trucking)
of any site evaluated, thus reducing hauling costs and the po-
tential for highway accidents.
• The site is used for wheat farming and presents no valuable
habitat for wildlife nor does it include any native species of veg-
etation. No surface water is found on or adjacent to the site.
These factors result in little potential for adverse environmen-
tal impacts associated with site development.

Disadvantages
• Permitting is expected to require a 5-yr lead time.
• Until detailed field data are available, it must be assumed that
this option will require the use of a more conservatively de-
signed repository to offset potentially marginal bedrock geol-
ogy. The characteristics of the underlying geology are un-
known.
• The cost of the more conservative design negates the savings
realized by the reduced hauling distance.
• There is a probable land use conflict with a newly proposed in-
dustrial park which is located adjacent to the site.
• There is a potential for increased public resistance due to the
site's relative proximity to Denver and comparatively greater
population density in the immediate area.

Commercial Facility at Beatty, Nevada/
Single-Use Site No. 11
Site No. 11 is located approximately 7 mi east of Limon on 1-70
near Cedar Point, Colorado. The relative advantages and disad-
vantages of this site include:
Advantages
• The site could be economically served by both rail or highway,
thus increasing competition and potentially reducing hauling
costs.
• Waste material would not have to be hauled through Limon,
Colorado. This may reduce local public resistance, especially
compared to the opposition that might arise for sites Nos. 12
and 13.

Disadvantages
• Permitting is expected to require a 5-yr lead time.
• Development of the facility could require special design con-
siderations to prevent impacts on surface water resources, since
the site is located adjacent to a drainageway feeding Big Sandy
Creek.
• The site appears to be subirrigated and consequently provides]
362 MINING & INDUSTRIAL WASTES
-------
productive grazing and, when compared to the other single-use
sites, a fairly productive wildlife habitat.
• The site would be readily visible from 1-70 and Colorado High-
way 86.
• Due to the relative proximity to water resources, this alterna-
tive may be difficult to permit, even with proven impact miti-
gation measures.

Commercial Facility at Beatty, Nevada/
New Rocky Mountain Compact Site
This alternative assumes that a new privately operated Rocky
Mountain Compact site would be co-located with the UMTRAP
facility in the vicinity of Grand Junction, Colorado. As with the
other options evaluated, the Denver Radium material that re-
quires immediate permanent disposal would be disposed of at
Beatty.
The relative advantages and disadvantages of this alternative
include:
Advantages
• Combining this facility with the UMTRAP site would avoid
the environmental impacts associated with developing a new
single-use site.
• Because the environmental and other permitting work would
be done for the UMTRAP facility, this effort would not have
to be completely duplicated, thus potentially reducing the time
required for permitting.
• This alternative could allow the Denver Radium material to be
hauled to the site by rail, thereby avoiding the potential for
additional vehicular accidents caused by trucking.
• Since the site is served by both rail and highway, competition
could result in lower hauling costs.
Disadvantages
• This option would result in the highest overall costs of any of
the alternatives considered in Task 3. This cost increase would
be due to operation at the facility, and disposal fees are ex-
pected to be comparable to Beatty.
• Similar to the single-use sites, this option could not be expected
to be on-line until 1989 to 1991. (It is possible that the new
Compact site, which will receive several types of low-level rad-
ioactive wastes, will be met with more public opposition than
a single-use site containing only the Denver Radium material.
Should this occur, the lead time for the Compact site could be
greater than that required for a new single-use site.)

PROGRAM IMPLEMENTATION
Scheduling Constraints
Based on the detailed alternative evaluation it became apparent
that resolution of the institutional issues facing all of the recom-
mended options would prohibit immediate action. Consequently,
it became evident that none of the alternatives identified to date
would fulfill program requirements, and that a more suitable
solution was needed. At this point, the need for temporary stor-
age became apparent. Previously, temporary storage had been
avoided due to the cost-inefficiencies and potential for accidents
resulting from double-handling of the material.
Unlike all of the other options evaluated, temporary storage
c°uld be implemented almost immediately. Under the National
Contingency Plan (NCP), no Federal or state permits are re-
quired to carry out CERCLA on-site response actions. More im-
POttantly, since all 31 of the Denver Radium properties are con-
sidered as on-site and as one Superfund site, temporary storage of
one property's waste material at another property can legally oc-
""f- Furthermore, the temporary storage option would eliminate
the need for the disposal of this material at the Beatty, Nevada,
facility. This, in turn, would avoid interstate transfer of waste
and the associated institutional problems as well as the high dis-
posal fees at Beatty.

Temporary Storage Alternatives
The temporary storage alternatives initially evaluated are given
in Table 7. The use of an asphalt pad and hypolon cover was con-
sidered publicly unacceptable and, while all of the options involv-
ing construction of a structure will be perceived as secure by the
public, all are costly.

Recommended Plan—Temporary Storage
Based on the information provided in Table 7, cost-effective
temporary storage alternatives were sought at the site-specific
"Feasibility Study" level. Analysis of the Card Corporation
Property (one of the Denver Radium Sites) revealed that two
warehouse-type structures and vacant parcels owned by Mentor
Corporation could serve as temporary repositories for up to
40,000 yd3 of material.
The U.S. EPA is negotiating a draft agreement with Mentor
Corporation, owners of the Card Corporation site, and has pro-
posed a State Superfund Contract with the State of Colorado re-
garding implementation of this alternative. A tentative agreement
with Mentor Corporation is presented in the form of a draft
CERCLA Section 106 Administrative Order on Consent.
The draft agreement with Mentor Corporation would allow
the U.S. EPA to establish a waste consolidation and storage or
staging area on a 4.9-acre parcel of land on the property. The
parcel would be available to the U.S. EPA for a period of 5 yr at
no charge to the U.S EPA or the State of Colorado. If the facility
is needed beyond that time, the parcel may be rented for a period
of up to 3 yr. All of the material must be removed from the site
by the end of 8 yr, at the latest. Under the terms of the draft
agreement with Mentor Corporation, the U.S. EPA would be
able to store between 24,000 and 40,000 yd' of contaminated ma-
terials from the properties that make up the Denver Radium
Site.
If this alternative is successfully implemented, the U.S. EPA
intends to consider a variety of ways to use the parcel. The U.S.
EPA may modify and use the existing building, known as the
True Truss Building. It may build a new structure on the open
ground included in the parcel, or it may use some combination of
the existing building and new structures. This option would in-
clude choices between bulk and containerized handling and stor-
age of the materials or some combination of bulk and container-
ized handling. Large capacity nylon fabric bags and construction
debris dumpsters with covers installed are included in the range
of potentially suitable containers for consideration.
If a permanent disposal site becomes available early in the 5-
yr life of the facility, it may be used as a transfer point for consol-
idating contaminated materials from several smaller locations and
preparing them for rail or truck transport to the permanent dis-
posal site. On the other hand, the facility could simply be used to
store wastes from sites with the highest priority for cleanup while
permanent disposal arrangements are being made. The various
options described above, together with public comments on them,
would be considered in the design phase of the project.

Cost of Temporary Storage
The cost of temporary storage of contaminated materials from
Card Corporation Property includes construction and demolition
of the temporary storage facility and the additional material
handling. Depending on the storage technique selected, the costs
will vary. For a concrete vault, the representative total capital cost
for storage of the Card Corporation Property material alone un-
MINING & INDUSTRIAL WASTES 363
-------
Table 7
Temporary Storage Alternative Evaluation
Denver Radium Site
Alternative A--
Aaphalt Pad With
evaluation Criteria By pa loo Cov«r
Deicrlptton Atphalt
curb, (o
contrail
piled oo
covered
pod with
r ruBoff
weitea
pad aod
with hypelon
Cone ro to Bunker
Covered concrete
btaifcor cooatnwtad of

Alternative C--
OM
Concrete «!•
Concrete alto con-
structed of poured

Altonutlwe 1*--
Pole Uro With
Concrete, fuoh Well
Coovootlonel etool
build log wltb o eoo-
waato ood rowvai ol
waata «•!•• coovoe-
tlonal tfrotit load log
Alternative I—
throe loo Cover
fcua>kor cooacnMted
of poured OHM role
•alto wit* reaweble
hyp* loo ««r
Alternative f~-
SolUIfy late
Concrete Mocka
Solidify U*e waoto
loco klocM end
atere 1*0 block* lo
«*letUg bulldlOM OB
U»e ilto
AltAnutivo C*-
Dee
Bulk Ufl low
floe* the v*at* lo
M to rial haadlUg
beg* oo4 atore hog*
t» a* fat la« Ml Idle**
oo. tJbo alto
*• |p»|rQUBe)q|j»l

a. Quit Control a. Nay bo very tit- a. MlolMt aloce tho a. MlnlMl, alocc a. Urn eo All. •- o. Mil ho loee of e. tow* control My o. Vary «l"ll»r to
during opera- flcult( will reuulro waite will be eo- the waeta will ho lotenol duet control a problem the* Alt. a he a pro*lo» at tke Alt. F. tool etmUml
tlon and delly covarlog with cloeed, there la oBeloood. However, will bo require*. fc*« tfco Bile will tUxtptle MO*nd t* Mr ta difficult i" -
etoraga bypolao ai wall ee little dual wring duat will Mod to be hove lo he etoblllae* feed the co****"- *"" " '
wetting. If the plocoMot and ator- controlled during dolly. »»to*tlel Covered c«o-v*y«re
kypoloB cover !• age- >«M Internal unloosing aod covered AOMJB of tho hyp* I OB will ho aoodod **d a
•UMgod by natural control will pro** coaveyora will be cover la tooo tfaep bag hoojao oo tbo lead
force*, BlgraitoD ably be needed, at e**ded. for Alt 1. ttortor hooker will bo re-
tro* tho alu !• la*it for worker protocxion wilt OM edited.
poaiiblo. aafety. he oa critical.

b. IbMoir b. Evaporation pood b. KlolMl control b. MlolMl control b. HlalMl coocrot »- OOOM oo«nl will b. MlelMl CMC re 1
will he Modod oeoded. oeodod. eoeded. ho M«ood. eo»4ed. llockalle
will oood
protect loo.

t. Ccoundwvter «. Mot a proolee. c. Hot a prufalea. c. not a proble*. c. Hoc a prt*la». c. «o* te«bolceily c
a arooleoi but tke
•vktlc 007 tool Uore
could b« M IB.11*4
e. Operation
and aafoty e. HlnlMl. d. Nay preaent a d. HlaiMl d. &aM aa Alt. B. d. llojlar EO d, PoteMlol for «. Stocking tM hag*
pr<4iloota problea due to petio- Alta. I «»d 0 but eccfOwoxa wk«» BO«le« exa? b» d**g«rewl.
tlal InhalatlM. Uu Owe to (b>e 00*0 tho "t—terete- bloche,
toy. atacJilog mtty bo 4*e-

«. Aoathotlca e. Mil bo cooaplc- a. Will ho cooaplc- •. lajw aa Alt. 2. a. T*o werebouao' 0. Could bo easily o. NlalM! oootkojlU «. SOOB eo fe* Alt* *.
uoui ood My oot op- uoua but appear M- typo focllltv will acre«Md fr«oi pvbllc lOf)*ct. alece the
pear euf (Ideally cure. kleiwJ U wick (ho vie*. bloc la would bo
aecure to public. eurroao)dLo| area aool atored IB omlatLkg
appear aocaro. helldlaga.
praklM.
to* «r On
• . Dailgn/ a. Slap leal oaalga a. Coae/aratlvoly a. Blallar to a. wSH bo toeo coer- e. timtita t* Alta. a. ftelatlvoly llttl* o.
Cuaatriwtloa aad conatructloB. covpteji deaiga aad Alt. B. plea, than Alta. I, B a»d C. u «*alg» or cae,- lo eeOtfB fa «•'
cooatrtkctloe. C, e*d I but oero eo atrwct. otrntrt.
then Alt. A.

b. Operatic* b. Deity cover tag b. Careful opera doe b. The coovevon My b. Operetlooe ehould a. Oporetora will b. The atrvecth of h. Fllllac iloi 1
to avoid duat ajgra- will ho rt90trod to pro** to be Mtatee- be coeftarably 00*7. hove> Ca oaortloo core iM "cooicroto* htockjt wot (ouo4 dlfflci
tlon will ho a Ait- avoid daetaglbg the aoce lotOBOlvo. Coo- Aa BeotloneJ earlier, to evotld di*a«§« of la wokBotoa. ChlpO)iog;, ot Xlevj* I
flcult Mnagaeat coltaBD* eupporLlog veyaace of aoM of lotormol cootrol My tho fantto w»U oo«- etc., com bo o*- Uikrt* OBM*» oo !«•*
proelM. tie roof. tbo Mtorlal (boerda. poao a ctullettfB. tiooo, Delly cower poctod. tboi Mlt- tor, rvkor, ellda*
rehar. etc.) My oot to o«old *MC prot- abtllcy of eel at leg etc.. lo 4l((lc»tt
The evaporation pood he poaalbU. IOM will ho re- atrvcturea to bowdle to h»|.
will bo e Mlote- vulred b»i wltl tbo r*aultla« a true-
BOACO headache. prafeoktv BM bo oo twrot load a la aot

M«t

c. Clooure c. Uaoval of tho c. Due to tho oo- c. KOODVO! of tho c. Clooure will be c. fteomal of tho «. Coow>or*tl*oty c. Vcrv aloJlaT t*
•atarlal could cauao cloouro, roonval of weata Mterlal will .l.ll.r to tk«x re- uoate trill bo eail.r tleyle cloowo. The Alt. t but probably
duat pro*lee,*..,but tho Mterlal will hove to occur aleul- quired for Alt. B. to coocrol thaa fnr blocka would ho ojowojd leu cloMMB uoold
tbli will bo true for preaaot a reduced taoooutly to deopll- Uowevor. doa0lltlon Alte. A end C. The by lort lift ta bo noulrod. M cm
all optloM to vary- duat praole*. tlcw. Both the weete of tU atracturo guaolta, oaplult. ky- trwcki oo4 hauled bag* wMld prooMi
lag degjreea. *od tbo atructural would bo loee coot- Mtoo), etc,, would away. Soewj cl«a«w oOaiooil teeftlu.
DaonlltloA will b. dobrlt would bo re- plicated then for he reaoved co»curr«at a«4 a*cMla*[t.ailo*i
DlaMndlog the re- wore eoe*lt* ctioa •ovod logether. TMa Alta. t and C. IB to <«ita roowvat. of tho ulatlog wer«-
pocltory ahould bo for Alt. A. will coualUale *iit feel, the bulldlmg houoo focllltleo
relatively atralght coatrol. My U worth etcoo- would Lo 0)o*4ad.
forward. taailnatloa aad reuoo. Tlw Jegree e>f thla
effort le i ii>nn»aii

*. iBpleMHtabnity

a. Public o. Probably tho a. Public will fool a. BOM ea Alt. B. A. BOM ae Alta. B o. flatllar to Alto. o. Sl«Iler to o. $1*4lor fc»
Acceptance leaat, alo«e tbo do- the waite to be aad C. B, C. a«d D, but Alto. B C, o*d D. Alta. I, C,
• lgj> doe* not eppoor alorod Meuroly. there My bo aoOM P,
•ecore. The coocept coocero lor |round-
of toBporary etoreie wotor.
U Denver ewy ho coo-
troveratat. Thla la
true for oil
aptlooa.

D. remitting b. Btate Bay quoi- B. Hloleiel coecaroa. b. Concerae regard- b. HlnlMl conceive, b. Concern* over k. CoBforne retard- h. Similar t«
tloo aafety. tag waato reoovol. f.w. cootaeilnet loo. lag bulldlag *)t. •,
cleanup.
HotaVf Co»t» for co«fiar(«oa puf|»otai only) Includoa coitatruct Ion, Illllng, clotura, and dlipuval of additional Miartala frnai alto roatttiot Ion.

364 MINING & INDUSTRIAL WASTES
-------
der this alternative is estimated at $1,410,500, including the costs
of excavation and permanent disposal. Annual costs for inspec-
tion and monitoring would be approximately $6,000 for 5 yr.
These costs do not include the costs of providing storage ca-
pacity for contaminated materials brought to the facility from
other Denver Radium Site locations. As shown in Table 7, the
estimated capital costs per cubic yard of storage capacity range
between $35 and $69. Applying these costs to the 24,000 to
40,000 yd3 of storage contemplated in the agreement with the site
owner, the lower and upper bounds of the incremental capital
costs for the additional storage would be $840,000 and $2,760,000,
respectively. Actual costs will, of course, depend upon the volume
and mode of storage that is finally selected. For purposes of its
negotiations with the State of Colorado, the U.S. EPA has pro-
posed a total project cost of $2,543,000 for the Card Corpora-
tion Property, and an additional $1,133,000 to provide additional
capacity for materials from other Denver Radium Site locations.
Recommended Plan—Permanent Storage
While none of the permanent disposal options is immediately
available, due to myriad institutional problems, they still remain
feasible over the longer term. The use of an UMTRAP site for
co-location of a repository for the Denver Radium material ap-
pears to remain as the best solution from environmental, institu-
tional and cost standpoints. At this point, the Union Carbide/
UMETRO Uravan site is the U.S. EPA's preferred choice, al-
though the Grand Junction site also would be acceptable.
However, if an agreement with DOE cannot be developed,
numerous technically suitable single-use sites can be identified
near Limon, Colorado. Use of any of these sites can be expected
to be met with institutional restrictions resulting from local public
resistance.
Under either option, the State of Colorado must initiate the
permitting for permanent disposal as quickly as possible if it is to
have a site within the time-frame permitted by the agreement with
Mentor Corporation.
Plate 1
Photographs—Site 11
Plate 2
Photographs—Site 11
View of Site 11 looking north from Colorado 86. Note 1-70 in the
background.
View from Site 11 to the Riparian Corridor of Big Sandy Creek approx-
imately Vi mile to the south.
v«w of Site 11 looking west from the 1-70 off ramp. Union Pacific Rail
s just west of the stock pond.
Evidence of drainage approximately 1/8 mile west of Site 11. Note
reaches upstream were dry. Numerous swallows were resting under the
bridge.

MINING & INDUSTRIAL WASTES 365
-------
Plate 3
Photographs—Site 12
View of Site 12 looking southwest.
View of Site 12 looking northwest. Note paved ro»d.
Plate 4
Photographs—Site 13
Big Sandy Creek near Boyero, Colorado
Union Pacific Line passing through Hugo. Colorado
Plate 5
Photographs—Site 13
View looking west in the vicinity of Site 13. Distant tree line represents
Riparian Corridor of Big Sandy Creek. Union Pacific Rail Line parallels
the creek.
View looking east from the Site 13 area.
366 MINING & INDUSTRIAL WASTES
-------
Plate 6
Photographs—Site 10
View of the Manila Site looking west. Note the Front Range Airport
terminal at the center of the photograph.
View of Boettchers Airport Industrial Park just south of the Manila Site.
MINING & INDUSTRIAL WASTES 367
-------
Heap and Dump Leaching and Management Practices
To Minimize Environmental Impacts
Robert L. Hoye
Robert L. Hearn
PEI Associates, Inc.
Cincinnati, Ohio
S. Jackson Hubbard
U.S. Environmental Protection Agency
Water Engineering Research Laboratory
Cincinnati, Ohio
ABSTRACT
Literature reviews, interviews and site visits were conducted to
compile information on the current status and extent of the prec-
ious metals heap leaching industry and the copper dump leaching
industry. A detailed description of the design and operation of
heap and dump leach facilities was prepared. Information on the
toxicity and mobility of cyanides as related to help leaching and
the potential for acid generation as related to dump leaching was
summarized. Alternative management practices that could miti-
gate actual or potential impacts that may be caused by these in-
dustries were evaluated. The costs associated with these alterna-
tive practices were estimated on a hypothetical site basis. Eighteen
heap and dump leach operations were visited to obtain informa-
tion on current industry and site-specific practices.''2

INTRODUCTION
In 1976 Section 8002(0 of RCRA required the U.S. EPA to
conduct an investigation of all solid waste management practices
in the mining industry. That mandate specifically directed the
U.S. EPA to conduct "...a detailed and comprehensive study on
the adverse effects of solid wastes from active and abandoned sur-
face and underground mines on the environment, including, but
not limited to, the effect of such wastes on humans, water, air,
health, welfare, and natural resources."1
When Congress amended RCRA in 1980, it added Section
8002(p), which directed the U.S. EPA to conduct a "...detailed
and comprehensive study on the adverse effects on human health
and the environment, if any, of the disposal and utilization of sol-
id wastes from the extraction, beneficiation, and processing or
ores and minerals.'" Moreover, Congress required the U.S. EPA
to make a "regulatory determination" within 6 mo after submit-
ting the study to Congress, stating either that regulations would
be promulgated or that regulations were unwarranted for such
mining wastes. A report was submitted to Congress on Dec. 31,
1985.'
On July 3, 1986, the U.S. EPA published the regulatory de-
termination regarding the issue of mining wastes that heretofore
had been excluded from regulation under RCRA.' The deter-
mination indicated that RCRA's hazardous waste management
standards "...are likely to be environmentally unnecessary, tech-
nically infeasible, or economically impractical when applied to
mining wastes." The U.S. EPA indicated that it plans to develop
a special program for mining wastes under Subtitle D, but
acknowledged that, after additional study, some mining wastes
may have to be regulated under Subtitle C. The U.S. EPA spe-
cifically expressed continued concern about problems and poten-
tial problems associated with mining wastes having high-acid
generation potential, radioactivity, asbestos and cyanide. The
U.S. EPA's current policy as stated in the regulatory determina-
tion regarding active heap and dump leach piles and leach solu-
tions is that these materials are not wastes, but are raw materials
used in the production process and a product, respectively.' Only
leach solutions that escape from the production process and aban-
doned leach piles are wastes.
To develop a special mining waste regulatory program, the
U.S. EPA is collecting additional information on the nature of
mining wastes, mining waste management practices and mining
waste exposure potential. Toward this end, the U.S. EPA's Office
of Research and Development contracted PEI Associates, Inc.,
to conduct an evaluation of the precious metals heap leaching
and copper dump leaching industries. The objectives of the
efforts were to characterize the industries, describe current design
and operational practices, summarize environmental concerns
and evaluate alternative management practices that could miti-
gate impacts which may be caused by cyanide and acid contam-
ination from these industry segments.

INDUSTRY STATUS
The application of both heap and dump leaching has increased
in recent years because of the relatively low capital investments
and fast payouts involved. These techniques allow recovery of
low grade resources that otherwise could not be profitably ex-
tracted.

Heap Leaching
The mining industry first became interested in the U.S. Bureau
of Mines' developments in gold/silver heap leaching technology
in the late 1960s, and the first commercial cyanide heap leaching
process was used at the Carlin Gold Mine Company in northern
Nevada on mine cutoff material. Since the early 1970s, interest
in heap leaching has continued to grow primarily in response to
the high prices of gold and silver. Low-grade (e.g., 0.03 oz/ton)
gold deposits previously considered uneconomical to recover are
now being exploited at a profit.' Currently, 79 gold and silver
heap leaching operations are active in the United States. The ma-
jority (60) of these operations are in Nevada. Four of the active
368 MINING & INDUSTRIAL WASTES
-------
heap leaching operations are in California, three in Colorado,
two in Idaho, two in Montana, two in New Mexico, five in Utah
and one in South Carolina. The approximate locations of these
facilities are indicated in Fig. 1 and 2.
In 1984, 524,553 troy ounces of gold were recovered from
19,857,613 tons of ore treated by cyanide heap leaching.' By com-
parison, 1,136,504 troy ounces of gold were recovered from con-
Jhunder Mm. * •Zortman-Landusky
KEY
I) BATTLE MOUNTAIN
2) BIHGHAM CANTON
3) MINERAL PARK
4) BAGDAD
5.6) INSPIRATION. OXHIDE
7.B) BIAHI LEACH. PINTO VALLEY
9| RAT
10.11) LAKESMORE. VANDYKE-
12) SAN MANUEL/KAI.AMAZOO
13) SILVER BELL
l<) SIEB.RITA/ESPERANZA
IS) COPPER QUEEN
16) JOHNSON
17) MORENCI/METCALF
16) TYRONE
19) CHINO
20) NACIMIENTO-
' \South Carolina
Mexico \^
Figure 3
Geographic Distribution of Active Copper Leaching Operations in the
United States
0 220 miles

Approximate Scale

' See Figure 2-3.
Figure 1
ngurc i
Approximate Location of Active Gold/Silver Heap Leach Operations
KEY
ARBITERTTACH
BERKELEY
BUTTE MILL LEACH
RIO TINTO
BIG HIKE
YERRIN&TOM
KIMBERLV/SUNSN1NE
OHIO COPPER
MILLFORD
LISBON VALLEY
UNITED VERDE
ZONIA
U) BLUEBIRD. INSPIRATION
COPPER CITIES
RAY
LAKESHORE
OLD RELIABLE
NCW CORNELIA (AJO)
21) SAN IAV1ER. TWIN BUTTES
PEACOCK
BURRO FOUNTAIN
Figure 4
Geographic Distribution of Inactive Copper Leaching Operations in the
United States
Figure 2
Distribution of Gold/Silver Heap Leach Operations in Nevada
Figure 5
Conceptual Flow Diagram of Typical Heap Leach Operation
MINING & INDUSTRIAL WASTES 369
-------
ventional cyanidation extraction in vats, tanks and closed con-
tainers.' The estimated average ore grades recovered, based on
these figures, are 0.03 ounces of gold/ton of ore by heap leaching
and 0.09 ounces of gold/ton of ore by conventional cyanidation.
The application of cyanide heap leaching has grown in recent
years, and this trend is expected to continue.

Dump Leaching
Currently, there are 18 active copper dump leach operations
in the U.S. Arizona has the majority of these facilities (14), New
Mexico has two and Utah and Nevada each have one. Copper
dump leaching has been conducted in this country for about 100
yr. There are 22 inactive and abandoned dumps located in the
western states The approximate locations of active and inactive
dump leach operations are indicated in Fig. 3 and 4, respectively.
Although total primary production of copper has declined over
the last 10 yr, the percentage of copper produced by leaching
operations has increased during the same period. In the future, as
lower grade ores are mined and the costs of conventional milling
and smelting continue to rise, leaching is expected to account for
an increasing percentage of the total primary copper production.
Some researchers estimate that by 1990 leaching will account for
as much as 24 to 30% of the annual copper production.'

DESIGN AND OPERATION OF
LEACHING FACILITIES
Heap and dump leaching are hydrometallurgical processes,
percolation leaching, that recover a product metal from low grade
ore, tailings and mine waste rock. The product metal is then re-
covered from solution by precipitation, carbon adsorption (gold)
or electrowinning (copper). Leaching permits recovery of metal
values in low grade ores or waste that could not be economically
recovered by conventional beneficiation. Heap leach operations
involve the use of liners and specially constructed leach pads and
solution ponds. Dump leaching involves larger piles leached on
native terrain.

Heap Leaching
The basic design and operational layout of heap leach projects
are similar at all facilities. Low-grade ore (typically from a surface
mine) is stacked 15 to over SO ft high in engineered heaps on
sloped (1 to 6%), impermeable pads, and a weak alkaline cyanide
solution is sprayed over the ore. The solution percolates through
the heap and dissolves finely disseminated free metal panicles
(gold and/or silver). Care is taken during the construction of
heaps to ensure that the material is uniformly permeable.
The design engineering and construction of liners in this indus-
try have reached a high level of sophistication. Pads, V* to 50
acres, are constructed of native or modified clays, synthetic lin-
ers (e.g., HOPE, PVC or Hypalon) or asphalt. This design helps
ensure that product and reagents are not lost through seepage.
The pads must be capable of providing structural support without
suffering damage from deflection due to the weight of the ore or
equipment traffic. Selection of pad materials and specifications
is determined by site-specific parameters such as availability of
local materials, slope, geotechnical properties of the sub-base,
temperature variations and operational considerations (i.e.,
single- or multiple-use pads).
The pregnant solution flows over the pad to a lined collection
ditch. The ditch carries the gold-bearing cyanide solution to a
lined pregnant solution pond. Pregnant solution then is pumped
to a recovery plant, where the metal product is removed by car-
bon adsorption followed by stripping and electrowinning or by
precipitation with zinc followed by filtration (Merrill-Crowe zinc
dust precipitation). The barren solution then is pumped to a lined
holding pond, where it is treated with additional NaCN and caus-
tic (e.g., lime or caustic soda). Sodium cyanide is the only com-
mercially proven lixiviant. It is added to maintain a concentra-
tion in the barren solution of approximately 1 Ib/ton of solution.
The optimal pH for the gold dissolution is 10.3. From the bar-
ren pond, the solution is pumped to the heap again and sprayed
over it to complete the closed-loop cycle. Heap leach operation!
are typically zero discharge facilities.
The leaching cycle is relatively short (20 to 90 days), especially
when compared to that of dump leaching (decades). At comple-
tion of leaching operations, the leach ore is rinsed with fresh
water to remove residual cyanide and dissolved metal. With few
exceptions, heap leach residue (the barren ore remaining after
precious metal values have been extracted) is left in place on the
pad. At a very few operations it is excavated, hauled by truck
and disposed of in an on-site disposal area.
A conceptual flow diagram of the heap leach operation is pre-
sented in Fig. 5. Although the basic process just described is sim-
ilar at all operations, each site is unique, and several alternative
approaches exist. Specific leaching times, reagent use, flow rates,
heap dimensions, pad construction, pond capacities, liner ma-
terials and other design and operational parameters vary from
site to site depending on the characteristics and quantity of the
ore and the climate, topography, hydrology and hydrogeology of
the site.

Dump Leaching
Dump leaching is used to extract copper from overburden and
waste rock generated by conventional (principally open pit) min-
ing operations. This material contains sulfide minerals and typi-
cally has a copper content of less than 0.3%." These materials
are too lean in copper to be smelted directly or to be beneficiated
to produce concentrate. These sulfide bearing materials normally
are deposited directly on the ground in large dumps adjacent to
the mining operation to minimize hauling costs and increase the
efficiency of the operation. Naturally sloping terrain typically is
selected to facilitate the placement of the material and to allow
collection of leach solutions.
Leach dumps typically cover hundreds of acres, rise to heights
of 200 ft or more and contain several million tons of uncrushed,
low-grade ore and waste rock. Leach dumps are constructed by
end-dumping in lifts on top of previously leached material. Large
dumps are raised in lifts of SO to 100 ft. An aqueous sulfuric acid
solution is used as the lixiviant. The pH of the leach solution is
usually between 1.5 and 3.0. This pH range is favorable for both
the dissolution of copper minerals and the maintenance of an ac-
tive bacterial population; it also minimizes the precipitation if
iron salts." Frequently, only makeup water is needed because the
oxidation of the sulfide minerals within the dump replenished
acid in the leach solution.
Typical application rates range from 0.01 to O.OS mVday/m' of
horizontal surface.10'11 Leach solutions are introduced onto the
dumps by a variety of methods:
• Flooding the surface in a series of small diked ponds
• Spraying from hoses or through sprinkler heads
• Injection through holes cased with perforated pipe
• A combination of the above
The solution distribution method depends on the climatic con-
ditions, dump height, surface area, mineralogy and permeability
of the leach material. In practice, most dumps are leached in
sections.
Pregnant leach solution drains from the base of the dump into
a collection channel and/or pond. This solution is pumped to a
precipitation or solvent extraction plant where the copper in solu-
370 MINING & INDUSTRIAL WASTES
-------
tion is recovered. Barren solution is returned to the dump where
makeup water and acid can be added before application.

ENVIRONMENTAL CONCERNS
Heap Leaching
Because cyanide is the lixiviant used in heap leaching of prec-
ious metals, there is concern over the potential for release of toxic
cyanides into the environment. The cyanide content in leach solu-
tions is maintained at about 250 mg/1. Because an alkaline pH is
maintained in the solution, most of the cyanide is present as free
cyanide, as required in the leaching reaction. The barren solution
pond typically holds hundreds of thousands of gallons of this
solution. The pregnant solution pond contains lesser concentra-
tions of free cyanides because of the destruction and complexa-
tion that occur in the heap; however, a significant concentration
of free cyanides may be present. The solution in these impound-
ments represents the greatest source of free cyanide at a leach
operation. Failure of the containment system, liner failure or
overtopping of the pond would result in the release of free cy-
anide in an alkaline solution to the environment.
Cyanide in leach residue occurs in combinations of various
metallo-cyanide complexes and possibly free cyanides. Cyanide
complexes vary from strongly bound forms to others that disso-
ciate more readily. The complexes in a given heap are determined
by the mineralogy of the ore. Essentially no data are available on
the content and fate of cyanides in leach residue. There are no
reports of cyanide contamination or migration from properly
constructed and operated heap leach operations. However, there
have been a few reported incidents involving pond failure or over-
topping and contamination resulting from clandestine operations
that did not use typical operational practices.

Dump Leaching
As with heap leach operations, the following potential mechan-
isms for environmental releases exist at dump leach operations:
• Seepage from the vase of the dump
• Leakage from solution ponds and ditches
•Spills
• Pond failures and overtopping
Unlike heap leach operations that use engineered pads, dump
leach operations are constructed on native ground and cover
much larger areas, thus offering greater potential for seepage.
The contaminants of concern are those metals contained in the
leach material that are soluble at low pH and, perhaps more im-
portant, are chlorides, sulfates and TDS. Several case studies
have implicated dump leach operations as sources of ground-
water contamination.

ALTERNATIVE MANAGEMENT PRACTICES
Heap Leaching
A limited number of alternative management practices can be
applied to minimize the potential for cyanide contamination from
heap leach operations. These practices include alternative liner
construction, oxidation of cyanide during post-leach flush-
out and use of reagents other than cyanide. Most heap leach oper-
ations are relatively small and their only sources of potential
contamination are the heaps themselves and the two process solu-
tion ponds. Once operations cease, the heap remains as the only
potential source of pollution; the ponds must be emptied during
closure. Additionally, most obvious controls (such as pond and
leach pad liners, surface water diversions and post-leach rinsing)
are already standard practice in the industry. Although the need
for controls beyond those currently in use has not been demon-
strated, the concerns related to potential releases of cyanide may
warrant additional controls or overdesign of existing controls.
The management practices that were evaluated are listed below:
Operational Phase
Pre-operations
Operational

Closure

Post-closure
Management Practice
Installation of French drains beneath
pads and pond liners.
Use of RCRA double liner systems with
leak detection in ponds.
Use of alternative lixiviants
More extensive groundwater monitoring
Flush heaps with cyanicide
Recontour and cap heaps
Long-term maintenance of heaps and
monitoring systems and site security
Most of the controls listed above have been incorporated into
the design and operation of at least one existing heap leach facil-
ity. The feasibility and cost of these controls at other locations
would have to be determined on a site-by-site basis. They would
depend on differences in mineralogy, topography, geology, hy-
drogeology, climate and design and operational characteristics.
The Pinson Mining Company has installed a system of French
drains beneath its leach pads. Individual pads measure about
300 ft by 300 ft. Ultimately, as many as 60 contiguous pads will
be constructed. Slotted PVC pipe is installed in a gravel bed be-
neath the clay pad. These drains connect with sumps that are
monitored for the presence of seepage. Another facility, the Stib-
nite Project, incorporates seepage collection drains sandwiched
in gravel between an upper asphalt liner and a lower PVC liner.
These drains discharge any seepage to the pregnant solution
pond. The use of such drains allows immediate detection of leak-
age and can provide for collection. The cost of a conceptual sys-
tem similar to that used at the Pinson Project was evaluated. It
was estimated that incorporation of the drain system would
double the cost of the pad ($1.80/ft2 vs $0.93/ft2).
The use of double liners in solution ponds is technologically
feasible and is a demonstrated practice at some heap leach sites.
A double-liner system consisting of two layers of 40-mil HDPE
separated by a leachate detection and collection system was eval-
uated. The pond was assumed to be 300 ft by 150 ft (approxi-
mately 1 acre). For the purpose of comparison, the costs asso-
ciated with a single 40-mil HDPE liner system, believed to be
common in the industry, were estimated. The cost comparison in-
dicates that the double-liner system increased the cost of the pond
by a factor of at least two. The cost of constructing the solution
ponds at a site can represent a significant percentage of the total
capital cost of the operation.
Cyanide is the only lixiviant currently used at commercial heap
leach faculties. Because of the actual or perceived toxicity asso-
ciated with cyanide, the question of the availability of suitable
substitutes for cyanide is raised. The development of alternative
lixiviants (e.g., thiosulfate, malononitrile and thiourea) is still in
the laboratory or pilot-scale testing stage. If alternative lixiviants
are developed, the environmental impacts associated with their
use must be fully evaluated. While thiourea can rapidly leach
gold from leach ore, it requires a very acidic medium (pH 1) that
would be an environmental concern. Additionally, reagent con-
sumption and cost are high.
The type and sophistication of groundwater monitoring sys-
tems vary considerably in this industry. The requirements for
these systems are specified on a site specific basis by state regu-
latory personnel. The installation cost of a detection monitoring
MINING & INDUSTRIAL WASTES 371
-------
system will vary greatly from site to site. The primary factors that
influence costs are the size of the operation and the complexity
of local hydrology. The principal factors are: diameter, depth and
components of the wells; drilling specifications; geologic ma-
terial; sampling and analytical requirements; and site access. Es-
timates made for an example site indicate that the costs of in-
stalling a system of 10 to 13 wells to depths of 25 to 300 ft would
range between $12,500 and $195,000. Consultant fees for a qual-
ified hydrogeologist could be expected to range from $6,000 to
$50,000. Analytical costs could amount to $12,000 to $16,000
annually plus reporting and recordkeeping. These costs empha-
size the great variability due to site-specific conditions.
During the post-closure period, the heap leach residue is the
only potential source of cyanide contamination. Current practice
is to rinse the leach residue with fresh water for a predetermined
time or until some preset cyanide concentration (e.g., 0.2 mg/1)
or pH (e.g., pH8) in the rinse water is achieved. An additional
control option could be the addition of a cyanicide, a strong oxi-
dant, to the rinse water. Alkaline chlorination is proven technol-
ogy for cyanide destruction and is the most highly developed of
the available methods in terms of experience, simplicity, con-
trol, availability of equipment and engineering expertise. This
process destroys all cyanide except iron cyanide and the more
stable metallo-cyanide complexes. Treatment of heap leach res-
idue by alkaline chlorination has been carried out at a few oper-
ations.12-13 When this system is used during the operational
period, the facility must incorporate at least one additional pond,
a neutralization pond, in its solution management system. If it is
used only at closure, the existing process solution ponds would be
adequate.
Application of a clay or synthetic cap over leach residue could
prohibit infiltration and run-on and thereby preclude leachate
formation. However, it would hinder the natural degradation of
cyanide by limiting volatilization and photodecomposition.
Sixty of the 79 heap leach operations are located in Nevada in
arid climates where capping may provide even fewer marginal
benefits. In order to place a cap, the side slopes of the heap would
have to be reduced to at least 3:1 or more from the 1:1 slopes
existing during operations. Assuming that a suitable source of cap
material exists near the site, recontouring and capping a 1 acre,
15-ft high heap would cost about $40,000 while a 50 acre, 100-ft
high heap would cost about $2 million.

Damp Leaching
Alternative controls for dump leach operations are much more
costly to implement than those for precious metals heap leaching
because of the huge size of the dumps. As mentioned previously,
typical dumps cover hundreds of acres. Square miles of ground
may be covered by leach dumps at a single mine site. Additional-
ly, the leaching of these dumps is a long, slow process lasting up
to 100 yr. During the operational phase, controls are limited to
enhanced groundwater detection monitoring, design redundan-
cies in the solution management system and run-on/run-off con-
trols. By definition, dumps are leached on native ground; engi-
neered impervious pads are not used. However, some copper min-
ing operations are now using heap leach methods (i.e., construct-
ing synthetic or clay pads) to leach copper ores in special situa-
tions.
Seepage through the soils that underlie the leach dumps and
solution collection systems is the most significant potential mech-
anism for the release of contamination. Most copper leaching
operations have implemented a system of management practices
that includes one or more mitigative measure designed to min-
imize solution losses. Historically, such management practices
were implemented solely for economic reasons (to improve cop-
per recoveries). The potential for long-term impact exists even
after production leaching operations have ceased because of the
acid formation potential that may be associated with the mineral-
ogy. To prevent leachate formation and migration, inactive
dumps could be recontoured and capped; the dumps could be
treated with lime or other alkaline material to neutralize acid
that is formed; or physical (i.e., grout walls) or hydraulic barriers
(e.g., extraction and injection wells) could be used to control
groundwater flow. The size of the dumps makes any of these con-
trols costly and difficult, if not impossible.

CONCLUSION
The low production costs, relatively short startup time and rela-
tive simplicity of heap and dump leaching techniques have led to
an increased use of these methods to recover precious metals and
copper that otherwise are not economically recoverable. Current
state-of-the-art design, construction and operation of precious
metals heap leach facilities incorporate obvious controls includ-
ing impervious leach pads, lined collection trenches and process
ponds, and closed loop zero discharge solution management.
Depending on site specific considerations, it may be beneficial to
incorporate redundancies and overdesigns into these systems.
However, the need for additional controls currently is not docu-
mented. Additionally, research to determine the presence, fate
and toxicity of cyanide in heap leach residue is just beginning.
By virtue of their sheer size, there are few alternative man-
agement practices applicable to copper dump leach operations.
Impacts on groundwater by dump leach operations have been
identified. Because most of these operations have been operating
for decades and have decades of life remaining, it is difficult
to retrofit controls. Some newer copper leach operations have
been designed as heap leach facilities, but only when econom-
ically permissible. The long-term impact of inactive dump leach
piles has not been fully assessed, but concern exists over the
potential for these piles to generate acidic leachate and cause
long-term impacts to the local environment.

REFERENCES
1. PEI Associates, Inc., "Gold/Silver Heap Leaching and Conceptual
Management Practices 10 Control Cyanide Releases." Draft Re-
port. Prepared for U.S. EPA, Office of Research and Development,
Water Engineering Research Laboratory, Contract No. 68-02-
3995 (024). Sept. 1986.
2. PEI Associates. Inc.. "Mitigation Techniques for Copper Heap and
Dump Leaching." Draft Report. Prepared for U.S. EPA, Office
of Research and Development, Water Engineering Research Labor-
atory. Contract No. 68-02-3995(025). Sept. 1986.
3. Resource Conservation and Recovery Act of 1976, PL94-580,
Oct. 21, 1976.
4. Solid and Hazardous Waste Disposal Act Amendments, PL-96-482.
Oct. 21. 1980.
5. U.S. EPA, "Report to Congress—Wastes From the Extraction and
Beneficiation of Metallic Ores, Phosphate Rock, Asbestos, Over-
burden From Uranium Mining, and Oil Shale," EPA/S30-SW-85-
033, Dec. 1985.

6. U.S. EPA, "Regulatory Determination for Wastes From the Extrac-
tion and Beneficiation of Ores and Minerals," 40 CFR Part 261,
Federal Register Jl(\28):24496-24502, July 3. 1986.
7. Lewis, A., "Leaching and Precipitation Technology for Gold and
Silver Ores," Engi. and Mining J.. June 1983.
8. Bureau of Mines, "Gold." Preprint from the 1984 Bureau of Mines
Minerals Yearbook.

9. Hiskey, J.B., "Arizona Bureau of Geology and Mineral Technol-
ogy." Tucson, AZ. Personal Communication. Aug. 20, 1986.
372 MINING & INDUSTRIAL WASTES
-------
10. Biswas, A.K. and Davenport, W.G., Eds. Extractive Metallurgy 12. Milligan, D.A., "Cyanide Destruction. Chapter 14 in Short Course
of Copper. 2nd ed. Pergamon Press, New York, NY, 1980. on Evaluation, Design, and Operation of Precious Metal Heap
Leaching Projects," AIME, SME, Conference. Albuquerque, NM,
Oct. 15, 1985.
13. Statts, W.G. "Handling Cyanide at Superior Mining Company's
11. Sheffer, H.W. and Evans, L.G. "Copper Leaching Practices in the Stibnite Heap Leaching Operation, in Conference on Cyanide and
Western United States," Bureau of Mines Information Circular the Environment," Tucson, AZ. Dec. 1984, Published by Colorado
8342,1968. State University, Fort Collins, CO.
MINING & INDUSTRIAL WASTES 373
-------
Abandoned Steel Manufacturing Site — A Case History
G.J. Anastos, Ph.D., P.E.
T.M. Legel, P.E.
Roy F. Weston, Inc.
West Chester, Pennsylvania
J.M. Perdek
U.S. Environmental Protection Agency
New York, New York
ABSTRACT
This paper discusses the potential environmental considerations
which may exist at an inactive steel plant. In addition, a case study
is presented of an abandoned New Jersey steel plant with wastes
stored on-site that were previously listed under RCRA re-
quirements, now listed on the U.S. EPA's National Priorities
List. The ongoing case study identified many problem areas
relating to the investigation of large abandoned industrial
facilities, i.e., the presence of baghouse dust and petroleum hy-
drocarbons over major areas of the site and numerous drums,
tanks and transformers scattered throughout the site.

INTRODUCTION
The demise of the steel industry in the United States has forced
the closing of many steel plants. Some of these plants are now
classed as abandoned or uncontrolled hazardous waste sites. The
U.S. EPA has added some of these sites to the NPL for character-
ization and eventual remediation. The investigation and cleanup
of these sites can be extremely costly due to the size and complex-
ity of the plant facilities. Prior to the enactment of Pollution Con-
trol laws, steel plants rarely saw the need to dispose of process by-
products and wastes in a controlled manner. The sites and the
volumes of waste material usually were so large that on-site dis-
posal of waste material was the most cost-effective method. When
the plants were closed, it usually happened without warning and
under economic duress; therefore, no effort was made to stabilize
or clean up the site prior to closing. In many cases, process tanks,
plating baths and process piping were left in an operable mode.
This paper discusses the potential contamination sources which
may be found at an abandoned steel plant.
One such New Jersey plant listed on the NFP that presently is
being investigated is discussed. Preliminary site characterization
activities identified: baghouse dust (an RCRA waste) and
petroleum hydrocarbons on major portions of the site;
laboratories with various hazardous chemicals (e.g., picric acid);
over 2,500 drums containing various raw and spent chemicals
ranging from strong oxidants (e.g., potassium permanganate) to
chlorinated solvents (e.g., methylene chloride); plating and other
chemical treatment baths; and a wastewater treatment system.
This case study identifies many problem areas relating to the in-
vestigation of large abandoned industrial facilities. Approaching
the site through detailed site characterization was not feasible. In-
stead, the site was divided into units which could be independent-
ly screened and characterized in depth.

374 MINING & INDUSTRIAL WASTES
CONTAMINATION SOURCES WHICH ARE
EXPECTED AT ABANDONED STEEL PLANTS
The potential sources of contamination that are expected at
steel plants can be divided into two general categories: those
associated with specific processes and those common to large in-
dustrial operations. The specific process wastes vary widely and
are dependent upon the type of plant. In general, those wastes
that may be encountered include:
Slags that can result in leaching of heavy metals
Baghouse dust that can result in the leaching of heavy metals
Coke oven tars that can result in the release of PNAs and VOAs
Ash that can result in the leaching of heavy metals
Metal treatment solutions that can result in the leaching of
heavy metals

In addition to these specific process wastes, other wastes
associated with large manufacturing operations include:

• Wastewater treatment plant sludges that include heavy metals
and hydrocarbon contamination
• Landfills that may include solid process by-products, plant
trash, drums of spent chemicals, asbestos, transformers and
capacitors
• Lagoons that may have been utilized to equalize wastewater
flow or disposal or bulk spent chemicals
• Laboratories that may contain strong acids and bases, solvents
and oxidants
• Fuel storage facilities above and below grade that may have
leaked over the years
• Paint shops and repair shops that may contain solvents, oils
and grease and other miscellaneous chemicals
• Transformers that may contain PCBs
• Asbestos on pipes or in buildings

This section of the paper does not provide an all-inclusive list of
expected wastes and resultant contaminants, but does identify the
many different types of contamination that can result from "nor-
mal operation" at industrial facilities. The following sections
discuss the types of waste identified in the preliminary characteri-
zation of the Roebling Steel site.

CASE STUDY: SITE LOCATION
AND HISTORY
The Roebling Steel site is located in the town of Roebling in
Florence Township, Burlington County, New Jersey. The site is
-------
an approximately 200-acre, partially abandoned industrial facility
bounded by the Delaware River to the north, Crafts Creek to the
east and the residential community of Roebling to the south and
west.
The John A. Roebling's Sons Company established the site in
1906 as a steel mill manufacturing steel wire and cables from scrap
steel and pig iron. The site operated under the management of
four different companies until June 1981; portions of the site cur-
rently are leased to several tenants.
The site is owned by the John A. Roebling Steel Company but,
due to financial obligation defaults, the site is under the control
of the U.S. Economic Development Authority. The site had been
used primarily to produce steel and wire for approximately 75 yr.
In recent years it also has served a variety of industries including
use as a polymer reclamation facility, a warehouse, a repair/
refurbishing facility for refrigerated trailers and shipping con-
tainers and an equipment storage facility for a construction com-
pany.
The waste products generated by the past steel manufacturing
and various other operations, along with some raw materials and
products, were stored and/or buried in many different locations
around the site.

SITE STATUS
The Roebling Steel site is a U.S. EPA federal remedial (fund-
led) site with ongoing enforcement action for cost recovery. It is
an NPL site, listed in Group 6. NPL sites are ranked into 11
groups based on hazard potential, with Group 1 being the most
hazardous.
The initial basis for listing the site on the NPL was the presence
of two hazardous wastes listed under RCRA:
• Baghouse Dust—listed as: KO61-Emission control dust from
the electric furnace production of steel
• Wastewater Treatment Sludge—listed as: FOO6-Wastewater
treatment sludges from electroplating operations except for
tin-, zinc- or aluminum-plating on carbon steel

The wastewater treatment sludges subsequently have been de-
listed under RCRA. Several remedial response activities have
occurred at the site since June 1979. Major removal actions con-
ducted at the site include the removal of 4,500 gal of non-
recoverable waste oil in November 1979, and the removal of
20,000 gal of waste oil (600 drums) and 60 yd3 of contaminated
soil in July 1981. Also, during December 1985, picric acid was
removed from the laboratory.
The review of background information and the initial site re-
connaissance also identified the following concerns:
• Contamination of surface soils with the baghouse dust and
petroleum hydrocarbons from past operations
• Presence of abandoned drums, tanks, transformers and rail-
road cars
• Presence of three large waste disposal areas
• Contamination of groundwater resulting from infiltration of
contaminants through surface soils

ENVIRONMENTAL SETTINGS
The Roebling Steel site is located along the south bank of the
Delaware River, just west of the Crafts Creek discharge into the
river. Except for a few low-lying areas along the Delaware River
and Crafts Creek, the site generally is above the 100-yr flood
plain. The site is relatively flat, with surface elevations varying
from 20 to 50 ft above sea level.
Site soils are from the Galestown-Klej association of sand and
sandy loam. Most of the site has been reworked and "urbanized"
over the years, and large areas of the river bank have been filled
with slag and other residues from the steel-making process.
Stormwater runoff is collected in storm sewers which discharge to
the Delaware River and Crafts Creek.
Geologically, the Roebling Steel site is located on the western
edge of the Atlantic Coastal Plain within the outcrop area of the
Raritan and Magothy Formations. The Raritan/Magothy Forma-
tions are the most productive aquifers in Burlington County.
Possible influences on local groundwater flow patterns include
Crafts Creek on the eastern margin of the site and a former
stream channel in the north central portion of the site. Although
no longer active, the old stream bed may represent a buried zone
of permeable alluvial sediments.
The effect that the site has had on the environmental quality of
the area's surface water and groundwater is not presently known.
Virtually no quantitative monitoring of the environment has been
done. Neither the nearby wells nor the receiving surface waters
have been sampled. In fact, no groundwater or surface water data
exist to either document or disprove contamination suspected to
be associated with the Roebling Steel site.

CONTAMINATION PROBLEM DEFINITION
In its current, partially active state, the Roebling Steel site con-
tains numerous possible sources of hazardous materials and
hazardous/nonhazardous waste materials, all posing potential
threats to public health, safety and the environment. These poten-
tially detrimental sources include two inactive wastewater treat-
ment sludge lagoons, steel furnace slag disposal areas, emission
control (baghouse) dust, electrical transformers potentially con-
taining PCB oil, an abandoned landfill, various tanks and drums
containing oil and other potentially hazardous substances, demoli-
tion debris and tires and buildings containing assorted chemicals,
including acids. Furthermore, it is anticipated that additional on-
site wastes may be discovered.
A phased remedial action feasibility study (RI/FS) was initiated
by the U.S. EPA at the site: Phase I—Preliminary Site Character-
ization, Phase II—Complete Site Characterization and Phase HI
—Feasibility Study. The Phase I field activities have been com-
pleted and Phase II activities are underway. The following section
describes the Phase I findings.

PRELIMINARY SITE CHARACTERIZATION
The preliminary site characterization of the Roebling Steel site
began in September 1985.
The first priority was to obtain a complete inventory of all
potential on-site contamination sources. These sources included
drums, tanks, transformers, railroad cars and baghouse dust
deposits.
In all, 202 acres of land area and 55 buildings were searched for
contaminants. The objectives of the inventory included the
following:

• Enumeration and location of all abandoned drums, tanks,
transformers, pits/sumps and railroad cars. Materials presently
under control of the tenants were not inventoried.
• Characterization of the above through visual observation and
inspection.
• Evaluation of sampling and analytical programs needed to de-
termine the potential hazards associated with each of the
above.

In addition to this inventory, a ground penetrating radar (GPR)
survey and a magnetometry survey were performed at selected
areas throughout the site to identify possible buried materials.
MINING & INDUSTRIAL WASTES 375
-------
These two geophysical techniques were used in a complementary
manner to locate disturbed areas and metal objects which could
be indicators of possible waste burial sites. The areas for in-
vestigation were chosen subsequent to the examination of
historical information, examination of aerial photographs dating
back to 1940 and personal interviews with former steel plant
employees.
The second priority was to evaluate other potential uncon-
tained sources of contamination. Samples of slag were collected
from the slag piles on the western side of the site and from the
railroad cars which are located throughout the site. Samples of
baghouse dust were collected from a large pile. These samples
were analyzed for EP toxicity, petroleum hydrocarbons and
priority pollutants.
The third priority was to determine the areal extent of con-
tamination on-site. A sampling grid system, 200 ft x 200 ft, was
collected from each grid sector to a depth of 6 in. These samples
were analyzed for EP toxicity and petroleum hydrocarbon
analysis; in addition, selected samples were analyzed for priority
pollutants.
In addition to the surface soil sampling, subsurface sampling
was performed at 45 locations throughout the site; 17 of these
sample locations were monitoring wells. As a result of the investi-
gation activities described above, many contamination sources
were discovered. The following section describes the major
sources encountered.

RESULTS OF SURVEY

Drums
A total of 2,621 abandoned drums was found above grade at
the Roebling Steel site. Approximately 15% of the drums were
located inside buildings.
Many of the drums were badly corroded and were leaking.
Many drums were half-buried in the ground in unknown condi-
tion. Other drums were stacked two and three high. The types of
materials identified in the drums included debris/trash, slag/
dust, chemical powders, oils, grease and chemical liquids.
The contents of the subsurface drums have not been identified
yet. However, the geophysical survey identified two areas where
unknown buried objects may be located. The first area includes a
cluster of targets in an area approximately 100 ft by 75 ft, and
other individual targets were scattered around the wastewater
treatment buildings and lagoons. The second area of potential
drum burial is the large open space southeast of the M.A.
Polymer Building where several individual objects were located. It
is suspected that most of these targets are buried drums or tanks.

Tanks
A total of 106 abandoned tanks was found at the Roebling Steel
site. Approximately 65% of the tanks were located inside
buildings. The contents of the tanks include oils, water and
chemical powders.
Many of the tanks were in poor condition with rusted walls,
leaky valves and open roofs. There is one relatively new fiberglas
tank which, according to its labeling, contains approximately
2,500 gal of phosphoric acid. A few underground tanks were
located, but their condition could not be determined. Other
underground tanks may exist, but none are visible at the ground
surface.

Transformers
A total of 222 transformers was found at the Roebling Steel
site. Approximately 65"% of the transformers were located inside
buildings. The size of the transformers varied with oil storage
capacities from 34 to 1950 gal.
Pits/Sumps
Eighteen pits/sumps were found inside the various buildings at
the Roebling site. The pits/sumps typically were associated with
the steel-making and forming activities and are constructed of
reinforced concrete. All the pits/sumps contained liquids, most of
which appeared to be aqueous or oily in nature. The amount of
any residues/sludges in the bottom of these pits/sumps is not
known. The pits/sumps ranged in size from large quench pits (ap-
proximately 25 ft wide x 30 ft long x 15 ft deep) to small
drainage collection sumps (approximately 2 ft wide x 2 ft long x
2 ft deep).

Railroad Cars
Fifty-two abandoned railroad cars were found at various out-
door locations across the site. The railroad cars were all open
freight cars containing a variety of solid materials including slag,
fly ash, debris/trash and dry sludge.

Baghouse Dust
Baghouse dust appears to be scattered throughout the site.
There are dumpsters full of the dust in the baghouse area, and the
ground surface is covered with at least 2 in. in the area of these
dumpsters. A large pile of baghouse dust exists beneath an awn-
ing behind a building, and the entire surface area between that
building and the landfill is covered with the dust. In addition,
there are random piles throughout the landfill. The potential
hazard associated with this dust is increased because of its mobili-
ty; it is easily transported by wind or other surface disturbance.
This dust is a listed hazardous waste: KO61 baghouse dust from
electric furnace steel-making.

Landfill
The landfill has a large variety of debris on the surface, in-
cluding baghouse dust (as described above), drums, tires, shred-
ded rubber, shredded plastic and building nibble. Drums prob-
ably are buried in the landfill. However, it is not known whether
they are full, empty or contain hazardous material. An internal
memorandum found at the site written by a former tenant gives
evidence that large quantities of waste oil were dumped into the
landfill.

Laboratories
Four chemical laboratories/storerooms were found in relatively
unsecured areas of the site. The chemicals found included acids,
bases, inorganic salts, alcohols and halogenated and non-halo-
genated organics. One bottle of picric acid was removed from the
site.

Miscellaneous Items

• Plating Baths—Two plating baths located in one building still
contain the plating solution which is believed to be copper
sulfate.
• Asbestos—Much of the piping throughout the site is, or was,
covered with insulation which appears to be asbestos; this in-
sulation has fallen or is falling off. Therefore, there is a possi-
bility that asbestos fibers are in the air and floor dust.
• Tires—Large piles estimated to be 50,000 to 75,000 tires were
found on-site. These tires present a potential fire hazard and
provide a mosquito breeding area (in retained and pooled
water).
• The building interiors, including floors, walls and machinery,
may be coated with any or all of the contaminants found thus
far (e.g., PCBs and asbestos) due to the wide-scale use of these
substances during plant operations and subsequent spreading
since the various operations closed.
376 MINING & INDUSTRIAL WASTES
-------
Surface Contamination
Surface soil samples were taken at a depth of 0 to 6 in. through-
out the site and were analyzed for EP toxicity and petroleum hy-
drocarbons content. These analyses revealed the following:
t 26% of the samples exceed EP toxicity limits; 1% of the
samples exceed allowable limits of 5.0 mg/1 for chromium;
2% of the samples exceed allowable limits of 1.0 mg/1 for cad-
mium; and 24% of the samples exceed allowable limits of 5
mg/1 for lead.
• Petroleum hydrocarbons were present in every sample taken.
76% exceed the NJDEP allowable limit of 100 /ig/g, and the
remaining 24% ranged from 23 /tg/g to 95 /tg/g.
• Selected stained areas of soil were sampled for priority pollu-
tant analysis; some samples contained high concentrations of
organic contaminants, particularly semi-volatile compounds
including phthalates, pyrene, chrysene, anthracene and fluor-
anthene.

ONGOING WORK
As a result of these preliminary findings, a sampling and
analysis program has been developed to complete the evaluation
of the full extent of abandoned waste and potential environmen-
tal contamination on-site. Key elements of this program include:
• Sampling wells to determine if groundwater is contaminated
• Sampling soil from nearby playgrounds to determine whether
contamination has spread by surface sediment or airborne
paniculate transport to these critical receptor areas
• Sampling building interiors to characterize their contents,
which may eliminate them from further evaluation
• Excavating and sampling landfill areas and suspected burial
areas to characterize their contents
• Sampling the contents of drums, tanks and transformers to
identify items for potential removal action
• Further sampling of subsurface soils to determine depth of
contamination identified during Phase I surface sampling

CONCLUSIONS
The RI/FS is ongoing at the site. However, the following con-
clusions already have been made:
• The potential contamination sources identified at the Roebling
Steel site are similar to those that would be anticipated at any
large industrial site (i.e., abandoned laboratories, numerous
tanks, drums and transformers and a waste water treatment
plant with associated contaminated sludges).
• The RI/FS at a large industrial facility should be conducted in
several phases. The phased approach allows initial screening
and identification of potential contaminant sources with subse-
quent detailed characterization.
• The characterization and subsequent remediation of aban-
doned industrial facilities under Superfund is a new area that
is only in its formation stages.
DISCLAIMER
The above information has not undergone formal peer review
by the U.S. EPA and therefore does not represent a technical or
policy position on behalf of the agency. The information included
herein is restricted to use only by its authors.
MINING & INDUSTRIAL WASTES 377
-------
Ocean Disposal Risk Assessment Model

Joseph G. Karam
ICF Incorporated
Washington, D.C.
Martha J. Otto
U.S. Environmental Protection Agency
Office of Solid Waste
Washington, D.C.
ABSTRACT
A risk assessment model was developed to estimate human
health risks and environmental risks due to ocean dumping of
hazardous waste at the 106-mi dump site or due to ocean incinera-
tion of hazardous waste at the North-Atlantic proposed burn site.
For a given ocean disposal practice, the model specifies the ex-
pected steady-state releases of chemicals to the ocean surface. The
physical transport submodel then calculates areas of the water
column and ocean floor associated with various chemical concen-
tration ranges in the water column and sediments, respectively,
taking into account two characteristics of the chemical released:
(1) the adsorbed-dissolved partitioning coefficient and (2) the
decay rate in ocean waters.
Using fish catch and consumption data and chemical-specific
dose-response curves, the model transforms the results of the
physical transport model into measures of human health risk due
to ingestion of contaminated seafood. Using chemical-specific
ecosystem damage functions, the model transforms the results of
the physical transport model into measures of damage to the
water column ecosystems and sediments ecosystems.
The model was applied to a number of hazardous waste
streams. Preliminary results suggest that human health risks, if
any, are primarily due to ingestion of benthic organisms (i.e., bot-
tom dwellers) and that environmental damage, if any, is suffered
by the sediments ecosystems. The model appears to be an ade-
quate tool for preliminary screening of hazardous wastes for ac-
ceptability for ocean dumping and/or ocean-based incineration
on the basis of expected risks to human health and the environ-
ment.

INTRODUCTION
This paper presents a model for estimating human health risks
and environmental risks due to ocean dumping or ocean incinera-
tion of hazardous waste. It is motivated by the ongoing debate
within the U.S. EPA, the regulated community and the public in
general on whether and how to allow the ocean dumping or ocean
incineration of certain hazardous wastes. This model is described
in more detail in a U.S. EPA report which also presents a land
disposal risk assessment model based on similar or comparable
assumptions.'
From a risk assessment perspective, the distinction between
ocean dumping and ocean-based incineration lies mainly in the
amounts and locations of releases to the marine environment. For
ocean dumping, the model considers only releases resulting from
deliberate open-ocean dumping of hazardous wastes at the 106-mi
dump site. For ocean-based incineration, the model considers only
certain stack emission releases occurring at the proposed North-
Atlantic burn site and accidental or fugitive releases occurring at
the proposed burn site or in the coastal zone.
While ocean dumping releases enter the marine environment
directly at the dump site, stack emission releases are transported
in air over some distance before they settle down to the ocean sur-
face and accidental and fugitive releases can occur anywhere in
the open ocean. The distances over which chemicals travel in air
before they settle to the ocean surface (less than 30 mi) are small
relative to distances travelled in ocean waters (several hundred
miles). Furthermore, because the North-Atlantic burn site is adja-
cent to the 106-mi dump site, one may use the physical transport
model originally developed for the 106-mi site to simulate
transport of chemicals released from ocean-based incineration.
The model estimates the area! distribution of ambient chemical
concentrations in the water column and in the sediments follow-
ing a constant release of a chemical during 100 yr at the 106-mi
site. Using fish catch and consumption data and chemical-specific
dose-response curves, the model transforms the results of the
physical transport model into two measures of human health risk.
Using chemical-specific ecosystem damage functions, the model
transforms the results of the physical transport model into two
measures of environmental risks for water column ecosystems and
two measures of environmental risks for sediments ecosystems.
Results of applying the model to a number of selected hazar-
dous waste streams are presented in the last section of this paper.

OCEAN TRANSPORT MODEL
For a given continuous rate of release at the 106-mi site, the
physical transport model calculates areas of the water column and
ocean floor associated with various chemical concentration ranges
in the water column and in the sediments, respectively. The model
assumes background contaminant concentrations equal to 0. The
model takes into account two characteristics of the chemical
released: (1) the adsorbed-dissolved partitioning coefficient and
(2) the decay rate in ocean waters.
The physical transport model is based on generic results obtained
by running ASA's ocean pollutant transport model.2-3 This sec-
tion briefly reviews the basic approach and assumptions of the
ASA model, presents the generic results of running the ASA
model and the parameter values used to obtain these results, and
explains how the ocean disposal model uses these generic results
to calculate, for any release scenario, areas of the water column
and ocean floor associated with various chemical concentration
ranges.

The ASA Ocean Pollutant Transport Model
ASA's transport model computes the movement of pollutants
in the water column and in the sediments. Empirical or modeled
measures of mean seasonal currents provide the basis for trans-
378
HEALTH & ASSESSMENT
-------
port estimates. Since mean surface and bottom currents may be
quite different, a two-layered representation is used. Transport
calculations are performed on a spherical coordinate grid system.
The grid system breaks up the physical domain of the model into
discrete regions or grid cells. The size of the grid cells used defines
the limit of the model to represent small-scale processes. Grid cell
size varies, according to the geography of the region simulated,
from about 100 km2 to approximately 2,500 km2. Once the model
grid has been established relative to land and open ocean boun-
daries and bottom topography, a mean current velocity vector is
determined for each grid cell.

Water Column Transport Model
Most of the man-made pollutants of concern in the marine en-
vironment, including hydrocarbons, heavy metals, radionuclides,
chlorinated hydrocarbons and other organic compounds, are
readily adsorbed onto suspended particulate matter and may
eventually settle to the sea floor. Therefore, the model takes into
account adsorption, desorption and settling dynamics.
The model uses a mathematical approximation method special-
ly developed for use in ocean modelling to describe the movement
of wastes in the water column. A waste disposal scenario at a
specific location is simulated by releasing one or more particles at
discrete time intervals, with each particle representing a specific
mass of pollutant.
Total pollutants are distributed between adsorbed and dis-
solved phases assuming linear partitioning. The fraction of pollu-
tant in dissolved form, f, is a function of the chemical's adsorbed-
dissolved partitioning coefficient, Kp (dimensionless), and the
concentration of suspended solids in ocean water, Css (dimen-
sionless):
f= 1/(KPCSS+
(1)
The particulate adsorption and settling processes as well as the
diffusion processes in the model result in a long-term flux of
pollutants from the water column to the ocean- floor. Pollutants
which adhere to particulates with nonzero settling velocities de-
scend at the associated settling velocity, modified by a random
diffusion term. Dissolved-phase pollutants as well as pollutants
that adhere to nonsettling particles are associated with neutrally
buoyant particles. Movement of these particles results primarily
from horizontal advection by the current motions with random
diffusion in both the horizontal and vertical planes.

Sediment Model
In the water column model, each pollutant particle eventually
settles to the ocean floor, is transported outside the model domain
or decays to background levels. Through resuspension, particles
which settle to the ocean floor in depths of 100 m or less may be
reintroduced to the water column. Resuspension results from
wave and storm-generated turbulence which is governed by wave
height and period frequency distributions. The frequency with
which sediments are resuspended at various depths is based on ex-
isting data.
The governing differential equation for vertical pollutant
transport within the sediments, inclusive of a first-order decay
term, is:
5C
5t
= D
£C_
8Z2
- kC
(2)
in which C is pollutant concentration, D is a dispersion coefficient
which includes bioturbation effects, t is time, z is distance (posi-
tive downward) into the sediments and k is the pollutant decay
rate.
Generic Results and Parameter Specifications
ASA's ocean transport model was run to stimulate the fate and
transport of persistent chemicals dumped at the 106-mi site at a
rate of 1 kg/day for a period of 100 yr. Due to the nature of the
model, steady-state chemical concentrations are observed pre-
cluding any assessment of transient or short-term effects. A rate
of 1 kg/day was used as a basis arbitrarily to be scaled up propor-
tionately for actual calculated rates of release. Table 1 lists the
ocean parameters and their values as specified for the ocean trans-
port model runs.

Table 1
Ocean Parameters and Their Values

Ocean Parameter Value
Disposal Location: the 106-mile site

Transport Simulation Period

Vertical Dispersion Coefficients

— upper water column
-- lower water column

Horizontal Dispersion Coefficient

Particle Setting Velocity
Suspended Solids Fraction, C
38.8 deg. N Lat.
72.25 deg. W. Long.

100 years
100 cm'/sec
10 cm'/sec

400 m1 /sec
10%
20%
50%
20X
0.002 m/sec
0.001 m/sec
0.0001 m/sec
0.00001 a/sec

"6
0.3 * 10" mg/mg

(i.e. , 0.3 ppm)
Since the generic runs assume no chemical decay in ocean
waters, the transport model results depend only on the product of
the adsorbed-dissolved partitioning coefficient of the chemical,
Kp, multiplied by the fraction of suspended solids in ocean water,
Css. In all cases (i.e., for all possible values of KpCss), chemical
concentrations in the water column range from 10-19 ppm to
10-8 mg/1 (12 concentration ranges), depending on the
geographical location. Chemical concentrations in the sediments,
averaged over the top 10 cm, range from 10-14 to 10- 3 mg/1 (also
12 concentration ranges), depending on the geographical loca-
tion.
Table 2(a)
Areas Associated with Various Concentration Ranges
for Various KpCss Values:
Continental Shelf Water Column
Concentration
Range (ppm)
io"19
10"18
ID'17
io"16
ID'15
ID'14
ID'13
ID'12
ID'11
10'10
10-'
lO'8
£ C £
£ C £
£ C £
£ C £
£ C £
£ C £
£ C £
£ C £
£ C £
£ C £
£ C £
£ C £
10'18
ID'17
ID'16
ID'15
ID'1*
io-13
ID'12
lO'11
ID'10
ID'9
io"8
ID'7
K C dimensionless
p ss
io-3
0
0
0
1
.00
.34
.63
.50
0.87
1
2
2
2.
1.
0.
0.
.62
.20
.09
.74
.99
.29
.00
ID'2
0
0
0
1
1
1
.00
.29
.81
.68
.28
.87
2.40
1
1.
1.
0.
0,
.87
.41
.40
.23
.00
lO'1
0.00
0.47
0.64
1.16
1.00
1.58
1.64
1.29
1.35
0.35
0.00
0.00
10°
0.00
0.06
0.24
0.35
0.53
0.53
0.47
0.42
0.30
0.00
0.00
0.00
io1
0.00
0.18
0.30
0.41
0.42
0.41
0.47
0.48
0.24
0.00
0.00
0.00
IO2
0.00
0.00
0.12
0.24
0.47
0.18
0.53
0.30
0.24
0.00
0.00
0.00
IO3
0.00
0.18
0.30
0.47
0.36
0.41
0.47
0.48
0.24
0.00
0.00
0.00
(areas in units of 10 m2)
Generic results, i.e.:
dumping rate = 1 kg/day
decay rate = 0 (no decay)
HEALTH & ASSESSMENT 379
-------
Table 2(b)
Areas Associated with Various Concentration Ranges
for Various K-C^ Values:
Deep Ocean Water Column
Table 2(d)
Areas Associated with Various Concentration Range*
for Various KpCH Values:
Deep Ocean Sediments

Concentration
Rang* (ppoj
- 1 0 -\H
10 " S C S 10 '"
- Ifi - 1 7
10 18 S C S 10 "
- 1 7 - 1 ft
10 17 S C S 10 "
io'16 s c s io"15
io'15 s c s io"1*
ID'14 S C S ID'13
-13 -12
10 i3 s c s 10 1Z
io"12 s c s io"11
-11 - 10
10 " S C S 10 I0
-10 -9
10 '" S C S 10 '
io"9 s c < io"8
io'8 s c s io'7

ID'3

0
0
0
1

1
2

8
7
0

.00

.12

.24
.90
.67
.15

.45
.17

.77

.91
. 11
.54

ID'2

0.00

0.12

0.30
0.84
0.24
0.83

1.20
1.74

3.88

6.17
7.41
0.54
(areas in
Generic results, i e.:
dumping rate 1 kg/'

day

K C
p 9S
ID'1

0.00

0.12

0.25
0.72
0.37
0.72

1.02
1.86

3.42

7.45
3.36
0.24
units of

dlmensionless
10°

0.00

0.18

0.24
u.67
0.60
0.79

1.03
1.03

1.09

1.44
1.20
0.00
IO10,

10'

0
0
0
1

0
0

1
0
0
.')

.00

.06

.18
.67
.61
.09

.85
.91

.27

.32
.90
.00

IO2

0
0
0
0

1
0

1
0
0

.00

.12
.24
.97
.30

.15
.66

.97

.68
54
.00

io3

0
0
0
0

1
0

1
0
0

.00

.12

.24
.79
.46
.97

.09
.97

.97

.50
.72
.00

decay rale = 0 (no decay)
Table
Areas Associated with
2(c)

Various Concentration

Ranges
for Various tifCK Values:
Continental Shelf Sediments

Range (pp»)
-14 -13
10 '* S C S 10 iJ
-13 -12
10 JJ S C S 10 12
- 12 - 11
10 " 1 C t 10 "
io'11 s c s io'10
- 10 -9
10 *U S C i 10 *
io'9 s c s io"8
io"8 s c s io"7
io"7 s c s 10"'
10'* i c s io's
-5 -4
10 S C S 10
-4 -3
10 S C S 10
-3 -2
10 S C S 10

io-3

0
0

0
1.
1
2.
4.

.22

.45

.69
.63

.87
04
28
35
58

ID'2

0.34

0.11

0.63
0.69

1.21
1.34
1.51
2.46
4.45

3.45

1.35

0.00
(ar««s in
K C
p «
ID'1

0.11

0.17

0.35
0.69

1.16
1.24
1.48
2.89
3.95

2.54

0.00

0.00
units of
dlnensionless

10°

0.12

0.18

0.23
0.29

0.35
0.60
0.90
1.09
1.28

0.55

0.06

0.00
io10.

io1

0
0

rj
0.
0.
1
0.

0.
,,,

.06

.00

.12
.35

.48
.54
.85
.15
85

IO2

0.
0

0.
0.
1.
0.
0.

.06

.00

.12
35

36
78
03
97
85

io3

0
0

0.
0.
1
1.
u.

.00

. 12
48

.29
72
03
.09
66

Generic results, i.e.:
dumping rate = I kg/day
decay rate = 0 (no decay)

For each of seven KpCss values (from 10-3 to 103, including
100 = 1), the generic results are expressed in terms of areas of the
water column or sediments, over the continental shelf or in deep
ocean, associated with various chemical concentration ranges.
Tables 2(a) to 2(d) indicate the area ( in 10'° m2) associated with
each of the 12 concentration ranges for each of the 7 KpCss pro-
duct values in the following four specific media:

• Continental shelf (depth ^200 m) water column
• Deep ocean (depthi200 m) water column
• Continental shelf sediments
• Deep ocean sediments

Use of the Generic Results
The generic results are valid for a unit dumping rate (1 kg/day),
cover only seven values of the KpCss produce (10'3, 10-2, . . ._
IO3) and do not take into account the chemical's eventual

ID'1*
lo'13
io"12
10""
.o-10
ID'9
10-"
io'7
!0-6
ID'5
io"4
10'3

~«. (PI
s c s
i c *
s c s
s c s
s c s
s c s
s c s
s c s
s c s
s c s
s c s
s c s

p->
io-13
io"12
ID'"
10-'°
ID'9
10-"
ID'7
ID'6
ID'5
ID''
ID'3
ID'2

10'3
0
1
1
1
2.
2.
2.
3.
4.
2.
0.
0.
.18
.14
.62
81
33
45
86
69
29
98
59
00

ID'2
0.24
0.60
1.69
1 68
2.40
2.33
2.75
3.40
5.37
3.09
0.29
0.00
(«!*••« In

K C
P ••
10-'
0
0
1
1
1
2
2
2
4.
4.
0.
0.
.61
.65
.59
.39
.76
.12
.24
.41
.33
69
06
00
units of
dl»«nsionlcss
10°
0.25
0.43
1.11
0.81
1.49
1.42
1.48
1.85
2.33
3.55
2.33
0.00
io10.
!0l
0.31
0.62
1.01
1.31
1.24
1.69
1.99
1.99
2.36
3.65
4.14
1.84
,,,
IO2
0.06
0.56
0.94
1.38
1.37
1.62
1.68
2.05
2.29
3.28
4.20
2.03

io3
0.06
0.81
0.76
l.SO
1.25
1.62
1.80
1.99
2.36
J.15
4.26
2.03

Generic result*, i.c
dumping rate I kg/day
decay rate - 0 (no decay)
degradation (decay rate = 0).
The ocean dumping model uses the generic results to calculate
areas associated with ranges of ambient chemical concentrations
resulting from dumping at a rate, R (in kg/day) a chemical char-
acterized by an adsorbed-dissolved partitioning coefficient, Kp,
and a decay rate in ocean waters, •> (in day-').
The model first multiplies the generic ambient concentrations
by the actual dumping rate, R. To account for the chemical's
decay in ocean waters, the model assumes that the average
residence time of a chemical particle in the water column is equal
to one year and that the chemical stops decaying once it has
reached the ocean floor.' The model thus multiplies the generic
ambient concentrations by a second factor representing the first-
order exponential decay of the chemical during 1 yi,
exp( - 365 y). The model finally calculates areas associated with
the new concentration ranges by extrapolating linearly with re-
spect to the logarithm of the KpCss product (Fig. 1).
ATM Ms
with • gtvmn
concentration
rang*
-3 -2 -1
* This example corresponds to the concentration range 10 ° ppm lo 10 " ^ ppm for the continen-
tal shelf sediments (generic case).
Figure 1
Linear Extrapolation Over Log (KpC^) to Calculate Area
Associated with a Given Concentration Range*
380 HEALTH & ASSESSMENT
-------
HUMAN HEALTH RISKS
The model calculates two measures of human health risk due to
ingestion of contaminated fish:

• Human population risk or expected number of cases
f Risk to the most exposed individual or MEI risk
This section presents the approach for estimating dose levels and
converting them into measures of individual risk (dose-response
curves), explains how the model specifies population exposure to
contaminated fish, calculated the expected number of cases,
defines the individual most exposed to fish contamination and
calculates the MEI risk.

Dose and Individual Risk
The concentration of contaminant in fish (Cfjsh, in ppm) is the
product of contaminant concentration in the aquatic environment
where the fish live (Cenv, in ppm) multiplied by a chemical-
specific bioconcentration factor (BCF, dimensionless):
= CenvBCF
(3)
The model assumes that the water column is the natural aquatic
environment for pelagic and demersal fish while the sediments are
the natural environment for benthic organisms. The ambient
chemical concentration, Cenv, in Equation 3 is thus equal to the
ambient chemical concentration in the water column for pelagic
and demersal fish and to the ambient chemical concentration in
the sediments for benthic organisms.
The model assumes an effective average individual consump-
tion figure of 14.3 g of seafood per day, as provided by the Na-
tional Marine Fisheries Service (NMFS).4 The model does not use
the standard U.S. EPA fish consumption value of 6.5 g/day
because that value was derived for freshwater and estuarine
species, not seafood.
Assuming that the average individual weighs 65 kg, the model
calculates the dose of contaminant ingested daily as:
d = (14.3 / 65,000) Cfish
(4)
where
d is does expressed in mg of contaminant per kg of body weight
per day
Cfjsh is the contaminant concentration in fish, expressed in mg
of contaminant per kg of fish tissue
Risks to human health are a function not only of exposure (or
dose levels), but also of the toxicity of the chemicals to which
humans are exposed. The model calculates the lifetime risk to an
individual from ingesting a given chemical at a daily dose d using
the lexicological approach described by the U.S. EPA.1

Human Population Risk
To calculate the expected number of cases, the model specifies
human consumption of contaminated seafood. To estimate the
number of persons eating fish contaminated by a given consti-
tuent and the average daily dose to which they are exposed, the
model calculates, for each waste stream constituent separately,
the total annual catch of contaminated seafood and the average
chemical concentration in contaminated fish caught.

Annual Contaminated Catch
and Average Chemical Concentration
The total annual catch of contaminated seafood is the sum of
amounts of contaminated pelagic (including demersal) and ben-
'hic fish caught annually over the continental shelf and in deep
water. The amount of contaminated fish caught in a specific
medium (water column or sediments, over continental shelf or in
deep water) is the sum, over all concentration ranges found in that
medium, of amounts of fish caught in the areas associated with
these concentration ranges. The model calculates the annual fish
catch in a given area of the specific medium as an area-prorated
fraction of the total annual catch in the largest area of that
medium that theoretically could be contaminated (the total an-
nual catch in a given area is based on 1983 NMFS catch data).1
The average chemical concentration in contaminated fish
caught is derived knowing the annual catch of seafood in each
specific medium (4 specific media) and for each concentration
range (12 concentration ranges for each specific medium). The
average chemical concentration over all contaminated catch is the
average over all specific medium-concentration range pair (48
pairs total) of the corresponding concentration (middle point of
the concentration range) weighted by the associated seafood
catch.

Number of Persons Exposed and Average Dose
The number of persons exposed to a given chemical through in-
gestion of contaminated seafood is equal to the total annual catch
of contaminated seafood divided by the average annual individual
consumption of seafood (14.3 g/day or 5.22 kg/year).
The average daily contaminant dose to an exposed individual is
proportional to the average concentration of the chemical in con-
taminated seafood caught and eaten (Eq. 4).

Average Individual Risk and
Expected Number of Cases
The model calculates the average individual risk due to a given
chemical by applying the dose-response curves described in Ap-
pendix D of the U.S. EPA report to the average dose calculated
above.' The expected number of cases due to one constituent is
equal to the average individual risk multiplied by the number of
persons exposed to that chemical.
The total expected number of cases (i.e., for the whole waste
stream) is the sum of expected number of cases due to each waste
stream constituent. The overall average individual risk (i.e., for
the whole waste stream) is the ratio of the total expected number
of cases divided by the maximum size of populations exposed to
each chemical.

Risk to the Most Exposed Individual
The individual most exposed to contamination through fish in-
gestion is a person who eats the average daily dose of 14.3 g of the
most contaminated seafood. The total MEI risk from ocean
dumping a given waste stream is the sum, capped at 1, of the MEI
risks due to each waste stream constituent.

ENVIRONMENTAL RISKS
The model estimates damage to two saltwater sub-ecosystems:
(1) water column ecosystem and (2) sediments ecosystem. For
each saltwater sub-ecosystem, the model calculates two measures
of damage due to chemicals released to the ocean.
To assess damage to the water column ecosystem, the model
calculates the:

• Weighted volume of water affected
• Damage to the most exposed water column ecosystem (water
column MEE risk)
To assess damage to the sediments ecosystem, the model
calculates the:
• Weighted area of ocean floor affected
• Damage to the most exposed sediments ecosystem (sediments
MEE risk)
HEALTH & ASSESSMENT 381
-------
This section reviews damage to saltwater sub-ecosystems as a
function of ambient chemical concentrations, explains how the
model calculates total damage to each sub-ecosystem (i.e., ecosys-
tem population risk) and the extent of damage to the most ex-
posed sub-ecosystem.'
Damage to Saltwater Ecosystems
The extent of damage, D, to a saltwater ecosystem (in the
ocean) due to a given chemical is a function of the ambient
chemical concentration, C, to which the ecosystem is exposed.
For the water column ecosystem, C is the ambient chemical con-
centration in the water column. For the sediments ecosystem, C is
the ambient chemical concentration in the sediments.
For a given chemical, the saltwater ecosystem damage function
is completely determined knowing the threshold. Or, and cata-
strophic, Cp, concentrations of the chemical in saltwater ecosys-
tems. For ambient chemical concentrations less than the threshold
concentration level Or, there is no damage to the ecosystem
(damage D = 0). Conversely, the ecosystem is completely dam-
aged (damage D = 1) when ambient chemical concentrations ex-
ceed the catastrophic concentration level, Cp. For ambient con-
centrations between Cj and Cp, damage is a linear function of the
logarithm of concentration. Estimated values of threshold and
catastrophic concentration levels in saltwater ecosystems are
documented in a U.S. EPA report.' We assume that these concen-
tration levels apply equally to the water column ecosystem and to
the sediments ecosystem. In other words, we assume the same
damage function for both sub-ecosystems.
Ecosystem Population Risks
The model calculates the weighted column of water affected
and the weighted area of sediments affected using the results of
the physical transport model, namely areas of the water column
and ocean floor associated with various concentration ranges.

Weighted Volume of Water Affected
The mass quantity of organisms (including pelagic and demer-
sal fish) living in a water column cylinder is equal to Mw (in mg
C):

Mw = ow Aw h (5)

where
Cw is the density of living organisms in the water column
(mg C/m3)
Aw is the cross-sectional area of the water column cylinder (m2)
h is the depth of the water column cylinder (m)

Of all organisms living in a water column cylinder associated
with a given concentration range (middle point Cw, in ppm) a
fraction, D(CW), is severely damaged due to ambient chemical
concentrations. Thus, a quantity of living organisms equal to
D(CW)M(AW) is severely affected in this water column cylinder,
where Aw is the area of the ocean associated with the concentra-
tion range with middle point, Cw. The total quantity of affected
organisms in the water column is therefore equal to the sum over
all concentration ranges (middle point Cw) of the term D(CW)M
(Aw):
expected mass quantity
of water column
organisms

affected (in mg C)
= £ ow h AW(CW) D(CW)
(6)
where
D(CW is the damage to the water column ecosystem for am-
bient concentration Cw (saltwater ecosystem damage func-
tion) and
AW(CW) is the area of water column associated with the con-
centration range with middle point C* (physical transport
model)

Asuming that the density of living organisms in the water col-
umn, cw, is uniform (i.e., does not depend on the location), ex-
pression (6) becomes:
expected mass quantity
water column organisms -
-------
sediments associated with the concentration range with middle
point Cs (AS(CS). The total quantity of affected organisms in the
sediments is therefore equal to the summation over all concentra-
tion ranges (middle point Cs) of the term D(CS)M(AS):

expected mass quantity of
sediments organisms affected = Z es AS(CS) D(Cs) (11)
(in mg C) Cs

where
D(CS) is the damage to the sediments ecosystem for ambient
concentration Cs (saltwater ecosystem damage function)
AS(CS) is the area of sediments associated with the concentra-
tration range with middle point Cs (physical transport
model)

Assuming that the density of living organisms on the ocean
floor, es, is uniform (i.e., does not depend on the location), Eq. 6
becomes:
expected mass quantity
sediments organisms
affected
(in mg C)
where
weighted area of
sediments affected
(m2)
= Cs
weighted area
• of sediments
affected
(in m2)
= £ A(CS)D(CS)
(12)
(13)
The weighted area of sediments affected (in m2) is a useful
measure of ecorisk since it gives, by simply multiplying it by the
density of living organisms along the ocean floor (es, in mg C/m2),
the expected mass quantity (in mg C) of organisms affected on the
ocean floor. For comparative risk analyses of ocean-based waste
management practices, the knowledge of es is not necessary
because the weighted areas of sediments affected (in seawater in
all cases) can be compared directly. This direct comparison is not,
however, possible whem comparing environmental risks from
ocean-based versus land-based technologies because: (1) the den-
sities of living organisms along seawater and freshwater floors are
not equal on average and (2) the comparative values of saltwater
and freshwater ecosystems are unknown.
A measure of average damage to the sediments ecosystem can
be derived from the expected mass quantity of organisms affected
(or, equivalently, from the weighted area of sediments affected);
it is equal to the average value of damage, D, over all points of the
ocean floor where D is strictly positive:
Average damage to
sediments ecosystem
(14)
The average damage measure is the equivalent of the average in-
dividual risk measure for human health risks. It takes values be-
tween 0 and 1.

Damage to the Most Exposed Ecosystem
For each waste stream constituent, the results of the physical
transport model indicate the highest ambient concentration in the
water column, Cw*, and in the sediments, Cs*. The extent of
damage to the most exposed water column ecosystem due to a
given constituent is equal to:

Damage to most exposed
= D(CW*) (14)
water column ecosystem
The extent of damage to the most exposed sediments ecosystem
due to a given constituent is equal to:
Damage to most exposed

sediments ecosystem
= D(CS»)
(15)
The damage to each most exposed sub-ecosystem is the sum,
not exceeding one, of damage due to each waste stream consti-
tuent.

SELECTED RISK RESULTS
The model was applied to a number of solvents, dioxins and
California list waste streams.' This paper presents risk results for
five selected waste streams (Table 3). Tables 4, 5 and 6 indicate
human health risks, risks to water column ecosystems and risks to
sediments ecosystems, respectively, due to ocean dumping or
ocean incineration of each of the five selected waste streams.

Table 3
Profiles of Selected Waste Streams
Wiitfl Stre»: OO.tiO.Ot
NAME: ORGANIC LIQUIDS
Conictcuentt or Concer
Wa* • Siraaa: 61. 02. IB
NAM : ACID WASTE FROM
Con tituanta of Concar
Wit • Strein: 62.01.21
HAH : AMMONIA STILL LI
Con tltuenti of Concer
Watt* 5tre»: 63.03.0U
EPA Hunbft
FROH DECREASING
n; TRICHLOROETHCNE
TCTRACHLOROETHENC
EPA Numb*
PLASTICS PRODUCTION
n: PHENOL
EPA Nunba
HE SLUDGE FROH COKING Of
n: CYANIDE
NAPHTHALENE
CFA Nunba
: F001
,1,1-TRICHLOROETHA
i 0002
ORHALOEHYOE
: K060
RATIONS
HENOL
LUORANTHENE
: P1Z3
QUANTITY: (MT/VR)
OICHLOROHETHANE
QUANTITY: (HT/VR)
QUANTITY: (HT/YR)
ARSENIC
309. UV7
1J56954.000
1026.461
Wilt* Stream: 61.02.09
NAME: ARSENIC--EP TOXIC
Conitituenti of Concern:
EPA Nuntaar: D004
INORGANIC FLUIDS
ARSENIC
QUANTITY: (HT/YR) 30BUU2.062
Format of Risk Results
For each line in Tables 4 to 6, the first three entries identify the
waste/technology combination for which risk results are in-
dicated. The first two entries identify each waste stream with a
number specific to this analysis and with the applicable U.S. EPA
code. The last entry identifies the technology applied to the waste:
OD for ocean dumping and OI for ocean-based incineration.

Human Health Risks
For each waste/technology combination (Table 4), two types of
human health risks are assessed:
• Human population risks
• Risk to the most exposed individual
There are five items under the human population risks heading:
• Average individual risk, i.e., expected number of cases
weighted by the severity factor, and divided by number of
people exposed
• Expected number of cases, weighted by the severity factor
• Medium of concern, i.e., air, groundwater or surface water
(including fish)
• Percentage of total risk attributed to the medium of concern
• Constituent of concern, i.e., the waste stream constituent
causing the highest risk summed over all media
The same items apply for risks to the most exposed individual
(MEI risks) except that: (1) the average individual risk is replaced
with the MEI risk and (2) the expected number of cases becomes
irrelevant.

Environmental Risks
For each waste/technology combination (Tables 5 and 6), two
types of environmental risks are assessed:
• Ecosystem population risks
• Damage to the most exposed ecosystem
There are five items under the ecosystem population risks

HEALTH & ASSESSMENT 383
-------
Table 4
Human Health Risk Results
Human Population
Waste
Stream
Number
00. U0.01
OO.U0.01
61.02. 1 a
61.02.18
62.01.21
62.01.21
63.03.0U
63.03.0U
6U.02.05
6U.02.05
EPA
Code
F001
F001
0002
D002
K060
K060
P123
P123
DOOU
DOOU
Tech
00
01
00
01
00
01
00
01
OD
01$
Average
Individual
Risk
0.26997E-06
O.OOOOOE+00
0. 00000 E>00
O.OOOOOE+00
0.20772E-OU
0. 11025E-OU
0.70812E+00
O.OOOOOE+00
0.33125E-03
0.33125E-03
Expected
Number
of Cases
O.UU021E+02
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
0.33870E+OU
0.65219E+03
0.115U6E+09
O.OOOOOE+00
0. 1959UE+05
0.1959UE+05
Hed 1 urn
of
Cone
CS/S
None
None
None
CS/S
OW/S
CS/S
None
ow/s
OW/S
Rl«k«
Pet
Hed
Cone
0.92U
0.000
0.000
0.000
0.92U
0.753
0.92U
0.000
0.753
0.753
ftlikt to Holt Expoted
Const 1 tuent
of Concern
TETRACHLOROCTHENE
None
None
None
FLUORANTHENE
ARSENIC
TOXAPHENE
None
ARSENIC
ARSENIC
MEI Risk
0.55316E-05
O.OOOOOE+00
O.U9I41I4E-03
O.OOOOOE+00
0.62938E-03
0.28579E-03
0. 10000E+01
O.OOOOOE+00
0.85523E-02
0.85523E-02
Hed lull
or
Concern
CS/S
None
CS/S
None
CS/S
CS/S
CS/S
None
CS/S
CS/S
Pet
Hed
Cone
1.000
0.000
1.000
0.000
1.000
1.000
1.000
0.000
1.000
1.000
Individual
Conitl tuent
of Concern
TETRACHLOROCTHENE
None
PHENOL
None
FLUORANTHCNE
ARSENIC
TOXAPHENE
None
ARSENIC
ARSENIC
Table 5
Environmental Risk Results, Water Column Ecosystems
Water Column Population Rliki
Waste
Stream
NuBbe r
OO.U0.01
00. MO. 01
61.02. IB
61.02. 18
62.01.21
62.01.21
63.03.0U
63.03.0U
6U.02.05
64.02.05
EPA
Code
F001
F001
0002
0002
K060
K060
P123
P123
000 U
DOOU
Tech
OD
01
00
01
00
01
00
01
00
01$
Average
Ecosystem
Damage
O.OOOOOE+00
O.OOOOOE+00
0.15573E-01
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
0.76012E-02
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
Weighted
Vo 1 ume
of Water
(1E10 «3)
O.OOOOOE+00
O.OOOOOE+00
O.B5U08E+03
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.U1688E+03
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
Hed 1 urn
or
Cone
None
None
DW/W
None
N/A
N/A
DW/W
None
None
None
Pet
Hed
Cone
0.000
0.000
0.999
0.000
0.000
0.000
1.000
0.000
0.000
0.000
Const 1 tuent
or Concern
None
None
PHENOL
None
Data Unava 1 lable
Data Unava liable
TOXAPHENE
None
None
None
Risks to
HEE Oaauge
O.OOOOOE+00
O.OOOOOE+00
O.U2760E+00
O.OOOOOE+OO
O.OOOOOE+OO
O.OOOOOE+00
0.38600E+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
Host Expo led
Medium Pet
or Hed
Concern Cone
None
None
DW/W
None
N/A
N/A
DW/W
None
None
None
0.000
0.000
1.000
0.000
0.000
0.000
1.000
0.000
0.000
0.000
Water Column
Const 1 tuent
of Concern
None
None
PHENOL
None
Data Unava i lable
Data Unava i lable
TOXAPHENE
None
None
None
Table 6
Environmental Risk Results, Sediments Ecosystems
Waste
Stream
Number
EPA
Code
Average
Ecosystem
Tech Damage
Sediments Population Rlski
Weighted
Area of Hed I urn Pet
Sediments or Hed
(1E10 n2) Cone Cone
Constituent
of Concern
Risks to Host Exposed Sediments
Hed I urn Pet
or Hed Constituent
HEE Damage Concern Cone of Concern
OO.U0.01
OO.U0.01
61.02.1B
61.02.18
62.01.21
62.01.21
63.03.0U
63.03.0U
6U.02.05
6U.02.05
F001
F001
0002
D002
K060
K060
P123
P123
DOOU
DOOU
OD
01
OD
01
00
01
00
01
O.OOOOOE+00
O.OOOOOE+00
00 0.7775UE+00
01 O.OOOOOE+00
0.23937E+00
O.OOOOOE+00
0.58279E+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00
O.OOOOOE+00

0.32175E+02
O.OOOOOE+00

0.63935E+01
O.OOOOOE+00

0.2U116E+02
O.OOOOOE+00

O.OOOOOE+00
O.OOOOOE+00
None
None
OW/S
N/A

CS/S
None

None
NONE
0.000
0.000
0.965
0.000
0.508
0.000
0.000
0.000
None
None
CS/S 0.505 PHENOL
None 0.000 None
CYANIDE
Data UnavaIlable
TOXAPHENE
None
None
NONE
O.OOOOOE+00 None 0.000 None
O.OOOOOE+00 None 0.000 None

0.10000E+01 CS/S 1.000 FORHALDEHYDE
O.OOOOOE+00 None 0.000 None

0.10000E+01 DW/S 1.000 CYANIDE
O.OOOOOE+00 N/A 0.000 Data unavailable

0.10000E+01 DW/S 1.000 TOXAPHENC
O.OOOOOE+00 None 0.000 None

O.OOOOOE+00 None 0.000 None
O.OOOOOE+00 None 0.000 None
heading:

• Average ecosystem damage, i.e., average damage over points
where damage occurs
• Weighted volume/area of water/sediments affected: although
both the weighted volume of water column affected (in
1010 m3) and the weighted area of sediments affected (in
10'° m2) are calculated, a direct comparison of the two
measures is not possible
• Medium of concern: the medium of concern is either contin-
ental shelf (CS/W) or deep water (DW/W) for water column
effects, and either continental shelf (CS/S) or deep water
(DW/S) for sediments effects
• Percentage of total risk attributed to the medium of concern
(by definition, it is equal to 1 if the risk is not equal to 0)
• Constituent of concern, i.e., the waste stream constituent
causing the highest risk summed over all media

Summary of Selected Risk Results
Based on the results presented in Tables 4 to 6, it appears that:

• Human health risks, if any, are primarily due to ingestion of
384 HEALTH & ASSESSMENT
-------
benthic organisms
• Environmental damage is mostly afflicted upon sediments
ecosystems
CONCLUSIONS
The model appears to be an adequate tool for preliminary
screening of hazardous wastes for acceptability for ocean dump-
ing or ocean-based incineration on the basis of expected risks to
human health and the environment. This model cannot, however,
qualify a waste stream for ocean dumping or ocean-based in-
cineration on the basis of risk results. Eventual ocean dumping or
ocean incineration of specific hazardous wastes must satisfy
various criteria developed in the ocean dumping and ocean in-
cineration regulations and will be permitted only on a case-by-
base basis.

REFERENCES
1. U.S. EPA, "Assessment of Impacts of Land Disposal Restrictions on
Ocean Dumping and Ocean Incineration of Solvents, Dioxins, and
California List Wastes," prepared by ICF Incorporated, Oct. 1986.
2. U.S. EPA, "Comparing the Risks and Costs of the Land and Ocean
Disposal of Hazardous Waste: A Preliminary Analysis," Draft Re-
port, prepared by ICF Incorporated and Applied Science Associates,
Inc., et at.. Mar. 1985.
3. U.S. EPA, "Land and Ocean Assessment Models Applied to Five
RCRA Hazardous Waste Streams," Draft Report, prepared by ICF
Incorporated and Applied Science Associates, Inc., Nov. 1985.
4. National Marine Fisheries Service, "National Seafood Consumption
Survey," Department of Commerce, Washington, DC, 1981.
HEALTH & ASSESSMENT 385
-------
Human Exposure Estimates Using U.S. EPA
Guideline Models: An Integrated Approach
George J. Schewe
Joseph Carvitti
Joseph Velten
PEI Associates, Inc.
Cincinnati, Ohio
ABSTRACT
A methodology was developed whereby dispersion modeling
using the U.S. EPA guideline models was combined with the pop-
ulation exposure and risk calculations of the Human Exposure
Model (HEM). Two dispersion models were used, one for com-
plex sources in relatively flat terrain and one for more complex
terrain. Each model was selected on the basis of previous use
frequency and compatibility with the HEM techniques.
The dispersion models are operated independently of the ex-
posure model and offer the complete range of options and source
characterization normally available. The special versions de-
scribed herein give output files that are wholly compatible with
the input requirements of the exposure model.
The use of this hybrid modeling technique allows a greater
attention to modeling detail than either model alone and allows
consideration of dispersion and exposure phenomena on an over-
all basis.

INTRODUCTION
For the last several years, the U.S. EPA has been performing
background assessments to assist in decisions regarding the need
for standards for control of potential toxic or hazardous air pollu-
tants. These assessments were deemed necessary by the U.S.
EPA due to the concern that toxic pollutants emitted by produc-
ers, users and transporters could pose risks to the public health.
Development of the impact assessment methodology has prev-
iously been hindered by the difficulty in defining the nature and
extent of the potentially toxic pollutant problem. The U.S. EPA's
goal remains, however, to reduce to the extent possible adverse
health effects resulting from community exposure to ambient
concentrations of airborne toxic pollutants.
The purpose of this paper is to present a set of techniques for
performing population and maximum individual exposure and
risk assessments due to toxic air pollutant emissions. The tech-
niques are based on currently available models and involve the
integrated use of the U.S. EPA's recommended dispersion mod-
els with population models or worst-case population assump-
tions.' This type of modeling approach is needed to assess the
potential adverse impacts from air emissions. Some of these emis-
sions have been determined to produce critical off-site population
exposures with a high potential for adverse effects on humans.
Effects of concern are based on maximum individual 24-hr
exposures and 70-yr (lifetime) population exposures. Calculated
air pollutant concentrations in ambient air for known or candi-
date pollutants are linked to population effects (chronic or acute)
through unit cancer risk factors or modified Threshold Limit Val-
ues (TLVs) [modified to put the TLV time-weighted averages
(TWA) on a consistent basis with dispersion model calculated
concentration).
The focus of this paper is on the integrated use of long-term
dispersion models with population data to estimate lifetime risk
for a collective population exposure to potential carcinogens and
the individual risk for a critical population group.

ASSESSMENT METHODOLOGY
The assessment techniques have grown somewhat from the in-
itial stages of simplified screening modeling and have been inte-
grated with population data into a model called the Human Ex-
posure Model.1 This model is somewhat limited by the simplified
dispersion algorithm which uses collocated sources in an ap-
proach similar to the Climatological Dispersion Model.' Given
the purpose of the U.S. EPA's use of HEM, to assess risk over
a great many facilities, the model has performed adequately as a
screening tool and is valuable in relative ranking to assess facilities
of concern. To adequately characterize an individual facility, as
will be required prior to development of a control strategy, the
dispersion modeling approach in HEM may yield less than fully
acceptable results. The need to review the impacts of multiple
sources at a single facility in order to derive cost-effective control
strategies has required replacement of the HEM dispersion calcu-
lations with more refined modeling procedures. The population
routines of HEM were, therefore, combined with detailed disper-
sion calculations found in more sophisticated U.S. EPA models.

Scope of Toxk Emissions Problem
The primary air releases of concern in the developing U.S.
EPA program are the potentially toxic air pollutants associated
with manufacturing, industry and consumers in individual com-
munities. This concern does not limit the scope of the desired risk
assessment methodology to any specific industry type. Rather,
risks also may be associated with steam and power production
facilities, storage areas and small local operations such as dry
cleaners. A uniform risk assessment methodology should ad-
dress the complexity of individual sources as well as allowing con-
sistency in emission and source characterization, selection and
use of meteorological data, model approach, use of population
statistics and health risk factors (e.g., what concentration of con-
taminants breathed over what time period will cause one cancer
incidence over what time intervcl).
The current status of the Federal air toxics policy is that many
states are assuming responsibility for review of toxic air pollu-
tant emissions and related ambient air impacts. Initial state permit
review guidelines have stressed simplified modeling. Subsequent
drafts have stressed more detailed dispersion modeling and risk
assessment analyses. To arrive at acceptable and representative
risk assessment methodologies, existing techniques should be
386
HEALTH & ASSESSMENT
-------
assessed and identified for consistency with sources and toxic
emission characteristics.

Integrated Approach
Many available dispersion models recommended by the U.S.
EPA Guideline on Air Quality Models' are applicable for deter-
mining annual concentrations which can then be used for deter-
mining lifetime risk. The basic assumption is that the concen-
tration estimated at a given location due to a particular source,
group of sources, facility or group of facilities on an annual basis
is a good estimate of the concentration to which a population
would be exposed over a lifetime. Thus, a population centroid
(representative of an area defined by census tracts and other cen-
sus subdivisions) is related to a concentration estimate represen-
tative of the same location. Assuming an average lifetime breath-
ing rate and relating the likelihood of cancer incidence to a certain
ambient concentration (unit risk), the collective population ex-
posure can be estimated. Table 1 describes the basic steps in this
approach, including detail related to the actual combining of
model-required inputs and outputs.
Table 1
Integrated Risk Modeling Steps

Step
No. Description
1. Review-source emission characteristics for facility of interest.
2. Perform source screening analysis of individual sources to deter-
mine critical downwind distances.
3. Determine applicability in study area for rural and urban treatment,
flat or complex terrain, paniculate or gaseous effluent, property
boundaries, downwash calculations, etc.
4. Select most appropriate dispersion model(s) for analysis.
5. Prepare joint frequency distributions (STAR) of windspeed, wind
direction and atmospheric stability class using on-site data or near-
by National Weather Service data.
6. Format dispersion model output concentrations into Human Ex-
posure Model (HEM) compatible input data.
7. Perform test runs on receptor grid in HEM to assess population
distributions; select appropriate unit risk factors.
8. Unify computer control commands to coordinate combined disper-
sion/risk modeling.
9. Input all source and emission characteristics.
10. Perform risk modeling for each complete year of meteorological
data.
11. Summarize 70-year risks and annual repeat intervals for substance
of interest.
Model Selection
The complexity of individual sources or facilities necessitates
the selection of detailed models for toxic pollutant risk assess-
ment studies. These models are more consistent with the Guide-
lines on Air Quality Modeling,1 U.S. EPA Regional modeling
guidance4 and PSD modeling procedures than simple screening
techniques. The models also offer more flexibility in treating the
wide variety of emission sources encountered in manufacturing,
industry and research facilities.
Models which correspond to the long-term averaging times
used to assess lifetime risk generally fall into two categories. The
first category includes those models requiring hourly meteoro-
logical data which then calculate a long-term average concen-
tration over all hours. The second category includes those mod-
els which calculate ambient concentrations for specific meteoro-
logical categories and then weigh each concentration by the an-
nual frequency of occurrence for that condition. This results in
weighted ambient concentration estimates at each receptor point
selected for calculation. Because the joint meteorological-fre-
quency-weighted models consider wind-sector-averaged concen-
trations (rather than point estimates in the hourly models), the
estimates are more applicable to a disperse population. The cal-
culated concentration then can be combined easily and accurate-
ly with the U.S. Census Bureau data reported as a collection of
population centroids. These models also require fewer resources
to operate than the hourly models.
Several U.S. EPA models already have been tested in this inte-
grated scheme. Some of the more applicable ones include:
• Industrial Source Complex Long-Term Model5—A flat terrain
model with steady-state Gaussian transport and dispersion for
continuous sources; allows spatially varying sources which
can be characterized as point, area or volume emission re-
leases; treats aerodynamic downwash; assumes concentrations
are dispersed throughout a wind sector due to wind variabil-
ity; uses available or derived point frequency distributions of
windspeed, wind direction and atmospheric stability class.
• LONGZ Model'—A steady-state Gaussian transport and dis-
persion model that can be used in areas with complex terrain;
allows treatment of point and area sources; assumes sector-
averaged concentrations; uses available or derived joint fre-
quency distributions of meteorological data.
• Other Models—A number of other models may be appropriate
for integration into this model/risk methodology. These in-
clude COMPLEX I, Air Quality Display Model (ADQM),
Climatological Dispersion Model (CDM) and Valley Model.
These models have some shortcomings including the use of
hourly meteorological observations, lack of treatment of de-
sired variables and inappropriateness due to undesired applic-
ability.
Table 2
Options Used in the Modeling of Annual Average Concentration
Patterns for Risk Assessment
Models

ISCLT,
LONGZ
ISCLT,
LONGZ
ISCLT,
LONGZ
ISCLT
ISCLT

ISCLT
ISCLT
ISCLT,
LONGZ
ISCLT,
LONGZ
ISCLT,
LONGZ
ISCLT
ISCLT,
LONGZ
ISCLT,
LONGZ
Description of the Option

Wind direction and speed, and Pasquill-Gifford (P-G) sta-
bility data are input in the form of the STAR data.
Annual morning and afternoon mixing heights are obtained
from the Holzworth (1972)' mixing height study.
Ambient air temperatures are selected from climatological
records and are distinguished by stability class.
Wind profile exponents are selected by stability class.
Vertical potential temperature gradients and entrainment
coefficients are ISCLT default values.
The rural or urban mode must be specified.
Wind system measurement heights may be set.
Terrain effects may be included.

Polar coordinate receptors used.

Stack, area or volume sources may be modeled.

Aerodynamic building downwash analysis.
Time-dependent exponential decay of pollutants may be
used.
Annaul concentrations are calculated and output for any
sources or combination of sources.
HEALTH & ASSESSMENT 387
-------
Important options used in the ISCLT and LONGZ models
are noted in Table 2. These options allow consideration of source
detail and the source-receptor interactions in the atmosphere
without sacrificing consideration of important factors which
more simplified techniques may require. Use of these models with
which regulatory agencies are already familiar allows a modeling
methodology consistent with ongoing modeling guidance.

Population Risk Analysis
The estimated ambient impacts are combined with the popula-
tion data in HEM to determine the total lifetime (70 yr) popula-
tion exposure within 50 km of a facility. For the calculation of
the maximum exposed individual, risk calculations are per-
formed using discrete receptors placed where individuals may be
exposed to the maximum concentrations estimated near a facility.
For other receptors, the model results are input to HEM. This
program provides output for each wind direction and gives an
average ambient impact at each ring of receptors (downwind
distance), the number of population centroids intersected and
the total number of persons included in each centroid.
A cumulative exposure table (Table 3) is prepared that pre-
sents the total number of persons exposed to selected concentra-
tion levels and the maximum and minimum concentration to
which any person is exposed. For example, the maximum con-
centration to which any person is actually exposed is given in
Table 3 as 1.01E-01 ug/m'. This maximum falls within the l.OOE-
01 concentration level (i.e., is greater than l.OOE-01 ug/m' and
less than 5.36 >ig/ms). Thirteen people are exposed to a concen-
tration equal to or greater than the lower bounds of this concen-
tration level (i.e. l.OOE-01). The next concentration level (5.00E-
02) shows an exposed population of 53, including the 13 people
given for the previous concentration level, and so on. The lowest
concentration level given in the table provides the cumulative
population exposure within the 50-km radius of the facility.
• Number of sources
• Unit risk estimate
• Distances from the plant centroid to each ring
• Latitude and longitude of the plant centroid
Individual sources or source groups (e.g., combined by com-
mon process, building, etc.) may be assembled and input to
HEM. The output from HEM then can be used to ascertain the
risk due to a variety of individual sources and sourge groups.
Risk factors can be input to the model to estimate the risk to
the surrounding population. Several sets of factors are available
for use with air dispersion model results, depending on the pollu-
tant of concern and time-averaging period of interest. Because the
main purpose of HEM is to generate a population/exposure
distribution around the modeled facility, a lifetime unit risk fac-
tor is most appropriate. This type of unit risk factor presents an
estimate of the excess probability of contracting cancer as a result
of constant exposure over a 70-yr period to an ambient concen-
tration of 1 ug/m'. This factor applied to HEM results thus esti-
mates the probability of cancer and the number of expected can-
cers in the population surrounding the facility.
The most readily available factors of this type are published
by the U.S. EPA's Carcinogen Assessment Group. Examples
are shown in Table 4. A factor should be applied for each ma-
terial expected to be emitted so that the full scope of effects can
be estimated. Other factors should be applied to the maximum
concentration estimates on an annual basis when estimating the
risk to the critical population or maximum exposed individuals.
These factors must reflect a short-term exposure over a year.
Sources of data can be found in a variety of places including
TLVs modified for annual exposures or the new U.S. EPA Chem-
ical Profiles. In applying unit risk factors, it is assumed that the
exposure is to 1 ug/m' and the average daily inhalation rate is
20 m', which produces a dose of contaminant of 20 ug/day. Ex-
posure is assumed to occur both indoors and outdoors at levels
predicted by the dispersion model.
T«ble3
Example Cumulative Exposure Table from HEM
Concentration
level,
ug/m'
5.36E+00
l.OOE-01
5.00E-02
2.50E-02
l.OOE-02
5.00E-03
2.50E-03
l.OOE-03
5.00E-04
2.50E-04
l.OOE-04
5.00E-05
2.50E-05
l.OOE-05
9.43E-06
Population
0
13
53
221
476
1,140
2,560
14,600
27,600
142,000
766,000
1,370.000
1,560,000
1,570,000
1,570.000
Exposure
0.00
1.33E+00
4.11E+00
9.31E+00
1.31E+01
1.80E+01
2.31E+01
4.20E+01
5.06E+01
8.69E+01
1.83E+02
2.27E+02
2.35E+02
2.35E+02
2.35E+02
Maximum concentration to which any person 1s actually exposed
1.01E-01
Minimum concentration to which any person 1s actually exposed
1.84E-05
Inputs to HEM include the concentrations predicted by the dis-
persion model and the following:
• Number of concentration rings desired (polar coordinates re-
quired)
T»bl«4
Example Unit Risk Factor1
Pollutant
Acrylonitrile
Arsenic
Chloroform
Ethylene
Methyl chloride
Perchloroethylene
Styrene
UnltRbk
6.8 xlO-5
4.3 xlO-3
I.OxlO-5
2.7xlO-«
1.4 xlO-7
1.7xlO-«
2.9x10-'
The latitude and longitude of the plant centroid (or other ref-
erence point used in the modeling) should be determined. HEM
uses these coordinates to extract site-specific population patterns
from the U.S. Census Bureau files using data at the Enumera-
tion District/Block Group (ED/BG) level. (When using HEM
with its own dispersion model components, the plant coordi-
nates also are used to identify the nearest set of meteorological
data.) All files accessed by HEM contain 1980 population statis-
tics. By combining the population data with the concentration
array, HEM calculates the cumulative population exposure with-
in a 50-km radius of the plant to produce specific exposure/dos-
age totals. Exposure is defined by HEM as the occurrence of con-
tact between humans and pollutants. Dose is the total amount of
material received.
388
HEALTH & ASSESSMENT
-------
Tables
Example Combined Dispersion Model/HEM Calculations
Unit Risk of 4.IOE-03
-Ha

Cone
.OJE«00
.»4E-01
.31E-OI
.53E-01
.03E-OI
.34E-01
.7IE-01
.7CE-01
.40E-01
.01E-OI
.»4E-02
.0»E-02

X 1 II

People
13
13
13
13
13
13
13
13
13
13
13
13

u a -

Exposure
4.03E+01
1.05E«01
5.4BE»00
4.44E+00
4.00E+00
3.09E«00
2.23E»00
2.24E+00
I.B5E+00
1.33E*00
7.84E-01
4.72E-01

Liretine
Incidence
2.74E-03
7.14E-04
3.B4E-04
3.17E-04
2.72E-04
2.IOE-04
1.S3E-04
1.32E-04
I.24E-04
V.04E-03
3.33E-03
4.57E-03

Max Risk
2.07E-04
S.41E-03
2.»3E-05
2.40E-05
2.04E-03
1.3TE-05
1.14E-03
I.14E-03
T.53E-04
4.87E-04
4.04E-04
3.44E-04

- - II

Cone
3.02E-04
1.35E-04
B.I7E-05
4.27E-05
5.1»E-03
4.11E-03
2.»2E-03
3.13E-03
3.14E-03
1.B4E-03
1.08E-05
1.0IE-03

1 n 1 H

People
,370,000
,370,000
,370,000
,370,000
,370,000
,370,000
,370,000
,370,000
,370,000
,570,000
,370,000
,370,000
1,370,000
u « - -

Exposure
4.40E+03
1.70E+03
I.04E»03
7.T7E+02
4.47E+02
S.27E»02
3.73E»02
4.00E+02
3.«OE»02
2.35E+02
1.38E«02
1.38E+02
1.2SE»04

Annual
Incidence
.0042
.0014
.0010
.0008
.0004
.0003
.0004
.0004
.0004
.0002
.0001
.0001
0.012

Repeal
Interval
140.
410.
W.
1,300.
1,300.
2,000.
2,700.
2,400.
2,400.
4,400.
7,400.
7,400.
BO.

Source
30KN RINO DIST.. SOURCE*)
30KH RINB IIST..SOURCEI2
SOKH RIKB IIST. .SOURCES 13
30KN RIN8 IIST.,SOUftCER7
SOKH RIR6 IIST. .SOURCE!?
SOKH RIN6 IIST..SOURCEI4
SOKH RING IIST.,SOURCEI8
SOKH RIN8 IIST.,SOURCEI3
SOKH RIMB IIST. .SOURCES 11
SOKH RINB IIST.,SOURCER3
SOKH RIH6 IIST.,SOURCER4
SOKH RINB IIST.,SOURCE*t«
Overall
Output is obtained from HEM for each source and source
group. The output contains the following:
• Input ambient estimated concentrations
• Exposure calculations
• Overall summary
• Calculation of annual incidents
• Incident repeat intervals
Table 5 shows an example output from HEM incorporating
the dispersion modeling results, populations and unit risk factors.
Sources or source groups are presented in the table. The variables
listed include:

• Unit risk—the estimates excess probability of contracting can-
cer as the result of a constant exposure over a 70-yr period to
an ambient concentration of 1 jig/m3
• Concentration—the maximum or minimum ambient concen-
trations, ug/m3, predicted to result due to the emissions from
each source
• People—the population exposed to concentrations equal to or
greater than the minimum concentration in each level
• Exposure—the occurrence of contact between humans and
pollutants
• Lifetime incident—equal to maximum exposure times unit risk.
This represents the estimated excess probability of contracting
cancer as the result of a constant exposure of the number of
people exposed to the maximum concentration over a 70-yr
period.
• Annual incidence—equal to cumulative population exposure
(to concentration equal to or greater than the minimum con-
centrations) times unit risk divided by 70 yr. This represents the
annual estimated excess probability of one of the 1,570,000
people in the 50-km area contracting cancer as a result of all
concentration levels predicted to occur in the 50-km region.
• Repeat interval—equal to the inverse of the annual incident
rounded to the nearest 10 yr. This represents the estimated
number of years between excess cancers produced as a result of
exposure of the people within 50-km of the plant to predicted
concentrations.
' Source—the source group for which the summary pertains.
CONCLUSION
A modeling methodology has been presented which combines
available and recommended air quality dispersion models with
population distributions and unit risk factors. The applicatory
value of such a hybrid technique is in the analysis of the air con-
centrations and potential associated health effects due to releases
of toxic emissions. Using models with which regulatory agencies
are familiar is an advantage to this methodology that is high-
lighted by the fact that these models offer capabilities to examine
variable source, emission, transport and dispersion phenomena.
Because the U.S. EPA is investigating and performing analyses
on a wide variety of potentially hazardous pollutants, including
radionuclides and their associated health effects, the need for
this type of approach is apparent.
Table 6
Summary of Repeat Intervals (One Cancer Incidence Per Repeat
Interval) by Source Group
Year of analysis Repeat interval
1981 '
2
3
1982 '
2
3
1983 l
2
3
1984 '
2
3
1981-
1984 l
2
3
1 Using a unit cancer risk
2 Using a unit cancer risk
3 Using a unit cancer risk
17
390
2.2 x 10J
26
590 ,
3.3 x 10J
24
550 ,
3.1 x 10J
20
460 ,
2.6 x 10J
21
480 ,
2.7 x 10J
factor of 4.1 x 10"6.
factor of 1.8 x 10"7.
factor of 3.2 x 10" .
HEALTH & ASSESSMENT 389
-------
Principal uncertainties in the methodology lie in the charac-
terization of emissions, the model applicability and performance,
the population data and its integration with modeled concentra-
tions and the selection of unit risk factors. The former areas are
those experienced in typical dispersion modeling studies and the
importance of emission characterization cannot be overstated.
Alternative population distributions and integration methods
may be applicable with the sensitivity to this factor being uncer-
tain at this time. The choice of unit risk factor is very important
as can be seen in the summary of repeat intervals presented in
Table 6, which shows a great range of repeat years for a cancer
incidence.
REFERENCES
1. U.S. EPA, "Guideline on Air Quality Models" (Revised), Research
Triangle Park, NC, Nov. 1984, Draft.
2. Anderson, G.E., Liu, C.S., Holman, H.Y. and Killus, J.P., "Human
Exposure to Atmospheric Concentrations of Selected Chemicals,"
U.S. EPA Contract No. 68-02-3066, U.S. EPA, Research Triangle
Park, NC, Mar. 1980.
3. Busse, A. and Zimmerman, J., "User's Guide for the Climatological
Dispersion Model." EPA-R4-73-024, U.S. EPA, Research Triangle
Park, NC, Dec. 1973.
4. U.S. EPA, "Regional Workshops on Air Quality Modeling: A Sum-
mary Report," U.S. EPA, Research Triangle Park, NC, Revised
Aug. 1983.
5. Bowers, J.F., Bjorklund, J.R. and Cheney, C.S., "Industrial Source
Complex (ISC) Dispersion Model User's Guide," Volume I, EPA-
450/4-79-030, U.S. EPA, Research Triangle Park, NC, Dec. 1979.
6. Bjorklund, J.R. and Bowers, J.F., "User's Instructions for the
SHORTZ and LONGZ Computer Programs" (Volume I), EPA-
903/9-82-0048, U.S. EPA, Region 3, Philadelphia, PA, Mar. 1982.
7. Holzworth, G.C., "Mixing Heights, Wind Speeds, and Potential for
Urban Air Pollution Throughout the Contiguous United States,"
AP-101, U.S. EPA, Research Triangle Park, NC, Jan. 1972.
8. U.S. EPA, "Summary of Data on Specific Pollutants and Cancer
Risk Estimates Excerpted from EPA Draft Study on Air Toxics Prob-
lems in United States," U.S. EPA, Office of Air and Radiation,
Washington, DC. July 1984.
390 HEALTH & ASSESSMENT
-------
A Multimedia Exposure Assessment Model for
Evaluating Land Disposal of Hazardous Waste

Atul M. Salhotra, Ph.D.
David R. Gaboury, P.E.
Woodward-Clyde Consultants
Walnut Creek, California
Peter S. Huyakorn, Ph.D.
GeoTrans, Inc.
Herndon, Virginia
Lee Mulkey
U.S. Environmental Protection Agency
Technology Development and Application Branch
Athens, Georgia
ABSTRACT
A unified multimedia model to simulate the fate and transport
of organic pollutants emanating from hazardous waste land dis-
posal facilities is described. The model accounts for the move-
ment of contaminants in the subsurface environment, air and sur-
face water. The subsurface model consists of one-dimensional
flow and transport in the vadose zone under the land disposal unit
coupled with a three dimensional, semi-analytical transport model
for the saturated zone. Contamination of a surface stream due to
the complete interception of the groundwater plume is included as
well as emissions of toxic chemical vapor into the atmosphere.
The fate and transport of contaminants in the various media
depends on the chemical and biochemical properties of the con-
taminants as well as the environmental parameters that describe
the multimedia environment. The uncertainty, spatial and tem-
poral variability in these parameters is quantified using the Monte
Carlo simulation technique.
The model can be used as a quantitative tool to assess the long-
term exposure from the disposal of hazardous waste in land dis-
posal facilities. It can be used for screening-type analyses to sim-
ulate "generic" land disposal facilities or for site specific studies.

INTRODUCTION
This paper describes the conceptual development and imple-
mentation of a multimedia model to simulate the release and sub-
sequent fate and transport of organic pollutants placed in a haz-
ardous waste disposal facility. Release to air and soil including
the unsaturated and the saturated zone and possible interception
of the contaminant plume by a surface stream are included. The
model can be used as a tool to quantify the exposure to humans
and aquatic organisms due to the land disposal of hazardous
wastes.
The development and application of such unified multimedia
models are still in their infancy. The available models can be
broadly classified into two groups: compartmentalized models
and spatial models.' The former simulate the multimedia environ-
ment as a set of well mixed compartments such as air, water and
soil. These models lack spatial resolution but provide quantita-
tive estimates of the overall partitioning of the pollutant in the
various media and are useful for large scale regional studies.
Spatial models provide a one-, two- or three-dimensional spa-
tial and often temporal resolution of the distribution of the pollu-
tants hi the environment. Typically, these models consist of a
series of coupled media-specific, differential mass balance equa-
tions that are solved using analytical, semi-analytical or numer-
ical techniques. Although complex three-dimensional spatial
models for each medium including a variety of fate and transport
processes are available, the inter-media links have not been ade-
quately addressed.2-3 The model presented in this report is a com-
bination of these two approaches.
The application of multimedia models has significantly lagged
behind the conceptual as well as software developments. This is
primarily due to the lack of sufficient data to calibrate and/or
verify these models. Nevertheless, the application of such models
provides an order-of-magnitude quantitative estimate for chem-
ical concentrations in the various environmental media from
which the exposure dosage can be estimated.

PHYSICAL SCENARIO AND SOURCE
CHARACTERIZATION
The source of toxic pollutants considered in the model is the
post-closure period of a hazardous waste land disposal facility.
Operation and maintenance of the facility during the pre- and
post-closure periods is assumed to provide adequate protection
to human health and the environment based on the design per-
formance standards and monitoring requirements. The failure of
the engineering controls (i.e., caps and liners) is assumed to occur
after the post-closure period and results in the simultaneous re-
lease of the leachate to the groundwater and emission of vapor to
the atmosphere. Note, however, that the model does not attempt
to simulate the failure modes of the facility.
Detailed descriptions of the fate and transformation of chem-
icals in the landfill prior to the release are not included. However,
variations in chemical properties such as adsorption coefficients
and degradation rates are accounted for using the Monte Carlo
technique.

OVERALL FRAMEWORK FOR THE MODEL
At the outset, it was realized that this multimedia model would,
over time, be modified and enhanced by the addition of new algo-
rithms and/or by replacing existing algorithms by more realistic
and complex ones. This required the code to be modular in struc-
ture to facilitate future enhancements. Further, it was considered
desirable to incorporate the ability to run the code in both a batch
and an interactive mode.
With these requirements in mind, a conceptual framework of
HEALTH & ASSESSMENT 391
-------
the multimedia model was developed. This framework, shown in
Fig. 1, essentially consists of 15 high-level modules. Of these, the
first 10 are utility modules; the remaining five modules include
the computational algorithms for each medium and the source.
»MMJLT« FAMCO THMOUOM COMMON OUTVAL
Figure 1
Conceptual Framework and Software Architecture for the
Multimedia Model
UNCERTAINTY ANALYSIS
The fate and transport of contaminants in the multimedia mod-
el critically depends on a number of model parameters. These
parameters can be broadly classified into two categories whose
uncertainties arise principally from different sources. The first set
of model parameters consists of chemical, biochemical and toxi-
cological properties of the hazardous constituent such as the octa-
nol-water partition coefficient, acid, base and neutral hydrolysis
rate constants, etc. Uncertainty in this set of parameters primar-
ily arises from laboratory measurement errors or errors due to ex-
trapolation from controlled laboratory measurement conditions
to uncontrolled environmental conditions.
The second set of model parameters consists of media-spe-
cific parameters, examples of which include the groundwater
velocity, wind speed, etc. Uncertainty in these parameters pri-
marily arises due to inherent natural, spatial and temporal vari-
abilities. In addition, the effect of measurement errors and ex-
trapolation errors is also included.
The effect of these uncertainties and variabilities on the model
output can be quantified in a variety of ways. A detailed review,
advantages and disadvantages of available methods have been
presented in a report to the U.S. EPA/ For the current version
of the model, the Monte Carlo simulation technique was adopted.
The use of this model necessitated the development of an input
processor with the capability to accept probability distributions of
input parameters (in addition to constant values if desired) and
to sequentially generate parameter values drawn from these dis-
tributions. Further, the model also includes an uncertainty post
processor that computes the cumulative frequency distributions
and other desired statistics of the model output.

THE VADOSE ZONE MODULE
In the event that the hazardous waste disposal facility is lo-
cated at the ground surface and the water table is deep, it is im-
portant to analyze the fate and transport of the contaminant
within the vadose zone (Fig. 2). The vadose zone module con-
sists of two algorithms—one for flow computations and the other
for the transport. Water contents as a function of depth are gen-
erated by the flow algorithm and are used to compute the pore
water velocities. These values are required by the vadose zone
transport algorithm.

The Vadose Zone Flow Algorithm
Flow in the vadose zone is considered to be steady and one-
dimensional. The governing equations are given in Appendix 1.
To solve the flow problem, it is necessary to specify the relation-
ships between the relative permeability and water saturation and
between pressure head and water saturation. The code allows
these two relationships to be included as analytical expressions9
or as piecewise linear curves.

LAND DISPOSAL
UNIT
FLOW
SATURATED
ZONE
Figure 2
A Schematic Diagram of Leachate Migration through the Vadose
Zone and into the Saturated Zone
The flow equations are solved (to yield pressure and water con-
tent variation with depth) using the Galerkin finite element tech-
nique. Application of this method results in a set of nonlinear,
simultaneous algebraic equations' that are solved using the
Thomas algorithm. Note that the code allows the non-linearity in
the equation to be treated using either the Picard or the Newton-
Raphson scheme.

Vadose Zone Transport Algorithm
The transport of contaminants within the vadose zone is treated
as a one-dimensional problem. Important fate and transport
mechanisms considered by the model include longitudinal disper-
sion, linear adsorption, chemical and biochemical first order de-
cay and advection. The source boundary conditions at the waste
disposal facility are represented as either a constant concentra-
tion, decaying concentration or a finite (constant) pulse concen-
tration source. The relevant mathematical equations and boun-
dary conditions are shown in Appendix 1.
The effect of varying degradation rates, dispersion coefficients
and seepage velocities within the vadose zone is accounted for by
dividing the vadose zone into a number of horizontal layers,
each of which is assumed to be homogeneous. The transport
equations are solved for each layer ensuring continuity of con-
centrations at the layer boundaries.
The solution to the layered vadose zone transport problem is
derived using Laplace transform techniques to transform the gov-
erning partial differential equations and the boundary conditions
to an ordinary differential equation in the Laplace domain. The
392 HEALTH & ASSESSMENT
-------
solution is obtained in the Laplace domain and then inverted us-
ing either a convolution theorem or the Stehfest numerical inver-
sion scheme.7-8 In general, the Stehfest method is computationally
faster than the convolution scheme but may exhibit some numer-
ical problems at high Peclet numbers.
Using the above method, the concentration at as well as the
mass flux to the water table can be computed. These values were
used to derive the source boundary conditions for the saturated
zone transport model. This provides the coupling between the two
modules.

THE SATURATED ZONE TRANSPORT MODULE
Contaminants can enter the saturated formation by direct
leaching from the land disposal unit (in the absence of a vadose
zone) or by flow and transport through the vadose zone. The
multimedia model allows the user to specify either of the above
options. Note that in both cases the governing equation and,
hence, the semi-analytical solution for transport in the saturated
zone are similar.
The saturated transport model considers uniform flow in a
homogeneous aquifer of uniform thickness with dispersion in the
longitudinal, transverse and vertical directions. In addition, linear
adsorption, first-order decay and dilution due to direct recharge
into the plume are included in the model. The flow domain is re-
garded as semi-infinite in the longitudinal, infinite in the trans-
verse and finite in the vertical direction. At the source, down-
stream edge of the land disposal unit, the contaminant concentra-
tion is assumed to be Gaussian in the lateral direction and uni-
form over a vertical mixing or penetration depth. A schematic
description of the flow domain and the source boundary con-
dition is shown in Fig. 3.
The vadose zone and the saturated zone transport modules are
coupled together by ensuring continuity of mass at the water table
as shown in Fig. 3. The penetration depth into the saturated zone
is computed by equating the flux of contaminant leaching out of
the vadose zone (or the land disposal unit) to the advective mass
flux flowing into the saturated formation.
A semi-analytical solution to the saturated zone transport equa-
tion has been presented by Huyakorn, et al.9 The numerical in-
tegration is performed using the Gauss-Quadrature Integration
Scheme.10 The model yields downgradient concentrations of the
contaminant as a function of space coordinates and time. In the
event of a steady infinite source, maximum concentration is ob-
tained at the plume centerline after a long time period, i.e., when
the plume is fully developed. Human exposure to the contaminant
can occur due to the consumption of water from a well located
within the contaminated, saturated zone.
.c =
X-Z PLANE
THE SURFACE WATER MODULE
Contamination of a surface stream can occur through the in-
terception of the saturated zone contaminant plume as shown
in Fig. 4. Exposure to human beings can occur through two
routes; exposure due to drinking stream water and exposure due
to the consumption of fish exposed to the contaminated water.
Further, stream contamination can result in toxicity to aquatic
organisms. These various exposure routes as well as different
scenarios resulting in stream contamination are presented by
Ambrose, et al."
The current version of the multimedia model assumes com-
plete interception of the groundwater plume. It is restricted to the
case of upland watersheds only where the assumption of complete
mixing within the stream would be applicable. Transport within
the stream is assumed to result in first order decay due to bio-
chemical and chemical degradation as well as volatilization.
Under these assumptions, the equations for the surface zone
module are presented in Appendix 1.
For the case of exposure to toxicants from drinking water, it is
assumed that water is pumped by a municipal water treatment
plant at a downstream location. The effect of primary sedimen-
tation on the removal of toxic chemicals is accounted for by the
model. Efforts are underway to develop a realistic model to sim-
ulate the removal efficiencies of various primary, secondary and
tertiary municipal water treatment processes.
Figure 4a
Groundwater Contamination Plume Interception by the Surface Stream
Plum* Boundary.
.
f Not Fhld
/ Mining Zona
Matujramant
•own
Groundwater
Loadina.
Figure 3
A Schematic Diagram of the Source Boundary Conditions for the
Saturated Zone Transport Module
Figure 4b
Downstream Contaminant Transport from the Edge of the
Initial Mixing Zone
To estimate the human exposure due to the consumption of
fish, it is necessary to estimate the bioconcentration factor for the
fish. This is obtained with the conservative assumption that the
concentration of the contaminant within the lipid and the non-
lipid phase of the fish is in equilibrium with the dissolved concen-
tration in the blood. The latter is assumed equal to the near-field
dissolved stream concentration. The overall fish partition coeffic-
ient is obtained by weight averaging the lipid and nonlipid bio-
mass of the fish.

THE AIR EMISSIONS MODEL
Emission of toxic chemicals from the waste disposal facility is
HEALTH & ASSESSMENT 393
-------
treated as a diffusion controlled process using Pick's Law for
steady state diffusion. Recall that the model is being developed
for the post-closure period of a hazardous waste disposal facility.
The case of co-disposal of waste is not treated here. Further, the
wastes contained within the facility are assumed to be suitably
segregated into cells so that there is no chemical or biochemical
activity within the facility that would generate gaseous by-pro-
ducts. With these assumptions, diffusion into the atmosphere
emanates from a plane surface with a constant concentration.
The model equations are presented in Appendix 1.
In addition to the diffusive transport mechanism, vapors can be
pumped out by barometric pressure fluctuations. Models that
simulate this transport mechanism12-13 have not been sufficiently
calibrated and verified.
This effect has been incorporated into the model by using an
empirical factor, 'CUP'. Further, laboratory studies have indi-
cated that the diffusive models typically underestimate the emis-
sions by a factor of about three. This anomaly has been ascribed
to the process of surface diffusion." This effect can be empir-
ically included by means of calibration factor 'ef*.
An important variable in the diffusion model is the pore vapor
concentration. This concentration depends on the properties of
the chemical species, quantity in the cells and the state of the
chemical with respect to other materials in the cells. Table 1 shows
a categorization of the various states of chemical substances in
landfills and the equilibrium laws that can be used to determine
the pore space chemical concentration. The current version of the
model takes the conservative approach and assumes that the
chemical exerts pure component vapor pressure. This assumption
is equivalent to assuming the existence of high concentrations of
solid wastes in pure form or as mixtures of solid flakes and gran-
ules.

MODEL RESULTS AND CONCLUSIONS
At the time this paper was written, individual algorithms for
each module as well as a sophisticated interactive processor had
been developed. Each algorithm has been individually tested and
has been linked to the others. Detailed testing and verification of
the code is in progress.
Typical results that can be expected from such a model are
shown in Fig. 5. The results indicate the composite effect of vari-
ous parameter values and their uncertainties on the concentra-
tions in the groundwater, surface stream, fish and atmosphere
(once the air dispersion model is implemented). In addition to the
graphic results, tabulated values of various percentiles and other
statistics can be generated by the model. These results can be used
to perform both exposure and risk assessments due to the disposal
of hazardous materials on land and their subsequent transport
in the various media of the environment. In this respect, the mod-
el represents an important tool for a comprehensive, quantita-
tive risk assessment.

ACKNOWLEDGEMENTS
Acknowledgements are due to J. Kittle of Aqua Terra Consul-
tants, Mountain View, California for developing the software
1.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 05 1.0
NORMALIZED CONCENTRATION

Figures
Typical Concentration Distributions Obtained Using the
Multimedia Model
State

High concentration*
Table 1
Stale of Chemical Substances In Landfills and Vapor Density Eqnfflbrtam Law*
Equl11br1um Law
Pure substances, solid form
Mixture, solid flakes, powder, granular
Pure substance, liquid
Liquid oil mixture, homologous mixture
Non-homologous mixture of liquids (non-aqueous)
Water solution of dissolved chemicals
Liquid chemicals solidified with fill material, sub-soil*
Liquid chemicals solidified with fly ash, sand*
Liquid chemicals solidified with waste paper, garbage,
organic materials*
01l-1n-water sludges, emulsions
Water-based mlcells*

Low concentrations

Pure substance mixed with fill material, fly ash, sand,
other Inorganic material (no water)*
Contaminated soil or small quantities of chemical mixed
with soil, water and organic matter present
Small quantities of chemical mixed with waste paper or
other organic material such as garbage*
Pure Component Vapor Pressure (PCVP)
PCVP, Oalton's Law of partial pressure (OL)
PCVP
OL, Equilibrium Concentration for hydrocarbon mixtures (ECHH)
Activity Coefficient (AC)
Henry's Law (HL), AC
PCVP, OL, BET Theory"
PCVP, BET. OL

PCVP. OL
Solubility. HL, AC
BET

Henry's Law/Soil Water partition coefficient (HCPC)

HCPC
*H1gh concentration 1s defined as approximately 5 percent of the
waste material.
"Brunauer, Emmett and Teller Adsorption Theory.
+Methodology unclear, research needed.
Source: Reference (14)
394 HEALTH & ASSESSMENT
-------
architecture and to P. Mineart of Woodward-Clyde Consultants
for developing the utility subroutines.
This work was sponsored by U.S. EPA under Contract No.
68-03-6304.
REFERENCES
1. Cohen, Y.( "Organic Pollutant Transport. Improved Multimedia
Modeling Techniques Are the Key to Predicting the EnvironmentaT
Fate of Organic Pollutants." Environ. Sci. Technol., 20, 1986,
538-544.
2. Anderson, M.P., "Using Models to Simulate the Movement of Con-
taminants Through Groundwater Flow System." CRC Crit. Rev.
inEnviron. Control, 9,1979, 97-156.
3. Swann, R.L. and Eschenroeder, A., Eds., Fate of Chemicals in the
Environment. ACS Symposium Series 225, American Chemical
Society, Washington, DC, 1983.
4. Woodward-Clyde Consultants and Vaneziano, D., "Development of
an Approach for Conducting Uncertainty Analysis in Multimedia
Modeling." EPA Project No. 68-03-6304, U..S. EPA Technology
Development and Applications Branch, Athens, GA, 1986.
5. Brooks, R.H. and Corey, A.T., "Hydraulic Properties of Porous
Media." Hydrology Papers No. 3, Colorado State University, Fort
Collins, CO, 1964.
6. Lester, B.H., Huyakorn, P.S., White, H.O., Wadsworth, T.D.
and Buckley, I.E., "EPACML Composite Analytical-Numerical
Model for Simulating Leachate Migration in Unconfined Aquifers."
Prepared by GeoTrans, Inc., Report for Woodward-Clyde Con-
sultants, Walnut Creek, CA, 1986.
7. Moench, A.F. and Ogata, A., "Numerical Inversion of the Lap-
lace Transform Solution to Radial Dispersion in a Porous Medium."
Water Resources Res., 17, 1981, 250-252.
8. Stehfest, H., "Numerical Inversion of Laplace Transforms." Com-
mun. ACM, 13, 1970,47-49.
9. Huyakorn, P.S., Ungs, M.J., Mulkey, L.A. and Sudicky, E.G.,
"A Three Dimensional Analytical Method for Predicting Leachate
Migration," submitted to Groundwater, 1986.
10. Carnahan, B., Luther, H.A. and Wilkes, J.O., "Applied Numer-
ical Methods," John Wiley, New York, NY, 1969.
11. Ambrose, R.B., Mulkey, L.A. and Huyakorn, P.S., "A Method-
ology for Assessing Surface Water Contamination due to Land Dis-
posal." U.S. EPA Draft Report, Athens, GA, 1985.
12. Thibodeaux, L.J., et al., "Models of Mechanisms for the Vapor
Phase Emission of Hazardous Chemicals from Landfills." J. ofHaz.
Mat., 7, 63-74, 1982.
13. U.S. EPA, "Evaluation and Selection of Models for Estimating Air
Emissions from Hazardous Waste Treatment, Storage and Disposal
Facilities." EPA-450/3-84-020, Office of Air Quality Planning and
Standards, RTF, NC, 1984.
14. Thibodeaux, L.J., Springer, C. and Hildebrand, G., "Transport of
Chemical Vapors through Soil—A Landfill Cover Simulation Ex-
periment." Presented at 1986 Summer National American Institute
of Chemical Engineers, Boston, MA, Aug. 1986.
15. Cussler, E.L., "Diffusion: Mass Transfer in Fluid System." 1st Ed.,
Cambridge University Press, 1983,188-189.
Appendix 1
WDOSE ZONE FLOW MODULE

Governing Equation:
Boundary Conditions:

V(0,t) = I

iji(L,t) = 0

Functional Relationships:

fS S ]n
'Yw
Sw
(1
Swr
S )n
wr;
f *
for i|>4i
(2a)

(Zb)

(3)

(4)
VADOSE ZONE TRANSPORT MODULE

Governing Equations:
dt
D
dz

R = 1 + pbKd
±_ \i "^ \ r
2 s dz v
(5a)

(5b)

(5c)
Initial and Boundary Conditions:

C(z,t) 0

C(o,t) CQ

C(o,t) = CQ exp(-Yt)

C(o,t) C0[l-s(t-T)]

C(-.t) 0

SATURATED ZONE TRANSPORT MODULE

Governing Equation:
(6a)

(6b)

(6c)

(6d)

(6e)

xle
d_
dz
i
(D
Initial and Boundary Conditions:

C(x, y, z, o) 0
(7b)
(8a)
HEALTH & ASSESSMENT 395
-------
C(0, y. z, t) - CQ exp|-y2/(2o2)| , 0 < z < H (8b)
C(x. +-, z. t) 0
C(-, y, z. t) 0
||(x, y, 0, t) - 0

|§(x. y, B, t) 0
SURFACE ZONE MODULE
•
m
m mL exp (- R y L ) exp ( ^-)
C(o.t) IA
"R "s
AIR EMISSIONS MODULE
E1 Dei(Csi Cai>eBPef
(8c)

(8d)

(8e)

(8f)

(9a)

(9b)

(9c)
<10b>
P* M,
cs1
C_^ * Pore space concentration of the chemical, 1,
Ig/rn3)
0 = Longitudinal hydrodynamlc dispersion 1n the
2
vadose zone [m /day)
De1 = Effective diffusion coefficient for chemical
'1' 1n soil |m2/day|
°xx» Oyy °zz Hydrodynamlc dispersion coefficient
in the x, y and z directions in saturated
zone |nr/day|
ebo " Empirical factor to account for emissions due to
fluctuations in atmospheric pressure
ef Empirical calibration factor

EJ Emission rate into the atmosphere for the
chemical 'i' |g/day|
H <* Depth of source penetration within the saturated
zone |m|
I Infiltration rate through the land disposal
facility |m/day|
k Saturated hydraulic conductivity for the vadose
zone [m/dayl
Kd = Distribution coefficient for chemical in the
liquid and solid phase.

*rw * Relative permeability of the vadose zone

PQ^ - Vapor pressure of chemical i |mm of mercury)

q Infiltration Into the groundwater plume |m?day)

R = Retardation factor for vadose zone [dimensionless)

R' Retardation factor for the saturated zone
Idimenslonless]
Rg Universal gsa constant [mm mercury-l/°K-mol |

Sw - Water phase saturation in the vadose zone
Idimenslonless)
396 HEALTH & ASSESSMENT
-------
Swr Residual water phase saturation for the vadose
zone [dimensionless]
XR
Duration of pulse source [days)

Oarcy velocity in the vadose zone [m/dayl

Seepage velocity in the vadose zone [m/day]

Seepage velocity in the saturated zone [m/day1

Mole fraction of chemical i in the landfill
[dimensionless]

Travel distance within the saturated zone [m]

Travel distance within the stream [m]

Vertical coordinate pointing downwards [m]
n Empirical coefficients in the vadose zone
functional relationships [dimensionless]
Air filled porosity
Porosity of the saturated or unsaturated soil
[dimensionless]

Bulk density of the soil [g/cc]
Biological decay coefficient for the vadose zone
[I/day]

Biological decay coefficient for saturated zone
[I/day]

Decay coefficient within the stream [I/day)
Overall decay coefficient within the vadose zone
[I/day]

Liquid phase chemical decay coefficient [I/dayI

Solid phase chemical decay coefficient [I/day]

Standard deviation of the gaussian contaminant
source [m]

The pressure head [m]

Air entry pressure head [ml
HEALTH & ASSESSMENT 397
-------
OSHA Standards or Risk Analysis: Which Applies?
Arthur D. Schatz
Michael F. Conway, P.E.
Gradient Corporation
Cambridge, Massachusetts
ABSTRACT
Federal agencies historically have relied on two distinct regula-
tory approaches for controlling exposure to hazardous chemicals
for workers and the general public. Occupational standards have
been developed by OSHA to protect industrial workers, while
risk-based guidelines more recently have been developed by the
U.S. EPA to protect public health. Each approach has been ap-
plied effectively to situations which distinctly involve either oc-
cupational or general population exposures. More recently,
however, cases have emerged where it is not clear which standard
applies.
This paper presents two examples which illustrate a dilemma
potentially created by these two regulatory approaches. Each case
concerns the exposure of workers to chemicals originating from a
hazardous waste site adjacent to the workplace. The approaches
used to evaluate the acceptability of the air contamination present
are presented, and the need for a consistent assessment policy is
discussed.
INTRODUCTION
The Occupational Safety and Health Administration (OSHA)
and the U.S. EPA were created to oversee the quality of two dis-
tinct environments. With regard to chemical hazards, OSHA
focuses on workers' health, primarily indoors, whereas the U.S.
EPA focuses on public health, primarily outdoors or in the am-
bient environment. In the past each agency has functioned inde-
pendently in regulating its environmental sphere. As a result, two
distinct approaches toward evaluating the health effects of
chemical exposures have evolved.
The distinction between the occupational and the ambient en-
vironments is not always as clear as the legislators responsible for
creating OSHA and the U.S. EPA and their original ad-
ministrators may have thought. As analytical advances increase
are ability to measure chemical contamination and as our abilities
to model transport pathways and routes of exposure grow more
sophisticated, the potential interactions between ambient and oc-
cupational environments become more apparent. Whereas the
potential for industry to affect the ambient environment has long
been recognized, the impact which environmental releases of con-
taminants may have on occupational settings has not.
This paper discusses a dilemma posed by the divergent
regulatory approaches of OSHA and the U.S. EPA for evaluating
chemical exposures. Two case studies are presented to illustrate
the problem and discuss how particular situations have been ad-
dressed. Finally, the need for an unambiguous uniform policy is
discussed.
BACKGROUND
The Occupational Safety and Health Act of 1970 authorized
the Secretary of Labor to set exposure standards for chemicals in
the workplace as follows:
"The Secretary, in promulgating standards dealing with
toxic materials or harmful physical agents under this sub-
section, shall set the standard which most adequately as-
sures, to the extent feasible, on the basis of the best
available evidence, that no employee will suffer material
impairment of health or functional capacity even if such
employee has regular exposure to the hazard dealt with by
such standard for the period of his working life."
The "feasibility" provision of this section of the Act assured
that the economic impact of standards would be considered and
that no undue hardship would befall industry as a result of these
occupational standards. The standards which were subsequently
developed were designed to prevent clearly recognized health ef-
fects of the regulated chemicals. As a result the standards for
some chemicals are based on a chronic health effect (such as vinyl
chloride: carcinogenesis), while the standards for the other chem-
icals are based on acute effects (such as acrolein: eye and upper
respiratory tract irritation).
While the OSHA permissible exposure limits (PELs)1 are the
enforceable national occupational standards, many industrial
hygienists consult additional guidelines developed by other
groups such as the American Conference of Governmental In-
dustrial Hygienists (ACGIH) or the National Institute of Occupa-
tional Safety and Health (NIOSH). For instance, the ACGIH
threshold limit values (TLVs)' are frequently utilized instead of
the OSHA standards by industrial hygienists, because the ACGIH
guidelines are considered more up-to-date (they are revised an-
nually).
The Clean Air Act of 1970, the first legislation administered by
the U.S. EPA, authorized the setting of the National Ambient Air
Quality Standards (NAAQSs) on the basis of health alone; the
economics of attainment was intentionally omitted from the for-
mula. The formulation of risk analysis methods in the late 1970s
as a U.S. EPA method for evaluation environmental chemical ex-
posures was a natural extension of the spirit of the Clean Air Act
NAAQSs.
The U.S. EPA issued its first set of health risk-based guidelines
in 1980, the Ambient Water Quality Criteria.3 Each water quality
criterion for suspected carcinogens was an estimate of the water
concentration which, based on conservative assumptions, posed a
given risk of cancer incidence (10-5, 10-6or 10-"0 over a lifetime
of consumption. These cancer risk estimates were based primarily
on animal bioassay results extrapolated to humans, extrapolated
398 HEALTH & ASSESSMENT
-------
from high doses to low doses and with safety factors applied. In
other words, these criteria were,conservative guidelines designed
to prevent potential health effects; health effects which would be
difficult or impossible to measure at the incidence rates con-
sidered to be acceptable (e.g., 10-4 to 10-7).
As the 1980s have progressed environmental agencies (state and
federal) have requested and conducted more and more risk assess-
ments. This increased demand has stimulated the development of
more sophisticated exposure and risk assessment methods and
their application to a broad range of potential exposure
pathways. CERCLA called for risk or endangerment assessments
to document the presence and extent of health hazards at hazar-
dous waste sites being considered for remedial or enforcement ac-
tion. To this end, the U.S. EPA issued guidelines for exposure
and risk assessments4-5 and developed methodology handbooks
for Superfund assessments.6'7 As a result risk assessment has
grown to be widely accepted and currently is applied extensively
to a variety of environmental situations.
A risk analysis approach generally will yield much more strin-
gent exposure guidelines than the recommended occupational
limits. This point is illustrated in Table 1 which shows examples of
the air concentrations which would be considered virtually safe
based on risk analysis for three chemicals compared with their oc-
cupational limits. The risk-based air concentrations were
calculated using the most recent U.S. EPA carcinogenic potency
factors (CPFs, also called potency slopes):8
C =
R(BW)
(CPF) (I)
(1)
where C = air concentration, mg/m3
R = incremental carcinogenic risk due to lifetime ex-
posure to the chemical at concentration C, dimen-
sionless
BW = body weight, kg
CPF = carcinogenic potency factor (mg/kg/day) -1
I = inhalation rate, mVday

The following values were assumed for calculations of the con-
centrations in Table 1:

• R = 1 x 10-6
• BW = 70 kg
• I = 20 m3/day

Table 1
Comparison of TLVs and Risk-Based Air Concentrations
10-6 Risk
Chemical
Benzene
PCBs
Trichloroethylene
OSHA
PEL*
(jig/m3)
30,000
soot
540,000
ACGIH
TLV*
0*g/m3)
30,000
soot
270,000
Air Concen-
trationf
0*g/m3)
0.12
0.0008
0.32
Notes:
*8 hour time-weighted average guidelines.
^Air concentration estimated to increase an individual's risk of manifesting cancer by 1 X 10-6,
if exposure is for a lifetime.
{Occupational limits are for Aroclor 1254.

It is not surprising that the allowable chemical exposures are
quite different for the two sets of guidelines. Indeed, there are
physical and philosophical reasons why a worker might be ex-
pected to withstand higher exposures to chemicals than the
general public, as shown in Table 2. A worker generally is con-
sidered to be healthier than the average individual. He or she is
exposed to chemicals for a shorter duration and later in life than
may be the case for the public exposed to environmental con-
tamination. Also, the worker theoretically is compensated for ac-
cepting additional risk, while the public is not. Still, the difference
is very large, with the TLVs up to five orders of magnitude larger
than the risk-based exposure guidelines.
Table 2
Characteristics of Worker and Public Exposures to Chemicals
Factor

Exposure Duration

Maximum (Lifetime!
Exposure Span

Exposed Population

Age

Compensated for Risk
Occupational

e hr/day
5 days/week

40 years

Healthy workers

Adult

Yes
24 hr/day
7 days/week
70 years
Sensitive
individuals
Any (child,
adult, elderly)
In cases where the potential human exposures cut across the
traditional occupational/environmental boundaries, the implica-
tions of which criteria are utilized to evaluate exposure may be
significant. Depending on which criteria are utilized, for instance,
a hazardous waste site remedial action may be accepted, rejected
or delayed.

CASE EXAMPLES
Industries Adjacent to the Hyde Park Landfill
Approximately 80,000 tons of hazardous chemical waste were
disposed of at the Hyde Park landfill in Niagara Falls, New York,
during the 1950s and 1960s. As part of a multi-agency effort to
determine the environmental and health impacts of the Hyde
Park landfill site, beginning in 1979 the National Institute of Oc-
cupational Safety and Health (NIOSH) conducted Health Hazard
Evaluations of three industrial plants abutting the landfill.9-10
There was concern that chemicals from the landfill had migrated
to the industries in the air, on dust or as vapors, or via water con-
duits beneath the plants thought to carry leachate from the land-
fill. Samples of water and sediment from the suspected conduits
and samples of dust from the inside of the plants were collected
and analyzed for three semivolatile chemicals: perchloropentacy-
clodecane (mirex), 7-hexachlorocyclohexane (lindane) and 2,3,7,
8-tetrachlorodibenzo-p-dioxin (TCDD). Substantial quantities of
all three chemicals were known to be present in the landfill. At
about the same time, OSHA conducted an air sampling survey to
determine worker exposure to organic vapors from the landfill.10
A mixed approach was used to evaluate the chemical analysis
data. OSHA determined that workers were potentially exposed to
small amounts of many organic chemicals in air originating from
the landfill. The measured concentrations of these chemicals were
compared with OSHA standards, and it was deemed that the
chemicals did not present an inhalation hazard to workers.
The three chemicals measured in water, sediment and dust
samples were neither subject to a risk analysis nor evaluated by
occupational guidelines. Relatively low levels of all three com-
pounds were found in sediment and dust. For example, dust
samples from one plant contained a maximum of 17 ppb of mirex,
79 ppb of lindane and 0.7 ppb of TCDD. No modeling was per-
formed to estimate occupational inhalation or dermal exposures
to these chemicals. Only one of the chemicals, lindane, had an oc-
cupational exposure guideline (ACGIH TLV = 0.5 mg/m3).
Nonetheless, the chemicals were determined to present an unac-
ceptable hazard based on their presence in measurable quantities
HEALTH & ASSESSMENT 399
-------
and their suspected carcinogenic and reproductive effects. It was
recommended that contaminated dust and sediment be removed.
Thus, exposure to volatile organic compounds from the landfill
was deemed acceptable, based on OSHA standards, while ex-
posure to the three semivolatile compounds were found unaccep-
table based upon neither occupational guidelines nor a scientific
evaluation of the risks posed.
In 1985 NIOSH revisited one of the industrial plants to collect
dust samples for analysis of Hyde Park chemicals." The dust
again was analyzed for mirex, lindane and TCDD plus hexachlor-
obenzene and polychlorinated biphenyls (PCBs). Only PCBs and
TCDD were found at low levels.
This time, the evaluation of exposure to dust contaminated
with PCBs and TCDD was based on a risk assessment. The
measured concentrations of PCBs (2 Mg/m2) and TCDD (0.2
ng/m3) in dust were compared with risk-based surface dust
criteria developed by the New York State Department of Health
(DOH) (PCBs: 100^g/m3; TCDD-equivalents: 3.3 ng/m*). Since
the measured concentrations were lower than the DOH criteria,
the conditions were deemed acceptable. Occupational guidelines
were not considered, although the measured concentrations
presumably would have satisfied them, also.
This case suggests that the NIOSH personnel evaluating the site
felt that the OSHA standards provided adequate protection for
volatile organic compounds, such as those commonly used in in-
dustry, while greater caution was warranted for the non-industrial
compounds. Some precaution would have been expected, since
the TLVs do not directly address exposures via contaminated
dust. Nonetheless, the unusual compounds were treated much
more conservatively than the common volatile compounds, al-
though they all originated outside the plant. The conservative
(this time risk-based) guidelines were applied once again in 198S.

Office Space Adjacent to a Niagara Falls Waste Site
Office space located on the upper floor of a municipal building,
a former water treatment plant pump house currently used for
storage and vehicle maintenance, was offered to U.S. EPA per-
sonnel for use during the remedial investigation activities at
another hazardous waste landfill in Niagara Falls, NY. The
building is located near the landfill and is within the study area of
the investigation.
A preliminary walk-through of the building revealed the
presence of standing water in the building basement and raised
questions regarding the environmental quality of the building's
interior. The standing water possibly contained chemicals
migrating from the site in groundwater which could volatilize to
contaminate the air inside the building. In order to address these
concerns, a sampling program was conducted which included
analyses of samples of air and the standing water in the building
basement.
Table 3
Concentrations of Chemicals Detected In the Proposed Office Space
with Evaluation Criteria
10** Itik Mr

f«i i V»r
Standing water samples were Analyzed for the presence of those
compounds and parameters previously identified as indicators of
contamination from the adjacent waste site. The air samples were
analyzed for a different but overlapping group of the same com-
pounds in both the gaseous and paniculate phase. Results of the
analyses indicated relatively low (ppb) levels of some of the com-
pounds in one or both of the environmental media. Analytical
results presenting concentrations of compounds observed above
the method detection limit in air of the proposed office space are
reported in Table 3.
The U.S. EPA requested that the sampling results be evaluated
based both on occupational guidelines and risk analysis. The oc-
cupational evaluation found that no OSHA standards or ACGIH
guidelines were exceeded by the contaminant levels present. The
upstairs offices were deemed acceptable for occupation based on
this evaluation, although actions were recommended to reduce
carbon monoxide levels from idling vehicles in the garage bay and
to reduce continued air contamination due to the standing water
in the basement.
Both carcinogenic and non-carcinogenic effects were evaluated
in the risk analysis. Lifetime risks for the two carcinogens, tetra-
chloroethylene (TCE) and o-hexachlorocyclohexane (er-HCH),
were estimated by equations (2) and (3):

R = (CPF) (LADD) (2)

where LADD = lifetime average daily dose, mg/kg/day
LADD - (C)
LADD~
(3)
(BW)(L)
where I = inhalation rate, mVhr
E = cumulative chemical exposure, hr
L = lifetime, days

The following values were assumed for calculations with equa-
tions (2) and (3):
• I = 0.833 m3/hr (20 mVday)
• BW = 70 kg
• L = 25,550 days (70 yr)

For 1 yr of occupational exposure, an individual would breathe
the air in the location of interest for 8 hr/day, 5 days/wk, for 50
wk/yr. Thus, for 1 yr of exposure, E = 2000 hr.
Based on the air concentrations shown in Table 3, the lifetime
carcinogenic risk for an individual due to one year's occupation
of the proposed office space is estimated as 1.7 x 10-' from ex-
posure to TCE and 5.0 x 10-6 from exposure to o-HCH. As-
suming the risks are additive, the combined risk from exposure to
both chemicals is 6.7 x 10-6. if an individual were to work in
this office for more than one year, his or her risk would increase
proportionally with exposure duration, assuming the same
chemical contamination levels. Thus, the lifetime carcinogenic
risk after two years employment is estimated as 1.3 x 10-5; after
5 yr employment, 3.4 x 10-5; etc.
The health risk due to the two non-carcinogens, chlorobenzene
and 1,2,4-trichlorobenzene, was evaluated by comparing the
average daily dose (ADD) of each chemical for an office worker
in the proposed space with the corresponding acceptable daily in-
take values (ADI) published by the U.S. EPA.1Z The ADD is
calculated by equation (4):
(4)
a) Calculated from CPFs shown to the right, and assuming the bodyweight, inhalation rate and
annual exposure duration described in the text.
b) Calculated from ADls shown to the right, and assuming the inhalation rate and daily exposure
duration described in the text.
ADD
where E = exposure duration, hr/day
Chlorobenzene dominates the non-carcinogenic risk, since its
air concentration in the building is over 500 times greater than tri-
400 HEALTH & ASSESSMENT
-------
chlorobenzene's and its ADI is lower. Therefore, only chloroben-
zene was evaluated. With I = 0.833 m3/hr as assumed previously
and E = 8 hr/day, the ADD for chlorobenzene is estimated as
611 jtE/day. This ADD is approximately 60% of the ADI. While
lower than the ADI, the ADD for chlorobenzene is high enough
for concern, particularly with the concurrent presence of other
compounds.
The carcinogenic risk estimated for the current building condi-
tions exceeds the 10-6 rjsk ievei commonly accepted by the U.S.
EPA. Occupation of the building for 2 yr or more would raise an
individual's estimated risk to the 10-5 range. There was addi-
tional concern that chemicals not specifically analyzed for might
be present in the building, based on the measurement of 1300 /tg/1
of total organic halogens (TOX) in the standing water in the base-
ment, most of which could not be attributed to the indicator com-
pounds.
The U.S. EPA has not announced a decision regarding oc-
cupancy of the building or announced which criteria, occupa-
tional exposure guidelines or risk analysis, will be the basis for the
decision, as of the tune this paper was written.

CONCLUSIONS
Two case examples have been presented which illustrate the
disparate recommendations which may be made and the confu-
sion which may exist when regulatory worlds collide. Undoubted-
ly, these cases are not unique, and such situations are bound to
recur as our understanding of chemical health effects and ex-
posure pathways and our concern for intermedia chemical ex-
posures grows.
An interagency policy is needed to assure that situations such as
these are treated equitably and not on an ad hoc basis. Both cases
described above involved environmental releases, although other
scenarios are conceivable (such as an industrial chemical release
affecting a neighboring industry). Several difficult questions must
be answered in the development of a comprehensive policy such
as:

• Should workers not customarily exposed to chemicals (such as
office workers) be protected with guidelines more stringent
than OSHA's or ACGIH's?
• Should workers normally exposed to chemicals as part of their
jobs receive more stringent protection against chemicals origin-
ating outside their workplace than provided by OSHA or
ACGIH? What if the incoming chemical is the same as one
used in the plant?

Until a clear policy is developed, decisions will be based on the
predisposition of the decision-maker and the influence of contrac-
tors and affected parties.

REFERENCES
1. Code of Federal Regulations, Title 29, Part 1910, Subpart Z, July
1985.
2. American Conference of Governmental Industrial Hygienists,
"Threshold Limit Values and Biological Exposure Indices for 1985-
86," 1985.
3. U.S. EPA, "Water Quality Criteria Documents, Notice of Avail-
ability," Fed. Reg. 45, (231), Nov. 28, 1980, 79318-79.
4. U.S. EPA, "Guidelines for Carcinogenic Risk Assessment," Fed.
Reg. 51 (185), Sept. 24, 1986, 33992-34003.
5. U.S. EPA, "Guidelines for Exposure Assessment," Fed. Reg. 51
(185) Sept. 24, 1986, 34042-54.
6. Versar Inc., "Superfund Exposure Assessment Manual," Final
draft report to U.S. EPA, Washington, DC, Aug. 1984.
7. ICF, Inc., "Superfund Health Assessment Manual," Draft report
to U.S. EPA, Washington, DC, May 1985.
8. U.S. EPA, "Health Assessment Document for Nickel and Nickel
Compounds," Office of Health and Environmental Assessment,
Research Triangle Park, NC, Sept. 1986.
9. National Institute of Occupational Safety and Health, "Summary
of Mirex, Lindane and Tetrachlorodibenzo-p-dioxin Sam pie Analy-
sis, Hyde Park Landfill Chemical Disposal Site, Niagara Falls, New
York," Hazard Evaluation and Technical Assistance Branch, Pro-
ject No. TA 79-22, Jan. 1980.
10. NIOSH, "Synopsis of Investigation: Hyde Park Landfill Chemical
Disposal Site, Niagara Falls, New York," Hazard Evaluation and
Technical Assistance Branch, Project No. TA 79-22, Dec. 19, 1979.
11. Kominsky, J.R. (NIOSH), Letter reporting results of a health haz-
ard evaluation at Greif Brothers Corporation, to J.C. Gibson (United
Steel Workers of America), Apr. 28, 1986.
12. U.S. EPA, "Summary of Current Oral Acceptable Daily Intakes
(ADIs) for Systemic Toxicants," Draft report prepared by the U.S.
EPA Environmental Criteria and Assessment Office, Cincinnati,
OH, May 1984.
HEALTH & ASSESSMENT 401
-------
A Waste Reduction Program and Assessment
Of Current Status for Illinois

David L. Thomas, Ph.D.
Daniel D. Kraybill, P.E.
Gary D. Miller, Ph.D.
Illinois Department of Energy and Natural Resources
State Water Survey Division
Hazardous Waste Research and Information Center
Savoy, Illinois
ABSTRACT
The Illinois Hazardous Waste Research and Information Cen-
ter (HWRIC) was established in 1984 to solve the state's haz-
ardous waste problems through research, information dissemina-
tion and direct technical assistance. One objective is to reduce the
volume of hazardous wastes generated in the state and the threat
that these wastes pose to human health and the environment.
This paper describes the components of HWRIC's waste reduc-
tion program and the status of industrial waste reduction efforts
in Illinois based on recent information from industry. This in-
formation includes results from a survey of approximately 200
companies which identified their current and alternative manage-
ment options for 385 waste streams; information received from
applicants for the 1986 Illinois Governor's Innovative Waste Re-
duction Awards; and an analysis of waste minimization state-
ments submitted by industry to the Illinois Environmental Protec-
tion Agency (IEPA). From these three sources of information, it
is clear that some generators have substantially reduced the
amount of hazardous wastes they generate. Most generators,
however, do not appear to have given waste reduction thorough
consideration. It is apparent that considerably more can be done
to minimize waste generation.

INTRODUCTION
Waste reduction is a national policy, but it has only very re-
cently gained the support in Congress and in the regulatory agen-
cies to make it a true priority. As Joel Hirschhorn of the Con-
gressional Office of Technology Assessment (OTA) stated at a
waste reduction conference in June 1986, "For 20 years we've
been saying that waste reduction is a top priority, and we haven't
been doing anything about it.'" Under the Hazardous and Solid
Waste Amendments (HSWA) of 1984, Congress declared that it
was "...the national policy of the United States that, wherever
feasible, the generation of hazardous waste is to be reduced or
eliminated as expeditiously as possible." The U.S. EPA2 stated
that, in the broadest sense, HSWA defines waste minimization
as any action taken to reduce the volume or toxicity of waste.
OTA preferred a more restrictive definition of waste reduction:
"in-plant practices that reduce, avoid, or eliminate the genera-
tion of hazardous waste so as to reduce risks to health and the en-
vironment."1 Whereas the U.S. EPA's waste minimization in-
cludes source reduction, treatment and recycling (both on- and
off-site), OTA considers off-site recycling and treatment as waste
management. Although we generally have followed the broader
U.S. EPA definition of waste management in this paper, we be-
lieve that OTA's distinction is a valid one; it reflects a prioritiza-
tion of waste reduction first and waste management second.
Geiser4 stated that 20 to 80% of the total hazardous waste
streams could be reduced by source (waste) reduction. He went
on to state, however, that source reduction still is not fully ac-
cepted either in industrial practice or in public policy debates.
"This slowness can generally be attributed to three factors: (1)
lack of comprehensive planning to encourage source reduction;
(2) lack of institutions to assist industries wanting to treat their
toxic by-products; and (3) an absence of capital for process and
product changes."
In response to similar concerns about hazardous waste man-
agement in Illinois, Governor James R. Thompson and the Illi-
nois legislature created the Hazardous Waste Research and In-
formation Center (HWRIQ in 1984 as part of the state's Chem-
ical Safety Initiative. A brief history of the Center and its pro-
grams has been given by Thomas, Miller and Kamin.9 HWRIC
was established within the Illinois Department of Energy and
Natural Resources (DENR) with a well-defined mission. The new
Center would combine research and education; information
collection, analysis and dissemination; and direct technical assis-
tance to industry, agriculture and communities in a multidisti-
plinary effort to solve Illinois' hazardous waste problems.
The Center also was charged with specific objectives. Those
directly related to waste reduction are:

• Reducing the volume of hazardous wastes generated and the
threat they pose to human health and the environment
• Helping develop and implement a comprehensive hazardous
waste management program for Illinois

This paper describes the basic Center program aimed at reduc-
ing the amount of hazardous waste generated in Illinois. The pro-
gram not only is geared toward those generators of hazardous
waste regulated by the U.S. EPA and IEPA, but also is con-
cerned with the proper management and reduction of wastes pro-
duced by the very small (generally unregulated) generator. The
Center's program is meant to supplement and enhance other
Federal and state waste reduction programs, such as the Illinois
Industrial Materials Exchange Service, which is run by the IEPA
and the State Chamber of Commerce, and the regulatory waste
minimization requirements of both the state and Federal govern-
ments.
In addition to these ongoing efforts, Illinois industrial data
402 STATE PROGRAMS
-------
from three sources also are discussed. One source is a survey that
asked companies what disposal alternatives they were consider-
ing in response to the upcoming (Jan. 1, 1987) restriction in Illi-
nois on landfilling hazardous waste. This second source is in-
formation submitted by industries applying for state-sponsored
awards, which were given to companies with innovative waste
minimization technologies or management strategies. The third
source consists of data from waste minimization statements in-
cluded in the annual hazardous waste generation reports sub-
mitted to IEPA for the year 1985.

HWRIC'S WASTE REDUCTION PROGRAM
HWRIC, having made waste reduction a priority, is helping
Illinois industries reduce the amount of hazardous waste they
generate. The following activities are a part of the Center's over-
all waste reduction program.
• Provide Technical Assistance to industry to help it improve
general housekeeping; recycle waste when appropriate, either
internally or through material exchanges; propose and provide
.-process modifications to reduce hazardous waste generation;
detoxify waste; and substitute non-hazardous for hazardous
materials.
• Create a Technology Transfer/Clearinghouse Data Base on
alternative technologies. It will be used by various industries
and trade groups to reduce hazardous waste generation.
• Administer a Matching Grant Program of up to $100,000 for
industry to support modification of existing equipment or pro-
cesses or to develop new technologies that minimize the gen-
eration of hazardous waste.'
• Encourage Waste Reduction by working through the Gover-
nor's office to solicit from industries a description of their
waste reduction initiatives. As a part of this effort, the annual
Governor's Innovative Waste Reduction Award for industries
was initiated this year.
• Conduct Research. For example, a study was conducted to
assess the feasibility of having a central recovery facility for
electroplating wastes in the Chicago area. One option from
this study that appeared promising was to develop local treat-
ment facilities that would serve a number of similar type gen-
erators in an area.7 Another project is underway to develop
in-plant treatment/destruction techniques for certain organic
wastes.
Although much waste reduction can be accomplished through
better housekeeping, recycling/reuse of materials, product substi-
tution and process modification, major long-term reductions of
waste often will require moderate to major process changes, the
use of add-on equipment and the development of new technol-
ogy. A long-term research and development program is needed
to help industry devise new non-hazardous substitute materials
and to evaluate the effectiveness of new equipment or techniques
prior to their use. Full implementation of these activities in Illi-
nois will be facilitated by the Center's new Hazardous Materials
Laboratory, which is now in the design phase and is scheduled to
be operational in 1989.

HWRIC'S VERY SMALL QUANTITY
GENERATOR PROGRAM
HWRIC considers working with very small quantity generators
an important component of its waste reduction plan for the state.
Although under HSWA the U.S. EPA recently has begun regu-
lating Small Quantity Generators (those who produce 220 to
2,200 Ib/mo of hazardous waste), most businesses that produce
less than 220 Ib/mo of hazardous waste still are unregulated.
Besides small businesses, unregulated generators include house-
holds, farms, high school laboratories, some hospitals, many
laboratories (research, industrial and others) and other institu-
tions that produce small quantities of hazardous waste. To date,
much of this material has gone into municipal landfills, into
public treatment works or septic tanks, into storm sewers or has
been poured onto the land. These practices have led to ground-
water contamination in some areas. At present, the scope and
magnitude of the potential problems posed by this group's cur-
rent waste disposal practices are unknown. Working with this
group offers IEPA an opportunity to teach people to properly
handle and dispose of hazardous wastes. The educational process
ultimately will help reduce the amount of hazardous waste these
businesses generate.
HWRIC presently is reviewing successful household hazardous
waste programs in other states and will help initiate collection
drives in Illinois. Specifically, the Center will: (1) initiate interest
among community groups; (2) help coordinate activities and pro-
vide technical assistance; (3) help with publicity; and (4) identify
sources of funding. Similar assistance will be provided to the
agricultural community. The educational component that is con-
current with these efforts is important in further reducing the
amount of hazardous waste generated by these groups.
In 1986, HWRIC cosponsored a project with the IEPA to in-
ventory hazardous chemicals found in high school laboratories.
The ultimate goal of the project is to conduct a one-time collec-
tion of unwanted chemicals for proper disposal. Beyond this are
the educational possibilities for students and teachers alike, who
learn from the study what materials are hazardous and how they
should be properly handled and disposed.
HWRIC's Industrial and Technical Assistance staff are con-
ducting a continuing program to assist Small Quantity Generators
(SQGs). In the process of giving seminars, talks and workshops
to various industries and trade groups, we also reach the very
small generators of hazardous waste (less than 220 Ib/mo). As
with the larger generators, we believe that waste reduction will be
an important component of the solution to this group's disposal
problems in the future.

ALTERNATIVES TO PRESENT
DISPOSAL METHODS
The Hazardous Waste Advisory Council was created by the
General Assembly in 1983 to serve as a body to review the imple-
mentation of the state's hazardous waste program. As part of its
mandate to review the state of the art of alternative technologies,
it developed a questionnaire that was sent to approximately
9,600 individuals in 5,800 industrial organizations. The question-
naire, which was distributed by the Illinois Manufacturers' Asso-
ciation and the Illinois State Chamber of Commerce, asked ques-
tions about waste streams, present disposal methods and any
alternatives being considered to these disposal methods. The ob-
jective was to determine what alternatives to landfilling were
being considered by companies in anticipation of the Jan. 1,
1987 Illinois restriction on land disposal of hazardous waste. The
returns were received and analyzed by HWRIC and included in
the Annual Report of the Illinois Hazardous Waste Advisory
Council.'
Responses were received from 203 companies (3.5% of those
solicited), representing 459 waste streams. However, only the 385
waste streams determined to be hazardous were used for this
analysis. Of these waste streams, about 91% were being disposed
of off-site. Of the 350 waste streams disposed of off-site, 130
(36%) were being incinerated, 97 (28%) were being landfilled,
81. (25%) were being recycled and 42 (11%) were being treated.
Identified alternatives under consideration included none (60%),
other treatment than that presently being used (14.5%), incin-
STATE PROGRAMS 403
-------
eration (13%), recycling (7.5%) and landfilling (4.9%). For waste
currently being landfilled, the alternatives being considered in-
cluded none (62.9%), incineration (16.5%), other treatment
(13.4%), recycling (4.1 %) and landfilling (3.1 %).
A few general statements can be made about the responses re-
viewed. First, many of the respondents did not understand the
questionnaire. For example, 67 waste streams listed (of the 459)
were non-hazardous. Second, the large number of respondents
who listed "none" under alternatives to landfilling and the even
larger number of companies contacted who did not respond at all
indicate at least one of the following conclusions:

• Companies feel they have no alternative to landfilling and
either believe they will get a variance or that the land ban will
not occur.
• Companies have not given much consideration to disposal
alternatives or to the implementing land ban; some are con-
fused about what the land ban will mean.
• Companies may not know what alternatives are available or
how to get this information.
• Companies are reluctant or do not wish to respond to a gov-
ernment agency.
• Companies have been saturated with questionnaires from gov-
ernment agencies and do not respond.

From this information, we have concluded that incineration is
the most commonly identified viable alternative to landfilling be-
ing considered for combustible hazardous wastes. Unfortunately,
waste reduction was not indicated as a viable alternative by those
companies that responded to the questionnaire.

PRESENT WASTE REDUCTION EFFORTS
IN THE STATE

Governor's Innovative Waste Reduction Award
Industries have been making efforts to reduce their waste and
are now required, by both the U.S. EPA and IEPA, to certify
that they are doing so. In an effort to learn more about indus-
tries' waste reduction accomplishments in the state, to recognize
the efforts that have been made and to encourage other industries
to increase their waste reduction activities, HWRIC worked
through the Governor's Office to create the Innovative Waste
Reduction Award. Industries throughout the state were solicited
and asked to apply for the award (Fig. 1).
Some 33 applications were received. Other companies called
and were interested in applying, but either did not have enough
Dapartaant of Enaroy and Natural Raaourcaa
Hazardous Maata Reaaarch and Information Cantar
IHHOVATIVE HASTE RIDUCTION AWARD
Kama of Company:

Addraaa:

SIC Coda:
Nuabar of Pull-Tima Evployaaa:

Typa of Maata Ganaratad!
Maata Raductlon Tachnology or Innovation Bain? Daacribad:
Estlaatad Amount (Parcantaga) of Waata Raducad:

Attach Mora datailad daacrlptlon of tachnology or Innovation for
waata raduction and submit to:
Dr. David L. Thomaa, Dlractor
Hazardous Waeta Raaaarch and Information Cantar
1608 Hoodflald Driva
Savoy, Illinois 61674
Figure 1
Application Form for Governor's Innovative
Waste Reduction Award
time to apply or could not get approval through their manage-
ment. It was clear that with a greater lead time, many more
companies would have responded.
Company subrnittals were looked at for a number of different
waste reduction and waste management strategies. These strate-
gies and the number of companies employing them include:
source segregation or separation (1), process modification (6),
chemical substitution or elimination (8), material recovery and
recycling (15), treatment (17), material exchange (2), replace-
ment of old and/or installation of modern process equipment
(4) and management strategies (17). The latter category includes
a number of the processes above, but more specifically refers to
corporate plans for waste reduction or introduction of non-toxic
processes. Many of the applicants had used two to five of the
above strategies. By their estimates, they had reduced their haz-
ardous waste production by 32 to 100%. The primary wastes be-
ing treated were solvents, degreasers and heavy metals.
Although there are those who do not consider waste treatment
a waste reduction technique, this was the most common method
mentioned. It generally included concentration or detoxification
treatment by the generator or a disposer (either at the generator
site or the disposal site). Material recovery and recycling was the
next most commonly used method of waste reduction. In some
cases, this was combined with process modification and treat-
ment; for example, a company used one waste to treat another
waste. Process modification (change in equipment or the way
chemicals are handled) and chemical substitution or elimination
were the next most frequent methods of hazardous waste re-
duction. In the latter case, some companies were able to replace a
hazardous material with one that did not produce a hazardous
waste.
Illinois Generators' Waste Minimization
Statements
In Illinois, generators of hazardous waste are required to sub-
mit annual reports to IEPA summarizing their hazardous waste
activities. This report must include basic information about the
company, the amounts and types of waste generated and the ul-
timate destination and disposal method for the wastes. Reports
generally are submitted early in the year for wastes generated dur-
ing the previous calendar year.
Beginning in 1986 (when data for calendar year 1985 were sub-
mitted), large quantity generators (greater than 2,200 Ib/mo) were
required to submit statements describing the steps taken to reduce
the amount of hazardous waste they generated.
We decided initially to examine a small quantity of the reports
to see if they contained potentially useful information on current
waste reduction practices. The first 116 reports, sorted by federal
generator number, were examined. Of those, 35 (30%) had no
statement, 23 (20%) had a statement merely repeating the re-
quired regulatory wording and 58 (50%) contained more explicit
information about waste reduction efforts.
Based on this information, the decision was made to examine
a larger number of randomly selected reports. Of approximate-
ly 1,300 generators required to submit waste minimization state-
ments, 21% or 275 were chosen at random by computer. A list of
summary data extracted from the annual reports is shown in Fig.
2. All of the information required on this list was assembled us-
ing IEPA computer data tapes, except for the items indicated on
Fig. 2.
Assembling these data required examination of each report.
Each waste minimization statement was placed into one or more
of the eight listed categories. Definitions of waste management
strategies are listed in Table 1.
The data were examined in several ways (Table 2). Additional-
ly, several waste streams mentioned in an IEPA memorandum
404
STATE PROGRAMS
-------
as being predominant wastes generated in Illinois were separated
out and examined individually.
Company Name_

Address

City
SIC Code_
State
zip_
•Haste Minimization statement Submitted? (Y/N)
• Is statement substantive?
1) Repeats Regulatory Hording 2) Brief statement (< 1 page)
3) Large Statement
Hastes (RCRA I)
Quantity(gal/mo)_
Man.Method(RCRA •)_
•Managed on-site? (Y/N)

Haste Management Firms
Name
Address_

City

Name
State Zip_
Address
•Haste Management Strategies
(circle one or more)

1.Source Segregation/sep.
2.Process Modification
3.Raw Material Substitution
4.Material Recovery & Recycling
5.Material Exchange
6.Treatment
7.New Process Equipment
8.Corporate strategies
City_
Zip_
•Information that could not be extracted directly form IEPA
computer tapes

Figure 2
Waste Minimization Information Summarized From
Generators Annual Reports to IEPA
Table 1
Waste Reduction and Waste Management Strategies
Source Segregation/Sep—Separation of hazardous and non-hazardous
constituents of a mixed waste stream into separate streams.
Process Modifications—A change in a process or a change in opera-
tional parameters of an existing process intended to reduce generation
of hazardous waste.
Raw Material Substitution—Substitution of ingredients for the purpose
of reducing hazardous waste generation.
Material Recovery & Recycling—Recovery or recycling of a material for
reuse in-house or by others, on-site or of f-site.
Material Exchange—Exchange of wastes with another company or in-
dividual for their use generally as a raw material.
Treatment—Treatment of a waste to either eliminate the exhibited haz-
ardous characteristic or separate and concentrate the hazardous con-
stituent in the waste, on-site or off-site.
New Process Equipment—Purchase of new equipment with at least the
partial intention of reducing waste generation.
Corporate Strategies—Statements of company policy, descriptions of
training plans for employees, descriptions of possible future actions, etc.
Fifty-nine percent of the generators examined responded with a
waste minimization statement. Of these, 6.5% simply repeated
the wording in the regulation and provided no documentation of
their waste reduction efforts. Of the 51% that did provide waste
minimization statements, most provided a brief explanation (less
than one page). This short report was not necessarily undesir-
able, however, since a simple explanation of actions taken (e.g.,
"We purchased and are using a solvent still") was often suffic-
iently descriptive.
Some companies gave quantified results on waste reduction,
but these numbers should be used with caution, as they may re-
flect changes in production levels or some other effect rather than
true waste reduction. Of the 162 companies that responded (out
of 275 generator reports examined), 36% used material recovery
and recycling, 32% used corporate strategies, 23% used process
modification, 14% used raw material substitution and 12% used
new process equipment. A similar group of strategies was used
by generators for specific waste streams F006 and D006-D008
(Table 2).
Extensive details and documentation of a particular company's
waste reduction efforts usually are not present in these reports,
although there are some exceptions. Since the material in these
reports is open for public inspection, inclusion of specific de-
tails of a waste reduction plan could provide an advantage to
any competitor who chose to examine these files.
Waste minimization statements included with the annual re-
ports also must be viewed in the context of the entire annual re-
port in order to be properly understood. They often refer to data
that are in other sections of the report or to the previous year's
report.

Table 2
Results of 275 Generator Reports Submitted to IEFA for
Calendar Year 1985
Survey
Sample
Total
F006
D006
D007
D008
NO.
275
26
5
23
20
No. Of
Responses Type
1
162
16
2
16
14
18
3
0
2
1
s of Response*
23 1
139 5
11 2
2 0
13 1
13 0
7
1
0
0
1
Waste Reduction and
Management Strategies**
2345678
37
4
2
6
5
22
1
0
3
0
59
5
1
6
5
2
0
0
1
0
12
5
0
2
2
20
3
1
3
2
53
5
0
7
5
*Note: Types of responses
1) Simply repeats regulatory certification wording
2) Substantive response of less than one page
3) Substantive response of greater than one page
"(see Table 1)

Usually, these types of data will yield general information
about large numbers of companies. More meaningful and reliable
data will only become available in the next several years as waste
reduction programs are more fully implemented by the regula-
tory agencies and the private sector.

DISCUSSION
The U.S. EPA2 listed a number of waste minimization tech-
niques:

• Recycling (On-Site/Off-Site)
-Use/Reuse
-Reclaim
• Source Reduction
-Product Substitution
-Source Control
—Good housekeeping practices
—Input material modification
—Technology modification
• Treatment

OTA3 concluded that there were five approaches industry
could take to reduce hazardous waste (OTA's definition of waste
reduction includes only in-plant practices):

• Change the raw materials of production
• Change production technology and equipment
• Improve production operations and procedures
• Recycle waste within the plant
• Redesign or reformulate endproducts

The U.S. EPA2 stated that the potential for future reduction
of waste appears to be significant, and that within the next 25 yr,
aggregate waste generation volume could be reduced an addi-
STATE PROGRAMS 405
-------
tional 15 to 30% by the extension of existing source control tech-
niques. OTA1 cited a Congressional Budget Office study that in-
dicated a total of 18% RCRA hazardous waste reduction nation-
wide might occur between 1983-1990. However, OTA stated that
there are few or no data on the extent of industrial waste reduc-
tion, and that it would be more useful to focus on a waste reduc-
tion goal, such as 10% annually.
OTA1 concluded that state programs will need to focus their
activities on waste reduction if it is to become a significant factor
in environmental protection at the state level and if these activ-
ities are to effectively prevent pollution. The U.S. EPAa stated
that its primary role was to support and encourage the states to
develop their own programs. HWRIC has recognized that one of
the best places to attack the state's hazardous waste problem is
at the source. It is clear that at both the Federal and state levels,
waste reduction will be given greater priority. Large quantity
generators are being asked to certify that waste is being reduced
and to provide an annual report to the IEPA detailing the waste
minimization programs. This policy has not yet been extended to
small quantity generators.
HWRIC is working not only with the regulated generators of
hazardous waste, but with very small (unregulated) generators
such as households and farms as well. It is with this latter group
that education and information dissemination can have the great-
est impact in waste reduction. However, even among the regu-
lated generators, we have found a definite need for information
and technology transfer.
In this paper, we examined three sets of data from Illinois in-
dustry to determine the waste reduction and management efforts
that were taking place. For those companies asked about which
alternatives to land disposal they were considering, most (96.5%)
did not respond; of those that did respond, 63% said "None."
Incineration appeared to be the major viable alternative to land
filling; unfortunately, few appeared to be considering waste re-
duction.
Of the 33 companies that submitted entries to be considered
for the Governor's Innovative Waste Reduction Awards, treat-
ment, management strategies and material recovery and recycling
were the dominant strategies used for waste reduction and waste
management. Based on generator reports to IEPA for 1985,
material recovery and recycling, corporate strategies and process
modification were most commonly used. Even for these reports,
however, only about 50% of the generators indicated that waste
reduction actually was occurring.
We conclude that technologies now are available for industries
to make substantial reductions in their generation of hazardous
waste. Most of the companies that applied for the Governor's
Innovative Waste Reduction Award were able to achieve over
50% waste reduction or waste minimization through convention-
al treatment, recycling and reuse, process modification and chem-
ical substitution or elimination. However, it appears that over
50% of the generators in the state have not yet begun serious
waste reduction and waste management efforts.
Although it is clear that some generators have reduced a con-
siderable amount of waste, it is not possible to determine what
percent reduction is being achieved. This conclusion is the result
of a rather poor response given to our voluntary survey, to un-
enforced waste reduction requirements of regulatory agencies and
to complicating economic factors. Only by monitoring waste pro-
duction over several years and by accounting for changes in the
associated industrial processes can a more accurate measure of
waste reduction be made.
As companies look more closely at the life cycle of chemicals
in their plants, including a mass balance or "cradle-to-grave"
analysis of the fate of materials, the processes they are using and
the products they are producing, waste reduction methods will
be employed more frequently and at an economic savings to the
company. This concept is, perhaps, becoming clearer to many
companies as they realize the true long-term costs of disposing
of their hazardous waste. The potential long-term liability and
cost for cleanup of land-disposed hazardous wastes are driving,
and will continue to drive, many companies to reduce their waste
and to find alternatives to land disposal. Information dissemi-
nation and education can help speed this process, and they should
be a priority of state and federal governments, trade groups and
others serving industry.

REFERENCES
1. Aspen System Corporation (Publisher), "Waste Reduction Confer-
ence Focuses on Financial, Regulatory Incentives," Hazardous Waste
Report, Rockville, NY, 1986,8-9.
2. U.S. EPA, Minimization of Hazardous Waste, Executive Summary
and Fact Sheet, Report to Congress, Washington, DC.
3. OTA, "Serious Reduction of Hazardous Waste for Pollution Pre-
vention and Industrial Efficiency," Congress of the United States,
Office of Technology Assessment, Washington, DC, 1986.
4. Geiser, K., "Source Reduction" in The Toxics Crisis; What the
State Should Do, J. Tryens, Ed., Conference on Alternative and
Local Policies, Washington, DC, 1983.
5. Thomas, D.L., Miller, C.D. and Kamin, J.M., "Hazardous Waste
Management," Government Infostructures, K.B. Levitan, Ed.,
Greenwood Press, Westport, CT, 1986 (in press); Chapter 14.
6. "HWRIC, Annual Report, May 1, 1985-April 30, 1986," Hazardous
Waste Research and Information Center, Department of Energy and
Natural Resources (HWRIC 86-008), 1986,106.
7. Huff, J.E. and Huff. L.L.. "Feasibility of a Central Recovery Facil-
ity for the Metal Finishing Industry in Cook County," ENR Con-
tract No. WR5,1986.
8. "Hazardous Waste Advisory Council, Annual Report," A Report to
Governor James R. Thompson and the 84th General Assembly, 1986.
406 STATE PROGRAMS
-------
The Effect of Federal and California Regulations on
Firms' Decisions to Minimize Waste in California
Barry Garelick
Julia Gartsef f
Versar Inc.
Springfield, Virginia
Michael Neely
Versar Inc.
Sacramento, California
ABSTRACT
Changes in both Federal and California regulations tend to
make disposal of hazardous wastes by various methods either
illegal or costly. The rising costs of both virgin solvents and the
disposal of solvent wastes have resulted in firms investigating
alternatives that reduce or eliminate waste generation as part of
their overall waste management strategy. The increasingly strin-
gent regulations also have resulted in firms investigating the feas-
ibility of either on-site or off-site recycling.
Although many laws and regulations are designed to encourage
alternatives to land disposal (including waste reduction), com-
pliance with regulations to control pollution of one medium may
result in non-compliance with pollution standards for other
media. For example, operation of some solvent waste treatment
and recovery equipment may result in increases in emissions of
solvents to the air. Also, the combustion of hazardous waste for
purposes of energy recovery can result in release of air pollutants.
The degree to which the other environmental regulations are en-
forced can have a significant effect on the decision which a com-
pany makes regarding waste reduction.

INTRODUCTION
Land disposal of hazardous wastes has been a popular waste
management option for many years. However, changes in both
Federal and California regulations tend to make land disposal of
hazardous wastes either illegal or costly. The rising costs of both
virgin materials and the disposal of wastes have resulted in firms
investigating alternatives that reduce or eliminate waste produc-
tion as part of their overall waste management strategy. The in-
creasingly stringent regulations also have resulted in firms inves-
tigating the feasibility of either on-site or off-site recycling.
The recent changes have caused businesses to seriously consid-
er these alternatives. Waste minimization refers to a reduction in
the volume and toxicity of either the waste generated or waste
that is subsequently treated, stored or land disposed. The primary
goal of waste reduction is to lower or eliminate waste generation
through material substitutions, process modifications or good
operating practices. Waste that is generated then can be man-
aged by a number of methods that reduce the volume and toxicity
of wastes that are land disposed. One method is recycling; a waste
is either used or reused directly without processing, or it is re-
claimed.
The Federal and California hazardous waste regulations may
contain disincentives for practicing certain waste reduction op-
tions. For example, a facility owner who installs equipment that
could reduce the generation of a solid waste may need to obtain
a hazardous waste permit.
This paper explores the interplay between legislation, regula-
tions and policy. It addresses environmental problems associated
with the phase-out of land disposal, including the possibility of a
treatment and incineration capacity shortage. The degree to
which the public and governmental agencies are willing to make
sacrifices in other areas of the environment will determine the
success of many new and innovative programs to encourage waste
reduction.

EFFECT OF REGULATIONS AND
PROGRAMS ON FIRMS' DECISIONS
The Hazardous and Solid Waste Amendments of 1984
(HSWA) establish as a goal and national policy the minimization
of hazardous waste generation and land disposal. In addition,
HSWA requires land disposal restrictions for certain wastes, in-
creased requirements for landfills and corrective action for "prior
releases." These factors act together as an impetus for change by
companies that otherwise may not have considered waste min-
imization methods and techniques as an alternative. In particu-
lar, generators now are required to certify on their manifests that
a program is in place to reduce the volume and toxicity of waste
generated and/or land disposed.
California's regulations also may serve as a basis for change.
They are, in general, more stringent than Federal requirements.
Phase-out of land disposal for example was enacted by Execu-
tive Order in late 1981, 3 yr before Congress enacted HSWA.
Under California's regulations, certain wastes are deemed "re-
cyclable." Generators must recycle these wastes or must provide
written justification for not recycling them.
The agency that enacts and implements hazardous waste regu-
lations in California is the Department of Health Services (DHS).
DHS has initiated a four-phase waste reduction program as part
of its hazardous waste management activities that consists of:
(1) technical assistance, (2) economic incentives, (3) information
transfer and (4) regulatory incentives. In the latter category, DHS
recently modified its regulations to operation of hazardous waste
facilities in an effort to promote recycling. These modifications
include changing the title of facilities that recycle hazardous waste
from "hazardous waste facility" to "resource recovery facility"
and simplifying permit requirements for facilities that recycle
non-RCRA wastes.

STATE PROGRAMS 407
-------
In addition to hazardous waste regulations, Federal and Cali-
fornia air regulations also have an important role in firm's de-
cisions to employ waste management techniques. The federal air
permit program is concerned primarily with pollutants for which
the National Ambient Air Quality Standards (NAAQS) are not
violated. These are called "attainment" pollutants. "Non-attain-
ment" pollutants come under the jurisdiction of California's air
program. Air permits are granted through the authority of air
pollution control districts. The critical requirement for non-
attainment pollutants is that emission increases of such pollu-
tants over specific amounts (usually ISO Ib/day) must be miti-
gated by emission reductions of the same pollutant. The reduction
is referred to as an emission "offset" and must be shown to result
in a net air quality benefit. Offsets are obtained by creating emis-
sion reductions from existing equipment that the permit applicant
currently owns or by reducing emissions from another facility.
In the latter case, the applicant would have to arrange a transac-
tion with the owner of the other facility.
The remainder of this section explores how the various haz-
ardous waste and air programs may affect firms' decisions (pos-
itively and negatively) to reduce wastes. It also examines how the
regulations may conflict with or complement each other and the
possible effects of that interplay.

Waste Minimization and the Fear of Regulation
Although generators must certify on their manifests that they
are minimising waste (per HSWA 1984), the U.S. EPA does not
have the authority to ascertain that individual waste minimization
programs are in place or whether they qualify as such. Determina-
tions of "economically practicable" and "practicable method
currently available" are made by the generator, not the U.S.
EPA (50 CFR 28734).
This provision of the HSWA may have several effects upon
generators. One is that generators may be more aware of waste
minimization, and the act of certifying by itself may become an
incentive for generators to make an effort to reduce volume and
toxicity of wastes. Another possible effect is that companies
may do only the minimum to certify that they have instituted a
waste minimization program. Without U.S. EPA monitoring and
enforcement, the quality of waste minimization programs may be
dependent on what generators fear may happen in the future.
Some companies feel that it may be only a matter of time be-
fore waste minimization is required as a U.S. EPA regulation
rather than in its current form. Thus, fear of future regulation
may act as an incentive for some generators to undertake waste
management programs involving some degree of innovation.
Companies often feel that they know the most economical and
feasible alternative, and they would rather make that decision
than be told to conform to an inefficient and costly standard.
They may, therefore, be motivated to take the first step. If waste
minimization regulations were enacted, the company that had
done as much as it could might be in a better position than the
company that had done only the minimum. Those who have com-
plied with the law may feel that they have done all that is possible
to minimize their waste and do not need to do more.

Waste Minimization and the Fear of Liability
Under the CERCLA, generators may be held liable for dam-
ages from subsequent treatment, storage or disposal of their
wastes. Generators that do not know the reliability of recyclers
may fear future costs that they may incur for damages due to
subsequent recycler handling of their wastes.
Liability issues and the cost of liability insurance also will affect
the owners of land disposal facilities. Because the owners of land-
fills must demonstrate financial responsibility under RCRA, the
unit price of land disposal is likely to increase because of liabil-
ity, creating another incentive for generators to consider the al-
ternatives.
The risk of future liability resulting from disposal of hazardous
waste may serve as an incentive for instituting on-site waste min-
imization practices. By reducing the volume of waste shipped
off-site, there is a corresponding reduction in risk. Fear of lia-
bility is most likely to be an incentive to reduce waste for compa-
nies that have access to capital and/or on-site technical expertise.
Other companies that must ship wastes off-site are left in a
quandary by the liability issue. To the extent that they can, and
desire to do so, they will reduce the volume of wastes generated.
For wastes that are generated, however, the lack of capital and
expertise leaves off-site waste management as the only option for
these companies. Larger companies can spend the time to pre-
qualify their off-site recyclers and regularly audit the recycling
facilities. Smaller firms do not have the means for such prequal-
ification and followup audits, and thus would have to rely on the
work of others.
When no on-site waste management is possible, the issue of lia-
bility can make off-site recycling preferable to off-site landfilling
because off-site land disposal offers the potential for improper
landfill designing and also does not reduce the volume of dis-
posed waste. Off-site recycling, on the other hand, provides an
opportunity for a reduction in the volume of land disposed waste.
California's regulations may help to relieve some of the fear of
liability associated with off-site recycling because off-site recyc-
lers are required to obtain permits. Under the Federal program,
and in many states that have adopted the Federal regulations,
permits are not required for the recycling activity itself. Al-
though permitting a facility provides no assurance that future
damages will not occur (just as there is no guarantee that a per-
mitted landfill will never leak), the recycler permit system pro-
vides some knowledge of the procedures that are being followed;
recyclers are subject to inspection and must comply with oper-
ating conditions specified in the permit.
Liability concerns may affect the degree to which recycled ma-
terials are used or purchased in lieu of virgin materials. Such con-
cerns are likely to result in a preference for the use of raw or vir-
gin materials in processes, as opposed to waste products serving
as substitutes, especially those that must be reclaimed prior to
use. Under CERCLA, transporters of wastes can be held liable
for future damages from wastes transported. The additional cost
of insuring transportation of wastes for reuse could offset the sav-
ings in raw material costs.

Economic Issues
Prior to the 1980s, waste reduction was used as a means to re-
duce manufacturing costs through the maximization of produce
yields and operating efficiency. Although Federal and California
hazardous waste regulations introduce the secondary goal of de-
creasing the amount of waste produced and managed, a company
decision to minimize waste is based on profit and risk. Even the
driving force behind the fear of liability is the fear of the cost of
future damages. The fear of future regulations also is economical-
ly based. Fear of liability and regulations are, therefore, indirect
economic issues.
Other more immediate economic issues are explored in this sec-
tion. These issues center around the costs of waste management
alternatives. While there is reason to believe that landfilling may
become less desirable for wastes that are still permitted to be land
disposed, other factors must be considered. In the case of off-
site recycling, for example, the distance to a recycling or treat-
ment facility and the resulting costs of transportation are im-
portant. In some cases, recyclers or treaters will not accept
408 STATE PROGRAMS
-------
amounts below a minimum volume. For off-site recycling to be
cost effective, sufficient volumes must be accumulated prior to
shipping off-site. This may be a problem for some generators
since storage for more than 90 days may require a hazardous
waste storage permit. For some generators, landfilling (where
allowed) may be a less costly and more attractive waste manage-
ment alternative to recycling or treatment. Some generators have
initiated recycling programs by consolidating their wastes, thus
creating enough volume to make recycling economical. The suc-
cess of such cooperative arrangements is dependent on the density
of similar industrial operations in an area serviced by a commer-
cial recycling firm.
Other economic factors affect the decision to reduce wastes by
material substitution, that is the use of different material in a pro-
cess or of a recycled substance in lieu of virgin material. Product
quality is a critical economic consideration for such investments.
If the substitution of materials will cause lower production costs
or improve product quality, the firm has a definite incentive to
invest. On the other hand, if product quality is sacrificed as a re-
sult of material substitution, firms are not likely to make such in-
vestments. For example, manufacturers of electronic equipment
(e.g., printed circuit boards) require a high degree of purity in
their solvents. Many choose to use virgin solvent rather than re-
cycled material. Although the cost of solvent purified and re-
cycled on-site could be less than for virgin materials, the risk of
inferior product quality represents a potential loss in profits. The
company would conclude that the risk of losses outweighs the
cost savings resulting from the use of recycled solvents.

Regulatory Gridlock
The regulations that govern environmental protection are writ-
ten by a large number of people within different sections of the
same agencies or by different agencies entirely. Each person has
his own perspective and objectives. In some instances, these ob-
jectives may conflict, resulting in a "gridlock" as perceived by
the regulated community. Similar conflicts sometimes arise be-
tween Federal and state agencies and local communities. The re-
sult is a cumbersome assemblage of regulations and procedures
that can adversely affect the decisions that businesses make with
regard to waste management alternatives. Potential regulatory
gridlock situations related to the phasing out of land disposal and
the permitting and siting of alternative treatment facilities are dis-
cussed below.

Land Disposal Bans and Capacity Shortage
The land disposal bans promulgated by the U.S. EPA on Nov.
7, 1986 may cause short-term problems with disposal of wastes
from recycling facilities. In particular, the Federal regulations ban
some of the RCRA F-code wastes which include still bottoms
from solvent reclamation.
The banning of solvent reclamation still bottoms from land dis-
posal will have an effect on some California recycling facilities.
The initial effect most likely will be an increase in the amount of
waste handled at recycling, treatment and incineration facilities.
This increase in solvent recycling will result in an increase in vol-
umes of still bottom residues generated by such facilities. Those
residues are subject to the land ban. The increase in volume of
still bottom residues that cannot be land disposed, coupled with
an increase in the amount of other banned wastes incinerated,
may result in a short-term shortage of incinerator capacity.
Commercial recycling facilities, faced with an incinerator
capacity shortage, may be forced to reject solvent wastes for re-
covery. Although waivers up to 2 yr may be granted in cases of
insufficient capacity, it is not known if additional capacity can be
achieved in that time.
Permitting
The problem of short-term capacity shortages discussed above
could be alleviated by expanding the capacities of existing treat-
ment and incineration facilities. These industries are concerned,
however, with the long and sometimes unpredictable delays asso-
ciated with multiple permits.
Incinerators and treatment facilities are required to obtain per-
mits under California's hazardous waste regulations. In addition,
such facilities also may need air permits. The granting of a haz-
ardous waste permit for incinerators or treatment facilities re-
quires careful review of the permit application. The application
must contain details of the facility's design, and incinerators must
have successfully completed a "test burn." Both types of permits
involve a significant investment of tune and money by the permit
applicant and of time and resources by the permitting agency.
Additional capacity is not likely to be achieved at the rate needed
to accommodate the increasing volume of waste to be treated.
Air permits also present special problems. Emission offsets are
required, as described above. The requirement for offsets is a sig-
nificant barrier because, in areas such as Los Angeles or the San
Francisco Bay Area, offsets are difficult to obtain unless a com-
pany already has facilities in place. The permit requirements thus
may prevent the installation of waste reduction equipment that
emits air pollutants.
Although the capacity shortage situation could provide an in-
centive for generators to use waste reduction technology, similar
permit requirements for generator treatment facilities may be-
come a disincentive for them as well.

Siting
Obtaining the proper permits for new facilities (recycling, in-
cineration and treatment) will be dependent on a siting process
which involves the public (especially the community in which the
facility would be located) and all governmental decision-making
agencies. This process is provided through the California En-
vironmental Quality Act (CEQA). Generally, CEQA directs that
these parties be informed of the potentially significant environ-
mental consequences of a proposed project, prior to the granting
of permits and approvals by the governmental agencies. Any time
a governmental permit is required (such as a county land permit,
air pollution control district permit or other state or county ap-
proval), CEQA requires assessment of the possible positive and
negative environmental effects of the project.
If the governmental agencies and/or the public consider en-
vironmental effects to be significant, then CEQA requires one of
the governmental agencies (termed the "lead agency") to prepare
an Environmental Impact Report (EIR). The EIR is open to pub-
lic review, and public hearings are held to discuss the project.
Based on the response of the public and governmental agencies,
a decision is made as to whether the project should be permitted
and the type of conditions to which the facility should be subject.
Historically, waste treatment facility siting has been controver-
sial, with most opposition originating from the local community.
The siting problem is the familiar one of "not in my backyard,"
and there is no guarantee that the project will be accepted. Pro-
ject review may go beyond the EIR stage and be subject to a com-
munity referendum.

Definition of Solid Waste
The U.S. EPA published a revised version of the definition of
solid waste in the Jan. 4, 1985 Federal Register. The revised
definition introduces new tests by which a substance may be
deemed to be: (1) a solid and hazardous waste; (2) legitimately
recycled; and (3) subject to regulation under RCRA. Unlike the
previous definition which exempted from regulation certain
STATE PROGRAMS 409
-------
wastes that are recycled in any manner, the revised definition
asserts that RCRA jurisdiction is determined by what the material
is and how it is being recycled.
California's regulations and laws recently have been modified
to embody the general concept of this definition. In general, how-
ever, the definition results in certain requirements which apply to
solid and hazardous wastes management.
A waste that is reclaimed prior to being recycled usually is
deemed to be a solid waste under the definition and subject to
regulation. If, however, it is used directly as a feedstock or ingred-
ient in a process or is used as an effective substitute for a commer-
cial chemical product (e.g., a solvent waste from a production
process used for degreasing), it is not deemed to be a solid waste
and is not subject to RCRA regulation.
California'r hazardous waste regulations, unlike Federal regu-
lations, require that recycling activities be regulated. Because of
this, the U.S. EPA's definition of solid waste has had little effect
upon recycling operations in the state. In addition, California's
regulations allow recyclers' operating permits to be called "Re-
source Recovery Permits" rather than hazardous waste treatment
facility permits, thus removing some degree of stigma associated
with such operations.
California requires, however, that end-users of material obtain
Class C Resource Recovery Permits. This requirement may be a
built-in disincentive in California's recycling regulations. Given a
choice between using a virgin material that does not require a per-
mit and waste material that does, companies may choose the ma-
terial that does not require a permit.

IMPLICATIONS FOR FUTURE POLICY
As shown in the preceding discussion, the various environ-
mental regulations and laws can both directly and indirectly affect
firms' decisions to reduce solvent and other types of wastes. In
some cases, there is a direct incentive to reduce waste or to seek
alternatives to landfills because of a reduction in operating and
production costs as well as a reduction in liability. Because of the
multiple objections of the various environmental agencies, how-
ever, there may be conflicts that can create obstacles and disin-
centives to employ such practices. The general public is concerned
with the effect of projects in their community; if they believe
(correctly or incorrectly) that their health will be threatened, the
people in the community will oppose the project. The public's
perceptions about various projects "builds in" a fear and also can
cause reported values in a community to diminish.
Coordinating goals and objectives among regulatory agencies
is an important step toward resolving "regulatory gridlock"
issues. For example, although an air pollution control district
can allow emission reduction of a pollutant to "offset" increases
of the same pollutant from a new source, it does not have the
authority to allow a decrease in hazardous waste generated to
offset air pollution emission increases. The air districts could,
however, plan for increases in certain types of air pollution
sources by building in reductions elsewhere. The degree to which
this is done is dependent on the importance of the industry growth
for which allowance is being made. Thus, in the case of solvent
treatment equipment, there will need to be a determination of the
importance of air emission sources such as air or steam strippers
used for waste reduction. Similarly, projects which combust haz-
ardous waste for energy recovery may be given special attention.
This type of consideration has been given to some extent for
refuse-fired power plants and cogeneration facilities. To what ex-
tent these considerations can be extended to hazardous waste
combustion is a matter of negotiation and social acceptance.
To succeed, future policies of the Federal, state and local en-
vironmental agencies will need an integrated approach. There is
unavoidable risk involved. If contamination of soil and ground-
water is to be reduced, there is apt to be some increase in pollu-
tion of other media. The problem facing agencies is to decide
how the risk should be minimized and distributed.
California already has taken steps to alleviate some of the
"roadblocks" to projects that help to reduce the land disposal of
hazardous waste. Labeling facilities as "Resource Recovery Facil-
ities" rather than hazardous waste treatment facilities alleviates
some of the problems associated with siting. This also could
create an incentive for the construction of incineration facilities
that recover energy in order to qualify as a "resource Recovery
Facility."
Changing the name of the type of facility is only a first step,
however. Successful siting efforts generally occur where there is
not only consistent government leadership, but also effective pub-
lic education and participation.
The California legislature has introduced bills that provide loan
guarantees, tax credits and limited liability to encourage innova-
tion in waste management. Enacting these bills will help provide
demonstrations of technical and economic feasibility. The incen-
tives must be complemented by regulations and policies that allow
the innovations to evolve with minimal administrative costs and
delays.
Advantageous policy decisions also may be made by develop-
ing regulations which avoid some of the disincentives discussed
here. The existing laws and regulations can be revised with re-
spect to questions of equity and unintended results. Some of the
inequities and biases in regulations may. in fact, stem from the
statutes. (For example, a transporter carrying virgin trichloro-
ethane for use in manufacturing is subject to neither the same
degree of manifesting requirements nor liability for future dam-
ages as a transporter carrying spent trichloroethane to a solvent
recovery facility for reclaiming.)
The appropriate environmental balance is not an easy goal to
achieve. The U.S. Congress and the State of California law-
makers have shown concern over the landfilling of hazardous
wastes and have taken major actions to phase out the practice.
Achieving this goal will require additional risks and environ-
mental sacrifices and an integrated approach toward future regu-
lation, legislation and policies that implement these changes.
Without such integration, deadlocks may continue and result in
capacity shortages and illegal disposal.
410 STATE PROGRAMS
-------
Siting Hazardous Waste Facilities in New York State

Peter A. Marini, P.E.
NYS Environmental Facilities Corporation
Albany, New York
ABSTRACT
Over the past few years, the State of New York has attempted
to define its needs for proper commercially available hazardous
waste management facilities. Recent state studies' support the
need to construct commercially available treatment facilities that
provide a higher degree of waste treatment than basic treatment
and land-burial. The state and Federal phased reduction of land-
burial of certain hazardous waste creates an even greater need for
new facilities. The use of a higher degree of technology, recovery,
reuse and maximum treatment of hazardous wastes in New York
would require future construction of the following commercial
.faculties:
• A 40,000-ton/yr incinerator to handle liquids, solids and
sludges
• An 11,000-ton/yr acid recovery facility to recover sulfuric and
hydrochloric acids
• A 21,000-ton/yr hydrometallurgical recovery facility with ion
exchange to recover heavy metals from sludges
• A 30,000-ton/yr advanced thermal destruction facility
• A 26,000-ton/yr landfill with stabilization capacity
• An oil re-refinement facility to convert waste oils into usable
multipurpose re-refined oils; it is assumed that waste oils will be
designated hazardous waste in the future.
At present, the major challenge in constructing these facilities
is gaining public confidence and government approval. Locating
the best potential site is a challenging and expensive undertaking.
Public opposition to these facilities remains the primary siting
obstacle.
Gaining approval for a site to handle hazardous wastes is as
challenging as finding the best private technology and operating
expertise. With public financial and regulatory incentives, the
private sector can successfully develop the commercially avail-
able higher technology hazardous waste treatment facilities
needed to protect New York's environment.
This paper will examine the need for commercially available
hazardous waste treatment facilities, the current best available
technology, a siting process and administrative procedure to
secure private sector technology and operating expertise.

INTRODUCTION
Over the next 15 yr, the Federal Superfund Amendment and
Reauthorization of 1986 and New York State Superfund Law
along with the passage of the Environmental Bond Act of Novem-
ber 1986 will create funds which will be used to investigate and
cleanup an estimated 1,400 inactive waste sites in New York State.
It is expected that a total of $4.3 billion will be spent by respon-
sible parties, the state Superfund and the Federal Superfund to
COMPLETION SCHEDULE FOR
REMEDIAL PROJECTS FUNDED BY
FEDERAL SUPERFUND
,% ^
'*
/•" CONSTRUCTION
FISCAL YEAR

(flpril 1 - M.ircli 31)
Figure 1
Completion Schedule for Remedial Projects Funded by
Federal Superfund
COMPLETION SCHEDULE FOR
REMEDIAL PROJECTS
FUNDED BY STATE SUPERFUND/BOND ACT

CONSTRUCTION
9»ia7 B8(0» 10/91 92103 9419* 9*197 90(99
FISCAL YEAR
(ftpril I . March HI
Figure 2
Completion Schedule for Remedial Projects Funded by
State Superfund/Bond Act
STATE PROGRAMS 411
-------
COMPLETION SCHEDULE FOR
REMEDIAL PROJECTS FUNDED BY
RESPONSIBLE PARTY
I I I I I I I I I I I I I I
IMtm Milt •*'•' Hill M'tl Mrlf M'M
FISCAL VIAM (*».i i . *•.<• pi

Figure 3
Completion Schedule for Remedial Project! Funded by
Responsible Party

complete this cleanup program. The rate of completion of de-
sign and construction stages of New York State/Bond Act, Fed-
eral and Responsible Party Remedial Projects is illustrated in
Fig. 1 through 3.
The form that remedial work can take varies considerably from
site to site. For a sample site where wastes are contained in drums,
remediation may be simply the complete removal of those drums.
The wastes stored in the drums would be treated or disposed
elsewhere. For other sites, however, remediation may be a com-
bination of waste removal, containing contaminants which can-
not be removed, capping the site to preclude the percolation of
rainwater and pumping leachate or groundwater within the sys-
tem to further control chemical migration.
At present, there is a maximum reliance on land filling in New
York State. However, on Apr. 30, 1984, the State of New York
initiated a phased reduction of the land burial of certain haz-
ardous organic wastes. On Nov. 8. 1986, the U.S. EPA imple-
mented a requirement of the 1984 RCRA amendments which pro-
hibits the landfilling of spent solvent and dioxin-containing
wastes. For concentrated spent solvents, the prescribed treatment
level will be incineration. The purpose of this phased reduction
was to take advantage of higher technology management capabili-
ties for hazardous waste treatment and disposal for the future.
A minimum landfill utilization policy with a maximum high
technology treatment policy must be implemented to protect New
York's environment. Advanced waste treatment technologies
should be used to destroy, treat and solidify or otherwise neutral-
ize hazardous wastes from inactive hazardous waste sites and ex-
isting industrial and manufacturing operations. These advanced
waste treatment methods are not available at the present time.
Hopefully, they will be in the future.
Recently, SCA Chemical Services, Inc. of Model City, New
York, received the necessary permit to construct and operate a
facility for the thermal destruction of whole PCB capacitators.
The PCB destruction efficiency is expected to exceed 99.9999%.
The planning and construction of these advanced technology
treatment facilities must begin soon to have them commercially
available for the Federal and state Superfund cleanup programs.
At the same time as the Superfund program is being imple-
mented, the state hazardous wastes enforcement program will be
fully implemented. Enormous quantities of hazardous waste will
have to be properly treated and managed within the state.
This paper will demonstrate the immediate need for hazardous
waste treatment facilities capable of providing high technology
for the destruction, treatment and management of hazardous
wastes. The needs for higher technology facilities, the process for
gaining approval of sites and the procedures for securing private
sector operating expertise are discussed.
The needs for new hazardous waste treatment facilities in New
York can be demonstrated best by examining the shortfall be-
tween "supply" and "demand", with demand being the future
quantity of waste expected to be generated in the State of New
York to be managed off-site.
It is assumed that essentially all existing commercial facilities
in New York will continue to operate and receive the same quan-
tity of waste imported from other states in 1988 as in 1984.
"Demand" is equal to the 1988 forecasts of manifested waste
presented in Fig. 4. Data in Fig. 4 illustrate all waste expected to
be manifested in 1988 that will serve as demand for New York
faculties. Although this analysis overstates actual demand be-
cause the exporting of some waste to facilities in other states will
continue in the future, it reflects a goal not to place undue reli-
ance on facilities in other states.
The following stacked bar chart best describes and documents
the latest inventory of hazardous waste in the State of New York
by waste categories.
HAZARDOUS WASTE OCHfRATIOM
100-
YEAR
ea
S3
BB
ES
Other Li

Or|iflic Solirfl 1.4 Sludf»

Inorganic Soli«i •** Slwdg*t

SolUi »< SI »li» i
Figure 4
Hazardous Waste Generation
As can be seen in Fig. 4, approximately 280,000 tons of a vari-
ety of hazardous wastes were generated in New York in 1984.
The various management methods for the total hazardous waste
stream are illustrated in Fig. 5. As can be seen from the pie chart,
the aqueous treatment and landfill method are the two most pop-
ular forms of waste treatment in New York. The aqueous treat-
ment technology provides for biological, chemical treatment and
separation, all of which can be described as simplistic treatment
412 STATE PROGRAMS
-------
technologies. Neither the landfill or aqueous treatment technol-
ogy represents higher technology methods for the management of
hazardous wastes.
undel.rmlnodl1o.4XI
Iandfllll34.3*l-
eneroy recoveryls.0%1
aqueous
— Irealmentl37.0%l
solvent recoveryl4.3%l

fuel blendlngl4.3XI
lncJnere.tlonl4.8XI
Figure 5
Hazardous Waste Management Methods
Table 1 reflects the shortfall in commercial facility capacity in
thousands of tons per year by management technology.
Table 1
Shortfall in Commercial Facility Capacity in the State of New York
Management
Requirenents-
Hsnagement Maximum
RECOVERY TECHNOLOGIES:
Solvent Recovery
Oil Recovery
Acid Recovery
Hydronetallurigal Recovery
Fuel Blending
Energy Recovery
Metals Recovery
THERMAL DESTRUCTION TECHNOLOGIES:
Incineration - Liquids
Incineration - Solids
Wet Air Oxidation
Advanced Thermal Destruction
CHEMICAL. PHYSICAL AND BIOLOGICAL
TREATMENT TECHNOLOGIES:
Conventional Aqueous Treatsient
- Cbemical Oxidation/Reduction
- Neutralization/Precipitation
- Solids/Liquid Separation
Oil/Water Phase Separation
Ion Exchange
Total Conventional Aqueous
Treatment
Chemical Dehalogenation
Conventional Biotreatment
LANDPIU. DISPOSAL TECHNOLOGIES:
Secure Landfill

11,400
3.500
11 , 300
7,600
8,400
17,700

24,900
14,900
17,700
30,000

0
73,000
18,000
2,000
8,300

101,000
1,200
46,000

2,000
'
Available
Facility

64,000
230,000
—
—
220,000
3,300

0
0

625.000
6,600
—

'
Net Available
or (Shortfall)
Capacity Maximum
Treatment

55,000
229,800
(12,200)
(15,300)
219,000
(1,400)

(55,200)
(18,200)
(28,400)

N/A

5,300
(46,800)

(45,400)

Source: New York State Department of Environmental Conservation, Hazardous Waste Treat-
ment Facilities TaskForce, Final Report, September 1986.

The energy recovery, solvent recovery, fuel blending and incin-
eration management higher technology methods account for a
total of only 19.4% of the total waste stream.
Assuming the need for high waste reduction and maximum
treatment technology (minimum landfill utilization), the major
hazardous, waste facility needs in the State of New York in 1988
would include:
• A 40,000-ton/yr incinerator that can handle liquids, solids and
sludges
• An 11,000-ton/yr acid recovery facility to recover sulfuric
and hydrochloric acids from spent pickle liquors
• A 21,000-ton/yr hydrometallurgical recovery facility with ion
exchange, as well as other processes, to recover chromium,
nickel, copper and zinc (and cadmium combined with the zinc)
from heavy metal sludges and hex-chrome wastes
• An 18,000-ton/yr wet air oxidation facility
• A 30,000-ton/yr of advanced thermal destruction capacity
• A 26,000-ton/yr stabilization capacity
• A 10,000,000 gal oil re-refinement facility to convert waste
oils to usable multipurpose refined oils; it is assumed that waste
oils will be designated hazardous waste in the future.
The estimated total costs to construct all facility needs over the
next 5 to 15 yr could be approximately $78 million.

FACILITY ASSESSMENT AND NEEDS
Incinerator
There is no commercial incineration capacity in New York for
the disposal of hazardous waste solids. There are, however, two
existing industrial rotary kilns (one producing aggregate, and the
other cement) that presently burn liquid hazardous wastes as sub-
stitutes for conventional fuel. There also are 12 generator-owned
hazardous waste incinerators in New York featuring liquid injec-
tion, rotary kiln, fluidized bed and single chamber thermal oxida-
tion processes. There is no indication that any company intends to
offer its incinerator for commercial use in the future.
A minimum of one 40,000-ton/yr rotary kiln incinerator to
burn liquids, solids, sludges and drums should be built. A sche-
matic diagram and description of the rotary kiln incinerator can
be found in Fig. 6. The estimated current capital cost for the in-
cinerator is approximately $20,000,000.

Acid Recovery
There is no commercial acid recovery facility in the State of
New York. The 600,000-ton/yr of aqueous treatment capacity in
New York represents simplistic treatment technologies such as
biological and chemical separation and treatment.
A minimum of one 11,000-ton/yr acid recovery facility is
needed to recover waste sulfuric and hydrochloric acid for re-
use. A schematic diagram of the recovery system can be found in
Fig. 7. The estimated current capital cost for the acid recovery
facility is approximately $3,000,000.

Metal Recovery Facility
There is no commercial metal recovery facility in the State of
New York.
A minimum of one 21,000-ton/yr hydrometallurgical recovery
facility with ion exchange, to recover chromium, nickel, copper
and zinc from sludges and wastes, should be built. A schematic
diagram of the recovery system can be found in Fig. 8. The esti-
mated capital cost for the metal recovery facility is approximately
$2,000,000.

Wet Air Oxidation Facility
There is no commercial wet air oxidation facility in the State of
New York.
A minimum of one 18,000-ton/yr wet air oxidation facility
should be built. A schematic diagram of the wet air oxidation fa-
cility can be found in Fig. 9. The estimated current capital cost
for the wet air oxidation facility is $3,000,000.

Advanced Thermal Destruction Facility
There is no commercial advanced thermal destruction facility
in the State of New York. The issuance of a permit, in Nov. 1986,
to construct and operate the SCA Chemical Services Inc.'s arc
pyrolysis project will be the first advanced thermal destruction
facility to be built.
A minimum of one 30,000-ton/yr advanced thermal destruc-
tion facility should be built. A schematic diagram of the advanced
thermal destruction facility can be found in Fig. 10. The esti-
STATE PROGRAMS 413
-------
FrMl)n« itir WMU. Into feniUf llM «••*• CM»H| lh« MM* •"* TrvMdtf llM Hw «»*.

•wMt •«! n.* iMtuprttr ih» man k4* «dl pn*m a«M (M* ihr .M* hm r««*«*t
« • ill ht M mm ihr I «*TMT •( iMm **•«**• n» r«m*.nin| |*m (llM kHb* MH« tW HIM §H
4 Ml»»v »«w* it*
MtaUMf*»IWIftlto rmvtrrtaln »h.* .Jl BMM •« mw iht vn
PMHI fchrrt ««anA*n I* M (^ ih, n«* ftMi i* i^ui •« •!•*« llw *
-------
Figure 8
Metal Recovery Facility
Figure 9
Wet Air Oxidation Facility
Figure 10
Advanced Thermal Destruction Facility
continue to press industry to implement and install more ad-
vanced waste management practices and technologies. The state
should leave the initiative for locating, proposing, developing
and managing treatment, storage and disposal facilities primarily
with private industry responding to economic and market forces.
DcmaaUlunt
liiumfiliau Sloragf
Lube* Fuel
fueUoMUoa
Hydmutaint
Figure 11
Oil Re-Refinery

However, locating the best potential site for hazardous waste
management facilities is a challenging and expensive undertak-
ing. Public opposition to siting hazardous waste facilities in New
York (or anywhere else) continues to be a difficult undertaking.
The threat of intervenor litigation in the State Environmental
Quality Review process and after the issuance of the necessary
permits discourages private applicant interest in these facilities.
However, the recent decision of the New York State Siting Board
for the application of the SCA Chemical Services Inc.'s arc pyro-
lysis project, indicates that the process can work expeditiously.
New York State Department of Environmental Conservation
(NYSDEC) Regulation Part 361 stipulates the provisions which
apply to the siting of new hazardous waste management facilities.
A new hazardous waste treatment facility in the State of New
York may have to meet and obtain the following requirements
and permits:

• Hazardous Waste Treatment, Storage and Disposal Facility
Permit
• Air Emission Permit
• State Environmental Quality Review Process
• Certificate of Environmental Safety and Public Necessity

The State Environmental Quality Review process is a review
process for all projects that go before New York State public
agencies that require an approval or permit. The Certificate of
Environmental Safety and Public Necessity consists of submitting

STATE PROGRAMS 415
-------
the hazardous waste and air permits, the Draft Environmental
Impact Statement (DEIS) and certain additional information
about the facility and its public necessity. The procedures and
time frames for each step of the application process and public
hearings are governed by the Uniform Procedures Act.
After thorough review of the permits and certificate by the
NYSDEC, a notice of complete application can be published by
NYSDEC in the Environmental Notice Bulletin. Concurrently,
the Governor's office also will be notified of the complete appli-
cation, and copies of these notices will be sent to the Chief Execu-
tive Officer of the municipality where the project is located and to
property owners within 3,000 ft of the facility. In addition,
notices are published in two local newspapers. Within 15 days of
notification, the Governor must constitute the Siting Board.
The Siting Board is composed of the Commissioners of Envi-
ronmental Conservation, Transportation, Health, Commerce and
the Secretary of State and three ad hoc members to be nomi-
nated by the Governor. Two of these ad hoc members must be
residents of the county of the project's location. The Governor
appoints the Chairman of the Siting Board, and the Commis-
sioner of Environmental Conservation makes staff available to
support the Board in carrying out its responsibilities.
Table 2
Siting Considerations Average Weights and
Special Case Weight Changes
SITING COHSIDCRATIONS
Population Density
Population Adjacent
to Transport tout*
lick of Accident in
Transportation
Proximity to Incom-
patible Structures
Utility Lift*.
Municipal Effect!
Contamination of Ground
and Svrface Waters
Water Supply Sources

Air Quality
Are** of Mineral
Exploitation
Preservation of
Endangered, Threatened,
and Indigenous Specie*
Conservation of
Historic and

Open Space,
•ecreational ,
and Visual Israels
Average
10
7
10
3
1
4
10 -U-J) -(2-3)
0
11 *(2-*> »(l-3)
12 *(4-6)
3 * * -(1)
6

3 +(1-2) *(!-:) -(1-2)
Tool Scor* - 100
Source: New York Stale Department of Environmental Conservation. 6 NYCRR Part 361, Siting
of Industrial Hazardous Waste Fac/lllles.
Public Hearing Procedures
The Commissioner of Environmental Conservation appoints
an administrative law judge for the State Environmental Quality
Review who conducts an adjudicatory public hearing upon the
application. The hearing will be held jointly with the Siting
Board's hearing. This hearing has to commence within 60 days of
constituting the Siting Board. The hearing will be published in the
same manner as in the case of a Notice of Complete Application.
The hearing has to commence within 30 days of such notice. The
members of the Siting Board may, at their option, participate in
the Adjudicatory hearings. Such participation may include, but is
not limited to, examining witnesses and requiring the production
of documents or witnesses.

Siting Board Decision Siting Criteria
The Siting Board shall make the final decision on an applica-
tion for a Certificate of Environmental Safety based upon the
record made before the hearing officer, after receiving briefs from
the parties to the hearing and exceptions to the recommended
decisions of the hearing officer and after hearing oral argument.
The Siting Board shall mail its decision to the applicant, to the
NYSDEC and to all parties of the hearing on or before 60 calen-
dar days after receipt by the Siting Board of the complete record.
The Siting Board will deny an application to construct or oper-
ate a facility if residential areas and contiguous population will
be endangered, if such construction would be contrary to local
zoning or land use regulations in force or if it otherwise does not
conform to the siting criteria established for such facilities. These
various siting considerations and their weighted averages in arriv-
ing at a Siting Board decision are contained in Table 2.

JOINT VENTURE CONCEPT
Faculties for the management of hazardous waste could be con-
structed through a joint venture concept. A regional hazardous
waste management facility, such as an incinerator, could be con-
structed to serve the shortfall in treatment capacity for a group of
large companies in New York. The joint venture would not only
provide for its own shortfall in hazardous waste treatment capac-
ity but also would provide for capacity available for commercial
use by other companies. The joint venture would provide up-
front costs to obtain the necessary approvals and permits for the
treatment/disposal facility.
The New York State Environmental Facilities Corporation
(EFC) is a public benefit corporation of the State of New York,
which provides a broad range of services in connection with envi-
ronmentally beneficial projects, including solid and hazardous
waste disposal throughout the State of New York.
EFC may plan, finance, construct, operate and maintain haz-
ardous waste disposal faculties for municipalities, state and pri-
vate entities. EFC can make loans to private industry for haz-
ardous waste treatment facilities by issuing bonds with the interest
exempt from federal and state taxes, thus allowing favorable in-
terest rates and terms. The bonds are special obligations of EFC
payable by industry from income from the project by a loan
agreement between EFC and industry.

REFERENCE
1. New York State Department of Environmental Conservation, Hta-
ardous Waste Treatment Facilities Task Force, Final Report, Sept.
1985.
416 STATE PROGRAMS
-------
Impact of State Environmental Laws
On Property Acquisitions

Daniel K. Moon
Robert H. Clemens
Diane P. Heineman
Alliance Technologies Corporation
Bedford, Massachusetts
ABSTRACT
A growing number of states including New Jersey, Massachu-
setts, Connecticut and New Hampshire have passed legislation
affecting the purchase and sale of commercial and industrial
properties. These regulations have prompted many property own-
ers, buyers, lenders and title insurance companies to become
aware of the potential liabilities associated with owning or pur-
chasing contaminated real estate and to insist that a site assess-
ment be conducted prior to closing. These site assessments have
helped to uncover many contamination problems prior to sale
and have helped to reduce the potential for future abandoned
waste sites. By protecting their investments, the buyers, lenders
and insurers are providing unique enforcement support to the
state hazardous waste programs.

INTRODUCTION
Most real estate investors, builders, developers, commercial
lending institutions, law firms and title insurance companies are
becoming increasingly aware of laws which have been enacted
that hold property owners responsible for any hazardous sub-
stances found on their property. A number-of states have passed
laws that closely parallel the federal "Superfund" legislation that
was passed in 1980 and more recently amended in 1986.

STATE PROGRAMS
New Jersey
New Jersey's Environmental Cleanup Responsibility Act
(ECRA) is perhaps the most stringent of the state programs. The
ECRA legislation requires that commercial and industrial facili-
ties undergo a thorough site assessment before the property can
be transferred. ECRA applies to industrial establishments in-
volved in the generation, manufacture, refining, transportation,
storage, handling or disposal of hazardous waste or substances,
and whose activities fall within certain Standard Industrial Class-
ification (SIC) categories.
As a first step in the ECRA process the seller must notify the
New Jersey Department of Environmental Protection (NJDEP).
After notification, the seller must submit general information and
site evaluation information and then must file either a "nega-
tive declaration," verifying that the property is free of environ-
mental problems, or he must outline a "cleanup plan," describ-
ing methods of site remediation. The site evaluation information
requires an extensive amount of information including: a descrip-
tion of facilities and operations, an inventory of hazardous wastes
and substances, maps of spills or discharges, description of any
enforcement actions or permits, a sampling and analysis plan
and procedures for decontamination. Prior to implementation,
the sampling and analysis plan must be reviewed and approved
by NJDEP. If sampling and analysis indicate the presence of haz-
ardous materials on the property, the current owner is required to
remediate the problem before the property transaction can be
completed.

Massachusetts
Other states have passed laws similar to New Jersey's ECRA.
In New England, Massachusetts, New Hampshire and Connecti-
cut have enacted legislation to address the liabilities and costs
that may be incurred by the state in cleaning up contaminated
commercial and industrial properties. Under Massachusetts Gen-
eral Law, Chapter 21E, strict liabilities are imposed upon pres-
ent and past owners of contaminated real estate. In addition, the
Commonwealth of Massachusetts has the power to place a lien
upon the property in order to recover the costs of testing, penal-
ties and cleanup expended by the state. In commercial real estate,
this hen takes priority over mortgages and is commonly referred
to as a "Superlien." The Massachusetts Department of Environ-
mental Quality Engineering (DEQE) is responsible for the imple-
mentation of Chapter 21E. However, it is not the state, but the
title insurance companies and lending institutions that apply the
full impact of this legislation to builders, developers and inves-
tors. Banks and lending institutions are wary of the significant
costs which may be associated with having a "Superlien" im-
posed upon a property. To protect themselves, the lenders seek
the attachment of a rider to the buyer's title insurance policy.
As a result, the title insurance companies often require a state
assessment for the purchase or refinancing of a commercial or in-
dustrial property.
New Hampshire
Two other New England states—New Hampshire and Connec-
ticut—have enacted legislation similar to Massachusetts'. New
Hampshire's Hazardous Waste Laws establish a cleanup fund in
Chapter 147-B. A section of that Chapter, 147-B:10, addresses
liability of responsible parties to recover the costs of containment,
cleanup or removal of hazardous waste expended by the State's
fund. As with Massachusetts, it is the lenders and title insurers
who are requiring site assessments for New Hampshire proper-
ties to reduce the risk that the property will be affected by a lien.

Connecticut
Transfers of real estate in Connecticut are affected by two pub-
he acts. The first, Public Act 85-568, provides for a civil fine of
$100,000 for failing to report the existence of a known release.
However, since a properly conducted cleanup can easily exceed
$100,000, crafty or irresponsible property sellers could realize that
it might be easier to risk payment of the fine and leave the buyer
STATE PROGRAMS 417
-------
with the more costly cleanup obligation. Connecticut's legislation
differs from that of its New England neighbors because it requires
filing declaration forms which attest to the property's environ-
mental condition, with the Connecticut Department of Environ-
mental Protection (DEP). The forms are intended to be com-
pleted and certified by the seller and signed, acknowledging re-
ceipt, by the buyer. So, an informed and wary buyer will know
that it is in his best interest to have a professional site assessment
of the property to minimize the risk of making a hazardous in-
vestment. In effect, by signing the forms, the buyer declares that
the property was clean when acquired; any releases discovered
subsequently are his responsibility.
The other statute affecting real estate transfers in Connecticut
is Public Act 85-443, which allows the state to place a priority lien
on the proper'y. The lien takes effect upon expenditure of state
money toward property cleanup. As in Massachusetts and New
Hampshire, it is the title insurers and lenders, not the regulatory
agencies, who are requiring environmental assessments in
Connecticut prior to closings.

SITE ASSESSMENTS TO MANAGE RISKS
In response to the needs of buyers, lenders, investors and in-
surance companies, the site assessment has evolved as a useful
tool to manage risks. An environmental assessment, conducted
prior to closing on a property, can avert a potentially hazardous
investment for a buyer. A smart buyer will insist on having a site
assessment conducted before taking title and, ideally, prior to
signing a purchase and sale agreement. If hazardous wastes are
found on the property early enough in the process, the buyer can
avoid a risky purchase. However, he also might choose to nego-
tiate a better price, knowing that the property is tainted, or even
accept the property subject to the seller's obligation to fund the
necessary cleanup.
At this time most states have not set up specific guidelines for
these site assessments; thus, the cost, quality and level of detail
vary significantly from one firm to another. The title insurance
firms usually will specify what type of assessment they will re-
quire, to be assured of the property's integrity.
Some states are considering reforms to their current require-
ments to provide for a stronger state role in requiring site assess-
ments and to furnish guidance for the performance of site eval-
uations. Eventually, most states may implement requirements
such as New Jersey's ECRA to require site assessments, at least
for major real estate transactions.
Perhaps the most significant impact of the various state pro-
grams is the increase in the number of "dirty" properties and
"problem" sites that are discovered and promptly brought to the
attention of the appropriate state regulatory agencies. With the
increasingly large number of industrial and commercial real estate
transfers taking place every day, site assessments provide a pow-
erful mechanism to add to the government's capability to discov-
er unreported spills, improper waste disposal practices, leaking
underground tanks and to uncover members of the regulated
community who are not in compliance.
CONCLUSION
Several states have enacted legislation that addresses the liabil-
ities and costs that may be incurred by the states in cleaning up
commercial and industrial properties. These state programs have
provided an additional dimension to the state hazardous waste en-
forcement programs. The burden of investigating real estate for
hazardous waste problems prior to sale lies not with the over-
worked, under-staffed state office personnel, but with the liable
and therefore cautious buyers, sellers and lenders. In a sense, pri-
vate parties, title insurance companies and lenders have evolved
into unique "enforcers" of these state regulations.
In addition, the incentive to complete site investigations prior
to closing on a real estate transaction often can produce rapid re-
sults; this is in contrast to the often extensive delays experienced
in the implementation of the Federal Superfund program.
In summary, state legislation impacting real estate transactions
has provided a unique additional component to state hazardous
waste enforcement programs. Real estate site assessments often
provide rapid results and have helped to curtail the growing num-
ber of abandoned waste sites in a number of states.
418 STATE PROGRAMS
-------
1987 Exhibitors
Alliance Technologies Corporation
213 Burlington Rd.
Bedford, MA 01730
617/275-9000
Alliance Technologies Corporation provides environ-
mental consulting and analytical laboratory services to
industry and government. Consulting services include
development of waste minimization programs, alterna-
tive concentration levels, comprehensive correction
active programs, fate and transport studies, air and
groundwater modeling and rust assessments. Compre-
hensive environmental testing services are provided by a
state-of-the-art laboratory certified by New York, New
Jersey, Massachusetts and which participate in the
U.S. EPA contract lab program. Formally OCA Tech-
nology Division, Alliance is now a subsidiary of TRC
companies.

Allis-Chalmers Corporation
P.O. Box 512
Milwaukee, WI53201 414/475-2690
Allis-Chalmers Corporation provides pyroprocessing
systems, equipment, and technical services for the
thermal destruction of bulk hazardous wastes; sys-
tems, equipment, and technical services for the removal
of heavy metals from liquid effluents. The technology
offered is an extension of the company's experience of
more than a century in minerals processing.

Art's Manufacturing and Supply
105 Harrison
American Falls, ID 83211 208/226-2017
AMS, the leader in hand-operated soil sampling equip-
ment for over 40 years, is now the leader in sampling
equipment for the hazardous waste industry. Stop by
Booth #1504 and see the new patent-pending soil recov-
ery auger and our full line of completely stainless
steel soil sampling equipment.

Association of Engineering Geologists
P.O. Box 368
Lawrence, KS 66044 913/843-1234
The Association of Engineering Geology, dedicated to
the advancement of the science of engineering geology,
promotes public safety and welfare; promotes public
understanding of the engineering geology profession;
and provides a medium for discussion and distribution
of technical subjects, problems, and information. The
Association has approximately 3000 members.

Autumn Industries, Inc.
518 Perkins-Jones Road
Warren, OH 44483 216/372-5002
Autumn Industries, Inc. is a licensed hazardous material
and waste transporter specializing in bulk solids. Skilled
management and trained personnel possess the ability
to provide 200 units to serve accounts in the Midwest
and Northeastern states. Autumn Industries, Inc. is fast
becoming a major force in the hazardous transportation
field.
BCM Engineers, Inc.
1 Plymouth Meeting Mall
Plymouth Meeting, PA 19462 215/825-3800
Quality engineering in hazardous waste management
and control: • Groundwater Studies • Geophysical
Surveys • Remedial Design Engineering • Superfund
Site Investigations • Facility Permitting • Closure
Plans • Real Estate Contamination Assessments
• Asbestos Surveys • Analytical Services.
BES Environmental Specialists, Inc.
8286 Boston Hill Road
Larksville, PA 18651
717/779-5317
BES-total environmental services emergency response,
waste removal, transportation and site restoration.
BES has provided removal services for a number of
governmental and private industrial clients. Our staff
has extensive field experience in site and safety manage-
ment. Our project managers are skilled problem solvers.
We have the hardware, the personnel and the ability to
respond quickly and professionally.

BONDICO, Inc
2410 Silver St.
Jacksonville, FL 32206 904/358-2606
BONDICO, Inc. has introduced a 90-gaUon container
system designed for transportation, storage, treatment
and disposal of hazardous materials and low-level rad-
wastes. A dual laminate composite of polyethylene and
fiberglas, the container provides superior safety and ex-
traordinary cost-effectiveness with multiple reuse, in the
hazardous waste market. For radwaste, as a 7A type A
package, BONDICOL's system provides benefits and
performance that will set new standards. This unit may
be utilized as a salvage container, on-site storage, trans-
fer container, tool crib or as an encapsulate via its inno-
vative on-site sealing system. BONDICO's container
with its fully removable lid, is rustproof, leakproof, cor-
rosion resistant, lightweight, nestable and reusable.

Battelle Pacific Northwest
Laboratories
P.O. Box 999
Richland, WA 99352 509/375-2867
Battelle Pacific Northwest Laboratories offers a wide
variety of R&D and technical application services in-
cluding site characterization and assessment for active
and inactive sites, health effects assessments, and pro-
cess control and remediation technologies. Battelle of-
fers advanced technology coupled with a cost-effective,
multi-disciplinary approach for solving waste-site clean-
up problems.

Bird Environmental Systems, Inc.
100 Neponset Street
South Walpole, MA 02071 617/668-0400
Bird Environmental Systems, Inc., manufactures five
models of mobile dewatering, oil recovery and pretreat-
ment systems. Each is based on difference separation
technology, including decanter and disc centrifugen,
belt filter and plate and frame filter presses. Two facil-
ities support sales and marketing of these systems, along
with special services provided in permitting, laboratory
analysis, and technical assistance.

The Bureau of National Affairs, Inc.
123125th St., N.W.
Washington, DC 20037 202/452-4452
BNA publishes regulatory, legal and working guides
providing the latest information concerning the manu-
facture, transportation, safe handling and disposal of
hazardous materials.

COM Federal Programs Corp.
7611 Little River Turnpike, Suite 104
Annandale, VA 22003 703/642-0544
CDM provides comprehensive engineering and manage-
ment services to public and private clients. Our haz-
ardous waste services include remedial investigations,
feasibility studies, site cleanup management, RCRA per-
mitting, computerized groundwater modeling, aquifer
restoration, risk assessment, underground storage tank
evaluation and remediation, environmental audits,
waste reduction and expert testimony.

CECOS International, Inc.
2321KenmoreAve.
Buffalo, NY 14207 716/873-4200
CECOS International, Inc., is a company specializing in
the treatment and disposal of hazardous chemical waste.
CECOS makes these services available to industry
through a network of regional treatment centers across
the United States and Puerto Rico. CECOS offers
specialized hazardous waste capabilities, research and
analytical and consulting services.

CarbonAir Services, Inc.
P.O. Drawer 5117
Hopkins, MN 55343 612/935-1844
CarbonAir Services provides treatment design and in-
1987 EXHIBITORS' LIST 419
-------
stallation for system removal of dissolved organic or in-
organic contaminants in groundwater, surface water or
process streams. Treatment alternatives include carbon
adsorption, packed column airstrlpping, oil/water sep-
aration (membrane or physio-chemical), heavy metals
precipitation, and ancillary equipment for turbidity re
moval, solids dewatering, etc. No PCB or Dioxin water
streams handled at this time.

Chem-Met Services
18550 Allen Rd.
Wyandotte, Ml 48192 313/282-9250
Chem-Met Services provides environmentally safe treat-
ment for liquid and solid wastes. Our process is ecolog-
ically sound and economically efficient. We have been
processing waste streams for industry since 1966. Client-
oriented, we know that dependable service is essential.
Our waste disposal treatment is proven sound for our
environment. We stay current with changing govern-
ment regulations. Analysis of all hazardous waste is re-
quired and maintained. Remedial action, on-site pro-
cessing, transportation in 36 stales.

Clayton Environment*! Consultants, Inc.
22345 Roethel Drive
Novi, MI 48050 313/344-1770
Clayton Environmental Consultants Waste Manage-
ment Services: • Environmental Risk Assessment •
Hazardous and Solid Waste Management • Under-
ground Storage Tank Management • Water Resources
and Wast «_ter Engineering • Environmental Audits
for Preacquisition, Mergers, Foreclosures and Internal
Review • PCB Sampling, Cleanup and Decontamina-
tion • Groundwater Contamination Studies • Water
and Wastewater Sampling • Statistical Sampling of
Soils and Waste Piles • Health and Safety Plans •
Regulatory Interpretation and Negotiation.

Chcmflx Technologies, Inc.
2424 Edenborn Ave., Suite 620
Metairie, LA 70001 504/831-3600
Chemfix Technologies, Inc. (CTI) offers the patented
CHEMFIX process for chemical fixation/stabilization
of both hazardous and nonhazardous liquids and
sludges. Complete mobile services are offered, as well as
fixed plant facilities for continuous generation waste
streams. CTI services include site assessments, waste
stream characterization and permitting support.
CompaCbtm Laboratories
3308 Chapel Hill/Nelson Highway
Research Triangle Park, NC 27709 919/549-8263
CompuChem provides comprehensive inorganic and
organic analytical laboratory services, utilizing GC, CC/
MS and inorganic techniques. We focus in the areas of:
Superfund analysis; RCRA; priority pollutant analysis;
identification of unknown wastes; groundwater moni-
toring; dioxin analysis; waste site screening and waste
characterization. CompuChem is introducing the Envi-
ronmental Site Profile (ESP) System, a proprietary data
management system which provides on-line access to
laboratory test results, plus the capability for flexible
analysis and presentation of downloaded data.

Dorr-Oliver Inc.
77 Havemeyer Lane
Stamford, CT 06904 203/358-3430
Dorr-Oliver manufacturers fluid bed and membrane-
based biological treatment systems to solve a broad
range of wastewater treatment problems. The fluid bed
Oxitron® system is available with an active media which
effectively combines biological treatment with physical
adsorption to provide high contaminate removal effic-
iency and low operating costs. Atmospheric emissions
of volatile organic compounds are also eliminated.

Dravo Engineering Companies, Inc./
Glbbs&HIII
111'enn Plaza
New York, NY 10001 212/216-6626
Dravo Engineering Companies, Inc./(Gibbs & Hill) has
been working with industry in solving their environmen-
tal problems for over 40 years. Dravo has a long success
in providing industrial treatment/process engineering,
hazardous waste (RI/FS) investigations, hydrogeologic
studies, air pollution control technology, and resource
recovery design/construction/operation services.

Do Pont Company/Teflon
1007 Market, Room NA 235
Wilmington, DE 19898 302/774-2692
Components manufactured from 100% virgin DuPont
Teflon* fluorocarbon resins used in the Groundwater
Monitoring Industry ensure that representative ground-
water samples are obtained. Displayed will be dedicated
bladder pumps, bailers, transfer tubing, casing and
screen of 100% TEFLON* manufactured by various
companies serving the Groundwaler Monitoring Indus-
try.

Dynamic Corporation
11140 Rockville Pike
Rockville, MD 20852 301 /468-2500
Dynamac Corporation, an engineering and scientific
services firm, has performed approximately 600 haz-
ardous waste/materials projects since its inception in
1970. We can solve your problems in areas such as com-
pliance auditing; RCRA permitting; corrective action;
waste site investigation, evaluation and remediation;
risk/endangerment assessment; and occupational safety
and health.
E.C. Jordan Co.
261 Commercial Street
Portland, ME 04112
207/775-5401
Solid and hazardous waste management services pro-
vided to industry and government agencies include geo-
physical and geohydrological investigations, record
searches, chemical characterization, contamination risk
assessment, identification and evaluation of remedial
action alternatives and implementation plans at haz-
ardous waste sites. Hazardous waste TSD facilities are
developed from initial planning stages, through site
selection and investigation, design, permit application
and construction management.

E.I. du Pont & Co.-
Envtronmenlal Services
Chambers Works, Technical Laboratory
Deepwater, NJ 08023 609/540-3884
Through its Environmental Services operation, Du Pont
treats wastewater and contaminated equipment on a
contract basis at its EPA-permitted Chambers Works
facilities, Deepwater, New Jersey. The 40-million-gallon
per day Wastewater Treatment Plant destroys industrial
wastewater chemicals through the patented Powdered
Activated Carbon Treatment (PACT) process. The
Thermal Decontamination Unit removes organic chem-
ical residuals from equipment through high-temperature
treatment.
ENSCO, Inc.
1015 Louisiana St.
Little Rock, AR 72202
501/375-8444
The company provides integrated hazardous waste man-
agement services to private industry, public utilities and
governmental entities. These services include chemical
analysis, collection, transportation, storage, processing
and incineration of hazardous waste. The company also
provides transformer decommissioning and other waste
management services, including the reclamation of
abandoned or problem waste sites, and engineering and
construction services for itself and others. The company
has developed and is currently marketing modular in-
cinerators for use at hazardous sites.

The ERM Group
999 West Chester Pike
West Chester, PA 19382 215/696-9110
The ERM Group, a nationwide network of consulting
firms specializing in environmental engineering and
science, brings together engineers, geologists, hydro-
geologists, biologists, economists, planners, and experts
in conflict management and public participation. ERM
offers services in the areas of hazardous and solid mole
management, hydrogeological evaluations, ecological
evaluations, air quality control, risk assessments, en-
vironmental compliance reviews, environmental data
management, design and construction assistance, under-
ground tank leak prevention and remediation, regula-
tory compliance training, (technical representation with
regulatory agencies and training of personnel in RCRA,
CERCLA and other regulatory acts) assistance with per-
mit applications, and assistance with negotiations and
public participation issues.
Earth Resources Corporation
P.O. Box 616961
Orlando, FL 32861-6961
305/295-8848
Earth Resources Corporation (ERQ is a full-service
hazardous materials management firm specializing in the
containment, treatment and removal of all types of haz-
ardous materials. ERC has a highly trained professional
and technical staff experienced in the design and imple-
mentation of innovative solutions to today's waste prob-
lems. ERC's remedial action capabilities include but are
not limited to soil, groundwater, facilities, containerized
wastes and pressurized gas cylinders.
The Earth Technology Corporation
3777 Long Beach Blvd.
Long Beach. CA 90807 213/595-6611
The Earth Technology Corporation offers comprehen-
sive hazardous waste management services. Those which
we are most frequently requested to provide include: en-
vironmental auditing/compliance assessment, haz-
ardouz waste permitting, remedial investigations, re-
medial/corrective action, design and engineering, geo-
technical investigations, waste stream reduction and
recovery, facility closure and laboratory and special
technical services.
ELF TechjMtofks, IK.
High Ridge Park. P.O. Box 10037
Stamford, CT 06904
203/358-5119
ELF Technologies, Inc., a subsidiary of Elf Aquitaine,
Inc., the largest petroleum group in France, presents the
Elf "Cinclus" process of incineration, and is interested
in a partnership with a U.S. company. Cinclus is suitable
for all gas, liquid and slurry waste, from all industries. It
can also be used on solids, if ground and pumpable.

Ecova Corporation
15555 ME 33rd
Redmond. WA 98052 206/882-4364
Ecova Corporation is « unique interdisciplinary com-
pany providing advanced technology solutions for the
on-site remediation of toxic and hazardous wastes.
Ecova combines advanced biological techniques with
chemical and physical technologies in conjunction with
on-site engineering and site management capabilities. As
a result, Ecova can offer site-specific treatment systems
that result in the terminal destruction of toxic and haz-
ardous wastes.
Engineering-Science
57 Executive Park South, Suite 590
Atlanta, GA 30329
404/325-0770
Engineering-Science (ES) is a major, full-service, na-
tional and international environmental engineering firm
providing complete services in hazardous waste man-
agement. ES has offices in 15 domestic locations con-
veniently located to serve industry, military and govern-
mental clients. ES is active in supporting industrial and
military clients in all phases of site/remedial investiga-
tions, feasibility studies, remedial action plan prepara-
tion, site cleanup/closure and postclosure activities.
420 1987 EXHIBITORS' LIST
-------
Environmental Enterprises, Inc.
10147 Springfield Pike
Cincinnati, OH 45215 513/772-2828
Disposal and cleanup services. Consulting, laboratory
testing, chemical treatment, waste encapsulation, PCB
treatment and disposal, emergency response.

Environmental Management News
225 N. New Road
Waco, TX 76714 817/776-9000
Environmental Management News magazine is dedi-
cated to the management of air, water, waste water,
pollution and hazardous materials. Environmental Man-
agement News Action Pac offers information on pro-
ducts and services for the control of industrial pollution
and hazardous materials.

Envirite Field Services, Inc.
600 W. Germantown Pk., #221
Plymouth Meeting, PA 19462 215/825-8877
Envirite Field Services provides solidification/fixation
services for organic and/or inorganic industrial wastes.
The company offers three proprietary delivery systems
to stabilize waste liquids, sludges and contaminated soils
with selected additives—the VR/S™ system for low-
range solids, the PF-5™ system for mid-range solids,
and the HSS™ system for high-range solids. The com-
pany also provides dewatering services using mobile
filter presses.
First Systems
134 Middle Neck Rd.
Great Neck, NY 11021
516/829-5858
U.S. EPA, Coast Guard Hazardous Material Available
on a PC—The Oil & Hazardous Material Technical
Assistance Database (OHM/TADS) of the U.S. En-
vironmental Protection Agency (U.S. EPA) and the
Chemical Hazards Response Information System
(CHRIS) of the U.S. Coast Guard are available from
FIRSTsystem on a personal computer. microOHM/
TADS lists 1,400 compounds selected for spill history,
production volume and toxicity. It includes 15,000 ma-
terial, brand and trade names, and 118 of the fields from
the full OHM/TADS.

GAI Consultants, Inc.
570BeattyRd.
Monroeville, PA 15146 412/856-6400
GAI Consultants, Inc. and its subsidiaries provide tech-
nical consulting services in the areas of solid and haz-
ardous waste management, federal and state permitting
assistance, and disposal site design services including
remedial investigations and feasibility studies, hydrogeo-
logic investigations, site selection and cost optimization
evaluations, and site operation and closure plans.

Colder Associates, Inc.
3772Pleasantdale Rd., Suite 165
Atlanta, GA 30340 404/496-1893
Colder is an international consulting company special-
izing in the technical fields of geo-engineering, primarily
covering the fields of energy, hazardous wastes, natural
resources, transportation, civil, and offshore engineer-
ing. The company maintains 26 worldwide offices
located in the United States, Canada, Australia and the
United Kingdom.
Gundle Lining Systems Inc.
1340E. RicheyRd.
Houston, TX 77073
713/443-8564/
8
-------
ICF Technology
1850KSt.,NW
Washington, DC 20006 202/862-1100
ICF Technology—the scientific and engineering subsidi-
ary of ICF Incorporated—consults on environmental
and hazardous waste management issues. We provide
our clients with a full range of technical services, from
site-specific investigations and risk assessments to
remedial design and construction monitoring. The firm
is headquartered in Washington, DC.

Industrial Chemicals Dlvblon/3M
Bldg. 225-3S-05, 3M Center
St. Paul, MN 55144-1000 612/733-9732
3M Foams have proven their suppression effectiveness
during hazardous material clean-up that involves re-
lease of volatile organic compounds (VOC), air toxics,
odors and dust. These water-based foams conform to
the terrain and last hours, days, and even weeks de-
pending on the site requirements.

In-Sltn Inc.
P.O. Box 1
Laramie, WY 82070-0920 307/742-8213
In-Situ's programmable data-acquisition instruments
record water level changes for aquifer studies or long-
term remote monitoring and can be downloaded to
mainframe or micro computers via an RS232C port or
an optional telephone modem. Forthcoming enhance-
ments will allow data collected by In-Situ instruments to
be easily transferred into your office system or PC.

Industrial A Environmental
Analysts, Inc.
P.O. Box 12846
Research Triangle Par, NC 27709 919/467-9919
Industrial A Environmental Analysts, Inc. (IEA) is a
multidisciplinary certified analytical and consulting
laboratory working primarily with industrial and gov-
ernmental clients in environmental and chemical moni-
toring, testing, research and management. lEA's na-
tionally recognized capabilities combine EPA Contract
Laboratory Protocol with State-of-the-Art instrumen-
tation in order to provide efficient solutions to analytical
problems.

Island Press
1718 Connecticut Ave., NW, WOO
Washington, DC 20009 202/232-7933
Island Press is a national non-profit environmental
organization dedicated to the proper use and man-
agement of our natural resources and the environment.
Island Press offers a full range of publishing and distri-
bution services, presenting the most advanced thinking
in the conservation and management of land, soil,
water, forests, wildlife, and hazardous wastes.

Lachit Instruments
10500 N. Port Washington Rd.
Mequon, WI 53092 414/241-3872
The QuikChem Automated Ion Analyzer, using meth-
ods formally approved by U.S. EPA for NPDES Mon-
itoring, automates the determination of many ionic
species in waters and wastes. These include most non-
metallic inorganics plus some metals and organics listed
in the U.S. EPA publication "Methods for Chemical
Analysis of Waters and Wastes." The system processes
90-120 samples per hour for up to 4 analytes. Each de-
termination takes less than 1 minute and costs under 1
cent. Up to 3 decades of concentration range can be
accommodated in a single run. Data handling is fully
automated including calibration, data reduction, report
generation and post-run processing.

Lam Lundf Manufacturing Co.
11615 N. Shore Rd.
Whitmore Lake, MI48189 313/449-4116
Manufacture extractor! which meet the requirements of
the U.S. EPA for evaluating solid waste. They also meet
requirements established in the new "Toxlcity Charac-
teristic Leaching Procedure."

Liw Environmental, Inc.
1000 Abernathy Rd., 11800. Bldg. 400
Atlanta, OA 30328 404/392-0585
Law Environmental, Inc. assists clienu with their haz-
ardous and nonhazardoui waste concerns, specializing
in seven core business areas: Hazardous and Solid Waste
Management • Groundwater Hydrology and Resource
Development • Surface Water Hydrology and Water
Quality Protection • Land Treatment of Wastes •
Ecological Assessment • OeophyiicaJ Exploration •
Seismic Hazard Evaluation.

metiTRACE. Inc.
13715 Rider Trail North
Earth City, MO 63045 314/298-8566
metaTRACE offers full service analytical capabilities in-
cluding: complete analytical services for organic* and in-
organics; analysis of toxics, Including dioxins and fur-
ans; analysis of co-contaminated waste; radiological
analysis; quick turnaround analyses for remedial pro-
grams; hazardous waste analysis; Appendix VIII and IX
analysis; U.S. EPA priority pollutant analysis; industrial
hygiene sample analysis; air quality analyse).

Mary Ann Liebert, Inc.,
Publishers
1651 Third Ave.
New York. NY 10128 212/289-2300
HAZARDOUS WASTE A HAZARDOUS MA-
TERIALS, a quarterly journal published by Mary Ann
Liebert, Inc., is the official journal of the Hazardous
Materials Control Research Institute. It is the central
source of information for advancing technology and
ultimately, of providing economical and ecological
methodology for the regulation and management of
hazardous waste and related hazardous material. Other
Mary Ann Liebert publications include Genetic Engi-
neering News. Journal of The American College of
Toxicology, Alternative Methods in Toxicology Series,
AIDS Research and Human Relrovinaet....Stop by our
booth to find out more about our other journals and
newsmagazines.
MetropoUuu Environmental Inc.
P.O. Box 609
Celina, OH 45822
419/586-6638
The company offers a total service package for today's
industrial waste. Transportation offers a full line of
waste hauling vehicles complete with roll-off equipment.
Mobile filtration units can be brought on-sitc and within
a few hours begin dewatering various types of sludges,
greatly reducing disposal cost. Completing the package
is excavation equipment that allows the company to do
complete EPA lagoon closures from start to finish.
Mllllport Corporation
80 Ashby Rd.
Bedford, MA 01730
617/275-9200x8425
Millipore will exhibit the new Zero Head Space Extrac-
tor for determining the mobility of volatile organics
called out In the new RLP procedure. The 142mm Haz-
ardous Waste Filtration system for EP toxicity testing
and semi-volatile filtration will also be shown.
NERI/PETREX
309 Farmington Ave., Suite B-109
Farmington, CT06032 203/677-9666
The Petrex Technique is an innovative geochemical
method for identifying and mapping volatile organic
compounds from soils and groundwater contamination.
The Technique utilizes Petrex monitors which are
placed in the ground in a strategic pattern. After a repre-
sentative subsurface sampling period, the monitors arc
removed and analyzed by mass spectrometry.
NL Barold, NL Industrie*
P.O. Box 1675
Houston, TX 77251 713/987-4850
NL Baroid'» high quality bentonite products offer cost-
effective solutions to soil permeability control. Whether
soil sealing, slurry trenching or other environmental ap-
plications, our technical services will determine op-
timum treatment to meet your project requirements. For
60 years, NL Baroid has devoted ils resources to pro-
viding leading edge bentonite technology.

N US Corporation
Park West Two
Cliff Mine Rd.
Pittsburgh. PA 15275 412/788-1080
NUS Corporation is a broad-based energy and environ-
mental engineering and consulting firm providing pro-
fessional services to industry, utilities and government in
the areas of air and water pollution control, solid and
hazardous waste management, site cleanup and closure,
groundwater quality protection, energy management,
health and safety training and laboratory services.

Nanco Labs, Inc.
RD*6 Robinson Ln.
Wappingers Falls. NY 12590 914/221-2485
Nanco Labs provides quality, cost-effective environ-
mental analyses to clienu throughout the United States.
Nanco is a contract laboratory to the U.S. EPA, the
New York Slate DEC and the New Jersey DEP. Nanco
specializes in HSL, priority pollutant and RCRA
analyses.

National Dneger, Inc.
P.O. Box 120
Pittsburgh. PA 15230 412/787-8383
National Draeger, Inc.. a U.S. subsidiary of Draeger-
werk AC, located in Luebeck, West Germany, has
earned a worldwide reputation for being the leader in
manufacturing specialized equipment and systems which
enable, support and protect human breathing safety. By
introducing innovative new products for gas detection
and warning systems, breathing protection, filter tech-
nology, diving equipment, air-supplied systems for avia-
ation and space technology as well as medical equip-
ment. Draeger helps to upgrade safety standards for the
entire safely industry.

National Urne Association
3601 N. Fairfax Dr.
Arlington. VA 22201 703/243-5463
The National Lime Association exhibit provides infor-
mation on the use of lime for hazardous waste treat-
ment. The principal use of lime is the neutralization of
inorganic acidic waste and the precipitation of heavy
metals. Lime and fly ash form a pozzotanic material
which can be used to solidify a hazardous sludge.

Occupational Hazards Magazine
1100 Superior Ave.
Cleveland, OH 44114 216/696-7000
A trade magazine serving the industrial safety, health,
hygiene and plant protection market.

Orion Research, Inc.
529 Main St.
Boston, MA 02129 617/242-3900
Orion designs and manufactures ion-selective electrodes
and direct readout pH/ISE meters for fast, accurate and
cost-effective chemical measurements. The ORION 960
Autochemistry System expands and enhances measure-
ment by electrode. Featuring ammonia, cyanide and
nitrate electrodes. And introducing a new, easy-to-use
Nitrate Test Kit.

P.E. LaMoreaux A Associates
P.O. Box 2310
Tuscaloosa, AL 35403 205/752-5543
Hydrological, geological, environmental, hazardous
422 1987 EXHIBITORS' LIST
-------
waste consultation services. Geophysical testing, sam-
pling, laboratory analysis, monitoring programs devel-
opment and well installation, reclamation. Permitting,
court testimony, graphics and communication capabili-
ty. Contact: James W. LaMoreaux, President.

Peroxidation Systems, Inc.
4400 E. Broadway
Tucson, AZ 85711-3558 602/327-0277
Pollution Abatement Services and Equipment.

Photovac International, Inc.
741 Park Ave.
Huntington, NY 11743 516/351-5809
Photovac's full line of environmental monitoring equip-
ment including TIP I, a hand-held real-time Photoioni-
zation detector and auto TIP which has continuous
monitoring and alarm capability. The popular 10S series
of portable GCs will also be shown with new capillary
column capability and the new range of multiple energy
photoionization detector lamps.
Poly-America, Poly-Flex Div.
2000 W. Marshall
Grand Prairie, TX 75051
817/640-0640
Poly-Flex is a polyethylene geomembrane liner (20-100
mils thick) which provides a cost-effective method of lin-
ing hazardous waste disposal facilities and preventing
groundwater pollution. Poly-Flex is manufactured by
Poly-America, one of the most modern extrusion facili-
ties in the U.S., producing 150 million pounds of poly-
ethylene per year. Poly-America's state-of-the-art qual-
ity control laboratory assures the finest quality poly-
ethylene liner available.

Recra Environmental, Inc.
10 Hazelwood Dr., Suite 106
Amherst, NY 14150 716/691-2600
Recra Environmental, Inc. is an independent firm which
provides industry and government with consulting and
testing services in the area of chemical waste analysis,
prevention and control. Waste minimization programs,
chemical control management, site and plant assess-
ments, real property evaluations coupled with environ-
mental analysis and waste characterizations are a por-
tion of services offered.

Resource Analysis, Inc.
P.O. Box 4778
Hampton, NH 03842 603/927-7777
Resource Analysis and its affiliates provide comprehen-
sive environmental testing service to industrial and com-
mercial clients and to all levels of government. Special-
ties include organic chemistry using IR, GC and GC/MS
analytical methods; inorganic and heavy metals chem-
istry; freshwater and marine aquatic toxicology; and
field sampling. The laboratories occupy a 10,000 sq. ft.
facility with a staff of 25 professionals.

Resources Conservation Co.
3101 NE Northup Way
Bellevue, WA 98004 206/828-2400
Resources Conservation Co. (RCC) is an environmental
systems engineering firm specializing in resource
recovery and environmental reclamation technology.
RCC provides complete services including laboratory
testing, process development, design engineering and
turnkey construction. RCC's 16 years of experience has
led to the development of several patented processes
such as B.E.S.T.™ for hazardous sludge processing;
the Brine Concentrator for water recovery and waste re-
duction. Dow Chemical's AquaDetox is also available
for air/steam stripping of volatile pollutants from waste-
water.

Rexnord Environmental Management (REM)
5103 W. Beloit Rd.
Milwaukee, WI53214 414/643-3641
Rexnord Environmental Management (REM), formerly
EnviroEnergy Technology Center, is a unit of Rexnord
Technologies with 20 years of extensive experience with
in-plant environmental assessments, laboratory analyti-
cal and bench treatability studies, pilot plant systems
and operations utilizing the Treatment Van of the 90's,
underground tank management/closure programs, and
full-scale design and operation. Services also include
SPCC plans, contingency plans and right-to-know pro-
grams.

Rollins Environmental Services, Inc.
P.O. Box 2349
One Rollins Plaza
Wilmington, DE 19899 302/479-2768
Incineration, PCB management and destruction, on-and
off-site cleanup, remediation, monitoring, secure land
disposal, chemical, physical, biological treatment, deep
well injection, consultation, analysis, engineering,
evaluation, classification, collection, packaging, trans-
portation, storage, Lab-pak service, energy recovery,
contract and liability protection.
Roy F. Weston, Inc.
Weston Way
West Chester, PA 19380
215/692-3030
Managers of major environmental projects including fa-
cilities design, construction, operation, remedial ac-
tions, permitting, closures, investigations and analytical
chemical analyses.

SAIC
8400 Westpark Dr.
McLean, VA 22102 703/821-4600
SAIC is an employee-owned company principally in-
volved in the application of scientific expertise to solve
complex technical problems. SAIC offers a full range of
environmental consulting services including hazardous
waste site investigations and feasibility studies, waste-
water engineering, analytical chemistry, groundwater
modeling, risk assessments, regulatory compliance and
specialized laboratory studies.

Sevenson Containment Corporation
2749 Lockport Rd.
Niagara Falls, NY 14302 716/284-0431
Sevenson Containment Corporation provides turnkey
services to government and industry in the areas of haz-
ardous waste management and waste site cleanup. Sev-
enson's full-service capabilities include: site restoration;
secure landfill constructions; slurry wall & trench con-
struction; sludge solidification & fixation; waste
recovery and treatment; drum removal and waste ex-
cavation; transportation and disposal.

Shirco Infrared Systems, Inc.
1195 Empire Central
Dallas, TX 75247 214/630-7511
Shirco, Inc. incineration systems featuring the use of in-
frared heating and conveyor belt transport of waste
material through an efficiently insulated, modularly
constructed waste disposal system. Since no fossil fuel is
required, the reduced gas flow is economically treated to
meet requisite emission standards. Systems are excellent
for intermittent operation and have transportable capa-
bility. Shirco Portable Pilot Test Unit is available for
on-site testing at your facility.

Skolnik Industries, Inc.
4601 W. 48th St.
Chicago, IL 60632-4896 312/735-0700
Manufacturer of hazardous material containers from 8
to 85 U.S. gallon capacities, including "Big Mouth,"
the 85-gallon salvage drum and "Quad Pak," the assort-
ment of 4 nested containers. Also, the Security Drum ac-
cessories.

Solinst Canada Ltd.
2440 Industrial St.
Burlington, Ontario L7P 1A5 416/335-5611
Manufacturers of high quality groundwater monitoring
instrumentation, including the Waterloo Multilevel
System and water level meters. The Waterloo System is a
simple, modular system which uses unique, self-activat-
ing packers in a flush-jointed casing string. The packers
are used to isolate multiple, discrete locations within a
single borehole for sampling and pressure monitoring.
Soiltest, inc.
P.O. Box 931
Evanston,IL 60204 312/869-5500
Soiltest offers a large variety of equipment to the envir-
onmentalist for sampling and testing soils and ground-
water. Featured at our booth are stainless steel samplers,
bailers, hand-held instrumentation and the Ainlay Tank
'Tegrity Tester™. This new tank tester provides the op-
erator with the accuracy needed at an affordable price.

Stock Equipment Company
16490 Chillicothe Rd.
Chagrin Falls, OH 44022 216/543-6000
Hazardous waste compaction equipment including a
1,500-ton supercompactor for 55 gal drums • Hazardous
waste cement solidification equipment (manual and fully
automatic) • Hazardous waste incineration systems •
Hazardous waste dewatering equipment • We build
equipment to print (single component or entire systems
including controls) • Leading supplier of nuclear waste
solidification and processing equipment - Design and
fabricate remote manipulators for special purpose use •
Design and fabricate tank and waste processing systems.
TAMS Consultants, Inc.
655 Third Ave.
New York, NY 10017
212/867-1777
TAMS, a leading international engineering and scientific
firm, offers comprehensive services in solid and haz-
ardous waste management. Capabilities include: RI/FS;
Health/Safety; Risk Assessment; Community Relations;
Remedial Design; Construction Oversight; Site Closure;
Waste Geotechnics; Chemical/Process Design; Water-
shed Management; Hydrogeology/Mathematical Mod-
eling. TAMS provides services to clients in government,
military and private sectors through offices in major
cities.

Target Environmental Services
8940-A Rte. 108
Columbia, MD 21045 301/992-6622
Soil gas surveys will locate and define areas of subsur-
face contamination as the first step in a site investigation
or cleanup project. They are fast, accurate, low cost and
low risk, and result in a detailed report and map of the
extent, composition and concentrations of subsurface
contaminants. Target was a pioneer in the development
and application of soil gas surveys in the environmental
field, and continues to be a leader in the use of this tech-
nology throughout the United States.

Tennessee Valley Authority
2D44 Old City Hall Bldg.
Knoxville, TN 37902 615/632-7421
The Tennessee Valley Authority is a multi-purpose fed-
eral agency. Its broad mandate includes agricultural, in-
dustrial and economic development, and natural re-
source development and conservation. TVA's Waste
Management Institute supports and enhances private in-
dustry and government through environmental assess-
ment, feasibility evaluation, commercialization and in-
formation transfer.

Tetra Tech, Inc.
(A Honeywell Subsidiary)
630 N. Rosemead Blvd.
Pasadena, CA 91107 818/449-6400
Tetra Tech provides water resource and hazardous
material management services for industrial, institu-
tional and governmental clients throughout the U.S.
The firm's hazardous waste management services in-
clude: materials management, waste characterization,
emergency response, risk assessment, feasibility studies,
1987 EXHIBITORS' LIST 423
-------
geohydrologic investigations, SPCC/contingency plans,
health and safety training.

Thermo Environmental Instruments. Inc.
108 South St.
Hopkinton, MA 01748 617/435-5321
Manufacturer of instruments both portable and sta-
tionary to measure organic toxic chemicals in air, water
and soil. Also manufactures ambient and stack monitor-
ing systems. On display will be instrumentation for the
determination of trace organic materials, including
equipment for on-site measurements at hazardous waste
sites, leaking underground storage tanks and field
analysis of PCBs both in soil and water.

U.S. Pollution Control, Inc.
2000 Classen Ctr., Suite 400 S.
Oklahoma City, OK 73106 405/528-8371
For two decades USPCI, Inc., has been a leader in en-
vironmental management. Through its subsidiaries
—U.S. Pollution Control, Inc., PPM, Inc., and
Hydrocarbon Recyclers, Inc.—a complete package of
services is offered including: transportation, recycling,
and resource recovery, treatment, hazardous waste
disposal and remedial action.

United Resource Recovery/Solldltech. Inc.
99 Detering, Suite 230
Houston, TX 77007 713/864-1983
United Resource Recovery, Inc. (URR) was incor-
porated in Sept. 1982 for the purpose of developing geo-
logical repositories, particularly salt, for the disposal of
industrial hazardous waste materials. Acquired in
part by GECOS, a wholly owned subsidiary of GTM-
ENTROPOSE, the technology utilizes solution mining
techniques to create engineered caverns before using a
proprietary chemical fixation process (URR1CHEM) to
solidify organic and inorganic industrial waste
byproducts prior to final disposal. In Dec. 1986. URR
was issued all necessary EPA and State of Texas
operating permits to begin construction.
Solidilech, Inc. is a wholly owned subsidiary of
GECOS, Inc., formed as a field service extension of
URR to focus specifically on the growing national need
for hazardous site remediation and cleanup, especially
organic waste contamination. The corporation em-
bodies a turnkey management to environmental prob-
lems, from initial survey and analysis through separa-
tion, treatment, solidification and disposal.
Unocal Chemicals Division
376 South Valencia
Brea, CA 92621 714/528-8142
Unocal's Applied Technology Group offers licenses for
Unipure process technology. Unipure Process Technol-
ogy was developed by Unocal for the removal of heavy
metals from industrial waste waters. Unipure is guaran-
teed to make the user compliant with relevant heavy
metal discharge limitations. Unocal's Applied Technol-
ogy Group offers design engineering and complete turn-
key engineering services for waste water generators. In
addition, on-going services such as operator training,
compliance monitoring and consultation are offered.

VNR Information Servkes
(Div. Van Nostrand Reinhold Co., Inc.)
115 Fifth Ave.
New York, NY 10003 2127254-3232
CHEMTOX™ DATABASE—A cost-effective, easy-
to-use, and portable means to access data on over 3,600
hazardous chemicals. Books, journals and newsletters
on hazardous chemicals and industrial toxicology.

Vanghan Company, Inc.
364Monte-ElmaRd.
Montesano, WA 98563 206/249-4042
Vaughan Company, Inc., manufactures severe-duty
chopper pumps, lagoon pumpers and diesel "mini-
dredges' designed to pump thick concentrations of
highly abrasive or corrosive material without plugging.
Patended designs using ductile cast iron and special alloy
steels allow Vaughan pumps to effectively handle the
toughest hazardous waste pumping applications

Viking Pnmp-Hondallle
406 State
Cedar Falls, IA 50613 319/266-1741
Manufacturer of positive displacement rotary pumps
made of cast iron, steel & stainless steel. Newest product
is a line of "sealless" pumps, magnetically driven for
"no-leak" pumping of hazardous liquids.

Waste Conversion, Inc.
2951C Advance Ln.
Colmar, PA18915 215/822-2676
Waste Conversion, Inc. offers services including clean-
up, consulting, oil reclamation, transportation, analysis,
treatment and disposal of hazardous and non-hazardous
materials. Support services include a complete transpor-
tation fleet comprised of vacuum tankers, box and
dump trailers, permitted to transport hazardous and
non-hazardous materials throughout the eastern half of
the country.
Wehran Engineering
666 E. Main St.
P.O. Box 2006
Middletown, NY 10940 914/343-0660
Wehran, an environmental consulting firm of engineers
and scientists, is recognized as one of the top firms na-
tionally in the field of solid and hazardous waste man-
agement. Located in Middletown, New York, with
branch offices western New York, New Hampshire,
Massachusetts, Indiana, Vermont and Buenos Aires,
Argentina, Wehran employs 150 professionals who pro-
vide a multi-disciplinary approach to developing in-
novative and cost-effective solutions to waste manage-
ment problems.
WUton Laboratories
525 N. 8th. P.O. Box 1884
Salina, KS 67401 913/825-7186
Full service analytical laboratory specializing in environ-
mental monitoring and the analyses of hazardous waste
samples. Expertise includes GC/MS, GC, HPLC and
Industrial Hygiene.
York Laboratories Div. YWC lac.
200 Monroe Tnpk.
Monroe. CT 06468
203/261-4458
A multidisciplinary environmental laboratory providing
air, soil, water and waste characterization for full-pro-
file site assessments. York Laboratories is a participant
in EPA's Superfund Contract Laboratory Program.
Additional YWC services include contract operation of
wastewater treatment facilities and interim sludge de-
watering services. Laboratory facilities are in Whippany,
NJ and Monroe, CT. with other offices in Atlanta,
Boston and Houston.

Zotprolne,
Military Rd.
Rothschfld, Wl 54474 715/359-7211
Zimpro Inc. provides treatment systems and services for
hazardous and toxic wastes. Technologies include
PACT™ treatment systems and package plants, wet air
oxidation systems and pre-engineered skid-mounted
units, combustion, filtration. Services include labora-
tory analysis, audits, treatability studies, design and
construction, startup and operation.
424 1987 EXHIBITORS' LIST
-------

-------
    PROCEEDINGS OF THE
 NATIONAL CONFERENCE ON
   HAZARDOUS
   WASTES AND
   HAZARDOUS
    MATERIALS
March 16-18,1987 • Washington, DC
          AFFILIATES
     U.S. Environmental Protection Agency
   Hazardous Materials Control Research Institute
        Department of Defense
  Agency for Toxic Substances and Disease Registry
    National Environmental Health Association
        National Lime Association
       U.S. Army Corps of Engineers
     Association of Engineering Geologists
      American Society of Civil Engineers
   National Solid Waste Management Association
        U.S. Geological Survey

-------
                                                        Preface
    The RCRA Amendments of 1984 (the new RCRA) have radi-
    cally  changed hazardous  waste  management  throughout
    the country and considerably improved the control of hazar-
dous wastes. Whereas it was estimated that the old RCRA pro-
visions would cost the regulated industry between $1 billion and
$3 billion per year, the new RCRA amendments, when completely
implemented and fully in effect, may cost as high as $20 billion
per year.
  The new RCRA includes  58 congressionally mandated sta-
tutory deadlines that go into effect by  1987 or early 1988.  Con-
gress wrote, into  the statute, provisions that are detailed and
directive. The bill  does not wait for EPA to write regulations or
issue  guidance; provisions go into   effect  automatically  on
prescribed dates.


Land Application Management
  The law establishes a hierarchy of management practices. There
is a very strong presumption against the disposal of hazardous
wastes on the land and a very strong preference for treatment and
destruction of hazardous wastes.


Site Remediation
  The bill  provides EPA with "corrective action authority" at ex-
isting hazardous waste  facilities which are very similar to the en-
forcement  authorities under Superfund. The bill also closes the gap
in the existing program by expanding the regulated community.


Leaking Underground Tanks
  A major new subtitle of the legislation dealing with leaking
underground storage tanks went into effect on May 8, 1985. This
section probably affects two to five million underground  tanks
across the U.S. including tanks storing petroleum products and
gasoline as well as very complex tank storage systems storing
hazardous chemicals. Funding for this to the tune of $500 million
was appropriated as part of SARA '87.
  As  of May 8, 1985, there is  a ban on new tanks  that are not
designed to prevent releases due to corrosion or structural failure.
An estimated 100,000  new tanks  must meet these requirements.
By May 1986, a nationwide registration program will  require state
notification of the age, size, type and location of the tank, as well
as its uses.

Burning Hazardous Wastes
  The  provision concerning burning hazardous wastes and the
blending of hazardous wastes into fuel requires notification of the
fuel user that they are receiving hazardous waste. Some cannot be
burned in residential and commercial boilers.

RCRA Permit Program
  The  bill makes permitting of a hazardous waste facility far
more difficult than it has been.  Permit applicants must now sub-
mit exposure information on the potential for public exposure to
hazardous substances from landfills and surface impoundments.
This exposure information will be used to write new permit condi-
tions. EPA  has the authority  to write any permit condition
necessary to protect human health and the environment,  inde-
pendent of whether or not regulations are in place  for that pur-
pose.
  Congress has also set some very strict deadlines for the issuance
of permits. Final  determination on all land disposal permit ap-
plications must be made by November 1988.
  By November 1986, all incinerator applications must be submit-
ted and EPA must make  final decision on these by November
1989. All remaining applications have to be in by November 1988,
and EPA must make a decision on these by November  1992.

Other RCRA Amendments
  There are  many other provisions of the new RCRA. There is a
set of provisions that deals with non-hazardous solid waste and
small quantity generator wastes disposed of in landfills and sur-
face impoundments.  The bill also provides for federal enforce-
ment of the  requirements for lagoons. There is a series of waste
minimization requirements for generators to certify that they are
doing everything economically  feasible to reduce the amount of
wastes generated. There are requirements  for listing additional
hazardous wastes. Delisting of wastes is much more difficult and
complex under the new bill. New tests for ascertaining toxicity are
included in the new RCRA.
  The details of these important areas plus others are described
within the contents of these HWHM Proceedings.

-------
                         Acknowledgement
  The National Conference and Exhibition on Hazardous Wastes and Hazardous Materials
required dedication and talent from many individuals and commitment from a number of
organizations. We gratefully express our thanks and appreciation to the following conference
affiliates for all their contributions to a successful conference.

U.S. Environmental Protection Agency
Hazardous Materials Control Research Institute
Department of Defense
Agency for Toxic Substances and Disease Registry
National Environmental Health Association
National Lime Association
U.S. Army Corps of Engineers
Association of Engineering .Geologists
American Society of Civil Engineers
National Solid Waste Management Association
U.S. Geological Survey

  We also wish to express our gratitude to all of these knowledgeable individuals for their
advice and guidance in planning and producing a highly effective and informative program:

Wayne Adaska, Portland Cement Association
Hal Bernard, Hazardous Materials Control Research Institute
Mary Ann Curran, U.S. Environmental Protection Agency/Hazardous Waste Engineering
  Research Laboratory
Andre Dupont, National Lime Association
Terry Johnson, National Environmental Health Association
Loren Lasky, Association of Engineering Geologists
Denny Naugle, Department of Defense
Suellen Pirages, NatioaatSolid Waste Management Association
Stephen Ragone, U.S. Geological Survey
Jerry Steinberg, Hazardous Materials Control Research Institute/Water and Air Research
Andres Talts, American Society of Civil Engineers/Defense Environmental Leadership Prp/ect
Joan  Warren, U.S. Environmental Protection Agency
Bob Williams, Agency for Toxic Substances and Disease Registry

  Producing a document of the magnitude of this proceedings requires a highly skilled team,
much cooperation and communication, and a tremendous amount of effort by all involved.
We are fortunate to have such .a team and would like to convey our special thanks to Dr. Gary
Bennett, Professor of Biochemical Engineering, The University of Toledo, and Hal Bernard,
Executive Director, HMCRI, for the excellent editing; to the typesetters, proofreaders,  and
graphic artists who completed a tremendous amount of work in an incredibly short period of
time;  and to the staff of HMCRI for coordinating the  myriad details and activities of this
conference.

-------
                                                   Contents
    ALTERNATIVE CONTAMINATION LIMITS

A Cost-Effective Approach to ACL Development	1
  Alfred C. Leonard & Robert H. Clemens
Alternative Contamination Limits for the
Natural Environment	4
  Michael J. Donate & Andrew Baldwin
Alternate Concentration Limits—The Link
Between CERCLA and RCRA	7
  Shawn L. Sager, Ph.D. & Paul C. Chrostowski, Ph.D.
Alternative Concentration Limits Under the
RCRA Program	12
  Vernon B. Myers, Ph.D.,  Mark A. Salee &
  Jerry Gorman
     LIABILITY/INSURANCE/DEREGULATION

An Overview of Liability Insurance Under RCRA  	14
  Jackie Tenusak & Paul E. Bailey, J.D.
Small Quantity Generator Liability and
Regulatory Compliance	19
  Robert Deyle & Rosemary O'Leary,  J.D.
Financial Liabilities for Natural Resource Damages	26
  Stephen  Wyngarden & Michael Goldman
Generalized Risk-Based Decision-Making for
Hazardous Waste Sites  	32
  Charles L. Vita, Ph.D., P.E.
       CONTAMINATED AQUIFER CONTROL

Development of an In Situ Biodegradation
Process for Remediation of a Gasoline Spill	
  J. V. Lepore, D.S. Kosson, Ph.D.,  & R.C.
  Ahlert, Ph.D., P.E.
Pilot Interceptor Drain Performance:  A Method for
Remedial Investigation and Aquifer Evaluation	
  John R. Mildenberger & Brian V. Moran, P.E.
The Use of Extraction Trenches to Remove
Industrial Solvents from Shallow, Low
Permeability Alluvial Aquifers	
  Lance D. Geselbracht, P.E., Thomas A. Donovan,
  P.E. & Richard J. Greenwood, P.E.
Benefits of Mitigating Releases from Underground
Gasoline Tanks	
  Donald W. Anderson
.38
 44
.49
                   MODELLING GROUNDWATER &
                           SURFACE WATER

         Use of the HELP Model in Evaluating the Cover
         Design for a Uranium Mill Tailings Disposal Site	
            Will Wright, A. Keith Turner, Ph.D. &
            C.E. Kooper
         Transport of Mercury in the North Fork Holston River
           David R. Cogley, Ph.D. & Neil Ram, Ph.D.
         Geologic and Hydrogeologic Characterization of a
         Hazardous Waste Disposal Site, Arlington, Oregon	
           Stephen M.  Testa & Frederick G.  Wolf
           INNOVATIVE TECHNOLOGIES

Waste Treatment and Reduction Applications of
Freeze Vaporization	
  Val Party ka
The Treatment of Solvent-Contaminated Waste
Using Liquefied-Gas Extraction	
  William E. McGovern & John M. Moses
Design Innovations Employed in the U.S. Air Force
Live Fire Training Facility	
  Bryce E. Mason
Composting Explosives Contaminated Soil	
  Richard C. Doyle, Ph.D., Jenefir D. Isbister, Ph.D.,
  George A. Anspach, David Renard & Judith
  F. Kitchens, Ph.D.
Use of Vapor Extraction Systems for In Situ
Removal of Volatile Organic Compounds from Soil	
  Magnus B. Bennedsen, P.E.,  Joseph P.  Scott, Ph.D.
  & James D. Hartley, P.E.
Macroencapsulation: New Technology Provides
Innovative Alternatives for Hazardous Materials
Storage, Transportation, Treatment and Disposal  	
  Mark D. Shaw
Photozone Destruction of Cyanide Waste at Tinker
AFB, Oklahoma Pilot Plant Results	
  Martin F. Herlacher & F. Robert McGregor, D.Sc.
                                                    .58
                                                    .64
                                                    .69
.77


.80


.84

.86





.92





.96
                                                            .101
.54
                    TREATMENT

The Treatment of Contaminated Groundwater and
RCRA Wastewater at Bof ors-Nobel, Inc	
  John A. Meidl & Ronald L. Peterson
Innovative Electromembrane Process for
Recovery of Lead from Contamianted Soils	
  E. Radha Krishnan,  P.E. & William F. Kenner
                                                             .106
                                                             .111

-------
Dewatering and Solidification/Stabilization
of Industrial Waste
Mark L. Allen & Steven D. Liedle
Effectiveness of In Situ Biological Treatment of
Contaminated Groundwater and Soils at
Kelly Air Force Base, Texas
Roger S. Weizel, Donald H. Davidson, Connie M.
Durst & Douglas J. Sarno
On-Site Destruction of Organic Contaminants
in Water
D.G. Hager & C. C. Loven
Waste Management Plan for Veterans
Administration Medical Centers
J.I. Bregman, Ph.D., Edward Findley & Wayne
Warren
Fixation of Metallic Ions in Portland Cement
Muhammad S. Y. Bhatty, Ph.D.
Mineralization of Recalcitrant Environmental
Pollutants by a White Rot Fungus
John A. Bumpus & Steven D. Aust
The Effect of Volatile Organic Compounds on the
Ability of Solidification/Stabilization Technologies
to Attenuate Mobile Pollutants
James H. Kyles, Kenneth C. Malinowski, Ph.D.,
Judith S. Leithner & Thomas F. Stanczyk
Treatment Technologies for Dioxin Wastes
M. Pat Esposito
Stabilization/Solidification of Hazardous Wastes
M. John Cullinane, Jr., P.E., R. Mark Bricka
& Larry W. Jones, Ph.D.

LAND DISPOSAL

Construction of Multiple-Lined Hazardous
Waste Landfills
Michael G. Ruetlen, P.E., Philip P. Sleeker, P.E,
& Wendell W. Lattz
An Improved Protective Cover Design for
Hazardous Waste Landfills
G. Stephen Mason, Jr. & Robert T. Pyle
Common Problems Associated with the Design and
Installation of Groundwater Monitoring Wells
David M. Nielsen, C.P.G.
Determination of the Capacity of Landfill Site
Soils to Attenuate Leachate Components
Michael L. Crasser
Construction Quality Assurance for Hazardous
Waste Land Disposal Facilities with Emphasis on Soil
Barrier Layers and Final Multilayer Cover Systems
Richard C. Warner. Ph.D. & Nathaniel Peters
MONITORING & SAMPLING

Comparative Analysis of Soil Gas Sampling Techniques
Peter P. Jowise, Jeff D. Villnow, Lazar I.
Gorelik & John M. Ryding
Comparison of In Situ Soil Pore Gas Screenings, Soil
Sample Head Space and Laboratory Techniques
John E. Adams
Screening for Characterization of PCB-Conlaining
Soils and Sediments
Bruce A. Fowler & Joseph T. Bennett
.118

.123

.129

.134

.140

.146

.152

.158

.167
.169
.174
.178
.185
.191
.193
.200
208
214
Performance of the Toxicity Characteristic
Leaching Procedure
Lynn R. Newcomer, W. Burton Blackburn
& Todd A. Kimmell
Sampling Strategies for Site Evaluation
Jeffrey C. Myers & Rex C. Bryan
RCRA SITE REMEDIATION

RCRA and Its Implications for CERCLA 215
Stephen M. Smith
Development of Engineering Design Strategies for
Remediation and Retrofitting of Existing
Waste Disposal Sites 219
Thomas J. O'Brien & Benjamin G. Siebecker
RCRA Corrective Action at a Large Chemical
Manufacturing Facility; Process and Results 221
George A. Furst, Ph.D.
The Remedial Investigation Process:
An Industry Perspective 228
James E. Gagnon, P.E.
Radial Wells and Hazardous Waste Sites 232
Wade Dickinson, R. Wayne Dickinson &
Peter A. Mote

SITE MANAGEMENT & CLOSURE
.204
Closure of a Hazardous Waste Landfill
Incorporating a Leachate Collection System
Thomas A. Kovacic & Keith C. Mast, P.E.
Effective Closure of Surface Impoundments
Requires a Multi-Phase Plan
Albert K. Langley, Jr.
A Successful Hazardous Waste Landfill Siting—
Maryland's Experience
Thomas D. McKewen & Anne C. Sloan

INCINERATION
Permitting for Hazardous Waste Incinerators—
Preparation of the Trial Burn Plan
Robert P. Newman, P.E. & Arthur B. Nunn, III
Planning and Design for On-Site Incineration at
Two Illinois Hazardous Waste Sites
James F. Frank
Full-Scale Rotary Kiln Incinerator Field Trial:
Phase I, Verification Trial Burn on Dioxin/
Herbicide Orange Contaminated Soil
Major Terry L. Stoddart & Jeffrey J. Short

RISK ASSESSMENT
The U.S. Environmental Protection Agency's
Guidelines for Risk Assessment
Peter W. Preuss, Ph.D., Alan M. Ehrlich, Ph.D.
& Kevin G. Garrahan, P.E.
Comparative Risk Assessment of Sources
Of Groundwater Contamination
Hope Pillsbury
Application of Expert Systems for Environmental
Engineering Decisions
Donald R. Brenneman
Communicating Risk Assessment Findings to the
Public: Approaches that Don't Work and
One that Might
Robin Sandenburgh & Marion Cox
.238
.243
.247
.252
.256
.260
.265
.273
.278
.281

-------
               UNDERGROUND TANKS
Tank Testing Method for Detecting Leaks in
DOD's Large Underground Storage Tanks	284
  L. Peter Boice, H. Kendall Wilcox, Ph.D.
  & Douglas E. Fiscus, P.E.
Safety Procedures for Testing and/or Removing
Underground Storage Tanks	291
  Fred Halvorsen, Ph.D.,  P.E.
Underground Storage Tank Corrective Action
Technologies	296
  Douglas C. Amman & S. Robert Cochran
Compliance Strategy for Underground Tanks:
A Cost Assessment	301
  Alan Lamont, Ph.D. & James D. Hartley, P.E.


               WASTE MINIMIZATION

Minimization of Industrial Oil Wastes  	307
  Franco E.  Godoy & Douglas L.  Hazelwood
Developing a Minimization Strategy Effective in
Reducing Metal Hydroxide Sludge	311
  Thomas F. Stanczyk
Small Generator Cooperative Effects
Economical Recycling	316
  M.E. Malotke
Financial Analysis of Waste Management
Alternatives	318
  Richard W. MacLean
Waste Reduction in the Semiconductor Industry	324
  Steve Pedersen & Mary Ann Keon
Department of Defense Hazardous Waste
Minimization 	328
  Capt. Michael J. Carricato, USN, Andres Tails,
  P.E., Joseph A. Kaminski, P.E. &  Thomas E.
  Higgins, Ph.D., P.E.
Reducing the Generation of Hazardous Waste:
Actions by Government and Industry	332
  Joel S. Hirschhorn, Ph.D. & Kirsten U. Oldenburg
Hazardous Waste Minimization by the U.S. Navy	335
  Thadeus J. Zagrobelny,  P.E.
Comprehensive Model for Hazardous Waste
Management Alternatives	340
  William M. Sloan, Jean Tilley & Stephen W.
  Bailey, P.E.
Characterization and Segregation of Waste Oils,
Solvents and Fuels at Naval Installations	345
  Alan J. Kaufman & Kendall M.  Jacob, P.E.
Faster Sample Throughput for Environmental
Analyses	350
  James A. Poppiti, Ph.D.
Hazardous Waste Reduction—Cornerstone of a
Waste Handling Strategy	
Robert P. Bringer, Ph.D.

          MINING & INDUSTRIAL WASTES

Institutional Realities of Locating a Repository
for Low-Level Radium Waste in Colorado	
  H. Donald Ulrich, Joan van Munster, P.E.
  & John Brink
Heap and Dump Leaching and Management Practices
to Minimize Environmental Impacts	
  Robert L. Hoye, Robert L. Hearn & S.
  Jackson Hubbard
Abandoned Steel Manufacturing Site—
A Case History	
  G.J.  Anastos, Ph.D., P.E., T.J. Legel, P.E.
  & J.M. Perdek

              HEALTH & ASSESSMENT
.352
.356
.368
.374
Ocean Disposal Risk Assessment Model	
  Joseph G. Karam & Martha J. Otto
Human Exposure Estimates Using U.S. EPA
Guidelines Models: An Integrated Approach	
  George J. Schewe, Joseph Carvitti & Joseph Velten
A Multimedia Exposure Assessment Model for
Evaluating Land Disposal of Hazardous Waste	
  Atul M. Salhotra, Ph.D., David R. Gaboury, P.E.,
  Peter S. Huyakorn, Ph.D. & Lee Mulkey
OSHA Standards or Risk Analysis: Which Applies?	
  Arthur D. Schatz & Michael F. Conway, P.E.
                 STATE PROGRAMS
A Waste Reduction Program and Assessment of
Current Status for Illinois	
  David L. Thomas, Ph.D., Daniel D. Kraybill,
  P.E. & Gary D. Miller, Ph.D.
The Effect of Federal and California
Regulations on Firms' Decisions to Minimize
Waste in California	
  Barry Garelick, Julia Gartseff & Michael Neely
Siting Hazardous Waste Faculties in New
York State	
  Peter A. Marini, P.E.
Impact of State Environmental Laws on
Property Acquisitions	
  Daniel K. Moon, Robert H. ClemensA&
  Diane P Heineman
.378
.386
.391
.398
.402
.407
.411
.417
1987 Exhibitors'	419

-------
A Cost-Effective Approach to ACL Development
Alfred C. Leonard
Robert H. Clemens
Alliance Technologies Corporation
Bedford, Massachusetts
ABSTRACT
This paper presents a method for developing Alternate Con-
centration Limits (ACLs) as allowed in RCRA under 40 CFR
Part 264.94. Concentration limits serve as thresholds, beyond
which corrective action will be required. Because less stringent
concentration limits would reduce the need for costly corrective
action, a successful ACL application may have significant mone-
tary benefits. On the other hand, an ACL demonstration can em-
ploy costly data gathering and analytical efforts which may be
wasted if the application is rejected or if less costly procedures
would have sufficed. The methodology presented here features a
sequential decision making process designed to minimize the costs
of unnecessary analyses.

INTRODUCTION
According to the groundwater protection regulations for per-
mitted RCRA facilities (40 CFR Part 264, Subpart F), a per-
mitted RCRA facility is normally required to initiate corrective
action if monitoring detects concentrations of hazardous constit-
uents that are higher than the background levels or drinking water
standards. The detection monitoring program, required in 40
CFR Part 264.98, samples at the point of compliance, immed-
iately downgradient from the edge of the unit. The concentra-
tion limits have been established to assure protection of human
health and the environment. However, the regulations (40 CFR
264.94), allow that Alternate Concentration Limits (ACLs) may
be granted for a site upon demonstration that, due to natural or
man-made capacity of the site to reduce risks to human health
and the environment, these alternative limits will provide suffic-
ient protection.
The methodology presented in this paper consists of three
steps: (1) screening, (2) conceptual design and preliminary es-
timate, and (3) ACL demonstration.

SCREENING
The screening step in an ACL application is required to deter-
mine whether the circumstances at the site are such that an ACL
application is feasible. A decision not to continue could be based
upon a judgment that the site is not suited to ACLs, or a decis-
ion that the cost of an ACL application will exceed the cost of
corrective action. The screening procedure will vary from site to
site, but it will generally include the following elements: (1) re-
view of the site data for quality and completeness; (2) review of
hazardous constituents to determine the difficulty of characteriz-
ing health or environmental risk; (3) review of the reasonable-
ness of the fate and transport hypotheses upon which the ACL
demonstration would be based; and (4) review of recent Federal,
Regional and (if appropriate) state ACL permit application pol-
icies. After each of these screening procedures, a decision of
whether or not to continue will be made.

Review Site Data
At the beginning of the screening, all of the appropriate site
data must be gathered and examined for quality and complete-
ness. This procedure may not constitute a true screening pro-
cedure, because the decision to continue or abandon the ACL
demonstration will rarely be made at this point. A reasonably
complete list of data to be considered is provided in the ACL reg-
ulations (at 40 CFR 264.94) and in ACL guidance materials. The
areas where data shortcomings are both common and critical to
continuation of the analysis are:
• Groundwater flow conditions, including soil type, hydraulic
conductivities, depth to water table, hydraulic gradients and
the effects of seasonal variations and large withdrawals
• Monitoring system design and data, including QA/QC pro-
cedures
• Contaminant identification, including all Appendix XIII con-
stituents known to have been placed in the facility and all that
have been detected in the monitoring wells
• Estimated cost of corrective action, in the chance that the ACL
application is not successful. This information is needed as a
point of reference for deciding whether or not to proceed with
the ACL demonstration

Availability of Toxicity Data for Constituents
A review of the known health hazards of the contaminants
found at a site is suitable as an early screening procedure be-
cause the absence of such information is easy to determine and
might make an ACL application prohibitively expensive and time-
consuming. Moreover, the U.S. EPA may not accept ACLs for
constituents for which risk data are not available.
Suspected carcinogens for which the U.S. EPA has estimated
carcinogenic potencies are available from the U.S. EPA Carcino-
gen Assessment Group (CAG). The CAG calculates unit carcino-
genic risks based upon a rigorous evaluation of the body of avail-
able scientific information for a substance. Where a substance has
not been listed, existing data may be inadequate for a determina-
tion.
Non-carcinogens (i.e., systemic toxicants) have been character-
ized by the U.S. EPA's Environmental Criteria and Assessment
Office. That office has developed Oral Acceptable Daily Intakes
(ADIs) for many of the Appendix VIII constituents. These ADIs
are based No Observable Adverse Effect Levels (NOAELs) or
Low Observable Adverse Effect Levels (LOAELs). A "safety"
factor of usually 10, 100 or 1000 is applied according to the level
of confidence in the data. ADIs consider acute and chronic toxic
ALTERNATIVE CONTAMINATION LIMITS 1

-------
effects of chemicals. Any daily human exposure in excess of the
ADI is considered to be excessive.
  The most important screening procedure for ecosystem hazards
is a review to determine whether known endangered species habi-
tats might be exposed to the contaminants. This circumstance
could prevent the  ACL  application  from being accepted and
might justify abandonment of the analysis. Responses of aquatic
ecosystems to contaminant exposures can be screened by review-
ing the U.S. EPA  water quality criteria established for priority
pollutants. Where  no criteria exist, the  analytical requirements
may be more substantial.
 Fate and Transport Scenario Plausibility
   Screening for the plausibility of the fate and transport scen-
 ario requires information about contaminant loading character-
 istics,  ground water flow patterns,  contaminant fate and trans-
 port characteristics and potential human and ecosystem recep-
 tors. This information will be used to determine whether contam-
 inant concentrations will be  significantly  reduced  between the
 point of compliance and the receptor.
   The common fate and transport scenarios that are appropriate
 for ACL applications are:
 •  Demonstrations of no or low exposure due to discharge to a
   large surface water
 •  Demonstrations of no or low exposure due to discharge to an
   otherwise contaminated (Class HI) aquifer
 •  Demonstration of contaminant attenuation by dispersion, re-
   tardation or degradation
   If the appropriate scenario is either discharge to a surface water
 or discharge to a Class III groundwater, a qualitative review  is
 adequate for the screening step. However, if a successful ACL
 application depends upon contaminant attenuation, some rough
 calculations are in order.  A significant amount of dispersion
 usually requires a long distance between the facility and the recep-
 tor. The potential for significant degradation varies widely among
 constituents  and settings.  Hydrolysis, chemical oxidation and
 radioactive decay are reasonably well understood, but a defen-
 sible quantification of  a biodegradation  rate will usually require
 significant on-site testing.  Retardation behavior is better under-
 stood  than biodegradation; however, retardation alone is  less
 effective in preventing exposures because it may only serve to de-
 lay the transport of contaminant concentrations toward a recep-
 tor.
   The  regulations state clearly (40 CFR  Part 264.94{b)) that
 ACLs must protect against  "present or potential hazard  to
 human health and the environment," consequently the most crit-
 ical human receptor scenario is often a future water supply well
 located along the site property line in the direction of ground-
 water flow.
 Review of Applicable Permitting Criteria
   A review of recent ACL policies or application decisions in the
 same U.S. EPA Region will provide useful information about the
 criteria by which the applications will be judged. It may also pro-
 vide some insight to favored or disfavored arguments and analy-
 tical techniques. To date, very few ACLs have been granted na-
 tion-wide and acceptable  procedures  are not well established.
 However, it is likely that  with the evolution of  Federal guide-
 lines, that more successful applications will be made. At that
 time  acceptable  techniques and  arguments, and Regional and
 state  variations (if states are given  primacy for  ACLs) will  be
 more apparent.
CONCEPTUAL DESIGN AND
PRELIMINARY ESTIMATE
  The conceptual design of an ACL demonstration follows a suc-
cessful screening of the potential application. The result of the
conceptual design task is an outline of the elements of the ACL
demonstration and their associated schedules and costs. The pre-
liminary estimate is an execution of the conceptual design using
readily available information, conservative assumptions, and sim-
ple analytical techniques. In some cases, the preliminary estimate
may provide ACLs that are well below expected compliance point
concentrations making further analysis unnecessary.
  The two basic types of conceptual designs are: (1)  A predic-
tive demonstration that the constituent concentrations observed
at the point of compliance will be at acceptable levels by the time
they reach, a^receptor; and (2) A retrospective demonstration
that the current level  of contamination  is  acceptable and  the
source concentrations are diminishing, therefore the concentra-
tions currently observed at the point of compliance must be ac-
ceptable.

Predictive Demonstration
  A standard predictive demonstration will use detailed site in-
formation and defensible estimates of  contaminant  exposure
thresholds and transport  behavior to develop  ACL  estimates.
The elements of the ACL demonstration are not  described in this
paper, but a typical demonstration  will  contain the  following
elements:
• Site description—Land use, water use and users,  facility oper-
  ating characteristics, hazardous constituents which may be re-
  leased  on-site and hazardous  constituents  detected  in  the
  groundwater
• Hydrogeology—Regional and site geology, precipitation, sur-
  face water hydrology and groundwater hydrology
• Exposure pathways—Human exposures/ ecosystem exposures,
  endangered species and maximum allowable concentrations
• Contaminant transport analysis—Estimated reductions in con-
  stituent concentrations  as they are transported between  the
  point of compliance and each of the designated receptors
• Alternate concentration limit estimates

Retrospective Demonstration
  A retrospective demonstration may be  used at sites  where the
applicant can demonstrate that concentrations at both  the recep-
tor points and the point of compliance will not exceed current
levels. The applicant then demonstrates  that the receptors  are
not at risk with the current levels of contamination. Therefore,
the current concentrations are the proposed ACLs. This approach
is applicable to closed RCRA surface impoundments or cleaned-
up  CERCLA sites. It can avoid costly and uncertain contam-
inant transport analyses  in favor of current and historical con-
centrations at the receptors. However, the applicant may wish to
offer predictive models to explain an observed historical pattern
of receptor and compliance point concentrations.

Preliminary Estimate
  A preliminary estimate should be performed at this point to test
the conceptual model. This estimate can follow all of the steps
identified  in  the conceptual model, but worse case assumptions
can be used in the place of missing data or sophisticated model-
ing results. In most cases this estimate will identify the most crit-
ical of the receptors and allow the later analyses to be more
focused. The estimate will also serve as a baseline to which later
refinements in exposure and fate and transport  estimates can be
compared. In those instances where the worst case assumptions
      ALTERNATIV£ CONTAMINATION LIMITS

-------
are all defensible and the ACL estimates are acceptable to the
applicant (i.e., they are above current and expected concentra-
tion levels), the preliminary estimate may be sufficient for the
ACL application.

ACL DEMONSTRATION
The ACL demonstration should be viewed as a systematic re-
finement of the preliminary ACL estimate. The shortcomings of
the preliminary estimate (indefensible assumptions and/or overly
stringent ACLs) should be identified and resolved. The compon-
ents of an ACL demonstration will vary from site to site, but they
will commonly include additional hgdrogeologic investigations,
additional monitoring, fate and tcansport analysis and if eco-
system exposures are identified, ecosystem surveys.
An example of the cost decisions to be made in this step is
found in the selection of the approach toward the fate and trans-
port analysis, witH options ranging from a simple series of cal-
culations to a sophisticated numerical model. The critical ques-
tions here are how much improvement in the ACL estimate will
the more sophisticated approach achieve and at what cost? More
sophisticated models are often data-intensive and may require
additional hydrogeologic investigations to provide useful results.
Yet, because the ACL demonstration requires conservative
assumptions, those results may not differ significantly from the
results of a simpler approach.
If there is reason to believe that the desired ACLs cannot be
demonstrated, or that some of the necessary refinements will be
so costly or time consuming that corrective action would be pref-
erable, then the analysis should be abandoned.

CONCLUSIONS
An ACL demonstration can employ costly data gathering and
analytical efforts which may be wasted if the application is re-
jected or if less costly procedures would have sufficed. The
methodology presented features a sequential decision-making
process designed to minimize the costs of unnecessary analyses.
ALTERNATIVE CONTAMINATION LIMITS

-------
                        Alternative Contamination  Limits for  the
                                         Natural  Environment

                                                 Michael J. Donate
                                                  Andrew Baldwin
                                              E.G. Jordan Company
                                             Wakefield, Massachusetts
ABSTRACT
  Hazardous waste facilities regulated under RCRA are required
to incorporate design features and control measures into their fa-
cilities  to prevent the leaking of hazardous waste constituents
into ground water. For each hazardous constituent entering the
groundwater from a regulated unit, a concentration limit must be
established that will serve as a limit beyond which degradation
of groundwater quality will not be allowed. There are three pos-
sible concentration levels that can be used  in the groundwater
protection standard:
• Background levels of hazardous constituents
• Maximum concentration limits
• Alternate concentration limits (ACL)
  Although a great deal of data exist to support an ACL demon-
stration for assessing risk to human health, potential damage to
wildlife and vegetation in aquatic and terrestrial environments is
still in  the early stages of development. This paper discusses the
different approaches that can be used to develop ACLs for the
protection of the environment, in particular, aquatic organisms;
the paper also provides a scientifically sound methodology to do
so and presents a case study of a hazardous waste site showing
the application of this ACL methodology.

INTRODUCTION
  The  U.S. EPA has developed  guidance for Alternate Concen-
tration Limit (ACL) demonstrations. However, much of the ACL
approach is subject to reader interpretation and there is not much
concrete ACL methodology.
  As part of the ACL process, the initial step in assessing pos-
sible environmental impacts is to determine the probable exposure
pathways for hazardous  constituents to reach environmental re-
ceptors. Referred to as  the exposure assessment, this step in-
volves  examining the extent of  the hazardous contaminants in
various environmental media, the potential of migration of haz-
ardous constituents  and  the location of receptors and environ-
ments of concern. The exposure assessment will result in delinea-
tion of likely  exposure pathways.  Subsequent to the exposure
assessment, the toxicity and  bioaccumulation of hazardous con-
stituents to flora and fauna should be examined.  Organisms can
be exposed directly to contaminants through the assimilation of,
or contact with, contaminated groundwater discharge to surface
water bodies, or they can be exposed indirectly through food web
interactions. Toxicants can accumulate in exposed biota  in in-
creased levels that are lethal or have chronic effects. A compre-
hensive search  for toxicity and bioaccumulation values for the
ACL  constituents found in  the groundwater must be accom-
plished. Where information  is sparse  or nonexistent,  a more
thorough analysis of potential environmental impacts is needed.
  This paper describes the application of short-term methods to
estimate the chronic toxicity of effluents to  freshwater  organ-
isms (flora and fauna). The paper also shows how these bioassay
techniques are relatively inexpensive and technically sound and
could be used to support an ACL demonstration for a hazardous
waste site. Unfortunately, in the process of ACL development,
little or no data may exist on the hazardous constituents of con-
cern for a specific site. The methodology below will describe how
bioassay tests can enhance the ACL  process and protect the
natural environment at risk.
                                                 N
                         Figure 1
              Locus Map of the Picillo Farm Site'
     ALTERNATIVE CONTAMINATION LIMITS

-------
Table 1
Organic Pollutants Detected at the Picillo Site1

Category
Volatile
Aroanlie

Volatile
Organic*

Ketonec

Acid
Org.nic.

Compound
Benzene
Toluene
Xylenei
l,2-Diehlorc«r.h«n«
1,1,1-Trichloroethnne
l,l-Dichloroeth«ne
Ch oroethine
Ch oroforn
1, -Dichloroetttylene
HechyUne Chloride
FluorotricKloromf th«ne
Triehloroethylene
Cirbon Tetr«chloride
Acetone
Methyl Ethyl Ketone
Methyl Iiobutyl Ketone
Phenol
Z,ft-Di«ethyl Phenol

3/83
250
29.800
4,510
160
23.000
710
260
to
290
2,100
280
4,84
1,93
5.86
ND
9,080
2,920
890
1,800
680

970
8.300
56,000
ND
6,600
100
ND
NO
1.700
4 ,200
I,
-------
cnts and receiving waters are relatively inexpensive ($8,000-
$10,000/test)! and offer the site-specific results necessary for
ACL demonstrations at RCRA sites.
Table 3
Contaminants Detected in Surface Water and Groundwater Samples
at the Picillo Site1
Contanlnant
Sample Location
Bedrock Spring Unnamed Swamp Monitoring Well
1,1.1,-Trlchloroethane 540

Chlorofora 120

TrIchloroethane ND

1,1-Dlchloroethane ND
130

230
430

110

19
All value* in nlcrograns per liter (ppb)

ND- not detected
REFERENCES
1. CCA Corporation Technology Division, "Endangerment Assessment
and Feasibility Study," Picillo Site, Coventry, RH," Apr. 1985.
2. GCA Corporation Technology Division, "Alternate Concentration
Limit Guidance Case Study for a CERCLA Site—Picillo Farm, RI,"
Sept. 1985.
3. Springborn Bionomics, Inc., "Chronic Toxicity of Two Surface
Waters and One Groundwater to Daphnids (Ceriodaphnia dubia)
and Duckweed (Lemna gibba)," Sept. 1986.
4. U.S. EPA, "Short-Term Methods for Estimating the Chronic Toxic-
ity of Effluents and Receiving Waters to Freshwater Organisms,"
Dec. 1985.
5. GCA Corporation Technology Division, "Alternate Concentration
Limit for the Picillo Farm Site, RI," Aug. 1986.

Table 4
Summary of MATC, EC10, EC50 and EC90 values from static re-
newal toxicity tests with Ceriodaphnia dubia and Lemna gibba exposed
to surface water samples SW-1C and DH-1C and to groundwater sample
MW-27C. Ceriodaphnia were exposed for 7 days and Lemna were
exposed for 14 days.1
Test
Sample
SW-1C

DH-1C

MW-27C

Test
Statistic
MATC
EC10b
EC50b
EC90b
MATC
EC10b
EC50b
EC90b
MATC
EC10b
EC50b
EC90b
Ceriodaphnia
N.C.a(>100%)
>100% (— ,-)
>100% I—,-)
>100% (— ,-)
17%
16% «1%,33%)
24% (8%, 41%)
32% (16%, 50%)
71%
60% (25%,>100%)
80% I34%,>100%)
>100% (45%,>100%)
Lemna
N.C.a«6.2%)
<0% «0\,64%)
>100% (19%.>100%)
>100% ( >100% , >100%)
4.4%
<1% I—,-)
16% (— ,-)
>100% 1— ,-)
N.C.a(>25%)
12% «1%,>100%)
44% (2%,>100%)
>100% (9%,>100%)
aNot calculated because NOEC and/or LOEC were not determined.

b95\ confidence limits in parentheses.
ALTERNATIVE CONTAMINATION LIMITS
-------
Alternate Concentration Limits—
The Link Between CERCLA and RCRA

Shawn L. Sager, Ph.D.
Paul C. Chrostowski, Ph.D.
ICF-Clement Associates
Washington, D.C.
ABSTRACT
The process for determining appropriate cleanup criteria for
hazardous waste sites is based on a risk management strategy
designed to minimize future risk to human health and the en-
vironment. Cleanup levels are based on standards or criteria for a
population at an exposure point. These levels are then used to
determine the contaminant concentrations that can remain at the
site and that will result in acceptable concentrations at the ex-
posure points.
RCRA and CERCLA have established a process to determine
cleanup levels. The determination of alternate concentration
limits (ACLs) under RCRA and the regulatory basis for setting
ACLs will be discussed. The ease with which ACLs can be set will
be compared to the process established under CERCLA.

INTRODUCTION
The establishment of cleanup criteria at hazardous waste sites is
based upon a hazardous management strategy designed to
minimize future risk to human health, welfare and the environ-
ment. CERCLA and its successor, the Superfund Amendments
and Reauthorization Act of 1986 (SARA), and RCRA require
that hazardous waste which has contaminated the environment
must be cleaned up to a certain level. The basis for this cleanup is
the attainment of a target concentration based on standards or
criteria for a population at an exposure point and the determina-
tion of contaminant levels in environmental media at the site that
would enable target concentrations to be achieved at these points
of exposure.
The focus of this paper is on the determination of alternate con-
centration limits (ACLs) under RCRA and a comparison between
ACLs under RCRA and cleanup objectives under CERCLA/
SARA. The paper is divided into five sections: (1) the develop-
ment of cleanup criteria; (2) CERCLA guidance; (3) RCRA
guidance; (4) a case study using CERCLA guidance; and (5) a
comparison of the RCRA and CERCLA cleanup criteria pro-
cesses. It is concluded that the quantity of evidence required for
an ACL demonstration at a complex site is so large that it is pro-
hibitive to the process. The simpler, less data-intensive
methodologies used under CERCLA are suggested as an alter-
native.

DEVELOPMENT OF CLEANUP CRITERIA
The process of determining appropriate removal and/or
remedial actions is set by the NCP for Superfund sites and by
RCRA for hazardous waste treatment, storage and disposal
facilities. The level of cleanup is set either by reference to ap-
plicable or relevant and appropriate requirements (ARARs), by
action levels or by ACLs.
The applicable or relevant and appropriate requirements
(ARARs) considered to be potentially usable by the NCP' are:
drinking water maximum contaminant levels (MCLs) and max-
imum contaminant level goals (MCLGs), federal ambient water
quality criteria (AWQC), national ambient air quality standards
(NAAQs) and state environmental standards.
There are two types of remedial alternatives that can be devel-
oped during a remedial investigation/feasibility study (RI/FS) for
a Superfund site: source control and management of migration.
In assessing the ability of the alternatives to protect human
health, welfare and the environment, ARARs are considered. In
the absence of ARARs, cleanup levels based upon target risk
levels are considered for population exposure points.'
The groundwater protection standard under RCRA Subpart F
(40 CFR 264.92) requires concentration limits which may not be
exceeded in order to be protective. These levels can be set using
three possible criteria (40 CFR 264.94a): (1) background levels of
the hazardous constituents, (2) maximum contaminant levels or
(3) ACLs. The maximum contaminant levels are those set by the
National Interim Primary Drinking Water Standards and would
include the MCLs mentioned above. In addition, according to the
ACL guidance,2 any state or local laws must be considered.
An ACL is based on the establishment of a site-specific ground-
water protection standard. The standard sets a limit on the
amount of groundwater contamination that can be allowed
without endangering public health or the environment. ACLs are
based on a formalized demonstration which focuses around a
quantitative human health risk assessment. The ACL demonstra-
tion can show either that there is no exposure to contaminants or
that exposure at the exposure point does not pose a substantial
current or potential risk to human health and the environment.
Essentially, an ACL under RCRA is the same as the cleanup
criteria used under CERCLA. The philosophy behind the two
processes is the same: the reduction of contaminants to a level
that will not endanger human health or the environment. The ma-
jor difference between the two is that an ACL demonstration, as
outlined in the guidance manual2 and discussed later, requires ex-
tensive documentation. The institutional nature of CERCLA has
fewer data requirements than RCRA. Thus, Superfund cleanup
levels are simpler to derive and more cost-effective than RCRA
ACLs.

CERCLA GUIDANCE
As part of a CERCLA feasibility study, the potential impacts
of proposed remedial alternatives (source control and/or manage-
ment of migration) on public health, welfare and the environment
are evaluated. The standards or criteria (target concentrations)
used to determine whether a proposed alternative is acceptable
ALTERNATIVE CONTAMINATION LIMITS
-------
can be used as design goals for remedial alternatives and to de-
termine contaminant levels (action levels) in environmental media
at the site which would allow target concentrations to be reached
at the exposure points.
The target concentrations are used based on applicable or rele-
vant and appropriate requirements (ARARs) when ARARs are
available. When ARARs are not available for the chemical being
assessed at the site, the remedial alternatives should reduce am-
bient chemical concentrations to levels associated with a car-
cinogenic risk range oflO-4tolO-7.
The implementing guidance for developing acceptable concen-
trations under CERCLA is the Superfund Public Health Evalua-
tion Manual.' This guidance document recommends focusing the
evaluation of the remedial alternatives on potentially carcinogenic
contaminants rather than noncarcinogenic compounds since
potential carcinogen levels usually drive the final design. Once
target concentrations for potential carcinogens have been de-
termined, long-term exposure to the concentrations of noncar-
cinogens that would remain after remediation is then evaluated to
insure that acceptable levels are attained.
The CERCLA process involves the determination of health
risks and the development of performance goals for the remedial
alternatives. Fig. 1 summarizes the process.

£.»/
,„

T^i*'
C*-Kt*l'tll»*
•* AftAftl

Aiitiuilm u
tiMn Tktl
TMIH
U-rT

iJ^"r"..
Htihk tut *l

Figure 1
Flow Chart of Performance Goals Process

RCRA GUIDANCE
RCRA regulations (40 CFR Parts 264 and 270) require permit-
ted hazardous waste facilities to be designed and operated in a
manner that will prevent groundwater contamination. The con-
centration limits for hazardous compounds in groundwater at the
compliance point generally are set at background or RCRA max-
imum contaminant limits (similar to ARARs under CERCLA).
Variances from these standards can be obtained by demonstrating
that the constituents will not pose a substantial present or poten-
tial hazard to human health or the environment. These petitions
for variances must be presented in an alternate concentration
limit (ACL) document. The ACL guidance document requires
that at least 35 types of information be included in the ACL
demonstration.
The fundamental basis for development of ACLs is 40 CFR
264.94(b), which sets forth a number of criteria for evaluating
ACLs based on the contaminant's effects on both groundwater
and hydraulically connected surface water. As outlined in the
draft ACL guidance document:'

1. Potential adverse effects on groundwater quality, considering:
• The physical and chemical characteristics of the waste in the
regulated unit, including its potential for migration
• The hydrogeological characteristics of the facility and sur-
rounding land
• The quantity of groundwater and the direction of ground-
water flow
• The proximity and withdrawal rates of groundwater users
• The current and future uses of groundwater in the area
• The existing quality of groundwater, including other sources
contamination and their cumulative impact on the ground-
water quality
• The potential for health risks caused by human exposure to
waste constituents
• The potential for damage to wildlife, crops, vegetation and
physical structures caused by exposure to waste constituents
• The persistence and permanence of the potential adverse
effects
2. Potential adverse effects on hydraulically connected surface
water quality, considering:
• The volume and physical and chemical characteristics of
waste in the regulated unit
• The hydrogeological characteristics of the facility and sur-
rounding land
• The quantity and quality of groundwater and the direction
of groundwater flow
• The patterns of rainfall in the region
• The proximity of the regulated unit to surface waters
• The current and future uses of surface waters in the area and
any water quality standards established for those surface
waters
• The existing quality of surface water, including other
sources of contamination and the cumulative impact on
surface water quality
• The potential for health risks caused by human exposure to
waste constituents
• The potential for damage to wildlife, crops, vegetation
and physical structures caused by exposure to waste con-
stituents
• The potential for damage to wildlife, crops, vegetation and
physical structures caused by exposure to waste constituents
• The persistence and permanence of the potential adverse
effects

Once the above criteria have been considered, a risk assessment
to determine the allowable concentrations at a point of exposure
is performed. Fig. 2 is a flow chart outlining a methodology that
has been used. First, compounds were selected from the list of
chemicals detected at the site. Second, information on their toxic
effects in humans and their environmental impact was reviewed in
detail. Third, an estimate was made of the concentration of each
compound at the facility boundary line or other appropriate ex-
posure point that would provide adequate protection of human
health and the environment. Fourth, the fate and transport of
these contaminants were mathematically modeled. Last, ACLs
were calculated by taking the concentrations at the exposure
points and projecting them backward to estimate the concentra-
tions at site monitoring wells, taking natural renovation by
physicochemical processes into account. In principle, the
methodology is virtually identical to that used at a Superfund site.
The main difference is in the larger quantity and higher quality of
risk assessment data required under RCRA compared to
CERCLA. Often the data required under CERCLA are readily
available from RI/FS or FIT documents, whereas the data re-
quired under RCRA are not available from common RCRA
documentation such as a Part B Permit Application and need to
be developed as part of the demonstration.

CASE STUDY: TYSON'S DUMP
Tyson's Dump, a CERCLA site, is an abandoned septic and
chemical waste disposal site that was operated from 1962 to 1970.
During the period of active operations, several lagoons were con-
structed within old sandstone quarry pits, primarily in the center
and western edge of the dump site. Liquid wastes and sludges
8 ALTERNATIVE CONTAMINATION LIMITS
-------
Toxicologic and
Environmental
Impact Evaluation
Fate And
Transport
Evauation
ADI Acceptable Daily Intake
UCR Unit Cancer Risk
AWQC Ambient Water Quality Criteria
ACL Alternative Concentration Limit
Figure 2
Determination of Alternate Concentration Limits
were hauled to Tyson's Dump in bulk tank trucks and deposited
in these unlined lagoons. In 1973 the dump was closed by order of
the Pennsylvania Department of Environmental Resources. Dur-
ing closure, the lagoons were reportedly drained, backfilled and
hydroseeded.
The U.S. EPA responded to a citizen complaint about Tyson's
Dump in 1983. Several corrective actions were taken at the site.
Subsequently, a remedial investigation and feasibility study for
Tyson's Dump was conducted. The result of this study was the
recommendation, in the record of decision, for the excavation of
contaminated soil from the former lagoon areas for off-site
disposal.
ICF-Clement prepared a risk assessment to assess the environ-
mental health risks associated with exposure to chemical con-
tamination at the Tyson's Dump site.3 The risks associated with
potential exposure resulting from the current situation and from
future conditions if a no-action alternative were adopted were
estimated. Since the U.S. EPA recommended excavation and off-
site disposal of contaminated soils in its record of decision,
health-based action levels for soil removal were developed. The
action levels were based on information gathered during the risk
assessment of current conditions. The main steps in developing
action levels are shown in Fig. 3. This flow chart should be com-
pared to Fig. 2 which is representative at an RCRA site.
During the risk assessment, the potentially carcinogenic
chemicals (benzene, tetrachloroethylene, trichloroethylene and
1,2,3-trichloropropane) were found to be responsible for the pre-
ponderance of the risk. Thus, action levels for only these
chemicals were derived.

Action Levels in Soils for On-Site Exposures
Three on-site exposures were considered for action level devel-
opment: (1) use of the shallow aquifer in the lagoon area for
Select
Compounds
Determine
Migration Pathways
Determine Current or
Future Points of Exposure
(On-and Off-Site)
Determine Target
Concentration in Media at
Exposure Points
Backmodel from Exposure
Point Concentrations to
Soil Concentrations
(Action Levels)
Figure 3
Determination of Action Levels in Tyson's Dump Soils

drinking, (2) dermal contact with soils by trespassers or future
workers at the site and (3) ingestion of contaminated soils by
trespassing children.

Ingestion of Groundwater from Lagoon Area Well
Target concentrations in drinking water were developed as part
of the risk assessment. If it is assumed that contaminated ground-
water is in equilibrium with contaminated soils at the site, corre-
sponding concentrations of action levels in soils at the site (Cs)
can be calculated using the following equation:
Concentration of Contaminant in Soil
Gig/kg)
= Kn
(1)
Concentration of Contaminant in Soil
Water (Cw) Oig/1)
where Kp, the soil-water partition coefficient, is the product of
Koc and Foc, Koc is the organic carbon partition coefficient and
Foc is the fraction of organic carbon in site soils (assumed to be
0.01 for this site). The values for Koc were obtained from the U.S.
EPA.' The resulting action levels for a Cw corresponding to a
10 ~6 excess cancer risk are given in Table 1.

Table 1
Action Levels* for Protection of Human Health

Ben
Tet
eth
Tri
eth
Tri
Drinking
Water

ene 1
achloto- /.t
lene
hloro- i.u
lene
hloro- 3.8
Soil

117
67
310
54

560
320
1500
260
Drinking Drinking
Drinking River Water River Water

32 Z.7>10* 5.2x10*
77 6.6x10* 9.7x10*
130 1. IxlO5 1.8x105
120 1x105 i.3xlo5
* in mg/kg soil corresponding to 10 ° risk

ALTERNATIVE CONTAMINATION LIMITS 9
-------
Ingestion of On-Site Soil
For this exposure scenario, it was assumed that ingestion is only
important for children between the ages of 2 and 6, that an
average body weight for a child of this age is 17 kg and that the in-
gestion rate varies from 0.1 to 5 g/day, with the higher values
representative of pica behavior (assumptions recommended by
the U.S. EPA.1 The target dose value assumes that no other ex-
posures to these chemicals occur and is based on a 10 ~6 excess
cancer risk. The action levels in soil (Cs that correspond to a pica
exposure scenario are estimated using the following equation:4
Table 2
GMATC Values for Aquatic Life
Oig/liter)
Target Dose (mg/kg/day)
= Cancer Risk x Body Weight x Lifetime

Potency x Cs
(2)
The results of this analysis for a 10-* risk level are presented in
Table 1.

Action Levels in Soils for Ingestion of
Groundwater from Boundary Well
Action levels in soils for ingestion of boundary well water are
calculated using basically the same procedures as for the ingestion
of the groundwater for on-site wells. However, in the case of the
boundary well, it is necessary to model the migration of con-
taminants in groundwater from target concentrations at the site
boundary to the areas of soil contamination. This has been done
assuming a distance of 75 ft from the boundary well to the area of
soil contamination, using the target concentrations based on the
ingestion of 2 I/day of water by a 70 kg person, Kp values derived
previously and the vertical-horizontal spread (VHS) groundwater
model.' The resulting action levels in soil are given in Table 1.

Protection of Plant and Aquatic Life in
Floodplain Wetlands
The target concentrations for groundwater discharging to the
floodplain wetland area which would be protective of aquatic life
are given in Table 2. The method used [calculation of the geo-
metric mean of the maximum allowable toxic concentration
(GMATC) by the method of Suter, etal.'], using fish LC50 values
as input GMATC values, will be protective of chronic effects
against fish. The philosophy of the U.S. EPA which underlies the
ambient water quality criteria documents is that chronic fish
criteria will also protect invertebrates and plant life. GMATC
values also may be compared to aquatic toxicity parameters for
confirmation of this approach.
Again, as in the previous example, it is necessary to model from
the target concentrations (GMATCs) in floodplain wetland water
to determine corresponding groundwater contaminant levels and
thus action levels in soils at the site. This can be performed using
the model described by Mills, et a/.' The appropriate equation is:
C0 =
Kp (GMATC)
0.064
(3)
The resulting action levels for the GMATC values are given in
Table 1.
Action Levels in On-Site Soils for
Schuylkill River Exposure
Three types of exposure to site contaminants in two areas of the
Schuylkill River are considered—exposure by recreational users
of the river at the point of site groundwater infiltration into the
Chemical
Acute LCso Fish
GMATC
Benzene
Tetrachloroethylene
Trichloroethylene
Trichloropropane
5,300
5,300
44,700
42,000
120
120
6SO
620
river, exposure by human populations consuming water from the
drinking water intake in the river 2,000 ft downstream and ex-
posure by aquatic life to site contaminants in these two areas.
The recreational use scenario considers an individual swimming
in Schuylkill River water which has been contaminated by
discharge from Tyson's Dump. A swimmer is exposed by dermal
absorption and ingestion of a small amount of water. Based on
the literature, an individual swims 9 times per year for 2 hr each
time.' It is assumed that swimming occurs for 35 yr or one-half a
lifetime. Based on this scenario, it is unlikely that any significant
exposure would occur. A conservative estimate of action levels,
however, may be derived by assuming that drinking water is con-
sumed at this point. These results for the 10-6 excess cancer risk
level are presented in Table 1.
The Keystone Water Company intake is about 2,000 ft
downstream from the site. Action levels for site soil based on this
point are calculated considering that volatilization has occurred
from discharge to intake. Volatilization was modelled using
methods from Mills, et at.1 Results corresponding to a 10 ~6 ex-
cess cancer risk level are shown in Table 1.
The action levels for protection of aquatic life in the Schuylkill
River are based on the GMATC concept described earlier. Con-
centrations were adjusted based on further dilution and mixing on
transport of groundwater to the river. The results corresponding
to a GMATC are shown in Table 1.

Summary of Action Levels
Table 1 summarizes the action levels in soils developed for the
Tyson's Dump site corresponding to the 10-6 human health ex-
cess cancer risk level. As seen in this table, the drinking water ex-
posure scenario drives the cleanup levels. These levels are based
on the persistence and mobility of these compounds as well as
their potential carcinogenic potency factors.

CONCLUSIONS
The methodology used to develop action levels for the Tyson's
Dump site corresponds in principle to that recommended in the
RCRA ACL Guidance, although it was accomplished with a
minimal level of documentation. The development of the cleanup
criteria depends on the use of site-specific exposure information
in conjunction with lexicological information on the compounds
found at the site and commonly used models. All the data re-
quired for this action level determination were readily available
from the RI/FS documents or the Superfund Public Health
Evaluation Manual.' The approach under CERCLA is straight-
forward and could be transferred easily to RCRA.
REFERENCES

1. U.S. EPA, Superfund Public Health Evaluation Manual. Office of
Emergency Response, Washington, DC, U.S. EPA 540/1-86/060,
Oct. 1986.
10 ALTERNATIVE CONTAMINATION LIMITS
-------
2. U.S. EPA, "Alternate Concentration Limit Guidance Based On
Section 264.94(b) Criteria. Part 1. Information Required in ACL
Demonstrations," Office of Solid Waste, Washington, DC, July
1986.
3. ICF-Clement Associates, Inc., "Risk Assessment of the Tyson's
Dump Site, Montgomery County, Pennsylvania," Prepared for U.S.
Army Corps of Engineers and U.S. EPA, Feb. 1986.
4. Schaum, J., "Feasibility Study of Alternatives American Creosote
Works Site," Draft, July 1985.
5. U.S. EPA, "VHS Model," Fed. Reg. 50, 1985, 7896-7900.
6. Suter, G.W., Vaughan, D.S. and Gardner, R.H., "Risk (assessment
by analysis of extrapolation error: A demonstration for effects of
pollutants on fish," Environ. Toxicol. Chem. 2, 1983, 369-378.

7. Mills, W.B., Dean, J.D., Porcella, D.B., Gherini, S.A., Hudson,
R.J.M., Frick, W.E., Rupp, G.L. and Bowie, G.L., "Water Quality
Assessment: A Screening Procedure for Toxic and Conventional Pol-
lutants," Prepared for the U.S. EPA, U.S. EPA 600/6-82-004, 1982.

8. Halper, M.P., "Identification and Evaluation of Waterborne Routes
Exposure from Other than Food and Drinking Water," Prepared for
the U.S. EPA, U.S. EPA 440/4-79-016, 1979.
ALTERNATIVE CONTAMINATION LIMITS 11
-------
Alternate Concentration Limits
Under the RCRA Program

Vernon B. Myers, Ph.D.
Mark A. Salee
Jerry Carman
U.S. Environmental Protection Agency
Office of Solid Waste
Washington, D.C.
INTRODUCTION
Hazardous waste facilities permitted under RCRA regulations
(40 CFR Parts 264 and 270) must be designed and operated in a
manner that will prevent groundwater contamination. If leakage
of hazardous constituents is detected in the groundwater at a
RCRA facility, the regulations require the establishment of a
groundwater protection standard at that facility. This standard
establishes a limit on the amount of groundwater contamination
that can be allowed.
The groundwater protection standard is an essential element in
the Agency's strategy to ensure that public health and the en-
vironment are not endangered by any contamination of ground-
water resulting from the treatment, storage or disposal of haz-
ardous wastes. As such, the standard will indicate when ground-
water cleanup is necessary to control contamination that has
emerged from a hazardous waste facility.

GROUNDWATER PROTECTION STANDARD
The principal elements of the groundwater protection standard
are discussed in section 264.92 of the regulation. For each haz-
ardous constituent entering the groundwater from a regulated
unit, a concentration limit must be established that will serve as a
limit beyond which degradation of groundwater quality will not
be allowed. There are three possible concentration limits that
can be used to establish the groundwater protection standard:

• Background levels of the constituents
• Maximum concentrations listed in Table 1 of Section 264.94(a)
of the regulation
• Alternate concentration limits (Fig. 1)
The approach used by the regulation is to adopt widely ac-
cepted environmental performance standards, when available, as
concentration limits. However, because of the lack of currently
available standards, specific concentration limits for only a few
specific constituents have been included in the regulations. These
limits are those standards that were established by the National
Interim Primary Drinking Water Regulations. If a constituent is
not on e of these compounds, the standard becomes no degrada-
tion beyond background water quality. In such cases the con-
centration limit is set at a background level. However, variances
are available where the permit applicant can demonstrate that a
constituent will not pose a substantial present or potential haz-
ard to human health or the environment at an "alternate con-
centration limit" (ACL). Section 264.94 of the regulation con-
tains nineteen criteria that must be considered when establishing
ACLs. Alternate concentration limits are established in a facil-
ity's permit only after the U.S. EPA decides that the ACLs will
not adversely affect public health or the environment.
DEFAULT VALUE
MCL FOR 14 CHEMICALS
BACKGROUND
CORRECTIVE ACTION
AND ASSOCIATED
MONITORING
Figure 1
RCRA Groundwater Protection Process
ALTERNATE CONCENTRATION LIMITS
To establish ACLs, two points must be defined on an RCRA
facility's property (Fig. 2): the Point of Compliance (POC) and
the Point of Exposure (POE). The POC is defined in the Subpart
F Regulations (40 CFR Part 264.95). The POC is the point in the
uppermost aquifer, on the immediate downgradient side of the
regulated unit (waste management unit boundary), where ground-
water monitoring takes place and the groundwater protection
standard is set. The ACL, if it is established in the permit, would
be set at this point.
The point of exposure (POE) is the point at which it is assumed
a potential receptor can come in contact (either now or in the
future) with the groundwater (contamination). Therefore, the
groundwater quality at the POE must be protective of that recep-
tor. For example, a facility may have a groundwater contaminant
plume restricted to a part of its property.
\2 ALTERNATIVE CONTAMINATION LIMITS
-------
POINT OF EXPOSUSE-
F«lllly Boundlry —/
Figure 2
Definitions
Point of Exposure (POE)—Point at
which potential exposure to con-
taminants is assumed. Location is
site-specific. Health or environ-
mental based level is met here.
Facility Boundary—The prop-
erty boundary of the facility.
Regulated Unit—The area where
the hazardous wastes are kept
(landfill, surface impoundment).

Point of Compliance (POQ—The
point of the downgradient side of
the unit where the groundwater
protection standard is set. The ACL
is set here.
• Contaminant Plume—The volume
of groundwater that contains the
leaking pollutants.

In this case, it may be appropriate to assume that people will be
exposed, through a drinking water well, to the groundwater im-
mediately at the edge of the plume. The groundwater at that
point, the POE, must then be safe for human consumption.
Likewise, if the groundwater contamination is discharging to an
on-site surface water body, the potential receptor may, in some
cases, be aquatic organisms. In this example, the aquatic organ-
isms must be protected from adverse effects of the discharging
contaminants.

POC/POE
Understanding and identifying the spatial relationship between
the POC and the POE is critical in the establishment of an ACL.
Mechanisms that attenuate contaminants should be considered
only over the area between the POC and the downgradient POE.
If the POE is established at the POC, then no form of attenua-
tion should be considered in setting the ACL. In such a case
(POC = POE), the ACL, if applied for, would be equal to the
allowable health or environmental exposure level, with the as-
sumption that exposure would occur at the waste management
unit boundary.
However, if the POE is removed a specified distance from the
POC, then appropriate and conservative estimates of contami-
nant attenuation may be used to calculate the ACL. These mech-
anisms of attenuation would only be considered over that distance
between the POC and the POE. For example, if the POE is 50 m
downgradient of the POC, then dilution in the groundwater could
be conservatively estimated from the volumetric transport of
groundwater in relation to the mass loading of the leaking constit-
uents over that 50 m. A dilution factor could then be applied to
the allowable health or environmental exposure level at the POE
to derive the ACL.
POLICY GUIDELINES
Experience gained over the last several years has allowed the
Agency to develop a better understanding of groundwater con-
tamination problems. The Agency is developing general policy
guidelines for the use of ACLs at RCRA hazardous waste facil-
ities. These guidelines are designed to establish an ACL proced-
ure that will be protective of human health and the environment.
To provide national consistency in calculations used to estimate
the potential impacts of releases of hazardous constituents to
groundwater from regulated units, the Agency will establish lim-
its on where the pointsof exposure should be assumed at an RCRA
facility. The agency believes that the POEs must be set conserva-
tively enough to be protective of human health and the environ-
ment in situations that would be encountered during the setting
of ACLs. The POEs should also consider the persistent nature of
many toxic chemicals in the environment and the need to prevent
exposure to many of these compounds.
The type and amount of information needed for an ACL
demonstration depends on the placement of the POE and the
site-specific characteristics. For units where the owner or opera-
tor desires an ACL to be set directly at the allowable health or en-
vironmental exposure level, relatively little additional information
beyond that already supplied in the permit application normally
will be required. If the owner or operator wishes to take into ac-
count mechanisms of attenuation in deriving the ACL, more in-
formation will be required.
The simplest and quickest method for deriving the ACL, in all
situations, will be to establish the POE at the POC. Even for
those sites with gross contamination, setting the POE at the POC
may be the least expensive option because high levels of contam-
ination usually will require major source control and corrective
action measures, regardless of the ACL. By setting the POE at the
POC, the owner or operator may save significant amounts of
time and money by not having to gather and organize the addi-
tional information required for deriving an ACL that employs
mechanisms of attenuation.
No matter where the POE is established, an allowable health or
environmental exposure level for each hazardous constituent
must be determined for that POE. The appropriate allowable
health or environmental exposure level will be dependent on the
most vulnerable receptor near the facility. The most vulnerable
receptor is that receptor, human or environmental, that has the
lowest tolerance to the hazardous constituent(s) for which the
ACL is being requested. In most cases, this will be humans ex-
posed to the contaminated groundwater via ingestion (i.e., drink-
ing). However, at times, the most vulnerable receptor will be an
environmental one.
CONCLUSION
The U.S. EPA indicates its final decision on the merits of an
ACL demonstration when it issues an RCRA permit for a facility.
Approved ACLs will be included as the groundwater protection
standards in the permit. The groundwater protection standards
are the concentrations which hazardous constituent concentra-
tions are not to exceed. However, if the concentrations of the
hazardous constituents do exceed their standards, groundwater
corrective action will be required. In this case, the protection
standards become the cleanup goal.
ALTERNATIVE CONTAMINATION LIMITS
13
-------
An Overview of Liability Insurance Under RCRA

Jackie Tenusak
Office of Waste Programs Enforcement
U.S. Environmental Protection Agency
Washington, D.C.
Paul E. Bailey, J.D.
ICF Incorporated
Washington, D.C.
INTRODUCTION
The U.S. EPA requires owners or operators of facilities that
treat, store or dispose of hazardous waste to maintain financial
assistance for third-party liability claims. Currently, however,
some owners and operators are finding liability insurance cov-
erage difficult to obtain. This paper discusses how the RCRA
liability coverage requirements have evolved in response to chang-
ing conditions in the liability insurance market.
First, the paper briefly describes the Congressional mandate
for financial responsibility in RCRA and the regulations promul-
gated and implemented by the U.S. EPA. It then describes how
the U.S. EPA developed its liability coverage requirements and
the influence of the insurance market on these rules. Next, the
paper traces the decline of the liability insurance market in the
past few years and describes how the U.S. EPA has responded
with rulemakings and enforcement policies. Finally, the paper
identifies the liability issues facing the U.S. EPA in the near
future.

LIABILITY COVERAGE UNDER RCRA
The Solid Waste Disposal Act of 1965, RCRA's predecessor,
did not contain financial responsibility provisions. By the time
RCRA was enacted in 1976, however, Congress had become in-
creasingly aware of the potential dangers posed by hazardous
waste. Section 3004(a)(6), therefore, was added to the Act of
authorize the U.S. EPA to establish financial responsibility "as
may be necessary or desirable" for owners and operators of haz-
ardous waste treatment, storage, and disposal faculties (TSDFs).
Between 1978 and 1982, the U.S. EPA implemented its Con-
gressional mandate through a series of rulemakings that required
owners and operators of hazardous waste facilities to demonstrate
financial responsibility for third-party liability as well as closure
and post-closure care. Congress emphasized the importance of
liability requirements to hazardous waste management when it
added Section 3005(e) to RCRA as part of the Hazardous and
Solid Waste Amendments of 1984 (HSWA). This provision re-
quired the U.S. EPA to terminate interim status for all land dis-
posal facilities by Nov. 8, 1985, unless the owner or operator had
complied with certain regulatory requirements, including liability
coverage.
The RCRA liability coverage regulations (40 CFR 264.147 and
265.147) currently require owners or operators to demonstrate
financial assurance for liability coverage for third-party damages
for as long as the facility remains in operation. Specifically:
(1) Owners or operators of all types of TSDFs must
demonstrate liability coverage for sudden accidental
occurrences in the amount of $1 million per occur-
rence and $2 million annual aggregate, exclusive of
legal defense costs; and
(2) Owners or operators of surface impoundments, land-
fills, and land treatment facilities must demonstrate
liability coverage for non-sudden accidental occur-
rences in the amount of $3 million per occurrence and
$6 million annual aggregate, exclusive of legal defense
costs.
Coverage is required on a per-firm basis; that is, the num-
ber of facilities owned by a firm does not affect the amount of
coverage required.
The regulations require coverage for sudden and non-sudden
accidental occurrences. An "accidental occurrence" is an insur-
ance term of art that describes an accident that is neither ex-
pected nor intended from the standpoint of the insured. A sud-
den accidental occurrence is an accidental occurrence that is not
continuous or repeated in nature, e.g., an explosion or a fire. In
contrast, a non-sudden accidental occurrence is an accidental
occurrence that takes place over a period of time and involve!
continuous or repeated exposure to certain conditions. Non-sud-
den accidental pollution, such as hazardous wastes generally
leaching into groundwater, is also referred to as gradual pollu-
tion.
Two different types of insurance policies have provided the
type of coverage required in the regulations. The Comprehen-
sive General Liability (CGL) policy has been widely available for
decades as the standard liability coverage form for damages due
to sudden accidental occurrences. Since the 1970s, most standard
CGL policies have excluded damages caused by non-sudden ac-
cidental pollution coverage; this exclusion is referred to as the
"pollution exclusion." Coverage for both sudden and/or non-
sudden occurrences may be obtained through an Environmental
Impairment Liability (EIL) policy, designed specifically to cover
third-party damages caused by pollution.
The regulations also allow owners or operators to demonstrate
financial assurance for liability coverage by passing a financial
test or by obtaining a corporate parent guarantee.
14 LIABILITY/INSURANCE/DEREGULATION
-------
HISTORICAL BACKGROUND
In 1976, Congress authorized the U.S. EPA under RCRA
Section 3004 to develop standards for financial responsibility but
provided no specific mandate for the types of standards that
should be set. In fact, the only mention of financial responsibility
in the 1976 statute was included in RCRA 3004(a) as follows:
...Such standards shall include, but need not be limited to,
requirements respecting—
(6) the maintenance of operation of such facilities and re-
quiring such additional qualifications as to ownership,
continuity of operation, training for personnel, and finan-
cial responsibility as may be necessary or desirable.
Moreover, when RCRA was enacted, no well-established mar-
ket was available for insurance of hazardous waste facilities and
neither EIL policies nor coverage for non-sudden occurrences
were yet available. As a result, the U.S. EPA relied heavily on
the advice of experts in the insurance industry and members of
the regulated community in developing insurance requirements to
ensure that their regulations conformed to industry practices.
In December 1978 (43 FR 58987), the U.S. EPA proposed the
first liability requirements. In response to the insurance industry's
optimism about a developing market for non-sudden coverage,
the proposed rules required sudden coverage for all facilities and
non-sudden coverage for disposal facilities. However, while the
insurance industry was optimistic about the availability of non-
sudden coverage, because few EIL policies had been written and
claims and actuarial data were not yet available, it was difficult
for the insurance industry to state with certainty the extent to
which coverage would be available or the costs.
In analyzing the comments to the December 1978 proposal,
the U.S. EPA was made increasingly aware that although firms
with good management practices could obtain sudden accidental
coverage, the availability of non-sudden coverage was still highly
uncertain. In part because of limited actuarial data on hazardous
waste facilities, insurers expressed a special concern about offer-
ing non-sudden coverage to firms owning facilities operating
under interim status that were not yet subject to specific permit
conditions and particularly to smaller firms. Consequently, the
U.S. EPA decided to repropose the liability requirements on
May 19, 1980 (45 Federal Register 33263). Among other changes,
the U.S. EPA proposed to require sudden accidental coverage
for all facilities, but to drop the non-sudden liability coverage
requirements for land disposal facilities.
The U.S. EPA promulgated its first final rules on liability cov-
erage on Jan. 12, 1981 (46 Federal Register 2802). This final
rule struck a balance between the U.S. EPA's commitment to
rely to the greatest extent possible on private insurance to provide
liability coverage and its continued concern that some, firms might
be unable to obtain non-sudden coverage. In response to the con-
cerns expressed by members of the insurance industry that non-
sudden coverage would not be available immediately, especially
to smaller firms, the final rule provided a three year phase-in
period for non-sudden coverage based on a firm's sales or reve-
nues. Owners or operators whose sales or revenues totalled at
least $10 million in their last fiscal year were required to obtain
non-sudden coverage by Jan. 15, 1983. Firms with sales or reve-
nues between $5 million and $10 million were required to obtain
non-sudden coverage by Jan. 15, 1984. The deadline for firms
with sales or revenues of less than $5 million was Jan. 15,1985.
Initially the phase-in seemed to be effective. Owners and oper-
ators used EIL policies to satisfy the nonsudden coverage re-
quirement. In July 1981, a U.S. EPA staff member was informed
by a prominent member of the environmental risk assessment
business that approximately six to eight policies providing non-
sudden coverage were being written each week, that six to eight
more applications were arriving daily and that business appeared
to be increasing as the compliance effective dates approached.
The insurance industry was also optimistic about insurance ca-
pacity for the liability coverage required by the January 1981
rule. In an attempt to increase the capacity for pollution liability
coverage, the Pollution Liability Insurance Association (PLIA)
was established in October 1981 with 37 member companies to
write entire packages of EIL policies. PLIA was established to
provide reinsurance to its member insurance companies offering
pollution liability coverage, i.e., the member companies would
transfer all or a part of the pollution liability risks they assumed
in their sudden and non-sudden policies to PLIA. Because PLIA
assumed part of their liability, member companies were able to
underwrite higher policy limits.
During 1981, under a new administration, the U.S. EPA con-
ducted a broad review of its regulatory agenda. This review in-
cluded a re-evaluation of the RCRA liability requirement. On.
Oct. 1, 1981, the U.S. EPA deferred the effective date of the lia-
bility coverage requirement until Apr. 13, 1982 and announced its
intent to withdraw the liability coverage requirements altogether
(46 Federal Register 48197). In the Oct. 1 notice, the U.S. EPA
"questioned whether those requirements were necessary or desir-
able to meet the requirements of RCRA" (50 Federal Register
33902).
In response to the announcement of a plan to drop the liability
coverage requirement, the U.S. EPA received wide-scale support
for retaining the requirement. In contrast, virtually no support
was expressed for the proposal to withdraw the requirement. The
insurance industry's response to the U.S. EPA's doubts about
the necessity or desirability of the liability coverage requirements
was consistent with those from regulated industries. Congress and
the public. Insurers argued that private insurance and liability
coverage requirements were necessary complements. Private in-
surance supported the goals of the liability rules by offering low-
er premiums to those firms that met RCRA standards. In turn,
the federal requirements were likely to encourage the develop-
ment of the insurance market. Federal requirements were also
likely to promote consistency in state regulations, thus minimizing
the need for insurers to tailor their policies to many different
sets of regulations.
Responding to the public outcry to retain the liability coverage
requirements, the U.S. EPA concluded that liability coverage
regulations were at a minimum "desirable" and that, based on
the insurance market's response, insurance would be available.
By early 1982, the insurance market for non-sudden coverage
had begun to respond to increasing demand and there were indi-
cations that the market would expand considerably in the near
future. Reflecting this renewed optimism about the availability
of both sudden and non-sudden coverage on Apr. 6, 1982, the
U.S. EPA revised the Jan. 12, 1981 liability coverage require-
ments. In general, this rule reiterated the earlier requirements,
including the phase-in period for non-sudden coverage. The most
significant change in the Apr. 16, 1982 rule was the addition of a
financial test for demonstrating financial responsibility for liabil-
ity coverage. Those owners or operators with the requisite finan-
cial strength would thus have a choice between self-insuring
and purchasing insurance. By allowing the use of a financial test,
the U.S. EPA intended to ease the burden on the insurance mar-
ket of supplying liability coverage to TSDFs.

THE U.S. EPA'S RESPONSE TO A CHANGING
INSURANCE MARKET
In 1982 when the U.S. EPA amended its liability requirements,
the insurance market seemed promising for both sudden and
LIABILITY/INSURANCE/DEREGULATION 15
-------
non-sudden coverage for hazardous waste management facilities.
The insurance industry responded to the liability coverage re-
quirements by providing sudden coverage in CGL policies and
non-sudden and sudden coverage in EIL policies. Not long after
the first phase-in deadline had passed, however, the U.S. EPA
became increasingly aware that the insurance market, especially
for non-sudden coverage, was not developing as quickly as the
insurance market had hoped. In early 1983 the U.S. EPA spon-
sored a study to identify and evaluate possible options for deal-
ing with potential problems in insurance availability or afford-
ability.1 One of the conclusions of that report was that because
the insurance market was just developing, it was difficult to pre-
dict with certainty the extent of availability problems and that the
U.S. EPA should continue to monitor carefully the insurance
market. The report recommended that market assistance plan
programs be considered as a possible means of handling short-
term availability problems.
These concerns about growing constraints in the pollution lia-
bility insurance markets were confirmed further by speakers at a
conference attended by U.S. EPA Headquarters, Regional and
state staff in October 1984 on the RCRA financial responsibility
requirements. One speaker talked at length about insurance
issues, ranging from the limited availability of non-sudden EIL
coverage to the increased reluctance of insurers to issue CGL pol-
icies providing coverage for sudden accidental coverage. The
U.S. EPA learned that few, if any, non-sudden policies were be-
ing issued or renewed. Moreover, many insurers limited the avail-
ability of sudden and/or non-sudden insurance to firms who
carried all their coverage with them (e.g., workers' compensa-
tion and fire insurance). The U.S. EPA also realized that the
smaller firms subject to the last effective date of the phase-in,
Jan. 15, 1985, probably would be unable to comply with the
liability requirements by obtaining insurance or passing the finan-
cial test.
The reluctance of insurers to issue CGL policies providing
sudden coverage stemmed in part from a number of judicial de-
cisions handed down in the period from 1981 to 1984. In several
cases, courts found the pollution exclusion clauses in CGL poli-
cies to be ambiguous and interpreted the ambiguity in favor of
the insured, in accordance with principles of contract law. That
is, the courts found that the exclusions were not effective and that
the CGL policies covered both sudden and non-sudden occur-
rences. Other courts have interpreted the pollution exclusion
more strictly, finding that CGL policies only cover sudden and
accidental occurrences.
Reacting to the courts' broad interpretation of the pollution
exclusion, insurance companies increasingly sought to avoid the
potential liability for non-sudden pollution occurrences by ex-
cluding all pollution coverage, including sudden and accidental
coverage, from CGL policies. Pollution coverage for sudden
occurrences in a CGL policy is now available only under "buy-
back" coverage provisions, for an additional premium.
Many other factors have contributed to the reduced availabil-
ity and increased cost of pollution coverage, including large
underwriting losses in the property and casualty market due to
low premiums and large claims; difficulty in setting premiums
based on risk; lack of demand for the coverage; and the magni-
tude of potential liability claims stemming from recent sensa-
tional pollution events, such as the tragedy in Bhopal, India. It is,
however, important to note that the insurance industry is cyclical.
During periods when economic conditions result in large insur-
1 J.R. Hunter and ICF Incorporated. "Residual Market Mechanisms
and Their Applicability to the Hazardous Waste Insurance Market,"
Apr. 8. 1983.
ance industry losses, the insurance industry responds by curtailing
high risk policies and raising premiums. Conversely, during
periods of high interest rates when insurers earn substantial in-
vestment income, they are willing to write high-risk policies at
competitively priced premiums.
The insurance industry considers sudden and non-sudden cov-
erage of hazardous waste management facilities to be high risk
for several reasons. Lack of actuarial data limits an insurer's abil-
ity to establish realistic premiums to cover the assumed risk. The
absence of acceptable and uniform methods for analyzing the
risks of hazardous waste management has also deterred insur-
ers from offering this type of coverage. Finally, there is a social
perception that hazardous waste has not been, nor can be, ade-
quately managed. Insurers contend that this attitude will lead to
a proliferation of claims that will be costly to defend and, ulti-
mately, to pay.
By early 1985, the U.S. EPA was faced with both a constrained
insurance market and the implications of HSWA on the U.S.
EPA's hazardous waste financial responsibility requirements.
In November 1984, Congress enacted HSWA, which, among
other provisions, terminated interim status at all land disposal fa-
cilities by Nov. 8, 1985 unless the owner or operator applied fora
final permit by that date and certified that the facility was in com-
pliance with groundwater monitoring and closure, post-closure
and liability coverage financial responsibility requirements. In a
survey of over 400 hazardous waste facilities, the U.S. EPA
found that a large number of facilities were likely to be out of
compliance with the liability requirements and subject to the
"loss of interim status" (LOIS) provisions of HSWA.
The U.S. EPA acted on several fronts simultaneously in ad-
dressing this latest insurance "crisis." On Apr. 12, 1985, the
U.S. EPA issues enforcement guidance stating that if a firm could
demonstrate that it made a "good faith" effort to obtain or re-
new insurance, then it should not be found out of compliance
and forced to close. The "good faith effort" to obtain insur-
ance has to be carefully and thoroughly documented. The U.S
EPA found, however, that this "good faith" guidance was ex-
tremely difficult to implement. For example, the U.S. EPA found
it difficult and resource intensive to evaluate whether a particu-
lar firm had in fact contacted those insurance companies that
could reasonably be expected to provide coverage. Furthermore,
the information the U.S. EPA received to document a "good
faith" effort did not always show clearly whether the firm could
not obtain or renew insurance because of constraints in the in-
surance market or because of the high risks associated with the
particular facility's operation. The matter is further complicated
by the volatility of the insurance market, i.e., a risk that is con-
sidered uninsurable in a tight market may be considered an in-
surable risk in a stronger market.
As it became clear that the "good-faith" compliance approach
would be difficult to implement and that coverage was becoming
increasingly difficult to obtain, the U.S. EPA recognized thai 8
large number of facilities could lose their interim status on Nov.
8,1985 solely as a result of the unavailability of insurance.
Realizing that coverage was not available even for some facil-
ities with low risk of release, the U.S. EPA suggested to Con-
gressional staff that the U.S. EPA issue compliance schedules to
owners and operators who satisfied the groundwater monitorial
and permit application requirements but were unable to obtain
liability insurance. Congress, however, stated that full compliance
was expected by Nov. 8, 1985, and the U.S. EPA was required
to terminate interim status on Nov. 8, 1985 for those facilito
out of compliance with the liability coverage requirements (over
1,000 facilities).
Coincident with these efforts to address the specific problem!
16 LIABILITY/INSURANCE/DEREGULATION
-------
posed by the LOIS provisions, the U.S. EPA was developing a
notice of proposed rulemaking and a formal request for more
information on the availability of insurance to satisfy the existing
liability coverage requirements and on options for enhancing its
availability. This notice was published on Aug. 21, 1985 (50
Federal Register 33902). In the request for comments, the U.S.
EPA proposed five possible regulatory responses to the problem
of limited insurance availability:
Maintain the existing requirements
Clarify the required scope of coverage and/or lower the limits
Authorize other financial assurance mechanisms
Authorize waivers
Suspend or withdraw the liability coverage requirements
The notice requesting public comment highlighted the advan-
tages and disadvantages of each approach. For example, the
U.S. EPA considered that while the waiver approach would
avoid shutting down those firms that genuinely attempted to
comply with the liability requirements, it would give firms that
obtained a waiver an unfair economic advantage over firms that
purchased insurance.
Partially in response to information received on the Aug. 8,
1985 notice, the U.S. EPA promulgated an interim final rule on
July 11, 1986 expanding the options available for liability cover-
age. The regulations were amended to allow a firm to use a corpo-
rate guarantee to demonstrate financial assurance for liability
coverage if the guarantor is the parent corporation of the owner
or operator (51 Federal Register 25351). The U.S. EPA limited
guarantors to parent corporations because it believed that parent
guarantees are more clearly enforceable under state law and that
parent corporations are in a better position than other corporate
entities to ensure that the facility is being operated in accordance
with U.S. EPA regulations. In adding only the corporate guar-
antee to the allowable mechanisms of insurance and the finan-
cial test, the U.S. EPA did not preclude taking any of the other
approaches proposed in the August 1985 Federal Register notice.
To the contrary, as the preamble to the July 1986 rule stated,
the U.S. EPA was still considering the approaches raised in the
August 1985 publication.

LIABILITY ISSUES NOW FACING THE
U.S. EPA
Today, the U.S. EPA still faces the basic issue of the limited
availability of liability insurance. Meanwhile, new issues are
developing. Some of the more interesting issues include the use of
exclusions in insurance policies and the role of captive insurance
companies and risk retention groups in RCRA liability coverage.
The U.S. EPA continues to monitor the insurance industry to
keep apace of the availability of both sudden and non-sudden
coverage. In light of the continuing limited availability of insur-
ance, the U.S. EPA is studying additional alternative assurance
mechanisms that may provide adequate coverage. The U.S. EPA
is considering other mechanisms besides insurance, the financial
test and the corporate guarantee that might provide adequate
assurance for liability coverage.
The impact of insurance policy exclusions on the adequacy of
the coverage provided is currently being evaluated by the U.S.
EPA. The U.S. EPA is concerned that the only benefit provided
by an insurance policy with many exclusions is the perception
that third-party liability claims will be covered if an accident
occurs. The perception of coverage alone will not satisfy third-
party claims or the intent of the liability regulations.
Insurance policies typically contain statements of the risks both
covered and excluded. The RCRA regulatory requirements for
liability coverage acknowledge this practice, by providing that:
In the liability insurance requirements, the terms "bodily
injury" and "property damage" shall have the meanings
given these terms by applicable state law. However, these
terms do not include those liabilities which, consistent with
standard industry practices, are excluded from coverage in
liability policies for bodily injury and property damage.

Several different types of standard exclusions are used by the
insurance industry. Commonly, pollution liability policies exclude
coverage for workers' compensation, unemployment benefits and
other related coverage required by federal or state statutes, war
and nuclear material. "Buy-back" provisions provide coverage,
at an additional premium, not included in the policy. For ex-
ample, sudden and accidental pollution coverage may be made
available in a CGL policy through a "buy-back" provision.
The U.S. EPA is evaluating with particular care an exclusion
for pre-existing conditions that began or occurred prior to the
effective data of the policy. The exclusion of pre-existing con-
ditions is sometimes included in both CGL and EIL policies to
limit coverage to only those environmental occurrences discov-
ered after the effective date of the policy. Insurers that conduct
a site evaluation may describe specifically what pre-existing con-
ditions are excluded in the policy. Insurers that do not inspect
the site, however, may include a general pre-existing condition
exclusion in the policy. The following is a typical example of
such an exclusion:
This insurance does not apply to LOSS:
(1) arising from POLLUTION CONDITIONS existing
prior to the inception of this policy, if any officer, director,
partner or other management personnel of the NAMED
INSURED knew or could have reasonably foreseen such
POLLUTION CONDITIONS would give rise to a claim.

In its analysis of the exclusion of pre-existing conditions, the
U.S. EPA has begun to investigate whether insurance policies
containing such exclusions satisfy the regulatory requirements for
liability coverage. While the U.S. EPA intended to allow policies
with exclusions that were consistent with standard insurance in-
dustry practice, no evidence has been found to indicate whether
the U.S. EPA considered the exclusion of pre-existing conditions
to fall within the framework of "standard industry practices."
It is also not clear if the U.S. EPA meant "standard industry
practice" to include policy provisions that the insurance industry
subsequently adopted. Currently, the U.S. EPA is researching:
(1) if the exclusion of pre-existing conditions was a standard in-
dustry practice at the time the regulations were promulgated;
(2) whether the regulatory intent was to allow exclusions that
developed over time as industry practice; and (3) if the current in-
dustry practice is indeed to exclude pre-existing conditions. Once
these issues are resolved, the U.S. EPA can make a final deter-
mination on the exclusion issue.
The U.S. EPA also has been following another new develop-
ment in the insurance market, namely the provision of coverage
by captive insurers and risk retention groups. This kind of cover-
age could potentially increase the availability of liability coverage.
One type of a captive, known as a pure captive, is owned and
controlled by only one company, while a group captive is owned
and controlled by a number of companies. A risk retention group
is a special type of group captive that can only offer coverage to
its members who are limited to companies or professionals with
similar exposures. Generally, group captives are not limited with
respect to their membership.
The passage of the Liability Risk Retention Act of 1986 will
make it easier for risk retention groups and some group captives
providing commercial casualty coverage to form and operate na-
LIABILITY/INSURANCE/DEREGULATION 17
-------
tionwide. For example, the Act allows a risk retention group or
group captive that is chartered and licensed in one state to oper-
ate nationwide without having to obtain prior rate and policy
form approval from every state in which it does business. While
the U.S. EPA anticipates that the availability of insurance should
be enhanced by the potential growth of another market for pollu-
tion liability insumace, the specific impacts will depend on the
extent to which the captives satisfy the U.S. EPA regulations.
For example, the risk retention group being formed by NSWMA
proposes to include legal defense costs in its coverage, which is in-
consistent with the U.S. EPA's regulations.
CONCLUSIONS
The U.S. EPA is committed to an effective financial responsi-
bility program for RCRA facilities, including a workable set of
requirements for liability coverage. The U.S. EPA believes that
the financial responsibility requirements, in conjunction with the
technical requirements, are necessary to ensure that hazardous
waste treatment, storage and disposal facilities': are managed
soundly in order to protect human health and'the .environment.
The U.S. EPA recognizes, however, the need to balance its man-
date from Congress to require TSDF owners and operators to
demonstrate liability coverage with prevailing conditions in the in-
surance market.
In developing the liability coverage rules, the U.S. EPA has
learned a great deal about the insurance industry. One of the
most valuable lessons is that the insurance market expands and
contracts to respond to ever-changing economic conditions. The
U.S. EPA has also learned that within this shifting market, the
types of coverage available change in response to events such as
judicial opinions and major pollution accidents.
Although the U.S. EPA now realizes that the required insur-
ance coverage will not always be widely available, it continues to
believe that liability insurance is the most useful and efficient lia-
bility assurance mechanism offered. The U.S. EPA therefore will
continue to work with the insurance industry to adapt its regula-
tory requirements to the realities of the insurance market. At the
same time, the U.S. EPA will be exploring new regulations and
enforcement policies to bring into compliance even those owners
and operators who cannot obtain insurance, or who obtain insur-
ance that is inadequate as a full measure of financial responsibil-
ity.
18 LIABILITY/INSURANCE/DEREGULATION
-------
Small Quantity Generator
Liability and Regulatory Compliance

Robert Deyle
Technology and Information Policy Program
Rosemary O'Leary, J-D.
Maxwell School
Syracuse University
Syracuse, New York
ABSTRACT
Hazardous waste generators, including so-called "small quan-
tity generators," face substantial potential liability associated
with the generation, handling, storage, treatment and disposal of
hazardous wastes. An empirical study of small quantity genera-
tor regulatory compliance in New Jersey was undertaken to de-
termine the extent to which knowledge of hazardous waste liabil-
ity and organizational concern for liability are significant factors
in waste management and regulatory compliance decision-
making.
Results of a survey of small quantity generators currently in
and outside the New Jersey manifest program indicate that ac-
cess to and use of legal and/or environmental management ex-
perts is the principal factor associated with compliance and vol-
untary adherence with an array of hazardous waste regulations.
Liability knowledge of the person primarily responsible for haz-
ardous waste management hi an organization was found to be
significantly correlated with compliance and voluntary adherence
with certain hazardous waste regulations, while the organization's
overall concern with hazardous waste liability was significantly
correlated with other compliance measures. Results of the study
support the need for technical assistance and educational pro-
grams directed at small quantity generators with 20 or fewer
employees to supplement available legal and environmental man-
agement expertise and to educate waste managers about the spe-
cifics of hazardous waste liability.

INTRODUCTION
Generators of hazardous wastes face potential liability costs of
considerable magnitude based on a number of federal and state
statutes and regulations, common law and local ordinances. Reg-
ulations promulgated pursuant to the 1984 RCRA amendments
have extended the potential liability that pertains to small quan-
tity hazardous waste generators by lowering the regulatory thres-
hold for major RCRA requirements to generators of 100 or more
kilograms per month of hazardous waste. Small quantity genera-
tors are those that produce between 100 and 1,000 kg of haz-
ardous waste per month.
Several studies of hazardous waste management have indicated
that concern for legal liability can be a significant decision-mak-
ing factor for waste generating organizations. Small businesses,
which account for most small quantity generators, face signifi-
cant knowledge and expertise constraints. These constraints can
have a major impact on the extent to which firms make use of
information concerning liability in their waste management and
regulatory compliance decisions.
In this paper, we briefly examine the various sources of liabil-
ity that pertain to small quantity generators. We then analyze
the significance of liability concern, liability knowledge and avail-
able legal and environmental expertise in regulatory compliance
decision-making by small quantity generators in New Jersey. New
Jersey offers an excellent setting for such a study because small
quantity generators in the state have been subject to the same lev-
els of hazardous waste regulation as large quantity generators
since 1981. We conclude with a discussion of the implications of
our findings for programs designed to enhance small quantity
generator regulatory compliance.

LIABILITY OF SMALL QUANTITY
GENERATORS
Case in Point
Consider the case of one small quantity generator of hazard-
ous wastes, a university. This small, private university generates
hazardous wastes in the form of laboratory wastes and waste oil
from its vehicle maintenance facility. The university stores its
wastes on-site in 55-gal drums until enough have accumulated to
make removal cost-effective. The university administration does
not know where the wastes are taken or how they are actually dis-
posed. In fact, for years the drums have been disposed of at the
local landfill which has just been declared a Superfund site. In
addition, the stored drums have leaked several times, allowing
some waste liquids to run from the university's property (which
is leased) to a brook and ultimately into navigable water.

Liability
The potential liability facing this small quantity generator of
hazardous waste is great. First, suit can be brought against them
by a number of plaintiffs: the Federal government, the state gov-
ernment, a local government and/or an individual. Next, they can
be sued in a variety of capacities, including: (1) as a generator of
hazardous wastes, (2) as an owner or operator of the waste gen-
erating facility at the time of disposal of the hazardous wastes,
and/or (3) as one who arranged for disposal of the hazardous
wastes.
Last, they can be sued under a number of Federal statutes,
including RCRA, CERCLA and CWA. Suit also can be brought
under state nuisance, negligence and trespass laws as well as state
environmental statutes. Many municipalities also have local ordi-

LIABILITY/INSURANCE/DEREGULATION 19
-------
nances governing the disposal of hazardous wastes.

RCRA Liability
RCRA regulates hazardous wastes from generation to disposal
or "cradle to grave."1 There are several major sections of RCRA
which are of particular importance to small quantity generators
of hazardous wastes. These concern: (1) on-site storage of haz-
ardous wastes,' (2) a manifest system for tracking hazardous
wastes from generator to transporter to disposal facility' and
(3) Federal minimum standards for hazardous waste disposal,
enforced through a permit system for disposal facilities.4 Regula-
tions promulgated pursuant to the 1984 amendments of this act
have extended the potential liability pertaining to small quantity
hazardous waste generators by lowering the regulatory threshold
for major RCRA requirements to generators of 100 or more kilo-
grams per month of hazardous wastes.9
If convicted of violating any applicable requirements of RCRA,
a small quantity generator may face civil penalties of up to
$25,000/day.' Criminal penalties under the same statute can
range from as much as $25,000/day and/or 1 yr imprisonment
to $50,000/day and/or 5 year imprisonment.7 Additionally, the
act contains penalties for the offense of "knowing endanger-
ment" for which penalties can range up to $250,000 and/or 15 yr
in prison.' If the defendant is an organization, that organization
may be subject to a fine of up to $1,000,000 for "knowing en-
dangerment.'"

CERCLA Liability
CERCLA was enacted in 1980 to deal with issues not covered
by RCRA. The act established a "Superfund" to enable the fed-
eral government to clean up hazardous waste disposal sites.'" It
expanded the class of private parties liable for such sites, includ-
ing site owners and operators, waste transporters and genera-
tors." The act also authorized the U.S. EPA to go to court to
force the abatement of hazardous waste pollution at Superfund
sites that pose an "imminent and substantial" threat to public
health or the environment.''
Courts have universally held that CERCLA imposes strict lia-
bility," meaning a generator (including a small quantity gener-
ator) may be subject to liability even if they have not departed in
any way from the standard of reasonable care. Additionally,
where injury is indivisible (i.e., where it is not clear whose wastes
caused what damage), liability is joint and several.'4 Joint and
several liability means that the government can proceed against
any one or a group of potentially liable parties for the total costs
of the cleanup, for which each is jointly or individually liable. If
such liability is established, the only remedy for the defendants is
to seek contribution from other responsible parties, if they can
be found, through legal action.
Compounding the issue is the fact that the government need
not prove that a defendant's actions were the cause of the envi-
ronmental threat or harm in question. Courts have held that
CERCLA requires no more proof than showing that generators
sent out hazardous wastes for disposal, that the wastes ended up
at a site at which a release or threatened release of any hazardous
substance necessitated response action and that wastes of the type
the generator disposed of are present at the site.''
Other Federal Statutes
In addition to RCRA and CERCLA, a variety of other Fed-
eral statutes can be used to bring suit against the hazardous waste
generator. For example, if hazardous wastes are discharged into a
waterway, the Clean Water Act" may apply. If wastes are in-
jected into the ground, the Safe Drinking Water Act" may be
triggered. If wastes are burned, the Clean Air Act" may be vio-
lated. All of these statutes have been used, either alone or in con-
junction with other laws, to bring suit against small quantity gen-
erators of hazardous waste.

Common Law
The small quantity generator of hazardous waste also may be
sued under a variety of common law doctrines, including negli-
gence, trespass and nuisance. Negligence has been defined as
"conduct which falls below the standard established by law for
the protection of others against unreasonable risk of harm."" To
establish a cause of action in negligence, four elements must be
proved. It must be shown that the defendant was under a duty to
conform to a standard of conduct; that the defendant breached
that duty; that there was a reasonably close connection between
defendant's conduct and plaintiff's injury; and that the plaintiff
suffered actual loss or injury. A degree of care on the part of the
defendant can be imposed or implied by a statute such as
RCRA."
If a generator's wastes interfere with another's possessory in-
terest in land, a cause of action in trespass may be possible. In
this instance, there must be entry upon the land either by the
defendant or by something which the defendant has put into
motion (i.e., a discharge of hazardous waste that was either inten-
tional or the result of an accident or spill).21 The injury must be
related to the invasion of the land." The trespass may be inten-
tional, negligent or the result of ultrahazardous activity."
Last, an action based on the doctrine of nuisance also may be
brought. A nuisance is a substantial, unreasonable interference
with another's use and enjoyment of land.24 A nuisance action for
pollution may be brought under state law or, where pollution has
an interstate effect, under federal common law.29 The resolution
of a nuisance claim requires a balancing of the equities. The
court must weigh the utility of the activity that creates the inter-
ference against the gravity of the resulting harm.2' A nuisance suit
can result from an intentional or unintentional discharge of haz-
ardous waste.

State Statutes
State statutes also can be used to bring suit against small quan-
tity generators of hazardous waste. Today, most states have en-
acted hazardous waste regulatory programs that parallel RCRA.
By law, these programs must be at least as stringent as federal
programs. Many are more stringent. State statutes comparable to
CERCLA and other Federal pollution laws are common. A num-
Table 1
Potential Sources of Small Quantity Generator Liability

Resource Conservation and Recovery Act
Comprehensive Environmental Response, Compensation and Liabfliff
Act
Other Federal Statutes
Examples:
Clean Water Act
Safe Drinking Water Act
Clean Air Act
Common Law
Negligence
Trespass
Nuisance
State Statutes
Examples:
RCRA Parallels
CERCLA Parallels
Municipal Codes
20
LIABILITY/INSURANCE/DEREGULATION
-------
her of states also have enacted legislation intended to strengthen
a plaintiff's cause of action for hazardous waste injuries. Any or
all of these state laws have been used alone or in conjunction
with other state and federal laws to bring suit against small quan-
tity generators of hazardous wastes."
LIABILITY AND WASTE MANAGEMENT
DECISIONS
The potential magnitude of liability resulting from hazardous
waste generation and management is clearly substantial for small
quantity as well as large quantity generators. Studies of genera-
tor waste management practices have indicated that the costs of
hazardous waste liability are, in fact, considered by some genera-
tors in their waste management decisions.
A 1980 analysis of national hazardous waste generation, con-
ducted by Booz, Allen and Hamilton and Putnam, Hayes and
Bartlett, concluded that concern with corporate liability is a fac-
tor that motivates industries to reduce the amount of wastes they
generate." A recent study by Deyle found that liability concern
is not only a significant factor in source reduction decisions, but
also stimulated other changes in corporate waste management
practices, including the following:

• Shifts to waste disposal practices perceived as posing less liabil-
ity risk, e.g., changing from land disposal to incineration
• Changes to reduce reliance on second and third parties, includ-
ing development of on-site treatment and disposal capacity
and use of company-owned transport vehicles
• Initiation of regular inspections and evaluations of treatment,
storage and disposal facilities with which companies contract
for waste management services"
Deyle's study, which was based on interviews with waste man-
agement consultants and waste generators, suggests that liability
concern is likely to be greater in firms that have environmental or
legal staff who are aware of the potential costs posed by haz-
ardous waste liability. His results also indicate that liability con-
cern is more likely to be a significant factor in actual waste man-
agement decision-making where in-house or corporate environ-
mental managers or engineers are involved in hazardous waste
management.
Deyle's findings also imply, however, that mere expression of
concern for corporate hazardous waste liability does not imply
that liability is necessarily a significant decision-making factor.
While 10 of 13 waste managers interviewed said their firms were
"very concerned" about hazardous waste liability, only three
cited long-term corporate liability as a significant source reduc-
tion decision factor.

SMALL QUANTITY GENERATOR
DECISION-MAKING
Most small quantity generators are also small businesses. Ac-
cording to a study by ICF, 67"% of small quantity generators
have fewer than 50 employees and 78% have fewer than 100 em-
ployees.30
Real world decision-makers typically do not have the resources
or computational capacity to be knowledgeable about all decision
alternatives or their consequences and probabilities. Thus, rather
than seeking optimal choices, decision-makers compromise by
searching for satisfactory alternatives within the limits of their
knowledge and computational resources." The decision-making
of small businesses is particularly affected by knowledge and ex-
pertise constraints," and these constraints can affect their abil-
ities to comply with environmental regulations."
Smaller businesses usually have fewer financial resources to in-
vest in staff devoted to nonproductive functions such as waste
management and regulatory compliance." Companies with fewer
than 100 employees are less likely to be able to afford an engineer
who can stay abreast of regulations and interpret and imple-
ment them. Smaller businesses also are less likely to hire engineer-
ing or legal consultants to assist in determining the extent to
which environmental regulations apply to them or in solving
technological problems of regulatory compliance. Where small
business owners or operators do not have the resources to hire
in-house or consultant environmental experts, their own time con-
straints preclude attaining a thorough understanding of regula-
tions and the technical means of complying with them.35
Based on qualitative studies of small quantity generator haz-
ardous waste management, several authors have concluded that
educational and technical assistance programs without the threat
of enforcement sanctions are essential to improving small quan-
tity generator regulatory compliance.36

SMALL QUANTITY GENERATORS IN
NEW JERSEY: AN EMPIRICAL ANALYSIS
An empirical study was made of small quantity generators in
New Jersey in an effort to identify significant factors that affect
regulatory compliance decision-making. The study distinguished
between regulated small quantity generators (hereafter referred
to SQGs) which generate between 100 and 1,000 kg/mo of haz-
ardous waste and very small quantity generators (hereafter re-
ferred to as VSQGs) which generate less than 100 kg/mo. In New
Jersey, SQGs are subject to the same level of regulation as gen-
erators of more than 1,000 kg/mo (large quantity generators or
LQGs). Under Federal regulations promulgated pursuant to the
1984 amendments to RCRA, and in most other states, SQGs are
not subject to the same requirements as LQGs.

Survey Administration and Response
A mailed survey was administered to a sample of 1,000 estab-
lishments drawn from two populations in the state: (1) estab-
lishments currently in the state manifest system that generate haz-
ardous wastes in amounts that qualify them as SQGs or VSQGs
and (2) establishments outside the manifest system that are in one
of the Standard Industrial Classification (SIC) codes determined
to have a high proportion of SQGs and VSQGs.37
A stratified random sample of 106 VSQGs and 194 SQGs was
drawn from the population of small quantity generator organiza-
tions currently in the manifest system. A random sample of 700
establishments was drawn from the population of approximately
39,550 organizations in the small quantity generator SIC codes
not currently in the regulatory system. An overall response rate
of 41.4% (414 responses) was achieved: the response rate for the
manifest subsample was 37.5% (113 establishments); and the rate
for the nonmanifest subsample was 43.1% (301 responses).
Surveys were directed to the person most knowledgeable about
the organization's waste management practices. Thirty-one per-
cent of the respondents whose firms generate some hazardous
waste were owners of the firm, and 23% indicated that they were
the president or a member of upper-level management. Only 34%
reported that they were operating managers or engineers. The
median number of employees per firm was 18.

Regulatory Compliance Measures
Four measures of small quantity generator regulatory compli-
ance were employed in the New Jersey study:
• Use of the state manifest system for off-site shipments of haz-
ardous waste
• Provision of notice to haulers for off-site shipments (combined
measure of use of the manifest plus container labelling)
LIABILITY/INSURANCE/DEREGULATION 21
-------
• Development and maintenance of hazardous waste contingency
plans
• Employee hazardous waste training
As noted above, SQGs in New Jersey are subject to the same
regulatory requirements as LQGs, including development and
maintenance of contingency plans and implementation of em-
ployee training programs." Adherence to these requirements by
VSQGs is essentially voluntary in terms of statutory obligations.

Liability Concern and Regulatory Compliance
Forty-eight percent of the New Jersey survey respondents
whose organizations generate some hazardous wastes indicated
that their organizations have a relatively high level of concern
with hazardous waste liability, while 18% reported a moderate
level of concern and 34% indicated relatively little concern.
As shown in Table 2, the expressed level of concern is signifi-
cantly correlated to the use of environmental management and
engineering consultants for advice in waste management decis-
ion-making. Organizational liability concern is not, however,
significantly correlated to use of legal experts, organization size
or profitability, the respondent's job category or the respondent's
level of knowledge about hazardous waste regulations or actual
hazardous waste liability under CERCLA or RCRA.
The results of this analysis of New Jersey small quantity gen-
erators are consistent with Deyle's finding that alleged organiza-
tional liability concern is not always directly reflected in organ-
izational behavior. The expressed level of organizational concern
for liability was not found to be significantly correlated to re-
ported compliance or voluntary adherence to manifesting and
container labelling requirements. Voluntary adherence to contin-
gency plan requirements is, however, significantly correlated to
liability concern, as shown in Table 3. Voluntary adherence to
employee training requirements also was significantly correlated
to liability concern for the nonmanifest subsample.
Table 2
Correlations Between Liability Concern and
Respondent Characteristics
Table 3
Correlations Between Liability Concern and
Measures of Regulatory Compliance
Respondent
Character I atIc

use of legal
consuItants

access to In- house
legal experts

*use of environ-
mental management
or engineering
consuItants

access to In-house
envIronmentaI
management or
engineering experts

total employment

prof I tab I IIty

respondent liability
knowledge

respondent regulatory
know ledge
KendaiI's tau-b
Correiat ion*
ProbablIIty
Value"
O.OO

0.06

O.OO

0.12

0. 10
O.O4

0.98

0.27

O.97

O. 14

O.36
• The Kcndall'i lau-b correlation coefficient measures the degree of concordance between pairs
of observations for the two variables of concern. A correlation of 1.0 indicates that both vari-
ables vary in iht same direction for all possible pairings of the observations in the data set. A
correlation coefficient of -1.0 indicates lhat all the pairs are discordant, i.e., the variables
vary in opposite directions.

•• The probability value is • measure of the significance of the measured relationship. For a
probability value of 0.10. one can be 90* certain that the measured relationship is correct.
1 Correlation is significant at the 0.10 lev el or better.

22 LIABILITY/INSURANCE/DEREGULATION
CompI Iance
Measure
KendaiI's tau-b
CorrelatIon
use of man Ifest
system O. 16

voluntary use of
manifest system -0.05

hauler notice 0.09

voluntary hauler
notice 0.03

contingency plan
maintenance -O.01

Tvoluntary contin-
gency plan
maintenance O.26

employee training 0.05

''voluntary employee
traInIng* O.24
ProbablIIty
Value
0.11

0.73

0.35

0.79

0.69

0.02

0.55

0.04
+ Correlation is significant at the 0.10 level or better.
• Correlation reported is for the nonmanifest subsample. Pooled sample did not show signlflcul
correlation.

Liability Knowledge and Regulatory
Compliance
While the level of CERCLA and RCRA liability knowledge of
an organization's waste manager is not significantly related to the
alleged level of organizational concern with hazardous waste lia-
bility, liability knowledge is significantly correlated with regula-
tory compliance behavior, including voluntary adherence with
some hazardous waste regulatory requirements.
Survey respondent knowledge of hazardous waste liability
under CERCLA and RCRA was measured with a series of four
true-false questions (Table 4). Variables based on three of these
questions (LITKNOW1, LITKNOW3 and LITKNOW 4) were
combined to provide a composite measure of liability knowledge
(LITKNOW) use of the Guttman scaling procedure."
The scale for the composite liability knowledge measure ranges
from 0 to 3 based on the number of individual liability knowl-
edge questions answered correctly. Only 13% of the respondents
whose organizations generate some hazardous waste answered all
three questions correctly, 36"% answered them all incorrectly,
44% answered one correctly and 7% answered two correctly.
Table 4
Hazardous Waste Liability Knowledge Measures
Survey Question
Variable Co*
Organizations that generate hazardous wastes would LITKNOW 1
not be liable if their wastes were removed from their
property by a second party.
Organizations that generate hazardous wastes would LITKNOWJ
be liable if their wastes were delivered to a second party
by themselves or by another transporter, but the
liability would not extend indefinitely.
Organizations that generate hazardous wastes would LITKNOWJ
not be liable if their wastes were disposed of at a treat-
ment or disposal facility with a permit to handle
hazardous wastes.
Organizations that generate hazardous wastes would LITKNOW4
be liable regardless of how their wastes were disposed,
and the liability would extend indefinitely.
-------
The availability or use of in-house or consultant environmen-
tal management or engineering experts in hazardous waste man-
agement decision-making is strongly correlated to the composite
measure of liability knowledge (Table 5). There is not, however,
a similarly significant correlation to access to or use of in-house
or consultant legal experts. Liability knowledge is significantly
correlated to organization size, which probably reflects the find-
ing that larger organizations are more likely to have access to or
use environmental management experts (Table 6). Persons re-
sponsible for hazardous waste management in larger firms also
may have more time to educate themselves about liability issues.
Liability knowledge is not, however, significantly correlated to
the respondent's job category. This suggests that even where the
person responsible for waste management is the president or own-
er of a small firm, hazardous waste liability knowledge may be
effectively transferred from environmental management or engi-
neering experts to waste managers.
Liability knowledge is significantly and relatively strongly cor-
related to compliance and voluntary adherence to state manifest
regulatory requirements (Table 7). It also is significantly corre-
lated to compliance with hauler notice requirements, although
not with voluntary adherence to these regulations. Liability
knowledge is not significantly related to compliance or voluntary
adherence to contingency plan or employee training require-
ments.
Table 7
Correlations Between Liability Knowledge and
Measures of Regulatory Compliance
CompI Iance
Measure
*use of man If est
system

'"'voluntary use of
man Ifest system

+ hauIer not Ice

voluntary hauler
not Ice

contingency plan
ma Intenance

voluntary contin-
gency plan
ma Intenance

employee training

voluntary employee
traInIng
Kenda I I's tau-b
CorreI at Ion
0.34

0.34

0.26

O.28

0.18

-0.01

0. 16

O.O5
Probabl I Ity
Value
0.01

O.O7

0.03

0.12

0.13

0.96

0.14

O.73
+ Correlation is significant at the 0.10 level or better.
Table 5
Correlations Between Liability Knowledge and
Respondent Characteristics
Respondent
Character IstIc
use of In-house
or consultant
legal experts

''use of In-house
or consultant
envIronmentaI
managers or
engIneers

+total employment
Kenda\\'* tau-b
CorrelatIon
0. 10
0.21

0.27
Probabl I Ity
Value
O.30
O.04

O.OO
The availability of legal or environmental experts appears to be
the major significant factor underlying both compliance and vol-
untary adherence to hazardous waste regulations by small quan-
tity generators (Table 8). Analysis of relationships with the sev-
eral measures of legal and environmental expertise produce sig-
nificant correlations to voluntary use of the manifest system,
voluntary adherence to hauler notice requirements and compli-
ance and voluntary adherence to contingency plan and employee
training requirements.
Analysis of the relationships between establishment size, rates
of compliance and voluntary adherence to regulatory require-
ments and use of and access to legal and environmental expertise,

Table 8
Correlations Between Compliance Measures and Expertise Variables +
'orrelation is significant at the 0. 10 level or better.

Table 6
Relation Between Establishment Size and Access to Expertise

Source of Kendall's tau-b Probability
Expert Ise Correlation Value

+ 1 n-house
lawyers O.26 0.00

+Envl ronmenta 1
management or
eng 1 neer 1 ng
consul tants 0.15 0.02
* I n-house envl -
ronmenta 1

managers or
eng Ineers 0.15 0.02
I n-house

use of
man 1 f est
system -u.06 -0.37 -0.0* 0.05
vol untary
use of
man I f est
system 0.32- u.24 -0. 14 u.28""
hau 1 er
notice -0.05 -0.09 u.OO u.07
voluntary
hau 1 er
notice 0.29"" u.14 -O.06 u.29""
cont 1 ngency
plan main-
tenance 0.27* O.29 0.28" 0.24"
voluntary
cont 1 ngency
tenanc* 0.22"" 0.09 0.36" 0.23"*

emp loyee
training O.14 0.35"" 0.15 0.18""
vo 1 untary
emp loyee
training O.23"" 0.14 u.24"" 0.22""

+ Correlation is significant at the 0.10 level or better.
+ Correlation coefficients are Kendall's tau-b values.
' Coefficient is significant at the 0.05 level of significance.
>* Coefficient is significant at the 0.10 level of significance.
LIABILITY/INSURANCE/DEREGULATION
23
-------
indicates that organizations with fewer than 20 employees are
most affected by these resource constraints. Approximately 63%
of all SQGs and VSQGs in New Jersey have fewer than 20 em-
ployees.

CONCLUSION
The results of this study indicate that access to and use of legal
and environmental expertise are positively correlated to several
measures of compliance and regulatory adherence to hazardous
waste regulations by small quantity generators. This suggests
that initiatives to make such expertise more available to small
quantity generators, especially those with 20 or fewer employees,
could bolster efforts to enhance compliance and voluntary ad-
herence to regulatory requirements.
The measured levels of CERCLA and RCRA liability knowl-
edge of persons primarily responsible for hazardous waste man-
agement in small quantity generator organizations are higher for
organizations with access to or use of environmental manage-
ment or engineering experts, but not necessarily higher for those
with access to or use of legal experts. This suggests that environ-
mental experts may be more likely than legal experts to transfer
their liability knowledge to waste managers. The results indicate,
however, that for regulatory requirements that govern direct
handling of hazardous waste (i.e., manifest and container label-
ling requirements), the waste manager's knowledge of liability is
correlated to compliance and voluntary adherence. Access to or
use of legal or environmental experts is correlated only to volun-
tary adherence with these regulations.
These results suggest that efforts to educate waste managers
about the specifics of the liability their organizations face as waste
generators under CERCLA and RCRA, particularly where they
do not have access to legal or environmental experts, may have
pay offs in terms of increased compliance levels.
The lack of correlation between reported organizational con-
cern with hazardous waste liability and six of the eight measures
of regulatory compliance and voluntary adherence to regula-
tions indicates that such concern is not necessarily a determinant
in actual waste management decision-making. Again, there is evi-
dence that environmental experts may have more influence than
legal experts on how organizations perceive their hazardous waste
liability, but the difference may not be significant from the per-
spective of achieving higher levels of regulatory compliance.
Liability concern is, however, correlated to higher levels of vol-
untary adherence with contingency plan and employee training
requirements. This suggests that efforts to raise the hazardous
waste liability consciousness of waste managers in smaller organ-
izations that do not have access to or use of legal or environmen-
tal management experts may lead to higher levels of voluntary
adherence with these hazardous waste management requirements.
The potential liability faced by small quantity generators can,
therefore, be substantial. There is evidence that where that liabil-
ity is understood, either by waste managers or by their legal or
environmental advisors, it can be a significant factor in waste
management and regulatory compliance decision-making. The re-
sults of this study of small quantity generators in New Jersey
also support the need to develop educational and technical assis-
tance programs, particularly for smaller firms with 20 or fewer
employees that are likely to be most constrained in their capacity
to afford legal and environmental management or engineering ex-
pertise.
REFERENCES
1. Goldfarb, "The Hazards of our Hazardous Waste Policy, 19 Nat.
Resources Journal 256 (1979); Note, "Liability for Generators of
Hazardous Waste: The Failure of Existing Enforcement Mechan-
isms," Geo. L.J. 69, (1981), 1051.
2. 42 U.S.C. Sections 6922-6924.
3. 42 U.S.C. Section 6924.
4. 42 U.S.C. Section 6925.
5. For an excellent discussion of hazardous waste regulations pertain-
ing to small businesses, see "Does Your Business Produce Hazardous
Wastes? Many Small Businesses Do." U.S. EPA, Washington,
DC, 1985.
6. 42 U.S.C. Section 6928(g).
7. 42 U.S.C. Section 6928(d).
8. 42 U.S.C. Section 6928(e).
9. 42 U.S.C. Section 6928(e).
10. 42 U.S.C. Section 9604.
11. 42 U.S.C. Section 9607(a).
12. 42 U.S.C. Section 9606.
13. United States v. South Carolina Recycling and Disposal, Inc. D.S.C.
Feb. 23, 1984; United States v. Conservation Chemical, W.D. Mo.
Feb. 3, 1984; NEPACCO, 579 F. Supp. at 844; United States v.
Argent Corp., D.N.M. May 4, 1984; United States v. Caufman,
C.D. Cal. Oct. 23, 1984; United States v. Ottati & Goss, D.N.H.
July 25, 1984; Bulk Distribution Centers v. Monsanto Co., 589 F.
Supp. 1437 (1985).
14. Colorado v. ASARCO, Inc., May 13,1985.
15. United States v. South Carolina Recycling and Disposal, Inc., D.S.C.
Feb. 23,1984; United States v. Wade, E.D. Pa. 1983.
16. United States v. OttatiA Goss, D.N.H. July 25,1984.
17. For a discussion of this issue see DiBenedetto, J. "Generator Lia-
bility Under the Common Law and Federal and State Statutes,"
Bus. Law 39, (1984), 611.
18. Id. at 623.
19. Restatement (Second) of Torts, Section 282 (1965).
20. For a discussion of this issue see DiBenedetto, J. "Generator Lia-
bility Under the Common Law and Federal and State Statutes,"
Bus. Law. 39, (1984), 611.
21. Prosser, W., Handbook of the Law of Torts (4th ed. 1971); see
also City of Philadelphia v. Stepan Chem. Co., 544 F. Supp. 1135
(1982).
22. Id.
23. Id.
24. Id.
25. DiBenedetto, J. supra note (20).
26. Prosser, supra note (21).
27. DiBenedetto, J. supra note (20).
28. Booz, Allen and Hamilton, and Putnam, Hayes and Bartlett, Haz-
ardous Waste Generation and Commercial Hazardous Waste Man-
agement Capability, U.S. EPA, Washington, DC, 1980.
29. Deyle, Robert E., "Source Reduction by Hazardous Waste Generat-
ing Firms in New York State," Syracuse University Technology and
Information Policy Program Working Paper No. 85-010.
30. ICF Incorporated, Economic Analysis of Resource Conservation
and Recovery Act Regulations for Small Quantity Generators,
U.S. EPA, Washington, DC, 1985.
31. Cyert, R.M. and March, J.G., A Behavioral Theory of the Firm
(Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1963); Simon, Herbert
A., "Rational Decision-Making in Business Organizations," Am.
Econ. Rev. 69. Sept. 1973.
32. Rice, G.H. and Hamilton, R.E., "Decision Theory and the Small
Businessman," Am. J. of Small Business, 4, July 1979.
33. CONSAD Research Corporation, Environmental Regulations and
Small Businesses: An Overview of Issues Concerning the Economic
Impact of EPA Regulations on Small Businesses (Washington,
24 LIABILITY/INSURANCE/DEREGULATION
-------
DC: U.S. EPA, 1983); Environmental Resources Management,
Inc., Hazardous Waste Facilities Needs Assessment Appendix D.
Small Quantity Generator Study (Albany, NY: New York State
Department of Environmental Conservation, 1985); Small Business
Association of New England, "Hazardous Waste Survey" (June
1983); United States Chamber of Commerce, "Survey on Regula-
tion" Washington Report, Aug. 3,1982.
34. Sloan, W.M., Hunt, G. and Walters, R., "An Approach to Techni-
cal Assistance for Industrial and Hazardous Waste Generators,"
in Redmond Clark, ed., Massachusetts Hazardous Waste Source
Reduction Conference, Massachusetts Department of Environ-
mental Management, Bureau of Solid Waste Disposal, Boston, MA,
1983.
35. CONSAD Research Corporation, Environmental Regulations and
Small Businesses: An Overview of Issues Concerning the Economic
Impact of EPA Regulations on Small Businesses, U.S. EPA, Wash-
ington, DC, 1983.
36. U.S. EPA Small Business Task Group, "U.S. EPA Small Business
Initiatives—Strategy for Improved Regulation and Compliance,"
Washington, DC, Sept. 28, 1984; Environmental Resources Man-
agement, Inc., Hazardous Waste Facilities Needs Assessment Ap-
pendix D Small Quantity Generator Study (Albany, NY: New York
State Department of Environmental Conservation, 1985); Nemeth,
John C. and Kevin L. Kamperman, The Georgia Tech Hazardous
Waste On-Site Consultation Program: Approach and Results
(Atlanta, GA, 1985).
37. Small quantity generator SIC codes were derived from the National
Small Quantity Hazardous Waste Generator Survey (U.S. EPA,
Washington, DC, 1985), prepared by Abt Associates.
38. New Jersey Administrative Code, Title 7, Department of Environ-
mental Protection, Chapter 26, Section 7:26- 9.3- 9.4, 9.6-9.7.
39. Mclver, J.P. and Carmines, E.G., Unidimensional Scaling, (Sage
Publications, Beverly Hills, CA, 1981).
LIABILITY/INSURANCE/DEREGULATION 25
-------
Financial Liabilities for Natural Resource Damages
Stephen Wyngarden
Michael Goldman
ICF Incorporated
Washington, D.C.
ABSTRACT
CERCLA and the Superfund Amendments and Reauthoriza-
tion Act (SARA) authorize government officials to sue respon-
sible parties for damages to natural resources caused by the re-
lease of oil and hazardous substances. Responsible parties are
liable for two basic categories of costs: (1) the dollar value asso-
ciated with an injury to natural resources; and (2) the costs of as-
sessing natural resource damages. These general categories have
been translated into more specific liabilities by recently issued
regulations by the Department of the Interior (DOI).
The DOI regulations offer several opportunities for respon-
sible parties to take certain actions which can affect their liabil-
ities. In particular, responsible parties can make inputs to the
selection of the procedures by which natural resource damages
will be assessed as well as play an active role in the damage assess-
ment.

INTRODUCTION
CERCLA creates many potential liabilities for parties respon-
sible for releases of oil or hazardous substances that cause dam-
ages to natural resources. Although CERCLA was passed 6 yr
ago, liabilities for natural resource damages are just now being
made clear by recently developed regulations and completed dam-
age assessment cases.
This paper focuses on responsible party liabilities for natural
resource damages in accordance with the natural resource dam-
age provisions of CERCLA. An increased emphasis has been
placed on such responsible party liabilities by the Superfund
Amendments and Reauthorization Act of 1986 (SARA), which
prohibits future payments from the Hazardous Substance Re-
sponse Trust Fund (the Fund) for natural resource damages. The
remainder of the paper is organized into four main sections: a
brief outline of the key natural resource provisions of CERCLA,
SARA and associated regulations; a description of the costs re-
sponsible parties may have to bear for natural resources (the
costs are broken down into damages for natural resource injur-
ies and costs for performing damage assessments); an outline of
the opportunities available to responsible parties that may affect
their natural resource damage liabilities; and conclusions.

STATUTORY AND REGULATORY
PROVISIONS
CERCLA creates general responsible party liabilities for
natural resource damages, provides for the recovery of natural
resource damages through the Fund and calls for the issuance of
regulations governing the natural resource damage assessment
and claims procedures. Some important amendments to these
provisions are contained in SARA. Regulations that translate
the CERCLA provisions into specific liabilities and procedures
have been developed by the U.S. EPA and the Department of
the Interior (DOI). CERCLA and SARA, the EPA regulations
and DOI regulations are described separately below.

Key Provisions of CERCLA and SARA
Natural resources are defined by CERCLA to include land,
fish, wildlife, biota, air, water, groundwater, drinking water sup-
plies and other such resources belonging to the government. If
these natural resources are injured as the result of a release of oil
or hazardous substances, that injury can be translated into a
dollar amount (i.e., damages) that can be recovered by govern-
ment officials acting on behalf of the public as trustees of natural
resources. As defined under CERCLA, trustees can recover these
damages from two possible sources: responsible parties and the
Hazardous Substance Response Trust Fund.
Section 107 of CERCLA establishes liability for responsible
parties. As called for in this section, responsible parties are liable
for "...damages for injury to, destruction of, or loss of natural
resources, including the reasonable costs of assessing such injury,
destruction, or loss...." Section 107 goes on to state that the
measure of natural resource damage shall not be limited to the
amount of money needed to restore or replace the injured re-
sources. This latter statement has been interpreted to mean that
responsible parties are also liable for sums needed to compen-
sate for the public's loss of use of the injured resource. Respon-
sible party liability for natural resource damages under CERCLA
is limited to $50,000,000 for each release of a hazardous sub-
stance.
If the responsible party is not known or is unable to pay for the
entire damage, CERCLA's Section 111 allows natural resource
damage claims to be asserted against the Fund. Section 517 of
SARA, however, precludes the use of any Superfund money for
natural resource damage claims asserted against the Fund after
Jan. 1, 1987. Accordingly, trustees desiring to restore or replace
injured natural resources will be forced to seek damage awards
from responsible parties or to pay for restoration actions out of
their own appropriations.
SARA amends the natural resource damage provisions of
CERCLA in three additional respects. First, CERCLA's statute
of limitations has been amended so that natural resource dam-
age assessments must now commence within 3 yr of the date of
discovery of the loss and its connection with the release in quei-
tion or the date on which DOI's natural resource damage assess-
ment procedures are promulgated (see discussion below), which-
ever is later. CERCLA's previous statute of limitations required
damage assessment to commence within 3 yr of the date of dis-
covery or 3 yr from the enactment of CERCLA, whichever il
later. The new statute of limitations further requires that dam-
26 LIABILITY/INSURANCE/DEREGULATION
-------
age assessments must be commenced within 3 yr after comple-
tion of the remedial action at a site. CERCLA has also been
amended to clarify that damage assessments performed by state
trustees in accordance with the DOI damage assessment proced-
ures are entitled to receive a rebuttable presumption; that is, evi-
dentiary weight in an administrative or judicial proceeding. Pre-
viously, CERCLA had been interpreted to afford a rebuttable
assumption to only those natural resource damage assessments
performed by Federal trustees. Finally, SARA clarifies the role
of Federal and state trustees by requiring the Governors of states
to designate trustees for natural resources and authorizing Fed-
eral trustees to assess damages for natural resources under a
state's trusteeship upon request and reimbursement from a state.

U.S. EPA Regulations
On Dec. 13, 1985, the U.S. EPA promulgated final regula-
tions governing procedures for filing natural resource damage
claims against the Fund (50 FR 51205) and prolcedures for arbi-
trating disputes over those claims (50 FR 51196). While these
regulations will be important for existing claims against the Fund,
the U.S. EPA rules will not apply to future claims because of
SARA's prohibition against the use of Superfund money to pay
for natural resource damages.

DOI Regulations
DOI is responsible for issuing procedures for assessing natural
resource damages. CERCLA requires two types of natural re-
source damage assessment procedures to be developed:
• Standard procedures for simplified assessments requiring min-
imal field observation, including established measures of dam-
ages based on units of release or units of affected area (these
procedures are called "Type A" procedures); and
• Alternative protocols for conducting assessments in individual
cases to determine the type and extent of short- and long-term
injury, destruction or loss ("Type B" procedures).
DOI issued final Type B regulations on Aug. 1, 1986, (51 FR
27674) and proposed Type A regulations for coastal and marine
environments on May 5, 1986, (51 FR 16636). At present, DOI is
under court order to issue final Type A regulations by Feb. 4,
1987. Trustees are free to use any natural resource damage assess-
ment procedures they desire; however, as mentioned previously,
CERCLA and SARA specify that Federal and state trustees must
follow the DOI procedures if their damage assessment is to be
accorded a rebuttable presumption.
A simplified flow diagram illustrating the damage assessment
procedures, as well as the interaction between the final Type B
procedures and proposed Type A procedures, is shown in Fig. 1.
After a potential natural resource injury is recognized and appro-
priate trustees are notified, trustees initiate an assessment by per-
forming a preassessment screen to determine whether an assess-
ment is warranted. If this screen results in a positive determina-
tion, the next phase is to prepare an assessment plan in which
the methodologies for the assessment are to be documented in
detail. The purpose of the assessment plan is to assure that only
reasonable costs will be incurred in the assessment and, there-
fore, only reasonable costs will be passed on to the responsible
party. As part of this planning phase, trustees are to decide
whether to use the Type A or Type B procedures. The proposed
Type A regulations state that, in making this decision, prefer-
ence should be given to the Type A procedures. Specifically, trus-
tees must use the Type A procedures to evaluate releases in a
coastal or marine environment unless one or more listed con-
ditions are not satisfied.
To use the Type A procedures, trustees need only apply the
necessary site-specific data needed to run a natural resource dam-
age assessment model. These data inputs include information on
the identity and quantity of the substance released, information
on the extent of any cleanup activities and environmental char-
acteristics (e.g., ocean currents, wind speeds and directions and
temperatures). Using the inputs, the model determines what (if
any) natural resource injury has occurred, quantifies the extent of
that injury and translates the injury into a dollar amount. No
confirmation of natural resource exposure or injury is required to
verify the model's results. In the post-assessment phase, trustees
use the model's damage estimate as the basis for a demand for
payment by the responsible party.

Potential
Natural Resource
Injury
1
Yes |
H
i_

Model
Application

Post
Assessment

Type A Assessment
Type B
Assessment
Figure 1
CERCLA Natural Resource Damage Assessment Procedures
The Type B procedures are much more lengthy and compli-
cated to execute than the Type A procedures. In a Type B assess-
ment, trustees are required to confirm natural resource exposures
and to determine injuries by conducting field sampling and test-
ing. The extent of the injury must be measured and compared to
a baseline condition to quantify injuries caused by the release or
discharge. Once the injury is quantified, it is translated into a
dollar amount (i.e., an estimate of damages) using an appropriate
economic methodology which must be developed on a case-by-
case basis. Finally, as part of the post-assessment phase, the trus-
tee presents to the responsible party a demand for a sum that rep-
resents the determined damages.
LIABILITY/INSURANCE/DEREGULATION 27
-------
COSTS TO BE BORNE BY
RESPONSIBLE PARTIES
As mentioned above, CERCLA's natural resource damage pro-
visions specify that parties responsible for releases of oil or haz-
ardous substances are liable for two basic categories of cost: (1)
damages for injury to natural resources; and (2) the costs of
assessing natural resource damages. These general categories,
which have been translated into more specific liabilities by the
DO! regulations, are discussed below.

Natural Resource Damages
There are two key points that must be considered to deter-
mine natural resource damage liabilities. First, because respon-
sible parties are liable for damages for "injury" to natural re-
sources, a party is liable only if a release causes a change in a
natural resource that meets the definition of injury under
CERCLA. Second, given that an injury has occurred, the way the
injury is translated into a dollar amount or measure of damage
will ultimately determine responsible party liability.
A natural resource injury is defined under CERCLA by a series
of exclusions, threshold criteria and miscellaneous provisions.
The primary exclusions, which are embodied in CERCLA's
Section 107, are that responsible parties are not liable for:

• Injuries that were identified as an irreversible and irretriev-
able commitment of natural resources in accordance with a
permitted activity in an environmental analysis
• Releases and subsequent injuries that occurred wholly before
the enactment of CERCLA
• Injuries resulting from the application of a pesticide product
registered under the Federal Insecticide, Fungicide and Roden-
ticide Act
• Injuries resulting from Federally permitted releases (although
recovery for such injuries may be allowed under other laws)
• Injuries resulting from actions taken in the course of render-
ing assistance in accordance with the National Oil and Haz-
ardous Substance Pollution Contingency Plan (40 CFR Part
300) or at the direction of an on-scene coordinator

If any injury is not excluded by one of the above exemptions,
the responsible party may be liable if the injury also meets certain
threshold criteria embodied primarily in the DOI regulations.
These threshold criteria depend on whether a Type A or Type B
damage assessment is performed. For both Type A and Type B
assessments, a preassessment screen is required to determine that
a CERCLA-covered incident has occurred and that natural re-
sources have been affected. This preassessment screen does not
require a rigorous confirmation of an injury, but rather involves
a rapid review of readily available information to ensure that
there is a reasonable probability that an injury has occurred. For
a Type A assessment, no additional injury threshold criteria must
be met and no field sampling is required to confirm that the injury
has in fact occurred; the natural resource damage assessment
model simply estimates an expected injury based on information
on the release and the area in which the release occurred. The
model defines injury as direct and indirect mortality to fish, shell-
fish and other species, as well as closure of fishing areas and pub-
lic beaches.
Type B assessments, however, build on the preassessment
screen findings by requiring field sampling to confirm that an
injury has occurred and that the injury meets several acceptance
criteria. While DOI states that absolute scientific certainty is not
required, meeting these acceptance criteria is intended to elimi-
nate claims for speculative and intangible injuries that cannot be
measured. If at any point in the Type B assessment an injury can-
not be confirmed, the assessment is to halt and the potentially re-
sponsible party can request to be released from all liability.
Other provisions of CERCLA and the DOI regulations outline
two other concepts that are important to the definition of injury.
First, responsible parties are liable for natural resource injuries
that are left after a response (removal or remedial) action has
been completed. Therefore, if a trustee determines that a response
action sufficiently remedies all natural resource injuries, the re-
sponsible party will not be held liable for any natural resource
damages separate from the response costs. According to the DOI
regulations, however, responsible parties may be liable for natural
resource damages if no response action is planned or taken.
Second, responsible parties are liable for damages for injuries
from the time that the injury first occurs to the time that the
natural resource is returned to its baseline condition. If the trus-
tee does not plan to restore or replace the injured resource, this
recovery time may be the natural recovery period or a shorter-
than-natural recovery time if it proves technically infeasible or
is not cost-effective to quantify damages over an extremely long
natural recovery period. If restoration or replacement actions are
planned, the recovery period is determined by considering the
cost-effectiveness of various alternatives, including a no-action
period wherein recovery proceeds naturally without intervention
and a recovery period that reflects restoration or replacement
activities.
Once an injury has been determined, it is translated into a
dollar amount to determine responsible party liabilities. Under
CERCLA, this dollar amount represents compensatory rather
than punitive damages for public instead of private losses; how-
ever, a variety of other statutes may allow punitive or private
damages to be sought for natural resource injuries. The measure
of damages will depend in part on the methods by which they are
determined and, as discussed above for the determination of
natural resource injury, different methods for determining dam-
ages may be used for the Type A and Type B assessments.
In Type A assessments, the natural resource damage assess-
ment model estimates damage in terms of the decrease in value to
the public of recreational or other public uses of the injured re-
source. For example, the model equates damages to the value of
commercially harvested fish and shellfish not caught or the value
of beach days lost due to the release of oil or hazardous sub-
stances. To determine these values, the model uses a data base of
market prices for those resources that are traded in a market
(e.g., fish and shellfish) and nonmarket prices for such things as
waterfowl, recreational fishing and shorebirds that are not traded
in a market.
In Type B assessments, damages are the lesser of: (1) restora-
tion or replacement costs; or (2) decrease in use values. This con-
cept follows the principle that it is not cost-effective to spend,
for example, $1,000,000 restoring or replacing a resource thatii
valued at only half that much by the public. Similarly, it is not
cost-effective to compensate the public $1,000,000 in decreased
use value if the injured resource can be restored or replaced fe
half that price. Restoration or replacement costs can include the
cost of any actions to restore, replace or rehabilitate an injured
resource as long as the actions are carried out in a cost-effective
manner and are needed to return the resource to no more than itt
baseline condition. If restoration or replacement costs are to form
the basis for the damage determination, trustees also may claim
damages for the diminution of public use values over the time re-
quired to perform the restoration.
Decrease in use values for Type B assessments are the same at
those described above for the Type A assessments and include
the value of public benefits forgone due to the oil or hazardoui
substance release (e.g., the value of fish not caught or recreation
days lost). In determining use values, it is important to recogWB
28 LIABILITY/INSURANCE/DEREGULATION
-------
that natural resources may be used for multiple purposes, and
responsible parties are liable for loss of use values for all uses to
which a resource may be put (although double counting of such
natural resource uses is prohibited). In addition, the DOI regula-
tions state that recoverable use values include fees and other pay-
ments made to the trustee for the private use of the public re-
source. However, trustees cannot collect for taxes forgone (be-
cause these are transfer payments from individuals to the govern-
ment) or income lost by private individuals and subsequent in-
direct effects (because these values do not accrue to the trustee
and can be the subject of private law suits).

Damage Assessment Costs
CERCLA states that responsible parties are liable for the "rea-
sonable" costs of assessing natural resource damages and that a
trustee must use "cost-effective" methods to carry out the dam-
age assessment. The DOI regulations have supplemented these
provisions by adding that responsible parties are liable only for
reasonable and "necessary" costs of performing a damage assess-
ment. Several requirements have been built into the DOI regula-
tions to assure that these principles are followed including:
• Requirements for trustees to document a plan for each step of
the damage assessment and to solicit comments on the plan
from the public, potentially responsible parties and others be-
fore undertaking those steps
• Requirements for trustees to weigh the cost-effectiveness of al-
ternative actions throughout the assessment process
• Requirements to document costs and the reasons why they were
incurred
Given these general requirements, there are no fixed rules for
what damage assessment costs responsible parties may have to
bear; a cost that is reasonable and necessary for one situation
may not be for another situation. However, the DOI regulations
do provide guidance on categories of costs for which responsible
parties can and cannot be held liable.
The DOI regulations state, in general terms, that damage
assessment costs are considered reasonable when: (1) the various
phases of a damage assessment have a well-defined relationship
to one another and are coordinated; (2) the incremental bene-
fits obtained by using a certain procedure in the assessment ex-
ceed the incremental costs of the procedure; and (3) the antici-
pated cost of assessment is less than the anticipated damage
amount. As long as these general criteria are met, responsible par-
ties can be required to pay several categories of assessment costs
including the costs of:
• Release detection and identification
• Trustee identification and notification
• Potentially injured resource identification
• Initial sampling, data collection and evaluation
• Site characterization and preassessment screen activities
• Potentially responsible party notification
• Public participation
• Assessment methodology identification and development
• Application of the natural resource damage assessment model
if a Type B assessment is conducted
• Confirmation of exposure, injury determination and quantifi-
cation and damage determination if a Type B assessment is con-
ducted
• Post-assessment activities, including the costs of documenting
the assessment findings, presenting a demand for payment by
the responsible party and planning actions to be taken with
the damage award
• Administrative expenses on the part of the trustee throughout
the damage assessment
There are, in addition to the above, certain categories of assess-
ment costs that are specifically excluded from responsible party
liability. First, responsible parties are not liable for assessment
costs that are not directly associated with deriving a dollar value
for natural resource injuries. Responsible parties therefore are not
liable for costs incurred by trustees to carry out their "regular
activities" to manage the natural resource. The DOI regulations
also state that responsible parties should not be required to pay
for new developmental research necessary to determine whether
an injury meets the acceptance criteria of the Type B procedures.
In addition, responsible parties may not be liable for any dam-
age assessment costs if at any point during the assessment a trus-
tee determines that no natural resource injury has occurred.

OPPORTUNITIES FOR RESPONSIBLE
PARTIES TO AFFECT LIABILITIES
Once oil or a hazardous substance has been released and a trus-
tee initiates a natural resource damage assessment, the DOI regu-
lations offer the opportunity for responsible parties to take cer-
tain actions which can affect their liabilities. In particular, re-
sponsible parties will be offered the chance to: make inputs to the
selection of a Type B damage assessment over a Type A assess-
ment; and play an active role in the damage assessment itself.
Both of these options available to the potentially responsible
party are described below.

Choosing Type B Damage Assessment
Procedures Over Type A Procedures
If a trustee chooses to conduct a damage assessment assoc-
iated with a release in a coastal or marine environment, the trus-
tee must conduct a Type A assessment unless the limitations of
the natural resource damage assessment model make it inappro-
priate to a given incident. However, even if the model is deemed
appropriate for a particular incident, the proposed Type A pro-
cedures allow the potentially responsible party to request that a
Type B assessment be conducted. The potentially responsible
party will be offered this opportunity to contribute to the choice
of Type B over Type A procedures at two separate points in the
assessment process: (1) when the preassessment screen has been
completed; and (2) when the trustee has determined that a Type
A assessment is appropriate and should be performed. If a poten-
tially responsible party prefers the Type B procedures, the trus-
tee could then conduct a Type B assessment if the potentially re-
sponsible party agrees to advance and ultimately bear respon-
sibility for all reasonable costs of the assessment regardless of the
outcome.
Some of the key considerations that may be taken into account
by the potentially responsible party in contributing to the choice
of Type B over Type A assessments are highlighted in Table 1.
As shown in this table, responsible parties are liable for the same
categories of costs—natural resource damages and damage
•ssessment costs—regardless of whether a Type A or Type B
assessment is conducted. However, the required sampling, analy-
sis, and record of conclusions of a Type B assessment are much
more detailed and thus is expected to be much more costly than
a Type A assessment. One of the most significant differences in
the assessment procedures is that Type A assessments require only
a relatively superficial preassessment screen to confirm natural
resource injuries, whereas Type B assessments require a rigorous
confirmation of an injury that has to meet certain acceptance
criteria. The key issue for the responsible party, therefore, is that
the confirmation of injury step in the Type B assessment will be
more costly, but is also the step which could result in a determina-
tion that no injury has occurred. Whenever such a determina-
tion is made (whether in a Type A or Type B assessment), the
LIABILITY/INSURANCE/DEREGULATION 29
-------
assessment must halt. The potentially responsible party may then
request to be released from all liability. According to DOI's pro-
posed Type A procedures, a trustee may not re-select the Type A
procedures if natural resource exposure cannot be confirmed dur-
ing a Type B assessment.
Another key difference between Type A and Type B assess-
ments lies in their potentially different methodologies for valuing
natural resource injuries. The Type A assessment bases its dam-
age estimate on diminution of use values, while the Type B assess-
ment bases its damage estimate on the lesser of diminution of
use values or restoration/replacement costs. Therefore, if a
potentially responsible party has sufficient reason to believe that
valuation based on restoration or replacement costs will yield a
significantly smaller damage estimate, selecting the Type B pro-
cedures could lead to a smaller liability. Regardless of the valua-
tion methods used, trustees are required to use recovered dam-
ages to restore or replace the injured resources.
Table 1
Key Characteristics for Choosing Type B Assessments over
Type A Assessments
Cherecterlstle
Proposed Type A Procedures
Flnel Type B Procedures
Scop, of Neturel N.tursl resourcei
Resource Coverage end urine envlr<
fined by CERCU
Generel Type of
Assessment
Appllcetion of Model with
•InlMSl field observation
Detailed csse-by-
eveluetion

tlon Methodology

of DaMge Avard

Role of Public

tielly Responsible
Petty

co.»p 1 e t e

Notified of key Milestones;
given opportunity to review

given opportunity to review
end coesssunt ; given opportun-
ity to conduct eai>..»ft.m«nc
under approve! of trustee

restore tlon/ rep laceaent
velues
coaplete If trustee
selects Type B; If re-
Type B. payment of essess-
SeM AS for Type A assess-
ments

••nts

The timing for payment of a natural resource damage claim
may differ depending on whether a Type A or Type B assessment
is performed. For either a Type A assessment or a Type B assess-
ment selected by a trustee, a responsible party cannot be required
to pay a natural resource damage claim until the damage assess-
ment is complete. However, if the potentially responsible party
requests and the trustee, in turn, selects a Type B assessment, the
potentially responsible party must pay the reasonable cost of
assessment in advance. The natural resource damages, if any are
determined, are in either case to be paid after the damage assess-
ment is complete.
Finally, the roles of the public and the potentially responsible
party are identical in both Type A and Type B assessments. Both
groups are to be notified of key findings and assessment mile-
stones and are to be given the opportunity to review and com-
ment on assessment plans and associated documentation. As dis-
cussed below, a responsible party may also be given the chance to
perform the damage assessment (either Type A or Type B), sub-
ject to review by the trustee. Because a Type B assessment in-
volves more steps and a generally more rigorous evaluation than
a Type A assessment, the potentially responsible party stands to
reduce assessment costs more by taking a more active role in a
Type B assessment. Therefore, in deciding between the Type A or
B procedures, the responsible party should keep in mind that the
difference in cost between the two assessments might be made
smaller by active participation and cooperation on the part of the
responsible party.

Playing an Active Role in the Damage Assessment
The DOI regulations give potentially responsible parties several
opportunities to take part in natural resource damage assess-
ments. At a minimum, trustees are required to give potentially
responsible parties the chance to review and comment on key
planning documents, such as plans which identify the methodolo-
gies and specific steps to be followed during the assessment and
the natural resource restoration activities. In addition, trustees are
required to invite the participation of potentially responsible par-
ties in the development of the type and scope of the assessment
and in the performance of the assessment. At the option of the
trustee, and if the potentially responsible party agrees, the po-
tentially responsible party can implement all or any part of the
damage assessment or natural resource restoration. However, re-
gardless of the level of involvement of the potentially responsible
party, the DOI regulations give the trustee final authority for all
decisions in the assessment and restoration activities.
A potentially responsible party can greatly affect his or her
liabilities by using these opportunities to play an active role in
the damage assessment or restoration actions. By playing an ac-
tive role in the development of all planning documents and by
thoroughly reviewing those documents once prepared, potentially
responsible parties can help ensure that costs are reasonable be-
fore they are incurred. In developing and commenting on these
documents, however, responsible parties should be careful to
participate in a manner that will not cause the trustee to spend
unnecessary funds that ultimately are to be paid by the respon-
sible party. For example, voluminous comments that are not ger-
mane to an assessment may lead to unnecessary assessment costs
for the trustee and, hence, unnecessary liabilities for the reason-
able party. Similarly, responsible parties should vie to conduct all
or any part of the assessment or restoration if they believe they
can do so at a significantly lower cost than the trustee. However,
in deciding whether to agree to conduct an assessment or restora-
tion, responsible parties should keep in mind that their actions
will be conducted under the direction, guidance, monitoring and
approval of the trustee.

CONCLUSIONS
In accordance with SARA, trustees desiring to recover natural
resource damaged and damage assessment costs will, in the
future, be able to seek recovery only from responsible partiei
rather than the Fund. Additional responsible party liabilities for
natural resource damages separate from the cost of an emer-
gency or remedial response can be reduced by incorporating
natural resource concerns in the planning of response actions.
However, even well-planned response actions are likely to leave
residual natural resource injuries that can be translated into a
dollar amount and recovered by trustees.
As discussed in this paper, the individual items for which
costs can be recovered under CERCLA are numerous and can
total up to a significant liability (up to $50,000,000 per release
as specified by CERCLA) for responsible parties. Therefore, it
is prudent for responsible parties to take advantage of the re-
cently developed natural resource damage assessment procedure!
which offer several opportunities for responsible parties to take
part in the assessment process. By assisting in the selection of
damage assessment procedures, reviewing and helping develop
plans for key steps in the assessment process and conducting parts
or all of the assessment and natural resource restoration, respon-
30 LIABILITY/INSURANCE/DEREGULATION
-------
sible parties can help reduce their liabilities for natural resource 2. CERCLA Natural Resource Claims Procedures. Final Rule, 40 CFR
damages. Part 306, 50 FR 51205 etseq. Dec. 13, 1985.
3. Comprehensive Environmental Response, Compensation, and Liabil-

ACKNOWLEDGEMENT ity Act of 1980'42 U'S-C Section 9601 etseq'
Th« o,,tx^e „*• *v,- i i j • , . • 4. National Oil and Hazardous Substances Pollution Contingency Plan,
The authors of th s paper acknowledge, with appreciation, the ^ CFR Part m 5QFR 4m2 et Nov 20 1985
review and helpful comments provided by Cecil S. Hoffman, , „ ^ A „-,„,„, ™D ».
U.S. Department of the Interior 5- Natural Resource Damage Assessments. Final Rule, 43 CFR Part
11, 51 FR 27674 et seq. Aug. 1, 1986.
6. Natural Resource Damage Assessments. Proposed Rule, 43 CFR
REFERENCES Part 11, 51 FR 16636 et seq. May 5,1986.
1. CERCLA Arbitration Procedures. Final Rule, 40 CFR Part 305, 50 7. Superfund Amendments and Reauthorization Act of 1986, Public
FR 51196 etseq. Dec. 13, 1985. Law 99-499, Oct. 17, 1986.
LIABILITY/INSURANCE/DEREGULATION 31
-------
Generalized Risk-Based Decision-Making
For Hazardous Waste Sites

Charles L. Vita, Ph.D., P.E.
Colder Associates, Inc.
Redmond, Washington
ABSTRACT
A generalized risk-based decision-making (RBDM)
methodology is outlined and discussed as a useful tool for prac-
tical application to hazardous waste facility management, in-
cluding siting, design, construction, operation, closure or
remedial action. Using a generalized definition of risk, RBDM
may or may not include health-based risk assessment. The
premise is that RBDM can efficiently, effectively and explicably
meet the multiple (often conflicting) needs of practice; RBDM
can lead to better realized outcomes. RBDM goals are presented
and an RBDM process is outlined in terms of: (1) criteria and
alternatives development; (2) risk measurement; and (3) alterna-
tive evaluation and selection. Emphasis is on processes, principles
and concepts rather than mathematical or technical details.

INTRODUCTION
This paper aims to introduce a risk-based decision-making
(RBDM) framework to complement, as a conceptual tool, ex-
isting siting, design or remedial action methodologies used for
hazardous waste technology selection, site selection or manage-
ment plan/policy selection.
RBDM can benefit practice through better understanding, bet-
ter communication, better decisions and better outcomes. It can
help elucidate and reduce biases, inconsistencies, incoherencies,
mixing of facts and values and mixing of relevant and irrelevant
factors. It can increase use of available data and help direct effort
to where it is most valuable. It can reduce prediction uncertainties
and increase decision defensibility. It can help elucidate true
needs and values (criteria); increase the base of alternatives, struc-
ture analysis and evaluation; improve decisions; and help achieve
desired ends. This paper discusses some general considerations for
achieving these benefits.
The paper is structured as follows. First, risk and RBDM con-
cepts are introduced. Then, some RBDM applications are pro-
posed. Next, goals comprising an ideal for RBDM are presented.
Following these sections, the RBDM process is outlined and dis-
cussed in terms of: (1) criteria and alternatives development, (2)
risk measurement and (3) alternative evaluation and selection.
Uncertainty and decision optimization, including some useful
ideas of multicriteria decision-making, are included. A summary
and concluding discussion ends the paper.

RISK CONCEPTS
Risk, though having many specialized meanings, is a com-
pound measure of the probability and severity of loss. Losses may
be economic/financial, environmental/health, ethical, personal/
psychological, organizational or societal. Losses may be pure, in-
volving no possibility of gain (e.g., aquifer contamination),
resulting from adverse effects from some activity or event. Losses
also may be speculative, involving less than an expected or
satisfactory gain from a set of actions (e.g., loss of expected
revenues due to a sub-optimal design) or missing opportunities
for profit. Risk implies the opportunity and need for decisions.
Management of hazardous wastes requires measurement of
risks and costs associated with risk management. Risk measure-
ment deals with discovery and quantification (numerical or
descriptive) of loss. (This may or may not include health-related
"risk assessment," as used by the U.S. EPA.4'8 Risk management
includes the whole process of loss control (risk measurement,
avoidance and reduction by design and operation) and loss finan-
cing, e.g., insurance. RBDM can be a risk management tool.
Risk can be measured (quantitatively or qualitatively) in terms
of (generalized) performance and (generalized) reliability. Per-
formance is measured relative to criteria (attributes, objectives,
goals) which reflect the multidimensional components of risk
(e.g., health, technical, financial, legal/regulatory, etc.). Reliabil-
ity is the probability the alternative or system performance will be
satisfactory in terms of established goals (i.e., levels of perform-
ance, targets, standards or limit [failure] states). Reliability is the
complement of the probability of unsatisfactory performance (or
"failure"). Risk, performance, reliability or failure can each be
multidimensional.
RBDM is formulated here as a special case of multi-criteria
decision-making, using a systems engineering problem-solving ap-
proach6 summarized in three steps (Fig. 1):

• Problem formulation, criteria and alternatives development
• Analysis and risk measurement
• Interpretation and alternative evaluation and selection

Within operational constraints of available budget, time, infor-
mation and talent, this process is nested, iterated and refined as
needed to produce a satisfactory, best or "optimum" alternative
for implementation. At any level of effort or stage (feasibility
study, preliminary or final design), the best choice alternative for
implementation is the one closest to holistically meeting the cri-
teria in terms, explicitly or implicitly, of decision-maker needs,
values, constraints and perceptions.

GENERIC APPLICATIONS
RBDM provides a general evaluation framework for inter-
preting and selecting hazardous waste management control alter-
natives at any point in a facility life-cycle: planning, siting, Of
ploration, design, operation, closure or decommissioning, post-
closure (including monitoring), remedial action or upgrading.
Where appropriate, RBDM can include health-based risk assess-
ment. RBDM could also be used for waste generation, transporta-
32 LIABILITY/INSURANCE/DEREGULATION
-------
RBDH PROCESS
RBDH INPUT
(As Needed and Available)
PROBLEM FORMULATION:
Criteria end Alternatives
Development
ANALYSIS: ftlsk
Measurement
INTERPRETATION:
Alternative Evaluation
and Selection
» Project and Decision Making
Situation, Context. History
I Decision Haker(s) Needs.
Values. Perceptions
(Regulatory/Legal,
.eerfng/Conitructlon/
• tfons. Environment!!/
th. Econoffllc/Coit.
unl)
te-speclflc
glonil or Global
• Risk Measurement
Techniques
- Audit and Ranking
- Eapert Opinion
- Scenario Analysis
- FHEA
- FTCTA
- Risk Assessment

• Engineering and Scientific
Principles, Models and
Experience

• Systems Principles and
Models (Operations Research/
Management Science)

• Social Science
Principles. Models
and Experience

• Expert Judgement
IMPLEMENT HIGHEST VALUED
ALTERNATIVE
Input to Next Systems Level
9r Decision Situation E-
Figure 1
Generalized Risk-Based Decision-Making (RBDM) Process
tion or control decisions, including waste minimization, destruc-
tion, containment or disposal.
RBDM can be used to support any objective, goal or process
which uses performance or reliability as external or internal (exo-
genous or indogenous) elements or criteria. This process may be
complex or relatively simple, depending on project-specific needs
and constraints (time, budget, information and talents of those
involved).

Costs
Performance and reliability can include estimated implementa-
tion cost or life-cycle costs. For example, a life-cycle cost objec-
tive function could be calculated (using financial analysis and
engineering economics) for each alternative as the sum of ex-
penses, minus revenues and recoveries, plus any salvage value.
Expenses would include capital costs (e.g., planning, siting,
design and construction) and operating costs, including risk man-
agement—loss control design, devices and activities and loss
financing instruments (e.g., insurance). In the event of system
failure (inadequate performance), costs also would include losses
from lost revenue or other opportunity costs, uninsured legal and
regulatory costs and any uninsured consequences of insured
losses.
Beyond economics, inadequate performance which leads to
adverse environmental or health effects or deaths poses ethical
costs, which can include criminal punishments. With this in mind,
the alternative which meets a level of acceptable performance
(after formulation-analysis-interpretation of alternatives and
trade-offs) at minimum cost would be the most overall cost-
effective alternative.

Performance
RBDM can be used to rationally evaluate or check the site-
specific performance adequacy of proposed alternatives (new
facilities, upgrades or remedial actions) in terms of reliability, R.
R may be composed of a single or multiple criteria. Also, time, T,
to a prespecified lower limit of reliability, R(T) or probability of
failure, F(T), can be estimated. Performance is predicted with the
appropriate site-specific, design-specific data. Sources of uncer-
tainty and corresponding needs for refining the analysis are iden-
tified. Results are compared (evaluated) against the appropriate
goal for acceptable performance.

• If R is below the goal, the proposed design is inadequate; it
constitute an underdesign. Either the design must be improved
or the analysis refined (with more data or better models), or
both.
• If at or above the goal, the design is adequate and, at least in
terms of R, acceptable (although as a possible overdesign, its
cost effectiveness may not be maximum and so could be im-
proved).

Results could be used, depending on needs, to support: siting,
design, monitoring, operation and maintenance, remedial or cor-
rective action, upgrading to increase T, closure considerations or
post-closure considerations.

Alternative Selection
In landfill or pond design, RBDM could be used to choose be-
tween alternatives: liner configurations, materials, treatments, in-
stallation details, levels of quality control, site improvement
measures (e.g., soil stabilization, reinforcement, grading and soil
replacement), monitoring schemes and so forth. Performance cri-
teria could be based on leakage (e.g., through primary or secon-
dary liners), reliability against undetected contamination, reliabil-
ity against given levels of contamination or biological receptor
toxicity (health) effects. Performance would be estimated using
each alternative design-dependent variable as input. Results
would be interpreted for selection of a best or preferred alterna-
tive or for refinement of alternatives and further analysis.
RBDM could be used for multi-criteria selection of alternative
remedial actions for existing uncontrolled hazardous waste sites.
Performance could be assessed for each alternative using
technical, environmental/health, economic, regulatory and social
criteria. Design, physical or social conditions could include: land
use; contamination conditions; site location, access, adjacent
population and land use; surface and subsurface (geotechnical,
geological, hydrogeological and groundwater) conditions; mete-
orological and climate factors; biota factors; local values (culture
and history); and so forth.
RBDM also could be used for multi-criteria site selection (or
evaluation of existing sites). Performance would be assessed for
each candidate site using relevant technical, economic, regulatory
and social criteria. In a learning process of iteration and refine-
ment, the highest valued alternative would be selected for imple-
mentation.

RISK-BASED DECISION-MAKING GOALS
The value and success of RBDM and subsequent implementa-
tion will depend on how the process is organized, applied, refined
and improved with experience and innovation. The following in-
terrelated goals, composing an ideal for RBDM, can provide
guidance to the process.

Goal 1: RBCM should be explicit in terms of values, purposes,
criteria and methods.
RBDM should define technical and policy needs and con-
straints. It should separate, for inspection and appreciation,
scientific (or factual) aspects from policy or personal values. It
LIABILITY/INSURANCE/DEREGULATION 33
-------
should enhance effective and innovative solutions to hazardous
waste problems through explicit consideration of: problem dis-
covery/identification and definition; goals and objectives; hypo-
theses, assumptions, criteria and procedures; products (output)
identification, including measures of attainment; information
flows, including data inputs, internal review and feedback, exter-
nal review and public communication; interpretation techniques,
including identification/prediction of implementation effects and
implications; conflict resolution approaches; and implementation
procedures.

Goal 2: RBDM should increase understanding and insight, and be
communicable, defensible and convincing.
RBDM should be effective and efficient but not excessively
complex or convoluted. It should increase flexibility for consen-
sus in multi-objective situations with conflicting viewpoints. It
should help clarify thinking and agreement as to what constitutes
adequate data, analysis, site selection/design or control measures.
It should enhance credibility, confidence and trust in the process
and implemented decisions.
RBDM should identify and address safety and risk perception
issues. This process may require consideration of pertinent
political, sociological and psychological factors such as: personal,
organizational or community values, goals, history and culture;
cognitive styles and biases; uncertainty of risk events; dread of
risk events; issues of equity in risk exposures; personal control
(voluntariness) over risks; catastrophic potential of risks; and
so on.

Goals 3: RBDM should help prioritize efforts.
RBDM should help quantitatively answer such questions as:
"What are the risk/cost trade-offs for this alternative? Is this an
effective alternative for achieving these (given) objectives? Is it ef-
ficient? Implementable?" For example, "Is this an optimum (ef-
fective, efficient and so on) exploration program, monitoring
program or remedial action program?"
Since uncertainties are pervasive and unavoidable, RBDM
should delineate uncertainties and how more information could
help. It should include sensitivity analyses and error analyses. It
should make full use of existing data (field, lab, analytical and
policy), clearly define data gaps and help prioritize future data
collection and development.

Goal 4: RBDM should function in practice.
RBDM should be able to operate under (and help identify and
resolve) adverse conditions or constraints that can include: con-
flicting policies, goals or attitudes; adverse public perceptions, ex-
pectations, attitudes or reactions; limitations in information,
know-how, budget, schedule (time) or talent. It should be prag-
matic, not dogmatic.

Goal 5: RBDM should be adaptable to changing conditions and
be evolutionary, improvable by learning and innovation.

RISK-BASED DECISION-MAKING FRAMEWORK
The following three-step formulation-analysis-evaluation pro-
cess, illustrated in Fig. 1, summarizes and discusses an RBDM
framework. Highly interactive, interdependent and evolutionary,
these elements can be nested and iterated or refined throughout
an RBDM process or given project from the initial investigation
and feasibility study through detailed, final design (plan or policy)
implementation. Each step can require its own formulation-
analysis-evaluation stages. The following summary assumes in-
formation/data will be obtained as appropriate, depending on
needs and availability.

34 LIABILITY/INSURANCE/DEREGULATION
Problem Formulation: Criteria and
Alternatives Development—Step 1
In a process of search and discovery, RBDM conditions, issues
and problems are identified and structured in terms of criteria and
alternatives. Relevant conditions, requirements and perceptions
must be determined, including: present situation, history and
potential future scenarios; option-alternatives; constraints and
alterables.

Criteria Development
Performance or selection criteria, i, measured on scale q (i =
1,2,111,1), must be determined. Criteria are, in general, a mix of at-
tributes, objectives and goals. Criteria can include many factors;
regulatory/institutional/legal, health/environmental, technical,
financial/cost, social, ethical or others. These criteria may be
multidimensional; for example, technical criteria can include ef-
fectiveness measures or indices for reliability, implementability/
constructability and scheduling, maintainability, availability,
usability, useful life, resilience, safety or cost-effectiveness.
The search for valid performance criteria will result in project/
context-specific criteria which may be a mix of qualitative and
quantitative, objective and subjective, well-defined and crisp or
ambiguous and fuzzy. Not all valid criteria are precisely measur-
able. Relative importance (value or utility) for each criterion is
estimated and may be refined throughout the RBDM process.

Alternatives Development
Identify and describe trial or candidate alternatives, j 0 =
1,2,...,J). Depending on project needs, values and so on, alterna-
tives can be exploration/research/data-based, analysis-based,
technology-based, management-based, siting-based, policy-based
or combinations thereof. During initial screening, many can-
didate alternatives may be rapidly formulated, analyzed and inter-
preted for suitability, with unsuitable alternatives dropped from
further consideration.
Structured problem-solving techniques can complement usual
engineering approaches to problem formulation. These include:
syenetics, nominal group technique, charette, backstep analysis,
cause-effect diagrams, problem clarification charts, force field
analysis, interaction matrices, decision trees, brainstorming or
brainwriting, E. De Bono's lateral thinking techniques, question-
naires, surveys, Delphi, systems definition matrices, scenario
writing, variations of these and more.3'6-9

Analysis: Risk Measurement—Step 2
Risk measurement deals with identification, description and
prediction or forecasting of loss or failure modes for each alter-
native. Analysis or measurement, ideally, is objective and scien-
tific. Done at appropriate detail levels during feasibility studies,
preliminary design or final design stages, risk measurement can
include screening-level analysis or fine-tuning (refinement or op-
timization) of each alternative (e.g., using engineering analysi*/
design and operations research techniques).
Risk measurement starts with the operational identification and
description (using available experience or systematic search and
discovery) of hazard or risk (or failure) mechanisms which may
lead to loss. These are then analyzed for performance, using
expert-opinion methods or formalized modeling methods.

Methodologies
Risk measurement methodologies vary in sophistication, scope
and use; they can draw from a variety of tools and procedure!,
depending on project needs and constraints. Procedures or
methodologies can include:
• Audit and Ranking—commonly, procedures are based on
-------
identification of hazard conditions which are then scored and
ranked using set procedures; an example is the Hazard Ranking
System used by the U.S. EPA to choose sites for inclusion on
the National Priorities List under Superfund.2
• Expert Opinion—uses individual or group subjective interpre-
tation of site or system risk conditions; systematic examples in-
clude Delphi or Delphi-like procedures.
• Scenario Analysis—starts from given events and conditions
then projects forward to possible consequences; these are
studied and interpreted for implications.
• Failure Modes and Effects Analysis (FMEA)—identifies, de-
scribes, examines and interprets component-by-component
failure causes, effects, probabilities, criticalities and corrective
or preventative actions.
• Probabilistic Reliability Analyses—uses various (network and
state space) classical and Bayesian techniques to estimate sys-
tem failure and non-failure probabilities.
• Fault Tree-Event Tree Analysis (FTETA)—used to simulate
and interpret system performance: with fault trees, modes of
failure are postulated, traced back to possible causes and as-
sessed qualitatively or probabilistically; with event trees (simi-
lar to decision trees), initiating events are projected forward,
combined with subsequent events and assessed probabilistically
for consequences. FTETA can be extended to include success
trees (duals of fault trees), process trees (system performance
measures are qualitatively modeled using basic performance de-
terminants) and influence trees (a quantitative extension of
process trees).
• Risk Assessment—specifically associated with toxicity (health)
effects (in hazardous waste applications), it has four basic
components:4'8 (1) hazard identification of toxic materials,
(2) dose-response assessment, (3) exposure assessment and (4)
risk characterization. Risk assessment requires characteriza-
tion and analysis of toxic material distributions, interactions
and pathways in environmental media (soil, water, air or biota)
and their impact on biological receptors.2'8
Predicting Performance
Analysis is used to predict alternative performance from the
identified risk mechanisms. Performance is measured, judged or
indexed in terms of the criteria, PJ(CJ), j = 1,2,...,J and i =
1,2,...,!; or simply: matrix P(c). For alternative j and criterion i,
Pj (q) represents the relative likelihood of achieving possible
levels of criterion performance q. P(c) can include health,
technical or economic effects (including, as appropriate, potential
effects on people, other species or aspects of the environment or
surrounding infrastructure, facilities or operations). P(c) is esti-
mated (predicted or forecasted) by:
plicit goals (targets, standards or limit states). For any goal-based
criterion having goal;, the reliability for any alternative, Rj(goalj),
P(c) = f(D)
(1)
where f(D) represents an appropriate set of relevant (e.g.,
engineering, scientific, economic or social science) models and ex-
pert judgment using input D. D-input includes project-specific
and relevant human/social factors or determinants, including
space and time effects. Project-specific determinants may include
conditions related to: the site and its effective environment; facili-
ty design or remedial action; waste characteristics and distribu-
tion; installation or construction; operation, including post-
closure; geographic factors; and so on. Human/social deter-
minants may include: financial/economic conditions, legal/
regulatory climate, ethical concerns, cultural/historical factors,
sociological factors and psychological factors.
An important measure of performance and risk, reliability is
the probability performance will be satisfactory in terms of ex-
Rj(goali) = Probability [(q igoalj)]
(2)
P(c) is (implicitly or explicitly) a joint multi-criteria fuzzy (i.e.,
imprecise) probability density (mass or mixed) function condi-
tional on f(D) and D. That P(c) is a fuzzy probability function
unavoidably results from lack of complete knowledge. For prac-
tical and theoretical reasons, the available D-information is
uncertain, incomplete and imprecise; and the response models,
f(D), and their interpretations suffer, to greater or lesser extents,
from simplifications and approximations, errors and omissions.
In practice, uncertainty, randomness, imprecision and com-
plexity are pervasive and unavoidable. Each influences RBDM.
Probabilistic methods can be useful for analysis of randomness
and uncertainty. These can be complemented with fuzzy set con-
cepts and techniques for analysis of imprecision and complex-
ity.i.'2
Uncertainty, randomness and imprecision in P(c) and its inter-
pretations results in risk. If, through omniscience, P(c) and all
consequences were certain, fixed and precise, rational decision-
makers could avoid risk. In practice, more information or knowl-
edge, more care or wisdom, or safer siting, design, construction
or operation, through improvement in D, f(D) or interpretations
of f(D) and P(c), can, at a cost, better define or decrease (but not
eliminate) risk. Thus, risk/cost trade-offs become a necessary
part of RBDM.
Interpretation: Alternative Evaluation
and Selection—Step 3
Interpretation deals with alternative-set evaluation and
(ultimately) an alternative selection for implementation and
follow-through. This is based on the relative desirability (impor-
tance, value or utility) of the predicted performance. There is no
unique way to evaluate alternatives. Evaluation may be struc-
tured, simply or complexly, but it is always subjective, judgmen-
tal and decision-maker goal/value/perception-dependent.

Utility Measurement
For dominant criterion decisions, evaluation and selection of
the alternative of choice can be based on maximizing expected
(average) utility, maximizing an up-side performance, maximizing
reliability, minimizing down-side performance or some combina-
tion of these measures. In these cases, performance desirability is
measured using a utility scale, Uj(q), estimated33 over the range of
potential variation in performance. The most desirable level of
performance is given a utility of one; utility for the "satisficing"
(minimally acceptable) level of performance is set equal to zero.
Performance is then rated (ranked or prioritized) by utility as the
product (for any i) of Pj(q) and Uj(q). Utility for any range of
performance can be computed. For example, a "down-side" utili-
ty measure of alternative performance, Uj(q)
-------
mance, Ej[ Uj(q)], is:
OO
Ej[Uj(q)] = J PJ(CJ) * Uj(ci) * dq
00
(4)
for repeated (i.e., recurring, identical) decision conditions, max-
imizing expected utility results in maximum long-term gain.
Reliability is a special case of utility measurement; implicitly,
utility is a step function passing through goalj or target value of
performance. Utility equals zero below the goal level of perfor-
mance, and one at the step and beyond.
Single-criterion utility measures can be conditional on meeting
other criteria requirements. For example, reliability can be maxi-
mized for a given budget or schedule constraint. To achieve cost
effectiveness, expected cost (negative utility) can be minimized for
a given level of performance or reliability.
Yet, these simple utility measures, though useful, are by them-
selves, inadequate for incommensurable multiple-criteria situa-
tions where conflicting criteria (which cannot be simultaneously
maximized or minimized) are to be judged together as a best-
available combined set. Multiple-criteria evaluation can be dif-
ficult because of trade-offs, uncertainty or multiple actors. The
generalization of utility-based evaluation to multi-attribute utility
theory3" is not, as Zeleny12 and other argue, a general solution to
multi-criteria situations. All-purpose prescriptions are not
available, as discussed next.

Multi-Criteria Scoring
Many possibly useful mathematical multi-criteria evaluation
techniques explicitly or implicitly use criteria weighting and aggre-
gating procedures in various decision-theoretic structures to score
each alternative (e.g., Zeleny12 and Voogd10). These may be repre-
sented symbolically by:
= 1,2,...,1
(5)
where Sj is the evaluation score for alternative j based on
calculating for each criterion, q, a performance dominance score
of Pj(q) using function gj[»], weighting the score by a weighting
CHIH8U (c
Weights
ReguUtory/Legal
| (1 - I) »|
j (1 • 2) -2
eilth/
tvlronnentil
Cj (I - 3)

echnlctl
Engineering/
(instruction/
Operation)

? I: : ii
Ininclil/Cost
c, (1 • 9)
SoCKl/Cthlcil
. c, (1 . 1)
.3
»4
£»AIUATIO« SCORE (Sj)
AUCRHATIVtS (j
J - 1 J • 2
93IP|(<:3>J
-
98[P|(CB)1
1.2.3,4)
J • 3
9S[P3(C5)
J • <
Highest Vjluw! tlternitlve • uilSj-, j - 1.2.3.4]

Note: Highest Vilued Alternitlve VI11 Ch«nge With f(l), «, Or g.f«J
Check for Sensitivity tnd 1*>llcitlons

Figure 2
Example Evaluation (Decision) Matrix (Generalized)
function, wj, then aggregating for all criteria (i = 1,2,...,!) using
function f i • }. Weighting functions may or may not be perfor-
mance level-dependent or correlated between criteria.

Evaluation Matrix
The maximum Sj (j = i,2,...,J) is the highest valued alternative. The
common decision matrix (Fig. 2) approach of performance rating,
weighting and summing is a simple example. Performance for each
criterion, Pj(q), is rated (e.g., on a scale of 1 to 10, or rank-
ordered), weighted by an estimate of criteria importance (e.g.,
weights of 1, 2 or 3), then summed to give a score, Sj. Maximum
Sj is assumed to indicate the highest valued alternative. Clearly,
however, the choice of g;[«], w, and f { • J can bias the evaluation
and reduce its effectiveness: fitting models to problems (and not
vice versa) can be as important in (subjective) evaluation as it is fa
(objective) analysis.
A general multi-criteria evaluation approach starts with for-
mulation of an evaluation or decision matrix (Fig. 2). Alternatives
(columns) and criterion performance estimates or effectiveness
score measures (rows) are filled-in from the results of problem
formulation and analysis. Matrix representations provide a unify-
ing format for Sj-kinds of evaluations, as illustrated in Fig. 2 (for
J = 4). Published, practical applications of the evaluation matrix
approach to selection of remedial actions at hazardous waste sites
have included St. Clair, McCloskey and Sherman,7 Walker and
Hagger" and Partridge.5 These studies have used relatively simple
rating, weighting and summing to calculate Sj values.
However, it is important to explicitly emphasize that alter-
natives can be valued and judged by the criteria set considered
holistically, integrating intuitive (visual-graphical) interpretations
and mathematical analyses. It also provides an effective format
for group communication.
Depending on project-specific constraints and needs, criteria
scores can be graphically or mathematically operated on to iden-
tify patterns, investigate implications and trade-offs between
criteria and alternatives and refine the decision-making process
(criteria alternatives and methodology). Using multiple evaluation
methods (Sj models) provides for an overall evaluation sensitivity
assessment. This process can help identify and reduce unwanted
biases in the criteria, weightings, evaluation models or other
limitations. It can improve understanding and insight. This pro-
cess is particularly useful in complex evaluations having unquan-
tifiable criteria.
This general approach can support an evolution of preferences
through a learning process. This may be required for successful
decision-making in complex situations (including groups) where
criteria, ideals, alternatives, conditions or perceptions evolve with
time and information. Ultimately, it can support better decisions
(in terms of achieving the criteria).

Alternative Selection
From initial alternative screening through final selection, the
RBDM formulation-analysis-evaluation process is iterated until a
satisfactory or "optimum" result is selected for implementation
and follow-through. In this process, alternatives are eliminated as
unsuitable or dominantly inferior or improved or invented to ef-
fect better sitings, designs, actions, plans or policies.

CONCLUSIONS
RBDM has been formulated using a systems engineering prob-
lem-solving approach, summarized in three steps: (1) problem
formulation (criteria and alternatives development), (2) analysis
(risk measurement) and (3) interpretation (alternative evaluation
and selection). Within constraints of budget, time and talents this
process is nested, iterated and refined as needed until a satisfac-
36
LIABILITY/INSURANCE/DEREGULATION
-------
tory, best or "optimum" alternative for implementation is ob-
tained. The best choice alternative for implementation is the one
closest to holistically meeting decision-maker needs, values, con-
straints and perceptions.
Caveats for RBDM are as for any serious decision-making pro-
cess. First is the paramount importance of discovering the right
problems, asking and answering the right questions and setting
the right goals and objectives. The second important considera-
tion is appreciating the limitations of available knowledge and
understanding the conditional nature of all forecasts, predictions,
risk estimates or evaluations. This appreciation includes under-
standing and distinguishing between facts, assumptions and
values, explicit or implicit. Third, data, models, interpretations
and levels of effort should be appropriate for the project scope,
needs, management and individual styles, constraints, and so on;
they should be technically balanced and compatible.
RBDM effectiveness is process-dependent; its value is decision-
maker needs/values/perceptions-dependent. There can be no
single correct RBDM approach; subjectivity, uncertainty and im-
precision in predicting and evaluating present and future condi-
tions lead, unavoidably, to different viewpoints and different
formulations, analyses, evaluations, resulting decisions, imple-
mentations and, ultimately, outcomes.
RBDM is a flexible, multi-dimensional tool, not a "cook
book." RBDM is science and art. Effective use requires skill,
understanding and judgment. Used intelligently, it can help effect
user goals and objectives.
REFERENCES
1. Brown, C.B., Chameau, J., Palmer, R., Yao, J. Eds., Proc. of
NSF Workshop on Civil Engineering Applications of Fuzzy Sets,
Purdue University, Lafayette, IN, 1985.
2. Chu, M.S.Y., Rodricks, J.V., St. Hilaire, C. and Bras, R.L., "Risk
assessment and ranking methodologies for hazardous chemical de-
fense waste: a state-of-the-art review and evaluation," Sandia Nat.
Lab. report SAND-0530, Albuquerque, NM, 1986.
3. Erickson, S.M., Management Tools for Everyone, Petrocelli, New
York, NY, 1981.
3a. Keeney, R.L. and Raiffa, H., Decisions with Multiple Objectives:
Preferences and Tradeoffs, Wiley, New York, NY, 1976.
4. National Research Council (NRC), Risk Assessment in the Federal
Government: Managing the Process, National Academy Press,
Washington, DC, 1983.
5. Partridge, L.J., "The application of quantitative risk assessment to
assist in selecting cost-effective remedial alternatives," Proc. Man-
agement of Uncontrolled Hazardous Waste Sites, Washington, DC,
1984, 290-299.
6. Sage, A.P., "Methodological considerations in the design of large
scale systems engineering processes," In Studies in Management Sci-
ence and Systems, Vol. 7: Large Scale Systems, North-Holland,
New York, NY, 1982.
7. St. Clair, A.E., McCloskey, M.H. and Sherman, J.S., "Develop-
ment of a framework for evaluating cost-effectiveness of remedial
actions at uncontrolled hazardous waste sites," Proc. Management
of Uncontrolled Hazardous Waste Sites, Washington, DC, 1982,
372-376.
8. U.S. EPA, "Guidelines for carcinogen risk management," Federal
Register, 51, No. 185, Sept. 24, 1986.
9. VanGundy, A.B., Techniques of Structured Problem Solving, Van
Nostrand, New York, NY, 1981.
10. Voogd, H., Multicriteria Evaluation for Urban and Regional Plan-
ning, Page, London, 1983.
11. Walker, K.D. and Hagger, C., "Practical use of risk management
in selection of a remedial alternative," Proc. Management of Un-
controlled Hazardous Waste Sites, Washington, DC, 1984, 321-325.
12. Zeleny, M., Multiple Criteria Decision Making, McGraw-Hill, New
York, NY, 1982.
LIABILITY/INSURANCE/DEREGULATION
37
-------
Development of an In Situ Biodegradation Process
For Remediation of a Gasoline Spill

J.V. Lepore
D.S. Kosson, Ph.D.
R.C. Ahlert, Ph.D., P.E.
Department of Chemical and Biochemical Engineering
Rutgers University
Piscataway, New Jersey
ABSTRACT
An accidental release of gasoline occurred in a public well field
alongside a major highway in the coastal plain of southern New
Jersey. Immediate remedial action included containment of re-
leased gasoline and removal of contaminated surficial soils. A
remedial investigation revealed substantial contamination in the
vadose zone, as well as a floating hydrocarbon lens on the
groundwater table. A long-term remediation program consisting
of groundwater recovery, treatment and surface application of
treated groundwater to flush the contaminated area was in-
stituted. This cleanup process successfully reduced contaminant
concentrations from several hundred mg/1 to several hundred
Mg/1- Long-term operation has not reduced petroleum concentra-
tions further. In addition, observed concentrations of marker
species in groundwater tend to increase in response to rainfall. We
believe that roadway runoff may be a major contributing factor
to the persistence of groundwater contamination.
In situ microbial degradation is being developed to mitigate the
continuing problem. An appropriate treatment process must
mineralize contaminants at a rate greater than the rate of
transport of contaminants to the groundwater. Batch acclimation
experiments have been performed to demonstrate the viability of
aerobic degradation of gasoline components. Soil column studies
then were implemented to show the applicability of the minerali-
zation process under simulated field conditions.

INTRODUCTION
In July 1984, a tank truck overturned and spilled approximately
7,000 gal of unleaded gasoline; the gasoline flowed across a lawn
and infiltrated an area on the east and north side of a potable
water pumping station. Some gasoline entered the pump house
through a door on the east side of the building, while more traveled
down a slope in a northerly direction and entered a creek which
flows from a pond in a northwesterly direction across the property.
Two small temporary dams trapped approximately 1,000 gal of
gasoline and allowed it to be collected. The fuel which entered the
building was pumped out. Surficial soils in the spill area were ex-
cavated to a depth of approximately 3 ft and were replaced with
clean sand and topsoil.
A subsequent hydrogeologic investigation involved the installa-
tion and sampling of six 4-in. PVC monitor wells within and out-
side the perimeter of the spill area. The purpose of this investiga-
tion was 10 determine hydrogeologic conditions present and the
extent of groundwater contamination. Groundwater from each
monitoring well was sampled and analyzed for aromatic hydro-
carbons. Preliminary data indicated that the contamination of the
groundwater was considerable, ranging from approximately 8 to
17 mg/1.
In March 1985, five soil cores were taken from the spill site to
define the horizontal and vertical extent of remaining contamina-
tion. A set of continuous split-spoon samples, in 12-in. in-
crements from the soil surface to 6 in. below the water table, was
taken in the immediate vicinity of Well 2 (Fig. 1). All soil samples
Figure 1
Site of Gasoline Spill
38 CONTAMINATED AQUIFER CONTROL
-------
were placed immediately in glass jars which were filled with dis-
tilled water, sealed with Teflon-coated lids and refrigerated for
transport to an analytical laboratory. These precautions were
taken to minimize losses of organic compounds through
volatilization.
In addition Well 2 was sampled for liquid with a notched
stainless steel cup attached to the end of a 12-ft pole. This sam-
pling was performed to verify the existence of a floating lens of
hydrocarbon. The notches in the cup allowed only surface liquid
to be collected. The samples obtained by this method indicated
the presence of a continuous floating hydrocarbon phase which
smelled strongly of gasoline, as well as dispersed hydrocarbon
phase in the aqueous phase.

LONG-TERM REMEDIAL ACTION
A series of recovery wells was installed to capture floating and
dissolved product. The placement of the wells is indicated on Fig. 1
as numbers M-2, M-4 and M-7. Cones of depression for operation
of the three wells at a pumping rate of 20 gal/min each are shown
in Fig. 2.
100 150
SCALE IN FEET
PU-31E3
CLEAR I
WELL
PV-29Q
LEGEND

B-2 • MONITORING WEIL
O-DEEP S'SHALLOW

PW-23 O PRODUCTION WELL

A SURFACE SAHPLE

i-2 O SOIL CORES
Figure 2
Cones of Depression for Long-Term Recovery System

The discharge from the recovery wells was manifolded to an
airstripping system and the effluent from the air stripper was
treated by carbon filtration before discharge to the surface of the
contaminated area. The purpose of the remedial plan was to have
the static level of the water in the contaminated zone lowered to
allow flooding of the vadose zone. Flooding was accomplished via
a series of sprinkler heads supplied with treated water from
holding tanks. Such loading of the unsaturated zone facilitated
transport of sorbed hydrocarbon fractions into recovery well
cones of depression.
The recovery process was tested for aromatic hydrocarbons.
Major reductions in contaminant concentrations were attained
very quickly. This period was followed by one of large fluctua-
tions around relatively low average concentrations on the order of
0.1 to 0.2 mg/1. This level of organic species in groundwater is
unacceptable.

CONTAMINANT MITIGATION STUDY
This investigation was comprised of three separate phases. The
first phase was designed to analyze correlations in contaminant
concentration fluctuations and local rainfall patterns. The second
phase was designed to demonstrate the treatability of petroleum
species through batch microbial degradation studies. The third
phase was designed to perform laboratory soil column ex-
periments to determine the ability of the biodegradation process
under simulated field conditions to reduce petroleum concentra-
tions to acceptable levels.

CORRELATION OF CONTAMINANT
CONCENTRATION WITH RAINFALL DATA
Correlations between groundwater contaminant concentrations
and rainfall patterns were evaluated to support or discount the
hypothesis of roadway runoff contributions to long-term aquifer
contamination. Weekly reports of contaminant concentrations in
the groundwater were received and compared to local rainfall
data. The concentration data for wells 2A, 4A and 7 and rainfall
data were plotted against the total weeks of operation for the
petroleum recovery process.
Figs. 3 through 5 show the concentrations of benzene, toluene
and mixed xylenes present in the groundwater at the spill site.
Note the high concentrations at the onset, followed by a steady
decline to approximately 0.5 mg/1. In Fig. 3, which reports data
from Recovery Well 2A, note the fluctuating values after very low
initial concentrations at the 25-wk mark. These pulses of hydro-
carbons were especially apparent after Week 35, and between
Weeks 56 and 70. Recovery Well 4A (Fig. 4) also had an organic
species concentration spike at 35 wk, but subsequently was most
stable. The data from Recovery Well 7 (Fig. 5) indicate the same
apparent fluctuations as Recovery Well 2A. The exception was a
very large pulse in hydrocarbon concentration at Week 42.
I 10 II tO U SO ft 40 IS tt SI 10 IS TO 71
wms UHC* jucorur ornunata occur
n 6«n*«tu + toItMfM O xylene
Figure 3
Concentrations of Contaminants in Recovery Well 2A

Figs. 6 through 8 give the same results as those previously men-
tioned, however, the concentration scale has been expanded to
better define the patterns of pulses in aromatic species concentra-
CONTAMINATED AQUIFER CONTROL 39
-------
14

U -

n -

it •

14 -

II •

II -

I -

I
ii tc u u u 4t 4i it is 10 u n n
ma ana ucanu omunturs mix
Figure 4
Concentration of Contaminants in Recovery Well 4A
f.»-

IJ -

I.I -

I •

0.1 -

0.7 -

0.1 •

O.I -

t.4 -

0.3 -

0.* •

O.I -

0 -
°a
a
< 10 ii to ts 10
4t 4i to u to u n n
mis sura ncomtr orxiumm fteur
D 6m**rw + iotusn* o xyltnt

Figure 6
Concentration of Contaminants in Recovery Well 2A
t t ie ti it u 10
ma sorer xrcomr oniunom tiuit
O b«tu*n« + (otufru o
Figure 5
Concentration of Contaminants in Recovery Well 7
4t 4i it n to n 70 n
nM|iii«iiiii i i i»i«mi i i i i
I 10 II 10 Ii 30 IS 40 41 It II
ma ana Ktcomr ortiunaia item
O b*ru«n« ^ ColtMrw 0
Figure 7
Concentration of Contaminants in Recovery Well 4A
» l*'*^
u n n
tion. Using these graphs in conjunction with the rainfall data
from the area of the spill (Fig. 9), a notable correlation between
organic species concentration and rainfall quantity is evident. The
response delay in organic species concentration to a high rainfall
period appears to be on the order of one week or less. Comparing
the data from Recovery Well 2A (Fig. 6) with the local rainfall
data, the correlation is good, especially starting from Week 40.
Note especially the high organic species concentrations en-
countered in Week 60 in response to high rainfall between Week
59 and 60. Well 7 (Fig. 8) appeared to be approaching "clean"
levels up to Week 23. However, Recovery Wells 4A and 7 appear
to have the same fluctuation pattern starting at about Week 27.
Data from both wells indicate strong correlation between the
amount of contaminant present and the rainfall at the site.
It is suggested that the continuing contaminant problem could
be at least partially attributed to roadway runoff. The spill site is
located on a heavily used state highway; seepage from motor
vehicles can provide a number of the same organic species found
at the spill site.
o +
D O
0 0 « •
tfti cP a " » a
+ * "oo o a .
™ rf.°* "" o»« oo
.0° °I o O « •
ma sore* xrcomtr ortiunom net*
a ttnntu » lotiun* o
Figure 8
Concentration of Contaminants in Recovery Well 7
40 CONTAMINATED AQUIFER CONTROL
-------
I -
I.I -

1.1 •
1.4 -

I -
O.I •
0.1 -
0.4 -
O.t
I It II 10 14 U 32 31 40 44 41 SI SS S» 13 17 71
Figure 9
Variation in Reported Rainfall with Time

DEVELOPMENT AND ACCLIMATION
OF AN AEROBIC CULTURE
Batch reactors were designed to demonstrate the feasibility of
aerobic microbial degradation of the components of gasoline.
Several investigators have confirmed activity of aerobes on
petroleum-based substrates.1^ The working volume of each
fermentor was 1000 ml. Each fermentor was aerated with instru-
ment air sparged through a coarse-fritted PYREX dispersion
tube.
A large ethylene glycol condenser was fitted into the off-gas
plumbing from each fermentor to control volatilization of lighter
molecular weight aromatics through reflux. The organic vapor
traps were operated at 2°C, corresponding to a partial pressure of
7 mm of mercury for benzene, the most volatile component pre-
sent. Off-gas from each reactor was passed through a trap and
analyzed for carbon dioxide production by two Horiba PIR-2000
infrared detectors. Mixing within each reactor was accomplished
using a magnetic stirrer. There were no temperature control pro-
visions on the batch reactors because temperature fluctuations
were not significant during the course of the acclimation ex-
periments; temperature averaged approximately 25 °C. Fig. 10
shows the basic configuration of the batch reactors used in this
phase of study.
-Cold fluid
, Inlit
Port for ftldlngi
-.0. fret.
Return to
cooling bith
Vr

i HftMriMll
1-1 Her
firavntor
BOO m\ working
VOllM
Dliptrilon
Tube
— 1 %
O
Htanttlc Sllrrtr
COOUUT STSTW OITMl
Wittr bith/dreulitkrCool
Figure 10
Fill and Draw Batch Reactor for Development
of Microbial Culture
Two fermentors were seeded with identical microbial cultures,
obtained from secondary sewage liquor from a local sewage au-
thority. The buffer used in this phase of study was 0.075 molar
potassium phosphate adjusted to a pH of 7 with sodium hydrox-
ide as necessary. The nutrients present in the medium included:
500 mg/1 NH4NO3
100 mg/1 MgSO4. 7H2O
0.5 mg/1 FeCl3. 6H2O
10 mg/1 Ca(NO3)2
10 mg/1 yeast extract
Feeding the reactors was implemented by a fill-and-draw tech-
nique.5 After allowing the cell mass in the reactor to settle, 30%
of the reactor volume was drained daily and replaced with fresh
medium and substrate.
The acclimation of the microorganisms from organic carbon
supplied by the glucose (GOC) to organic carbon supplied by the
petroleum (POC) was accomplished by feeding only GOC to the
reactor, followed by a slow increase in feed POC commencing on
the second day, eventually reaching 100% POC by the end of 14
days. To determine comparative performance levels, one batch
reactor was fed glucose for the duration of the study. The target
total organic carbon concentration for combined petroleum and
glucose fractions was 1000 mg/1.
The POC substrate fed to the reactor was prepared through
water extraction of unleaded gasoline. An Erlenmeyer flask was
filled with equal amounts of gasoline and water and was placed on
a stirrer for 24-hr under a fume hood. The stopper on the flask
was fitted with a pressure relief valve to control internal pressure.
After agitation, the two phases were allowed to separate. The
water soluble fraction was drained off and analyzed by Total Car-
bon analysis followed by chromatographic assays for benzene,
toluene, xylenes, ethylbenzene and other water extractables which
had been found in the field.
Carbon dioxide production was evident from the onset of the
batch acclimation experiments in the one-liter fermentors. The
microbial activity remained high and stable until petroleum or-
ganic carbon (POC) was the sole carbon source fed to the reac-
tors. At that point, the activity decreased slowly, falling to nil in a
period of 3 days. It is suspected that either the agglomeration of
individual cells or cell growth did not occur in the absence of
organic carbon as glucose (GOC), causing inadequate settling of
the cell mass prior to withdrawal of 30% of the reactor volume at
the time of daily feeding. This sequence of events led to "wash
out" of the microbial population and necessitated starting the ex-
periment again.
These results indicated the possibility that co-metabolism plays
a significant role in the acclimation of a microbial culture to a
petroleum-based substrate and, further, is necessary in the con-
tinued viability of that culture. Co-metabolism in this sense ap-
plies in a case where organic carbon as glucose is used for
microbial growth, while the petroleum fraction is metabolized for
cell maintenance.5 With this in mind, a new experiment was con-
ducted utilizing a slower approach to 100% petroleum organic
carbon. In addition, the target culture cell concentration ap-
peared to stabilize at a GOC:POC ratio of approximately 1:6.
The acclimation reactor showed total organic carbon concentra-
tions of approximately 200 mg/1 after feeding and typically de-
creased to approximately 40 mg/1 after 24 hr, an 80% removal
rate in that time period. Upon analysis of the carbon dioxide
evolution curve over a 24-hr period, it appeared that glucose was
metabolized very quickly, followed by a lag phase which after
several hours, led to sporadic carbon dioxide generation
suspected to be gasoline degradation. The acclimation reactors
have continued to operate in this manner for 3.5 mo. Experiments
to determine the time scale of microbial degradation presently are
CONTAMINATED AQUIFER CONTROL 41
-------
in progress.
To address the volatilization question, a control reactor of the
same type as the other ferrnentors was assembled, except a solu-
tion of 1 % sodium azide was used as a bactericide. The ethylene
glycol condenser trap for volatile organic species was im-
plemented as before. The reactor was fed organic carbon accord-
ing to the same schedule as the others. The results indicated that
minor losses of approximately 0.5% were present, but such losses
would be difficult to distinguish from analytical error.

SOIL COLUMN EXPERIMENTS
This phase of study was carried out to simulate field condi-
tions. The experiment was implemented using 6-in. diameter, 4-ft
long PYREX columns. Features of the column included hydraulic
control of the saturated zone level, aeration provisions within the
column and subsurface feed of volatile substrates (to control
back-diffusion).
The column packing materials included glass wool and screen-
ing to prevent clogging of the column outlet, glass beads to evenly
distribute the fluid flow and soil from the spill site. Analysis of
the local soil indicated that it was 94% sand, 2% silt and 4% clay.
A detailed flow diagram for the soil column construction is shown
in Fig. 11.
Table 1
Operating Parameters for Laboratory Soil Columns
StndAGliss betds
Glass woffl
Alt colunns constructed fro*
6-Inch 1.0. PYREX process
pipe with polypropylene end
flanges
Effluent Collection
Figure 1 1
Soil Column Employed for Gasoline Biodegradation

Instrument air was injected subsurface to each column at an ap-
proximate rate of 0.9 1/min. Thirty milliliters of concentrated
nutrient solution were added to the soil surface daily. The com-
ponents of the concentrated solution are as follows:
500 mg/1 (NH4)2SO4
100 mg/1 MgSO4 x 7H2O
500 mg/1 FeCl3 x 6H2O
100 mg/1 glucose
0.5 M KH2PO4 buffer
An 0.2 m calcium chloride solution was the bulk liquid used to
provide transport for the components through the column. The
operation parameters defined at startup are shown in Table 1.

42 CONTAMINATED AQUIFER CONTROL
Colinm Huaber
iW.W'iJ.LT *
•OT*it£T1SOLDTIOi X
DAILY
_ a V

SIHOLI FEED Of
FUEL IWULSIOH AT
ORSET OF STODT
X
X
x

X
x
x

X
X
X
x

X
x x

X
During operation, the permeability of the column (i.e., effluent
volume) is monitored along with the effluent pH. Periodic atomic
absorption spectroscopy assays were performed on the column ef-
fluent to demonstrate the absence of heavy metal contamination
and monitor proper sodium adsorption ratios (SAR).
Total organic and inorganic carbon measurements were made
for both influent and effluent streams. This analytical methodol-
ogy was sufficiently sensitive to concentrations of 20 mg/1. Provi-
sions were made for the analysis of volatile species, also. Defini-
tion of the organic species prevalent in the column was done by
high performance liquid chromatography (HPLC) when hydro-
carbon concentrations were above SO mg/1. Gas chromatography
including a purge and trap concentrator and flame ionization
detector was employed when analysis in the /tg/1 range was
necessary. Where possible, the analyses were in conformance in
Standard Methods.6
Operation of the soil columns for approximately 45 days has
yielded some preliminary results. The column permeabilities have
been high and relatively constant. Those columns which were
Table 2
Effluent Concentrations from the Soil Columns
INFLUENT: 200 ppM TOTAL ORGANIC CARBON, WITH POC:GOC OF 6:1
COLUMN DATE FLOW pH TC TOC TIC
1 ml ppH PPN ppN
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
1
2
3
4
5
6
9/30
9/30
9/30
9/30
9/30
9/30
10/1
10/1
10/1
10/1
10/1
10/1
10/3
10/3
10/3
10/3
10/3
10/3
10/11
10/11
10/11
10/11
10/11
10/11
10/16
10/16
10/16
10/16
10/16
10/16
10/19
10/19
10/19
10/19
10/19
10/19
1020
1050
650
1075
700
375
775
975
1000
1100
650
625
880
460
450
925
375
7SO
1040
855
700
830
945
200
775
1000
675
885
750
650
1200
1200
475
1200
1100
820
6.2
6.12
6.47
6.48
6.67
6.95
6.35
6.33
6.68
6.85
6.67
6.99
6.22
6.47
6.84
7.01
6.73
6.82
7.21
6.79
6.82
6.74
6.92
6.41
7.22
7.13
6.86
7.01
7.28
7.28
7.16
7.07
6.89
6.97
7.23
7.2
132.48
105.9
117.42
102.56
48.77
79.77
128.11
146.53
44.7
175.31
42.5
74.68
191.74
191.01
141.56
142.35
31.8
83.79
54.06
87.54
95.16
93.56
33.9
40.46
52.57
87.54
98.19
96.62
32.98
40.38
55.54
87.54
92.16
90.5
39.96
40.54
121.4
97.99
107.77
95.43
44.49
73.7
119.33
137.42
36.59
167.16
38.44
70.71
183.61
181.43
130.8
134.51
26.48
78.8
46.95
76.51
90.2
90.66
27.87
31.52
45.55
76.04
93.64
90.17
28.22
34.88
48.36
76.99
87.08
84.16
34.67
35.15
11.01
7.91
9.65
7.1)
4.31
6.07
1.71
9.41
1.11
1.19
4.01
).»7
1.1)
9.M
10.7(
7.14
5.)]
4.»
7.1
ll.CJ
I.M
T.W
SM
I.M
7.W
u.»
4.95
1.45
4.«
5.9
7.11
10.99
9.M
f.M
9.»
9.»
-------
seeded with microorganisms have been maintaining an effluent
pH of about 7.1. Effluent from the columns which were not seed-
ed have been slightly more acidic, with pH averaging approx-
imately 6.5. This observation indicates that biodegradation is pro-
ceeding, although it is unlikely that the columns are being
operated in a stable regime.
Some difficulties became evident beginning at the 30th day.
Feed to the columns was comprised of 3.5 ml of pure unleaded
gasoline in the case of daily feeding and 35 ml of pure gasoline for
the single batch feed case. Some reduction in total organic/in-
organic carbon concentrations was evident.
One matter that required thorough investigation was the validi-
ty of the influent/effluent comparison. Pure gasoline was fed to
the columns through the 30th day. Some of it is water soluble,
and the remainder, which is less dense than water, forms a
predominantly petroleum phase. The effluent from the columns
has had only one phase from the onset of the study, which made it
necessary to question the fate of the petroleum fraction. It is
suspected that these components, which make up the heavier frac-
tion of the gasoline, were becoming trapped within the column
matrix via capillary forces. This result is reflected in the total
organic carbon (TOC) results, which show that the effluent con-
centrations are substantially comparable to those of the aqueous
feed. Table 2 shows preliminary results of the soil column ex-
periments. It is possible that bacterial activity within the columns
may decrease significantly if capillary forces hold petroleum at
sites where the organisms would normally function.

Table 3.
Effluent Concentrations from Soil Columns, Effluent Recycle

INFLUENT: 120 ppM TOTAL ORGANIC CARBON, WITH POC:GOC OF 6:1
COLUMN
1
DATE
FLOW
ml
PH
TC
ppM
TOC
ppM
TIC
ppM
11/19
11/19
11/19
11/19
11/19
11/19

11/21
11/21
11/21
11/21
11/21
11/21

11/24
11/24
11/24
11/24
11/24
11/24
600
740
750
790
810
760
400
475
495
800
700
1215
550
750
850
950
200
500
6.85
6.92
7.01
6.98
6.62
6.54
6.86
6.93
7
6.96
6.67
6.52
6.85
6.87
7.09
6.99
6.7
6.59
62.71
38.95
58.65
40.07
88.5
95.65
42.96
52.22
61.5
58.05
75.25
81.03
59.64
48.23
45.21
39.41
100.6
65.29
49.4108
29.9236
48.3913
34.8047
85.9026
92.6254
31.7531
43.2233
46.9372
46.0994
72.0932
76.3716
47.0922
38.3877
35.0648
26.4965
97.1727
62.5677
13.2992
9.0264
10.2587
5.2653
2.5974
3.0246
11.2069
8.9967
14.5628
11.9506
3.1568
4.6584
12.5478
9.8423
10.1452
12.9135
3.4273
2.7223
To address this problem, the feeding of pure gasoline to the col-
umns was discontinued. The effluent was recycled through the
columns, and each column was reinoculated. At the present time,
the columns are in an equilibration period. Two preliminary
analyses indicate that the current biodegradation process is pro-
ceeding more successfully at this point. Table 3 lists the results of
these analyses. After the current supply of moderate TOC ef-
fluent is exhausted, the substrate feeding will be continued with
water-extracted gasoline (the preparation of which has been
heretofore mentioned). In addition, a continuous feed of a trace
glucose solution will be introduced to the columns, as it is be-
lieved the microbial population within the columns is utilizing the
glucose supplied much too rapidly, a factor which tends to hinder
the co-metabolism process.

PRELIMINARY CONCLUSIONS
The investigation of the gasoline spill site has yielded some in-
teresting results. Comparison of contaminant concentrations with
local rainfall data has provided a positive, though not well-
defined, correlation. Results of this phase of the experiment sup-
port the validity of the roadway runoff hypothesis. The role of
runoff in the persisting contaminant levels at the spill site is not
accurately defined; study will continue in this area.
The acclimation of secondary sewage sludge microbes appears
to have been successful. Relatively long-term experiments have
demonstrated that substantial reduction of total organic carbon
as petroleum is attainable in the presence of co-metabolism.
The preliminary results from the soil columns point to a
positive conclusion. It is difficult to define the performance to the
microbial mineralization process before an equilibration time has
passed. The presentation at the Hazardous Wastes/Hazardous
Materials 1987 meeting will include additional data obtained after
the time of this writing.

ACKNOWLEDGMENTS
The able and dedicated assistance of Mr. Jonathan Mokrauer
and Ms. Alison Redmond has been greatly appreciated. The
author is very grateful to Mr. Art Rosenbaum of OH Materials
for his assistance. In addition, Mr. Gene Sturdyvin of Sayre and
Toso is acknowledged for his interest in this research.

DISCLAIMER
The citation in this document of commercially available pro-
ducts does not constitute the approval or endorsement either of
Rutgers University or that of the author for the use of such pro-
ducts.

REFERENCES
1. Atlas, R.M., "Microbial Degradation of Petroleum Hydrocarbons:
An Environmental Perspective," Microbiol. Rev., 45, 1981, 180-203.
2. Tabak, H.H., Quave, S.A., Mashni, C.I. and Earth, E.F., "Biode-
gradability Studies With Organic Priority Pollutant Compounds,"
JWPCI, 53, 1981, 1503-1518.
3. Grady, C.P.L., Jr., "Biodegradation: Its Measurement and Micro-
biological Basis, Biotech. Bioeng., 27, 1985, 660-674.
4. Wodzinski, R.S. and Johnson, M.J., "Yields of Bacterial Cells from
Hydrocarbons,"/lp/7/. Microbiol., 16, 1968, 1886-1891.
5. Venkataramani, E.S. and Ahlert, R.C., "Role of Cometabolism in
Biological Oxidation of Synthetic Compounds," Biotech. Bioeng.,
27, 1985, 1306-1311.
6. Greenberg, A.E., Trussel, R.R. and Clesceri, L.S., Standard
Methods For the Examination of Water and Wastewater, Port City
Press, Baltimore, MD, 1985.
CONTAMINATED AQUIFER CONTROL 43
-------
Pilot Interceptor Drain Performance: A Method for
Remedial Investigation and Aquifer Evaluation
John R. Mildenberger
Brian V. Moran, P.E.
Geraghty & Miller, Inc.
Annapolis, Maryland
ABSTRACT
Groundwater interceptor drains (French drains) are being used
increasingly to abate shallow aquifer contamination problems. In
this paper, the authors present results of the use of this remedial
technique for an underground storage tank leak at a former
chemical packaging and distribution facility which had contam-
inated groundwater with halogenated volatile organic compounds
(VOCs). Results of a drain performance test indicated that de-
watering equations can be used with confidence to predict the dis-
charge and capture area of a French drain used for contaminant
control in shallow groundwater. The paper presents a comparison
of actual drain performance to theoretical predictions.

INTRODUCTION
A limited initial investigation identified the problem and de-
scribed the hydrogeologic setting as a shallow watertable aquifer.
Preliminary calculations predicted that a pilot interceptor drain
could be installed in lieu of the usual remedial investigation en-
tailing a monitoring network, aquifer tests and evaluation of
various remedial alternatives. This process was proposed in order
to expedite remediation of the property at the client's request.
Several monitor-well clusters and piezometers were installed to
evaluate pilot drain performance and help determine the hydraul-
ic characteristics of the shallow aquifer. This action, at a signif-
icant cost savings to the facility owner, accomplished the objec-
tives of the remedial investigation and established a pilot remed-
ial system to begin recovery of contaminated groundwater.
The setting of this investigation is the Piedmont, where very
low permeability bedrock is overlain by a veneer of relatively
moderate permeability saprolitic residuum. Watertable levels are
shallow, making this an ideal situation for a drain. After esti-
mations of discharge rates and concentrations of VOCs (several
hundred mg/1) were made, negotiations were initiated for direct
discharge of the contaminated groundwater to the local POTW
during the performance testing of the drain. A 100-ft long, 16-ft
deep drain and monitoring system was installed in 5 days at a cost
of just under $20,000. At the completion of the pilot drain con-
struction, a 2-mo constant drawdown performance test was con-
ducted to characterize the shallow groundwater aquifer hydraul-
ics and, thus, determine the feasibility of full-scale remediation
via hydraulic controls. The hydraulic characteristics of the re-
siduum (saturated saprolite) were calculated using classical de-
watering equations. After evaluating the hydraulic parameters for
the residuum and water-quality data collected during the per-
formance test, recommendations were made to extend the drain
an additional 100 ft in order to ensure complete capture of the
plume.

44 CONTAMINATED AQUIFER CONTROL
SETTING
The Piedmont, where the facility is located, contains largely
residual soils overlying a saprolite (highly weathered rock) reser-
voir, overlying bedrock which may contain interconnected joints
and faults. The thickness of each zone depends on the lithology
of bedrock and topography. Topographically, the site is near the
crest of a hill; however, there is little more than 5 ft of relief over
the property. Surface runoff drains to a swale running east along
the northern fence line of the property. This swale, which gen-
erally flows only when it rains, eventually empties into the head
waters of a local creek some distance from the site. The area sur-
rounding the site is generally characterized by light industrial and
commercial use.
The study area is underlain by a sheared granite which is coarse,
pink schistose and gneissic rock containing green schistose and
slatey dikes. A fairly uniform 30 to 35 ft of overburden overlies
the bedrock. During drilling at the site, bedrock was encountered
as an abrupt transition from the overlying unconsolidated ma-
terial. The overburden consists primarily of residual soils formed
in situ from the parent bedrock. However, a thin veneer of fill
material 1 to 2 ft thick occurs over most of the north end of the
site.
A well-developed soil horizon comprises the first 3 to 4 ft of re-
sidual soils. This initial three- to four-foot, brown to orange/
brown, silt, clayey horizon exhibits some soil structure (i.e., frac-
tures) which is lost as a gradual transition is made to a micaceous
sandy silt varying from orange/brown to green, black, and white.
These transitional soils are occasionally saprolitic (i.e., demon-
strate structure inherent to parent rock) and become more so
with depth. The saprolite below a depth of 10 to 15 ft is green,
black and white micaceous, fine, sandy silt and is fairly uniform
beneath the site. The lack of evidence of fracturing in the sapro-
lite indicated relatively unfractured parent bedrock. Fig. 1 i* •
cross-section of the subsurface lithology.
The water table was encountered at depths ranging from 9 to
11 ft below land surface, resulting in a saturated thickness of 20
to 25 ft above the relatively low permeability bedrock. Inferred
flow direction for the shallow groundwater beneath the site a to
the north at a moderate gradient of just over 0.01. The shallow
groundwater flowing beneath the site presumably discharges to
the head waters of a local creek which occurs some distance from
the study area.

FIELD PROGRAM
A drilling and soil-sampling program was conducted in con-
junction with the construction of a pilot groundwater intercep-
tor drain at the facility. Three, two-monitor-well clusters (shallow
-------
(NO
100-
9O-
80-
70-
6O-
t.
ITHl
°I5J
"I
LE p-
HAH DRAIN
/

:->!
|
1

-I P'
saaMOJtt
Littl. fintSond
SOondCUY
UffltToSoiM
FhtSond.Mkmow
SAPROUTE
(McacwMSitaml
FinSond)
-2 P'

R5
A'
. (SOUTH)
CORROSIVE TANK
DOCK

i
EXPLANATION
SCREENE6 INTERVAL
0 HofL 100
5l — VERTCAL EXAGGERATION
BEDROCK
—
Figure I
Cross-Section Illustrating Subsurface Lithology
and deep) and 10 piezometers were installed to depths ranging
from 14 to 30 ft (Fig. 2). The monitor wells and piezometers were
constructed to facilitate the collection of water-quality and pilot
drain hydraulic performance data. Boreholes were drilled to
depths of 15 ft (for shallow wells) or 30 ft (for deep wells). Hand-
auger boreholes for piezometer installations were drilled to 14 ft.
Core samples were taken at 5-ft intervals in the 30-ft borings for
purposes of lithology description.
After completion of the monitor-well and piezometer installa-
tion, the pilot drain construction began. Fundamentally, the
drain is a trench penetrating the water table and backfilled with
a high permeability material to enhance flow to a collection sump
installed at one end of the drain. The initial construction con-
cern was to install the sump so that dewatering could begin. This
inhibited trench wall cave-in during the remainder of the con-
struction. A 20-ft length of 30-in. diameter corrugated aluminum
pipe was used for the collection sump. Before the sump was low-
ered into the trench, the bottom 2 ft were slotted to facilitate
ground water flow from the gravel envelope. When the sump was
in place, a pump was lowered into the pipe so that collected water
could be pumped out.
A 4-in. diameter perforated plastic pipe was installed into the
sump at an elevation IS ft below ground surface and laid on a
gravel envelope at a 2% grade leading to the sump. The 0.5- to
1.0-in. gravel envelope was maintained at a minimum 4-ft thick-
ness and capped with a permeable geotextile fabric to prevent
clogging with silt and clay from the backfill material. After the
100-ft trench was completed, the remaining open excavation was
backfilled with excavated soil. The pilot drain construction de-
tails are illustrated in Fig. 3.
The sump is equipped with a submersible pump with an auto-
matic level controller adjusted to maintain the water level a min-
imum of 16 ft below ground surface. A plastic discharge line car-
ries the collected groundwater to the surface where a sampling
spigot and flow meter (recording total gallons discharged) were
installed. From the flow meter, the discharge line carries the
drain effluent to a nearby sanitary sewer access.
PLOT DRAW
NOTEl
PPEauUOMA4FT>4ndUMCt.
ENVELOPE «TZ%«UDC
Figure 2
Map of the Facility Illustrating Monitor Well and Piezometer Locations
:!— "&S38P-

—'%"^"tK
Figure 3
Cross-Section of the Pilot Drain Detailing Construction
CONTAMINATED AQUIFER CONTROL 45
-------
Table 1
Results of Water Quality Analyses

Carbon Tetrachloride
Chlorobcnzene
1 ,2-Dlchloroeth*ne
1,1,1 -Tr lchloroeth»ne
1 , l-Dlchloroeth»na
1,1, 2-Trlchloroeth»ne
1,1,2, 2-Tetr»chloroethane
Chloroeth*ne
2-Chloroethyl Vinyl Ether (Mixed)
Chlorof oro
1, 1-Dlehloroethylene
1 ,2-Tr«n»-Dlchloroethylene
1 , 2-Dlchloropropane
1 , 3-Dlchloropropylene
Methylene Chloride
Methyl Chloride
Methyl Broalde
Bromoform
DlchlorobrODomethane
ChlorodlbroBooethane
Tetrechloroethylene
Trlchloroethylene
Vinyl Chloride
Detection
Llnlt
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
FT-1S
SDL
BDL
BDL
BDL
100
BDL
BDL
3,400
BDL
BDL
200
BDL
100
BDL
BDL
BDL
BDL
BDL
BDL
BDL
5,200
8,000
BDL
PT-1D
BDL
BDL
BDL
3,600
600
BDL
BDL
BDL
BDL
BDL
4,900
BDL
8,100
BDL
400
BDL
BDL
BDL
BDL
BDL
2,400
90,000
BDL
PT-2S
BDL
BDL
BDL
200
300
BDL
BDL
37,000
BDL
BDL
700
BDL
1,400
BDL
BDL
BDL
BDL
BDL
BDL
BDL
600
14,000
BDL
PT-2D
BDL
BDL
BDL
34,000
500
BDL
BDL
BDL
BDL
BDL
14,000
BDL
20,000
BDL
29,000
BDL
BDL
BDL
BDL
BDL
2,600
91.000
BDL
PT-3S
BDL
BDL
BDL
8,000
600
BDL
BDL
500
BDL
BDL
5,400
BDL
19,000
BDL
3,100
BDL
BDL
BDL
BDL
BDL
1,400
81,000
BDL
PT-3D
BDL
BDL
BDL
70,000
600
BDL
BDL
BDL
BDL
BDL
20,000
BDL
5,100
BDL
41,000
BDL
BDL
BDL
BDL
BDL
8,000
69.000
BDL
PD Discharge
BDL
400
BDL
9,700
200
BDL
BDL
2,300
BDL
BDL
6,900
BDL
4,100
BDL
4,900
BDL
BDL
BDL
BDL
BDL
11,000
89.000
BDL
All Concentration! In Mlcrograms/Llter
BDL - Below Detection Limit

GROUNDWATER QUALITY
Six cluster wells and 10 piezometers were sampled in addition to
the pilot drain discharge. The cluster wells and drain discharge
samples were analyzed for halogenated volatile organic com-
pounds by U.S. EPA Method 601 and total organic carbon
(TOC). Piezometer samples were analyzed only for TOC, as a
surrogate parameter to screen for additional plume-area defi-
nition. Results of analyses via U.S. EPA Method 601 for all clus-
ter wells and the drain discharge have been summarized in Table
1. Nine compounds were detected at levels ranging from 100 >ig/l
(1,2-dichloropropane and 1,1-dichloroethane in Well PT-1S) to
91,000pg/l [trichloroethylene (TCE) in Well PT-2D]. Of the nine
compounds, TCE was the most prevalent, occurring consistently
in concentrations exceeding 8,000 >ig/l in all cluster wells and the
drain discharge. 1,1,1-Trichloroethane was detected at the second
highest concentrations, with values ranging from 200)ig/l in PT-
2S (none was detected in PT-1S) to 70,000;ig/l in PT-3D.
At the request of the local municipality, and as an additional
means of monitoring drain performance, water-quality analyses
were performed on the discharge via U.S. EPA Method 601 for
halogenated volatile organic compounds.
Parameters encountered in the discharge analyses essentially
mirrored those encountered in the cluster wells, as discussed in
the Plume Area I section. A general increase was observed in
VOC concentrations over the 2-mo period. This increase in VOC
concentrations was expected, since the plume was drawn toward
the drain under artificial gradients induced by pumping. Further-
more, contaminant concentrations are greater at depth, and less
contaminated shallow water is removed during the early stages of
pumping.
An increase in concentration trend is evident. This trend will
eventually peak and be followed by decreasing concentrations as
the contaminants are removed by the drain.

DRAIN PERFORMANCE AND AQUIFER
EVALUATION

Theoretical Basis for Groundwater Drain
Hydraulics
The hydraulic performance of a groundwater interceptor drain
can be analyzed using an anology to radial flow to a well discharg-
ing at constant rate when equilibrium is reached. The drain then
is considered to be essentially a "line of pumping wells" in this
case. The condition for a water-table aquifer and one pumping
well is:
Q =
K(H2-h2)
458 In Ro/rw
(1)
For a line source in a water-table aquifer, the rate of flow per
unit length of trench for one side of the trench may be expressed
as:
Q/X =
K(H2-h2)
2880L
(2)
where:

Q
H
h
Ro

X
K
= Drain discharge (gal/min)
= Saturated thickness of residual soils (ft)
= Saturated thickness at trench during pumpage (ft)
= L = Approximate radius of influence or distance to
zero drawdown (ft)
= Radius of well, one-half of which is located at each end
of the trench (ft)
= Length of the trench between the well halves (ft)
= Permeability of residual soils (gal/day/ft1)
For a long narrow system, such as the interceptor drain in
question for this application, the ratio of length to width u suffic-
iently large that a combined model can be used; the line source**
given by Equation 2 above is coupled with the Jacob Well Equa-
tion 1 which accounts for the end effects of pumping the drain.
It is important to note that the end effects have greater pump-
ing effects than those in the center of the drain. The combined
equation is as follows:
46 CONTAMINATED AQUIFER CONTROL
-------
Q =
K(H2-h2)

458In Ro/rw
+ 2
XK(H2-h2)

2880L
(3)
As shown in the above equation, the line source equation
(Eq. 2) has been doubled to account for withdrawals from both
sides of the drain.
An initial estimate of radius of influence (Ro) is also neces-
sary to determine the capture area provided by the drain system.
The most reliable way to determine Ro is to use a pumping test,
as was done in the field evaluation. An initial theoretical approx-
imation of Ro can be made, however, from estimated aquifer
parameters utilizing the following equation:
R0 = rw +
Tt
C4Cs
(4)
where:

T
t
C4
= Transmissivity (gal/day/ft)
= Time(min)
= Storage coefficient
= Constant (unit dependent)
The values used for €4 and Cs are 4790 and 0.2, respectively,
for a water-table situation.4 The above equation indicates that
RO is a function of time, which is usually selected from pumping
duration considerations. Ro, in this case, makes no allowance for
recharge; the equation is only valid for confined situations. Ac-
cording to Powers, however, it can be used as a rough approxima-
tion to make first-cut approximations for water-table aquifer de-
watering.
In order to obtain a temporary POTW discharge permit, the
drain discharge rate (Q) had to be estimated. From the field pro-
gram, the aquifer thickness (H) was determined to be 20 ft. The
drain design established the length of the trench between well
halves (x) at 96 ft and the radius of the well, one-half of which is
located at each end of the trench at 2 ft. The thickness of the
aquifer at the drain during pumping will be constant at 14 ft. A
mid-range permeability for a silty sand of 20 gal/day/ft2 was
assumed in lieu of permeability testing.2 The radius of influence
was calculated to be 268 ft assuming 1 mo to reach steady state.
These parameters were used in Equations 3 and 4 to estimate a
drain discharge of 2.8 gal/min.
In the field testing program, it was determined that a "system"
permeability, K, rather than a series of permeability tests via slug
tests or other means which might vary depending on location and
procedural error, was preferable. In order to accomplish this and
provide a valid "K" value for aquifer evaluation, the drain was
operated until quasi-steady-state conditions were achieved. The
actual system K could then be evaluated by rearranging Equation
3 as follows:
K =
(H2-h2)
1
(5)
458 In Ro/rw
Hydraulic Evaluation of Drain Performance
At the completion of the pilot drain construction, a 2-mo con-
stant drawdown performance test was conducted in order to char-
acterize the shallow groundwater flow regime and thus determine
the feasibility of remediation via hydraulic controls. The pumping
level in the drain was set at approximately 16 ft below land sur-
face. Water from the drain was disposed of via metered discharge
to the sanitary sewer, with permission from the city.
Flow rates from the drain were averaged weekly and ranged
from 4000 to 6000 gal/day with an average rate of 4,600 gal/day.
The response to pumping was monitored weekly in all monitor
wells and piezometers for the duration of the 2-mo test. In addi-
tion, a water-level recorder was installed on PT-1S to continuous-
ly monitor the water-table response. Periodic water-level meas-
urements were made in all cluster wells and piezometers. At the
end of the 2-mo period, a final set of water levels was collected
and the resulting water-table map was constructed. The cone of
depression and approximate area of influence that the pilot drain
developed during steady state pumping are illustrated in Fig. 4.
To illustrate the interception of groundwater flow in the resid-
ual soils overlying the bedrock, a flow section (Fig. 5) was con-
structed. Flow is from high to low hydraulic head contours shown
by the arrows in this figure, indicating the generalized direction
of groundwater flow to the drain. In addition to interception of
flow from the south, gradient reversals extending for a similar dis-
tance north of the drain are also expected, since the water table
is lowered on both sides of the drain during pumping. A conser-
vative radius of influence for the pilot drain of 215 ft (Ro) was
estimated from semi-logarithmic distance/drawdown plots. These
drawdown plots for wells and piezometers constructed in prox-
imity to the drain were extrapolated to zero drawdown as indi-
cated in Fig. 6, 7 and 8.
Figure 4
Water Table Map Illustrating Generalized Flow to the Drain During
Steady State Pumping
The hydraulic characteristics of the shallow unit (permeability
of the saturated residual soils) were approximated using Equa-
tion 6:
K =
(H2-h2)
1
(6)
458 In Ro/rw
Using this equation, a K of 20 gal/day/ft2 (~10~3 cm/sec) is
estimated for the residual soils. In estimating Q prior to the per-
formance test, a permeability for the observed residuum was ap-
proximated and a radius of influence was calculated using that
value for K. The calculated permeability was equal to the approx-
imated value. Although the measured Q was greater than the esti-
mated Q, this difference was offset by a smaller measured radius
of influence than had been estimated prior to the performance
test.
CONTAMINATED AQUIFER CONTROL 47
-------
Figures
Conceptual Flow Section Illustrating Flow to the Pilot Drain During
Steady State Pumping
4
i
T
i
3.3 I-
4
i
2.3 i-
i
i
1
Z L.
1
1.3 -

1 h
•4
0

—i — r— n-rrrn 1 — r— mrrn i i i i i MI
PTIS

"
.
'T2S
\
\ -

\PT3S
°\ -
\
\
\
\
\
\ J?.= 250FT
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 \/l 1 1 1 M |
18 IBB IB
Distance From Pilot Drain (Feet)
Figure 6
Distance vs. Drawdown Plot of PT-1S, PT-2S, and PT-3S During Steady
State Pumping of the Pilot Drain
HA-2
Distance From Pilot Drain (Feet)
Figure 8
Distance vs. Drawdown Plot of HA-4, HA-S, and HA-6 During
Steady State Pumping of the Pilot Drain
CONCLUSIONS
Based upon the testing and results of the pilot groundwater
recovery drain system, a groundwater drain was proven to be an
effective control device for recovery and prevention of migra-
tion of contaminated groundwater in and around the site, as
evidenced by the cone of depression developed by the system
and favorable hydraulic response to pumping. Hydraulic evalua-
tion of the drain performance exhibited a permeability of 10~J
cm/sec for the saturated residual soils overlying the low perme-
ability bedrock. Water-quality testing near the pilot drain ex-
hibited a well-defined plume emanating from the area of the
former buried tanks which is intercepted by the pilot drain
system.
The pilot drain was proven to be a viable remedial device for
interception and discharge of contaminated groundwater from
the site via the sanitary sewer. Continued operation and monitor-
ing now is being carried out to generate trend data on the water
quality discharge. These data will provide insight on the time
necessary for the groundwater to achieve "clean" conditions.
\HA-7
•I65FT.
\W I
1_1 I Illl £J L-.l.l Mill
IBB ta
I IB
Distance From Pilot Drain (Feet)
Figure 7
Distance vs. Drawdown Plot of HA-8, HA-7, and HA-2 During Steady
State Pumping of the Pilot Drain
REFERENCES
1. Daniels, C.D., III and Sharpless, N.B., "Groundwater Supply Poten-
tial and Procedures for Well-Site Selection Upper Cape Fear Fiver
Basin," U.S.G.C., North Carolina District Office, Raleigh, NC,
1983.
2. Freeze, R.A. and Cherry, J.A., Groundwater, Prentice Hall, Engl*
wood Cliffs, NJ, 1979.
3. LeGrand, H.E., "Groundwater of the Piedmont and Blue Ridp
Provinces in the Southeastern States," U.S.G.S., Circular 538,19£T.
4. Powers, J.P., Construction Dewatering, John Wiley & Soni, Ne»
York, NY, 1981.
48 CONTAMINATED AQUIFER CONTROL
-------
The Use of Extraction Trenches to
Remove Industrial Solvents from Shallow,
Low Permeability Alluvial Aquifers

Lance D. Geselbracht, P.E.
Thomas A. Donovan, P.E.
Richard J. Greenwood, P.E.
Canonic Environmental Services Corp.
San Mateo, California
ABSTRACT
A hydrogeologic investigation was completed to determine the
lateral and vertical extent of industrial solvents in a shallow,
alluvial groundwater aquifer. The groundwater gradient was de-
termined in order to predict chemical plume migration in the
affected aquifer. Based on extensive evaluation, it was deter-
mined that an extraction trench, excavated perpendicular to the
direction of plume movement, was the most appropriate remed-
ial measure for the site.
Two trenches were installed at depths ranging from 15 to 24 ft
below ground surface. A shorter trench was installed immediate-
ly downgradient of the source area and a longer trench was in-
stalled at the property line. The two trenches were designed to
prevent continued off-site migration of the chemical plumes. Fol-
lowing construction, each trench was operated for a 30-day
period to determine their hydraulic impact on the groundwater
aquifer and their initial effect on the mitigation of the chemical
plume.

INTRODUCTION
Hydrogeologic investigations were performed for a semicon-
ductor manufacturing facility in Northern California's Santa
Clara Valley. During these investigations monitoring wells were
installed and sampled to determine the extent of chemical con-
tamination. The results of analysis on water samples provided
evidence that industrial solvents had contaminated the underly-
ing shallow aquifer. Additional investigations were implemented
to determine the extent of contamination at the site. The data
obtained through these efforts was utilized to design and imple-
ment interim remedial measures. The additional remedial investi-
gations included: installation of three small diameter extraction
wells downgradient of the source area and a complete series of
hydrogeological tests to determine the relative permeability of the
affected aquifer.
Pump testing of the three extraction wells revealed a sustain-
able yield of only 0.1 to 1.0 gal/min. This low production rate
could be expected to produce an extremely limited cone of influ-
ence. As a result of the pump testing, groundwater extraction
utilizing wells was determined not to be feasible. Based on the
limited cones of influence for the wells, it would be nearly im-
possible to create a hydraulic barrier with an array of extraction
wells. Also, the mechanical and operational difficulties assoc-
iated with constant cycling of a well pump would become a sig-
nificant cost factor. Therefore, extraction trenches were selected
as the most appropriate remedial measures to extract and treat
the contaminated groundwater and prevent continued off-site
plume migration.
This paper discusses the design, construction and initial opera-
tion (monitoring) of the extraction trenches at the semiconduc-
tor manufacturing facility. In addition, site information on the
design and construction of extraction trenches installed at other
sites, under the supervision of the authors, has been included.

DESCRIPTION OF THE SITE
The site of this work is located in the central portion of a gently
sloping alluvial plain that extends from the base of the Santa Cruz
Mountains northward to the San Francisco Bay. The alluvial
plain is composed of stream-deposited materials ranging from
coarse-grained materials (sands and gravels) to fine grained ma-
terials (silts and clays). In general, the shallow soils present be-
neath the site are predominantly fine-grained (clay size) materials.
The "A" aquifer, the shallowest water-bearing zone beneath
the site, was encountered in borings at depths ranging from 7 to
13 ft below ground surface. This aquifer varies in thickness from
less than 5 to 15 ft. The regional groundwater gradient, defined
by measured piezometric surfaces within the "A" aquifer moni-
toring wells is 0.005 (dimensionless). The groundwater flow direc-
tion is northeast toward the San Francisco Bay.
The "B" aquifer is more permeable, but less consistent than
the "A" aquifer and varies more in depth and thickness. In gen-
eral, the "B" aquifer is comprised of interbedded layers of silts
and clays in a sand and gravel matrix. The permeable and effec-
tive water bearing zones lie between 30 to 60 ft. Some areas, how-
ever, are comprised of continuous silt and clay zones which, sim-
ilar to the "A" aquifer, provide preferred flow paths. The "B"
aquifer is confined and is under a higher hydrostatic head than
the "A" aquifer.
The 15-ac site under remediation is within a semiconductor
manufacturing facility that produces silicon wafer chips. As part
of the manufacturing operation, the industrial solvent, trichlor-
ethylene (TCE), was used as a degreaser during the manufactur-
ing processes. The virgin solvent and waste solvent were stored in
underground tanks at the site.
Hydrogeologic investigations were initiated to determine if the
underground tanks leaked, resulting in TCE leaching into the
"A" aquifer. Subsequent to these investigations, which deter-
CONTAMINATED AQUIFER CONTROL 49
-------
mined that leaks had occurred, the tanks and surrounding con-
taminated soils were excavated and disposed as a hazardous
waste.

HYDROGEOLOGIC INVESTIGATION
During and following removal of the leaking underground
storage tanks, 30 monitoring wells and three small diameter (less
than 6 in.) extraction wells were installed. Boring logs from the
well installations were used to construct geologic cross-sections
of the site. Undisturbed soil samples were obtained during the
drilling operation and were selectively tested for permeability.
A hydrogeologic description of the site was developed based on
the results of the permeability testing and evaluations of the bor-
ing logs. In general, the site consisted of a continuous "A" aqui-
fer with preferred flow paths or "channels." These channels,
alluvial fans, are separated by finer grained materials of lower
permeability. The entire "A" aquifer is underlain with a contin-
uous aquiclude ("A-B") with an estimated permeability of
1 x 10~6 to Ix 10~7 cm/sec.

Groundwater Sampling
A sampling and monitoring program was developed and imple-
mented as part of the initial hydrogeologic investigations; this
program included measurement of piezometric surfaces and col-
lection of groundwater samples. Groundwater samples, obtained
from each well, were chemically analyzed. Interpretation of the
chemical data revealed two distinct and separate plumes existed
within the "A" aquifer. The "B" aquifer also contained TCE
and 1,1,1-trichloroethane (TCA). Evaluation of on-site and off-
site data concluded that the second "A" level plume was from an
off-site, upgradient source that was migrating onto the site. Both
of the "A" level chemical plumes are depicted on Fig. 1. Chem-
icals within the "B" aquifer were determined to be the result of
vertical migration from both "A" aquifer plumes.

Pump Testing
To supplement previous hydrogeologic testing step-drawdown
and 48-hr continuous discharge pump tests were conducted on the
three small diameter extraction wells. These tests were conducted
to determine the average permeability (k) and transmissivity (T)
of the "A" aquifer. Also, "B" aquifer wells were monitored
PUMP CONTROL PANEL
WITH LEVEL CONTROL
WATER TIGHT MANHOLE SWITCHES
rnvFR SFT Tn HRADF 1

vi
N
g
•2
tn
a
^

1 ," i A.C PAC€MENT WITH
_i_jf_" / ROCK SUBASE
|| 'nf-f.hio* • 0-,Jt^ •'„'.•> ••'*'•„* ^.c

SYSTEM
^
ffl"l ^ ''
g.^ ^NATIVE BACKFILL-'^ ;
52 / 1 FROM UNSATURATED ZONE) i

V
Jk""*" °*l
•* ^PERMABLE BACKFILL ?"°
i . Y MATERIAL £
-------
no response to the pump tests. No response was noted in any of
the "B" aquifer wells during the pump testing.

EXTRACTION TRENCH DESIGN
Due to the low permeability of the "A" aquifer and the high
cost of multiple extraction wells, extraction trenches were selected
as the most appropriate remedial alternative for the ground-
water cleanup. Initial chemical and hydrogeological evaluations
had established two distinctive and separate chemical plumes,
each with different characteristics. Based on this information,
two separate trenches were selected.
Based on the available data, preliminary alignment and total
depth was established for each of the two trenches. To confirm
these preliminary design conditions, additional soil borings and
monitoring wells were completed. The soil borings were drilled
to verify the thickness of the "A" aquifer and to obtain undis-
turbed soil samples from the underlying "A-B" aquiclude. Per-
meability tests were conducted on the undisturbed soil samples
to confirm that the "A-B" aquiclude had a minimum permeabil-
ity of 1 x 10~6 cm/sec.

Selected Design
Each trench was designed to penetrate 2 ft into the low perme-
able "A-B" aquiclude. The length of each trench was established
such that the limits corresponded to the horizontal limits of the
chemical plumes. Based on previous pump testing an estimate of
the sustainable yield for each trench was calculated and collec-
tion sumps were sized accordingly to the corresponding flow
rates. Each pump and the mechanical piping network were de-
signed to enable operation between 3 and 20 gal/min. The se-
lected design section of the sump is provided in Fig. 2.
The design width of each trench was established based on con-
struction constraints, including the shoring technique to be used
during the excavation process and the minimum width required
by construction equipment. Based on assumed construction con-
ditions, a minimum three foot bottom width and a 0.75H:1V
trench side slope were selected. However, in the event that the side
slopes could not be maintained during excavation, optional
trench protection utilizing a trench shield was also selected. A
typical cross-section of the selected extraction trench, showing the
two options, is provided in Fig. 3.

Soil Disposal
Provisions were also incorporated into the design for disposal
of the contaminated soils excavated from the saturated zone. The
TCE contained in the soils were volatile and considerations were
given to mechanically aerating them. However, two factors neces-
sitated the disposal of the soils as a hazardous waste. First, the
trenches were going to be installed during the winter rainy sea-
son which would have limited the number of dry, sunny days and
all important, warm weather. Second, there was a limited amount
of paved surface to spread the soils for the mechanical aeration
process.

Final Details
Once the final alignment, depth and length of the trenches were
established, some secondary design parameters were determined.
A small diameter gravel (D50 less than pea gravel) was selected as
the permeable backfill material to be utilized within the satur-
ated zone. This type of material was locally available at a rea-
sonable cost and had been used successfully in the past for trench
drains under nearby building foundations. The low permeability
and resultant low trench infiltration rates expected, provided evi-
dence that filter fabric would not be required. Therefore, the
trenches were designed without filter fabric to reduce cost.
The pump within each trench and the associated control panel
were wired with process control type relays to prevent overflow at
the activated carbon treatment unit. A final set of design draw-
ings and specifications were assembled to allow final cost estimat-
ing and construction.

CONSTRUCTION
General
Construction equipment required for the project included a
Caterpillar 235 hydraulic track excavator, a Caterpillar 950 front
end loader, a small (10-ton capacity) hydraulic crane, a sheeps-
foot compactor, steel shoring shields, dewatering pumps and mis-
cellaneous small tools and supplies. Construction began with the
longer and deeper trench near the property line. The approxi-
mate excavated width of the trench was laid out on the asphalt
parking lot and then the asphalt was saw cut.

Sump Installation
The sump was excavated prior to the trench to allow ground-
water to flow toward the sump as the trench excavation pro-
gressed. After the final depth of approximately 25 ft was reached,
1.5-in. drainage gravel was used as a solid base for the first 4-ft
diameter concrete manhole section. Rather than pour an invert in
the first section, a special section was used that contained a con-
crete invert. After the first section was set, subsequent concrete
manhole sections were installed until the final conical section was
installed near the ground surface. The excavation was then back-
filled with 3 to 5 ft of drain rock. The unsaturated soils, initial-
ly removed from the excavation and stockpiled, were placed over
the gravel. Saturated soils were stockpiled and later sampled to
determine if they required disposal as hazardous waste.

Trench Excavation
The first portion of the trench excavation was approximately 30
ft long, to allow the installation of the steel trench shield. Exca-
vation was attempted with vertical slideslopes, prior to shield
placement to depths of approximately 23 ft. Repeatedly, portions
of the vertical sidewalls sloughed off and fell into the excavation
before the trench shield could be installed. In order to remedy
this condition the excavation was sloped to 0.75H:1V on both
sides. All future trench excavation utilized benching of the side-
walls for the first 5 vertical feet, prior to excavating vertically to
the final subgrade of the trench.
The trench shield was installed into the open excavation with a
small crane. A temporary dewatering pump was placed into the
concrete sump, plumbed with hoses to a settling basin and
allowed to gravity discharge through a temporary carbon treat-
ment vessel. Then, backfill material (small diameter gravel) was
placed into the trench to create a bedding for the infiltration pipe.
After the trench shield was in-place, workmen were allowed into
the trench to place the bedding material and infiltration pipe.
An OVA meter was used to periodically monitor for the pres-
ence of potentially explosive vapors. Also, calculations were per-
formed to verify that the OSHA 8-hr exposure standard of 100
ppm for TCE would not be exceeded by the workmen in the bot-
tom of the shield. Standard safety equipment included rubber
boots, gloves, eye-protection and hard hats.
Installation of the trench was completed by excavating approx-
imately 5 ft in front of the shield, pulling the shield forward with
the excavator by the large cables attached to the front of the
shield. As the shield was moved forward, additional gravel ma-
terial was placed, creating bedding for the infiltration pipe. Back-
filling operations continued behind the trench shield as gravel ma-
CONTAMINATED AQUIFER CONTROL 51
-------
terial was placed within the saturated zone. After the gravel had
been placed, the native unsaturated soil was utilized to backfill
the remaining portions of the trench. A sheepsfoot compactor
was used to compact the native soils to prevent settlement from
creating deflection of the resurfaced asphalt.
A few dewatering problems occurred during the excavation
operations. Large quantities of silts and fines periodically clog-
ged the infiltration pipe and concrete sump. These silts also clog-
ged the temporary dewatering pump placed in the sump. A de-
watering pump was attached to the trench shield to prevent fu-
ture clogging of the infiltration pipe. Although there were some
difficulties with the discharge hoses conflicting with the con-
struction operation, the installation of the dewatering pump pre-
vented future problems with the infiltration pipe and sump. The
settling basin and carbon filter, for the groundwater treatment
system, received an excess amount of silt, but operated satisfac-
torily until the end of the project.

Confirmation Testing
When the trench excavation reached its designed terminus
point, soil samples were obtained from the "A" aquifer material.
The results of chemical analysis on these samples confirmed the
aquifer was not contaminated, therefore the trench was termi-
nated at that point. During the backfill operations, three piezo-
meter points were installed with 4-in. diameter PVC well screen
along the length of the trench. These piezometer points were
used to measure the water surface within the trench and remove
groundwater samples for chemical analysis.
The saturated soils were placed in small, segregated stock piles
so that confirmation testing could be completed prior to disposal.
Soils that contained less than 10 mg/kg of TCE were returned as
backfill to the trench. The remaining saturated soils were dis-
posed as a hazardous waste. The trench was surfaced with a layer
of base rock and asphalt cement. The permanent pump with float
switches was installed and wired to the pump control panel. A
conveyance pipeline was then installed to allow discharge of the
extracted groundwater to the activated carbon treatment system.

Source Remediation Trench
The second trench was installed utilizing the same methods as
the initial trench. Since the "A" aquifer is not very permeable
at the second trench location, dewatering did not pose the same
problems as with the larger trench. No sidewall sloughing oc-
ccurred during excavation work within the second trench. The
trench was backfilled, the permanent sump and pump installed
and the surface restored within 1 wk. All saturated soils from the
second trench excavation were disposed to a Class I facility be-
cause of the elevated groundwater concentrations of TCE in the
groundwater.

PERFORMANCE TESTING AND MONITORING
Testing Procedure
A 30-day performance test was conducted to determine the im-
pact of the trenches upon the "A" aquifer. Prior to the perfor-
mance tests, static water levels for the "A" and "B" aquifers
were measured in the monitoring wells and piezometer points
along the trench. The static water levels were compared to the
levels obtained the previous year. During the 30-day extraction
program, water level readings in selected "A" level wells were
recorded on a weekly basis. Since the trenches were approximate-
ly 350 ft apart, the zone of influence for the two trenches did not
approach each other during the 30-day period.
Numerous water level readings were taken to later enable con-
struction of distance-drawdown graphs for the monitoring wells
surrounding the trenches. The distance-drawdown graphs were
used to calculate the permeability (k) and transmissivity of the
"A" aquifer in the vicinity of the trenches. These data were then
used to compare the calculated yield from the trenches to the ac-
tual yield. In addition, the resulting information was used to pre-
dict plume movement in the "A" aquifer and verify that trenches
were acting as hydraulic barriers. A typical drawdown curve from
the performance testing is provided in Fig. 4.

Interpretation of Performance
During the 30-day test period, the groundwater elevation in
nearby "A" level monitoring wells gradually and steadily de-
creased. Near the larger trench along the property line, greater
hydraulic influence was observed for wells along a north/south
axis. Monitoring wells along an east/west axis were not influ-
enced. The same type of hydraulic response was noted for the
shorter trench. This observation coincides with the results of in-
itial hydrogeologic evaluations.
- SUBMERSIBLE PUMP

- COLLECTION " SUMP"
Figure 4
Typical Drawdown Curve
At the end of the 30-day period, water level readings for the
monitoring wells and trenches were used to construct distance-
drawdown diagrams on semilog paper. The slope of the line plot-
ted on these graphs was then used to calculate aquifer transmis-
sivity (T) in units of gal/day/ft of drawdown. Transmissivity
valves were used to calculate permeabilities (K) for the aquifer in
the vicinity of nearby monitoring wells. The thickness of the
aquifer, estimated from the boring logs, was used in the calcula-
tion of K.
Permeability values ranged from 0.01 to 0.0075 cm/sec for the
large trench and from 0.004 to 0.0015 cm/sec for the small trench.
The calculated values showed good correlation to permeability
values developed from evaluations of materials identified from
boring logs.1 The average extraction rate for the large trench
was 8.3 gal/min and 0.7 gal/min for the small trench.

CONCLUSION
Concentrations of TCE in groundwater samples from the
large trench remained constant during the 30-day period at ap-
proximately 10,000 jig/1. Concentrations in the small trench in-
itially started at 10,000 ug/1 but eventually rising to 100,000
pg/1 at the end of 30 days. The increase in TCE concentration
was attributed to higher concentrations within the chemical plume
near the source area and a general movement of chemicals to the
extraction trenches.
Groundwater elevations in the monitoring wells surrounding
the trenches and piezometer points within the trenches indicate
that a hydraulic barrier had been formed along the length of the
trench. The sustainable yields of the respective trenches indi-
cates groundwater extraction with wells would not have created «
effective of a hydraulic barrier as the extraction trenches created.
52 CONTAMINATED AQUIFER CONTROL
-------
The results of the 30-day performance testing and the final
cost to install the trenches ($350 to $450/ft) provided evidence
that extraction and treatment of groundwater, utilizing extrac-
tion trenches, is a more cost-effective alternative than establish-
ing a hydraulic barrier with an array of wells. The trenches also
have the added advantage of a lower yearly operation and main-
tenance cost. The use of extraction trenches should be closely
compared to the use of extraction wells when remediation of
shallow, low permeable alluvial aquifers is required.
RELATED WORK
Since the installation and testing of the trenches presented with-
in this report were completed, application of these techniques
have been utilized at several sites throughout the United States
and specifically in California. Trench depths have varied between
10 and 25 ft with lengths of up to 700 ft.
Construction techniques have also improved, thereby reducing
cost and construction time. Recent techniques have included
utilization of geofabric drain material in place of gravel back-
fill. Utilization of the fabric allows backfill of the trench with
native material. This technique reduces the disposal volumes
(the most significant cost factor) and reduces the required use of
costly gravel drain material. In addition improved production
rates of up to 40 gal/min have been produced in similar trenches.
Continued applications of these techniques should produce addi-
tional ways to decrease costs and provide multiple use for on-site
treatment.

REFERENCES
1. Matthewson, C.C., Engineering Geology, C.E. Merrill Publishing
Company, 1981.
CONTAMINATED AQUIFER CONTROL 53
-------
Benefits of Mitigating Releases from
Underground Gasoline Tanks

Donald W. Anderson
Research Triangle Institute
Research Triangle Park, North Carolina
ABSTRACT
The U.S. EPA is considering a set of regulatory alternatives de-
signed to mitigate releases from existing and new underground
storage tanks (USTs). A Regulatory Impact Analysis (RIA) is be-
ing prepared to evaluate the benefits and costs of the regulatory
alternatives under consideration.
In support of the RIA effort, a case-study examination of 33
actual UST leaks in 7 states has been conducted. Case files from
state environmental offices have been collected and reviewed,
and in-person interviews have been conducted for several UST
incidents. The intent of the study is to learn from affected parties
themselves the nature and magnitude of the welfare losses they
have incurred. The information gathered is being used to estimate
the benefits of mitigating releases from underground storage
tanks.
It is apparent that tank owners, owners of nearby businesses
and nearby homeowners all may incur costs when leaks occur.
Leak prevention would result in both pecuniary and non-pecun-
iary benefits. The magnitude of these benefits would vary great-
ly from case to case, and further research is necessary to draw
meaningful conclusions about the relative wisdom of alternative
regulatory strategies from a net-benefits perspective.

INTRODUCTION
Congress created a program under Subtitle I of the 1984 Haz-
ardous and Solid Waste Amendments to address the problems
posed by leaking underground storage tanks. The U.S. EPA
estimates that as many as 1.5 million underground storage tanks
(USTs) are used in the United States to contain hazardous sub-
stances and petroleum products. As many as 300,000 of these
tanks may be leaking.
Subtitle I requires the U.S. EPA to develop standards for new
and existing tanks. Specific regulations are now being developed,
and the regulations may be quite costly. Executive Order 12291
requires that proposed major regulations be subjected to a formal
Regulatory Impact Analysis (RIA). An RIA is, in essence, a bene-
fit-cost analysis. It requires a calculation and comparison of a
proposed rule's benefits and costs. Subject to other laws, regula-
tors are directed to propose rules that maximize the net benefits.
Net benefits analysis requires the projection of a rule's bene-
fits. Benefits are changes in individuals' well-being for which they
are willing to pay. They may include not only pecuniary improve-
ments, but also "quality of life" improvements. Consequently,
benefits analysis should examine how a regulation may affect
society in all of these ways.
This paper presents some preliminary findings of a benefits
analysis of regulations intended to reduce the number and severity
of leaks from USTs. It comprises an account of how particular
individuals were affected by actual leaks, as well as exposing how
individuals, including tank owners were adversely affected by
leaks. The potential benefits of reducing such leaks are explored.

METHODOLOGY
A case study methodology was selected for this analysis. The
case study approach assesses types and magnitudes of damages
that can occur by examining damages from leaks that have
occurred. Actual UST leaks have been identified and affected
parties questioned about the costs they incurred. The approach
has a number of distinct advantages, including:
• Affected persons may identify damages that would otherwise
be overlooked
• Reported damages are "real" rather than abstract
• Affected persons may be able to provide monetary estimates
of damage costs
• Affected persons can indicate how they actually felt following
incidents
Damage estimates have been compiled from case files and from
personal interviews. Owners of leaking tanks, owners of affected
nearby homes and owners of affected nearby businesses were in-
terviewed. Damages resulting from contamination of both
groundwater and air were examined.
Leaking UST incidents were identified in the following manner.
Officials in a number of East Coast state environmental regula-
tion offices were contacted by telephone. They were asked to de-
scribe their UST programs and file maintenance procedures; they
were asked if they would be willing to open their files for examin-
ation. Cooperation was quite good, and UST leaks in seven states
were examined.
RTI file examiners reviewed hundreds of files and selected i
subset for further study based on several practical criteria:

• Incidents that are less than 5 yr old
• Cases that are well documented, particularly regarding
damages
• Cases with files that are open to public inspection
• Incidents that resulted in documented environmental damages

The case study selection process clearly was not statistically
random. Indeed, UST incidents were selected primarily for the
richness of damages they exhibit. The bias introduced by the
selection process is obvious. The selection process may have iden-
tified and profiled incidents causing more than normal damage-
Consequently, the findings reported here should be interprettd
with caution. They are not necessarily indicative of damages that
are imposed by an "average" or "typical" leak. They do indicate
types of damages that can occur in certain circumstances.
54 CONTAMINATED AQUIFER CONTROL
-------
At the same time, it is not necessarily accurate to assume that
the case studies represent "worst cases." The sample of incidents
is very small, and we were not allowed to examine files for certain
leaks that currently were under investigation by state environ-
mental officials. It is likely that some of these cases were closed
precisely because they were serious enough to warrant thorough
investigation and litigation. Furthermore, we know informally of
other incidents not profiled that have resulted in damages sub-
stantially more serious than those profiled.
The case study selection process yielded a total of 33 incidents
for examination. The 33 incidents are located in 7 states, and
most involve gasoline leaks at retail gasoline establishments
(Table 1). The availability of information about damages varied
greatly from case to case. For some, it was possible to interview
involved parties by telephone or in person. Many individuals,
including "victims", were unwilling to be interviewed. Others
could not be located, even though all of the leaks occurred rela-
tively recently. For these cases, damages had to be assessed by ex-
amining information in the files alone.

RESULTS
The discussions of damage costs are arranged around three
categories: on-site, off-site business and off-site residential. On-
site damages are those that occur on the property of the tank own-
ers and normally are borne by the tank owner. Off-site business
and residential damages are those affecting persons and proper-
ties of other businesses and households, respectively.

Table 1
Case Study Coverage
Area Covered Number
Leak sites 33
Towns and cities 27
Counties 19
States 7
Groundwater regions 5

ON-SITE DAMAGES
The cost of product loss is the most frequent, if one of the more
minor, costs borne by owners of leaking tanks. Of the 33 cases
studies, the smallest product loss was 500 gal while the highest was
62,500 gal. The mean and median losses were 11,700 gal and
3,500 gal, respectively. The large discrepancy between the two
measures of central tendency is explained by the presence of sev-
eral very large leaks. Available evidence suggests that the product
loss estimates are very crude. For example, product recovery from
groundwater sometimes exceeds the inventory record-based loss
estimate.
An appropriate approximation of the cost of product (gasoline)
loss is the retail price of gasoline. This estimate reflects not only
the cost of the gasoline to the tank owner who bought it, but his
opportunity cost as well. Given the uncertainty inherent in loss
estimates, the convenience of assuming a cost of $l/gal should
not be overridden by concerns about inaccuracy so introduced.
Loss estimates then translate directly into cost estimates.
It is reasonable to assume that most if not all of the leaks re-
sulted in temporary or permanent business loss since tanks rou-
tinely are emptied following leak verification. At least 5 of the 33
businesses examined in the case study closed permanently as a re-
sult of their UST leaks.
When gasoline retailers suspend gasoline sales, revenues ob-
viously decline. At the same time, though, certain costs decline.
The relevant measure of closure cost is revenue loss.
Available data from several incidents imply a revenue loss of
$1,400 to $l,700/week of closure. The longest "temporary"
closure in our cases was 6 yr at an estimated cost of $500,000.
Other known closures lasted several days, several weeks and sev-
eral months. There appears to be a great deal of variability in
closure periods, but it seems reasonable to assume that any leak
will result in a closure of 1 wk or more.
Another on-site damage involves tank evacuation and replace-
ment. Leak verification by state authorities is always followed, at
minimum, by removal of contents. Most often, this removal step
is followed by excavation of the leaking tank and sometimes of
surrounding tanks of similar age. Before normal operations can
resume, new tanks are installed.
Cost estimates for this entire operation range from about
$10,000 to $40,000 per incident. One incident reported a cost of
$180,000, but this included some (unknown) amount of soil re-
moval as well.
Remedial actions involve attempts to restore affected soil and
groundwater to their initial quality. Most remedial action pro-
grams investigated involved some combination of bioreclamation,
recovery wells and barrier trenches.
The range of estimated costs is very large—from $10,000 to
$1.5 million. Most operations, however, were estimated to cost
from $60,000 to $250,000.
A candid interview with one station owner revealed other on-
site costs. Lost time and interest on outstanding debt were re-
ported to be significant. The time spent by the tank owner as a
direct result of the leak dealing with authorities, affected parties,
contractors, station managers and others is estimated at about 700
hr. The estimated interest cost on outstanding debt is $30,000.
While it may not be appropriate to generalize about these types
of costs based on a single incident, the particular incident in ques-
tion is not abnormal in any apparent, major way.
Table 2 summarizes the on-site damage results from the 33 case
studies.
Problem
Product Loss

Business Loss
Tank Evacuation/
Replacement
Remedial Actions

Other Costs
Table 2
On-Site Damages

Range: 500-62,500 gal
Mean: 11,700 gal
Median: 3,500 gal
Permanent closure: 5 cases
Temporary closure: 1 wk to 6 yr
Cost per week: $1,400-$ 1,700
Cost range: $10,000-$40,000

Range: $10,000-$!.5 million
Typical: $60,000-$250,000
Lost time
Interest on debt
Legal costs
Impact
OFF-SITE BUSINESS DAMAGES
Business establishments within contamination plumes may be
affected by vapors or by contamination of their wells. Potential
detrimental effects include reductions in sales, increased costs of
production, lost time and health and safety risks.
The case studies confirmed at least four incidents with eight
businesses affected by vapors.
Businesses may elect or be forced by authorities to suspend
operations when vapor levels in their establishments reach
nuisance, unhealthy or dangerous levels. The resulting business
CONTAMINATED AQUIFER CONTROL 55
-------
losses can be substantial. Closures may last only hours, or they
may last days. In some instances, establishments may close per-
manently. Of the eight establishments known to be affected by
vapors, seven closed temporarily and one closed permanently.
The owner of a restaurant and bar that never reopened follow-
ing a leak about 1 yr ago is claiming $67,000 in damages from the
station owner.
Several other businesses affected by the same leak closed for
between 4 and 7 days. A copy center closed for 1 wk at an esti-
mated total net cost of $3,000. The majority of this loss is prob-
ably lost net revenues. A fast food restaurant closed for 7 days
and, once reopened, experienced a decline in business for about
2.5 mo due to parking lot disruptions. The store manager esti-
mates the loss of gross revenues at about $70,000. That same
incident also temporarily closed a large grocery store, a laundro-
mat and a photo booth.
Separate incidents closed a clothing store for 3 days and a
municipal middle school for 1 day.
None of the business cases examined revealed expenditures by
owners on additional productive inputs. Undoubtedly, one cost
imposed on all owners was lost time. The owner of the copy store
estimated that he spent about 60 hr dealing with authorities, the
station owner and suppliers as a direct result of the leak. The
owner of the fast food restaurant reported lost time totaling
about 16 hr.
The case studies revealed not less than six businesses affected
by well contamination in three separate incidents.
Damages from well contamination mainly involved increased
production costs. A lawn and garden center has incurred inven-
tory related costs that very likely exceed $3,500 and continues to
incur such costs. As a result of the same incident, a day care
center purchased bottled water for several weeks before connect-
ing to a municipal water line.
In a separare incident, three small businesses responded to well
contamination by purchasing filters at an estimated cost of over
$750 each. Bottled water also had to be purchased for consump-
tion purposes.
Following a leak in the Northeast, the owner of a diner pur-
chased bottled water for about 1 yr and then incurred the cost of
establishing a new, deep well.
Table 3 summarizes off-site business damages.
Table 3
Off-Site Business Damages
Problem
Vapor Damages
Well Damages
Business Loss

Other Costs
Impact
4 leaks, 8 businesses
3 leaks, 6 businesses
Permanent closure: 1 business
Temporary closure: 8 businesses
Lost time
Inventory loss
Bottled water
Filter systems
OFF-SITE HOUSEHOLD DAMAGES
Like businesses, households can be affected by vapors, well
contamination, or both as a result of a tank leak. The case study
revealed about 100 households affected one way or the other by
18 incidents.
Vapors from eight leaks affected about 35 households. Costs
were incurred for evacuations and remedial actions.
The most costly incident caused the evacuation of 25 homes for
10 mo at a cost (to the tank owner) of about $600,000. This cost
only reflects expenditures for rental lodging, food, laundry, etc.
Transactions and psychological costs have not been estimated
but are likely to be substantial. The implied lower-bound cost
per household is $2,400/mo.
Another incident caused the evacuation of four homes for 7
mo. The estimated cost of the evacuation alone, using the above
data, is approximately $67,000. Three of the families ultimately
sold their homes for about $150,000 each. The four family is seek-
ing compensatory damages of $246,000.
In another incident, the damages at two condominiums were so
severe that the tank owner elected to purchase them from the
owners at a cost of approximately $300,000. It is not at all clear
that the new owner will be able to re-sell them at any price.
Other documented vapor incidents include evacuation of four
condominiums for several days and evacuation (non-use) of an
apartment building garage and storage area for several days.
Vapor threats sometimes result in remedial actions costs.
Following one New York incident, costs directly related to
eliminating vapors from a single home were approximately
$53,000. In another incident, damages to an apartment building
garage and storage area were repaired with a sand and cold patch
at an unknown cost.
The case study has documented at least 50 cases of residential
well contamination from 10 leaks. Typical responses to well con-
tamination include some combination of using bottled water,
using filtered water, establishing a new well and connecting to a
municipal system.
The best data are from a Florida leak that contaminated 20
residential wells in a suburban area with a municipal water line
nearby. Households purchased bottled water for consumption for
1 yr at an average cost of $40/household/mo. Households also
purchased filter systems so they could use well water for non-con-
sumption purposes. The cost per household was approximately
$650. At the end of 1 yr, households connected to nearby water
mains at a cost of about $750/household. The total cost to date is
about $36,000. Of course, households will pay for monthly water
service from now on.
In another incident, one family evacuated their home for an un-
known period. They then consumed bottled water for several
months before purchasing a carbon filtration system for approx-
imately $2,000.
The apparently most costly incident affected 22 residential wells
in an area not serviced by municipal water. All 22 families have
been using bottled water for several years at a cost of about
$30,000. The planned permanent solution involves establishing a
new well field and connecting the homes at an estimated cost of
$2.2 million.
The psychological cost to homeowners affected by gasoline
vapors or well contamination appears to be significant. Interest-
ingly, homeowners expressed more frustration than anger. Home-
owners worried, of course, about long-term health effects but
usually did not blame their illnesses on exposure to vapors or con-
taminated water. They were extremely uneasy about not knowing
how long the contamination had been present before detection.
They were frustrated about their state's unwillingness or inability
to address the problem once it occurred. Owners of contaminated
wells worried about using filtered well water for bathing and
washing, even though they were told by authorities that it was
safe to do so. They also expressed concern that if it happened
once, it could happen again.
56 CONTAMINATED AQUIFER CONTROL
-------
Table 4 summarizes observed off-site household damages.
Table 4
Off-Site Household Damages
Impact
8 leaks, 35 homes
10 leaks, 50 homes
Duration: several days to 10 mo
Typical cost per household: $80/day
Bottled water: $40/household/mo
Filters: $650-$2,000/household
Municipal connections: $750 and up per house-
hold
Health concerns
Lost time
Aesthetic damages
Frustration, anger
CONCLUSION
The 33 leaking underground storage tank case studies indicate
that regulations to mitigate leaks would benefit tank owners as
well as business owners and homeowners near potential leaks.
The range of pecuniary damages of any particular description is
Problem
Vapor Damages
Well Contamination
Evacuations

Water Replacement

Other Costs
very wide. For example, the ultimate corrective cost for a contam-
inated residential well can be as low as a few thousand dollars or
as high as tens of thousands of dollars. The cost is a function of
many things, especially distance from a municipal water line or an
uncontaminated well field. Vapor damages are equally variable.
Non-pecuniary damages—mainly psychological costs—can be
considerable. These impacts have not been valued in dollar terms,
although it is within the realm of modern welfare economics to
do so. It is interesting that homeowners are more likely to exper-
ience (or at least express) such costs than tank owners or owners
of affected businesses. One might speculate that being adversely
affected in one's home is somehow more objectionable than be-
ing affected in one's place of business.
The results reported here are quite preliminary. It would seem
very tenuous to generalize from these results alone about the costs
of the "typical" leak or the "national benefits" of regulating
underground storage tanks. This study does, however, identify
many of the issues that need to be addressed in a comprehensive
benefits analysis.

ACKNOWLEDGEMENT
This study was supported in part by the U.S. EPA. Special
thanks to Dr. Glen Anderson of the Agency's Office of Policy
Analysis in Policy, Planning and Evaluation. Still, the views ex-
pressed here are solely those of the author and do not necessarily
reflect the U.S. EPA's.
CONTAMINATED AQUIFER CONTROL 57
-------
Use of the HELP Model in Evaluating the Cover
Design for a Uranium Mill Tailings Disposal Site
Will Wright
Colorado Geological Survey
Denver, Colorado
A. Keith Turner, Ph.D.
C.E. Kooper
Colorado School of Mines
Golden, Colorado
ABSTRACT
A computer model for the hydrogeologic evaluation of land-
fill performance (HELP) has been applied to the design of a uran-
ium mill tailings disposal cover. The code was developed by the
Waterways Experiment Station of the Army Corps of Engineers
specifically for use at RCRA sites. However, it is well suited as a
general model for simulating the hydrologic performance of vari-
ous types of waste disposal systems.
In this application, the generation of leachate from the base of
a planned tailings cover system was estimated by a series of simu-
lations which assessed the original design proposals and suggested
possible modifications to that design. Four types of modification
to the original design were evaluated alone and in combination:
(1) slope of the top surface, (2) evaporative zone depth, (3) per-
meability of the clay barrier and (4) length of the drainage paths.
When these parameters are combined in their most optimal con-
figuration, the HELP model predicts a significant decrease in the
percolation issuing from the base of the tailings impoundment.
The HELP model offers professionals in many areas of waste
disposal practice a powerful tool for predicting leachate genera-
tion from a disposal cell. The code is ideally suited for evaluat-
ing the mitigative effects that various design parameters may have
on the precolation from the base of an engineered impoundment.
By testing different design scenarios with the HELP model, the
overall design program for a waste impoundment can be opti-
mized in terms of cost and performance. The results of these
simulations have been presented to the Department of Energy for
review.

INTRODUCTION

The UMTRA Program
The Uranium Mill Tailings Radiation Control Act was passed
in 1978 in response to concerns for public health and safety. It
established a Federal program, called UMTRA, whose objective
is to permanently isolate, from human contact or exposure, the
various "abandoned" uranium tailings sites located within the
United States. Colorado has nine of the 24 designated sites sched-
uled for cleanup and permanent isolation. Colorado has more
than any other state in terms of tonnage and number of sites.
The authorizing legislation specified several technical and
financial considerations which greatly affect the design and accep-
tability of the engineered cover systems. The principal specifica-
tions are:
• Minimum design life of 200-yr for these disposal systems, dur-
ing which no appreciable leakage or radon emanation should
occur
• Design objective of 1,000 yr where reasonably achievable
• Zero to minimal maintenance regime for these structures over
the life of the project
• Individual state payment of 10% of the program costs dur-
ing the initial relocation and/or construction phase
• Possible ultimate state responsibility for maintaining the integ-
rity and successful performance of the various impoundments
in perpetuity once construction is completed
The first three requirements dictate a very stringent and con-
servative design standard, while the last two requirements place
Grand
Junction
.For! Collins

£ D«nv*r

COLORADO
•Colorado
Springs
'Pu.blo
.1-
Figure!
Location Map of the Town of Durango, La Plata County, Colorado
58 MODELL1NG.GROUNDWATER & SURFACE WATER
-------
potentially catastrophic future financial burdens on the various
states. As a consequence, the state review of DOE proposals is a
crucial task.

The Durango Site
The uranium mill tailings site discussed in this paper is at Dur-
ango, an important regional center in southwestern Colorado
(Fig. 1). The tailings presently are located within the town's cen-
tral business district, adjacent to the Animas River (Fig. 2). The
future disposal site for these materials is approximately 3 road
miles west of town, situated in a small basin on top of Smelter
Mountain.6-7'8 This site, known as Bodo Canyon, is at an eleva-
tion of 7050 ft, some 500 ft above the present position of the tail-
ings. These relationships are illustrated in Fig. 2
EXPLANATION

QUATERNARY

[~w"| Alluvial volley (ill
[ e« Mixed alluvium and collwium
i Q«p I Terrace or pediment

CRETACEOUS

I " I Lewis Shah

p^TI Cliff House St.
g a. | 1 Cliff Hauu/Menefee
y § Traniition Zone ^JO
SS[««]M.n.f..Fm. »
X I *fi | Point Lookout Ss.

I «• ] Mancos Shale

~^— Fault showing movement
•• Drill hob location m no.
r Tailing! ponds
• Proposed disposal site

Figure 2
Block Diagram of Study Area, Showing Relationship of Disposal Site

to Processing Site and Town of Durango, Colorado
The site is underlain by shallow valley fill and slope wash de-
posits overlying a silty shale bedrock (Fig- 2 and 3). The upper-
most portion of the bedrock is weathered and fractured. The cli-
mate is semi-arid with an average annual precipitation of 19 in.
Surface water runoff is intermittent. Groundwater is found be-
low the disposal site primarily in the shallow unconsolidated de-
posits. The elevation of the piezometric surface shows a marked
seasonal trend due to infiltration from spring snowmelt and rain-
fall during spring and summer (Fig. 4). Both surface and ground-
water flow from the site toward the Animas River, about 1.5 mi
away
6,7,8
Department of Energy Cover Design
The basic design of the proposed containment system for the
Bodo Canyon disposal site is shown in Fig. 5.' This design in-
cludes five materials which are, from bottom to top:
• Native material subgrade;
• Thick sequence of tailings (expected to be approximately 25 ft
thick)
• Approximately 6 ft of compacted clayey-silty material
• 1-ft layer of sandy material
• 1-ft layer of gravel

MODELING METHODOLOGY
The evaluation of the cover design was supported by a five-step
modeling methodology. The steps involve the estimation or
assessment of:
• Natural groundwater flow past the site
• Leachate production by simple Darcian calculations
• Leachate production by infiltration modeling
7100

7075

7050-

7O25

7000

6975 •

6950

6925
- Bottom of Weathered and Fractured Zone
Land
I Surface
/June 8,1983-
Potentiometric Surfaces 1 K«k/Kmf
'•March 9, 1983 ' Rch/*mf
"Rmf
2100
°ac- Valley Fill and Slop* Deposits
Kch-Clilf House Sandstone- Marine Sandstone, Shaly Sandstone, Silty Shale
Kmf-Menefee 'Formation-Sandstone,Slltstone, Shale with Thin Coal Seams
Kch/Kmf-Transitional Between Kch/Kmf

Data Derived from Draft EIS, Ocl.,1984,U.S. DOE
ff
Source: DUR RAP, Draft, June 1985
Figure 3
Cross section for Bodo Canyon, showing fluctuation in potentiometric
surfaces, cross-sectional area of flow, and rock units. Lower map shows
section line and site topography.
• Potential for leachate dilution
• Chemical attenuation effects

These steps are discussed in detail by Wright and Turner.10
Some involved relatively simple calculations which were per-
formed on hand-held calculators, but computer models were re-
quired to complete the process. In some instances, the values pro-
duced by these simple calculations were compared to the values
produced by the models. Such comparisons were valuable be-
cause they provided a relatively independent check on the
accuracy of the results produced from the computer models. In
this situation, when the credibility of any modeling effort was
going to be questioned, avoidance of any mistakes in the model-
ing process was extremely important. By using such comparisons,
errors in the modeling, due to incorrect data specifications for
instance, would be readily identified.
A second benefit from such comparisons was observed during
this study. Because the model results showed relatively close
agreement with the results obtained by simple, readily under-
standable, calculations in those areas where comparisons were
possible, the model predictions of conditions which could not be
MODELLING GROUNDWATER & SURFACE WATER 59
-------
'18B» '1983 yui
D*U [tarhwd front Or.fl Ernlronmvitil Impact SUUmMlt. Ocl..1«M
US. rMol. ol EMrgy,|>.PF-87

Figure 4
Sum of the Positive Slope Intervals on Hydrograph Indicate Water
Table is Rising 69% of the Record.
estimated by simple methods also were made more reliable. In
short, such comparisons improved the modeling credibility.

Simulation of the DOE Cover Design
The HELP model' proved useful in estimating the probable
volume of leachate produced from a proposed engineered im-
poundment in Bodo Canyon for uranium mill tailings. These
simulations showed that the permeability of the proposed cover
was the most critical factor controlling the generation of leach-
ate. If the suggested design permeability for the cover is in fact
achieved in the field, then HELP predicts 1 gal/min of leachate
will be produced from the site. However, larger cover perme-
abilities show correspondingly larger leachate volumes.
Estimates of groundwater infiltration values in this area sug-
gest that a dilution ratio of 30:1 may occur as this leachate moves
toward the Animas River. With this dilution ratio, simple mixing
calculations show many metals remaining above U.S. EPA drink-
ing water standards. Very low calcium carbonate levels have
been measured in the soils beneath this site. Using conservative
assumptions concerning the neutralization of the acidic leachate
generated from these tailings, we calculated the time required to
exhaust the neutralization capacity will range from approximate-
ly 4 to 90 yr; the variability in the time period is due entirely to
the estimated volume of leachate. These results further emphasize
the importance of achieving the smallest possible in-place perme-
ability for the cover.
Use of geochemical speciation models demonstrate that once
the natural neutralization capacity is exhausted, the attenuation
of the leachate contaminants is minimal because of the increased
mobility of the metals in an acidic environment. This mobiliza-
tion is further compounded by the natural occurrence of fluoride
in the shallow groundwater system. Under this scenario, other
significant attenuation mechanisms, such as sorption and ion ex-
change, are expected to be severely limited.
These simulations demonstrated that the proposed DOB de-
sign can be expected to work within the specified design life
only if the very stringent design permeability for the cover is in
fact achieved for the cover and additional carbonate material is
placed in the base of the cell and perhaps downgradient in the
Bodo Canyon area.

Preliminary Analysis of Alternative Designs
The HELP model analysis of the original DOE cover sug-
gested that some relatively simple modifications to this design
might substantially improve its performance. Four additional de-
signs were investigated, in a preliminary fashion, primarily to
determine if a more effective alternative design was obtainable
for an equal or lower cost. The HELP simulation of these ahem-
Figure 5
Proposed DOE Cover Design for the Bodo Canyon Disposal Sit*
60 MODELLING GROUNDWATER & SURFACE WATER
-------
ative designs was encouraging. Accordingly, a study of cover per-
formance was undertaken to determine the lowest infiltration flux
reasonably achievable by incorporating additional modifications
to the original DOE design.

SENSITIVITY ANALYSIS OF THE DOE
COVER DESIGN
The HELP model was used to evaluate the response of the pro-
posed DOE cover design to a sequence of design changes. The
objective was to identify the importance of various physical
parameters in improving the cover's performance. Two criteria
were used:

• Minimization of the volume of contaminants percolating from
the base of the impoundment
• Cost

Four design components were selected for study. Each was in-
vestigated while holding all others constant. The four design com-
ponents included:
1.0
• Top slope of the cover
• Thickness of the evaporative zone
• Permeability of the clay barrier layer
• Length of the drainage path
Effects of Length of Precipitation Record
on Simulation
In common with many infiltration models, the HELP model is
sensitive to the period of precipitation used during the simulation.
Actual daily precipitation records are used by the program, up to
a maximum period of 20 yr. The length of time required to com-
plete each simulation is proportional to the length of th is clima-
tological record. With the large number of simulations required
by this study, it was desirable to use a shorter simulation period,
provided adequate accuracy in the simulation results could be
maintained.
Accordingly, the model was executed using varying lengths of
time up to the maximum allowable 20-yr period. Using site-spe-
cific daily precipitation data recorded in Durango, the infiltra-
tion flux determined by a 20-yr record was typically larger than
the flux determined by a 5-yr period (Fig. 6). This difference is
due to year to year variations in the distribution of annual precip-
itation and to the equilibration of water within the modeled pro-
file.
Each separate design component parameter was simulated in-
itially for the full 20-yr time period and for a 5-yr period. Sub-
sequent simulations of the same parameter used the 5-yr dataset
and then were scaled up to a 20-yr precipitation period by the use
of the appropriate scaling factor. The results obtained by this pro-
cedure were randomly checked against computations obtained
with full 20-yr simulations; no disagreements were observed.
When extensive modeling is anticipated, this procedure can be a
significant tune-saving step.

Response to Top Slope Modification
The local availability of cobble-sized aggregate, the size of the
disposal area and the cost of redesigning the drainage channels
suggested that any changes to the specified 5:1 side slopes were in-
feasible. This was not a significant limitation to the design opti-
mization study. The flatter top slopes account for the largest por-
tion of the cover's surface area, while the steep side slopes encour-
age runoff of incident precipitation. Thus the largest contribu-
tion to infiltration of precipitation and the resulting generation
of leachate is through the top of the disposal cell, not the sides.
During the initial simulations, the top slope was set at 3%. The
a
J?
x
_3
U.
E
a
«rf
c
o
o
5 10 15
Duration of Simulation lyears]

Figure 6
Containment Flux Begins to Stabilize After Model Uses
10 Years of Daily Precipitation R
20
1.0
». .5
c
o
o
5 10 15
Degree of Slope (%)

Figure 7
Effects of Top Slope on Contaminant Flux
original design specifications allow a variation of this parameter
from 2 to 4%. However, since steeper slopes promote runoff and
help to decrease infiltration, this design element was thought to be
critical to better performance. Simulations of steeper top slopes,
up to a maximum of 12%, showed significantly lower percola-
tion rates through the base of the impoundment (Fig. 7). At 12%
top slope, the percolation volume was reduced by 40%. At 12%
top slope was considered the maximum feasible slope due to
long-term stability and erosion considerations, so no steeper top
slopes were evaluated.

Significance of the Evaporative Zone Depth
The proposed cover would be protected from the effects of
long-term erosion and denudation by a top layer of cobble sized
gravel. While this material certainly will minimize erosion, it has
a deleterious effect on the infiltration of incident precipitation.
Gravel mulches have long been used to minimize evaporation and
enhance infiltration in agricultural applications. For example,
Hillel2 states: "Gravel mulching is an age-old method and can be
very effective in water conservation (both in enhancing infiltra-
tion and in suppressing evaporation) even in layers as thin as
5-10 mm."
The proposed DOE cover initially was modeled with an evap-
orative zone depth of 4 in., which the HELP model documenta-
tion suggests is appropriate for bare ground.' However, recent
validation tests for the HELP model4 indicate that this value may
be much too low for semi-arid conditions, such as found at this
site. Furthermore, the potential for woody vegetation, such as
pinyons and junipers, becoming established on the gravel sur-
MODELLING GROUNDWATER & SURFACE WATER 61
-------
face of this cover within its design life is considered realistic by
several experts.' Such vegetation would increase the evaporative
depth.
Consequently the evaporative depth was changed through a ser-
ies of four increments until a maximum depth of 24 in. was
reached. As would be expected, increases in the depth of evapor-
ation cause decreases in the volume percolating through the base.
As shown in Fig. 8, this volume is decreased approximately 0.1
gal/min for each 4-in. increase in the thickness of the evaporative
zone. Thus, when the maximum 2-ft evaporative zone depth is
used, the volume is approximately halved.

Permeability of the Clay Barrier Layer
The permeability of any clay barrier layer will restrict the
amount of precipitation that infiltrates into the underlying tail-
ings. The original DOE cover design specified a permeability
of 1 x 10~7 cm/sec, which approaches the limit of what is achiev-
able under present day construction practice. For instance, Dan-
iel' has shown that the in situ or "as built" permeability of com-
pacted clay liners is often 10 to several thousand times higher
than predicted from current laboratory testing procedures. His
data show that field permeabilities are rarely lower than 1 x 10'6
cm/sec.'
a.
E
a
o
U
0 10 20
Evaporative Zone Depth (inches)

Figure 8
Effects of Evaporative Zone Depth on Contaminant Flux
Three simulations were undertaken to analyze the effects of
changing the cover permeability. The permeability was changed
by one order of magnitude for each simulation. As shown by
Fig. 9, the percolation volume increases from 1 gal/min when the
cover permeability is 1 x 10~7 cm/sec to 8 gal/min when the
permeability is 1 x 10"5 cm/sec. This result clearly demonstrates
the importance of verifying that the design permeability is actual-
ly achieved during construction.

Length of the Drainage Path
The HELP model currently does not permit the analysis of
changes in surface runoff collection systems. It does allow for the
modeling of changes in lateral flow distances to drainage collec-
tor systems within the cover profile. Accordingly, the HELP
model was used to evaluate how changes in lateral drainage dis-
tances, within a lateral drainage layer, affect the percolation vol-
umes from the base of the impoundment. In this simulation, the
sandy filter layer between the compacted clay and gravel layers
in the cover (Fig. 5) was modeled as the lateral drain.
Drainage distances ranging from 50 to 200 ft were used to eval-
uate the changes in the predicted contaminant flux as the length
of these drainage paths varied. The estimated volumes from these
simulations are shown in Fig. 10. For this particular site, when
the length of the drainage path was decreased to 50 ft, the flux

62 MODELLING GROUNDWATER & SURFACE WATER
volume was reduced by approximately 25%.

COMPOSITE SIMULATION INCORPORATING
THE ABOVE MODIFICATIONS
For the final simulation the various factors that improved the
performance of the cover were combined. This included a 50-ft
drainage path, a 24-in. evaporative zone depth and a 12% top
slope. The HELP model estimated that a cover with these mod-
ifications will produce a percolation flux from the base of the
impoundment of 0.14 gal/min; an 86% decrease over the orig-
inally estimated 1 gal/min flux rate. Due to the inherent uncer-
tainties in the modeling process, such comparisons give only a
relative measure of improvement and should not be considered
as absolute values. Nevertheless, these changes, or modifications
thereof to the extent reasonably achievable, will be recommended
for incorporation into the final design.
These modifications may not be totally obtainable in the field,
For example, without the benefit of vegetation on the surface of
this cover, the original DOE design may result in an evapora-
tive zone depth of much less than the 18 to 24 in. expected for this
area. This result is due mostly to the presence of the gravel cover,
which will tend to decrease evaporative losses while simultan-
eously increasing infiltration of incident precipitation.' Currently
there is strong opposition to a vegetative cover at any of the Col-
orado LJMTRA Project locations, even though Oregon and Penn-
sylvania sites were conpleted with this type of system. The impacts
posed by vegetation on the covers of these Colorado sites remain
under review.
I 10
£
o
o
o
10
-7
10
-6
10
-3
Cover Permeability
10
-4
Figure 9
Effects of Barrier Permeability Changes on Contaminant Flux
1.0
E
a.
E
o
c
o
o
.5
50 100 150
Drainage Length (feet)
Figure 10
Effects of Drain Length on Contaminant Flux
200
-------
CONCLUSIONS
The results of any model simulation should never be considered
absolute. Nevertheless, the HELP model is a valuable aid in
assessing design alternatives.
Site-specific data are critical in achieving any reasonable meas-
ure of predictive reliability. The HELP model contains default
data values for both climatological and cover material charac-
terization. These are useful in a preliminary design review. How-
ever, the limitations of such data must be clearly understood by
the users of the model. For example, the model's default perme-
ability values for a silty clay loam vary over three orders of mag-
nitude. Thus, the user of the model must be able to define the
site materials with a degree of precision greater than represented
in the default soils data.
The precision of model simulations increases when longer peri-
ods of precipitation data are used. The HELP model can use up
to 20 yr of such data to perform a simulation. However, the time
required to complete a simulation is proportional to the length of
the precipitation record. When many simulations are required, it
is possible to develop a scaling factor which allows conversion of
simulation results obtained using shorter precipitation periods to
those resulting from using the full 20-yr period allowed by the
model. This procedure can be a valuable time-saving steps when
many simulations are anticipated.
The chief benefit hi using HELP during the design assessment
process is its ability to make comparisons between design al-
ternatives or modifications. Used in this capacity, the HELP
model can be a valuable tool during the actual design process.
The relative accuracy of such comparisons is always better than
the absolute reliability of the estimated values.
ACKNOWLEDGEMENTS
The following people have offered significant support, assis-
tance or guidance during this study: Rahe Junge, Mark Sniff and
Robert Talbott.
REFERENCES
1. Daniel, D.E., "Predicting Hydraulic Conductivity of Clay Liners,"
/. Geotech. Eng., ASCE, 110, 1984, pp. 285-300.
2. Hillel, D., Applications of Soil Physics, Academic Press, New York,
NY, 1980.
3. Lynn, D., District Conservationist, personal communication from
District Conservation Office, Soil Conservation Service, USDA,
Durango, CO, Aug. 1986.
4. Payton, L. and Schroeder, P.R., "Verification of the Hydrologic
Evaluation of Landfill Performance (HELP) Model Using Field
Data," Course Notes, Part 3, HELP Model Short Course, Colorado
School of Mines, Golden, CO, Dec. 1986.
5. Schroeder, P.R., Morgan, J.M., Wolski, T.M. and Gibson, A.C.,
The Hydrologic Evaluation of Landfill Performance (HELP) Model,
Vol. I Users Guide for Version 1, U.S. EPA Report Number EPA/
530-SW-84-009; available from NTIS as report number PB 85-
100840, 1984.
6. U.S. DOE, "Remedial Actions at the Former Vanadium Corporation
of America Uranium Mill Site, Durango, La Plata County, Col-
orado," Draft Environmental Impact Statement DOE/EIS-0111D
(2 vols), prepared by the U.S. DOE, UMTRA Project Office, Al-
buquerque Operations Office, Albuquerque, NM, 1984.
7. U.S. DOE, "Remedial Actions at the Former Vanadium Corpo-
ration of America Uranium Mill Site, Durango, La Plata County,
Colorado," Final Environmental Impact Statement DOE/EIS-
011 IF (2 vols), prepared by the U.S. DOE, UMTRA Project Office,
Albuquerque Operations Office, Albuquerque, NM, 1985.
8. U.S. DOE, "Disposal Site Characterization Report for the Alternate
Disposal Site in Bodo Canyon near Durango, Colorado," Draft,
prepared by the Technical Assistance Contractor (Jacobs-Weston
Team), for the U.S. DOE, UMTRA Project Office, Albuquerque
Operations Office, Albuquerque, NM, 1985.
9. Weeks, E.J., personal communication from U.S. Geological Survey,
Water Resources Division, Denver, CO, Aug. 1986.
10. Wright, W. and Turner, A.K., "A Combined Modeling Program
for Evaluating the Cover Design at a Uranium Mill Tailings Disposal
Site," Proc. Use of Models to Solve Groundwater Problems,
NWWA, Denver, CO, 1987, in press.
MODELLING GROUNDWATER & SURFACE WATER 63
-------
Transport of Mercury in the
North Fork Holston River

David R. Cogley, Ph.D.
Neil Ram, Ph.D.
Alliance Technologies Corporation
Bedford, Massachusetts
ABSTRACT
An electrolytic chlor-alkali plant operating from 1951 to 1972
at a site on the North Fork Holston River in Virginia resulted in
high concentrations of mercury in river sediment and in a waste
disposal pond. The quantity of mercury present in the waste pond
has been estimated to total more than 50,000 Ib with concentra-
tions in the sediment up to 120 ppm. Sediment sampling shows
concentrations of mercury in the river bed exceeding 1 ppm at
locations up to 80 miles downstream from the chlor-alkali plant.
Sediment and fish sampling data show that sediment concentra-
tions of 0.5 ppm mercury are highly correlated with an FDA
limit of 1.0 ppm methylmercury in (bottom feeding) fish.
Column and batch leachability studies were performed on rep-
resentative samples of waste from the waste pond. The results in-
dicate that the percentage of mercury that can potentially be
leached from the waste is at least 10%, in contrast to the 2% re-
ported in earlier studies. The results support the conclusion that
mercury can be expected to leach from the pond for periods great-
ly exceeding 100 yr.
The U.S. EPA model TOXIWASP has been calibrated with
total suspended solids data available from USGS. The calibrated
model was used to calculate the time required for sediment-mer-
cury concentrations to decrease below 0.5 ppm for several levels
of mercury flux from the waste pond. Model simulations indicate
that at present discharge rates, mercury concentrations in sedi-
ment will remain above 0.5 ppm for extended time periods (14 yr
to greater than 20 yr) over the stretch of river extending at least
20 mi downstream from the chlor-alkali plant.

INTRODUCTION

History
Mercury discharged from an Oh'n Corporation chlor-alkali
plant contaminates the North Fork Holston River (NFHR) lo-
cated in Piney River, Virginia. The mercury contamination be-
gan in 1951. Mercury used in the chlorine-caustic process was re-
leased from the chlor-alkali plant into the process wastes and on-
to the plant grounds. A surface impoundment, designated Waste
Pond 5, was used to dispose of waste sludges from the chlor-
alkali processes. In 1963, another impoundment, designated
Waste Pond 6, was constructed to receive overflow from Waste
Pond 5. No wastes containing mercury supposedly were dumped
into Waste Pond 6, but structural components of the old chlor-
alkali plant reportedly were buried at the eastern edge of the
pond. In 1972, Olin ceased operation of the chlor-alkali plant
and initiated a series of actions at the site to decrease the flux of
mercury into the NFHR. These remedial cleanup actions in-
cluded demolition of the chlorine plant, drainage of Waste Ponds
5 and 6, river bank erosion control measures, dredging of a por-
tion of the river sediments, and capping of the site. While these
remedial efforts considerably reduced the quantity of mercury be-
ing discharged to the NFHR, mercury from the Saltville site con-
tinues to be discharged at an average rate of approximately
10 g/day.
Air, soils, surface water, sediments and biota from the NFHR
have been sampled during the past 15 yr, and mercury has been
detected in all media. Air monitoring has detected insignificant
levels of paniculate mercury; however, elevated mercury vapor
concentrations were detected in June, 1983, while remedial activ-
ities were being conducted at the site. Surface water sampling by
Oak Ridge National Laboratory in 1975 detected insignificant
levels (less than 0.2 /tg/1) of dissolved mercury in filtered water,
but significant (greater than 1 ppm) levels of mercury on sus-
pended particulates in the river. Concentrations of mercury in the
Waste Pond 5 effluent ranged from 10 to 120 jtg/1. Low levels of
cadmium, lead and arsenic (less than 10 /ig/1) also were found in
Waste Pond 5 effluent. The results of waste sampling and analy-
sis led to the conclusion that about 53,000 Ib of mercury are con-
tained in Waste Pond 5 with 92% (49,000 Ib) contained in the
upper 17.5 ft. The mercury is concentrated in the west end, the
northeast corner and the far east end of the pond. Mercury in the
upper 17.5 ft of these areas (29 ac total) represents about 69% of
the total mercury in the pond. Sediment sampling in the NFHR
showed that significant levels (above 1 ppm) are present at most
stations up to 80 river miles downstream from the Saltville Site.
The majority of fish samples collected below the Saltville site con-
tained edible-portion mercury concentrations greater than 1.0
ppm.
Mercury compounds present in the sediment and water column
of the North Fork Holston River (NFHR), the Holston River
(HR) and the Cherokee Reservoir have contaminated fish and
other biota. Fish data for the NFHR indicate that, at various
locations and times, methylmercury concentrations have exceeded
the Food and Drug Administration action level of 1 ppm methyl-
mercury. The concentrations of methylmercury in fish are be-
lieved to be highly dependent on historical and present concen-
trations of mercury in river and reservoir sediment.
The principal source of continuing mercury flux into the North
Fork of the Holston River is via seepage from the Waste Pond 5
outfall. Mercury flux into the NFHR from groundwater flow
from Waste Pond 5 cannot be quantified due to the absence of
analytical data. However, hydrologic measurements have shown
that the majority of groundwater flow through Waste Pond 5 h
discharged through the outfall. An additional source of mercury
is discharge of contaminated groundwater from beneath the
former chlor-alkali plant. However, the flux from this source \t
insignificant compared to the discharge from the Waste Pond 5
outfall.
64 MODELLING GROUNDWATER & SURFACE WATER
-------
This paper presents the results of a modeling effort which Alli-
ance conducted for the Saltville site examining mercury sediment
concentrations in the NFHR and corresponding fish tissue mer-
cury levels. The goal was to predict the time required for sedi-
ment mercury concentrations to decrease to less than 0.5 ppm
mercury and to achieve a corresponding fish concentration of less
than 1 ppm methylmercury. The model was used to study mer-
cury dynamics under a "No-Action" scenario, as well as for re-
medial actions which decrease the amount of mercury being dis-
charged into the NFHR. Model results indicate that even if mer-
cury discharge were to be completely eliminated, mercury laden
sediments would continue to represent a risk to human health
and the environment for about 10 yr. Existing levels of mer-
cury discharge would increase this estimated time to 14 yr.

Important Factors Governing Mercury Transport
in NFHR
In order to model the fate and transport of mercury in the
NFHR, one must first understand the equilibrium chemistry
which exists in Waste Pond 5 representing the major mercury
reservoir to the river. Alliance calculated equilibrium concentra-
tions of soluble mercury species in Waste Pond 5 sludge with the
aid of a pH-redox-equilibrium equations model4 developed by
staff of the U.S. Geological Survey. Thermodynamic data were
obtained from Faust and Aly.2 The model results, which com-
pared well to observed mercury concentrations from the Waste
Pond 5 outfall, indicated that mercury dissolution is limited by
the percolation or infiltration of oxygenated water into the Waste
Pond 5 sludge. These results indicate that over very long time
periods, essentially all of the mercury eventually will be leached
from the waste pond.
This hypothesis was confirmed by column and batch leach-
ability studies performed by Alliance on representative samples
of Waste Pond 5 sludge. The results indicated that at least 10%
(equal to about 5000 Ib of mercury) of the mercury could poten-
tially be leached from Waste Pond 5. The data clearly imply that,
in the absence of site remedial action, mercury likely will continue
to leach into the NFHR for centuries into the future.
Equilibrium concentrations of soluble mercury species also
were calculated for mercury entering the NFHR from the Waste
Pond 5 outfall. Results were consistent with the precipitation of
sparingly soluble hydroxide precipitates and with sorption of mer-
cury to river bottom sediments. River water and river sediment
monitoring data correlated well with modeled concentrations in
the sense that most of the mercury was found in sediments and
very little mercury was found in the water column of the NFHR.
Microbiological transformation of metallic and inorganic
forms of mercury to the organic compounds methyl and di-
methylmercury is thought to be a major factor in the environmen-
tal fate of mercury in the river. Alkylated mercury compounds,
especially methylmercury, are strongly bioaccumulated in biota,
forming over 90% of the mercury in fish tissues.
The river basin to be characterized extends from the old chlor-
alkali plant outfall located at NFHR River Mile 83 (NFHR RM83)
downstream to the Cherokee Reservoir Dam, a distance of 172
mi. The highest concentrations of mercury in sediment occur in
the 83-mi reach extending from the old chlor-alkali plant to the
confluence with the South Fork Holston River. The combined
flows of the NFHR and SFHR provide an approximately 8-fold
dilution of water relative to flows at the Saltville gaging station
near the old chlor-alkali plant. By the time the water reaches the
Cherokee Reservoir Dam, the dilution factor is approximately 15.
Mercury transport in the NFHR is believed to occur through
transport of suspended solids onto which mercury compounds
are sorbed (or coprecipitated). Transport of methylmercury is
thought to be a minor process relative to the transport of other
forms of mercury. The important physical factors controlling
mercury transport in the NFHR are those which control settling
of suspended solids and those which control erosion of bed sedi-
ment, including:
• Sediment particle size distribution
• Average water depth
• Water flow velocity
• Water flow distance along the river course
The first three factors are affected by the river geometry and
typical flow conditions. As a practical matter, one requires data
on sediment particle size distribution, river geometry, average
daily flow at several stations for the period 1951 to the present
and total suspended solids (TSS) levels as a function of flow.

MODELING
Model Selection
A primary objective of our assessment of the NFHR mercury
contamination was to estimate the period of time it would take
for mercury concentrations in sediment to decrease to less than
0.5 ppm. A concentration of 0.5 ppm in sediment was strongly
correlated with the Food and Drug Administration (FDA) action
level for mercury in fish, 1 ppm.
Modeling mercury transport requires the ability to simulate the
important physical processes governing mercury transport. Of the
available documented models, the U.S. EPA's Chemical Trans-
port and Fate Model TOXIWASP1 was selected. TOXIWASP
is a dynamic model for simulating the transport and fate of toxic
chemicals in water bodies. TOXIWASP combines the kinetic
structure adapted from the EXposure Analysis Modeling Sys-
tem (EXAMS) with the transport framework provided by the
Water Analysis Simulation Program (WASP) along with simple
sediment balance algorithms. TOXIWASP is capable of simu-
lating variable chemical degradation rates (hydrolysis, biolysis,
photolysis, oxidation and volatilization) from chemical charac-
teristics of a compound and the environmental parameters of the
aquatic system. For the present application, TOXIWASP was
used to simulate sorption onto sediments but not to model degra-
dation or chemical transformation.
TOXIWASP simulates a river as a series of water and sedi-
ment segments. TOXIWASP calculates total sediment and chem-
ical concentrations explicitly for every time step and every seg-
ment including water, sediment pore water and sediment bed.
TOXIWASP models the effects (on sediment) of advection, dis-
persion, mass loading, settling, scouring and sediment burial
within the river bed. For chemical concentrations, TOXIWASP
models the same processes plus degradation, sediment-water dis-
persion and percolation. Within any particular segment, concen-
trations are assumed to be uniform.

Model Setup
Selecting model parameters and the length of a simulation are,
in part, affected by the speed with which the simulation executes.
TOXIWASP is available in a mainframe computer version and in
a microcomputer version. Alliance performed modeling runs on
an 8-megahertz AT&T Model 6300 microcomputer with 640 kilo-
bytes of system memory, an 8087 numeric co-processor and a 2-
megabyte "RAM disk." Typical eight-segment TOXIWASP runs
took about 1-min of run time/day of simulation for a single flow
condition.
Conceptually, an ideal river simulation might include many
1-mi segments starting at the Waste Pond 5 outfall and ending
at the Cherokee Reservoir Dam. One would be able to precisely
MODELLING GROUNDWATER & SURFACE WATER 65
-------
model the transport of mercury from the source down to, and
through the reservoir. For two reasons, this approach was not
adopted for the NFHR. First, the microcomputer version of
TOXIWASP is limited to 80 segments; 40 water segments and
40 sediment segments. Second, U.S. EPA personnel indicated
that model accuracy was optimum when water segments were of
equal volume (not equal length). A glance at the USGS topo-
graphic quadrangles for the NFHR and the Holston River down
to the Cherokee Reservoir Dam together with stream discharge
data indicates that if the entire 170-mi reach is modeled, the first
(upstream) water segment would include all of the NFHR. Since
essentially all of the water quality data and sediment data apply
to the NFHR, it was decided that only the NFHR would be mod-
eled. The quantity of mercury advected from the NFHR to the
Holston River also would be computed.
A series of preliminary modeling runs indicated that the mer-
cury-contaminated stretch of river extending from the old chlor-
alkali plant downstream approximately 20 mi would be expected
to remain at high mercury concentration levels longer than down-
stream areas. In the final modeling runs, four water segments
each 5.7-mi long were modeled: segment 1 extended downstream
from the chlor-alkali plant and including the Waste Pond 5 out-
fall; segment 3 centered at the River Mile 77 (RM77) monitoring
point; segment 5 centered at the RM72 monitoring point; and
segment 7 centered midway between the RM72 and RM59 moni-
toring points. Each odd-numbered water segment was paired with
an even-numbered sediment segment.
Sediment particle settling rates and scour rates were estimated
based on particle size distribution data and power law functions
for total suspended sediment concentration data for the NFHR as
reported by Milligan (1979). Sediment scour rates were estab-
lished through an iterative process for the 30- and 70-percentile
flow rates, 97 and 327 cfs at Saltville. The power law functions
were for the Saltville and Gate City monitoring stations. The
functions were:
Saltville—TSS (ppm) = 70.6 (Q/A)0.61
Gate City—TSS (ppm) = 88.4 (Q/A)°-56
(1)
(2)
ing the period 1970 through 1984, it is known that the river dis-
charge is less than or equal to 1100 ft'/sec, 95% of the time. A
weighted average of these four runs was then taken to determine
the long-term average transport rates for mercury. Weighting
factors were 0.1, 0.4, 0.4 and 0.1 for the 5-, 30-, 70- and 95-per-
centile flow conditions, respectively.
Each of the TOXIWASP simulations was run for 20 days.
From each 20-day run, the long-term, average equivalent first-
order rate constant for changes in sediment mercury concentra-
tion was obtained for the first river reach. Also, the ratio of water
column mercury concentration to sediment mercury concentra-
tion was obtained for each river reach. These parameters were
used as input to a detailed mass balance from which mercury con-
centrations were extrapolated for a period of 20 yr. Calculations
were performed for three mercury load conditions: 0 g/day, 10 g/
day and 27 g/day discharged from Waste Pond 5. A load of
0 g/day was modeled to obtain an estimate of the fastest reduc-
tions in mercury concentrations possible under natural river flow
conditions (i.e., complete site remediation and a correspond-
ing total cessation of mercury discharges to the NFHR). Loads
of 10 and 27 g/day correspond approximately to the Pond 5 dis-
charges from the time of construction of the diversion ditch con-
sidering either: (1) only those time periods when the ditch was
functional or (2) the entire time period including those days when
the diversion ditch had failed.

Model Results
The 0 g of mercury/day scenario (Fig. 1) establishes the fastest
rate of dilution of mercury expected for long-term average flow
conditions for the NFHR. Dilution of sediment mercury by un-
contaminated sediment from upstream and scouring of sediment
from contaminated reaches of the river would be faster if NFHR
flows were higher than normal and slower if NFHR flows were
lower than normal. The 10 g mercury/day scenario corresponds
to the historical discharge rate for the post-ditch-construction
time period but does not include the time period when the diver-
sion ditch failed. The 27 g mercury /day scenario corresponds to
the historical discharge for the post-ditch-construction time
period including the time period when the diversion ditch failed
(Fig. 2). Results are summarized in Table 1.
where Q is the mean daily stream flow discharge in ft'/sec, and A
is the drainage area (222 mi2 for Saltville and 672 mi2 for Gate
City). Simulation runs for this calibration covered the river reach
from Saltville to Gate City in four water segments and four sedi-
ment segments.
Initial sediment mercury concentrations were derived from a
combination of sources. For segment 2 (the sediment segment
corresponding to water column segment 1), the mercury concen-
tration was set at 6.8 ppm, twice the segment 4 concentration, re-
flecting the trend shown in Fig. 3 of Milligan's paper.3 For seg-
ment 4, a concentration of 3.4 ppm was selected based on a linear
regression on RM77 monitoring data for the period 1977 through
1984. A linear regression technique was used to compensate for
the large year-to-year variations in observed mercury concentra-
tions. The intent was to obtain the best estimate of the present
(1986) concentration of mercury in sediment. For segment 6, a
concentration of 2.1 ppm was selected based on a linear regres-
sion on RM72 monitoring data. For segment 8, a concentration
of 1.9 ppm was selected based on an average of linear regressions
on the RM72 and RM59 data.
The selected river reaches were modeled under four flow con-
ditions: 42, 97, 327 and 1100 ft'/sec at Saltville, corresponding
to the 5-, 30-, 70- and 95-percentile flow conditions. For ex-
ample, from historical river discharge data for the NFHR cover-
MERCURY LOAD -0 grams /day
10 12 14 16 IB 20
2 4
Figure 1
Mercury Concentrations in Segments 2-8
Zero grams of mercury p«r day discharged from Waste Pond 5.
(River Mile) indicates the center of each 5.7-mile river segment.
66 MODELLING GROUNDWATER & SURFACE WATER
-------
MERCURY LOAD - 27 grams/day
NFHR, Segment 6
02 46 8 10 12 14 16 IS 20
Figure 2
Mercury Concentrations in Segments 2-8
27 grams of mercury per day discharged from Waste Pond 5. RM (River
Mile) indicates the center of each 5.7-mile river segment.
1 -
0 8
..07
0.
anfi
-0.6-
ae
•=> n «
a:
u
S 0 4 -
_j
? °-s
o
'"02

O-

\
\
\
\

V
\
\
\

\
*v
^

/
H
\

M
<

i'°

^
-10

\
7

n>y

-
gn

daj

-~—

— — .

me/day

—
i — o v
II

ami

/20
20
16
NFHR, Segment 4

6
£ 07

•3. °'6 -
0
£ 0 5 -
S
£
O O 3 -
fc. °-3

O -

/
— c

\
\

grc

^
' \

\
^

mi /day

\
N

/T^UL
L

^
L

*^.

^
gra

^
ms/

-10 grams A

^~-
day

— —

^—

— ,
24 6 8 10 12 14 16 18 20
TIME.yaart

Figure 3
Mercury concentrations in Segment 4 (RM77) for 0, 10 and 27 grams of
mercury per day discharged from Waste Pond 5. RM (River Mile) indi-
cates the center of a 5.7-mile river segment.
0.9
0.7
0.6
=> 0.5
0.4
0.3
0.2
\
\
\
27 grami/da '
-10 grami/doy
0 grami/day
0 2 4 6 8 10 12 14 16 18 20
TIME,y«ar«

Figure 5
Mercury concentrations in Segment 8 (RM66) for 0, 10 and 27 grams of
mercury per day discharged from Waste Pond 5. RM (River Mile) indi-
cates the center of a 5.7-mile river segment.
Natural dilution of mercury in sediment would require approx-
imately 10 yr for the four river segments to reach a concentration
of 0.5 ppm mercury if there were no discharge of mercury from
Pond 5 (Fig. 1). A concentration of 0.5 ppm is the mercury con-
centration in sediment which corresponds to a 1 ppm mercury
concentration in fish. At a mercury discharge rate of 27 g/day
(Fig. 2), segment 2 would still reach 0.5 ppm in approximately
10 yr, but segments 6 and 8 would require approximately 20 yr.
Segment 4 (centered on RM77) would not reach a concentra-
tion of 0.5 ppm mercury within 20 yr.
Segment 4 includes the river reach from river mile 80 (RM80)
to RM73. It is the river segment downstream of the Waste Pond
5 outfall which is located near RM82. A closer look at the seg-
ment 4 response to mercury load (Fig. 3) shows that recent histor-
ical rates of mercury discharge from Waste Pond 5 have a major
influence on the rate of change of mercury concentration in the
river sediment. At mercury discharge rates in the range of 0 to 10
MODELLING GROUNDWATER & SURFACE WATER 67
-------
g/day from Waste Pond 5, it would take 9 to 14 yr for the mer-
cury concentrations to decrease to 0.5 ppm. At higher mercury
discharge rates (e.g., 27 g/day), mercury concentrations in sedi-
ment would be expected to remain well above 0.5 ppm for periods
exceeding 20 yr.
Sedment 6 (Fig. 4) and segment 8 (Fig. 5) are the next two seg-
ments downstream from segment 4. They are not affected as
strongly as segment 4 by mercury discharges from Waste Pond 5,
but they are affected by discharges from Waste Pond 5, by mer-
cury eroded from segment 2 and by mercury eroded from segment
4. Segments 6 and 8 show gradual decreases to 0.5 ppm mercury
within 20 yr, even at discharge rates as high as 27 g/day.

Accuracy of Model Predictions
The accuracy of model predictions is a function of input data
including: mercury concentrations in sediment, river flow, par-
ticle settling rates, sediment scour rates and river geometry data.
In addition, the model output must be understood in terms of
average concentrations over the river reaches modeled.
The most serious data gap affecting the accuracy of model pre-
dictions is the lack of detailed data concerning mercury concen-
trations in sediment for the portion of the river extending from
RM83 at the old chlor-alkali plant past the Waste Pond 5 dis-
charge point down to the monitoring point at RM77. If the con-
centration of mercury in sediment is higher than assumed for
segment 2 (RM86 to RM80), the rates of decrease in mercury con-
centrations in segments 4, 6 and 8 will be slower than calculated
under the present assumptions.

CONCLUSIONS
Mercury from the Saltville Waste Disposal Site has been dis-
charged and continues to be discharged into the NFHR. While
the Olin chlor-alkali plant was in operation, and before the plant
area was capped, unquantified amounts of mercury were released
from this area to the NFHR, resulting in a reservoir of mercury
in the river sediments. At present, the primary input of mercury
to the NFHR is from the Waste Pond 5 outfall, which discharges
both surface runoff and groundwater seepage from the pond.
Mercury-contaminated groundwater from the former chlor-alkali
plant area and groundwater flow from Waste Pond 5 that is not
intercepted by the outfall also discharge to the NFHR. Alliance
has concluded, however, that their contribution to the total mer-
cury discharge into the NFHR is insignificant relative to the Pond
5 outfall. Mercury discharge from Waste Pond 6 also is relatively
insignificant. Thus, the primary risk associated with mercury dis-
charge into the NFHR is from continued mercury flux from
Waste Pond 5.
Mercury discharge from Waste Pond 5 is expected to continue
indefinitely (more than 1,000 yr) based upon laboratory column
and batch leachability studies conducted by Alliance. These stud-
ies show that at least 10% of the mercury in Waste Pond 5 wiD
leach into the NFHR.
Model simulations indicate that, at present discharge rates,
mercury concentrations in sediment will remain above 0.5 ppm
for extended time periods over the stretch of river extending at
least 20 mi downstream from the chlor-alkali plant. The calcu-
lated time to reach 0.5 ppm for the 0 g/day discharge scenario is
approximately 10 yr. The actual time could be shorter or longer,
depending on the NFHR flows and the actual quantity of mercury
present in the search of river extending from the chlor-alkali
plant to the monitoring point at RM77.
The impact of mercury discharges from Waste Pond 5 is signif-
icant. Discharge of 10 g/day of mercury increases the calculated
time to reach 0.5 ppm from 10 to 14 yr for segment 4. Discharge
of 27 g/day of mercury increases the calculated tune to reach
0.5 ppm to greater than 20 yr. Actual times could be shorter or
longer, depending on the NFHR flows and the actual quantity of
mercury present in the reach of river extending from the chlor-
alkali plant to the monitoring point at RM77.

ACKNOWLEDGEMENT
This paper presents work performed by Alliance Technologies
Corporation under subcontract to NUS Corporation (subcontract
No. Z0830913, Work Assignment No. 1) funded through EPA
Contract No. 68-01-6699 to NUS Corporation.

REFERENCES
1. Ambrose, R.B., Jr., Hill, S.I. and Mulkey, L.A., "User's Manual
for the Chemical Transport and Fate Model TOXIWASP, Version
1," EPA-600/83-005. U.S. EPA, Athens, GA, 1983.
2. Faust, S.D. and Aly, O.M., Chemistry of Natural Waters, Ann Arbor
Science Publishers, Inc., Ann Arbor, MI, 1981.
3. Milligan, J.D., Betson, R.P., Bales, J. and Bittman, R.M., "Trans-
port and Partitioning of Mercury in the Sediment of the North Fork
Holston River—1978, WR-57-1-1." Tennessee Valley Authority,
Division of Environmental Planning, Water Quality and Ecology
Branch, 1979.
4. Parkhurst, D.L., Thorstenson, D.C. and Plummer, L.N.,
"PHREEQE—A Computer Program for Oeochemical Calculations,"
U.S. Geological Survey, Water Resources Division, 12201 Sunrise
Valley Drive, Reston, VA, Nov. 1980.
68 MODELLING GROUNDWATER & SURFACE WATER
-------
Geologic and Hydrogeologic Characterization
Of a Hazardous Waste Disposal Site
Arlington, Oregon
Stephen M. Testa
Engineering Enterprises, Inc.
Long Beach, California
Frederick G. Wolf
Parametrix, Inc.
Bellevue, Washington
ABSTRACT
Chem-Security Systems, Inc. owns and operates a major hazar-
dous waste disposal facility near Arlington, Oregon, serving the
Pacific Northwest, Canada and Alaska. The Arlington site main-
tains numerous favorable environmental characteristics for siting
a hazardous waste disposal facility which include:
Semi-arid climate
Low precipitation
High evapotranspiration
Low rate of infiltration
Deep water table (uppermost zone of saturation)
Thick vadose zone
Abundant low permeable and moisture deficient soils
Low population density
Lack of nearby surface water bodies
The risk of contamination as a result of potential leakage from
a waste management unit via primary pathways to surface water,
groundwater or by direct contact and/or ingestion is thus
reasonably low. However, these same characteristics which make
the site most suitable for hazardous waste disposal often conflict
with (1) the demonstration of the groundwater monitoring sys-
tem's ability to adequately perform immediate leak detection
monitoring as mandated under RCRA, 40 CFR Part 264, Part F,
and (2) the level of demonstration required for the site to be
"properly characterized" which risks the integrity of some of
these characteristics.
Such favorable environmental characteristics warrant con-
sideration of other site-specific factors to develop a contaminant
detection groundwater monitoring system which include:
• Heterogeneous mixture of waste types
• Complex stratigraphic depositional environments
• Identification of hydrostratigraphic units
• Complex groundwater movement regimes
• Infeasibility to use conventional unsaturated zone monitoring
techniques which reflect low permeable and moisture defici-
ent soils
• Level of demonstration versus level of risk to site integrity

INTRODUCTION
During the period from December 1983 through November
1986, Chem-Security Systems, Inc. (CSSI), in support of its Part
B Application under RCRA, embarked on an extensive field in-
vestigative program in order to characterize existing site geologic
and hydrogeologic conditions.1 This program has included:
• Detailed geologic mapping including trenching to evaluate and
characterize the present faults
• Drilling over 102 boreholes and installing over 120 wells/
piezometers to depths ranging to 363 ft below ground surface
to evaluate subsurface hydrogeologic conditions and the
groundwater flow regime
• Collecting undisturbed samples for detailed laboratory evalua-
tion and testing
• Performing both pumping/packer and slug tests to assess in
situ hydraulic characteristics and possible intercommunication
between the uppermost aquifer and the underlying basalts
• Performing both surface and borehole geophysics
• Performing analytical testing of groundwater samples to evalu-
ate the groundwater quality, including the presence of tritium
to assess recent recharge

The results of the site characterization, specifically the site
geology and the existing groundwater recharge-discharge regime
are discussed in this paper. The foregoing study is then used to
assess the site's suitability for hazardous waste disposal and to
specify a groundwater detection system.

SITE DESCRIPTION
The Arlington facility is located 6.5 mi south of the Columbia
River and 7.5 mi southwest of the town of Arlington in Gilliam
County, Oregon (Fig. 1). The facility site is situated on a 640-acre
parcel, of which 320 acres (eastern tract property) currently are
used for waste management operations (Fig. 2). The property is
bounded on the south by the east-west trending Alkali Canyon at
an elevation of approximately 700 ft (site datum). The upland
plateau is at an elevation of approximately 145 ft. Waste manage-
ment activities are limited to that area above 920 ft on the eastern
tract property. In the southern portion of the property bounded
by Alkali Canyon, the relief between the valley floor and the
upland plateau is approximately 280 ft.
Adjacent tracts to the north, east, south and west of the site are
owned by others. Portions of the tract to the east of the Chem-
Security site are under cultivation and are irrigated. The facility is
remote from any residential, commercial or industrial develop-
ments. The nearest residence is approximately 1 mi by road west
of the western site boundary.
MODELLING GROUNDWATER & SURFACE WATER 69
-------
Portland
131 MI.
OREGON
ARLINGTON FACILITY
V
w
N
EASTERN TRACT
Figure 1
Chem-Security System, Inc. Disposal Site Location Map
LEGEND

DACIh« Wilt*
M«nce«m*nt Unll

C~ Computed Wist*
S, Manflgamvnt Unit
«n
L-4: Landfill 4
P-4: Pond 4
4
N
Figure 2
Facility Layout Map Showing Major Waste Management Units
FACILITY LAYOUT AND OPERATION
The Arlington facility, which was opened in 1976, provides
hazardous waste treatment, storage and disposal services primar-
ily to the Pacific Northwest, Alaska and Hawaii, although it also
receives hazardous wastes from other western states and Super-
fund-related activities. The Arlington facility presently operates
under RCRA Part A Interim Status authorization and a Hazar-
dous Waste Disposal Site License from the Oregon State Depart-
ment of Environmental Quality. PCB wastes regulated under
TSCA are accepted at the site under authorization from U.S.
EPA Region 10 and are stored, treated and disposed of separately
from the RCRA-regulated wastes. The facib'ty does not accept ex-
plosive, radioactive or infectious wastes. Wastes that cannot be
treated or disposed of at the facility, or that can be reused or
recycled are temporarily stored at the facility and then shipped
elsewhere for treatment, recycling, disposal or beneficial use.
The existing waste management units that require groundwater
monitoring under RCRA at the Arlington facility include surface
impoundments, reactive solids hydrolysis and landfills. Sixteen
waste impoundments (of which a minimum of eight are under-
going closure), nine landfills (of which four are complete), seven
container storage areas and four storage tanks comprise the ex-
isting major RCRA waste management units at the site; addi-
tional facilities are planned for the future. A liquid waste
solidification system and a truck-wash operation are also in use.
The layout of the existing waste management units of most im-
portance at the Arlington facility are shown in Fig. 2.

GEOLOGIC SETTING
The Arlington facility is located in the south-central portion of
the Columbia Plateau physiographic province within the
Deschutes-Umatilla Plateau.2 The area is characterized by upland
areas of sandy deserts separated by relatively wide, deep to
moderate ephemeral stream drainages such as the Alkali Canyon
which borders the south side of the property. In addition to the
semi-arid climate, several factors account for the physiography of
the area and include the presence of extensive flood-bault
bedrock, subsurface geologic structure and catastrophic floods of
glacial melt water.
The subsurface geology of the facility and surrounding aretl
consists of a thick, accordantly layered sequence of basalt flow
and sedimentary interbeds, collectively known as the Columbia
River Basalt Group. The basalt flows are part of the Columbia
Plateau geological flood-basalt province' of Miocene to lOMT
Pliocene age (8 to 17 million years old). This sequence is unco*
formably overlain by younger intercalated and suprabaaA
sedimentary units of Miocene to Holocene age.
Within the site area, the formations which comprise theCotaav
bia River Basalt and the EUensburg Formations include WCfri
members of regional extent. These are the Frenchman Spri»P
70 MODELLING CROUNDWATER & SURFACE WATER
-------
and Priest Rapids Members of the Wanapum Basalt, the Pomona
Member of the Saddle Mountains Basalt and the Selah and Rat-
tlesnake Ridge Members of the Ellensburg Formation. In addi-
tion to these formal stratigraphic units, several informal units
useful to comprehension of site structure and stratigraphy are de-
fined. These include several unnamed interbeds within the
Frenchman Springs Member, an interbed within the Priest Rapids
Member and an areally extensive vitric tuff which occurs at the
top of the Selah Member. In addition to the informal units, there
are several facies of local extent. These include three facies of dif-
ferent lithology within both the Selah Member of the Ellensburg
Formation and the Dalles Formation. A generalized stratigraphic
column is shown in Fig. 3. A discussion of the underlying geologic
units, oldest to youngest, is presented below.
1
1
ll
c

:
a
L

;•
C
«0
O

o
o
o
2

| EPOCH
9
c
S
X
Pleisto-
cene

Miocene

IU
0
<

=
E
«
to

ta
a
o
o
E
to
p>

STRATIGRAPHIC
UNIT
iiiiiiilP
^§^§§iP
^iilllilliP

Dalles Foundation

Ellensburg
Formation
M
sss!
wlm|
o c
3 °
•o 5
c E
^ 5
UJ U.
I 1
o
o
u.
E
3
a
c
to
S
Grande Honde Fm
GEOLOGIC UNIT
(Common Nam«)
Loess
Alluvium
Colluvium
Flood Gravels
"•..Channel unit
Upper Tulf Unit
Conglomerate Unit
Ra'tYle snake Ridge
... Member
Pomona Basalt
Vitric Tuff Unit
Selah Member
Priest Rapids Basalt
(Upper Flow)
Priest Rapids Interbed
Priest Rapids Basalt
(Lower Flow)
Frenchman Springs
Basalt Member
Grande Ronde Basalt
LITHOLOGY

Sand, gravel and silt water-laid
Slope wash-sHt.sand&rock (rag.
Sand and gravel, some silt.
some caliche
Poorly sorted silly grave)
Tan to light green, massive.
very solt
Poorly to moderately indurated
conglomerate
Weathered tuff
Dark grey. very hard, massive.
tine grained occasionally
vesicular
Light buff to cream, very soft
Tuflaceous slitsloneCsome clay
and sand interbeds). tight olive
green, very soft to soft
Dark grey, massive, line grained.
occasionally vesicular, very hard
Tuflaceous vitric to lithte tuff.
Hghl ottve green very toll to soft
Dark grey, massive fine grained
occasionally vesicular, very hard
Dark gray, massive, fine grained.
occasionally vesicular, very hard
Dark grey massive, fine grained.
occasionally vesicular, hard
LEGEND: ••"-•-. Unconformity
Figure 3
Generalized Stratigraphic Column
Grande Ronde Basalt
The Grande Ronde Basalt is the oldest formation within the
Columbia River Basalt group in the site area. In the site area, the
thickness of the Grande Ronde Basalt is not known although it is
probably 3,000 to 4,000 ft thick based on a review of well records
and regional outcrops." Because these large flows advanced to the
limits of the Plateau and then cooled while ponded, they formed
nearly continuous sheets of competent, black, glassy, fine-
grained, columnar-jointed basalt which are separated by vesicular
to rubbly flow breccias which occur at the top of most of the
flows. Only a few thin discontinuous sediments occur as interbeds
within the Grande Ronde. This unit was not penetrated by on-site
borings.

Vantage Member of the Ellensburg Formation
The Vantage Member of the Ellensburg Formation is an
arkosic sandstone deposited by the ancestral Columbia River on
top of the Grande Ronde Basalt. In most areas of the Plateau, the
Vantage sandstone is absent; however, a well developed saprolitic
soil generally considered to be part of the member occurs at its
stratigraphic position on top of the Grande Ronde. This unit was
also not penetrated by on-site borings but is approximately 22 ft
thick as noted in a boring drilled about 6 mi southeast of the
facility.5

Frenchman Springs Member
The Frenchman Springs Member of the Wanapum Basalt6'7 is
the most extensively distributed group of flows in the Wanapum
Basalt and perhaps in the entire Columbia River Group. There are
three to six flows of basalt within the member in the site vicinity
characterized by hard, dark gray basalt. This unit was penetrated
during drilling for an on-site water supply well and is completed
within a 10- to 40-ft rubbly zone at a depth of about 560 ft below
ground surface.

Priest Rapids Member
The Priest Rapids Member of the Wanapum Basalt consists of
two to six large and several smaller basalt flows. In the site area,
the member is comprised of two flows which are laterally con-
tiguous to the Priest Rapids Member north of the Columbia
River8 and generally separated by an interbed. A thickness of 136
ft is inferred from a geophysical log of the on-site water supply
well.

Lower Priest Rapids Flow
The upper part of the lower Priest Rapids flow and the Priest
Rapids interbed are exposed in several outcrops in the vicinity of
the site and in several on-site boreholes. The lower flow is approx-
imately 45 ft thick. It is very similar to the upper flow with respect
to petrography, jointing and joint linings; however, in most of the
boreholes the lower flow has a flow-top breccia. This breccia con-
sists of ash to block-size fragments of scoria and vesicular basalt
that is often loose; however, in some places it is slightly welded to
form a relatively competent rock. Open voids are typical of this
interval. In some boreholes, the voids are infilled with fine-
grained sedimentary material which effectively reduces the
permeability of the breccia.

Priest Rapids Interbed
The interbed thickness varies from 2 to 12 ft in boreholes at the
site and is characterized as a tuffaceous siltstone; however, it is
quite variable in lithology and includes clayey silt, silty clay,
weathered hyaloclastite and clayey silty sandstone.

Upper Priest Rapids Flow
The upper Priest flow at the site generally consists of hard, dark
gray to greenish gray basalt. The upper few feet of the flow con-
sist of vesicular basalt, flow breccia or scoria that is moderately
weathered or decomposed. The weathered basalt consists of soft
to medium hard rock with the consistency of a clayey, sandy silt-
stone because of alteration.
The upper Priest Rapids flow varies from approximately 60 to
95 ft in thickness at the site. In outcrops in Alkali Canyon south-
west of the site, the upper Priest Rapids flow is characterized by
well-developed columnar jointing ranging from 4 to 8 ft in
diameter and rising from the base of the flows for two-thirds to
three-fourths the total thickness. Nearly vertical and roughly hex-
MODELLING GROUND WATER & SURFACE WATER 71
-------
SOUTH

WATER LEVEL MONITOR WELLS
."ptrujMn siiRFAf^e
DALLES
POMONA BASALT
VTTWC
— I
SE
UP
PRCS
BA
TUFf
-AH
^*WATER LEVEL AT IA»E
^~ OF SELAH fUHCOffffO)

llAPOS WATER LEVEL AT BASE OF
SALT /""" ""EtT»A"Mi
| ( (UHCOMFKED)
l__l
1

\
SURFACE INFLTRATON
*a3/ ;-*iriiUi>B?^«^feE j^Viiisr^Q^yiVAi^Ufi'g;
VERTICAL FLOW
VECTOR (NO HORIZONTAL FLOW)
p
«.__
T ' ^ RESULTANT FLOW VECTOR
VERTICAL
FLOW VECTOR
(NO HORIZONTAL FLOW)
(DRY}
HORIZONTAL

VERTICAL FLOW
VECTOR
WATER LEVEL N
FLOW VECTOR ,_
NO VERTICAL FLOW

^'^"^••^gir^lji^
'

UHSATUA
SL 1
SATURA
-J '
UNSATUf
, 1
r^
•TEOJOX
TED ZONE
ATEOZOKEO)
[
1 — «—
SATURATED ZONE
^_<-_->r-_-i-_rK:<<<<-z-:
PHEST RAPOS
BASALT
LEGEND
rt
UONITOR ZONE
NOTES:
(I) NOT TO SCALE
(2) CROSS-SECTION VEW EAST
(3) UNSATURATEO ZONE H UPPER PRIEST HAPOS BASALT W
SOUTHERN PORTION OF SECTION IS LOCALLY SATURATED
Figure 5
Generalized Schematic Diagram Showing
Structure of a Basalt Flow
agona) in cross-section, these columns are cut into sections by
subhorizontal cross-joints with highly variable spacing (0.5 to 10
ft). Randomly oriented joints frequently extend through the up-
per part of the colonnade. These random joints increase in density
upwards, creating a brickbat jointing pattern which forms the up-
per one-third to one-fourth of the flow. A generalized schematic
diagram showing the structure of a typical basalt flow is shown in
Fig. 5.

Selah Member
The Selah Member of the Ellensburg Formation (Schminke,
1967) occurs as an interbed between the Priest Rapids Member
and the surface. It is comprised of weathered tuffs and fluvio-
lacustrine tuffaceous sediments that accumulated during the
volcanic hiatus between extrusion of the Priest Rapids and
Pomona Basalts. The thickness of the Selah varies from approxi-
mately 115 to 160 ft in boreholes at the site where the top has not
been eroded in channels of Dalles or glaciofluvial gravel. The
uneroded Selah is approximately 138 ft thick.
The Selah contains three facies which are readily dis-
tinguishable on geophysical logs. Due to severe weathering and
zeolitic and clayey alteration of the member, they are not as
discernible in outcrop or borehole samples. The facies include:
(1) a lower facies comprised primarily of flood plain deposits
derived from the Columbia Plateau and adjacent areas, but also
containing three to four airfall tuff units; (2) a middle facies
similar to the lower, but containing large amounts of silicic-
volcaniclastic material derived from volcanic areas; and (3) an up-
per or channel facies which results, in part, from reworking of the
lower two facies.

Vilric Tuff Bed of the Selah Member
A gray, dacitic, vitric tuff attaining a maximum thickness of
about 30 ft occurs at the top of the Selah Member of the Ellens-
burg Formation. Geologic mapping in the area indicates that the
ash most likely originated from eruptions in the Cascades Range.1
Texturally, the vitric tuff is soft to medium hard, gray, medium-
grained sandstone (tuff) which is well-soned and massive in the
central 20 to 24 ft of the deposit. The lower 1 to 5 ft consist of soft
to medium hard, gray to dark gray, laminated, silty sandstone
(tuff) and sandy siltstone (tuff). The finer laminae are weathered
to soft, clayey siltstone or silty claystone. The upper 3 to 8 ft con-
sist of silty, fine to medium, vitric sandstone that occurs as cross-
laminated and/or thinly graded beds. These fine-grained layers
consist of weathered clayey silt/silty clay similar to the bottom
laminated portion. Development of an open-pit pozzalana mine
and explorations for its development have provided nearly con-
tinuous exposure of the vitric tuff in the southern part of the site.

Pomona Member
The Pomona flow is restricted to the northern portion of the
site and consists of very hard, black, fine-grained to glassy pot-
phyritic basalt. Maximum thickness is about 40 ft while the
average thickness is about 20 ft. The Pomona member at the site
exhibits its characteristic jointing habit of lower colonnade, en-
tabulature and upper colonnade (Fig. 4). However, where the
flow is thin (< ft thick) the flow top consists of 3 to 5 ft of platdy
jointed, vesicular basalt with little columnar development. Also,
development of the distinctive fans of joints in the entablature B
attenuated where the flow is thin.

Rattlesnake Ridge Member of the
Ellensburg Formation
Overlying the vitric tuff in the north part of the site is a tin,
fine-grained tuff which varies from 0 to approximately 8 ft thick.
It averages 4 to 6 ft thick in areas where it was not scoured a«*f
72 MODELLING GROUNDWATER & SURFACE WATER
-------
prior to deposition of the overlying Dalles conglomerate. In the
southern part of the site, the tan tuff rests directly on the underly-
ing vitric tuff and has a sharp, conformable contact with it. The
tuff typically consists of weathered to decomposed fine volcanic
ash and has the texture of a plastic, silty clay or clay which is
largely montmorillonitic. The tuff is generally massive; however,
locally the top and bottom few inches have thin laminae of car-
bonaceous material.

Dalles Formation
The Dalles Formation overlies the Columbia River Basalt and
Ellensburg Formation throughout the site area. It consists of
fluvial gravel, conglomerate, sandstone, fluviolacustrine silt-
stone, tuffs and loess. Many fossil soil horizons represented by
zones of caliche are present in the Dalles, and local unconformi-
ties are common.
Two facies of different origin are present throughout most of
the site area: a basal cemented sandstone and conglomerate facies
and an overlying tuffaceous siltstone facies. In addition to these
two facies, a third facies consisting of poorly sorted silty, sandy
gravel occupies a channel which meanders across the site and is
cut to the top of the Selah Member.

Surficial Materials
Pleistocene glaciofluvial, slackwater, torrential flood, col-
luvial, alluvial and Holocene eolian deposits locally mantel the
bedrock units.

Structural Geology
The geologic structure of the site is relatively simple. The site
lies partially astride one of several small east-west anticlinal folds
that interrupt the floor of the required Dalles-Umatilla
synclinorium. The overall trend of this small anticline is east-west;
however, it is somewhat sinuous in detail. The strike of the anti-
curie varies from about N 70 W in the northwestern part of the
site to N 80 E in the northeastern part. North and south of the an-
ticline, the bedrock units dip relatively uniformly to the north and
south, respectively, toward adjacent synclines.
Structure contours of the top of the Priest Rapids Member
define the shape and amplitude of folding that has occurred in the
site area since extrusion of the member approximately 15 million
years ago. Maximum amplitude of the fold in the Priest Rapids
Basalt is approximately 130 ft between the local culmination on
the anticline in the northwestern corner of the site and the
synclinal axis south of the site in Alkali Canyon. Average dip of
the Priest Rapids Basalt is about 1.5 degrees, although higher dips
occur along the anticlinal axis (1.5 to 5 degrees), and lower dips
occur in the southern part of the site (0.5 to 1 degrees). Folding of
the east-west anticline is slightly disharmonic, in that the folding
of Priest Rapids, Selah and Pomona was not perfectly concentric.
Folding of the anticline was accompanied by minor faulting
along the fold axis. Faults and related shears on the Selah typical-
ly are characterized by gouge zones consisting of remolded,
plastic clay and clay and thus appear to be of relatively lower
permeability than the Selah Member that is cut by them. High
angle to vertical tension fractures that occur above the small
thrust faults are typically open or filled with loose glaciofluvial
deposits and thus appear to be of relatively higher permeability
than those of the Selah.
Evidence developed during investigations of the site and nearby
areas indicates a lack of Holocene deformation (i.e., since the
Spokane flood). This evidence includes: (1) that evidence
developed by geologic mapping which shows that unfaulted
glacial flood deposits cover the small faults on-site and in the
western tract where no waste management activities presently ex-
ist and (2) relationships exposed in excavated trenches whereas the
fault zones exposed are truncated by unfaulted glacial flood
deposits.
HYDROGEOLOGIC SETTING
Within the Columbia Plateau, groundwater recharge is derived
mainly from incident precipitation and surface runoff; however, a
small percentage also may be derived from irrigation and irriga-
tion water diversions from the Columbia River (Pacific Northwest
River Basins Commission, 1970). The climate of the Arlington
area is arid to semi-arid; consequently, potential rates of ground-
water recharge from precipitation are low in comparison with
those for the more-humid areas such as the Blue Mountains south
of the site. According to National Weather Bureau statistics, the
mean annual precipitation at Arlington, Oregon, is less than 10
in., mostly occurring between October and March. The low an-
nual precipitation and high summer evapotranspiration (40 to 55
in.) is evidenced by the sparse vegetation consisting of prairie
grass, low flowering plants, sage brush and occasional junipers.
Phreatophytes, including some deciduous trees mostly brought in
by early settlers, are found along the stream valleys, near springs
and in other areas where the groundwater table is at moderate to
shallow depths.2
The principal aquifers within the Arlington area are associated
with the interflow zones between basalt flows. The principal inter-
flow aquifers in the site area are within and between Priest
Rapids, Frenchman Springs and the upper part of the Grande
Ronde Basalt. Of these, Frenchman Springs is the principal
aquifer developed locally for irrigational purposes. The use of this
aquifer has decreased in recent years, due to the reduction in
yields as a result of overpumping.
Within the site area, the uppermost basalt aquifer of regional
importance occurs within the interbed zone between the two
Priest Rapids basalt flows at an elevation of 620 to 640 ft (site
datum). The static water level in test wells completed within the
Priest Rapids interflow zone is 100 ft or more above that reported
for wells completed in the Frenchman Springs. Priest Rapids has
also been severely depleted in recent years because of overpump-
ing. Grande Ronde, which underlies Priest Rapids and French-
man Springs, is tapped by the city of Arlington for its water sup-
ply and by several irrigation wells south of the Arlington facility.
Recharge to the basalt interflow aquifers occurs mainly along
outcrops and through fractures which provide hydraulic com-
munication to the surface. The principal areas of groundwater
recharge to the Priest Rapids, Frenchman Springs and Grande
Ronde aquifer systems are south of the Arlington facility, where
the edges of north dipping basalt flows are exposed and precipita-
tion is comparatively higher. Additionally, Priest Rapids and
Frenchman Springs are exposed locally along the major drainages
south of the site which are local areas of groundwater recharge.
Regionally, the direction of groundwater flow within the Colum-
bia River Basalts in the Arlington area is to the north toward the
Columbia River.
The interflow aquifers within the Columbia River Basalts
typically have high to very high permeability and low storativity
because of the open nature, but limited volume, of joints and
fractures. Furthermore, because of the generally impervious
nature of the intervening tuffaceous sediments and dense basalt,
stratigraphically adjacent interflow zones may be hydraulically
isolated over large geographic areas. This physical and hydraulic
separation is commonly reflected by differences in both piezo-
metric levels and water quality between adjacent interflow
aquifers.
MODELLING GROUNDWATER & SURFACE WATER 73
-------
Gronndwater Occurrence
The hydrogeologic conditions at the Arlington facility are com-
plex, consisting of multiple zones of saturation with varying de-
grees of interconnection (Fig. 4). The uppermost zone of satura-
tion beneath the Arlington facility is at the base of the Selah, 100
to 200 ft beneath the existing ground surface. This zone of satura-
tion at the base of the Selah is the first detectable zone en-
countered during drilling capable of yielding even small quantities
of water to an open borehole and it is, therefore, the uppermost
zone capable of being monitored. Because of stratification and
marked permeability contrasts that exist within and between the
overlying Dalles Formation, Pomona Basalt, Vitric Tuff, and the
Selah, it is reasonable to expect that isolated, perched zones of
saturation could exist above the base of the Selah. However, no
such zones have been detected. Furthermore, no perched zones of
saturation were identified from the downhole geophysical logs
although variations in soil moisture content with depth are evi-
dent.
Groundwater occurs under water table conditions at the base of
the Selah. It also occurs under both water table and partially con-
fined conditions within the upper Priest Rapids flow above the in-
terbed between upper and lower flows and within the interflow
zone between the two Priest Rapids basalt flows. The uppermost
zone of saturation is located physically within the Selah and
above the top of the Priest Rapids Basalt. This saturated zone is
continuous across the active southern two-thirds of the site (north
—Fanning columns
Either Chined. Brecclated -
Contact with Lower Flow
or
Pillow -Palagonlte Complex
Relatively sharp
contact
Dividing columns

Btocky Joints
— Vesicle Columns
Pipe Vesicle
— Vesicular base
Chilled base and
"Incipient pillow"
PalagonKe
Flow direction
Figure 4
Conceptual Flow Model Showing Flow Vectors
in Relationship to Observed Water Levels
of Alkali Canyon). It is either thin or absent beneath the north-
western corner and in the north-central part of the site. Ground-
water that would otherwise be present in these areas is believed to
flow downward from the Selah into Priest Rapids, in contrast to
other areas of the site where groundwater is perched on top of
Priest Rapids and forms a continuous saturated zone.
Beneath the northern two-thirds of the site and along its
southern margin, an unsaturated zone exists within the upper part
of the upper Priest Rapids basalt flow. The thickness of this un-
saturated zone ranges from a few feet near the southern boundary
of the property to greater than 80 ft in the northern portion of the
site. The lower portion of the upper Priest Rapids basalt flow is
saturated. In the southwestern; portion of the site, continuous
saturation appears to exist from the base of the Selah downward
to the top of the interbed. The existence of the saturated zone
within the Selah is thus a manifestation of the anisotrophic nature
of the Selah as well as the existence of a low permeability zone(s)
at the Selah/Priest Rapids interface.
Groundwater also occurs under both confined and water table
conditions within the interflow zone at the top of the lower Priest
Rapids basalt flow. Groundwater within the interflow zone
beneath the south-central portion of the site, in general, is con-
fined or partially confined. In the northern portion of the site
where the interbed rises toward the anticline, groundwater within
it exists under water table or unconfined conditions.
Groundwater Regime
The groundwater regime within the uppermost zone of satura-
tion at the base of the Selah is characterized by predominantly
lateral flow from the groundwater divide in the northwest corner
of the site to the south and southeast and toward Alkali Canyon.
Groundwater movement appears to be principally horizontal, al-
though available piezometric data indicate that there is also a ver-
tical (downward) hydraulic gradient within this zone of satura-
tion. This finding indicates recharge from the Selah to Priest
Rapids, albeit at a slow rate. The observed pressure head distribu-
tion within the Selah suggests the existence of a low hydraulic
conductivity zone at the base of the Selah/top of Priest Rapids.
Evidence of the presence of such a layer was found in rock core
samples from the Selah/Priest Rapids interface. In these core
samples, the fractures and vesicles within the top several inches of
Priest Rapids were infilled with secondary weathering products
(clay and silt).
Recharge to the saturated zone at the base of the Selah is be-
lieved to be from direct infiltration of incident precipitation
across the site and ponded surface runoff. In this regard, the doted
depressions in the northern portion of the site and the silty gravel-
filled Dalles channel are believed to be preferential areas for
groundwater recharge. Groundwater discharge is to Alkali Can-
yon as evapotranspiration, as lateral inflow to the glaciofluvitl
deposits in the canyon and by vertical recharge to the underlying
Priest Rapids Basalt. The water table at the base of the Selah is
depressed below the top of the Priest Rapids Basalt along theat-
treme southern boundary of the site. This depression of the water
table coincides with a groundwater mound within the upper Priest
Rapids Basalt indicating recharge from the Selah to Priest Rapids.
Tritium analysis indicates that groundwater within the lower por-
tion of the Selah predates 1953.

Aquifer Characteristics
The existence of a zone of saturation at the base of the Sdah
directly overlying an unsaturated zone within upper Priest Rapids
Basalt suggests either a permeability contrast between the Selth
and Priest Rapids or the presence of a lower hydraulic conductiv-
ity layer or zone. The latter hypothesis is consistent with the
observed pressure head distribution within the Selah.
74 MODELLING GROUNDWATER & SURFACE WATER
-------
In situ falling head permeability (slug) tests in open boreholes
terminated at the base of the Selah/top of Priest Rapids indicate
that the horizontal hydraulic conductivities at the base of the
Selah range over two orders of magnitude from about 1 x 10-6
to 1 x 10-4 cm/sec. Laboratory tests on undisturbed cores from
the saturated zone at the base of the Selah indicate that the ver-
tical hydraulic conductivity, which ranges from 1 x 10-8 to
1 x 10-5 cm/sec, is one to several orders of magnitude less than
the horizontal hydraulic conductivity as determined from slug
tests. The apparent difference between field horizontal hydraulic
conductivity and laboratory vertical hydraulic conductivity may
be attributed in part to the difference in measurement techniques.
Packer tests conducted within the Priest Rapids Basalt indicate
wide variations in hydraulic conductivities. The measured
hydraulic conductivities in the basalt range from less than 1 x
10-8 cm/sec to greater than 1 x 10-3 cm/sec. Although highly
variable, the hydraulic conductivity of the Priest Rapids Basalt is
generally of the same order of magnitude as that within the basal
portion of the Selah.
To assess their hydraulic characteristics and degree of hydraulic
connection, pumping tests were conducted in the Selah and Priest
Rapids. The Selah pump tests entailed pumping from the base of
the Selah and monitoring the response to pumping at the base of
the Selah, at an intermediate depth between the base and the
water table, and at the water table. Two locations were selected
and pumped at rates 0.25 and 0.167 gal/min, respectively, which
resulted in full drawdown in the pumping wells. The Selah pump
tests were, however, of limited duration because of the low
permeability of the Selah.
The calculated horizontal hydraulic conductivity of the Selah
for the Jacobs nonequilibrium' and the Theis' analytical methods
ranges from 1 x 10-6 to 1 x 10-4 cm/sec. These values are based
on the time-drawdown responses, which closely follow the Theis
curve predictions. Because of the limited duration of the tests, the
calculated results were considered order-of-magnitude estimates
only.
The Priest Rapids pump test had dual objectives: (1) the assess-
ment of the aquifer characteristics within the Priest Rapids in-
terflow zone and (2) the assessment of possible interaquifer com-
munication between the Priest Rapids interflow zone, the upper
Priest Rapids basalt flow and the saturated zone at the base of the
Selah. Observation wells/piezometers were installed at three
levels: within the saturated zone at the base of the Selah, within
the upper Priest Rapids Basalt flow and within the Priest Rapids
interflow zone. The Priest Rapids interflow zone was pumped at a
constant rate of 0.5 to 1 gal/min, which was the maximum sus-
tainable pumping rate over the planned duration of the tests.
The Theis and Jacobs analyses were applied on the assumption
that the observation wells partially penetrate an anisotropic, un-
confined aquifer composed of the Selah, the upper Priest Rapids
Basalts and the interflow zone. These analyses, however, would
not be applicable if an unsaturated zone existed between the base
of the Selah and the interflow zone. An unsaturated zone would
provide an effective discontinuity between the saturated zone at
the base of the Selah and the pumping zone within Preist Rapids,
and hence mask any response. The area in which the Priest Rapids
pump test was conducted is the only known area on the site that is
fully saturated from the base of the Selah to the interflow zone.
The drawdown response of several of the observation wells in-
dicates that the interflow zone in this area is hydraulically con-
nected with the saturated zone at the base of the Selah.
The transmissivity calculated on the basis of the Priest Rapids
pump test ranges from 0.3 to 19.1 ftVday. Assuming an average
saturated thickness of approximately 7.5 ft for the interflow zone,
the horizontal hydraulic conductivity of the interflow zone is
estimated to be 1 x 10-5 to 9 x 10-4 cm/sec. Due to the limited
duration of the test and the mathematical limitations of the Theis
and Jacobs analyses, the calculated pump tests results are con-
sidered order-of-magnitude estimates.
An analysis by Witherspoon10 for aquitard response to pump-
ing in an aquifer has also been applied on the assumption that the
Selah, the upper Priest Rapids Basalt and the interbed zone are an
aquitard and the interbed zone is an aquifer. In this analysis, the
observed drawdown responses in the aquitard are used to estimate
the vertical hydraulic conductivity of the aquitard. The estimated
vertical hydraulic conductivity of the aquitard estimated on this
basis ranges from 1 x 10-7 to 2 x 10 ~7 cm/sec, although these
values may be an underestimate due to the time lag response to
the piezometers.
DISCUSSION AND CONCLUSIONS
Geologic and hydrogeologic characterization of the Arlington
facility to the extent necessary to proceed with development of a
detection groundwater monitoring system was approved by
Region 10 of the U.S. EPA in November 1986. The studies per-
formed addressed criteria established under 40 CFR 264, Subpart
F and included identification of the uppermost aquifer and
aquifers hydraulically interconnected thereof beneath the waste
management area, evaluation of their respective groundwater
flow rates and direction and the basis for such identification.
These studies have also demonstrated that the geologic and
hydrogeologic conditions at the Arlington facility make it en-
vironmentally favorable for hazardous waste disposal. The
favorable conditions at the site, however, have been very difficult
to demonstrate in an unequivocal manner due to tenuous reliabil-
ity of testing methods available for use under these site condi-
tions.
During the development of the RCRA Part B Permit,
numerous meetings were held between the U.S. EPA, facility rep-
resentatives and various consultants to discuss the hydrogeologic
and groundwater monitoring potential of the facility. The
agency's posture has been that a site hydrogeology is not "proper-
ly characterized" until all alternative hydrogeologic hypotheses
have been ruled out. Such a standard of characterization is not
manifested in the specific regulations nor is it applied to other
groundwater monitoring programs including the Safe Drinking
Water Act (1424E), Sole Source Aquifer Review Program or the
CERCLA program involving groundwater issues.
Although the basic flow regime at the facility was understood
as early as January 1984, "characterization" was not completed
until November 1986. Over 120 borings and/or wells were con-
structed during this period. Many of these installations were re-
quired by the agency to verify predicted water table elevations or
to "demonstrate" the presence or absence of some hypothesized
factor conclusively. The agency expressed deep concern that,
despite the preponderance of hydrogeologic data, a "black hole"
existed beneath the site that would provide a conduit to vertical
flow. While the preponderance of geologic data was to the con-
trary, this hypothesis was maintained throughout the characteri-
zation period.
As a result, during the characterization process, U.S. EPA re-
quired numerous borings through the confining layer, one of
several factors making the site favorable for hazardous waste
management. The very data deemed critical to characterization in
turn required risking the integrity of the vadose zone, the lower
flow boundary and overall site suitability. It becomes evident that
the burden of test required for a site to be "properly character-
ized" may in fact introduce considerable uncertainty in relation
to site integrity and long-term suitability.
MODELLING GROUNDWATER & SURFACE WATER 75
-------
REFERENCES

1. Dames & Moore, "Report on subsurface investigation and moni-
toring well installation, CSSI Arlington facility, Arlington, Oregon,"
Unpublished report prepared for Chem-Security Systems, Inc.,
Bellevue, WA, Dec 1983.
2. Dicken, S.N., "Oregon geography," Eugene, OR, 1955.
3. Fenneman, N.M., Physiography of Western United States, McGraw-
Hill, New York, NY, 1931.
4. Rockwell Hanford Operations, "Geologic studies of the Columbia
Plateau—a status report," Report No. RHO-BWI-ST-4, 1979.
5. Foundation Sciences, Inc., "Geologic reconnaissance of parts of the
Walla Walla and Pullman, Washington and Pendleton, Oregon,
1 x 2 AMS quadrangle," Prepared for USAGE, Seattle District,
Contract No. DACW 67-80-C-0125, 1980.
6. Mackin, J.H., "A stratigraphic section in the Yakima basalt and the
Ellensburg formation in south central Washington," Washington
DOE, 1961.
7. Swanson, D.A., Wright, T.L., Hooper, P.R. and Bentley, R.D.,
"Revisions in stratigraphic nomenclature of the Columbia River
basalt group," U.S. Geol. Survey Bull. 1457-G, 1979.
8. Schminke, H.U., "Petrology peleocurrents and stratigraphy of the
Ellensburg formation and interbedded Yakima basalt flows, south
central Washington," Ph.D. disseration, Johns Hopkins University,
Baltimore, MD, 1964.
9. Lohman, S.W., "Groundwater Hydraulics," U.S. Geol. Survey
Prof. Paper 708, 1972.
10. Witherspoon, P.A., Javandel, I., Neuman, S.P. and Freeze, R.A.,
"Interpretation of aquifer gas storage conditions from water pump-
ing tests," American Gas Association, Inc., New York, NY, 1%7,
76 MODELLING GROUNDWATER & SURFACE WATER
-------
Waste Treatment and Reduction Applications
Of Freeze Vaporization

Val Partyka
CALFRAN International, Inc.
East Longmeadow, Massachusetts
ABSTRACT
CALFRAN's process of low temperature evaporation is cap-
able of reducing and treating a wide variety of hazardous waste
streams. This process has been successfully used to treat a number
of aqueous wastes. Three examples of waste reduction applica-
tions are presented. High purity product water was produced with
metals concentrations as low as 0.01 mg/1 in the reclaimed water.
The system's ease of operation and its insensitivity to feed con-
centration are discussed. Performance data on a number of pend-
ing applications also are discussed.

INTRODUCTION
The CALFRAN system of low temperature evaporation is be-
ginning to be applied in the wastewater treatment and waste
reduction field. Historically, freezing or low temperature
evaporation has been thought to possess a significant potential to
concentrate industrial waste streams to recovery valuable pro-
ducts from the wastes. The advantages of a low temperature pro-
cess over other treatment systems for industrial waste treatment
are: its immunity to fouling, negligible corrosion and lack of any
pretreatment requirements.

WASTE REDUCTION
There is increasing industrial emphasis on waste reduction
because of increasing cost and difficulty of disposal. One way to
reduce hazardous waste is to concentrate hazardous constituents
in aqueous solutions. The low temperature evaporation process
described in this paper does just that. In many cases, the concen-
trates have been reduced to a solid or semi-solid form. Many
wastes classified as hazardous actually contain a very small por-
tion or percentage of hazardous material compared to the overall
volume because they are suspended in an aqueous solution. The
low temperature evaporation process is a method which efficient-
ly concentrates these hazardous components down to saturation.
Most disposal facilities charge the same amount to dispose of
similar drums of hazardous waste even though their concentra-
tions may differ substantially. One need only contaminate a non-
hazardous solution with a small amount of hazardous material
and the entire volume becomes hazardous. Thus, concentration
becomes cost-effective.

WASTE TREATMENT
The low temperature evaporation process also offers a new ap-
proach to wastewater treatment. It is possible to feed a waste
stream directly into the system without using chemicals or filters.
The treatment can be placed upstream in the process at the point
of greatest concentration of pollutants, thereby greatly reducing
the volume of wastewater to be treated. In most cases, the pro-
duct water can be recycled to the process due to its high purity.
For example, in a typical counterflow/rinse operation, the
highest concentration of contaminants is in the first rinse tank.
Since the process is not dependent on concentration it is logical to
draw from the first tank and return the purified water to the last
counterflow rinse tank, thereby reducing the volume of
wastewater to be treated. This closed loop approach is fast
becoming the goal of all who are faced with stricter discharge re-
quirements.

THEORY OF OPERATION
Water at atmospheric pressure (760 mm Hg) boils at 212 °F.
When the pressure is lowered to below 4 mm of Hg, the boiling
point of water will equal its freezing point. Thus ice and water
Coolant
Chambe"
Coolant —*. —.*.

Feed .— , _ ,
Brine Ice

Waate
jj£^

ENERGY (~~) RECOVERY

CALFRAN'S

FREEZE VAPORIZATION PROCESS

* PATENT No.4406748
Figure 1
Freeze Vaporization Process
INNOVATIVE TECHNOLOGIES 77
-------
vapor are formed simultaneously. This is known as the triple
point effect. The CALFRAN process employs this principle. The
process can also be operated above the triple point where water
will boil between 32° and 50 °F at approximately 10 mm of Hg.
In a basic treatment unit, a condenser is placed within a sealed
chamber which is evacuated by a conventional vacuum pump.
The waste stream to be treated is pumped into the chamber and
then either sprayed or allowed to flow through an open pipe (Fig.
1). As long as the temperature of the condenser and the pressure
in the system are below the triple point of the feed waste stream,
the liquid will spontaneously form brine and/or ice and water
vapor. Below the triple point, liquid will split into two parts; an
approximately 88% ice and brine mixture and 12% pure water
vapor. The vapor travels upward in the reaction vessel toward the
condenser where it becomes the purified flash ice or condensed
water on the special heat exchanger. At the bottom of the reaction
vessel, a brine solution is formed; this solution contains the con-
centrated contaminants.
By cycling the process from freeze to defrost (melt), flash ice is
formed and then melted to produce the purified product water.
This melting of flash ice to pure liquid product can be ac-
complished in several ways. One method is to introduce excess
vapor into the system causing the pressure to rise above the triple
point. The vapor will condense on the flash ice, thus melting it
and producing more product water. Operating slightly above the
triple point is very effective in some cases.
The process finds broad application throughout the field of
wastewater treatment. For water reuse, the process can handle
even mixed effluents producing ultrapure water directly. Tests
have shown in a number of cases that TDS (total dissolved solids)
concentrations have been reduced by factors of up to 10,000. For
Table 1
Freeze Vaporization Performance Data
Wastewater Treatment Applications
Zinc Phosphate
and Nickel
Rlnsewater

ChroaAc Sulfrlc
Etchant
Electrolesa
nickel .
Strip Acid

Soldering
Mlnseuater
location

IN
25.0

0.00
III
OUT

in
OUT

III
OUT
*0,000
0.08
16,700 0.1
0.02 5.5
Composite Plating IH
Mlnsevater OUT
Selenlu.
Rlnsewater

Industrial
Laundry
ftinsewaler

Silver Coating
Eaulsion
•ater Base
Ink
II
OUT

OUT

II
OUT

II
OUT
13.9 29.9 0.99

Se_
10,000
0.5
Tot > Sol
105" OTT
0.01 0.00
1100 210 0.62
0.00 0.04 0.1

1.3
0.0

1.9
<.01
5.1 2.9 12.0

<.01 C.01 8.8
Color
dark blue
2
Mater
Soluable oil •
Cutting Oil

Zinc Cyanide
llnsevater
Iron rnospnau .
Chrcavjte Com.
Coating

tlec Iroc leaner*

in
OUT

II
OUT
II

OUT

11
OUT

ex
372 ~
0.00 2.0
us *.*

0.01 0.02

O.Ot 0.00 0.00

8.1

6.5
Surfactants

0.0 8.3
*2«.l

0.6
50/50 I

0.4
O-C - Oil and Crease
waste treatment, the process can concentrate a waste stream to
saturation, with the potential for resource recovery.
For waste reduction applications, one of the more attractive
features of the process is that the brine solution in certain cases
can be completely dried. Units can be designed to either concen-
trate and recover the brine solution for reuse or take the brine to
dryness.

WATER QUALITY
Listed in Table 1 are a number of industrial waste streams that
have been successfully treated by the low temperature evapora-
tion process. Metals have been consistently reduced to less than
0.1 mg/1 in the effluent. The process also removes color, total
dissolved solids, suspended solids, surfactants, oils, phosphates,
chlorides and organics responsible for chemical oxygen demand
and conductivity.

ADVANTAGES
Insensitivity to Concentration
Fluctuations in concentration do not affect process perfor-
mance efficiencies. As the feed concentration increases, the brine
solution concentration increases but the product water purity re-
mains the same. This makes the process particularly attractive for
those applications with widely varied waste stream concentra-
tions.

High Reliability
Since the system does not use filters, employs no surfaces that
corrode, and is insensitive to contaminant concentration or the
accidental introduction of unexpected pollutants, the process is
less susceptible to upsets than conventional treatment processes.
One need only flush the system with water if contamination
should occur, start feeding the waste stream again and the system
is back on line. Because it is a physical process, product water
contamination can only occur by one of the following
mechanisms:
• In a few feedwaters, excessive foaming can cause carryover
into the condensing area. Waste streams which have a tendency
to foam are metered at a controlled rate under vacuum. In
this manner, solutions such as electrocleaners, soap-cleaners
and phosphates can be successfully processed.
• Carryover can also occur if droplets of the concentrate reach
the condensing area. Carryover can be controlled by proper
mist elimination.
Both of these phenomena, if occurring, can be visually ob-
served, thus making the process very easy to diagnose.

Pretreatment Not Required
Pretreatment is not required and most wastewaters can be pro-
cessed directly without chemically altering their characteristic!.
However, some industries have requested pH adjustment before
processing, which has rendered the concentrate non-hazardoui
and given them an attractive disposal option.

Size and Ease of Operation
The units are compact, with all components mounted on a
single frame. A basic unit consists of an air-cooled compressor, >
vacuum pump and specially designed condensers (Fig. 2). Larger
systems consist of two or more of each of the foregoing.
The units are simple to operate. Once the vacuum pump &
turned on and the vacuum begins to decrease, one only hai to
open the chiller valve and the system operates with no further ad-
justments. The process can be stopped or interrupted at any tin*
78 INNOVATIVE TECHNOLOGIES
-------
and restarted without causing damage to the equipment or af-
fecting treatment and product water quality. No water, steam or
air supply is necessary. Wastewater streams can be hard piped for
continuous large volume operations, or units can be batch run for
small volumes.
Figure 2
CALFRAN PTU-100 Waste Reduction Unit
(Capacity 100 GPD)
Energy Requirements
Energy requirements are site-specific and depend on the type
and degree of treatment desired.

WASTE REDUCTION APPLICATIONS
Since the process is capable of concentrating a solution to
saturation, it has been utilized for waste reduction. The volume of
aqueous solutions can be reduced by as much as 99% with signifi-
cant savings in disposal costs.

Selenium Waste
One industry that generates 211 kg/day of a hazardous waste
which contains up to 1,000 mg/1 of selenium has utilized a 100
gal/day Portable Treatment Unit to reduce the amount of waste
to 8.0 kg/day containing 96% solids. At a $250.00 disposal cost
per drum, the savings is in excess of $50,000/year. The payback
on the unit is approximately 6 mo.
The treatment system utilizes approximately 0.44 kwh of elec-
tricity/gal of wastewater treated. A typical day's operation in-
volves 1.5 hr/day of operator time including startup and shut
down with manual concentrate removal.
Test data on the product water indicate that selenium levels
were reduced to between 0 and 0.5 mg/1. Once the unit has been
started, the wastewater is automatically fed into it from a 55 gal
drum until the drum is empty. At this point the unit will begin to
concentrate the solution to dryness. This step is accomplished
automatically by high and low pressure controls within the unit.
When the waste has reached its maximum concentration, either as
a liquid or dry material depending on the type of waste being pro-
cessed, the compressor will shut down but the vacuum pump will
remain on.

Nickel Solutions
Another example is treatment of nickel chloride solutions and
electroless nickel. The unit has been set up with its own feed tank
and product water holding tank. The unit reduces approximately
1600 gal of nickel chloride solution to 5.0 gal or 50 Ibs of dry
nickel chloride salt. In this application, some of the dry nickel
chloride product is periodically returned to the plating bath while
the excess is stored as a dry solid. The pure product water is
recycled back into the process, thereby completely closing the
loop with no regulatory permitting or monitoring required.

Spent Plating Baths
A 100 gal/day unit is being used for waste reduction on a vari-
ety of spent baths and solutions. The industry has processed spent
electroless nickel, nitric acid, sulfuric acid, alkaline cleaners and
water soluble oils in one unit and under one set of operating con-
ditions. Volume reductions range from 10% to 98%. In all cases,
the product water is acceptable for discharge to the sewer meeting
both categorical and local pretreatment limits. The product water
from these solutions would be acceptable for recycling, however
there is no recycling presently being done.

CONCLUSIONS
With the future on water reuse, resource recovery and waste
reduction, the low temperature evaporation process allows one to
attain all three goals simultaneously.
INNOVATIVE TECHNOLOGIES 79
-------
The Treatment of Solvent-Contaminated Waste
Using Liquefied-Gas Extraction
William E. McGovern
John M. Moses
CF Systems Corporation
Cambridge, Massachusetts
ABSTRACT
The impending Nov. 8, 1986 RCRA land disposal ban places
significant pressures on waste generators and management
facilities to separate and treat organic hazardous waste streams.
Liquefied gas extraction is a process designed to remove organics
from a variety of waste streams and is capable of reducing the
residual concentrations of industrial solvents to below regulatory
levels. This paper reports the calculated residual concentrations
that can be achieved for 10 organic solvents treated in a 25-tray
counter-current extractor, based on an organic feed concentration
of 1.0 wt.%.

INTRODUCTION
According to estimates from the National Association of Sol-
vent Recyclers,' nearly 10 million tons of solvent-contaminated
waste are generated in the United States annually. The nature and
origin of these wastes vary greatly and include both chlorinated
and nonchlorinated hydrocarbons from the electronics, polymer,
dry cleaning and metalworking industries. Other volatile organics
and nonchlorinated solvent wastes such as those containing
acetone and alcohols, are generated in chemical and degreasing
operations. Currently these wastes are handled in several ways: (1)
by recovery at some of the large chemical and electronics plants;
(2) by incineration in Chicago or in the South; (3) sent to landfills
in New York, Ohio, South Carolina and Alabama.2 Even in dilute
concentrations, the presence of these organic solvents often deter-
mines whether a waste is considered hazardous, and pretreatment
methods designed to remove organics are increasingly seen as a
favorable approach.

LAND BAN DISPOSAL RULE
The impending land disposal ban on the spent halogenated and
nonhalogenated solvents listed at 40 CFR, 261.31 as F001-F005,
the listed dioxin wastes, and the soon-to-be-proposed California
List (which includes liquid or solid halogenated organics in excess
of 1,000 ppm), as well as the recently proposed toxic constituent
leach procedure (TCLP), place significant pressures on waste
generators and management facilities to separate and treat
organic hazardous waste streams. The TCLP targets 38 organic
constituents that were not previously part of the characteristic
wastes on the "EP toxic" list at 40 CFR, 261.64. This, plus the
ban on land disposal of the solvents and California List, will
create significant demand for separate treatment and manage-
ment of solvent and halogenated organic waste streams.'

LIQUEFIED GAS EXTRACTION
The CFS process removes organics from a variety of waste
streams, and is capable of reducing the residual concentrations of
industrial solvents to below regulatory levels. The net effect of
this process is the separation of a waste into two fractions: an in-
organic fraction usually containing solids and water, and a con-
centrated organic fraction. The process recycles wastes when one
or both of these fractions is returned to industry for reuse, and is
a favorable method of compliance with the Nov. 8, 1986 RCRA
restriction on the use of landfills. In cases in which recycle is not
feasible, the inorganic fraction may be suitable for sewer
discharge or a sanitary landfill, whereas the concentrated organics
are incinerated or used as waste fuel.
The CFS process liquefies ordinary gases like carbon dioxide
and propane by compressing them above their room temperature
saturation pressure. In the extractor these liquefied gases are
capable of dissolving large quantities of organics to form an ex-
tract phase, which is continuously fed to a separation vessel. In
the separator, the extraction gas is vaporized and recycled as fresh
solvent while the extracted organics are continuously removed
and collected as product (Fig. 1).
CF Systems
Extraction Flowchart
Extrac
Separator
Extractor
Makiup
Compressor
Solvent
Water and Solids
Organics
Figure 1
CF Systems Extraction Flow Chart
Carbon dioxide is a prime candidate for the removal of
organics from aqueous waste and was the solvent gas used in all of
the extractions presented here. Carbon dioxide is inexpensive, en-
vironmentally safe (nontoxic and nonflammable) and can be u*td
effectively at room temperature. Most organics readily dissolve in
liquid CO2,' and a few examples of organics found
in CO2 at 25 °C and 65 alms are given in Table 1.
80
INNOVATIVE TECHNOLOGIES
-------
Compound
Table 1
Examples of Organic Compounds Miscible in
Liquid CO2 at 25 °C and 65 atm4

Class
Acetaldehyde
Acetic Acid
Acetone
Acrylonitrile
Benzene
Butyl Alcohol (sec-)
Chlorobenzene
Chlorophenol (0-)
Dimethylformamide (N,N-)
Ethyl Acetate
Isopropyl Alcohol
Toluene
Propylene
Pyridine
Aldehyde
Carboxylic Acid
Ketone
Nitrile
Aromatic
Alcohol
Aromatic
Phenol
Amide
Ester
Alcohol
Aromatic
Olefin
Heterocyclic
Performance data for acetonitrile and acrylonitrile were ob-
tained from extractions carried out in a continuous, counter-
current sieve tray extractor with a maximum feed capacity of 1.0
gal/min. The acetonitrile and acrylonitrile were prepared in a 2.0
wt. % ammonium sulfate solution to simulate the waste generated
in an acrylonitrile manufacturing plant. Other organic solvents
were chosen for their potential recovery as high volume wastes,
and were taken from several studies to determine the distribution
coefficients of industrial chemicals in carbon dioxide and water.
The objective of this research was to determine the degree of
removal of several organic solvents from water based on their
distribution coefficients and obtain actual performance data for
acetonitrile and acrylonitrile in a continuous extractor with eight
sieve trays. All of the work presented in this study was performed
by CF Systems Corporation.

RESULTS
Laboratory Data
The equilibrium distribution coefficient (DC) for any organic
compound distributed in a CO2-water mixture can be expressed
as:
DC =
(1)
(Mass of Organic in CO2 Phase)

(Total Mass of CO2 Phase)

(Mass of Organic in Water Phase)

(Total Mass or Water Phase)
The distribution coefficients for acetonitrile and acrylonitrile
are given in Table 2. The data indicated that the extraction of
acrylonitrile would be very easy with CO2 and that acetonitrile
would be the limiting component but would not present a difficult
extraction either. All of the distribution coefficients presented in
this report were obtained using an equilibrium cell containing the
organic solvent, liquid CO2 and water.

Table 2
Distribution Coefficients for Dissolved Organics
Compound
Distribution
Coefficient
Acetonitrile
Acrylonitrile
0.8
5.0
Performance Data for Dissolved Organics
All of the performance data for acetonitrile and acrylonitrile
was obtained using a countercurrent extractor with eight sieve
trays and a maximum capacity of 1 gal/min. A multipass extrac-
tion was conducted by using the raffinate from a previous pass as
a feed for the next pass, for a total of five passes. Based on
distribution coefficient data, the CO2-to-feed ratio was chosen to
be 1.5.
As predicted by the laboratory data, acrylonitrile extracted
much more easily than the acetonitrile. In one pass through the
extractor, 99% of the acrylonitrile was recovered while three
passes were required to recover the same percentage of
acetonitrile.
Tables 3 and 4 list the concentrations of acetonitrile and acrylon-
itrile in the feed and raffinate throughout the multipass experi-
ment. Also listed is the fraction of organics recovered per pass
through the extractor. Recovery here is defined as (Xf-Xr)/Xf or
more simply as l-(Xr/Xf).

Table 3
Multipass Extraction with Dissolved Organics Waste;
Acetonitrile Concentration History
PASS
NUMBER
T1HE
_

RAFFINATE
(pp.)

5
15
30
40
50
65
85
95
100
105
120
130
140
145
160
165
170
175
180
185
190
195
....
342
370
36B
340

43
43
45

17
18
19
IB
...
5.8
...
6.2

5.7
...
8.1
26
27
38
50
39
21
12
10
...
11
5.1
3.9
3.4
3.1
3.8

2.1

BMDL*
...
BHDL
2. 3
89

245
250
BMDL
8MDL
*BMDL: B«low Minln
Detection Lialc <2 ppn)
Since acetonitrile was more difficult to extract than
acrylonitrile, it was chosen as the solute of interest in the commer-
cial design of a 125 gal/min unit. It can be assumed that the final
concentration of acrylonitrile in the raffinate will be significantly
lower than acetonitrile, based on acrylonitriles lower feed concen-
tration and higher distribution coefficient.
To complete the material balance around the extractor, the
level of dissolved organics in the incoming solvent (CO2) and ex-
tract were measured using a total hydrocarbon analyzer. The
results presented in Table 5 have been converted tc a water basis
by multiplying the actual concentration of the organics in CO2 by
1.5.

ANALYSIS
The acetonitrile-water-CO2 equilibrium data combined with the
compositional analyses of the four streams entering and leaving
the extractor, gave the information required to perform an extrac-
tor analysis. The Kremser-Brown-Souder analytical solution of
INNOVATIVE TECHNOLOGIES 81
-------
Table 4
Multipass Extraction with Dissolved Organics Waste;
Acrylonitrile Concentration History
the use of the Kremser equation given below:
tlii
MUMBgl
1

•>MDL:
_HD:
UHfi
(•In)
5
15
10
10
JO
65
IS
95
100
105
120
130
140
149
HO
165
170
175
110
115
HO
195
245
2 SO
B.lov Mlcil.u.
Hoc D«c«cc«d
EMfi
(PP-)

2(7
292
211
261

9.0
«.S
! 8

2.3
3.2
2.«
2.7
...
HD

ID
...
no

D*c*etlon
{!••• than
BAFFIHATE
(PP-)
2 0
2 6
2 7
3 3
2 4
BHDL*
>HDL
>HOL
....
•HDL
HD_
HD
BD
HD
HD
...
8D
...
HD

HD
HD
HD
HD
Limit (2 ppn)
1 pp.)
RECOVERY
U-xr/xf>
t
99

Table 5
The Level of Dissolved Organics in the CO2 Streams
Around the Extractor
PASS

4
ORGANICS

(PP»>
fCO^ OUT1

633

12
ORGANICS
(ppm)
INI
"ch
- LOG
[(Xr - YS/DC)
(1 -
where:
DC
Nth
Q
LOG Q
= distribution coefficient
= number of theoretical trays
= S/F x DC
Using the Kremser equation and the data for acetonitrile, the
average tray efficiency (E) for the four passes was calculated to be
0.55. Further study showed that the Kremser equation was sen-
sitive to errors in the determination of the distribution coefficient,
and for that reason an efficiency (E) of 0.44, or 20% lower than
calculated, was used as the design basis for a 125 gal/min unit.'
The composition of feed coming into the 125 gal/min extractor
is given in Table 6; it was based on the actual concentrations of
dissolved organics in the effluent from an acrylonitrile manufac-
turing plant.
Table 6
Feed Composition Used as the Design Basis
for a Commercial Scale Extraction Unit
Component
Concentration
Acetonitrile
Acrylonitrile
Ammonium Sulfate
Water
1680 ppm
1200 ppm
2.0 wt. %
97.7 wt. %
Commercial Scale-Up
Analysis 1 (Table 7) shows the relationship between the raf-
finate and the number of actual trays required to achieve the
desired concentration. Also shown is the percent reduction and
the number of theoretical trays (Nth) corresponding to each
calculation.

Table 1
Commercial Scale Extraction of Acetonitrile;
Design Basis, Feed Concentration of 1680 ppm
the McCabe-Thiele graphical method can be used to analyze the
extractors performance for a given set of conditions.' The
number of theoretical plates were calculated using the Kremser
equation which gave the efficiency of the actual sieve trays.
The information needed to perform the analysis included the
equilibrium data given by the experimentally determined distribu-
tion coefficients, and the following four terms:
• Solvent-to-feed ratio (S/F)
• Feed composition (Xf)
• Raffinale composition (Xr)
• Solvent composition (Ys)
These four parameters were used to determine the number of
theoretical trays necessary for the extraction of acetonitrile in
each pass. The tray efficiency was then calculated for each pass
using the equation given below:
N (theoretical)
E = (2)
N (actual)

A constant distribution coefficient (DC) of 0.76 was used in the
extractor analysis. The use of a linear equilibrium line allowed for
SOLVENT-TO-FEED RATIO: 1.5 TO 1.0
RAFFIHATE (FPH)
X,
REDUCTIOH I
(1 - Xr/Xf>
2
5
10
2D
50
100
37.
27.
22.
17.
11.
7.
tt
6*
51
40
27
11
99.9
99.7
99.4
9>.l
97.0
94.1
AHALYSH ? SOLVEHT-TO-FEED "ATIO: J.O TO 1.0
RAFFIHATC (FFH)
X,
IBDUCTIOI %
(1 - Xr/Xf)
2
5
10
20
50
100
14.
11.
9.
7 _
5.
4.
14
27
22
It
14
to
99.9
99.7
99.4
9».«
• 7.0
94.1
AHALYS1S 1 X - 5.0 FFM; SOLVEIT-TO-FEED RATIO: 1.} TO 10
.13
.76
.tl
.37
.43
.ss
20.2
30.3
57.1
•«ct

55
RELATIVE t ERROR

20.1

50.7
82
INNOVATIVE TECHNOLOGIES
-------
Analysis 2 presents similar data, but uses a solvent-to-feed ratio
of 2.0 as the basis of design. This analysis shows that increasing
the solvent flow by 25% can reduce the number of trays required
by more than 50%.
Analysis 3 shows how sensitive the Kremser equation is to er-
rors in the determination of the distribution coefficient. A 10%
error in DC corresponds to much larger errors (20.3 to 50.7%) in
the calculation of the number of trays required for an actual pro-
cess.

ORGANIC SOLVENT EXTRACTION
Based on data from the extraction of acetonitrile, a tray effi-
ciency of 0.40 was used together with the distribution coefficients
for 10 organic solvents, to calculate their degree of removal from
water in a continuous, countercurrent extractor. It was assumed

Table 8
Calculated Performance of a 25 Tray Continuous Extractor
in every case that the concentration of organics in the wastewater
feed was 1.0 wt. %, and that the extractor contained 25 actual
trays, or 10 theoretical stages. The residual concentrations of the
organic solvents in the treated water (raffinate) given in Table 8
were obtained using an analytical solution to a McCabe-Thiele
type diagram; they are based on a CO2-to-feed ratio of 1.5.
For acryonitrile, trichloroethylene and carbon tetrachloride,
the residual concentration of organic solvent in the treated water
after one pass through the extractor, was less than the regulatory
level (TCLP) proposed by the U.S. EPA for landfill restriction.
Regulatory levels have yet to be proposed for the other solvents
listed in Table 8.
Finally, Table 8 shows the correlation between percent removal
and the distribution coefficient for each organic, and indicates
that laboratory data is a useful means of predicting the residual
concentration of organic when extractor efficiencies are known.
RAFFIHATE
COMPONENT
Kethanol

Isopropyl Alcohol

Acetonicrlle

Acetone

Acrylonltrile

HIBK

Triehloroethylan*

Carbon T«trichloride

Methyl Hethacrylate

Xylene
0.10

0.26

0.75

1.10

45
CONC.

(pp.)
8600

6200

500

0.16

0.06

0.05

0.Oli

0.04

0.02
(pp.)
0.07

0.07
REMOVAL
14.0

38.0

95.0

99.7

>99.9

>99.9
REFERENCES
1. National Assn. of Solvent Recyclers, Washington, DC,
personal communication.
2. Purington, J.A., personal communication, Purington Assoc.,
Braintree, MA.
3. Fortuna, R., personal communication, Hazardous Waste
Council, Washington, DC.
4. Francis, A.W., J. Phys. Chem. 58, 1954, 1099.
5. Treybal, R.E., "Mass-Transfer Operations," 3rd Ed., Mc-
Graw-Hill, New York, NY, 1980, p. 128.
6. Rice, P.N., McGovern, W.E. and Kingsley, G.S., "Near
Critical CC>2 Extraction of Hazardous Organics from Acrylonitrile,
Pesticide and Steel Mill Wastes," Contract 68-03-1956, U.S. EPA
Cincinnati, OH.
INNOVATIVE TECHNOLOGIES 83
-------
Design Innovations Employed in the U.S. Air Force
Live Fire Training Facility
Bryce E. Mason
Tyndall Air Force Base, Florida
INTRODUCTION
The U.S. Air Force has continually emphasized effective and
efficient firefighting techniques related to aircraft crash and res-
cue. To achieve high standards of readiness and effectiveness,
it is necessary to have live fire training facilities where controlled
situations can simulate emergency conditions under which per-
sonnel must operate. These facilities must include equipment and
materials that will provide safe, realistic practice scenarios—
with adequate quantities of burning jet fuel and approved ex-
tinguishing methods—along with protection of the groundwater
and liquid effluent treatment.
To meet these goals, the U.S. Air Force has been developing a
design for improved faculties that meet both practical and en-
vironmental requirements. The result is an innovative design that
provides a high degree of realistic simulation capability with en-
hanced personnel safety and acceptable environmental safe-
guards.

ORIGINAL DESIGN PROBLEMS
Along with fidelity to realistic scenarios, designs for a live fire
training facility (for simulated aircraft crash and rescue) must
consider the problem of jet fuel (JP-4) ignition and leakage in-
cluding both effluent and volatile organic compounds (VOCs)
regulation. In the past, designs for the containment liners have
incorporated concrete, a rigid material that is susceptible to both
cracking and spalling, thus creating leakage problems. Concrete
surfaces become slippery when covered with water and HP-4
(creating safety problems), prevent maintenance access to what-
ever they cover and provide no fuel retention capabilities on
windy days. Asphalt, when used near the fire area, ignites from
high radiant heat.
The new technology described here deals effectively with these
past inherent problems. The system itself is described in a fully
developed, site adaptable design package that includes a flexible
double liner containment pit, a leak detection technique, back-
fill layers, oil/water separator and an effluent holding system.
The design also incorporates a closed process configuration, en-
hanced safety factors, increased longevity and low initial and
life cycle maintenance costs. The result is an environmentally ac-
ceptable training unit.

BURN PIT LINER SYSTEM
The circular burn area provides optimum access to the entire
pit from the perimeter. The pit design consists of several layers of
material to provide containment of pollutant, protection of con-
tainment liner and an operating surface.
Possibly the most prominent containment characteristic is
found in the two flexible membrane liners of high density poly-
ethylene (HDPE), separated by a drainage net of the same ma-
terial. Selected to reduce the possibility of leaks caused by cracks,
the liners will be protected from heat during a fire by a layer of
stone and water. Further protection is offered by geotextiles and
sand, thus helping to ensure a 20- to 25-yr liner life; furthermore,
the liner is highly resistant to both chemical exposure and sun-
light. Unlike rigid concrete, HDPE is flexible and flexes with
earth movement. Thus, the chances for effluent discharge
through leakage are greatly reduced. Finally, the liner is easily re-
pairable, and its thickness (80 mil each layer) is more than twice
the U.S. EPA's current minimum.
The HDPE liner also provides a safety factor against the effect!
of construction and possible settlement (Fig. 1).
Figure 1
HDPE Liner Detail
LEAK DETECTION
If a leak develops in the upper liner, helium detection tech-
niques can be used to locate the leak for repair. Helium injected it
a low pressure into the screen between the liners will flow upwiti
through porous sand and crushed rock to the surface where it «•
be detected. This technique will permit evaluation of integrity of
liner system, provide for field weld repairs and measure effective-
ness of repairs accomplished.

BACKFILL LAYERS
The liners and burn pit area are enhanced and protected by
sand, geotextiles, stone and stabilized aggregate. JP-4 fuel the)
floats on top of a thin film of water covering the aggregate, emtf-
84 INNOVATIVE TECHNOLOGIES
-------
ing a final surface that offers stable footing, ease of mainten-
ance and the capability for efficient residue removal.
A layer of sand provides liner protection from the sharp edges
of the crushed stone and serves as a protective buffer zone if stone
removal or replacement becomes necessary. A geotextile filter
fabric separates the sand and coarse stone, thus preventing sur-
face settling caused by the sand working into voids. A layer of
well-graded crushed stone is used for the burn area surface. The
interlocking effect of the angular crushed stone provides a stable
surface for secure footing during training exercises and a porous
structure for rapid pit drainage. This surface also is mainten-
ance free and tends to hold the JP-4 in place on windy days.
Stabilized aggregate is used to backfill the pit's sloping perim-
eter. This uniformly graded material with a high percentage of
fines prevents fuel from floating into the pit's perimeter and con-
tains the fire within the designated burn area. Aggregate is one
of the few available materials that can tolerate the fire area's high
temperatures (1800-2400 ° F).

SEPARATOR AND EFFLUENT HOLDING POND
The three main system components outside the actual burn area
are a drainage trench, an oil/water separator and an effluent
holding pond. These units help ensure cleansing of the burn sur-
face, efficient separation of the oil/water residue and efficient,
environmentally sound reuse of both JP-4 and water (Fig. 2).
MAJOR COMPONENTS Of THE LIVE FIRE TRAINING FACILITY
BURN PIT WITH
MOCK AIRCRAFT
VEHICLE
MANEUVERING
WASTEWATER
HOLDING BASIN
Figure 2
Major Components of the Live Fire Training Facility
pie open tank, three-chamber, gravity type with a provision for
removing residual fuel from the first chamber for later use. This
design has demonstrated a capacity to reduce total oil and grease
in the effluent to a level acceptable for discharge into the holding
pond. The separated JP-4 fuel and foam are skimmed off the top
of the first chamber by raising the water level in the entire sepa-
rator. This unburned JP-4 fuel and residual foam will be pumped
to a holding tank. When sufficient quantities have accumulated,
a pump will return the liquid to the burn pit through a spray noz-
zle. The return spray system is activated only after the pit fire has
fully developed. Thus, JP-4 waste is significantly minimized.
The effluent holding pond provides a parallel, efficient sys-
tem for reuse of the water residue. Here, burn pit effluent col-
lects for evaporation and biological degradation. To contain all
pollutants, the pond is double lined with the same HDPE ma-
terial as the burn pit.

CLOSED PROCESS SYSTEM
The integrity of the liner configuration and the necessity for
reusability make this a closed process system. Even the water to
refill the pit conies from the storage pond. Water is added only if
evaporation exceeds rainfall and is discharged to permitted dis-
charge when the converse is true. And although virgin JP-4 starts
the fire, reclaimed JP-4 fuel is injected to increase or prolong the
blaze.

SAFETY
The burn surface is much safer. Walking is facilitated by the
stones, which do not retain the JP-4 as does concrete. This sur-
face also does not crack or spall to destruction. Another safety
improvement, an electrical ignition system, eliminates the need to
manually ignite the fuel with a hand held torch.

EQUIPMENT: COST AND MAINTENANCE
This system is relatively cheap to install and maintain. The
HDPE structure allows flexibility and strength—thus, fewer
problems to begin with—and areas for repair are easily located
and accessible; any repairs are inexpensive to make. Steel piping
is used, and only the water and fine aggregate are exposed to the
fire.
The durable liner meets all current and expected safety and
construction quality regulations. All underground valves are in
boxes and, along with the pumps and electrical equipment, are
standard and readily available.
The fuel supply system provides JP-4 to the burn pit, which is
divided into five zones. The JP-4, which can be fed to all zones
simultaneously or to any zone, is ignited by T-33 or turbine en-
gine spark plugs. This configuration allows optimum control of
bum placement while alleviating the present hazardous practice
of manual lighting with a torch.
After a burn, an outlet drain at the burn area edge allows the
JP-4 residue (and, thus, VOCs associated with the fuel) and fire
fighting foam used in training to be drained off of the stone sur-
face by raising the water level inside the pit. The runoff is fed by
'gravity through a trench to the oil/water separator which is a sim-
CONCLUSION
The live fire facility described here presents an innovative liner/
containment system that is initially inexpensive, highly reliable,
efficient, easy to maintain and service and has a long life expec-
tancy. The design presently exists, and prototypes are being site
adapted and tested to verify function across a spectrum of Air
Force commands, site requirements and geographical areas.
Finally, it is an environmentally acceptable system that depends
on both structural integrity and minimum waste through fuel and
water reuse to ensure accomplishment of military purpose while
adhering to safety and efficiency standards.
INNOVATIVE TECHNOLOGIES 85
-------
Composting Explosives Contaminated Soil

Richard C. Doyle, Ph.D.
Jenefir D. Isbister, Ph.D.
George A. Anspach
Atlantic Research Corporation
Alexandria, Virginia
David Renard
U.S. Army Toxic and Hazardous Materials Agency
Aberdeen Proving Ground, Maryland
Judith F. Kitchens, Ph.D.
ABSTRACT
Laboratory and pilot-scale studies were conducted to determine
if sediment contaminated with TNT, RDX, HMX and tetryl from
Louisiana Army Ammunition Plant (LAAP) and sediment from
Badger Army Ammunition Plant (BAAP) containing nitrocellu-
lose could be decontaminated by composting. Laboratory stud-
ies utilized '^C-tracers to follow the breakdown of explosives in
two types of compost (hay-horse feed and sewage sludge-wood
shavings) with contaminated sediment added at 10, 18 and 25%
of the compost mass. TNT and RDX degraded rapidly at the 10%
level, but breakdown was inhibited by increased sediment load-
ing. HMX breakdown was relatively slow. Both tetryl and nitro-
cellulose degraded rapidly and were not adversely influenced by
the higher rates of sediment loading.
Pilot-scale composting tests were conducted using 500 gal self-
sustaining composts. In the hay-horse feed, composts which con-
tained 11% LAAP sediment losses of explosive chemicals fol-
lowed firstporder kinetics with half-lives of 1.6, 3.0 and 4.7 wk
for TNT, RDX and HMX, respectively. Breakdown of explosives
in sewage sludge compolsts amended with 16% LAAP sediment
was insignificant. A third mixture of compost materials (manure-
hay-saw dust) amended with 12% LAAP sediment was tested.'/2
First order half-lives for TNT, RDX, HMX and tetryl were 1.0,
2.5, 3.3 and 1.2 wk, respectively. Nitrocellulose in the BAAP sed-
iment was rapidly broken down: 100% decontamination within
3 wk in hay-horse food composts, and 92-97% decontamina-
tion in the sewage sludge composts within 4 wk. Loss of ex-
plosives or heavy metals in compost leachates was low in all tests.

INTRODUCTION
Operations at a number of military installations have resulted
in the release of hazardous materials into soil and sediments.
These released substances include a variety of explosives, sol-
vents, pesticides, heavy metals and other organic and inorganic
materials. Concentrations of contaminants in the soil vary from
low ppm levels up to levels where more than one-half the weight
of the soil is composed of pollutants.
In some situations, contaminants have leached into ground-
water and are migrating toward potable water supplies. A num-
ber of these heavily polluted sites, either threatening or poten-
tially threatening public water supplies, are primarily contam-
inated with explosives.
Current cleanup technology at these sites is limited to excava-
tion and incineration of the soil, a process which is prohibitively
expensive. Biological treatments generally do not effectively
destroy most explosives. However, previous laboratory testing1'2
demonstrated that composting effectively degrades some of the
most common explosives used by the military (TNT and RDX).
Composting has significant advantages over incineration, in-
cluding being substantially less expensive. It is relatively easy to
set up and run a composting facility. The compost treated soil is
enriched by the composting process, rather than being rendered
useless by incineration, and can be backfilled and revegetated on-
site.
Previous composting work was limited to using a "clean" soil
spiked with up to 20,000 ppm of TNT or RDX. The current stud-
ies were undertaken to determine the utility of composting to re-
claim soils containing a number of explosives, as well as other
contaminants, with concentrations of explosives as high as 50%
of the soil weight. Actual contaminated sediments from Army in-
stallations were used for all tests.

MATERIALS AND METHODS
Contaminated Sediments
Sediments from lagoons previously receiving process waste-
water were collected from two U.S. Army installations. Sediment
from Louisiana Army Ammunition Plant (LAAP) was contam-
inated heavily with TNT (1,3,5-trinitrotoluene) and RDX (hex-
ahydro-l,3,5-trinitro-l,3,5-triazine) with lower level contamina-
tion of HMX (octahydro-l,3,5,7-tetranitro-l,3,5,7-tetrazocine),
tetryl (N-methyl-N,2,4,6-tetranitrobenzenamine), chromium and
lead. Lagoon sediments from Badger Army Ammunition Plant
(BAAP) contained nitrocellulose as the sole known pollutant
Separate samples of sediment were used for the laboratory and
pilot-scale studies. Concentrations of explosives in these samples
are given in Table 1.

Laboratory Composts
Separate experiments were set up to study the fate of individual
explosives (TNT, RDX, HMX and tetryl) in the LAAP sediment
and of nitrocellulose in the BAAP sediment during composting.
In each experiment, a measured quantity of one of the explosive*
containing a 14C-label was added to the LAAP or BAAP sedf-
ment. This spike did not measurably change the concentration of
explosive in the sediments (except for nitrocellulose where the
concentration was increased by approximately 20%), but did in-
troduce the l^C-label which was utilized to follow breakdown of
the single explosive.
86 INNOVATIVE TECHNOLOGIES
-------
Table 1
Concentrations of Explosives in Sediments Used in the
Composting Studies
-BlS.
THT
RDX
LAAP Uborxtory Co«po»ts IU.600 66.205 7,0*3 5,023
IAAP Pilot Coopasta 322.735-660.388 62,405-67.769 8,650-9,001 3,961-17.076
HAAP l^borctory ConpoBtB 10.252
RAAP Pilol r«»pn»t« 69,950-56,382

Each experiment was set-up with three independent variables.
The first variable was sampling time; compost samples were
extracted and analyzed for explosives immediately after being
set up (time zero) and after a given period of composting. The
second variable was compost materials; two types of composts,
alfalfa hay-Purina Sweetena horse feed and domestic sewage
sludge-wood shavings, were utilized. The third variable was sed-
iment loading level; the contaminated sediment was added at
levels equivalent to 10, 18 and 25% of the total dry weight of the
compost. Duplicates of all experiments were run except the time
zero composts which had only one replicate. Each compost con-
sisted of 55 to 90 g of dry matter.
All the composts were connected to an aeration system where
carbon dioxide-free air was pulled through each compost. The
gases given off by each compost pile were individually scrubbed
by activated carbon and solutions of sulfuric acid and sodium
hydroxide. All composts were incubated at 60°C for 6-10 wk.
Detection of the 14C-label was utilized to follow the degrada-
tion of the explosives. An outline of the analytical procedures is
shown in Fig. 1.
"C-COWOST
VOLATILE
LOSSES
NITROCELLULOSE
ANALYSIS USING
COL011
XETHOD
SOLVENT EXTRACTION
1AC-ACTIVITY by LSC I
conrosT SOLIDS
KIND (0.050 IN.)
CONCENTRATION
Figure 1
Outline of Laboratory Compost Analyses
The data were analyzed using a two-way analysis of variance
with compost materials and sediment loading rates as main fac-
tors. Differences between means were assessed where appro-
priate, using the Student-Newman-Kuels multiple range test.
Testing was at the 5% level of probability.
Pilot-Scale Composts
Pilot-scale testing was conducted in 500-gal cylindrical stain-
less steel tanks. The composters were fitted with an aeration sys-
tem to draw air down through the composter and to collect leach-
ate from the bottom of the tanks (Fig. 2). The rate and frequency
of aeration were monitored and regulated to maintain optimum
composting conditions. Composts were watered as needed and
were mixed periodically to assure uniform composting through-
out each tank.
^
1.1*1

HlLj-LJjJ'F
O^~ ^— ^\^~ — v.
/ S r^
^-A, ,~^
V r/^ '^'(r
_^ ' — — ' - K \
'
•C?.\ '.'.«/• '.'•'••' •"•••.••:•
r//h
'
IT
: r
1
I -
1
ft 1
( 1
- - H
ill ••HI
1
-U

r tie nn

- .• .ILKIf
*•• 11* •••
J1"- '

Figure 2
Schematic of Pilot Scale Compost Apparatus. Arrows Indicate the
Flow of Air.
Both the LAAP and BAAP sediments were composted in the
pilot-scale apparatus. Two types of compost material were tested:
hay-horse feed and sewage sludge-wood chips. Coarse wood
chips were mixed with the sewage sludge in place of the wood
shavings used in the bench-scale tests. A single rate of sediment
addition was used for each treatment combination. This loading
rate was based on the results from the laboratory trials. The
LAAP sediment was mixed into the hay-horse feed compost at
11% of the total dry weight and was added to the sewage sludge
compost at 16% of the mass. BAAP sediment was incorpo-
rated into both types of compost at the 15% rate. Duplicate
treated composts and unreplicated control composts were set up.
Composts were maintained for 4 to 7 wk.
Results from the LAAP composts indicated that a very high
activity compost was necessary for the biodestruction of the ex-
plosives. To verify this conclusion, an unreplicated compost con-
sisting of horse manure, hay and sawdust with 12% LAAP sedi-
ment was set up and composted for 8 wk.
The LAAP composts were sampled every 7-10 days. A large
composite sample was removed from each compost and then
replicate subsamples were taken for moisture and explosives
analysis. Leachate samples were collected daily and pooled by
week for analysis. Concentrations of all explosives were quanti-
fied by HPLC. Heavy metals in the leachate were determined by
atomic absorption spectrometry.'
Preliminary tests with BAAP composts indicated that large
quantities of substances which interfere with the nitrocellulose
analysis are produced during the initial stages of composting.
Therefore, BAAP composts were sampled after 0, 3 and 4 wk of
composting. Leachate samples were collected daily throughout
the experiment. Nitrocellulose levels were estimated by quantify-
ing non-ionic nitrate levels using a colorimetric procedure.3
Final composts from all hay-horse feed and sewage sludge-
wood chip composts were extracted using the U.S. EPA EP pro-
cedure (40 CFR 261 Appendix II). These extracts were subjected
to the Ames assay for mutagenicity using five Salmonella tester
strains with and without S-9 activation/
INNOVATIVE TECHNOLOGIES
87
-------
Table 2
Distribution of 14C in LAAP Sediment Composts Spiked with
Rlng-UL 1*C-TNT

iwr

«KF

01
n
sz
DZ
IZ
SZ
01
IZ
11
01
IZ
sz

•I .
91.
(Jl
•9
19
9| .
u o.;t." 7.
16 O.lb 10.
I* 0.9b 21.
W 7 . *. 1 .
It 1.9. 1.
U 11. 2
1 TKI
•4
(4
90.
•0.
77.
IS.
1.
21.
1*.

0.
lUiid
9
5.
II.
1.
7.
6.
b 100
i 75
• 77
Db*" 9*
W> 97
Jb 97
Ml Total
97.
98.
101.
98.
96.
105.
1* 101.
2« 106.
•• 100.
** 98.
4. 100
b. 10!
• HHF-hay-horse feed composu; SS-sewage sludge composts.
•• Values followed by the same letter are not significantly different at the 5^ level of probability
according to the Student-Newman-Kuel multiple range test. Data from samples composted
zero days were excluded from statistical analyses.
••• Nol detected.
RESULTS

LAAP Laboratory Composts
All explosives in the LAAP sediment were degraded by com-
posting, but the rates of breakdown varied significantly under
laboratory conditions. Loss of uniformly ring-labeled 14C-TNT
was most rapid. During 36 days of composting, more than 99%
of the 1^-TNT was lost in the sewage sludge composts. Average
losses in the hay-horse feed compost ranged from 79 to 98%
where the rate of TNT loss was apparently inhibited by the higher
rates of sediment loading. A summary of the 1*C distribution
in the 14C-TNT composts is presented in Table 2.
Accumulation of TNT transformation products (normally
found in environmental samples) was minimal. Less than 2% of
1*C label was associated with any one of the amino, diamino
or azoxy derivatives of TNT. No major metabolite of TNT de-
gradation was isolated. Most of the 14C from TNT losses was
recovered as unextracted residues. Losses of l^COj generally
were low «2.4% average value), indicating little or no cleavage
of the benzene ring in TNT. Other volatile losses of l^C were
inconsequential.
LAAP sediment spiked with uniformly ring-labeled RDX was
composted for 70 days. RDX was substantially reduced in all
composts, but the rate of breakdown appeared to be inhibited by
increased rates of sediment addition. This inhibitory effect was
readily apparent in the hay-horse feed composts. In the sewage
sludge compost, average losses of RDX were decreased somewhat
by increased sediment addition, but these differences were not
statistically significant (Table 3).
In composts where more than 50% of the RDX was degraded,
a substantial quantity of the J4C was evolved as 14CO2, demon-
strating that the breakdown of the RDX molecule was extensive.
The radiolabeled carbon from degraded RDX that was not vola-
tilized as 14CC>2 was largely recovered as bound residue (not
solubilized by the solvent extraction). Volatile losses of l^C from
the compost, other than 14CO2, were insignificant. No extract-
able degradates built up in the compost.
The degradation of uniformly ring-labeled HMX in compost
was relatively slow. The results of 14C-HMX breakdown in com-
posts are summarized in Table 4. Interpretation of the results is
somewhat difficult due to contamination of the '^C-HMX spik-
ing solution with 14C-RDX. Approximately 30-50% of the HMX
in the sewage sludge composts was lost during composting. The
breakdown of HMX appeared to be related to the rate of sedi-
ment addition to the compost, but differences in HMX degrada-
tion between sediment loading rates were not significant at the
5% level of probability. No HMX breakdown was observed in
the hay-horse feed composts.
Degradation of the l^C-RDX contaminant in the sewage sludge
composts was similar to, but somewhat slower than, that ob-
served in the composts spiked with l^C-RDX. The breakdown of
RDX in hay-horse feed composts was slow at all levels of sedi-
ment addition (less than 25% breakdown in 70 days).
In this study, it was not possible to distinguish between 1unL of
dlMDt
dd*d Co«p<

OZ
IZ
SZ
OZ
91
SZ
OZ
SI
sz
OZ
IZ
51 7

• tint

---
...
...

46.S«**
4.0c
O.lc
}7.2a
1 2*. 2b
26. Ob
i*c

94.0
16.5
83,1
91.1
89.2
98.6
5.6b
70 ?•
h9.S«
II. Ob
IS.Sb
I5.4b
R.cov.r.d (X of .«.<)

K.lr»ct
KDX IMt
90.1 1.6
m.s 1.1
80. 1 1.1
89.7 1.4
»S.B 1.7
92.1 4.9
4.6b O.lc
67.1 1.1.
fci.a J.l.
1.7 l.4bc
II. 8 1.9k
21.2 i.e..

UnuiratlM
*.*idu.i

IZ.6
U.4

2S.O
9.6
29. 7b
24. *b
21. Ib
S4.»«
U.I*
46.6*

TM«)

91.)
91*

114 I
l«.l
II. 1
M,i
H.?
101.1
IN 1
•1,1
* HHF-hay-horw feed compost; SS-sewage sludge composts.
*' Values followed by same letter are not significantly different at the 5% level of probability
according to the Student-Newman-Kuel multiple range test. Data from samples compacted
zero days were excluded from statistical analyses.

Table 4
Distribution of 14C in LAAP Sediment Composts Spiked with
Ring-UL 14C-HMX

Aoount of
StdlMnt
Common* (dry vt Z)
HHF OZ
II
SZ
SS QZ
8Z
SZ
HKF OZ
8Z
SZ
SS OZ
IZ
sz
'*C R.CoV.r.d (Z o

Tlw (d.ym) ' *CO j Eilr«cl R.
0 --- ai.l 1 .
17.5
79 R 1 .
81,0 I
76.6 1 .
81.7 1 .
70 0, b**» 76. 8« 19.0
70 0. b 74.8. 20.
70 0. b 7| 6b 27.
70 ]| . • 27 Id 42.
70 U b }*.2d %0.
70 11. b 41. Ic 41.
«dd«d J

I duel Tol
91.
96.
95.
91.
96.
b 96.
b 94.
• 101.
. 99.
* 100.
Maru1ll«4 l«BMrT{I]"

1 WO tOi

101. 1* •>-**
91.4* II 1»
94.1* tU
49.2k I1.»*
si. ik »*•
69. 7h «*•
• HHF-hay-horsc feed composts; SS-sewage sludge composu.
•* HMX and RDX normalized to average recoveries from time zero composts.
"• Values followed by the same letter arc not significantly different at the Wt level of prob-
ability according to the Student-Newman-Kuel multiple range test. Data from umple> com-
posted zero days were excluded from statistical analyses.
Uniformly ring-labeled 14C-tetryl was composted for 44 days.
Loss of tetryl was substantial and did not vary significantly with
composting treatments (Table 5). Extracted l^c-tetryl accounted
for an average of 3.7 and 2.6% of the total radioactivity added to
the hay-horse feed and sewage sludge composts, respectively.
However, some error in the final tetryl levels resulted from inade-
quacies of the extraction and TLC methodologies. Extraction
efficiency, as indicated by 14C extracted from the time zero com-
posts, was low (42%-68% recovery) in the sewage sludge com-
posts, and tetryl recovery from the 44 day composts may be ami-
88 INNOVATIVE TECHNOLOGIES
-------
larly low. The TLC procedure used for the compost extracts did
not separate l4C-tetryl from its primary impurity (unidentified)
found in the spiking solution (7.5% of the total activity). Thus,
it is not possible to determine if a portion or all of the 14C re-
ported as tetryl was tetryl or an impurity. Despite these errors,
90-100% loss of tetryl was demonstrated.
Degradation of tetryl was accompanied by substantial increases
in the bound *4C residues. Volatilization of 14CO2 was low,
and because of the impurities present in the spiking solution it
cannot be determined if the 14CC>2 evolution resulted from tetryl
breakdown or from the degradation of the impurities. Several
14C compounds were separated out of the solvent extract, but no
one of these unidentified compounds accounted for more than
0.6% of the total 14C.

BAAP Laboratory Composts
Composts amended with the BAAP sediment composted ex-
tremely well. Compost temperatures ranged from 66 to 77 °C in
an incubator held at 60 °C. After 44 days of composting, the
composts were very dark in color and substantially reduced in vol-
ume. Elevated temperatures of this magnitude in composts con-
sisting of < 70g of organic materials demonstrated that the BAAP
soil or some component(s) of the soil significantly enhanced the
composting process.
Table 5
Distribution of 14C LAAP Sediment Composts Spiked with
Ring-UL 14c-Tetryl

Coopost*
UHF

HHF

Aaount of
Sedlaent
Added
(dry ut Z)
10Z
18Z
25Z
10Z
18Z
25Z
10Z
18Z
25Z
10Z
18Z
25Z

CoMpostlnf
Tin (days)
0
0
0
0
0
0
44
44
44
44
44
44

"C02
...

---

5.0
3.0
2.1
3.2
3.4
3.9

Extract
84.0
78.4
81.8
45.6
56.9
68.1
0.1
4.6
7.6
3.2
4.3
1.8
14C Recovered
Tetryl
83.7
77.2
81.1
41.9
53.1
64.4
ND«*
4.0
7.1
2.7
3.8
1.2
(Z of Added)
Unextracted
Reildual
11.3
8.6
9.4
50.4
24.9
21.1
84.2
89.2
90.1
92.6
96.0
96.5

Total
95.3
87.0
91.2
96.0
81.8
89.2
89.
96.
99.
99.
103.
102.
* HHF-Hay-horse feed compost; SS-sewage sludge compost.
" ND = Not detected by TLC analysis.

70 -i
60-
O
o
in
o
ui
40-
o
IU
o
30-
20-
10-
HHF-1 8
10
20 30
TIME (DAYS)
40
Figure 3
14CO2 Evolved from BAAP Composts Spiked with Uniformly
Labeled 14C Nitrocellulose
Breakdown of the uniformly labeled 14C-nitrocellulose began
within the first week of composting as indicated by the release of
14CO2 (Fig. 3). Initial 14CC>2 evolution rates were very high.
After 3-4 wk, the rates of *4CO2 release began to decline. The
cumulative losses of 14C from the composts as 14CC>2 accounted
for 43-74% of the total activity added to individual composts.
Small quantities of ^4C were recovered in the acid traps and in
water condensates from the hay-horse feed composts.
After composting for 43 days, very little of the 14C could be
extracted into acetone. Essentially all of the 14C activity re-
mained as bound residue in the compost or had been lost as
14CO2- Analyses of the extracts for nitrocellulose showed that
less than 1.5% of the explosive initially added to the composts
remained after composting.

Pilot-Scale LAAP Composts
The LAAP hay-horse feed and sewage sludge composts were
the first to be tested in the pilot-scale composters. Minor prob-
lems in composter design and composting practices resulted in
less than optimum composting conditions for the first 4 wk of this
7-wk study. These problems caused areas of the compost within
any one composter to dry rapidly, and then the dry areas cooled
to below thermophilic temperatures.
Despite less than ideal composting, loss of all explosives in the
hay-horse feed compost was substantial. Rates of explosive de-
cline conformed to first order kinetics, and half-lives were cal-
culated for TNT, RDX and HMX. The TNT degraded most rap-
idly with an average half-life of 1.6 wk. Losses of TNT were in-
itially accompanied by small increases in the 2-amino and 4-
amino dinitrotoluene transformation products. However, neither
of these products accumulated in the compost, and both com-
pounds decreased after the second week of composting.
RDX degraded with an average half-life of 3 wk. HMX break-
down was the slowest, with an average half-life of 4.7 wk. Tet-
ryl was only quantified in the starting materials and in the 7-
wk composts; therefore, the rate of breakdown cannot be fitted
to a kinetic model. After 7 wk of composting, tetryl levels were
significantly reduced. Less than 7% of the tetryl was discovered
in the 7-wk composts.
The use of sewage sludge and woodchips as composting ma-
terials was unsuccessful. The LAAP sediment inhibited the com-
posting process. This inhibition coupled with the aeration prob-
lem discussed above resulted in very poor composting. No loss
of explosives during composting could be confirmed. Sampling
the sewage sludge composts was difficult, and variation between
subsamples was high. Concentrations of explosives varied some-
what randomly during the 7-wk composting period. No signifi-
cant differences in the explosives concentration over the 7-wk
period could be detected by analysis of variance testing at the
5% level of probability.
Ames testing of extracts of the final products from both the
sewage sludge-wood chip and the hay-horse feed composts did
not give any positive results.
All technical problems in aerating the pilot-scale composters
had been solved when the horse manure-hay-saw dust composts
were tested. However, mixing the compost on day 10 of the study
upset the system, and compost did not recover (reach thermo-
philic temperatures) until day 19 of the study. No further com-
posting difficulties were encountered during the 8-wk study.
Explosives levels in the compost decreased rapidly, approxi-
mating first order kinetics. TNT, tetryl and HMX concentrations
decreased substantially within the first 10 days of composting.
RDX levels remained relatively constant during the first 19 days
of composting and then dropped dramatically as the microbial
activity in the compost increased after day 21 of the experiment.
INNOVATIVE TECHNOLOGIES 89
-------
Half-lives calculated assuming first order kinetics are listed in
Table 6. The relative rates of loss of explosives were similar to
those found in the hay-horse feed composts, but overall decon-
tamination was noticeably faster in the manure compost. After
56 days of composting, RDX levels were reduced 94% and HMX
was decreased by 81%. Tetryl levels were reduced 91% within 27
days; after 31 days, concentrations were too low to be quanti-
fied. The TNT concentration was reduced from 31,021 ppm in
the compost at time zero to 138 ppm in 56 days, a 99.6% reduc-
tion. The 2- and 4-amino-dinitrotoluenes did not accumulate in
the compost as the result of TNT loss. Collectively, these amino
derivatives decreased 60% during the first 10 days of composting,
remained constant through the 27th day of composting and then
decreased to 124 ppm (97% overall decrease) by the 56th day of
composting (Fig. 4).
3S.OOO- -
10,000
35.000
20,000- -
• TNT

• 2- * 4- AMINO DNT'S
15.OOO- -
10.00O- -
5.OOO- -
Figure 4
Concentration of TNT and its Amino Transformation Products in
Manure Compost as a Function of Composting Time

The loss of explosives in compost leachate was very low, as
expected, given the low solubility of explosives in water. Interfer-
ence in the HPLC analyses of the leachate resulted from natural
organic products leached from the composts. This interference
could have falsely inflated the apparent explosives levels detected
in the leachate. Although accurate quantitation of leachate losses
was not possible, it is known that some explosives did leach from
the composts and that the amount leached was close to neg-
ligible.
Metal concentrations in the composts varied with the materials
used. The hay, horse feed and manure contained levels expected
for natural uncontaminated materials. Sewage sludge contained
moderately high levels of copper, zinc and lead. The LAAP sedi-
ment had elevated levels of lead and chromium, with low levels
of mercury. Generally, the leachate was alkaline and the leach-
ing of metals was limited to relatively low concentrations of cop.
per and zinc. These metals readily react with ammonia to form a
water-soluble species under basic conditions.
Table 6
First Order Half-Lives of Explosives in LAAP Composts
Half-Life (Weeks)
Compost
Hay-horse feed
Sewage Sludge
Woodchips
Mamire-Hay-Savdust
TNT
1.6
ND*
1.0
RDX
3.0
ND
2.5
HMX
4.7
ND
3.3
Tetrvl

ND
1.2
' No significant degradation.
Pilot-Scale BAAP Composts
As was observed in the laboratory studies, hay-horse feed com-
posts containing BAAP sediment composted very rapidly with ex-
cessive amounts of heat generated. Once thermophilic tempera-
tures were reached, temperatures remained in the 62-80 °C range.
Very high aeration rates were utilized to prevent temperatures
that would lead 'to charring. Under these conditions, nitro-
cellulose was rapidly degraded. Initial concentration of 6,655 tol
8,176 ppm in the compost were reduced to below the detection
limit (25 ppm) within 3 wk.
Composting of the sewage sludge-woodchip-BAAP sediment
mixture was marginal, with temperatures being maintained near
60 °C. Analysis of these composts was difficult due to interfer-
ence from the organic acids in the compost; as a result, the data
were highly variable. Despite this, it was possible to demonstrate
rapid nitrocellulose breakdown (greater than 92% degradation
within 4 wk).
Ames testing of both the hay-horsefeed and sewage sludge-
wood chip compost extracts did not yield any positive results, in-
dicating that the final product of the composting treatment is not
mutragenic.
Leachate from the BAAP composts did not contain any detec-
table nitrocellulose. Small quantities of copper and moderate
amounts of zinc were found in the leachate.

CONCLUSIONS
Composting is an effective means for decontaminating sedi-
ments and soils containing high levels of explosives. Although
different explosives degrade at different rates, all explosives tested
(TNT, HMX, RDX, tetryl and nitrocellulose) were broken down
rapidly enough to make composting a practical and cost-effec-
tive treatment. Laboratory studies demonstrated that no degrada-
tion compounds with structures similar to the parent explo»v«
accumulate during explosive destruction. Both RDX and nitro-
cellulose were completely metabolized, with most of the carbon
being released as CO2- Ames testing of EP extracts of the final
products reinforced the conclusion that the finished compost dott
not have any properties of a hazardous waste.
Rates of explosive destruction appeared to vary directly with
the microbial activity of the compost (measured by compost tem-
perature). Highly active composts maintaining high thermophilic
temperatures (>70°Q were most effective in rapidly degrading
the explosives. Selection of composting materials is, therefore,»
critical factor in obtaining efficient waste destruction. Material!
90 INNOVATIVE TECHNOLOGIES
-------
such as sewage sludge and wood chips are biologically too stable
to provide for accelerated composting. Fresh manure and other
materials (such as some food processing wastes) with a high con-
tent of readily available energy sources for bacterial growth are
more suitable for constructing highly active composts that can be
maintained for a sufficiently long period to ensure complete
waste destruction.
Some sediments or the hazardous wastes in some sediments
may inhibit the composting process. In these situations, the sedi-
ment loading rate can be adjusted to minimize the toxic effects.
The selection of composting materials can also be utilized to re-
duce the inhibitory effects.
Loss of hazardous materials via compost leachate is not a sig-
nificant problem. Small amounts of some explosives as well as
copper and zinc may be solubilized in the leachate. However, a
well-managed compost does not produce much leachate, and the
amount of leachate typically generated can be recycled to the
compost to maintain appropriate moisture levels.
Composting as a waste treatment process has several advan-
tages over current explosives treatment technologies. Costs for
the cleanup of large areas of land by composting typically are
a small fraction of the costs incurred by conventional physical-
or chemical-based technologies. Composting facilities are not
complex structures and do not require any significant investment
in engineering design. Equipment used in composting consists of
off-the-shelf items which do not generally require major modif-
ications to be used in a hazardous waste composting facility.
Manpower requirements are minimal and the skill requirements
for most personnel are limited to operating the materials hand-
ling equipment (front-end loader, truck, etc.). The expendable
materials used for composting are usually waste products which
can be obtained at minimal or no cost and the end product can be
used as a soil conditioner. No materials need to be disposed of in
a hazardous waste landfill.

ACKNOWLEDGEMENTS
This research program was funded by the U.S. Army Toxic
Hazardous Materials Agency, Aberdeen Proving Ground, Md.
under Contract No. DAAK11-84-C-0057. We also acknowledge
the assistance of Mr. George Wilson who provided guidance in
managing the pilot scale compost.

REFERENCES
1. Isbister, J.D., Doyle, R.C. and Kitchens, J.F., "Composting of Ex-
plosives," Contract No. DAAK11-80-C-0027, U.S. Army Toxic and
Hazardous Materials Agency, Aberdeen Proving Ground, Maryland,
NTISAD-A119276, 1982.
2. Isbister, J.D., Anspach, G.L., Kitchens, J.F. and Doyle, R.C.,
"Composting for Decontamination of Soils Containing Explosives,"
Microbiologica 7, 1984,47-73.
3. Doyle, R.C., Isbister, J.D., Anspach, G.L. and Kitchens, J.F.,
"Composting Explosives/Organic Contaminated Soils," Contract
No. DAAK11-84-C-0057, U.S. Army Toxic and Hazardous Ma-
terials Agency, Aberdeen Proving Ground, Maryland, NTIS AD-
A169 994, 1986.
4. Ames, B.N., McCann, J. and Yamasaki, E., "Method of detecting
carcinogenes and mutagenes with the Salmonella/mammalian-mi-
crosome mutagenicity test," Mutation Res. 31, 1975, 347-364.
INNOVATIVE TECHNOLOGIES 91
-------
Use of Vapor Extraction Systems for In Situ
Removal of Volatile Organic Compounds from Soil

Magnus B. Bennedsen, P.E.
Joseph P. Scott, Ph.D.
James D. Hartley, P.E.
Woodward-Clyde Consultants
Walnut Creek and San Diego, California
ABSTRACT
Applying vacuum to contaminated soils, through extraction
wells or perforated pipelines, can effectively remove volatile
organic compounds from the soils. The process can be a cost-
effective remedial measure. Its performance can be enhanced with
appropriately designed and placed air inlet facilities to replace soil
gas removed by vacuum application to the contaminated soil
mass. The major costs associated with operating the systems are
for sampling and analysis of the extracted soil gas. Process emis-
sions controls, if required, are simple and derived from common-
ly available industrial systems.

INTRODUCTION
Throughout the United States, there are numerous instances
where volatile organic compounds (VOCs) were discharged, spilled
or leaked on the ground surface, or released below the surface
from buried tanks and pipelines. Many of the discharged com-
pounds were chlorinated, and quite a few are known or con-
sidered to be toxic. The most commonly perceived threat from
these spilled materials is contamination of groundwater beneath
the spill area, which may be extracted for potable use. Fig. 1
presents a schematic diagram of the contaminant pathway to
groundwater.
Because of the costs of groundwater remediation, there are
benefits to recovering spilled VOCs from soil before they enter
groundwater. However, releases are frequently not detected until
contamination is detected in groundwater, and even prompt re-
sponse to a large and sudden spill may not prevent transport of
VOCs to groundwater. At the other extreme, leaks which are un-
detected by carefully maintained flow metering and mass-bal-
ance-accounting systems may unknowingly continue for years be-
fore they are first detected in underlying groundwater.
In the past, the most commonly applied soil remediation pro-
cess was excavation, with the excavated soil either treated and
returned to the excavation area, or disposed of off-site in a haz-
ardous waste management facility. However, excavation is costly,
and may be only a partial solution, especially if contamination
extends beneath buildings, across property lines or to depths too
great to be practical for excavation.
This paper describes an alternate remedial procedure that,
where applied, has been economical and effective and which has
broad applicability to VOC contamination. The procedure in-
volves applying a vacuum to a VOC-contaminated soil mass,
thereby inducing a flow of air through the soil and removing
Ground surface
VOCs entering vadose zone
and spreading by gravity flow
and vapor diffusion through
soil to groundwater
Contaminated
groundwater
Figure 1
Schematic of VOC Contamination Problem
Soil gas extraction well
Air inlet well
Local raising of water
table surface due to
applied vacuum

Figure 2
Schematic of Soil Gas Vacuum Extraction System
vapor phase VOCs with the extracted soil gas. Vapor extraction
systems (VESs) are not universally applicable. Moreover, they
rarely yield a complete solution to a VOC-contamination prob-
lem and supplemental remedial measures may have to be en>-
92 INNOVATIVE TECHNOLOGIES
-------
ployed, particularly for highly soluble or less-volatile compounds.
This paper provides information that will be useful to persons
who must evaluate potential remedial measures for specific con-
taminated soils and sites. The decision to include a VES in a re-
medial program will be based on the estimated total costs for the
complete remedial program. The program evaluation should con-
sider the quantity of groundwater that must be extracted and
treated and the probable duration of the groundwater treatment
operation, with and without the VES included in the remedial
program. VES capital and operating costs are usually low and the
systems make use of readily available engineering technology.

DESCRIPTION OF THE VAPOR EXTRACTION
PROCESS
Applying vacuum to a soil mass is a simple process. Typically,
the vacuum is applied through one or more extraction wells,
which differ from conventional groundwater monitoring wells in
that the screened sections must extend above the water table sur-
face. If groundwater is at a shallow depth or if contamination
is confined to the near-surface soils, then the vacuum can be ap-
plied through horizontal perforated pipes buried in trenches. In
order to enhance air flow through zones of maximum contam-
ination, it may be desirable to include air inlet wells in the installa-
tion. Fig. 2 indicates how Inlet and extraction wells might be in-
stalled at a typical contaminated site.
To develop a vacuum, ordinary positive displacement indus-
trial blowers are used. Commonly, the blowers have ratings of 100
to 1000 ftVmin, at vacuum ratings up to about 10 in. Hg gage.
Ratings of the electric drive motors are 10 hp or less. The cost
for a typical blower assembly is approximately $5,000.
If air emissions control is required for an installation, a vapor
phase activated carbon adsorber system probably will be the most
practical system, although catalytic oxidation units have pro-
duced promising results in recent applications. Utilization of
available on-site resources almost always produces the most cost-
effective air emissions control; at one industrial installation,
where the initial VOC extraction rate was over 200 Ib/day, the
extracted soil gas was piped to the combustion air intake of a
nearby boiler that was in continuous operation.

VAPOR EXTRACTION SYSTEMS
APPLICABILITY
This section of the paper contains general guidelines that
should be useful for deciding if a VES is applicable at a site.
Most important are contaminant and site characteristics, al-
though cost is also important.

Character of Spilled Materials
A vapor phase vacuum extraction system will be an effective
remedial measure only for compounds that exhibit significant vol-
atility at ambient temperatures in the contaminated soils. At this
time, no rigorous criteria for identifying such compounds are
known to the authors. However, it is suggested that as a first
approximation, any compound exhibiting a vapor pressure of
about 0.5 to 1.0 mm or more of mercury at 20° C is a likely can-
didate. Another proposed screening tool is the air-water partition-
ing coefficient, expressed in dimensionless terms as Henry's Law
constant. Compounds which have values of Henry's Law con-
stants greater than 0.01 are also likely candidates for VES.
Examples of compounds which can be effectively removed by
VES include trichloroethylene (TCE), trichloroethane (TCA),
and most gasoline constituents. Compounds which are less ap-
plicable to removal by VES include trichlorobenzene (TCB),
acetone, and heavier petroleum fuels (e.g., diesel).
Dispersion of Spilled Material
The extent to which spilled VOCs are dispersed in soil—both
vertically and horizontally—will be an important consideration
in deciding if a VES is preferable to excavation. If only a few
hundred cubic yards of near-surface soils are contaminated, then
excavation probably will be the preferred remedy. But if the spill
has penetrated more than 20 or 30 ft, or spread at depth over an
area of more than several hundred square feet, then excavation
costs may begin to exceed those associated with a VES.

Character of Contaminated Soil
The air conductivity of the contaminated soils will control the
rate at which air can be caused to flow through the soils by the
applied vacuum. Grain size, moisture content and stratification
probably are the most important properties in this regard. In a
stratified soil there generally will be significant differences in the
air conductivity of the various strata. A horizontally stratified
soil is usually favorable to a VES installation—the relatively non-
conductive (impervious) strata will limit the rate of vertical inflow
from the ground surface and will tend to extend the applied
vacuum horizontally to useful distances from the point of appli-
cation.
Experience with VES installations indicates that even clayey
and silty soils may be effectively ventilated by the usual levels of
vacuum developed by the systems. The success of VES in these
soils, however, may be dependent on the presence of more con-
ductive strata (e.g., sand or gravel), such as would be expected in
alluvial settings or on relatively low moisture contents in the finer-
grained soils.

Site Characteristics
The location of contamination on a property and the type and
extent of development in the vicinity of the contamination, may
favor the installation of a VES. For example, if the contam-
ination is beneath a building or an extensive utility trench net-
work, extensive demolition might be required before the contam-
ination could be removed by excavation, whereas a VES might be
installed to remove the contamination without significant surface
disruption. VES can also be successful in recovering contamina-
tion from beyond the property line.
The depth to groundwater also must be considered when eval-
uating a potential VES application. Groundwater at a depth of
more than about 40 ft, in an area where contamination extends
to groundwater, will favor installation of a VES; whereas if the
groundwater is less than about 10 ft below the site surface, then
excavation may be a practical alternative.

GENERAL APPROACH TO DESIGN OF A
VAPOR EXTRACTION SYSTEM
The design of a VES starts with compilation and evaluation of
all site contamination and chemical data. The site information
should be placed on scale drawings—both plans and sections.
Then flow nets should be constructed on the drawings to indi-
cate the expected movement of air through the soil as a result of
applying vacuum through VES wells and/or trenches. This pro-
cess should result in determining the optimum locations for the
inlet and extraction compounds of the VES. Based on the con-
clusions obtained from the flow nets, the system compounds can
be designed and the vacuum equipment specified. Installation
and short-term operation of a part of the system may be neces-
sary, to provide information on the appropriate spacing of wells,
the VOC extraction rate and the need for emissions control. The
results of the short-term operation may then be used to design the
complete system.
INNOVATIVE TECHNOLOGIES 93
-------
MONITORING VAPOR EXTRACTION SYSTEM
PERFORMANCE
To determine if a VES is performing satisfactorily, it will be
necessary to establish criteria for satisfactory performance and
then make measurements to determine if the criteria are satisfied.
Establishing criteria for satisfactory performance will be a site-
specific task, related to regulatory requirements including the
type, quantity and dispersion of VOCs and the time-frame in
which VOC removal should be accomplished which may be a few
months to several years.
The basic measurements required to assess VES performance
are the system flow rate (in units of either volume or mass per
unit time) and the concentration of VOCs in the extracted flow.
A flow meter consisting of an orifice plate and manometer, to-
gether with the appropriate rating curve, will yield the system
discharge air flow rate.
Several techniques are available to establish the concentration
of VOCs in the extracted gas. The technique adopted should be
based on the quality of the data obtained and the cost and ease
of obtaining it. In general, a gas chromatograph equipped with
an appropriate detector for the compounds expected to be pres-
ent in the gas will best provide the necessary data.

COSTS
The costs to install and operate a VES will be site- and
design-specific and will vary widely between sites. A complete
blower assembly can be purchased for less than $5,000. How-
ever, the blower assembly generally will represent a minor part of
the total system costs. Permitting, piping, valving and instrumen-
tation, emissions control and performance monitoring typically
cost substantially more.
Some savings can be realized by using existing or planned
groundwater monitoring wells as part of the VES, provided the
wells include screening above the groundwater table. Also, on
some projects, soil borings that are required for the contamina-
tion investigation may be completed as VES air inlet or extrac-
tion wells. The opportunities for multi-purpose use of required
borings should be examined at the project planning stage. If sev-
eral wells are to be installed exclusively for VES purposes, they
will add significantly to the system costs.
The costs for routine operation of a VES are primarily asso-
ciated with monitoring system performance. Records should be
maintained of gas flow rates in the various system components
over time and of the concentrations of VOCs in the various gas
streams. The costs for sampling and analysis of the gas streams
will dominate the system performance monitoring costs.
If an air emissions control system must be included with a
VES, it will add significantly to system installation and operating
costs. The required capacity of the control may vary over a wide
range and the costs for installation will vary correspondingly.

REGULATORY AGENCY ACCEPTANCE
AND PERMITTING
Installation and operation of a VES at a hazardous waste site
will require prior approval by the appropriate regulatory agen-
cies. Most regulatory personnel have had no experience on pro-
jects where a VES was installed, and their reaction and response
to a proposal to install a VES may range from totally negative
to greatly enthusiastic. Agency personnel will probably approve
the VES proposal if they are presented with a well-conceived
and explained proposal to install a VES on a trial program basis
and they are offered the opportunity to witness the installa-
tion and operation, and to receive complete data from the opera-
tion.
The proposal given to the regulatory agency should include an
estimate of the VOC extraction rate, a discussion of how the rate
will be determined and the actions that will be taken if the esti-
mate is significantly in error. Also, the proposal should include a
discussion of how the test results will be used to design a full.
scale VES for site remediation.
CASE HISTORY
An electronics manufacturing facility in the Santa Clara Valley,
California, removed a spent-solvents storage tank, and discovered
that the tank leaked. Several chlorinated solvents were discharged
to the tank, but an estimated 80% of the total chemical lost was
1,1,1-trichloroethane (TCA). An initial boring beneath the tank
revealed a high concentration of solvents at a depth of about 40 ft
below the ground surface and groundwater at a depth of about 90
ft. Two additional borings were drilled on opposite sides of the
tank location. One of the borings was completed for operation as
a vacuum extraction well and the other boring as an air inlet well.
The soils in the area are predominantly alluvial clayey silts and
sands which are considered relatively impervious. In both wells
the casing diameter was 2 in.
A nearby building was operated under a continuous low vacuum
to prevent accumulation of solvent vapors in the work area. A
duct to the building ventilation blower existed a few feet from the
completed extraction well. A connection to that duct created a
vacuum at the well head of about 2.5 in., water gauge, and a gas
flow rate from the well of about 10 ftVmin. Analysis of the ex-
tracted gas showed that it consistently contained over 2000 ppm
of organics. A vacuum gauge connected to the air inlet well, with
the well capped to prevent inflow, gave a reading of about 0.15 in.
water gauge. The gauge-collected data indicated that the applied
vacuum of 2.5 in. water gauge was effective in inducing a flow of
air through the soil at a distance of about 30 ft from the extrac-
tion well.
A dedicated blower assembly was added to the system and the
extraction rate increased to about 100 ftVmin. Also, an addi-
tional extraction well was installed nearer the leak source and was
connected to the blower. With both wells in service, a vacuum of
about 3 in. Hg was sufficient to induce 100 ftVmin. flow through
the soil.
This system continues to operate. Over the first 3 yr of opera-
tion, approximately 12,000 Ib of VOCs were extracted. During
that time, the concentration of TCA in the extracted gas de-
creased from more than 2000 ppm to about 50 ppm. It is expected
that the system will be continued in operation until the total con-
centration of VOCs in the extracted gas is approximately 20 ppm,

CONCLUSIONS
Vapor extraction systems are effective for removing volatile
organic compounds from unsaturated soils. There are many
places where toxic or hazardous VOCs were released on soils and
resulted in contamination of underlying groundwater. At some of
these places installation and operation of a VES would be a cost-
effective and satisfactory remedial measure.
VES systems require installation of extraction wells or trenches
to apply the vacuum to the contaminated soil mass. The applied
vacuum causes a flow of air through the soil and the vapor phase
VOCs are entrained in and removed with the extracted soil gas.
Strategically located air inlet wells can enhance air flow through
the areas of greatest concern. As appropriate, emissions control
may be added to the system discharge.
The VOC extraction rate may be in the range of a few pounds
to several hundred pounds per day, depending on the extent and
character of the contamination, the soil characteristics and the
94
INNOVATIVE TECHNOLOGIES
-------
scope and character of the extraction system. Typically, the
system power requirements will be 5 to 10 hp, the flow rate of ex-
tracted soil gas will be 100 to 1000 ftVmin, and the applied
vacuum will be 2 to 6 in. Hg gauge. The major system-operating
cost will be for sampling and analysis of the extracted gas, as re-
quired to monitor and evaluate system performance. Normally,
the system will operate continuously and unattended, except for
routine data recording, system inspection and servicing.
It is recommended that before a full-scale VES is installed, a
partial system be installed and operated on a short-term basis, to
determine if a full-scale system should be installed, and to collect
data for designing the full-scale system. Regulatory agency review
and approval of VES installation and operating plans should be
obtained before the start of installation, and the agencies should
be kept informed on progress and performance of the system.

To conclude, a concept has been presented that the authors
hope will stimulate further thought and research on the role and
application of VESs. The work to date has clearly demonstrated
that a significant volume of air can be induced to flow through
soils by the application of a modest level of vacuum. This process
tends to provide oxygen to the soil gas; therefore, aerobic condi-
tions are likely to be created in the soils being ventilated. And
aerobic conditions are known to be a necessary but not sufficient
condition for biodegradation of many VOCs. Based on the aera-
tion benefit, one recognized ancillary effect ot^xisting VES pro-
jects is a partial stimulation of aerobic biological activity. Con-
ceivably, if nutrients, and possibly selected bacteria and water
vapor, were introduced to the soil through the air inlet facilities,
biodegradation of the residual organic compounds in the soil—
those that are sorbed to the soil and cannot be removed by volatil-
ization—may be degraded to a nonhazardous state. However,
the principal difficulties associated with inducing unsaturated
zone biodegradation appear to be the vapor phase transfer of
nutrients to the zones of interest and performance monitoring.
We encourage others to consider and investigate this opportunity.
INNOVATIVE TECHNOLOGIES 95
-------
Macroencapsulation: New Technology Provides
Innovative Alternatives for Hazardous Materials
Storage, Transportation, Treatment and Disposal
Mark D. Shaw
Bondico, Inc.
Jacksonville, Florida
ABSTRACT
Encapsulation of hazardous materials has been discussed in
theory for many years. Theories hold that a macroencapsulate
load-bearing container would provide an impermeable barrier im-
mobilizing the hazardous material from the environment for an
indefinite period of time.
Millions of steel, plastic and fiber drums are used each year to
store, transport or dispose of hazardous materials. Many of to-
day's environmental problems and regulations have risen out of
the failure of conventional drums to contain the toxic wastes.
While exciting and innovative technologies have been developed
to treat hazardous materials, no developments have been made on
the basic tool of the industry, the drum.
Two new encapsulation technologies have emerged that show
great promise in providing industry with a macroencapsulate con-
tainer which will enable the generator to safely store, transport or
dispose of hazardous and toxic wastes.

INTRODUCTION
The term "encapsulation" has been generally used to describe
two different processes best identified generically as macroencap-
sulation and microencapsulation. Macroencapsulation processes
utilize a load-bearing container of unique physical properties to-
gether with a sound closure technique which will encompass a
hazardous waste form and effectively immobilize the waste from
the environment. Microencapsulation usually is identified as a
solidification process in which individual waste particles are
suspended in the solidification matrix.
Two companies have performed considerable research and
development of two macroencapsulation technologies that are
similar in approach but different in technique.

Environmental Protection Polymers
Environmental Protection Polymers, Inc. of Hawthorne, Cali-
fornia, has been developing various encapsulation technologies
since 1977 under grants from the U.S. EPA Research and Devel-
opment branch in Cincinnati, Ohio. Some of the earlier efforts
were jointly performed with TRW Systems Group. Although
some of the technologies developed were found to be unsuitable
for commercial development, one patented technology has shown
potential. This encapsulation system is based upon spin-welding
or friction-welding a polyethylene lid to a polyethylene drum.
Custom designed lids and drums are used in conjunction with an
engineered spin-welding machine to provide the user with a total
system for encapsulation of waste materials.

Bondico
Bondico, Inc. of Jacksonville, Florida, has developed a
patented encapsulation technology that utilizes a fusion system
that is built into the lid of its composite container system. The lid
is compressed to the container and attached to Bondico's encap-
sulation control unit which passes an electrical current through
the fusion system allowing the heat generated to fuse the
polyethylene portions of the lid and container together. This en-
capsulation technology has been in commercial use since
February 1986, and has been used successfully to encapsulate
hazardous and radioactive wastes in the United States and
abroad.
This paper will discuss the technical aspects of both processes,
their mechanics, machinery requirements, support needs and ap-
plications.

MACROENCAPSULATION
To develop a viable macroencapsulation process, several tech-
nical and operational factors must be addressed to provide a use-
ful and effective system for commercial application:
• Physical properties of the load-bearing container
• Compatibility with internal and external elements
• Meet the required U.S. Department of Transportation (DOT)
regulations for public highway transportation
• Closure must be easily performed, 100% effective and remain
functional for the life of the container
• A cost which will allow commercial use.

The two main components of a macroencapsulation process are
the load-bearing container and the technique for sealing the lid to
the container. The load-bearing container must be designed to
withstand the following conditions:

Compression due to burial depths from 3 to 50 ft
Internal and external chemical resistance from wastes con-
tained and trench liquids
Biodegradation due to fungus and/or bacteria
Freeze/thaw cycles
Ultraviolet light exposure for long periods of time
Internal pressures built up by the encapsulated wastes
Long-term creep under burial related stress loading
Performance requirements for transportation under DOT 49
CFR 178.

In addition to meeting the above criteria, the load-bearing
macroencapsulate must be designed to accept a variety of W8**
forms, solids, drums, contaminated soils, etc. This requirement
necessitates development of a closure method which will meet th*
following criteria:

• Will provide a 100% effective seal
• Will not diminish the performance or abilities of the encap-
INNOVATIVE TECHNOLOGIES
-------
sulate to withstand the items noted in the previous section
• Can be implemented in the field using a mobile system
• Can be sealed using a technique which requires no experienced
or skilled personnel and provides total safety to those personnel
• Does not require expensive machinery or lengthy process time.
A macroencapsulation system which meets the criteria and ad-
dresses the numerous factors involved will provide a needed op-
tion for waste generators and handlers confronted with special
applications.

APPLICATIONS
Macroencapsulation probably will be used in special applica-
tions. In the same manner that solidification is successfully util-
ized for specific waste applications, macroencapsulation can be a
viable and effective technology for the following applications:
• Overpacking and encapsulating damaged drums excavated
from remedial action sites prior to disposal or long-term storage
• Encapsulation of dioxin- and PCB-contaminated soils at re-
medial action sites and spill sites
• In-container solidification of liquids and sludges which will be
encapsulated and then landfilled or stored
• Encapsulation of solids and sludges, particularly heavy metals
such as lead, mercury, etc.
• Encapsulation of incineration slag and scrubbing residues
prior to disposal
• Encapsulation of radioactive wastes and mixed radioactive
wastes which are hazardous under RCRA guidelines
• Other specialty applications where current treatment technolo-
gies are ineffective, too costly or inappropriate.
In each application, the encapsulation technique is used to in-
definitely isolate the wastes from the environment. In some cases,
the wastes may be recovered and recycled at some future time; in
other applications, the wastes eventually may be recovered and
treated.

BONDICO ENCAPSULATION SYSTEM
The Encapsulation System (Fig. 1) involves a composite con-
tainer with a built-in encapsulation device, an Encapsulation
Control Unit and an operator.
The system has the following steps:
• The waste is loaded into the special container
• The lid (with the attached encapsulation sealing device) is
placed onto the container
Figure 1
Bondico Encapsulation System
• The galvanized bolted hoop ring is used to compress the lid to
the container during fusion
• The operator connects the hand portable Encapsulation Con-
trol Unit's leads to the leads on the lid
• The operator pushes the "start" button and the encapsulation
control unit automatically fuses the lid to the container, en-
capsulating the waste; after 15 min of fusion, the encapsula-
tion control unit shuts off
• A green light will signal the operator to disconnect the leads;
the encapsulation is complete, and the system is ready for the
next container.
The Encapsulation System was designed to meet all of the func-
tional and operational criteria discussed earlier. It meets the cri-
teria with a system that offers versatility, ease of use and safety.
The composite container is a dual laminate polyethylene/fiber-
glas reinforced plastic (FRP) composite overpack with a 90 gal
capacity. The system is designed to easily accommodate damaged
or leaking 55 gal steel drums or wastes in various forms.
The inner liner is constructed of a medium density polyethylene
resin which is rotationally molded into a liner for the overpack. A
patented process bonds a fiberglas-reinforced plastic casing to the
polyethylene liner. The bond is important as it successfully joins a
thermoplastic liner to a thermosetting casing, thereby causing
each material to become enhanced by the physical properties of
the other and effectively eliminating each material's performance
inadequacies.
Attempts to overlay these materials without the bond will pro-
vide minimal enhancement of the materials or overall container
performance. The bonded composite uses the FRP casing to pro-
vide structural strength, performance and exterior corrosion
resistance. There are many other benefits derived from the bond-
ing process, such as enhanced polyethylene chemical resistance,
long-term burial life, better physical properties and other advan-
tages that are unique to the composite material.
Independent testing has indicated that various Bondico com-
posite containers can withstand without failure tests such as:
130 Ib/in2 external pressure simulating 125-ft burial depths
25 ft drops onto compacted sand
18 ft drops onto concrete
25 lb/in2 internal hydrostatic test
Puncture tests from rifles to forklifts
Vibration tests of 18,000 impacts/hr.
An independent consulting firm, which evaluated the useful
lifetime of U.S. EPA-approved synthetic liners for landfills,
evaluated the Encapsulation System using the same criteria used
in its U.S. EPA liner evaluation. The results indicated the Encap-
sulation System could provide a quantum leap in the safe contain-
ment of buried wastes. Their report stated:

"... has recently conducted a study for the U.S. EPA to
estimate the service life of synthetic liners in landfills by
considering the following factors.
• Past performance
• Accelerated aging tests
• Analysis of possible failure mechanisms
"Ways in which the liner can fail are important limita-
tions to the overall life of the typical liner. These include:
» Quality of the initial installation
• Structural support by subgrade (initial support and
stresses due to differential settling)
• Integrity and durability of field and factory seams
• Damage by animals
• Use with incompatible wastes
INNOVATIVE TECHNOLOGIES 97
-------
• Mechanical failure due to loading, freeze/thaw, etc.

"Based on the above considerations, we estimate that the
average lifetime of a synthetic liner in a landfill is about 25
yr-
"The encapsulated drum is not subject to the above
limitations except for
• Use with incompatible wastes
• Mechanical failure
"In the absence of these two failure mechanisms, its
lifetime should approach several hundred years.
"Overall, our conclusions are that under favorable con-
ditions (i.e., limited mechanical stress) the lifetime of the
Bondico drum would be several hundred years compared to
a typical lifetime of 25 yr for a liner in a landfill."1

The composite container has also met the necessary U.S. DOT
and U.S. EPA regulations allowing for approvals such as:
• DOT-E 9430: Approved as a salvage drum under 49 CFR 173.3
(c) and authorizes the use of the Encapsulation System under
this exemption.
• DOT-E 9359: Approved to directly contain various hazardous
solids and semi-solids and authorizes the use of the Encap-
sulation System under this exemption.
• PCB Approval—U.S. DOT and U.S. EPA: First non-metallic
container approved for use for the direct containment of PCB-
contaminated materials or articles. Approves the use of the
Encapsulation System.
• Qualifies as a U.S. DOT 7A type A package for class A radio-
active materials.
The composite container can be manufactured into a variety of
shapes and sizes, incorporating the encapsulation technique into
each design. Fig. 2 indicates the design and construction of the
current 90 gal size container.
Figure 2
Bondico Composite Encapsulate

The encapsulation welding process enables the composite con-
tainer with a standard hoop ring to be totally encapsulated in an
on-site field application. An encapsulation sealing device is ap-
plied at the manufacturing facility to a standard lid for the con-
tainer. Once a leaking 55 gal steel drum or contaminated soil is
placed inside the container, the lid with PE gasket is placed on
top, clamped down with a standard hoop ring and leads from the
sealing device are attached to leads from a control unit. The con-
trol unit is connected to a 110 volt power source (portable
generator), and the operator activates the automated process.
After 15 min have elapsed, the polyethylene liners of the con-
tainer and lid have been fused together, completely encapsulating
and immobilizing the wastes inside.
The encapsulation technology was designed to allow versatility
and flexibility when used as a cleanup technology. The process is
so simple and automated that untrained personnel can operate the
process upon review of the instruction guide. Operator error has
been effectively eliminated through various features of the con-
trol unit. Extremes in air temperature do not affect the process,
allowing for a - 20 °F to 110°F operating range during encapsula-
tion activities.

ENCAPSULATION APPLICATIONS
Encapsulate containers were used successfully in an emergency
response cleanup of radioactive waste in Great Britain in
February 1986. An incident involving polonium required the use
of the composite containers to contain and encapsulate the con-
taminated materials. Following Great Britain's Ministry of the
Environment's instructions, the containers were encapsulated and
monitored for 6 mo. At the end of the monitoring period, the
containers indicated no signs of leaks in the welded area and were
subsequently disposed of in the government's nuclear disposal
site.
Other containers have been purchased for applications such as
PCB containment, emergency response and other specialty ap-
plications.
The Bondico encapsulate offers an on-site process for the
following applications:
• On-site overpacking and encapsulation of leaking or rusting
55 gal drums. The chemical resistance of the composite ma-
terial will allow safe containment of acids, bases, flammables,
solvents, salts, heavy metals and poisons.
• On-site encapsulation of dioxins.
• On-site encapsulation of PCB-contaminated soil, leaking trans-
formers and parts or fixated PCB liquids.
• Effective containment for ocean disposal applications.
• Encapsulation of low-level radioactive wastes and mixed rad-
wastes to immobilize the waste for long-term storage or dis-
posal.
• Encapsulation of incinerator slag and heavy metals such as
mercury, lead, antimony and arsenic.

The composite material and encapsulation process can be used
to manufacture various sized containers or tanks: 55 gal size or
210 ft3. Continued successful demonstration of the encapsulation
process may lead to further development of a whole family of
containers for industry's use.
The application of the encapsulation process and technology
provides relief from the problems of containment that have
plagued the environment. The RCRA standards for approved
landfills anticipated that steel drums would leak, necessitating
that the landfills have double liners, leachate collection systems
and monitoring wells. It has been reported that new U.S. EPA-
approved landfills might have an effective life of 25 yr before
failure. The encapsulated drum will last for hundreds of years.
This encapsulation process offers a "window of safety" to allow
further development and perfection of waste treatment technolo-
gies, including incineration and retrieval/recycling of encap-
sulated wastes.
98
INNOVATIVE TECHNOLOGIES
-------
Mobility of Equipment
The Bondico composite containers are tapered in design which
allows nesting and efficient shipment to the cleanup site. They are
lighter in weight than comparable steel drums, thus reducing
labor and freight costs.
The operation of the encapsulation process involves only one
person with no special skills. Safety features have been built into
the system to eliminate potential hazards to the operator.
The number of containers which can be encapsulated on-site in
an 8-hr workday is based upon a modular concept. One operator
can work three control units at a time and encapsulate 13 con-
tainers/hi. The chart below shows some modular examples of
volume:

Units Encapsulated

No. of
Operators
1
5
10
No. of
Control
Units
3
15
30

Per hr.
13
65
130

Per 8-hr.
Shift
104
520
1040
Each control unit is powered by a 110 volt electrical source. A
portable generator provides a good on-site source of power.
Terrain, temperature and external conditions have no
detrimental effect on performing the encapsulation. The size of
the cleanup site has no real bearing on the use of a Bondico con-
tainer. A one-drum cleanup can be accommodated as efficiently
as a 10,000-drum site, given the availability of sufficient control
units. The encapsulation process is clean, safe and leaves no
residual wastes.

Capital and Operating Costs

Capital Costs:
Purchase of the Bondico encapsulate in quantities of 200 or
more would cost approximately $200 per container. However,
market acceptance of this technology resulting in an increase in
production volume could further reduce this cost. The control
unit used to perform the encapsulation process costs less than
$1,000. Leasing for cleanup sites could become an available op-
tion. In order to determine capital costs of encapsulation at a site,
the number of encapsulates required combined with an adequate
supply of control units are the only items to be considered.

Table 1
Comparison of Encapsulation Technologies
(Source: Shooter, 1985)1
Tre.t.lle
U.it*i
Very ioluble. eont*»Jn«nti *r* totilly
lint(ltd fro* iht environment.
Inoriinlt Slroni Otldl
required, beceuie the co*t
•iCerlili in iironi ind
thcvlcilly Inert.
Technique! |tn«r*ll
-------
Figure 3
EPP Overpack Welding Unit
(Source: Unger, et a/.)1
Table 2
Operating Parameters of the Friction Welding Encapsulation Process
Processing Parameter
Range of Values
Rotation speed
Welding pressure
Spinning time
Curing time
Cure pressure
280-350 rev/min
75 Ib/in2*
(line pressure)
5 Ib/in2
(weld pressure)
30-45 sec
2-7 min
75 Ib/in2*
(line pressure)
5 Ib/in2
(weld pressure)
* Relates to line gauge on apparatus
Source: Unger, el at.*

The welds resulting from the friction welding process were sub-
jected to various tests and inspections. According to the report
"Fabrication of Welded Polyethylene Encapsulates to Secure
Drums Containing Hazardous Wastes,'" . . .

"Sections of the weld were subject to tensile testing, and
the welds were examined visually and micrographically.
Tensile pulls showed that a 'proper' weld was achieved by
friction welding. The specimen consistently and without ex-
ception yielded at the toe of the weld and not at the inter-
face.
"Thin cross-sections of the weld were examined by
preparing transmission optical micrographs. A 5x
magnification showed the weld to be a continuous and
void-free cohesive bond. The weld length measured 2'/z
times greater than the wall thickness of the overpack. No
distinct weld line was observed indicating that good wet-out
and thorough mixing of the 50 x magnification of a welded
specimen prepared from a natural (white) HOPE cover and
a UV-stabilized (black) HOPE receiver. This micrograph
shows the welded interface to be formed by a gradual and
homogenous mix of each parent material. The thickness of
the interfacial region was approximately 560 microns.
Moreover, no Turn-line' (distinct interface) between the
joined materials was observed. Interfaces featuring
uniform surface morphology without a 'hairline,' differen-
tiates friction welds from heat-seams and butt-weld. We
postulated that advanced mechanical performance ex-
hibited in hydrostatic and tensile tests was due to the for-
mation of cohesive bonds characterized by large,
homogeneous interfaces."s

Field tests were performed at the U.S. EPA's Research and De-
velopment facility in Edison, New Jersey. The purpose was to
determine the effectiveness of the friction welding process under
"field conditions" and to subsequently evaluate the polyethylene
encapsulates against the performance requirements of U.S. DOT
34 (polyethylene tight head containers).
Conclusions drawn from the results of those tests are as
follows:
• The friction welding process is a viable means for encapsulation
in specific applications.
• Further development is needed to address the following specific
areas:
-A need to seal off the waste contents from the area in friction
-A means to assure the weld has been complete with no voids
present
-Refinement of process to assure quality and consistency of
each weld.
• Preliminary test results indicated the polyethylene encapsulates
were susceptible to cracking and leaking when subjected to the
U.S. DOT-required drop tests. U.S. EPA officials involved in
the test have suggested the polyethylene drums may not be
capable of withstanding the U.S. DOT drop tests consistently
and proposed exploring more suitable materials for the drum
construction.
The friction welding technology shows promise and could pro-
vide generators and waste management companies with an addi-
tional encapsulation option for remedial actions and disposal
needs.

CONCLUSION
Macroencapsulation is now a commercially available option for
transportation, storage and disposal of a variety of hazardous and
radioactive wastes. As with other waste-related technologies, the
application of macroencapsulation will be in situations where its
costs, benefits and process are justified and useful. Macroencap-
sulation is another solution to be added to the list of technologies
designed to meet the challenges of today's hazardous waste prob-
lems.

REFERENCES
1. Shooter, D., March 1985, "Evaluation of the Bondico Multi-Purpoie
Container Drum for Hazardous Waste Disposal Services." A.D.
Little, Boston, MA, 1985, 5-7.
2. Ibid., 13.
3. Unger, S.L., Eliash, B.M., Telles, R.W. and Lubowitz, H.R., "Fabri-
cation of Welded Polyethylene Encapsulates to Secure Drums Con-
taining Hazardous Wastes," Environmental Protection Polymai.
Inc., Hawthorne, CA, 1983, 78.
4. Ibid., 83.
5. Ibid., 87.
100
INNOVATIVE TECHNOLOGIES
-------
Photozone Destruction of Cyanide Waste
At Tinker AFB, Oklahoma
Pilot Plant Results

Martin F. Herlacher
Oklahoma City Air Logistics Command
Tinker Air Force Base, Oklahoma
F. Robert McGregor, D.Sc.
Water Management Inc.
Englewood, Colorado
ABSTRACT
The difficulties of treating complex waste at one central In-
dustrial Waste Treatment Plant, the cost of hauling waste to a
suitable disposal site, Air Force policies regarding waste minimi-
zation and recent advances in ozone technology have led Tinker
AFB to consider Photozone Activated Oxygen for the treatment
of cyanide.
Photozone Activated Oxygen consists of ozone (66.7%),
hydroxyl radical (14.7%), hydrogen peroxide (5.9%), atomic oxy-
gen (4.4%) and nitrogen oxides (<..!%). Activated oxygen is
generated by passing ambient air over spectrally controlled ultra-
violet lamps in the range below 190 nanometers. The specifica-
tions for activated oxygen generator are the subject of patents in
the U.S. and Europe.
Tinker AFB conducted pilot tests of the Photozone Activated
Oxygen System on cyanide waste in September and October of
1985. These pilot tests will provide information for the full-scale
system budgeted for installation in early 1987.
The economics for the proposed system are compared with
alkaline chlorination addition in the following summary:
Cost Item
Alkaline
Chlorine
Photozone
Capital $140,500 $538,000
Annual Cost
Operation and Maintenance $146,700 $ 78,200
Debt Retirement (10%, 10 yr) 22,900 87,600
Total Annual Cost $169,600 $165,800
Cost per Pound of
Cyanide Destroyed $ 8.23 $ 8.05

INTRODUCTION
Tinker AFB operates one of the largest electroplating processes
in the world. It produces 20,600 Ib of cyanide waste/yr. Of this,
6000 Ib/yr are in the form of alkaline cyanide and 14,600 Ib/yr are
carried out in the dilute rinses.
Previous treatment/disposal methods include alkaline chlorina-
tion, corona discharge ozonation, treatment in the Industrial
Waste Treatment Plant (IWTP) and hauling by truck to landfills.
Pilot studies using actual wastes were conducted at Tinker AFB
in September, October and November of 1985. These studies were
directed toward optimizing the basis of design for the Photozone
system and to identify the effects of pH and copper on the reac-
tion rate.
Results from a pilot study indicate 60 standard Photozone
lamps and 60 UVOX lamps will result in 95% removal of total
cyanide, pH was the most important parameter in the process and
the addition of copper sulfate was not an effective catalyst. The
total capital cost was $428,000. With an annual maintenance and
operation cost of $32,200, the total annual cost for Photozone is
$165,800. The total annual cost of the most likely alternative,
alkaline chlorination, is $169,600 (4% more expensive).
Cost consideration, safety and convenience make Photozone
the best alternative for the treatment of cyanide waste at Tinker
Air Force Base.

HISTORY OF TREATMENT AT TINKER AFB
Over the years a variety of treatment processes have been used
at Tinker AFB. In 1978 a pilot-scale ozone treatment plant was
constructed to treat alkaline cyanide waste. The process was
found to be extremely effective in treatment; but had a number of
operational problems. Since then alkaline chlorination was tested
but abandoned for reasons of high operating and maintenance
cost, exposure to workers and side reactions in the waste stream.
The present means of handling alkaline cyanide waste (6,000
Ib/yr) is having it hauled off-site. The dilute rinses (14,600 Ib/yr)
are mixed with other waste flows from the entire base and treated
at the IWTP.
The current approach has disadvantages. Hauling alkaline
cyanide off-site is expensive, and finding a long-term site for
disposal is becoming more difficult. Mixing the dilute rinses with
the entire waste stream was inefficient and at times interfered with
the treatment process at the Industrial Waste Treatment Plant.
These disadvantages, combined with the Air Force's policy of
waste minimization and recent advances in ozone technology led
Tinker AFB to investigate the use of Photozone as a treatment
method for cyanide.
PHOTOZONE AND
CYANIDE DESTRUCTION
The electroplating industry generates a large quantity of
cyanide waste and heavy metal salts. These wastes are toxic and
can have adverse effects on the environment if discharged un-
treated into streams or groundwater. These wastes also can have
deleterious effects on industrial wastewater treatment. If the
treatment process is a biological one, the cyanide can kill the
INNOVATIVE TECHNOLOGIES 101
-------
bacteria. If the treatment process is a chemical one, cyanide can
interfere with the chemistry and cause upsets in the treatment pro-
cess (i.e., metal precipitation). To alleviate these problems, eco-
nomical and simple treatment processes must be used to treat the
cyanide at the source.

Traditional Treatment Technologies
Cyanide wastewater traditionally has been treated by means of
gaseous chlorine, oxidation with hypochlorides and ferrous
sulfate. In recent years, oxidation with ozone has been used.
Chemical oxidation with ozone is an effective method of
removing organic constituents from the waste stream. Many
organic materials which are resistant to other treatment methods
can be oxidized using ozone. Cyanide- and phenol-bearing wastes
are two examples where chemical oxidation using ozone has prov-
en to be advantageous. When highly complex metal cyanide
wastes are present, ultraviolet radiation can be used to break
down the complexes and speed the reaction rate with ozone. The
reactions are as follows:

Chemical Ozonation and Photolysis
• Metal Complexes:
MCN +
Complex cyanide
MCN +
Complex cyanide
• Free Cyanide:
FAST 2CN- +
SLOW 2CNO +
2CN - + H2C
uv

03
Ozone

30j =
H20 -i-
> 1- 5O3
- M+ +
Metal complex
- M+ +
Metal complex

2CNO + 2C
3O3 = 2HCO3
= 2HCO3 + N2 +
CN- (I)
Free cyanide
CN- (2)
Free cyanide

>2
+ N2 +302 (3)
502 (4)
Ozone is a form of triatomic oxygen, O3, a very powerful oxi-
dant. It is highly unstable and has a tendency to form oxidizing
radicals in water. Ozone and these radicals react with the contam-
inants in the water and revert back to the diatomic state. One of
ozone's principal virtues is that it is one of the most powerful oxi-
dants available for wastewater treatment.
Oxidation with ozone offers several advantages:
• Destruction with cyanide in wastewater is rapid and effective
• More complete oxidation can occur
• Ozone is generated on-site as needed
• Ozone produces no toxic by-products
• Ozone is well suited to automatic controls
Oxidation with ozone has some disadvantages:
• Higher initial cost
• Maintenance problems associated with traditional ozone gen-
eration

The classical technology used in Europe and the United States
is based on imposing a high voltage electrical current (8,000 to
20,000 V) across a dielectric in the presence of a gas containing
oxygen. The resulting reaction is strongly exothermic and requires
cooling water to remove heat from classical ozone generators.
This technology additionally requires dry air prior to entering the
ozone generator. This requirement for dry air is due to the
nitrogen oxides that are formed during the ionization of the gas
stream. When ionized nitrogen reacts with the water in the gas,
nitric acid forms. This corrosive material will destroy the
generator's ability to produce ozone. Therefore it is necessary to
dehumidify the air to a dewpoint less than - SOT in the corona
discharge units.
The maintenance of dehumidifiers and high voltage re-
quirements were the main reasons this approach was abandoned
by Tinker Air Force Base in 1978.

Photozone Activated Oxygen
An alternative process, Photozone, is available on a commer-
cial basis for generating ozone and other oxidizing radicals to
react with cyanide in wastewater. This process is based on a
patented process using spectrally controlled ultraviolet lamps.
This process has several advantages over conventional ozone
generation:
• It produces a gas known as Photozone Activated Oxygen. Ac-
tivated oxygen has a higher oxidation potential than traditional
ozone alone and is therefore capable of providing more rapid
and complete oxidation of cyanide waste.
• It requires similar energy, typically 7.5 to 10 kwh of energy to
produce a pound of Photozone versus 7.5 to 12 kwh for con-
ventional ozone generation.
• It does not produce nitrous oxides and therefore does not re-
quire dehumidified air.
• Variations on Photozone design can be combined with the
beneficial aspects of ozone and ultraviolet light to react with
the metal complexes of cyanide. This process is called UVOX.
• It produces other beneficial oxidants in addition to ozone.
The UVOX process is applicable to situations such as cyanide
treatment where direct radiation of the wastewater enhances the
treatment process. The UVOX reactor consists of two concentric
tubes:
• Outer tube acts as a chamber for the liquid to be treated.
• Inner tube contains ultraviolet light source.
In the outer tube, a greater UV exposure to liquid volume ratio
is produced, adding to the synergistic effects of photolytic oxida-
tion by activated oxygen in the UVOX lamp.

PILOT STUDY
Tinker Air Force Base initiated the pilot-plant study on the
treatability of cyanide with Photozone in September 1985. The
primary objectives were to:
• Define the flow and chemical characteristics of the dilute rinses
and alkaline cyanide waste streams.
• Determine the capabilities of Photozone to remove free cyanide
and total cyanide in the two waste streams.
• Determine the effects of copper and pH on the treatability of
the waste.
• Identify the major design parameters for a full-scale plant.
• Prepare a firm cost estimate for a full-scale Photozone plant
installation.

Waste Characteristics
All the wastes from the electroplating facility flow through
either the alkaline cyanide sump or the dilute rinse sump. The
characteristics of these waste streams are shown in Table 1.
Table 1
Sources, Amount and Concentrations of the Cyanide Waste*
Total Cyanide Flow
Sump
mg/l
Ib/day Ib/yr gal/mw
Dilute Rinse
Alkaline Cyanide
TOTAL
42
30,000

40
16
56
14,600
6,000
20,600
80
0.046
80.0
102 INNOVATIVE TECHNOLOGIES
-------
The total annual weight of cyanide was determined from pur-
chasing records for the electroplating facility. Assuming no
cyanide is lost in the system, these records indicate 20,600 Ib/yr of
cyanides are generated in the plating shop. Since the plating
operation is not seasonal, this amounts to an average of 56 lb/-
day.
The waste disposal records indicate 24,000 gal/yr of concen-
trated alkaline cyanide are hauled away for disposal. The
laboratory analysis reports the average concentration of cyanide
is 30,000 mg/1. Therefore, the total amount of cyanide disposed
of in the aqueous phase is 6,000 Ib/yr. The average flow rate is
0.046 gal/min if this waste were bled into the waste treatment
system. The volume of the alkaline cyanide sump is 5,000 gal,
thereby providing substantial holding capacity for storing any
short-term surges in waste flows.
All remaining cyanide (14,600 Ib/yr) flows to the dilute rinse
sump, averaging 40 Ib/day. A pressure gauge was installed on the
discharge line from the dilute rinse sump, and a timer was placed
on the control box to provide the data needed to determine an
average flow rate. The flow rate and concentration were
calculated to be 180,000 gal/day and 42 mg/1, respectively.
These high flow rates were considered to be excessive; there-
fore, water conservation was instituted in the treatment process
and sized accordingly. For design purposes, the flow and concen-
tration were assumed to be 80 gal/min (115,000 gal/day) and 46
mg/1 in the dilute rinse, respectively.

Pilot-Plant Setup
A pilot-plant was installed in the electroplating shop near the
sump tanks (Fig. 1). The reaction vessel was a nalgene 60 gal
open-top drum. Ten feet of triple-cut Hinde aeration tubing were
placed in the bottom of the tank to bubble Photozone gas into the
liquid to be treated. For each test, the tank was filled to a depth of
30 in. with 45 gal of rinse water.
FLOWMETER

UNIT
PH IS5 HW1
on
PH IV) Ht>
P006E5

TANK
40 GAL
OXYC.EN CONCENTRATOR
-AIR co/ipftrsso/z
CkAS
SPARGER
Figure 1
Pilot-Plant Schematic for the Destruction
of Cyanide Wastes by the Photozone Process

Photozone gas was generated by either a single 90 cm lamp
(Model PH190HD) or a single 45 cm lamp (Model PH145HDM).
The oxygen concentrator used was a Briox Model 2100B, capable
of providing a gas which is 90% oxygen at 61/min. The effective
transfer rate of the Photozone gas was calculated to be 10-11%.

Test Results
The pilot-plant was monitored with portable meters and probes
designed to measure pH, ORP and free cyanide. The pH of the
solution was determined using an Orion Model SA210, Oxidation
Potential Probe (Cole Farmer Chemcadet) and free cyanide (Cole
Farmer Chemcadet with Orion Electrode 940600). Samples were
taken at appropriate intervals for testing by an independent
laboratory. These samples were analyzed for metals, forms of
cyanide and other chemical parameters. Twelve separate tests
were conducted: six on a mixture of dilute rinses and alkaline
cyanide sump wastes, five on dilute rinse water and one on
alkaline cyanide sump waste. The pH was adjusted from 6.4 to
12.4 standard units. Copper sulfate was added to nine of the tests
to determine the catalytic effects of copper on the process. The
duration of the tests ranged from 240 to 2400 min. Samples were
taken at the beginning and the end and intermediate samples were
taken throughout the test.

Analysis of the Results
The most significant adjustment of the results from the pilot-
plant data to the design of a full-scale system was related to the
expected transfer rate of the Photozone gas. The transfer rate was
estimated to be in the 10 - 11% range. In a full-scale system, a
more efficient transfer was achieved using greater liquid depth (20
ft) and static in-line mixers or bubble diffusers. A transfer effici-
ency was estimated for the design purpose to be 90 - 95%.
The full-scale system will use 90 cm lamps. Correction factors
of 2.0 and 2.5 were used for 45 cm lamps and the absence of an
oxygen concentrator, respectively.
A detailed review of this data supports the following conclu-
sions:
• Total cyanide removal rates generally increase with pH.
• The dilute cyanide has a high fraction of cyanide complexed
with metals.
Table 2
Major Components of Photozone Cyanide Treatment System

Components
Photozone Lamps
UVOX Lamps
Gas Flow Rate
Static Mixers
Dilute
Rinse
40
60
14.1
4
Alkaline
Cyanide
20
0
2.8
2

Total
60
60
16.9
6
Treatment Tanks
Volume (gal)
Compartments
Circulation Rate (gal/min)
Average Cyanide Waste
Flow (gal/min)
Tap Water Dilution
(gal/min)
Treatment Detention Time
(min)
Turnover Time (min)

Compressors
Number
Total Flow (ftVmin)

Oxygen Concentrators
Number
Total Inflow (ftVmin)
Total Outflow (ftVmin)
48,000 12,000 60,000
426
400 100 500

80 0.046 80

0 4.15 4.15

600 2,860
120 120

3
274

3
274
16.9
INNOVATIVE TECHNOLOGIES 103
-------
• Free cyanide removal rates were mixed, based on pH and the
relative level of completing of the waste.
• The addition of copper sulfate did not significantly increase
the removal rate.
• In high strength waste, the side reaction demanding ozone in-
terferes with the cyanide removal in the early stages of treat-
ment.
• Over 90% removal of the free cyanide can be achieved in mix-
tures, similar to those produced by Tinker AFB in 480 min at
pH 11.
• Similar results occurred for total cyanide in 570 min at
the same pH.
• Dilution of the alkaline cyanide sump with tap water allows
effective removal of free and total cyanide at pH 9.5.

PRELIMINARY DESIGN
The preliminary design for the entire system is based on a com-
bination of Photozone and UVOX lamps. The dilute rinse and
alkaline cyanide sumps will be treated in separate tanks to allow
flexibility for different metal removal and metal recovery opera-
tions. Due to the relatively high concentration of complexed
cyanide in the dilute rinse, the addition of UVOX lamps will en-
hance the removal rate. Some of the energy from the ultraviolet
light will be used to break the metal-cyanide bond to free the
cyanide for treatment by Photozone. For a list of the units and
their respective sizes, see Table 2.
ECONOMIC COMPARISON WITH
TRADITIONAL TECHNOLOGY
Previous studies at Tinker Air Force Base demonstrated that
ozone is a viable means of treating cyanide. The technical ad-
vances embodied in the Photozone system make this even more
likely today since most of the operational expenses which
plagued the other ozone system designs are eliminated by the Pho-
tozone process. The purpose of this section is to make this eco-
nomic comparison, including the capital cost of the new installa-
tion.

Alkaline Chlorination
Alkaline chlorination is the most viable alternative to Photo-
zone for treating cyanide as a pretreatment step in metal removal.
The major cost for this treatment is the large amount of chemicals
that must be used to treat cyanide wastes. These chemicals include
chlorine and sodium hydroxide to control pH and avoid off-
gasing of highly toxic cyanogen chloride by-products of the reac-
tion. Industry-wide experience indicates a dosage requirement of
7.35 Ib of chlorine and 1.125 Ib sodium hydroxide/lb of cyanide
destructed. These values were used in this economic analysis. This
chemical dosage is based on the fraction of the waste which is tied
up in metal complexes. The fraction of metal complexes in the
waste stream was determined from the analytical data gathered in
the pilot plant study. This information is summarized in Table 3.
Tankage and mixing requirements are based on general in-
dustry experience. The costs of chemicals are based on current
quotes for bulk quantities of 50% sodium hydroxide and 2-ton
cylinders of chlorine gas. The labor requirements, based on
general industry experience, are 6 man-hr/day for routine opera-
tions, replacement of system components and on-site hauling of
chemicals. Miscellaneous equipment replacement is based on
$6,000/yr to replace mixing impellers plus 2% of the capital cost
for the non-tankage equipment.
The alkaline chlorination system will produce some sludge. The
dry weight of sludge is approximately equal to the amount of
chemicals added as part of the process. This sludge is ac-
cumulated in the treatment tank and will be removed in the metal
removal process. The sludge amounts to 170,000 Ib dry weight/yr
or 680,000 Ib/yr at 25% solids after belt press dewatering. The
current cost for hauling sludge from Tinker AFB is $0.032/wet Ib.
This disposal requirement results in a cost of $21,800/yr.
Debt retirement is based on a ten-year period at 10% annual in-
terest rate.
Table 3
Basis for Cost Estimate Comparison
Item
Cost or Requirement
Electric Power
Water
Alkaline sump dilution
Unit Cost
Chlorine
Volume
Cost
NaOH
Volume
Cost (50% solution)
Soda Ash
Adjust to pH 11
Dilute Rinse Sump
Alkaline Cyanide Sump
Cost (50% solution)
Chlorine Tanks
Volume
Mixing Requirements
Sludge Haul
Percent Solids
Cost
Labor
Mixed Skilled and Unskilled
Debt Retirement Capital Factor
(10%, 10 yrs)
$0.05/kwh

4.2 gal/min
$0.90/1000 gal

7.35 Ib Cl2/lb CN
$0.21/lb

1.1251bNaOH/lbCl2
$0.22/lb NaOH
51.21b/day
2.7 Ib/day
$0.22/lb Soda Ash

20,000 gal
2 HP/3,000 gal

25%
$0.032/wet Ib

$20.00/hr

0.1628
Table 4
Economic Comparison of Photozone with Alkaline Chlorination
(20,600 Ib Cyanide Per Year)
PHOTOZONE t UVOX
Gas Handling
Tanks
Chlorine Feed ft Mix
pH Adjustment
Hlscellaneoua
TOTAL CAPITAL

OPERATION ft MAINTENANCE

PHOTOZONE Lup Energy
Mixing I Gai Handling
Chlorine
pH Adjustment
NaOH
Soda Ash
Labor
Sludge Hauling
Dilution Water
'Equipment Replacement
TOTAL OPERATION ft MAINTENANCE

ANNUAL COSTS

Debt Retirement (101, 10 jr)
Operation ft Maintenance

TOTAL ANNUAL COSTS

UNIT COSTS

Cos: per Pound of Cyanide
ALKALINE
CHLORINE
0
0
37,000
79,000
20,000
4,500
* 140,500
0
4,900
31,800

37,400
0
43,800
21,800
0
7,000
I 146,700
I 22,900
146,700
$ 169,600
S.23
PHOTO ZOtlt
f 5M.OOO
I 71.200
104 INNOVATIVE TECHNOLOGIES
-------
Photozone System
The costs for the Photozone system are based on the prelimin-
ary design described in the previous section.
The major operating costs for Photozone are the electrical
energy and replacement of Photozone lamps. Electrical energy is
required to run the compressors (37 kw), pumps (7.5 kw),
Photozone lamps (6 kw) and miscellaneous items (1 kw). The
average life of a lamp is 1.5 yrs under continuous use; it costs $300
to replace a lamp for a total annual cost of $24,000. The cost to
replace other items is 2% of the original cost of the non-tankage
and Photozone items. The labor requirements are estimated to
average 2 man-hr/day since the system is fully automated and
does not require any on-site handling and storage of chemicals.
The sludge production is estimated to be about 20,000 Ib of dry
solids/yr or 80,000 If/yr total weight at 25% solids. At
$0.032/wet Ib, this totals $2,600/yr. This information is sum-
marized in Table 4.

CONCLUSION
Based on total annual cost, the alkaline chlorination system is
4% more expensive than the Photozone system. The cost advan-
tages of Photozone, when combined with the elimination of the
hazards associated with the chlorination system and the well oxy-
genated quality of the effluent from the Photozone system, make
Photozone the better alternative for the treatment of cyanide
wastes at Tinker AFB.
INNOVATIVE TECHNOLOGIES
105
-------
The Treatment of Contaminated Groundwater
And RCRA Wastewater at Bofors-Nobel, Inc.
John A. Meidl
Zimpro Inc.
Rothschild, Wisconsin
Ronald L. Peterson
Bofors-Nobel, Inc.
Muskegon, Michigan
ABSTRACT
Bofors-Nobel, Inc., a Michigan manufacturer of herbicides
and organic chemicals, is using an innovative approach to treat
production (RCRA) wastewaters as well as contaminated ground-
water at its site. A powdered carbon/biological system (known
commercially as PACT™) and wet air oxidation are used to en-
sure that Bofors' discharge is acceptable to the Muskegon County
wastewater system.
In the PACT process, powdered carbon is added to end-of-pipe
activated sludge treatment, to remove a wide range of organic
compounds not treatable solely by biological means. As demon-
strated at Bofors, such addition also eliminates the need for end-
of-pipe GAC adsorption. Currently, 1.8 million gal/day of con-
taminated groundwater and RCRA production wastewaters are
being treated.
When Bofors expanded organics chemicals manufacturing
operations, the quantity of powdered carbon used in the PACT
system justified regeneration. Wet air oxidation was selected as
the technology to regenerate spent carbon from the PACT sys-
tem. WAO also destroys toxics adsorbed on the carbon.
A second WAO unit has been installed at Bofors to detoxify
certain process waste streams.
More than 100 different organic compounds have been effec-
tively treated by the system since its start-up in March 1983.
Overall treatment results are shown, as well as information on
system selection and economic comparisons of treatment alter-
natives.

INTRODUCTION
It is not often that wastes regulated by the Resource Conserva-
tion and Recovery Act (RCRA) and Superfund (CERCLA) are
generated and treated at the same site. But Bofors-Nobel, Inc., a
manufacturer of herbicides and organic chemicals located near
Muskegon, MI, presents a rare opportunity to observe successful
technological solutions to both types of waste problems. A unique
combination of treatment processes has been in use there since
1983, detoxifying process wastewaters as well as contaminated
groundwater emanting from an abandoned landfill.
The key to the treatment process at Bofors is a powdered ac-
tivated carbon/biological system (PACT™) and wet air oxidation
(WAO) system. Manufacturing wastewaters, containing several
organics classified as toxic and hazardous under RCRA, are
treated in the PACT system.
PACT is also used, however, to decontaminate 1.2 million
gal/day of groundwater that is pumped from beneath an aban-
doned landfill located on plant property.
Removals of orthochloroanaline (OCA) and dichlorobenzidene
(DCB) are significant in the PACT system. COD reduction has
averaged greater than 98%, even though the system must take
heavy loads from ongoing chemical production.
The wasted biomass and spent carbon from the PACT system
are processed in a skid-mounted WAO unit; the biological sludge
is oxidized, while the powdered carbon is regenerated and reused
again in slurry form. Carbon recoveries of 91% or better have
been common.
A second skid-mounted WAO unit is used to detoxify produc-
tion wastes, providing roughing treatment of an acid waste stream
prior to a crystallization process that produces fertilizer.
The PACT system was selected for cost-effectiveness and treat-
ment capability following an engineering analysis of biological
treatment, adcorption by granular carbon, combinations of bio-
logical and granular carbon treatment, chemical oxidation and
sorption on such materials as bentonite and clay.
Actual capital and operating costs have confirmed the validity
of the earlier review and selection; and in conjunction with the
WAO units, PACT gives Bofors cost-effective total control of
both its liquid and solid waste streams.

BACKGROUND
In 1977 the AB Bofors Group of Sweden bought Lakeway
Chemical, located just a few miles east of Muskegon's city limitt.
The purchase was part of a plan to increase the company's service
level and participation in the United States' specialty and fine
chemical markets.
Shortly after the purchase of the property, however, severe en-
vironmental problems—resulting from past disposal pracu'cei-
surfaced at the site. More than 370 million Ibs of sludge had been
placed in on-site lagoons since 1971. Materials excavated from the
lagoon area were used to form berms or dams to contain the
sludges. Sludges consisted primarily of a water-calcium sulfitc
slurry containing several primary organic amine compound!—
each with concentrations as high as 5,000 ppm.
Due to the porous nature of surface and subsurface
groundwater beneath the site had become contaminated with
these organic compounds. This contamination was a source of
pollution of Black Creek—which ultimately flows to Lake Mich-
igan.
In 1978, when the State of Michigan filed suit against Bofon-
106 TREATMENT
-------
Nobel, the company voluntarily installed eight purge wells to in-
tercept the contaminated groundwater and direct it to a public
wastewaste treatment plant.
In addition to the accumulated waste materials, continued
chemical manufacturing at Bofors produced a variety of other
waste streams. Many were too toxic for biological treatment and
had required hauling by a commercial treater.
Bofors faced a difficult problem: how to expand production at
its facility while at the same time resolving complex environmen-
tal issues. Together, these requirements would place heavy
demands on capital resources.
A solution to the environmental problem could have involved
creation of a burial vault to contain 170,000 yds of primary
sludges and 350,000 yd3 of contaminated subsurface soils.
Some form of treatment of the contaminated groundwater,
such as carbon adsorption units, would also be necessary. How-
ever, Bofors felt that such a solution would not truly destroy the
bulk of the contamination—thus leaving a legacy which might
possibly have to be dealt with a second time.
After investigation it became apparent that costs of destroying
all toxic components through on-site treatment—while significant
—were not so different than those for mere containment of the
materials.
Thus, in 1982 Bofors joined with two other companies to form
a separate company, Environmental Systems Corp. (ESC), to
solve the environmental problems, Zimpro Inc. of Rothschild,
WI, and Chemical Waste Management, a Waste Management,
Inc. subsidiary, were the other members of the new firm.
ESC developed a combination of technologies and facilities
which could be used to perform the cleanup of the site and
simultaneously process the wide variety of wastes from continuing
and expanding chemical operations in an environmentally accep-
table manner:
• Wet air oxidation
• PACT system treatment
• Waste acid neutralization and detoxification
• Secure landfill
• Forced water-soil flushing
Each technology or system can be uniquely integrated with each
other to permit the use of the lowest cost treatment or detoxifica-
tion method for a wide variety of waste streams.
Ultimately, the environmental impact of ESC is to eliminate the
discharge of any form of pollutant—gas, liquid or solids—from
the plant site, and to completely rectify past disposal problems.

GROUNDWATER TREATMENT
Engineering Studies
The original site survey confirmed that a large volume of sludge
existed south of the manufacturing facility. Acid wastes had been
neutralized with calcium hydroxide, which precipitated calcium
sulfate solids. This sludge was landfilled, and contained high
levels of organics which were seeping into the groundwater.
As a first step, Bofors installed a purge well system to intercept
the contaminated groundwater. About 1.3 million gallons of
purge water per day were brought to the surface and discharged to
the Muskegon County wastewater system. Table 1 indicates the
results of a GC/MS organic pollutant wide scan analysis of the
purge water performed by the county wastewater authority in
1980. Fifteen different organic compounds are identified, the
most concentrated being 2-chloroaniline (OCA). The next most
abundant organic chemical was benzene.
In response to an OCA minimization plan put forth by the
county, Bofors began a series of treatment studies and pilot-plant
runs to determine the most cost-effective methods of reducing
Table 1
GC/MS Organic Analysis of Purge Water
Compound

2-Chlorophenol

Phenol

Cresol

2-Chloroaniline

1,2 Dichloroethane

Benzene

Perchloroethylene

Toluene

Chlorobenzene

Ethyl Benzene

Dichlorobenzene Isomer

3, 3-Dich.lorobenzidene

Bis (Ethyl Hexyl) Phthalate

3-Chloroaniline

Benzidine Isomer*
Concentration (ppb)

13,000

420

4,900

1,500

150

220

2,500

100

68
65
*Mass spectrum is very similar but retention time is two minutes earlier.
OCA and the other organic constituents, such as benzidene and
dichlorobenzidene (DCB) in the purge water.
The waste characterization strongly suggested that a combina-
tion of biological treatment and carbon adsorption would most
likely be required to achieve the desired results.
Studies performed by the Ada, MI, engineering firm of Fish-
beck, Thompson, Carr and Huber demonstrated that biological
treatment (activated sludge) could accomplish reduction of the
OCA, benzidene, chemical oxygen demand (COD) and total or-
ganic carbon (TOC). In addition, activated sludge treatment ap-
parently reduced ethylenedichloride and toluene concentrations
likely as a result of air stripping. However, little consistent
removal of DCB occurred.
When physical adsorption on granular carbon in column was
tested, DCB removals to levels near 5 fig/1 were achieved, but it

Table 2
Anticipated Water Quality
(Concentrations in mg/l Unless Otherwise Indicated)
Separate
Biological
BOD
COD
SS
TOC
DCB
OCA
Benzidine
EDC
Toluene
30 to 40
70 to 80
25
20 to 30
100 ppb
30
90 ppb
2A ppb
130 ppb
0 to 5
5 to 11
5 to 11
5
75 ppl
N.D.
N.D.
7 ppl
12 ppl
N.E.

N.E.

S ppb*

300 ppb

15 ppb

60 ppb

30 ppb
Biological
and Carbon

0 to 5

5 to 10

5 ppb*

N.D.

3 ppb

12 ppb
* Detectability Limit
N.D. • Non-Detectable
N.E. • No Estimate oc Data
TREATMENT
107
-------
was also apparent that considerably more carbon contact would
be required to reduce DCB levels further.
These findings led Bofors and its consulting engineer to pursue
additional testing of various combinations of carbon adsorption
and biological treatment. Raw purge water was treated with car-
bon in one series of tests. Other tests attempted to answer ques-
tions relating to the use of carbon both prior to and after
biological treatment (Table 2).
It became obvious that the optimum system for the Bofors
purge water would maximize hydraulic detention time, so that the
OCA could be virtually eliminated and would maximize carbon
contact time so that the DCB could be reduced to less than 5 jtg/1.
It was also obvious the conventional approach of using carbon
columns in conjunction with activated sludge would be very costly
under these conditions.
At this point, since both biotreatment and carbon adsorption
appeared to be necessary, Bofors began experiments in which
powdered activated carbon was added directly to the activated
sludge system so that physical adsorption and biological treat-
ment could occur simultaneously. Commercially, this treatment
process is known as the PACT system, and is marketed by Zim-
pro Inc. Such an approach increases the amount of time the waste
constituents are in contact with both the carbon and the biological
mass, and also exposes the wastes to treatment for the full solids
residence time of the system, as opposed to only the hydraulic
residence time which occurs in a biological process. Testing in-
dicated that this condition met the treatment objectives most cost-
effectively.

System Design
In the orignal plan, Bofors contemplated using a second stage
for additional powdered carbon contact, following the PACT
system, but results obtained in further pilot testing indicated the
second step was not necessary.
In the final design scheme, the purge water and process waste-
water are exposed to single-stage treatment, with wastewater
aerated in the presence of a high concentration of powdered acti-
vated carbon (PAC) and volatile biological solids in the aeration
basin.
Wastes are accumulated in an equalization basin. Phosphoric
acid is added as a nutrient for the biomass.
In the aerator, the PAC concentration may range from 4,000 to
12,000 mg/1, depending on the influent wastewater characteristics
and effluent quality required. Mixed liquor is composed of 50%
PAC, 40% biomass and 10% ash.
As a result of the concentrations of PAC and biological solids
maintained in the system, a high degree of reliable treatment is
obtained. Toxic materials or shock loadings can be accommo-
dated without upset; the carbon adsorbs materials which are non-
biodegradable and the biological organisms assimilate non-
adsorbable pollutants.
The aerator, a circular, above-ground tank, is mixed by two
100 hp downdraft turbine aerators. Two blowers are capable of
providing 1,600 ft3/min of air each. The aeration basin capacity is
1.5 million gal.
Following aeration, treated wastewater is settled in a circular,
above-ground, 1 million gal skimmerless clarifier. Effluent is
discharged to the Muskegon County Wastewater Treatment
System.
Spent carbon and biomass are wasted from the system
periodically, and treated in a separate titanium wet air oxidation
process. This consists of a prefabricated skid-mounted unit,
capable of processing 10 gal/min of slurry. At temperatures of
SOOT and under pressures of 1,500 Ib/in2, the biological material
associated with the carbon is oxidized to a small amount of inert
ash, while the carbon is regenerated for use again.

108 TREATMENT
The process operates autothermally on feed solids of 8 to 10%,
utilizing double-pipe heat exchangers to conserve on fuel costs
even further.
In addition to recovering virtually all of the carbon (Table 3),
the process is cost-effective since it eliminates the need to dispose
of sludge from the PACT system—estimated at the beginning of
the project to be in the neighborhood of $350/ton or $600,000 per
year.

Table 3
Powdered Carbon Usage
PACTTM/Wet Air Regeneration Systems
Carbon Dose, Ib/d

Regenerated
Virgin

Total

Virgin Carbon Makeup
>2,500«
50

2,550
2.0%
^Quality check against virgin carbon using DCB standards shows 90% adsorption recovery ef-
ficiency.

Performance
The facility has been treating approximately 1.2 million gal/day
of contaminated groundwater pumped from beneath the old land-
fill site and up to 600,000 gal/day of process wastewater. A total
of 780 million gal has been processed since startup.
More than 100 organic chemical components are received by
the PACT system during a year; 90 are biodegradable and 10 are
carbon adsorbable (Table 4).

Table 4
Partial List of Permitted Compounds
Bofors-Nobel Inc., Muskegon, Michigan
Acetone

Aliphatic Amine

Allyl Alcohol

Aaaoniuia Dithiocacbam

AjBBoniun Thiocyanate

Aniline

B-Chloroanil me

B-Napthylailne

Benzene

B nzidine

B nzoic Acid

B phenyl-OL

B phycidene

B s (ethyl hexyl) ptht
orobenzene

orobenzidene
chlorobiphenyla

Ethyl Acetate

Ethyl Benzene

Formaldehyde
Isophoron*

nethylene Chlorldf

Hethylpyildlni

Hitrocceiol

HltropthiUc Acid

perchloroethyltM

Phenol

phenoiybiphenyl

phenylnaphthileiM

pthallc Acid

2-Piopanol

sodlun Acetatf

TatrachletetthylffM

Toluene
System effectiveness is shown in Table 5. COD reductions have
averaged better than 98%, or from 6,000 mg/1 to well under 100
mg/1. OCA and DCB in the effluent average less than 10 mg/1 and
2 mg/1, respectively, despite influent levels that contain high con-
centrations and vary widely from day to day and hour to hour.
The PACT system has also been nitrifying the wastewater.
Table 5
Waste Treatment Performance

i
* Recent 7-d«'
low, HCD
OD, pp.
rthochloroanaline, ppb
uspcnded Solids. pp»
composite ample », ppm-.
L.9
6,000 • 10
53,000 1
12,000"
I
150-200 • 1
6/12/8C 3
fl/26/8* 2
9/9/9* 4
9/2J/B6 4
10/14/86 2
91. 13
99.99
99,91
94.lt
-------
Annual operating costs budgeted for the PACT plus wet air
regeneration systems for 1986, including solids disposal, neutrali-
zation, groundwater pumping and county wastewater treatment
charges, is less than $1.0 million/yr, or less than lOC/lb COD
treated. Regeneration of spent powdered activated carbon has
proven to be cost effective as well. Some 2,500 Ib of carbon are
recovered daily. Annual cost for regeneration, virgin carbon
makeup and solids disposal is budgeted at less than $300,000 per
year; without regeneration those same annual costs would exceed
$1,000,000—and the problem of contaminated solids would not
be eliminated.

RCRA WASTEWATERS
In addition to the accumulated waste materials, continued
chemical manufacturing at Bofors-Nobel produces a variety of
waste streams, some of which are too toxic for biological treat-
ment and have required hauling by a commercial treater in the
past. Expansion of manufacturing capacity, of course, produces
more quantities of non-biodegradable wastes. A list of com-
pounds contained in these production wastes is reported in Table 4.
Because these wastewaters are too dilute to incinerate
economically, yet too toxic for even the PACT system, a second
wet air oxidation unit is employed to reduce their toxicity before
further treatment. This unit operates alongside the unit for car-
bon regeneration.
Wet air oxidation destroys toxics contained in aqueous solu-
tions, or converts them to biologically degradable organics such
as acetic acid.
The WAO unit is designed to operate at a temperature of
SOOT. The unit has a capacity of 10 gal/min with a design
pressure rating of 2,000 Ib/in2.
Wastewater is pumped to the WAO system high pressure
pumps from existing storage. The high pressure pumps increase
the pressure of the wastewater to the 1,600-1,700 Ib/in2 range.
Following the high pressure pumps, a portion of the oxidizing gas
(air) is added to the wastewater. The remainder of the air
necessary for oxidation is added to the oxidation reactor down-
stream.
The wastewater-air feed is first preheated against hotter oxida-
tion reactor effluent. Preheating is such that when the waste-
water-air mixture is introduced into the downstream oxidation
reactor, the heat of oxidation will increase the mixture
temperature to the desired maximum. Included in the preheating
circuit is another heat exchanger using natural gas-fired hot oil
which is used for startup, and to sustain the process during
periods when oxidation is not autothermal.
The reaction of oxygen-demanding components takes place in
the oxidation reactor, which provides a 60-min residence time.
The reactor effluent, comprised of oxidized liquor and spent
air, is used to preheat the feed mixture in the feed heat exchanger
prior to being cooled indirectly against plant cooling water in the
cooler. Following cooling, the system pressure is released th rough
a pressure control valve and the oxidized liquor-spent air mixture
is separated in a separator vessel. A water spray in the scrubber-
separator serves to cool the off-gases before discharge to a stack.
Like the other WAO unit at ESC, this one is also a
prefabricated, portable package consisting of two 8 ft by 35 ft
skids and a reactor vessel.

Performance
The wet air oxidation unit began operating at Bofors on process
wastewater in April 1983, and averaged 99.8% destruction for the
toxic components in the feed stream (Table 6). These com-
ponents, produced during the manufacture of a pesticide pro-
duct, were non-biodegradable chemicals ranging in concentration
from 600 to l,200mg/l.
After wet oxidation, concentrations of the toxic components
were less than 10 mg/1 and often undetectable. The biodegradable
effluent was pumped through the PACT system before discharge
to the sewer.

Table 6
Wet Air Oxidation Unit
(Results on Original Waste Stream)
Design Flows: 10 Gallons/Minute, or 14,400 Gallons/Day
Toxic Component

COD

Actual Plows
Feed

600-1,200 ppm

70-80 g/L
Effluent

2-9 ppm

30-40 g/L
Percent
Reduction
99.8»

50+%
4,500-7,500 gpd
Currently, the WAO unit is used for another purpose. A 40%
sulfuric acid waste stream (6 gal/min) is first neutralized with am-
monia and then wet oxidized. Organic contaminants are
destroyed or reduced to short-chained compounds. The oxidized
liquor is sent to a proprietary waste acid neutralization and detox-
ification plant—which is a continuous crystallization process that
produces high nitrogen fertilizer which is sold for agricultural use.
Both WAO units at Bofors are operated 24 hr/day. Solvent
washes with caustic or nitric acid have proven effective for scale
control. The frequency of washings is higher for the carbon
regeneration unit, about once every 2 to 3 wk.
In addition, hydrocarbon emissions from the wet air oxidation
unit to the atmosphere have averaged 0.63 Ib/hr, and are well
within prescribed emission limits of 3.0 Ib/hr.
FILTRATION
Figure 1
PACT® Wastewater Treatment System General Process Diagram
AIR COMPRESSOR
Figure 2
Wet Air Oxidation General Flow Diagram
TREATMENT 109
-------
>LUT(
k**o»
u »*ITC. ,1—
lf*C»*rr NCUTAAUZAT
REFERENCES

1. Zadonick, L.A., "Waste Treatment Technologies on Display at
Theme Park," Waste Age Magazine, Oct. 1984.
2. Meidl, J.A. and Wilhelmi, A.R., "PACT: An Economical Solution
in Treating Contaminated Groundwater and Leachate," Paper pit-
sented, New England Water Pollution Control Association, Jan.
1986.
3. Fishbeck, Thompson, Carr & Huber, Consulting Engineers, Ada, MI,
"Treatability Studies for Groundwater Treatment," 1981.
4. Bofors-Nobel, Inc., Presidents Council on Environmental Quality
Entry Award, 1983.
5. Environmental Systems Corp., Plant Operating Records, 1986-87.
Figure3
4.5 MOD Wastewater Treatment Facility
110 TREATMENT
-------
Innovative Electromembrane Process for
Recovery of Lead from Contaminated Soils

E. Radha Krishnan, P.E.
William F. Kemner
PEI Associates, Inc.
Cincinnati, Ohio
ABSTRACT
This paper presents the results of a research program designed
to optimize, through bench-scale tests, the process variables for
the chelation and electroplating operations of an innovative elec-
tromembrane process employing a chelating agent for recovery of
lead from contaminated soils. The paper also discusses scale-up
design criteria and comparative economics for the process.
Ethylenediaminetetraacetic acid (EDTA) was used as the che-
lating agent for the process. Metals treatability tests conducted in
the laboratory on soil samples from two different sites showed
that the optimum EDTA/lead molar ratio for the chelation reac-
tion is in the range of 1.5 to 2.0 for the two materials, and that the
chelation reaction is essentially complete within 1 hr. A pH of 12
effectively prevented the chelation of other metals such as iron.
Tests conducted on a bench-scale electromembrane reactor
(EMR) unit showed lead removal efficiencies approaching 90"%
for chelate solutions with initial lead concentrations greater than
1%. The lead removal efficiency decreased with decreasing lead
concentration. The pH of the chelate solution had no apparent ef-
fect on plating efficiency. Higher current densities resulted in
faster plating rates and a spongy lead deposit on the cathode. The
as-plated material analyzed over 75% lead in two experimental
runs.
The process appears to be applicable to a wide variety of lead-
containing soils and wastes such as slags, dusts and sludges from
industrial processes. Economic analyses for the overall soil wash-
ing electromembrane process based on a preliminary process
design show its cost to be competitive with off-site disposal op-
tions.

INTRODUCTION
Numerous Superfund sites throughout the United States are
contaminated with toxic metals. Battery reclamation, lead
smelting and lead-based paint manufacturing are examples of
processes which could result in lead-contaminated soils. Soils
from defunct battery reclamation sites average about 5% lead.
Quantities of contaminated soils range from less than 5,000 yd3
per site to almost 100,000 yd3. Many of the sites are located over
key underground aquifers in populated areas, raising concerns
about contamination of water supplies.
The concentration range of lead in soils found at 436 con-
taminated sites has been reported to be 0.16 to 466,000 ppm, com-
pared with the natural background level of 2 to 200 ppm.1
The cleanup of such sites traditionally has involved excavation
of the wastes and contaminated soils with subsequent disposal at
an off-site, RCRA-approved landfill. In addition to increasing
costs and dangers to public safety from large-scale transportation
of wastes, long-term environmental liability is also a concern
associated with the landfilling approach. Many experts have
characterized this approach as simply "moving the problem" in-
stead of solving it. Thus, there is great incentive for the develop-
ment of alternative methods for cleanup of contaminated sites.
Fig. 1 summarizes the alternatives available for treating lead-
contaminated soil. It should be noted that only the soil-washing
option actually removes the lead from the contaminated soil. This
paper describes research conducted to investigate the process
characteristics, design and economics of a soil-washing process
employing an electromembrane reactor (EMR) for treatment of
contaminated soils for recovery of heavy metals such as lead. Fig.
2 provides a highly simplified overview of the soil-washing pro-
cess. The process uses ethylenediaminetetraacetic acid (EDTA) as
the chelating agent and recovers lead by electrodeposition. The
primary objective of the research was to optimize, via bench-scale
tests, the process variables for the chelation and electroplating
(EMR) operations of the process. The classification and dewater-
ing steps, though crucial to the overall process, represent existing
technology and were not studied specifically during this research.
This process results in a lead product containing about 90% lead
at optimum process conditions.
OH-ltTI
worn
H
"JS3"
Figure 1
Treatment Alternatives for Lead-Contaminated Soils
TREATMENT 111
-------
Table 2
Sieve Analysis of Soil from a Battery Reclamation Site
Figure 2
Simplified Process Flow Diagram of Overall Soil Washing Process
The applicability of the process is highly site-dependent. Fac-
tors such as soil fines content, clay content and lead solubility can
strongly influence the cost and performance of the process. Con-
sequently, both the soil treatability (chelation) and electroplating
tests were conducted on a variety of samples in order to make
preliminary assessments of process applicability.

SOIL TREATABILITY TESTS
The purpose of the soil treatability testing was to determine the
optimum conditions for soil-EDTA reactions to: (1) maximize
lead chelation, (2) minimize EDTA consumption and (3)
minimize reaction time.
A soil treatability test procedure was developed to evaluate the
effect of pH, EDTA consumption and reaction time at a constant
temperature. The treatability testing involved physical and
chemical characterization of the raw material followed by chela-
tion testing for lead recovery/metals interference.

Physical and Chemical Characterization
Soil samples typically consist of varying amounts of gravel,
sand, silt, clay and organic matter. A sieve analysis was used to
determine the distribution of particle sizes in the soil. The exact
test is described under ASTM Designation D 422. Material pass-
ing a No. 200 sieve tends to be composed largely of clays and silts
and is generally difficult to dewater. Screening the material prior
to reaction separates the material into fractions which can be
analyzed to determine the particle size distribution of the
material. Screening has shown a tendency for higher lead content
material to segregate in the fine fractions. Consequently, screen-
ing may be used to reduce the volume of material to be treated.
Samples of soil from two sites were screened and extraction
procedure (EP) toxicity tests were performed on each fraction to
determine if a toxicity gradient existed based on physical sizing.
The results shown in Tables 1 and 2 illustrate the tendency for
lead to segregate in the fine fractions for these soils. A similar
relationship, however, may not be expected for all soils.
Table 1
Sieve Analysis of Waste from an Industrial Site
Size fraction
»20 nesh
(-20)«3S nesh
(-35)*100 KSh
(-100)«200 nesh
-200 wsh
in
Range
45-63
9-12
18-29
5-8
5
Percent
size fraction
Mean
54
11
23
7
5
EP toxicity
value for Pb, rag/liter
67
186
174
248
344
Size fraction
>10 mesh
>20 mesh
>35 mesh
>70 nesh
>100 mesh
>200 mesh
9/ liter
7
22
37
42
51
49
55
Chelation Testing
Before describing the chelation tests in detail, it is helpful to
review briefly the properties and characteristics of EDTA. There
are many forms of EDTA. In this work, the tetrasodium salt of
EDTA was used as the chelating agent. By definition, a chelating
agent is a compound containing donor atoms that can combine by
coordinate bonding with a single metal atom to form a cyclic
structure called a chelation compound or, simply, a chelate.
A range of molar ratios of EDTA/lead was used at a selected
pH condition to determine the minimum ratio necessary for
essentially complete chelation. Liquid chelate was sampled from
the soil-EDTA reactor at specified time intervals to determine
chelation as a function of time.
These tests provide information on lead recovery, iron interfer-
ence, reagent needs and feasibility of treating a particular waste
by chelation. The ranges for pH, time and EDTA use can be
varied depending on the particular soil.
The soil treatability procedures developed for this study were
performed on lead-contaminated soil samples from two Super-
fund sites (Arcanum near Troy, Ohio and Lee's Farm in Wood-
ville, Wisconsin).2 Table 3 provides the analysis of the metals con-
tent of these two soils. Fig. 3 illustrates the relationship of chela-
tion efficiency versus time for two test runs on the Lee's Farm soil
and one test run on the Arcanum soil. It is apparent that the
chelation reaction is essentially complete within 1 hr for both the
Lee's Farm soil and Arcanum soils at each of the EDT/Pb motor
ratios. It cannot be predicted that other wastes or soils wiD
necessarily be chelated so rapidly.
O 60-

1 *,
u
a 40
pH. 11.B-I23
LESHB
A O.MEOTA/PS.IEWM*"**

I i - I-1 • 0.43EOWH/WW""*
24 6 t 10
REACTION TIME, hr«
Figure 3
Chelation Efficiency as a Function of Time
112
TREATMENT
-------
LESEHB
• LEE'S FARM SOIL
» AHCANUMSOIL
12345

EDTA/Pb MOLAR RATIO
Figure 4
Chelation Efficiency as a Function of EDTA/Pb Molar Ratio
Table 3
Chemical Analysis of Test Soils
(ppm on as-received basis)
Soil source
Element
Arcanum
Lee's Farm
Cadmium
Calcium
Chromium
Iron
Lead
Zinc
4
59630
19
20790
78950
110
1
47340
14
22010
38670
81
iFig. 4 presents final chelation efficiency as a function of
EDTA/Pb molar ratio. The optimum EDTA/Pb molar ratio ap-
pears to be approximately 1.5 to 2.0 for both the soils tested. The
optimum EDTA/Pb ratio may be different for other materials.
Chelation efficiencies exceeding 90% were observed for the Lee's
Farm soil at an EDTA/Pb ratio above 1.5. The apparent lower
chelation efficiency for the Arcanum soil may be due to the
presence of either metallic lead (as opposed to ionic lead) in the
sample or microencapsulation of lead.
Metallic lead is digested in the analysis procedure for total lead
but is not chelatable. Metallic lead is not extracted in the EP tox-
icity procedure which is conducted at a pH of 5 using acetic acid
and used to determine leachability characteristics. Since the basic
purpose of the chelation process is to render the soil non-
hazardous, lead recoveries must be based on the ability of the
chelation process to produce a residue that has an EP toxicity lead
content of less than 5 mg/1 (the federally allowable standard)
rather than the total lead removal.

EMR TESTS
Previous research on the electromembrane reactor (EMR) has
been performed in the context of regenerating ion-exchange
resins.3 The current research expanded upon this application.
Several variables are of importance in the experimental design of
the EMR tests.

Electrode Potential
The extent of chemical reaction occurring in an electrolytic cell
is directly proportional to the quantity of electricity passed into
the cell. For example, it requires 2 moles of electrons to produce a
mole of copper from Cu^+ and 3 moles of electrons to produce a
mole of aluminum from AP +:
Cu2+ + 2e- -* Cu (1)
A13+ + 3e- -* Al (2)
The electrical charge on a mole of electrons is called a Faraday
(F), equivalent to 96,500 coulombs. A coulomb is the quantity of
electrical charge passing a point in a circuit in 1 sec when the cur-
rent is 1 amp. Therefore, the number of coulombs passing
through a cell can be obtained by multiplying the amperage and
the elapsed time in seconds.

Current Density
Current density is calculated as milliamps (ma)/cm2 (amps/ft2,
etc.). Current density for the experiments was determined by
computing the ratio of the current flow on the power supply unit
to the cross-sectional area of the membrane.

pH
The pH in the electromembrane reactor is a very important pro-
cess condition which influences both the removal of metal from
the solution and the recovery of the chelating agent by regenera-
tion. The pH at the anode and the cathode varied during the EMR
experiments due to the production of hydrogen ions at the anode
and hydroxide ions at the cathode; the pH, however, was not ad-
justed during each experiment.

Current Efficiency
The energy requirement for ionic transport in the electromem-
brane process is a function of the electrical resistance of the solu-
tions and the membranes and the back electromotive forces caused
by concentration gradients. The current efficiency can be
calculated according to the following equation:
Current efficiency =
Metal ion removed (meq) x 96.5 (C/meq)

Time(s) X applied current (C/sec)
x 100% (3)
where:
meq = milliequivalent
C = coulomb
The current efficiency was determined as a function of time for
the tests.

Chelate Concentration
The concentration of the lead chelate in the cathode chamber
of the EMR affects current efficiency. As concentration de-
creases, power requirements to plate a given mass of lead in-
crease.

Experimental Procedure
Fig. 5 depicts the reactor system used for these experiments.
The rectangular unit was constructed from a commercial glass
aquarium with 1/4-in. thick plexiglass. It was divided into two
chambers by two 1/8-in. thick plexiglass pieces. The frames served
as supports for the cation exchange membrane. The membrane
was glued into place and the joints were sealed with silicone rub-
ber sealant to prevent leakage between chambers.
The membrane used was manufactured by Ionics, specifically
61CZL386 modacrylic fiber-backed cation transfer membrane.
The membrane has low electrical resistance and excellent resis-
tance to physical and chemical stress. Most importantly, it allows
sodium ions to pass from the anode to the cathode chamber while
preventing ionic transport in the opposite direction.
Lead electrodes were used in the EMR system. Both electrodes
had dimensions of approximately 7 in. by 10 in. by 1/16 in. They
were mounted on wooden dowel rods suspended across the top of

TREATMENT 113
-------
CATHODE (•)
Hi]CO] lOLUnON
BASKET-
LEAD AMOOE
COVER
PLATE '
MAGNETIC STIRRER
Fh-EDTA (OLUTION

CATION
TRANSFER
MEMBRANE
LEAD CATHODE •
IUQNETIC ST1RREH
1/4 h. U.I.J 1M kL
tkv-
Jin.
Figure 5
Schematic Illustration of EMR Test Unit
the aquarium. The power source supplied a potential of up to 40
volts and a direct current of up to 30 amp.
Once the reactor was operational, each run was started by add-
ing a 5% by weight sodium carbonate solution to the anode
chamber. In addition, an appropriate amount of metal chelate
complex solution was placed in the cathode chamber. Each elec-
trode was then placed in the EMR by suspending it approximately
1 in. from the membrane surfaces. The test began when voltage
was applied and the current was set at the proper amperage. The
voltage across the circuit was allowed to vary in such a fashion
that the current was maintained at a desired setting.
Considering the overall reactions involved in the reactor sys-
tem, the major reaction of concern was the one resulting in re-
moval of metal from solution; thus, the metal concentration
and reaction time were monitored regularly. This was done by
taking samples from the cathode chamber at regular tune inter-
vals. To enhance mass transfer, a magnetic stirrer was placed in
each chamber to cause mild turbulence throughout the opera-
tional period.

Experimental Design
The three primary control variables of interest in the EMR
bench-scale experiments were current density, lead concentration
in the chelate and cathode solution pH. Higher current density
generally produces a lower quality plated metal, but plated metal
quality is not of paramount importance in the soil-washing pro-
cess as long as its quality is not so inferior that it would inhibit
sale of the product. The maximum current density for the ex-
periments was kept below 30 ma/cm*. Effective operation at both
high and low lead concentrations is extremely important in order
to accommodate various levels of contamination in soil or waste
materials. Solution pH is of interest because of the need to elevate
pH to inhibit iron chelation in high iron wastes. The EMR should
thus be able to function well at both low and high pH.
The source of lead chelate solution for the experiments was ac-
tual chelate produced at the Lee's Farm site. This material con-
tained about 3% Pb, and portions were diluted with water to
create nominal 1% and 0.2% solutions. The solutions were ad-
justed to the desired pH using sulfuric acid or sodium hydroxide.
Table 4 summarizes the actual lead content and pH of the feed-
stock solutions.

Table 4
Lead Content and pH of Feedstock Solutions Used in EMR Experiment!
Feedstock
No.
1
2
3
4
5
6
7
8
Nominal Lead
Content
rag/liter
30,000
30,000
10,000
10,000
10,000
2,000
2,000
2,000
PH
11
4
11
8
4
11
8
4
Five experiments were performed on the 0.2% Pb solution, and
two experiments each were performed on the 1 % and 3% Pb solu-
tions. A partial factorial experimental design was adopted to
evaluate the effects of lead concentration, current density and
PH.
Theoretical plating time was calculated based on Faraday's law.
2e •• Pb
(4)
Two moles of electrons (2 faradays) are required to plate 1 mole
(2 equivalents) of lead. The grams of lead plated in 1 hr at 1 amp
at 100% current efficiency can be calculated as follows:

grams of Pb = (1 hr) (1 amp)
(3600 sec) (1 coulomb) (1 faraday) (1 mol Pb)
hr
mol
amp-sec 96,500 coulomb 2 faradays
= 3.86gPb/amp-hr
(5)
Given the total amount of Pb in solution and the desired cur-
rent density, theoretical plating time (at 100% current efficiency)
was determined. Current densities were calculated based on the
400 cm2 area of the membrane.

EMR Test Results
Figs. 6 through 9 illustrate plating efficiency (i.e., lead plated ai
a percent of total lead in solution) as a function of time. Ai ex-
pected, increased lead is plated with increasing time in all easel.
Extremely high lead recoveries and current efficiencies are ob-
served for the 3% and 1% lead solutions during the experimenttl
time period. It appears, however, that current efficiency (and
subsequent lead removal) at the starting lead concentration of
0.2% is low regardless of pH or current density. Figs. 6 and 7
show that lead recoveries are below 40% at the 0.2% lead level for
the experimental time period.
114 TREATMENT
-------
so
40-
»
I
a
10
0.5 1.0

PLATING •nME.hrs
1.S
LEGEND

A pH 11 CURRENT DENSITY 25 mtan2
* pHIICURHENTDENSITYISimtan2
•pH 11 CURRENT DENSITY Snmtem2
Figure 6
Lead Plating Efficiency as a Function of Time
(0.2% Pb in initial solution, pH = 11)
i
I
§
30-
20-
10-
0.5 1.0
PLATING TIME, hr»
LEGEND

T pH 8 CURRENT DENSITY 15 rmfcm2
• pH 4 CURRENT DENSITY 15 rratan2

Figure 7
Lead Plating Efficiency as a Function of Time
(0.2% Pb in initial solution, pH = 4, 8)
1.5
LEGEND
« pH 11 CURRENT DENSITY 15
• pH 11 CURRENT DENSITY 25 m/aif
1.0 1.5

PLATING TIME, hr>
Figure 8
Lead Plating Efficiency as a Function of Time
(1% Pb in initial solution)
pH11.CURRENTOENSITY.2Sm«*in2
pH 4. CURRENT DENSITY.
234

PLATING TIME, hre
Figure 9
Lead Plating Efficiency as a Function of Time
(3% Pb in initial solution)

Greater time periods should result in higher lead removal ef-
ficiencies for the low lead solutions. Figs. 8 and 9, however, show
lead removal efficiencies approaching 90% for the 1% and 3%
lead solutions. Fig. 8 shows the effect of current density at con-
stant pH for a 1% lead solution. As expected, the higher current
density produces a faster plating rate. Higher current density pro-
duces a spongy lead deposit on the electrode. Fig. 9 illustrates the
high plating efficiency achievable at higher initial lead concentra-
tions. The effect of current density on plating rate is confirmed
again by the results shown in Fig. 9. There is no apparent effect of
initial cathode solution pH on plating efficiency.
There was no noticeable difference in the visual appearance of
the lead product from the various experiments of a given initial
lead concentration. In the 0.2% lead experiments, the plated lead
was not visibly discernible on the electrode but was confirmed by
analytical results and the increase in the weight of the cathode.
Based on the experiments on the 0.2% lead liquor, the current
efficiencies are higher at lower current densities, decreasing from
40% at a current density of 5 ma/cm2 to approximately 20% at
25 ma/cm2. There is no apparent effect of pH on this relation-
ship. In the full-scale process, the current efficiency should not be
a controlling factor in the economics because power costs are in-
significant compared to other cost elements. Time, however, is an
important factor because it related to labor cost. Consequently, it
is desirable to run the highest possible current density.
Table 5 provides an analysis of the plated lead product for
those experiments where sufficient deposit could be scraped off
the cathode. The plated metal analyzed over 75% lead in Runs 1
and 2. As shown, the amount of other metals plated is insignifi-
Table 5
Analysis of Plated Metal from EMR Experiments
(all data in ppm as-received basis)
Experiment
No.
Cd
Ca
Cr
Cu
Fe
Pb
Hg
Zn
1
6.7
1128
<1.2
264
35.1
787700
74.1
54.0
2
3.1
1751
<1.2
226
25.7
755700
292
43.1
8
4.0
499
<1.2
175
23.9
497500
70.2
56.8
9
2.1
2709
<1.2
265
48.7
669500
180
75.8
9
(duplicate;
2.4
3015
<1.2
259
51.2
672000
182
84.7
TREATMENT 115
-------
cant compared to the lead. Although not shown, the moisture
content of the product is the other main constituent. After dry-
ing, therefore, the lead product is expected to have a purity in ex-
cess of 90<7o.
Hydrogen is generated at the cathode as a product of the elec-
trolysis of water. The hydrogen generation rate was not
measured, but the pH increase detected during the experiments in
the cathode chamber indicated a decrease in hydrogen ion con-
centration.
SCALE-UP
Although this research focused on the chelation and plating
steps of the soil-washing process, the design factors necessary for
scale-up must be considered for the overall process. The four ma-
jor process operations are solids handling, EDTA reac-
tion/washing, lead plating (EMR) and water treatment.

Solids Handling
Initial solids processing depends upon the specific site charac-
teristics. The material may be processed via screening, magnetic
separation and/or crushing. Metal and other bulk material must
be removed. If crushing is required, the material may be rescreened
and stockpiled for later feed to the system.

EDTA Reaction/Washing
The purpose of the EDTA-reaction step is to thoroughly mix the
soil and EDTA solution to chelate the lead. After chelation, the
lead complex is washed from the solids in a series of dilution steps.
The major parameters governing the operation of this phase are
the lead concentration leaving the system as a final product and the
moisture content of the solids as they move through the system. De-
watering characteristics of the material are critical in this step. The
amount of water to be used must be optimized through the use of
multiple stages.

Water Treatment
Water from the plating step is sent to a waste treatment system.
When the lead concentration decreases to a low level in the EMR, it
probably will be cost-effective to reconcentrate the water to max-
imize lead recovery. Eventually, dissolved solids will build up and a
blowdown stream will have to go to a waste treatment system. It is
essential to the economics of the process to recover and reuse the
chelating agent prior to final discharge of the water.

ECONOMIC ANALYSIS
Comparative economics for cleanup of a given site are highly
dependent upon site location, lead concentration and nature of the
material. Some sites contain lead only, and others are contaminated
with multiple pollutants, both inorganic and organic. In addition,
the dewatering characteristics of various materials vary widely,
which in turn affects processing cost. The comparative economics
of soil washing versus other alternatives must be determined
specifically and individually for each site.
A computerized cost model was developed to evaluate the effect
of site-specific process variables. Table 6 lists the variables included
in the model. The current cost model is based on the use of mobile
equipment, including cement mixers, trailer-mounted pressure
filters, and trailer-mounted EMR units. Preliminary cost analyses
indicate that the cost of the soil washing process is competitive with
landfilling. The process has an economy of scale, being more com-
petitive as the quantity of material to be treated increases above ap-
proximately 2000 yd^. Assumptions used for the cost computations
must be refined based on the specific considerations associated with
each site.
Table 6
Variables in Cost Model
Total Material (yd3)
It Material Dry Process
Equipment Rental, $/mo.
Mixers
Screens & conveyors
Filters
EMRs
Tankage
On-site Trailer
Others
Operation, hr/day
Number of Rinses
Water/Soil Ratio
Filtering Rate, gph/ft2
Plating Rate, hr/2000 gal
Reaction Time, hr/batch
Batch Size, yd
Analytical, $/batch
Operating Supply, %
Maintenance Supply, %
Lead in Soil, %
Lead Recovery, %
EOT A/Lead Ratio
EDTA Recovery, %
Capacity Utilization, %
Cost of EDTA, $/lb
Cost of Caustic, $/ton
Cost of Sulfuric acid, $/ton
Cost of Sodium Carbonate,
$/ton
Cost of Sodium Sulfide, $/ton
Lead Credit, $/lb
Blowdown rate, It
Water Treatment, POTW,
$71000 gal
Trans, to POTW, S/loaded mile
Distance to POTW, miles
Avg. Operating Labor Rate, J/hr
Avg. Maint. Labor Rate, $/hr
Labor Overhead, %
Per Diem, $/day
Electricity, $/kWh
Supply water, $/1000 gal
Connected Load, HP
Lighting Load, Kw
Plating Load, kWh/lb Pb
Total Crew Size (operating)
Maintenance Crew Size
Pounds lead recovered
Lead in Soil Processed (Ibs)
Supply Water, gal
Filtering Time, hr
Reaction Time, hr
Plating Time, hr
Number of Units
Mixers
Screens & Conveyors
Filters
EMRs
Tankage
On-site Trailer
Others
Filter Size, ft2
CONCLUSIONS
The bench-scale research has shown the feasibility of the two
essential process steps of an innovative soil-washing process: chela-
tion and electrodeposition.
A long-term pilot-scale demonstration at several actual sites is
necessary to develop the data required for commercialization. The
research will examine several critical issues that have been identified
through the bench-scale experimental studies the preliminary
engineering design/economic evaluation. These include:
• The concentration of low-lead-content wastes to achieve lead
levels in the EMR of 1% or more
• The dewatering of fine materials
• The water balance for the washing and rinsing process (i.e., de-
gree of recycle possible, blowdown requirements and total sup-
ply of water required)

The completion of the pilot-scale research will culminate in:

• Final process flowsheet
• Complete material balance
• Plot plan and equipment layout
• Mechanical flowsheets
• Equipment specifications
• Design calculations
• A detailed operating procedures manual
• Refined cost model

ACKNOWLEDGMENTS
The research reported in this paper was conducted under Grant
No. ISI-8560730 from the National Science Foundation's SmiD
Business Innovation Research Program. Special acknowledgment
goes to Dr. James Etzel, Chairman, Department of Environmen-
tal Engineering, and Dr. Dyi-Hwa Tseng, former Ph.D. student
116 TREATMENT
-------
at Purdue University in Environmental Engineering, whose earlier of Management of Uncontrolled Hazardous Waste Sites, Washing-
work on electromembrane reactors was a milepost toward this ton, DC, Nov. 1983, 226.
present work. 2. Castle, C., et al., "Research and Development of Soil Washing Sys-
tem for Use at Superfund Sites," Proc. of Management of Uncon-
trolled Hazardous Waste Sites, Washington, DC, Nov. 1985, 452.
3. Tseng, D-H., "Regeneration of Heavy Metal Exhausted Cation Ex-
1. Sims, R. and Wagner, K., "In-situ Treatment Techniques Applicable change Resin with a Recoverable Chelating Agent," Thesis submitted
to Large Quantities of Hazardous Waste Contaminated Soils," Proc. to the Purdue University, Lafayette, IN, 1983.
TREATMENT 117
-------
Dewatering and Solidification/Stabilization
Of Industrial Waste
Mark L. Allen
Steven D. Liedle
Bechtel National, Inc.
Oak Ridge, Tennessee
ABSTRACT
Recently, Bechtel National, Inc. performed field dewatering
and stabilization tests on the waste materials from five waste-
water ponds at a major chemical manufacturing facility. The
waste materials in these ponds varied in color, composition and
consistency. The tests were performed to evaluate methods for
dewatering and solidifying/stabilizing these materials.
Test results showed that all waste material present in the ponds
can be solidified to meet strength and compressibility require-
ments using cement or pozzolanic processes. This treatment
makes the material more manageable and reduces the likelihood
that leaching will occur.
Dewatering of the waste material prior to additive introduction
is indicated for only a portion of two ponds since they contain a
relatively low percentage of solid materials. The waste remaining
after dewatering can be solidified in place and then excavated,
transported and disposed of in an engineered containment facil-
ity. Additional investigations were performed to assess the feas-
ibility of in-place stabilization of the waste. These investigations
indicated that the material in one of the site ponds is amenable to
in-place stabilization.

INTRODUCTION
As a result of increasingly stringent Federal and state require-
ments for the handling of hazardous materials (e.g., RCRA,
TSCA, etc.), and the availability of improved waste manage-
ment techniques, many industries are reevaluating their waste dis-
posal practices. Many organizations are removing obsolete treat-
ment facilities from service and replacing them with state-of-the-
art units. Closure of an industrial waste treatment facility re-
quires a thorough knowledge of the chemical and physical char-
acteristics of the wastes and an understanding of suitable tech-
niques which can be used to minimize movement of these ma-
terials into the environment.
Beginning in 1986, several process wastewater treatment ponds
at a chemical manufacturing facility will be closed. Pond closure
will consist of:

• Liquid removal and treatment
• In-place stabilization and capping of sludges containing heavy
metals
• Dewatering and solidification of an organic sludge
• Removal and interment of the organic sludge

Laboratory and field tests were conducted on the sludges prior

118 TREATMENT
to full-scale work. The results of these experiments were used to
determine the most efficient and cost-effective sludge treatment
methods for this site.

SITE DESCRIPTION
The industrial site for which this work was performed is lo-
cated adjacent to a major navigable river and has several unlined
surface impoundments on it. These impoundments range in size
from less than 1 acre to approximately 4 acres in size. Dikes sep-
arating the ponds were constructed using stiff to soft clay and
silty clay and are lined with rip-rap. Underlying the ponds are ap-
proximately 2 ft of medium density sand and silty sand with inter-
bedded clay layers. Groundwater occurs 5 to 10 ft below the sur-
face and is in contact with the bottom of two of the ponds.
The depth of sludge in the ponds ranges from 1.5 to 12 ft
Standing water was present in all ponds, although sludges were
exposed in some. Sludges in four of the ponds (designated 0-1
through 0-4) are organic, while the material in the remaining
pond (designated I) contains an assortment of heavy metals. The
organic sludges are generally brown to black in color with variable
viscosity. Sludges from Pond I are grayish in color with high vii-
cosity. Table 1 summarizes the raw sludge characteristics of the
ponds investigated.

Table 1
Raw Sludge Characteristics
Pond
0-1
0-2
Readily
Pinpable
X
X
Not Readily
Pupable

CoMflts
High Solids Content mi WiewlJ.
Portions of Pond ire not tapHi
UM Solids Content and ViHHitr
0-3

M
High Viscosity,
(terterogeneous HltKUl
Tar-Like Oxaistacy

High Solids Content. CrmUf
Teiture
MECHANICAL DEWATERING PDLOT TESTS
The goal of this program was to determine if mechanicil *•
watering is appropriate for the pond sludges and, if so, wbiebd*'
-------
watering method is most efficient. Laboratory tests were per-
formed on samples of raw and dewatered sludge (cake) to gather
the necessary information.
Random grab samples from the site ponds were evaluated in
the laboratory to determine their Theological (pumpability) prop-
erties. The tests performed consisted of:
• Density
• Viscosity
• Particle size
• Total solids
• Specific gravity
Tests were performed in accordance with procedures estab-
lished by the American Society for Testing and Materials (ASTM)
and the American Petroleum Institute (API). Table 2 summarizes
the results of these tests.

Table 2
Raw Sludge Characteristics
Pond
0-1
0-2
0-3
0-4
I
pH
8.4
7.3
7.6
NT"1
6.3
Viscosity
(SEC)
33.0
28.5
33.0
27.0
30.5
Density
(Ib/ft'l
69.0
63.5
63.5
62.0
77.5
Specific
Gravity
1.10
1.01
1.01
0.99
1.24
Solids
(% by wt)
12.63, 12.60
4.46
6,50
NT"1
17.17, 41.22
'"Not Tested

Three mechanical dewatering methods were evaluated for re-
moving water from the raw sludge: the plate and frame filter
press, the belt filter press and the centrifuge.

Plate and Frame Press
The plate and frame filter press was a small, bench-scale unit
with a 0.026 ft3 chamber. Raw sludge was pumped into the press
using a double diaphragm pump without polymer additions or
premixing. Cake was obtained by manually operating the unit.
Density (wet unit weight) and percent solids determinations were
performed in the laboratory to assess the relative effectiveness of
the dewatering process. Sludge from Ponds 0-2, 0-3 and 0-4 were
not dewatered in the plate and frame filter press because their
high organic content fouled the equipment.

Belt Filter Press
A trailer-mounted belt filter press unit with a 3.3 ft (1 m) belt
equipped with a polymer conditioning system, polymer and
sludge feed pumps, belt wash booster pump and other auxiliary
items also was evaluated. The belt consisted of a tightly woven
cloth that typically is used for delicate or biological sludges. Both
anionic and cationic polymers were utilized in the tests.
Samples of cake from the belt filter press were analyzed in the
laboratory to determine various properties of the sludge including
percent solids. Material from Ponds 0-3 and 0-4 fouled the belt
and would not dewater. Additional belt cleaning and changes in
the belt cloth weave might have enhanced the dewatering of these
sludges.

Centrifuge
The final mechanical dewatering method considered was the
centrifuge. The unit was set up as shown in Fig. 1 with sludge be-
ing pumped from Pond 0-1 into a 10,000 gal holding tank from
which the material was withdrawn for dewatering. Cationic and
anionic polymers were evaluated during the test, and the internal
geometry of the centrifuge unit was adjusted to optimize the
percent solids in the cake. Percent solids determinations were
made for the final cake product. These results and similar in-
formation from the other dewatering methods evaluated are sum-
marized in Table 3.
10,000 GALLON
HOLDING/MIXING TANK
Figure 1
Centrifuge Flow Diagram
Table 3
Dewatered Sludge Characteristics
Pond
0-1
Raw Sludge Solids
(1 by Height)
12.63
12.60
Dewatered Sludge Solids
(I by Height)
Belt Filter Plate and Prate
37.2 51.6

Centrifuge
46
47
0-2

I
4.46

17,17
41.22
28.0

64.8
NT

75.3
NT

NT
SOLIDIFICATION EVALUATIONS
Solidification of materials involves physical and chemical pro-
cesses that produce physically stable solids with improved hand-
ling characteristics. Due to the form of the solidified mass, the
leaching potential also may be reduced. The ultimate objective of
this modification process is the transformation of the waste into a
more environmentally stable form. Two solidification procedures
were evaluated for treating the sludges either in-place or in the
cake form: cement-based processes and pozzolanic processes.
The cement processes which were investigated utilized a mix-
ture of Portland cement and sludge to produce a stable material
suitable for interment. When the cement is mixed with the waste,
the water in the material hydrates the silicates and aluminates in
the cement which then harden to bind the particles together.
Since some organic substances—specifically those containing oils,
solvents and greases—inhibit the hydration and hardening of ce-
ment, physical testing was necessary to determine the effective-
ness of this stabilization technique.
The pozzolanic processes examined involve the reaction of a
fine-grained aluminous and siliceous material with water to form
a stable solid. Industrial by-products such as fly ash, ground blast
furnace slag and cement (or lime) kiln dust are typically employed
in pozzolanic techniques. In comparison to cement-based stabil-
ization processes, pozzolanic processes generally require greater
quantities of dry materials to stabilize sludges. However, when
mixed with the waste, these greater quantities of dry materials can
reduce the free moisture content and improve the handling char-
acteristics of the product.
TREATMENT 119
-------
Testing was performed to evaluate the effectiveness of stabil-
ization in-place and after removal and mechanical dewatering.
This testing program was based on the Theological and chemical
characteristics of the materials, particularly those relating to then-
stabilization potential.
Since the solidification process usually results in an increase in
the material's strength, an unconfined compressive strength test
was selected as the primary measure of the effectiveness of the
stabilization agent. A pocket penetrometer also was used to esti-
mate intermediate gains in material strength.
The compressibility of the waste material was measured by a
standard laboratory consolidation test. Permeability was meas-
ured using a flexible wall penneameter under a constant head
condition. These tests were performed on cured samples repre-
sentative of the final solidified material.
The following sections of the paper discuss the effects of stabil-
ization agents on the engineering characteristics (e.g., strength
and compressibility) of the cake and the in-place waste material
as indicated by test results.

Solidification of Cake
To evaluate the solidifying effects of various cement-based and
pozzolanic-based stabilization processes, the following sludge/
additive mixtures were selected for testing (percentages are by
weight):
Mixtures
Mixture 1
Mixture 2
Mixture 3
Fly Ash (%)
30
0
50
Cement (%)
10
15
0
These mix designs were selected based on previous experience
with waste materials containing a similar proportion of solids.
Selection of additives was based in part on the availability of suf-
ficient quantities of material and on the cost (including transpor-
tation) of the additives. To maximize the probability of solidifica-
tion of the cake during the initial testing program, the propor-
tions of fly ash (class C) and cement (Type 1 Portland) desig-
nated for the mixtures were the probable maximum percentages
needed to solidify the material. Adjustments in the percentages of
the additive selected for use were performed in the field with the
final goal being a mixture which will meet stability criteria with
the smallest possible volume of additive (and therefore the lowest
possible cost) and minimal increase in total volume.
Belt filter press cake samples from Ponds 0-1, 0-2 and I were
mixed with additives and subjected to a battery of tests including
the following:
• Liquid limit and plastic limit
• Wet density
• Unconfined compressive strength
• Consolidation
• Volume increase
• Permeability

The strength gained by each mixture over a 28-day period was
determined by performing penetrometer strength determinations
at predetermined intervals and an unconfined compressive
strength test at 28 days. In addition, consolidation and permeabil-
ity tests were performed on a sample from each mixture follow-
ing a 14-day cure period. The volume increase caused by the addi-
tion of the additive was measured for all mixes. The results of
these tests are shown in Table 4.

In-Place Stabilization
Tests were also performed on raw sludge from the five ponds
(0-1 through 0-4 and I) to measure the effectiveness of in-place
stabilization. Mixtures of raw sludge from Ponds 0-1 through
0-4 and various proportions of fly ash and Portland cement were
prepared and subjected to the following tests:
• Wet density
• Unconfined compressive strength
• Volume increase

The results of these tests are summarized in Table 5.
Raw sludge samples from Pond I were sent to three inde-
pendent firms for solidification evaluations. The design criteria
for the solidified material were an unconfined compressive
strength of 1.2 tons/ft1 (tsf), a consolidation of less than 10%
under a design load of 1.2 tons/ft2 and, preferably, reduction
Table 4
Summary of Test Results of Solidified Filter Cake
Pond
0-1

0-2

LL
PL
NP
NC
Notes
Mix Cement
(X)
1
2
3
i
2
3
1
2
3
10
IS
0
10
IS
0
10
15
0
Fly
Ash
30
0
50
30
0
50
30
0
50
LL PL
55 54
73 NP
67 54
108 52
167 75
95 45
NP
NP
NP
Moisture
Content
49.2
65.0
44.9
76.7
141 .1
69.7
39.9
51.1
36.0
Maximum
Density"'
(PCf)
99
82
104
92
NC
97
109
104
111
Unconfined Strength
Days Cured
3 28

>4.5 11.0
>4.5 14.5
>4.5 7.2
2.2 1.1
<-5 .2
2.0 1.2
>4.5 15.5
>4.5 6.6
>4.5 10.6
Volume
Increase
<*>
+ 30.4
+ 13.4
+ 32.1
t 32.7
+ 14.9
+ 34.3
•f 46.1
+ 22.3
+ 51.4
Percent
Consolidation0'
(*)
4
1
3
10
27
12
2
2
1
.60
.55
.21
.8
.3
.0
.67
.84
.3
Permeability
(en/sec « 10")
9.25
0.722
4.S9
2.78
2.44
1.40
2.23
19.6
1.43
Liquid limit
Plastic 1 imi t
"on-plastic
Non compactable

(I) Manmum ton] density a at the moisture content of cake material u delivered from the filler prm. computed II a modified proctor energy.
(2) Percent coroobdauon u it • loading of 4 TSF which exceeds the design load.
120 TREATMENT
-------
Tables
Sludge Solidification Test Results
Pond
0-1

0-3

0-4

Solidification
Additive
Portland Cement

Fly Ash

Portland Cement
Portland Ceient
Fly Ash
Fly Ash
Portland Ceient
Fly Ash
Fly Ash
I Additive
(By ft.)
25

25
50
50
100
50
75
100
Unconfined Strength
7 Days (tsf) 28 Days (tsf)
2.75
3.50 4.79
1.51
No Strength 1.69
No Strength
4.00
Ho Strength
1.50
2.55
.40
1.2
Volune
Increase}
-27.80
7.60
-5,30
19.80

17.00

35.80
16.50
27.00
42.60
Bet Den-
Sity (pcf)
91,7

94.2

84.2

98,0
83.0
88,6
92,4
in teachability by a factor of two when measured by the Extrac-
tion Procedure (EP) Toxicity Test. Each firm performed the
following tests on the sludge:
Total solids (raw sludge)
Wet unit weight (raw sludge)
Specific gravity
Unconfined compressive strength
Permeability
Wet unit weight (solidified)
EP Toxicity Test (eight metals only)
Consolidation
Volume change
The results of these evaluations are summarized in Table 6.
Table 6
Sludge Stabilization Results—Pond I

FIRM
65
in
m
1.2
5.2 i 10"'
95.7
63
106
1.70
31,1-37,4
<1,4 x Id'1
110-113
56.9
97.8
1,72
1,4
6.3ilO'J
88.5
Total Solids (!) (Raw Sludge)
Bet Density (Ibs/ft1) (Rav Sludge)
Specific Gravity
Unconfined Strength (tsf)
Peneability (en/sec)
Bet Density (lbs/ft!) (Stabilized)

EP lOXICITJ - EXTRACT CONCENTRATION (tqll)

Pb <0,005 0.25 0.4
Kg 0,19 0.0004 0.00006
Consolidation (! at 1.2 tsf) 2.1 <1 1*
Voluw Change (!) 0 t!3-16 +25

•(at 3,2 tsf lbs/ft!)
RESULTS

Raw Sludge Testing
As indicated previously, the raw waste was tested to determine
whether the material could be easily pumped. The test results,
which were confirmed by field-scale testing, indicated the Pond
0-2 material could be easily pumped due to its low proportion of
solids content. Sludge from the upper portion of Pond 0-1 was
pumped with only slight difficulty during the field tests. The
other sludges, with greater solid content and higher Marsh Fun-
nel viscosities, were not easily pumped during the field test. The
Pond 0-4 waste could not be pumped at all because of its tar-like
consistency. Wastes from Pond I, with higher specific gravity,
were difficult to keep in suspension and therefore were not
pumpable.
These test results indicate that the Pond 0-1 material and the
sludge from Pond 0-2 can be mechanically dewatered success-
fully. While Pond 0-1 sludge is amenable to in-place solidifica-
tion, significant reductions in the volume of waste to be placed in
the engineered containment facility can be realized by mechan-
ically dewatering the upper, less dense, portion of sludge. In-
place solidification of the bulk of the Pond 0-2 material is not
practical due to its low solid content. The residual sludge in Pond
0-2 that cannot be extracted for mechanical dewatering can be
solidified in place and excavated prior to disposal. Sludge from
Pond 0-4 will require in-place solidification because its tar-like
consistency precludes pumping; furthermore, it would quickly
foul the mechanical dewatering equipment.
Wastes from the lower portion of Pond 0-1 as well as wastes
from Ponds 0-3 and I are better suited for in-place solidifica-
tion than for mechanical dewatering because of their relatively
high solids content.

ID-Place Solidification Testing
The results of in-place solidification tests indicate that the
sludges can be solidified in-place without dewatering, however,
it is more effective to mechanically dewater the waste from the
upper portion of Pond 0-1 and from Pond 0-2 for the reasons
given above. The alternative in the case of Pond 0-2 would re-
quire large percentages of additive to achieve the required
strength. The sludge in Pond I, which had the highest percen-
tage of solids, greatly exceeded the strength requirements with the
addition of minimum percentages of cement (25%) or fly ash
(50%) used during the original phase of the testing program.
Additional testing indicated the Pond I sludge could be stabilized
to achieve the required characteristics using only 10% cement
as an additive.
Relatively high percentages of cement and fly ash are required
to solidify the sludge in Pond 0-4 because of the difficulty inher-
ent in mixing additives with the tar-like material. It appeared that
the cement and fly ash were forming bonds around floes of the
sludge instead of actually mixing with the particles in the sludge.
Significant differences were noted in the volume change values
obtained by the two independent laboratories performing tests.
These differences were caused by variations in the test procedure
in each laboratory. In this case, one laboratory measured volume
change immediately after mixing, thus resulting in higher values.
The other laboratory allowed the mix to set until all free water
rose to the surface and then removed it prior to taking measure-
ments.
Mechanical Dewatering Demonstration
The results of the dewatering demonstrations show that sludge
from all ponds except 0-4 could be mechanically dewatered. Test
results indicate at least a three-fold increase in the percen-
tages of solids in the cake as opposed to the raw waste. Pond 0-2
was found to have the greatest increase in the percentage of solids
in the cake material.
Several problems occurred during the dewatering of the waste
materials. It was found that the high oil content of Ponds 0-1
and 0-3 caused the filter medium to become clogged during the
dewatering operations. In full scale dewatering operations using a
belt filter press, this would severely hamper efficiency and pro-
duction. Use of a full scale plate and frame filter press for pro-
duction dewatering of the wastes would be a slow and inefficient
process since the batch method would have to be used to load the
press for each dewatering operation. The centrifuge performed
well once the proper polymer and unit geometry were established
TREATMENT 121
-------
and the unit was able to dewater continuously.
Analysis of the rake material indicated the moisture content
was above the liquid limit for all three samples tested. This de-
termination is significant because it indicates that the cake ma-
terial exhibits the properties of a liquid and therefore has no shear
strength. The high moisture content of the cake necessitates that
this material be solidified to meet requirements for strength and
compressibility.

Cake Solidification Tests
An analysis of solidified cake material was completed to deter-
mine the effectiveness of various sludge/additive mixes. As pre-
viously indicated, the filter cake material exhibited the properties
of a liquid prior to the addition of the solidification agents. As
shown in Table 4, the moisture content of the solidified wastes
dropped substantially and, more significantly, fell below the ma-
terial's liquid limit. This indicates that the material is workable
and lends itself to replacement and compaction.
All solidified mixtures except one (Pond 0-2 with 15% cement)
achieved the required strength of 1.2 tons/ft2 in 28 days. Consoli-

Tible7
Summary of Actions
SoUdlCtettioa/StaBlllutlon
•oUdlflcation at cake after
•echanleal devatering (or
upper portion of pond iludo*
In-plaee eolidlficatlon for
lower portion of pond eludge
JM Portland
Caaant
1M Portland
High charge cattonic polymer uaed
to latprova dawataring efficiency.
Sludge diapoaed in an on-alte
containment facility.

Cemtnt added by pneuMtic
on. Sludge diapoaed In en
containment facility.
Solidification of cake
mechanical dexatanng
In-pl«e« solidification

In-«l«ct Mlldlf icition

ln-eUc« ItAbillutlon
effici
and 0-
ncy. Sludge fro> pond 0-3
«aa blandad prior to
Mt Portland
Mt Portluu)
ing to provide a man
uniform aludge for dawatering.
Sludge diapoeed .n an on-aite
containment facility.

Sludge dlapoaad in an on-aita
dupoial facility. CaMnt added
by i
Mot posiibla te dcuttir th<
actflrUl prior to Bolidif icBtion.
C«M«nt Bddvd by pntiMtie
injKtioa.
ain* priMrily htavy
Mt*l con»titu*nta. stablliMtion
producad • »«t«ri«l with
•ubBtantially rtducMl la*ch
pot«ntl*l. Stibiluad *«t«rial
laft in-plae* and upp«d.
dation requirements also were met by all but this mixture under a
load of 4 tons/ft1. The 4-ton/ft1 load exceeds the design criterion
for the solidified waste.

CONCLUSION
Full scale field activities were begun using the data obtained
during the pilot and laboratory tests. A summary of the actions
taken in each pond is given in Table 7.
The centrifuge was chosen for mechanical dewatering because
its production rate is approximately 35 gal of raw sludge per min-
ute. This rate is greater than that obtained with either the belt
filter press or plate and frame filter press.
Portland cement was chosen as the solidification agent for sev-
eral reasons: cement is a manufactured product and is readily
available in sufficient quantities; the cost of cement is lower than
other alternatives when transportation costs are considered; sub-
stantial volume reductions were indicated for several of the
ponds; and the ratio of additive to sludge is lowest for cement.
Mechanical dewatering was performed on a portion of Pond
0-1 and on Pond 0-2. The upper 8 ft of waste in Pond 0-1 were
mechanically dewatered because of the relatively low percent of
solids present in the material. Estimates showed that over
12,000,000 gal of water (filtrate) could be removed from the
sludge by the centrifuge and that the volume of material to be
placed in the engineered containment facility was reduced by
approximately 60,000 yd3.
Sludge from Pond 0-2 was combined with that from Pond 0-1
prior to mechanical dewatering. This blending produced a more
consistent waste stream for the centrifuge and provided a cake
that was more responsive to solidification with cement.
Cake produced by the mechanical dewatering process was
solidified with 25% cement. The remaining ponds (Ponds 0-3,0-
4, and I) were solidified in-place using various ratios of sludge
and cement.
Mechanical dewatering of the material from Ponds 0-1 and 0-2
is complete and was accomplished in approximately 4 mo of con-
tinuous operations. Solidification of the remaining wastes it pro-
ceeding and is expected to be completed in mid-May, 1987. A
total volume of 250,000 yd3 of sludge is expected to be stabilized/
solidified at this time.
122 TREATMENT
-------
Effectiveness of In Situ Biological Treatment of
Contaminated Groundwater and Soils at
Kelly Air Force Base, Texas

Roger S. Wetzel
Donald H. Davidson
Connie M. Durst
Douglas J. Sarno
Science Applications International Corporation
McLean, Virginia
ABSTRACT
In situ biological degradation of organic contaminants was
tested at a waste disposal site at Kelly Air Force Base (AFB),
Texas. A small-scale soil and groundwater treatment system was
used to circulate groundwater by pumping and gravity injection.
Oxygen was supplied by hydrogen peroxide, which was added
along with specially formulated nutrients to the subsurface to
stimulate microbial degradation of the organic contaminants. Ex-
tensive background studies were performed to characterize the
site geology and hydrology, to determine the contaminant pro-
file and to demonstrate treatability in the laboratory. This pro-
ject was the first field application of in situ biological treatment
at a site contaminated with a complex mixture of both organic
and inorganic wastes.
The injection/extraction treatment system was operated for a
period of approximately 8 mo during which hydrogen peroxide
was added to the subsurface. Nutrients were added for a period of
approximately 5 mo. Routine monitoring of the soil and ground-
water was conducted to provide data to evaluate the system's
performance and control its operation.
Problems associated with operation of an in situ biological
treatment system and useful monitoring parameters are discussed
in the paper. In addition, results collected from groundwater and
soil chemical analyses performed during operation of the system
are presented. The effectiveness of biodegradation treatment of
the organic contaminants was assessed by documenting changes
in the microbial populations. Costs for applying the technology
for a generic site were evaluated and compared with those for
equivalent technologies. The demonstration project at Kelly AFB
has provided a better understanding of the capabilities and limi-
tations of the technology for treatment of organic contaminants
in soils and groundwater.

INTRODUCTION
In situ biological treatment of soils and groundwater contam-
inated with organic compounds is based on stimulating the in-
digenous subsurface microbial population to degrade the organic
contaminants. Conditions for contaminant biodegradation are
optimized by providing the nutrients and oxygen which may be
limiting factors for the growth of aerobic microbes in the sub-
surface.
In situ treatment offers advantages over conventional methods
such as technologies in which groundwater is pumped to the sur-
face for treatment. Because the active treatment zone is in the
subsurface, in situ biological treatment has the potential to both
remove contaminants sorbed to the soil matrix and treat the con-
taminated groundwater. Pump-and-treat methods treat only the
groundwater, allowing clean groundwater to become contam-
inated because desorption of the pollutants occurs when the
groundwater contacts the untreated soil. In situ biological treat-
ment is more desirable environmentally than excavation, removal
and disposal of contaminated soils which merely transfer contam-
inants to a more secure disposal area without treatment. In situ
biological treatment offers treatment of organic-contaminated
soils and groundwater and can be less expensive than conven-
tional treatment of disposal methods.
A field demonstration of the technology was conducted at a
waste disposal site at Kelly Air Force Base (AFB), Texas. The
site originally was used as a disposal pit for chromium sludges
and other electroplating wastes. Prior to closure in 1966 it had
been used as a chemical evaporation pit for chlorinated solvents,
cresols, chlorobenzenes and waste oils.
The Kelly AFB waste site was selected to demonstrate in situ
biological treatment for a number of reasons, because: the sub-
surface contained biodegradable organics and a highly adaptive
and substantial population of microbes; and the perched aquifer
was not used as a supply of drinking water. This project was the
first field application of the technology at a site contaminated
with a complex mixture of both organic and inorganic wastes.

SYSTEM OPERATION
The configuration of pumping and injection wells used to cir-
culate groundwater during the demonstration is shown in Fig. 1.
Nine 4-in. diameter pumping wells and four 6-in. injection wells
were operated within a 60-ft diameter treatment area. The satur-
ated zone within this area was at a depth of 15 to 25 ft below land
surface. One upgradient and two downgradient monitoring wells
were used to determine any effects of the system outside the
perimeter of the treatment area. The system was operated and
monitored from June 1985 to February 1986.

Nutrient Addition
For the first 2 wk of operation, groundwater was circulated
without the introduction of nutrients or hydrogen peroxide in
order to stabilize the aquifer. Nutrient injection began in mid-
June. The nutrient solution used consisted of chloride, nitrogen
and phosphorous. Originally, nutrients were introduced contin-
uously via a laboratory-type feed pump. Almost immediately
TREATMENT 123
-------
upon introduction of nutrients, a thick, white precipitate of cal-
cium phosphate developed in the distribution box and injection
wells along with a corresponding decrease in injection capacity.
Addition of nutrients was changed to a batch mode, and the con-
centration of nutrients added was changed in order to reduce the
precipitation problem.
Efforts were made to rehabilitate the injection wells; these
efforts included manual brushing of the well screens and air surg-
ing, followed by pumping and bailing. These activities removed a
large amount of the precipitate and resulted in a short-term in-
crease in injection rates. Low injection rates continued to be a
chronic problem, however, and nutrient injection was suspended
permanently after 5 mo of operation. During the time of nutrient
addition, it is estimated that 500 Ib of chloride, 500 Ib of phos-
phate and 254 Ib of ammonium ion (NH4 + ) were introduced to
the aquifer. An unknown percentage of these quantities was re-
moved as precipitate during well-cleaning activities.
The treatment system was originally designed to operate con-
tinuously and provide uniform pumping of groundwater and in-
jection of nutrient and hydrogen peroxide solutions to the aqui-
fer. Following construction and development of the injection and
extraction wells, slug tests were performed to determine the hy-
draulic conductivity and expected discharge of each extraction
well. Results from the slug tests indicated that the hydraulic con-
ductivity of the aquifer ranged from 0.11 ft/day to 9.26 ft/day.1
Due to the variable soil permeability, uniform pumping and injec-
tion rates could not be maintained throughout all areas of the
demonstration site. Pumping rates were decreased to maintain a
constant injection rate without flooding the injection wells.
\
\

Site
POM" o Circuit
Line Box
O O
P2 P3
O
11
o S0"
np,
Overflow
Tanks
0
M
O O
P8 P7
1
1
1
0' 10' 1
L i 1 '
Sole V - W i
1

Office

O
P4
O
12
riburton
Box
O
P5

O
13
O
P6
Figure 1
Configuration of Pumping and Injection System

Due to low permeabilities and the reduced injection capacities
resulting from the precipitate formation, it also became neces-
sary to alternate operation of the wells so that only about one-
2000-
•o
a.
•§1500
e
u

I
•o
c
3
o
(9
o
1000-
I
500-
10 15 20
Weeks of Operation
25 30
Figure 2
Average Groundwater Circulation Rates for Kelly AFB
Treatment System

half of the wells were pumping at any one time. The overall re-
sult was a dramatic decrease in groundwater circulation rates over
the life of the project as illustrated in Fig. 2.
Modifications in system operation were made during the pro-
ject, based on groundwater monitoring results, in order to pro-
mote transport of nutrients to the areas which had been left un-
treated. During the last months of the project, groundwater was
pumped from, and injected into, only those wells which had not
shown evidence of nutrient transport in order to enhance micro-
bial degradation and treatment of the contaminants within those
areas.

Hydrogen Peroxide Addition
Hydrogen peroxide was selected as the source of oxygen be-
cause it can provide approximately five times more oxygen to
the subsurface than aeration techniques. Other in situ bioreclama-
tion projects have been conducted for treatment of gasoline-con-
taminated aquifers using hydrogen peroxide as a source of oxy-
gen.2 Although hydrogen peroxide can be toxic to bacteria at
high concentrations, studies indicate that it can be added to soil
or groundwater systems at concentrations up to 100 mg/1 with-
out being toxic to microbial populations. Concentrations as high
as 1000 mg/1 can be added to microbial populations without toxfc
effects if the proper acclimation period is provided for the bac-
teria.'
Injection of hydrogen peroxide commenced in late June 8t^»
rate sufficient to provide a concentration of 100 mg/1 H202 *
the groundwater. The concentration was increased by increment*
of 100 mg/1 every 2 wk in order to acclimate the bacteria to the
hydrogen peroxide solution until a final level of 500 mg/1 ***
achieved. The loss of a supplier of peroxide resulted in a suipeB"
sion of hydrogen peroxide injection for several weeks in Novem-
ber. When peroxide was introduced again, the acclimation p**-
124 TREATMENT
-------
cess was repeated. A total of 299 Ib of hydrogen peroxide were
introduced into the system over the life of the project.

Groundwater Monitoring
Groundwater monitoring was a critical component of the treat-
ment system operation at Kelly AFB. A detailed monitoring
schedule was implemented to routinely sample groundwater from
the wells in the system and soil from the treatment zone in order
to monitor the performance and effectiveness of the system. The
groundwater monitoring schedule is outlined in Table 1. The
groundwater and soil sampling program was implemented prior
to startup of the system in May 1985 and continued for the dura-
tion of the project.
Groundwater was routinely analyzed for chloride, ammonium
and phosphate ion concentrations to determine if nutrients were
successfully transported through the subsurface and made avail-
able for microbial growth. Regular monitoring of these param-
eters indicated that nutrient transport varied throughout the site
due to variability in the hydraulic conductivity of the aquifer.
Carbon dioxide concentrations were measured in groundwater
samples from each well to monitor biological activity and system
performance. Since CO2 is an end product of microbial degra-
dation and chemical oxidation of organics, changes in ground-
water C02 values can indicate changes in the activity of the micro-
bial population.
The system was operated for 257 days during which a total of
52,270 gal of groundwater were circulated. Groundwater was cir-
culated every day except during well-cleaning. There were no
mechanical failures or system shutdowns due to equipment
failure.

Table 1
Groundwater Monitoring Schedule

Sampling Frequency
Parameter
Temperature
Conductivity
PH
Dissolved Oxygen
Carbon Dioxide
Ammonia
Phosphate
Chloride
Hydrogen Peroxide
Nitrate
Sulfate
Acidity
Alkalinity
Total Hardness
Chromium
Lead
Extraction
Wells
2/Week
2/Week
2/Week
2/Week
2/Week
2/Week
2/Week
2/Week
Weekly
2/Month
2/Month
2/Month
2/Month
2/Month
Monthly
Monthly
Injection
Wells
2/Week
2/Week
2/Week
2/Week
2/Week
Weekly
Weekly
Weekly
Weekly
2/Month
2/Month
2/Month
2/Month
2/Month
—
---
Monitoring
Wells
Weekly
Weekly
Weekly
Weekly
Weekly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Monthly
Oil and Grease

Total Hydrocarbons
(alkanes)
Priority Pollutants
(volatile organics
and metals)

Microbiological

Groundwater
Elevations
Monthly Monthly Monthly

Quarterly Analysis of 10 Wells

Quarterly Analysis of 10 Wells
Monthly

Daily
Monthly

Daily
RESULTS
Inorganic Chemical Results
Communication of nutrient solution between injection and
pumping wells was evaluated by measuring the concentration of
chloride in the groundwater. Chloride was selected as the tracer
chemical because concentration levels in the nutrient were two
orders of magnitude greater than background levels. Chloride
communication was observed in all nine pumping wells. The level
of this communication varied significantly between wells and did
not correlate with initial permeabilities. Some areas of the treat-
ment zone where hydraulic conductivity was low and transport
of nutrients was expected to be minimal exhibited transport of
the chloride tracer similar to more permeable areas. Factors such
as well placement relative to the direction of groundwater flow
and the presence of gravel and clay layers seem to have influ-
enced interwell fluid communication to a greater extent than soil
permeability. The evidence of elevated chloride concentrations
in all nine wells indicates that the opportunity existed for oxygen
and nutrients to be transported to microbes throughout the en-
tire treatment area. It also was shown that injected chemicals were
contained largely within the treatment area because very low
chloride concentrations were observed in the monitoring wells.
Ammonia and phosphate communication was less than chlor-
ide ion communication. Ammonia communication was observed
in only two wells. Although free ammonia was not injected into
the groundwater, ammonia is used to report nitrogen levels added
as ammonium ion (NH4 + ). The low level of communication
observed indicates a substantial uptake of ammonia and phos-
phate by the soil and/or precipitation into the flow channels of
the soil. The uptake of nutrients varied significantly throughout
the treatment area and did not correlate with the direction of
fluid communication.
Precipitation could have been caused by reaction with calcium
ion exchanging clays in the soil. The calcium ion exchanges with
the ammonium ion and then reacts with phosphate to precipi-
tate calcium phosphate. This reaction would proceed until all
the calcium exchanging clays were saturated with ammonium ion.
The resulting calcium phosphate would reduce permeability by
clogging pores in the soil. It was estimated that only 24% of the
potential ammonia uptake was injected into the treatment area.
This factor combined with variable communication rates indicates
a highly variable exchange capacity throughout the well pattern.
Field measurements of dissolved C(>2 showed levels five to ten
times greater in the pumping wells than in the injection fluid.
This result indicates a net gain of C02 as groundwater circulated
through the treatment zone. There are several likely sources of
C02- First, C02 is an end product of microbial degradation of
organics which would provide evidence of contaminant degra-
dation. Second, C02 could be generated through chemical reac-
tions such as the reaction of oxygen (supplied as H202) with
organic carbon or acid with calcite in the soil. Variations in C02
levels between pumping wells were less than the variations ob-
served for nutrient communication. A possible explanation is
that peroxide decomposes in the soil to free oxygen gas and water.
This mixture reduces flow in the higher permeability zones rela-
tive to the lower permeability zones resulting in flow equaliza-
tion. Free gas also can be transported through the groundwater
and reach contaminated areas ahead of the injected fluid.
Based on metals analysis of soils prior to and during system
operation and metals analysis of sediments obtained from pro-
duced fluids, metals appear to have been mobilized from the soil
and transported by the circulating groundwater. In addition,
clay, calcite and iron minerals in the soil appear to be entrained
in the circulating groundwater. Sediments deposited in the surface
piping were found to contain these compounds. A laboratory
microcosm study with soil and groundwater from the Kelly AFB
site demonstrated the same effects of metals removal from the
soil by a hydrogen peroxide solution as observed in the field test.
Oxidation of metals by hydrogen peroxide is one possible meth-
TREATMENT 125
-------
od of initiating metals mobilization. The groundwater pH is too
high to maintain oxidized metals in solution, and they would
likely precipitate as oxides which could then be transported by
the circulating flow. However, the oxidation process could con-
sume a considerable fraction of the oxygen content of the per-
oxide and would then be unavailable for aerobic interaction with
organic contaminants.

Microbiological Results
A reduction in microbial cell numbers in the injection wells
was observed during the period of highest hydrogen peroxide
injection levels, suggesting that the level of hydrogen peroxide
used was toxic to the microbe population. However, subsequent
reductions in concentration of hydrogen peroxide did not restore
microbial counts. Another mechanism that may have contributed
to the decline of microbial counts was entrapment of microbes
on the precipitate that formed in the injection wells. A similar de-
crease in microbial numbers was observed for soils sampled in
the treatment area; however, microbial numbers in pumping wells
remained relatively stable.

Organic Chemical Results
Total hydrocarbon and oil and grease soil concentrations for
the treatment phase are shown in Table 2. Variations in both
measurements over the initial 200 mg/1 levels indicate both a sig-
nificant localized increase and a reduction in total hydrocarbon
and increases in oil and grease. Since these changes exceed the
level of chemical reactivity of the hydrogen peroxide, it is likely
that addition of peroxide and nutrients to the groundwater may
have resulted in mobilization and redeposition of the materials
within the well pattern indirectly by formation of an intermed-
iate chemical such as a surfactant.
Total hydrocarbon groundwater concentrations at selected
wells are given in Table 3. Although variations between sam-
pling points are large, comparison of samples collected Aug. 8,
1985 and Sept. 12, 1985 with those collected Dec. 5, 1985 and
Feb. 17, 1986 indicates a quantitative reduction in the ground-
water organic level.4
Table 2
Concentrations of Total Hydrocarbons and Oil and Grease
In Soil During Treatment
Table 3
Total Hydrocarbons In Groundwater, GC Scan (ppm)
Hell
SB-1 (20'-22'|
S«-2 (ZO'-JZ1)
Sl-3 (20'-22')
SI-4 (20'-22')
SI-5 (20'-22'J
SI-6 (20' -2!')
SI-7 (20'-22')
SI-7 (2S'-27')
S«-« (20--22')
Dili
08/08/85
08/08/85
08/08/85
08/08/85
12/05/85
12/05/65
12/05/85
12/05/85
12/05/85
ppa Total Hydrocarbon
NO (0.05)
NO (0.05)
0.071
HO (O.OS)
Ml (50)
HO (50)
1000
HO (50)
330
pp« Otl 1 Grease

900
HO (200)
1300
800
600
SB-1, SB-5 tame area of well pattern.
SB-2. SB-6 same area of well pattern.
SB-3. SB-7 iamc area of well pattern.
SB-4, SB-6 same area of well pattern.

The field data pertaining to changes in well pattern ground-
water concentrations of volatile priority pollutants prior to and
during treatment are shown in Table 4 and indicate direct evi-
dence of contaminant degradation. Tetrachloroethylene (PCE)
Staple ID
PI
P2
P3
P4
P5
P6
P7
P8
P9
11
12
13
14
Ml
M2
CC
Blank
Bailer Wash
8/8/8S
NO (0.05)
ND (0.05)
-
6.80
-
ND (0.05)
-
ND (0.05)
0.580
0.150
.
1.40
730.0
_
4700.0

9/12/85

-
49.0

-
-
280.0
ND (0.05)
ND (0.05)
-
1.70
ND (0.05)
ND (0.05)
18.0
ND (0.05)
15.0
ND (0.05)
12/5/85 2/17/8S
ND
ND
0.0|j »(,.„,
-
NO (O.OS) 0.36
0.31
NO
NO
NO
ND
0.05 0.076
0.05 1
0.05 0.20
0.05 O.J7
m
NO
ND
ND
0.05)
0.05 ND (O.OS)
0.05) NO (0.05)
„
ND
ND
ND
ND
0.05) ND (2.4)
0.05) 3:8
0.05) 0.055
O.OS) 0.061
- Well was not sampled.
ND (*)—Not detected (detection limit).

Table 4
Summary of Groundwater TCE, PCE and Trans-l,2-DCE Chingei
During Treatment
Priority
Pollutant
TCE
(Decrease)
PCE
(Decrease)
Trans-1,2 DCE
(Increase)
Vinyl Chloride
(No Change)
Chi orobeniene
(Initial Rise
Ho Net Increase)
Methylene Chloride
[Increase)
Date
OS/23/85
Prior to
Treatment
08/08/85
12/05/85
02/17/86
05/23/85
Prior to
Treatment
08/08/85
12/05/8S
02/17/86
05/23/85
Prior to
Treatment
08/08/85
12/05/85
02/17/86
05/23/85
Prior to
Treatment
08/08/85
12/05/85
02/17/86
05/23/85
Prior to
Treatment
08/08/85
12/05/85
02/17/86
05/23/85
Prior to
Treatment
08/08/85
12/05/85
02/17/86
Average All
Production
Hells (opal
2.3
1.7
0.69
0.47
1.7
1.8
0.74
0.46
0.027
2.4
0.76
1.4
0.26
0.32
0.11.
0.23
0.021
0.049
0.095
0.029
<0.005
<0.026
0.30
0.12
14
No Data
No Data
0.21
.009
No Data
No Data
0.21
0.023
No Data
No Data
0.4
0.026
No Data
No Data
No Data
No Data
No Data
No Oata
0.059
HO (0.015)
N2
No Data
No Data
5.S
7.1
No Data
No Data
1.0
10
No Data
No Data
1.0
2.0
No Data
No Data
NO (1.0)
0.14
No Oata
No Data
1.0
.1
No Data
No Data
•w
CC
No Data
No Data
O.ON
9.4
No Data
No Data
0.11
0.10
No Oata
NoUU
M
I.I
loData
No Data
NO (.001)
0.11
NO Data
•oOata
.on
0.1
MMa
MMta
»«r,
No Data—Sample not analyzed for this parameter.
ND (*>—Not detected (detection limit)
—Detected at a value less than that shown.
and trichloroethylene (TCE) were two of the contaminant! in-
itially present at the highest levels in the well pattern ground-
water. The total TCE and PCE concentrations averaged <•"
mg/1. A consistent decline in the sum of their concentrationi
was observed throughout treatment; the final total concentratioii
126 TREATMENT
-------
DAT* GENERATED
•LKll Hltlr tupplltl
I •Topogriphy
MntiUy unlttlvt irtts
•Greumlviltr depth fHydriuUC conductlvltltl
•PirMibllltUi •SiturilwJ »ni
•Groumhatir flow oSoU typli, hoaoaintUy
iTyp«i of •Icrobti pnitnt
iDlitrlbutlen of BlcrolMi pnlint
tConcintntlon of •Icrobti prount

•Fodtr«l, Stltt, ind locil rtflglltlo
Figure 3
Site Characterization for In Situ Treatment
of PCE and TCE was approximately 0.93 mg/1. A corresponding
increase in trans-l,2-dichloroethylene (DCE) from 0.03 to 1.4
mg/1 was observed; DCE is a decomposition product of PCE
and TCE. A similar trend in PCE, TCE and DCE concentra-
tions over a 100-day period was observed in the laboratory micro-
cosm tests which were conducted as part of the field test design.4

APPLICABILITY TO OTHER SITES
A major emphasis of the Kelly AFB project was to develop
guidance for the future use of biological degradation at contam-
inated sites. Specific attention was paid to site characterization,
engineering components and cost estimation.
Characterizing a hazardous waste site is one of the most com-
plex activities in any remedial action. The site characterization
requirements for in situ treatment are even more extensive than
for most other remedial alternatives. Fig. 3 outlines the site char-
acterization process associated with in situ treatment, presents the
associated tasks and lists the data that can be expected from each
task. All or most of these data will be necessary before an in situ
treatment system can be designed. Activities associated with the
general approach for characterizing DOD or CERCLA sites are
identified in Fig. 3 above the dotted line; activities more specific
to in situ treatment are listed below the dotted line. Major differ-
ences from those activities normally performed during a remedial
investigation are groundwater and soil treatability studies.
The engineering components for a groundwater circulating, in
situ treatment system will be similar regardless of the specific
site. The following paragraphs present a number of factors that
should be given consideration when planning an in situ project.

Well Design
The size, number and spacing of pumping and injection wells
will depend largely on site-specific characteristics. The ratio of
pumping to injection wells should be maintained so that no ex-
cess groundwater is generated. The cost of storage, transport or
disposal of large amounts of untreated groundwater would neg-
atively impact the cost-efficiency of in situ treatment. The spac-
ing of wells can be varied over a site to compensate for subsur-
face variabilities. Sufficient space should be provided when siz-
ing the wells to allow for the pump, piping, electrical equipment
and sampling and monitoring devices.

Transport and Mixing System
Many options are available for the mixing of nutrients and
hydrogen peroxide with the groundwater. The most desirable op-
tion is an in-line system whereby nutrients and peroxide are added
directly into the transport lines between pumping and injection
wells. At low flow rates, it is necessary to closely monitor pump-
ing and injection volumes and a batch-type mixing system will
probably be most feasible. Storage tanks should be available to
prevent overflow and provide a sampling point to analyze circu-
lating groundwater.
The size and location of a central mixing system(s) will depend
largely on the flow rate, size of the site and site topography. Grav-
ity is the least expensive and most reliable method of ground-
water transport. Pumping wells may provide enough pressure to
transport water to the central mixing tank in cases where the
topography prohibits gravity flow.

Treatment Chemicals
Precipitation of nutrients was a major problem at the Kelly
AFB demonstration site. It is important to quantify this problem
prior to system operation and take measures to prevent it from
adversely affecting the system. Hydrogen peroxide is an effective
supplier of oxygen and requires little special equipment. Other
methods of oxygenation, such as forced air, require additional
mechanical equipment.

Equipment and Material Selection
Durability is the primary factor to consider when selecting ma-
terials and equipment for use in an in situ treatment system. The
system will operate continuously and place a great deal of stress
on pumps, piping and electrical controls. All materials also must
be nonreactive and noninterfering. All metal surfaces should be
corrosion-free, and pump housings must be durable enough to
prevent oil leakage.

Operational Characteristics
The system will be operated continuously over a long period of
time; as high a degree of automation as feasible should be imple-
mented. More complex systems at very large sites could be de-
signed to regulate flow and adjust pumping rates automatically.
Although this system would be expensive to install, the cost prob-
ably could be recovered in reduced operating expenses.
Sufficient provisions for sampling and monitoring are integral
to any in situ treatment system. Wells should be designed large
enough so that groundwater samples can be collected. In addi-
tion, a great deal of on-site analysis may be performed. A suitable
laboratory facility and an area for decontaminating field equip-
ment and disposing of laboratory wastes will be necessary.

Construction of the System
Construction always will begin with well installation; it is im-
portant that wells be installed properly. A poorly developed well
will be of little use to the treatment system due to reduced pump-
' ing rates. Pumps, piping and electrical equipment placed in the
wells should be constructed so that they can be removed easily
without disrupting the entire system.

Startup and Operation
Once construction is complete, groundwater should be circu-
lated for a few weeks without treatment chemicals. This initial
circulation should provide sufficient time to adjust pumping and
injection rates and remove all residual fines from pumping wells.
With pumping and injection rates known, it then becomes a sim-
ple matter to calculate the desired nutrient and peroxide concen-
trations. This initial period also provides a shakedown of the
system so that any operational problems may be resolved before
treatment is commenced.
TREATMENT 127
-------
The first few weeks of treatment should be monitored care-
fully to determine system performance. Results of frequent on-
site testing will determine chemical transport and guide decis-
ions with respect to the concentrations of treatment chemicals
and/or the rates of pumping and injection. Once the system is
operating effectively, the frequency of on-site sampling and
analysis can be reduced to a reasonable level in order to provide
long-term performance analysis. On-site analysis is not expensive
or difficult, but it can be very time consuming. Field staffing
will be scheduled according to the quantity of tests to be per-
formed.

Cost Analysis
To facilitate costing of an in situ degradation project, a generic
cost model was developed. Our model consists of a number of
worksheets which detail major cost items for site characteriza-
tion, design, construction and operation. These worksheets use
site characteristics and provide the formulas necessary to develop
a first cut cost estimate at a proposed site.
Using these worksheets and the actual costs incurred at Kelly
AFB, a cost sensitivity analysis was performed. This analysis
showed that the major construction costs were incurred through
well installation and labor. Construction materials will vary in
cost depending on the site conditions, but should not constitute
a large portion of capital costs. During operation of a system, the
major costs incurred are due to sampling and analysis to monitor
system performance. The cost of treatment chemicals will be
minor in comparison.

CONCLUSIONS
A number of conclusions were reached based on the informa-
tion gained from the operation of this system. These conclusions
relate specifically to the Kelly AFB site and also to the applic-
ability of the technology to other sites.
• The mechanical system used at Kelly AFB circulated ground-
water with very little maintenance and no downtime due to sys-
tem failure. The low and variable permeability required alter-
ing the system operation and increasing the effort and time re-
quired for treatment. The ultimate treatment objective, how-
ever, was achieved. The circulation system was sufficient to
overcome the natural groundwater flow in providing commun-
ication to all wells.
• The introduction of nutrients and hydrogen peroxide to the
system resulted in an almost immediate and significant de-
crease in permeability and heavy precipitation of calcium phos-
phate. A large uptake of ammonia and phosphate by soil was
also indicated, while a significant production in C02 was ob-
served and could have accounted for as much as 30% of the
hydrogen peroxide supplied to the system.
• Chemically, the sum of PCE and TCE levels in site ground-
water decreased from 4.0 ppm to 0.93 mg/1, while trans-1,2-
DCE increased from 0.03 to 1.4 mg/1. Mobilization of heavy
metals (antimony, arsenic, cadmium, lead, silver and thallium)
from soils was indicated but was not detected in the ground-
water. However, heavy metals were found in sediments de-
posited in several surface pipes.
• Microbial numbers decreased in the soil and injection wells but
remained relatively constant in the pumping wells.
• The cost of performing in situ treatment is less than that for
removal and redisposal of soils. It is estimated that in situ treat-
ment can be performed for $50 to $100/ton of contaminated
soil for a typical site. The estimated costs for treating the Kelly
AFB site were $36/ton of contaminated soil; analytical costs
accounted for $9/ton or 25% of the total cost.

RECOMMENDATIONS
To effectively implement in situ biodegradation at a full-scale
application, a number of modifications in physical, chemical
and analytical procedures are recommended based on the opera-
tion at Kelly AFB. For long-term operation, it is necessary to in-
crease the amount of automation used for pumping, chemical
mixing and injection. Increased economy of sampling and chem-
ical analysis would need to be implemented, with an emphasis on
more cost-effective on-site analyses by limiting the use of off-site
laboratories. Additional experiments should be implemented to
delineate the treatment mechanism responsible for contaminant
reduction (aerobic vs. anaerobic vs. chemical oxidation).
Future biodegradation projects should be developed to further
investigate the pumping/injection well approach to groundwater
circulation and compare it to alternative methods such as infil-
tration galleries. The generic cost model developed as a result of
the Kelly AFB project should also be applied to these future pro-
jects to determine its usefulness and/or refine it to make more
accurate estimates of cost.
Technical issues requiring additional study include the deter-
mination of a nutrient and hydrogen peroxide formulation that
maximizes the supply to groundwater while minimizing soil per-
meability damage. It also will be necessary to identify the mechan-
ism and chemistry controlling the mobilization of heavy metals.
An optimal well pattern should be developed with regard to nu-
trient and hydrogen peroxide application to maximize contam-
inant reduction.

ACKNOWLEDGEMENT
This effort was jointly sponsored and funded by the Air Force
Engineering and Services Center and the U.S. EPA. SAIC per-
formed a significant portion of this work under contract to
EG&G Idaho, Inc.; the project manager was Barbra Broom-
field. The authors wish to acknowledge the assistance of Captain
Edward Heyse of the U.S. Air Force and Stephen C. James of
the U.S. EPA.

REFERENCES
1. Wetzel, R.S., Durst, C.M., Sarno, D.J., Spooner, P.A., Jama,
S.C. and Heyse, E. "Demonstration of In Situ Biological Degndi-
tion of Contaminated Groundwater and Soils." Proc. of the Sixth
National Conference on Management of Uncontrolled Htaardoui
Waste Sites, 1985,234-238.
2. Yaniga, P.M. and Smith, W. "Aquifer Restoration via Accelerated
In Situ Biodegradation of Organic Contaminants." Proc. of Iht
NWWA/AIP Conference on Petroleum Hydrocarbons and Organic
Chemicals in Groundwater Prevention, Detection and Restoration.
Houston, TX, 1984,451-472.
3. Texas Research Institute, Inc. "Enhancing the Microbial Degrada-
tion of Underground Gasoline by Increasing Available Oxygen."
American Petroleum Institute, Washington, DC, 1982.
4. Wetzel, R.S., Durst, C.M.. Davidson, D.H. and Sarno, D.J. "to
Situ Biological Degradation Test at Kelly Air Force Base, Volume
II." Engineering and Services Laboratory, Air Force EngineeriBI
and Services Center, Tyndall AFB, FL, 1987.
128 TREATMENT
-------
Oil-Site Destruction of Organic
Contaminants in Water

D.G. Hager
C.G. Loven
Rubel and Hager, Inc.
Tucson, Arizona
ABSTRACT
Chemical oxidation of phenolic contaminants in groundwater
was evaluated for three types of groundwater problems using an
ultraviolet light (UV)-hydrogen peroxide (H202) process. Com-
plete destruction of the phenolic compounds was accomplished
in 20 min or less of oxidation time.
The phenolic compounds oxidized by this method included
multisubstituted chlorinated phenols and 2,4 dichlorophenoxy
acetic acid. The phenolics were reduced from 100-200 mg/1 start-
ing concentration to non-detectable levels at a direct operating
cost of l.SC/gal. Process variables and economic projections for
the process are described.

INTRODUCTION
On-site destruction of toxic organic wastes is the inevitable re-
sult of the environmental legislation regarding land disposal
which has been adopted by the U.S. EPA. It is only a matter of
time for both small and large generators of toxic and hazardous
waste until off-site disposal alternatives reach cost levels assoc-
iated with on-site treatment alternatives. Besides future cost-
competitiveness, on-site distinction will avoid the uncertainty of
legal liability for proper disposal of organic wastes by others.
Organic wastes can be destroyed by various forms of oxidation
including biological, thermal and chemical treatment. Of these
three methods of treatment, only chemical oxidation offers ulti-
mate destruction of the organics without residual sludges, ash or
air discharge. Hydrogen peroxide (H202) is a readily available
chemical oxidant which, when catalyzed by ultraviolet light (UV),
reforms into hydroxyl radicals which are the strongest chemical
oxidants next to fluorine.
The UV-H202 process has been extensively studied over the
past several years for its applicability for destruction of a broad
spectrum of concentrated aqueous wastes, industrial effluents
and groundwaters containing toxic solvents, fuels and pesticides.
The results of these studies clearly conclude that the UV-H202
treatment process works well for most types of water and waste
problems. The economics of the process make it competitive with
incineration for concentrated wastes which are too dilute or too
small in volume for economic thermal oxidation. The UV-H202
process is competitive with granular activated carbon treatment
for industrial effluents and groundwater applications when ulti-
mate destruction of the organic contaminant is included in the
economic analysis.
The UV-H202 process has been selected for application at sev-
eral locations. Fig. 1 is a photograph of a recent installation in
which the UV-H202 chemical oxidation process is used to destroy
concentrated organic contaminants contained in a steam conden-
sate. At this facility steam is used to periodically regenerate gran-
ular activated carbon used for groundwater purification.'
Figure 1
UV-H202 System to Destroy Concentrated Organic Contaminants

Fig. 2 is a photograph of a small UV-H202 system designed for
use by small quantity generators of toxic and hazardous wastes.
In this system, the aqueous organic wastes are recirculated
through the UV reactor as a batch until purification objectives
Figure 2
Small UV-H202 System Designed for Small Quantity Generators of
Toxic Wastes
TREATMENT 129
-------
have been obtained. This unit presently is employed for the de-
struction of pesticide residues found in the washwaters from pes-
ticide application equipment.
Fig. 3 shows a UV-H^ system which has been in use since
1982 for the purification of groundwater contaminated with
various industrial solvents.1
Figure 3
UV-H20j System for Purification of Solvent Contaminated Groundwater
Phenol is an organic contaminant commonly found at low lev-
els in various groundwaters and waste effluents. Phenol at a con-
centration of 1 mg/1 or less typically can be found in a variety of
water sources and can readily be removed by oxidation or adsorp-
tion.
Phenol, however, when stored, used as a reactant or improper-
ly disposed of, can produce waste or groundwater with signifi-
cantly higher concentrations. Phenol concentrations ranging
from 10 to 2000 mg/1 are commonly found in process wastewater
and contaminated groundwater.
Data presented in this paper describe the successful destruc-
tion of phenolic compounds found in three types of contaminated
water.

UV-H202 PROCESS DEVELOPMENT
PROTOCOL
Testing Equipment
The UV-H202 process development equipment includes a 2.3-
gal cylindrical stainless steel reactor, variable speed and H^
metering pumps, flow rotameter and feed reservoir with cool-
ing coils for temperature control (Fig. 4).
The chemical oxidation reactor contains the UV lamp enclosed
in a quartz tube and a UV detector. The UV lamp and tube are
specially selected for each application and are mounted axially
through the reactor wall. The UV detector is mounted through
the reactor wall perpendicular to the lamp.
Ancillary equipment includes a power panel with UV lamp
ballast, thermocouples for monitoring process temperature and
pressure gauge for measuring reactor pressure.
All materials in contact with the liquid are either 316 stainless
steel, Teflon, ceramic or quartz glass.

Testing Procedures
In these batch studies the solution to be purified was recircu-
lated through the UV reactor as a batch. The organic contami-
nant destruction progress was monitored by periodic sampling
and analysis. The H202 oxidant concentration was measured
throughout the test. The resultant H^ data then provided per-
formance criteria relative to H^ and UV dose over a period of
time.
The selection of UV and H202 doses to be evaluated were based
on the raw water analyses. Both total organic and specific con-
taminant concentrations can affect \he UV-H2O2 demand. Inor-
ganic analyses also provide insight into potential reactions which
could inhibit or catalyze the organic contaminant oxidation.
The UV dose for each test was a function of the wattage of the
UV lamp being used and the reaction time. The H^ dosage was
adjusted throughout the testing period for each sample. Any sus-
pended solids were removed by settling or filtration before start-
ing each test.
HXH-CXJ—,
Figure 4
UV-H2U2 Process Development Testing Equipment
When the preliminary steps were complete, the water wtt
pumped into the UV unit and the UV lamp was illuminated. The
recycle flow was adjusted to the desired rate to insure turbulent
flow in the reactor, and the H^ metering pump was started.
The test solution temperature was controlled by the flow of
water through the reservoir cooling coil. The cooling water flow
rate was adjusted to maintain the reactor temperature in the range
of maximum UV production for each lamp. Samples were col-
lected for analysis at the reactor discharge at selected time inter-
vals throughout the test.
At the end of each test the UV test unit was shut down tad
drained. The lamp and quartz tube were removed and inspected
for materials which may have coated the surfaces and interfered
with subsequent tests.
Destruction of Phenol in Groundwater
The UV-H202 process can be both efficient and co»t-«fften>«
for the destruction of phenol in aqueous wastes. Proper ulertifl"
of process variables as well as treatment system design wfll en-
sure successful application of the UV-H^ process.
Fig. 5 illustrates the importance of selecting the correct degree
of UV energy for each waste. In this test, a 100 mg/11
130
TREATMENT
-------
o
o
o
5
CCL

z:
QJ
^ 0.4 .
o
P
o
u_
0.2 -

0 ,
HZUZ uust = luuu mg/1
\ UV ENERGY = 250 Watts/1
\
\
\
\

\
\

x*- .
0 10 20
OXIDATION TIME, MINS.
30
Figure 6
Photochemical Oxidation Rate for Phenol in Groundwater.
solution of phenol was exposed to four different levels of UV en-
ergy (1000, 250, 150 and 75 watts/1). Phenol destruction was
plotted against reaction tune. The reaction time for phenol de-
struction is inversely proportional to the quantity of UV energy.
The oxidation time for 99+ % destruction was 5 and 15 min re-
spectively for the higher UV dosage levels.
Fig. 6 illustrates the photochemical oxidation rate curve for an
actual groundwater sample containing 120 mg/1 of phenol. For
this test, a UV energy level of 250 watts/1 was selected. The meas-
ured phenol destruction rate, was, however, lower than predi-
cated from the data presented in Fig. 2 because of the higher
initial concentration of phenol (120 vs 100 mg/1) and the effects of
using actual groundwater in place of the distilled water used in
the previous tests. Typically, groundwater contains organic and
inorganic constituents which tend to reduce the oxidation rate
when compared to experimental tests using distilled water.

Destruction of Penta, Tetra and Trichlorophenol in
Process Wastewater and Groundwater
Pentachlorophenol (PCP) is used primarily as a wood preser-
vative and is used extensively for termite protection. PCP-con-
taminated wastewater is generated as a result of the manufacture,
application or disposal of PCP and is commonly found as a
groundwater contaminant as a result of current or past wood
treating operations.
PCP has been designated by the U.S. EPA as a dioxin waste.
Although carbon adsorption is an effective treatment process for
the removal of PCP, the disposal of the dioxin-contaminated car-
bon remains a serious problem. Photochemical oxidation using
the UV-H202 process offers the advantage of efficient on-site
1.0
0.8-
o
o
o
"
CJ
0.4.
0.2.
H202 DOSE = 500 mg/1
UV ENERGY = 250 Watts/1
A PENTACHLOROPHENOL Co = 20.1 mg/1
* TETRACHLOROPHENOL Co = 2.7 mg/1
• TRICHLOROPHENOL Co = 2.3 mg/1
0 10 20 30
OXIDATION TIME, MINS.

Figure 7
Photochemical Oxidation Rates Multisubstituted
Chlorophenols in Groundwater
TREATMENT 131
-------
destruction of PCP and its related compounds.
Fig. 7 shows the photochemical oxidation rates achieved using
actual on-site wastewater containing penta, tetra and trichloro-
phenol in concentrations ranging from 2-20 mg/1. The plot indi-
cates 95 % destruction of the chlorophenols after 20 min of oxida-
tion. Continued photochemical oxidation reduces the chloro-
phenol concentrations to below detectable levels.
Wastewater containing concentrations of up to 400 mg/1 of
chlorophenols also have been successfully treated at comparable
oxidation rates using the UV-H202 process.

Destruction of 2,4 Dlchlorophenoxy Acetic Acid
(2,4 D) and 2,4 Dkhlorophenol (DCP) in
Herbicide Leachate
Samples of wastewater containing 100-150 mg/1 of 2,4 D were
obtained from an industrial site where a leachate collection sys-
tem is used to intercept drainage from a herbicide-contaminated
area. This leachate waste also contains 10-15 mg/1 of DCP.
Preliminary photochemical oxidation studies using the UV-
H202 process were first conducted to ascertain the relative oxida-
tion rates of the two major components of the leachate waste.
The waste under study contained 335 mg/1 2,4 D and 10.4 mg/1
2,4 DCP. Fig. 8 illustrates the results of the photochemical oxida-
tion of this wastewater. The concentration of 2,4 D decreased
consistently as a function of time throughout the test. The con-
centration of 2,4 DCP, however, increased during the first 30 min
350
300
i
«\
a
4
O
250.
1200-
i
a.

o
_j
3:

5
a-
CN
1150.
100.
50.
60
H202 DOSE = 7200 mg/1
UV ENERGY = 250 WattS/1
* - 2,4 D Co = 335 mg/1
• - 2,<4 DCP Co = 10.a mg/1
D-
g
o
30
20 -
10
0 30 60 90 120
OXIDATION TIME, MINS.

Figures
Comparative Oxidation Rates for 2,4 D & 2,4 DCP
(Preliminary Tests)
150
of the test from 10.4 to 35 mg/1. The concentration of 2,4 DCP
then decreased over the remainder of the test at a rate which
approximated the rate indicated by the DCP curve. The results
of this preliminary test indicate the oxidation pathway for 2,4 D
may include partial oxidation to 2,4 DCP during the early stages
of the oxidation.
1.0 '.
0.8 -
o
o
H202 DOSE = 1200 mg/1
UV ENERGY = 250 WattS/1
X - 2,4 D CO = 144 mg/1
- 2,4 DCP CO = 1H mg/1
0.6
0.4
0.2
10 20
OXIDATION TIME, WINS,
Figure 9
Oxidation Rates for 2,4 D & 2,4 DCP in Herbicide Leachate
(Optimized Conditions)

4 -I
•x.
ID
O
o
1 .
CASE #3

CASE #1
0 10 20 30
OXIDATION TIME, MINS,

Figure 10
Estimated Operating Cost for UV-H^ Treatment
132 TREATMENT
-------
1000 .,
Q.
<£>
o

-------
Waste Management Plan for
Veterans Administration Medical Centers

J.I. Bregman, Ph.D.
Edward Findley
Bregman & Company, Inc.
Chevy Chase, Maryland
Wayne Warren
Veterans Administration
Washington, D.C.
ABSTRACT
A waste management plan has been developed to assist Veter-
ans Administration (VA) field facilities and, in particular, Med-
ical Centers (VAMCs) in complying with waste handling and dis-
posal requirements established under RCRA and its 1984 Amend-
ments. This plan focuses primarily on application of current
waste management approaches to the segregation, handling,
treatment and disposal of radioactive, infectious, chemical and
physically hazardous wastes generated during operation of a full
service health care facility. The paper discusses the processes by
which hospital hazardous wastes are identified and handled.
Problems associated with compliance at a large number of differ-
ent types of facilities around the country are discussed.
The final product of this study was a three part VA Waste
Management Plan as follows:
• Part I-Hospital Waste Handling Regulations/Characteristics
• Part II-Hospital Waste Management Methodology
• Part Ill-Hospital Waste Management Evaluation & Control

Parts I and II are discussed in this paper.

INTRODUCTION
The objectives of the project described in this paper were to
present guidelines for Veterans Administration Medical Centers
(VAMCs) and other VA medical facilities to use in establishing
waste management systems and to make specific recommenda-
tions concerning VAMC waste classification and handling.
The Veterans Administration operates 172 medical centers; 226
outpatient clinics; 16 domiciliaries; a prosthetic center; a pros-
thetic distribution center; 58 regional offices, including two in-
surance centers; 17 VA regional office activities; 109 cemeteries
with one under construction; 3 cemetery area offices; 3 canteen
field offices; a canteen finance center; 5 data processing centers;
a records processing center; a marketing center; and 3 supply de-
pots. These facilities are located in every state, the District of
Columbia, in Puerto Rico and the Philippines.
During Fiscal Year 1984, over 1.4 million inpatients were
treated either within VA medical care facilities or in non-VA facil-
ities under VA authorization. There were approximately
16,935,000 outpatient visits to VA staff and 1,681,000 visits to
non-VA physicians under VA auspices. To meet these patient
care demands, the Department of Medicine and Surgery had
over 199,000 full-time equivalent employees and budget obliga-
tions of almost $8.6 billion. In addition, over 82,000 volunteers
provided 12.1 million hours of service.
Under its operating statutes and Executive Order 12088, "Fed-
eral Compliance with Pollution Control Standards," the U.S.
EPA has been directed to enforce pollution control at Federal
facilities, including government hospitals and laboratories. These
facilities specifically include VAMCs and other VA medical facil-
ities. We shall refer to all VA medical facilities in this paper as
VAMCs for the sake of convenience.
Insuring a VAMC's compliance with waste management laws
is an on-going process and requires day-to-day effort at each
VAMC. A manual, therefore, was prepared to support the efforts
of VAMC staff in complying with rules on waste management.
THE NATURE OF VAMC WASTES
General
Waste is produced in achieving an aseptic environment. Num-
erous items that come in contact with patients must be sanitized
or be made germ-free. To achieve this goal, an item must be disin-
fected and sterilized or discarded after patient contact. Recep-
tacles of various types are provided for depositing disposable
items, including covered and uncovered cans, bottles and boxes,
as well as sinks/drains where liquid wastes are routinely depos-
ited.
There are many different wastes generated at a hospital. One
waste is putrescible sewage. Another waste consists of anatomical
parts or tissues removed for analysis or as a result of surgery. A
third waste results from cleaning and repairing the building)
floors, grounds and equipment.
Still another grouping of wastes includes excess supplies and
waste mixtures with valuable constituents. Excess supplies, such
as aged vaccines, become wastes after their useful life has ended.
Used items like silver and gold may be separated out and recov-
ered as valuable products. Items at the hospital become waste
when a person places them in a receptacle destined for destruc-
tion or disposal off-site or, in the case of regulated materials,
opened stocks which remain unused for more than 270 days. Un-
used regulated material may be stored anywhere at the hospital
while the material is not yet designated as waste. Once desig-
nated as waste, however, the material must be dated, container-
ized and stored according to specific standards.
In the discussion that follows, we first will describe medial
wastes and then examine other regulated hazardous wastes.
134 TREATMENT
-------
Medical Wastes
Medical wastes may be hazardous because of the possible con-
tamination by infectious agents. They may originate in patient
rooms, therapeutic services and in analysis units (laboratories,
autopsy rooms and research areas). Historically, all the wastes
from these areas have been grouped together because steriliza-
tion or incineration was *he preferred method of on-site waste
handling.
The U.S. EPA has recommended that hospitals consider the
following categories of waste for handling as infectious': isola-
tion wastes; cultures and stocks of infectious agents and asso-
ciated biologicals; human blood and blood products; pathologi-
cal wastes; contaminated sharps; contaminated animal carcasses,
body parts and bedding; miscellaneous contaminated wastes
(optional category); wastes from surgery and autopsy; contam-
inated laboratory wastes; dialysis unit wastes; and contaminated
equipment.

Other Regulated Hazardous Wastes
Common wastes generated in the hospital include: dusts, fumes
and other solids; vapors, mists and other liquids; and gases. These
wastes may be generated by using commercial products and by
working with biological matter. They become regulated hazard-
ous wastes through contamination with infectious agents, radia-
tion and/or chemical therapies. A description of each waste
follows.
Waste dusts are light, solid particles that escape from contain-
ers of powdered chemicals and drugs, or are produced from
stone, wood, metals, their coatings and other materials during
repairs, renovation, building or therapy. Waste metallic dusts can
be produced by grinding, cutting and polishing metals. Metal
fumes are produced in soldering, welding and metal casting.
These fumes can react with oxygen in the air to form metal oxide
fumes, which cool and condense on objects as very fine solid par-
ticles. Metallic salts heated in hospital laboratory work also pro-
duce metal oxide fumes. Dyes and organic pigments frequently
are present on solids as a coating. Waste dusts generated from
surface friction are composed primarily of these coatings.
A generic list of 22 chemotherapy drugs commonly used as
powders or solutions in hospitals follows:2-3-4-5 azathioprine, bleo-
mycin, busulfan, chlorambucil, cisplatin, cyclophosphamide,
cytarabine, decarbazine, dactinomycin, doxorubicin, 5-fluoro-
uracil, hydroxyurea, lomustin, ccnu, mechlorethamine, mel-
phalan, mercaptopurine, methotrexate, mithramycin, mitoycin,
procarbazine and vincristine.
Solvents are used in hospital facilities to dissolve medication
formulations, oils, resins, waxes, plastics, varnishes and paints.
Toxic solvents include acetone, isopropryl alcohol, ethyl alcohol,
ketones, esters, alcohols, petroleum distillates, hexane, methyl
butyl ketone, methyl cellosolve, nitrobenzene, phenol and most
of the aromatic and chlorinated organic chemicals. Aerosol
sprays may contain solvents such as petroleum distillates, toluene,
chlorinated organic chemicals, ketones, propane, butane and car-
bon dioxide.
Metallic salts, when heated in laboratory work, can produce
vapors which can be absorbed through the skin or inhaled. In
the pharmacy, highly toxic waste mists and vapors from prepara-
tion of chemotherapy drugs may contaminate absorbents, gloves,
masks, glove box and fume hood filters and lab ware. Medical
services work with human tissue and fluids and research services
work with a variety of biological secretions and fluids which may
become contaminated with molds, bacteria, viruses, parasites and
fungi.
Toxic gases are waste products of many hospital workshops,
medical services and laboratories. Some of these waste gases are
flammable. Anesthetic gases like cyclopropane are also flam-
mable and may leak from cylinders. Vapors and fumes act like
gases and present the same containment problems as gases. Irri-
tating waste gases arising from hospital work include ammonia,
ethylene oxide, fluorine, formaldehyde, hydrogen sulfide, ozone
and sulfur dioxide. Ethylene oxide results from degassing of gas-
disinfected objects; ozone comes from arc welding and carbon
arcs; and ammonia, sulfur dioxide, fluorine, hydrogen sulfide
and formaldehyde come from laboratory cleaning, fixing and
analyses.
Medical and research activities may produce waste contam-
inated with low-level radioactive wastes.6'7'8 Contaminated solid
waste includes protective clothing, spill cleanup absorbents, vials
and syringes, human and animal tissues and excreta, and acti-
vated charcoal and other filters from fume hoods and glove
boxes.
Liquid waste contaminated with radioactivity includes regu-
lated hazardous chemicals, solvents used in laboratory analysis
(such as alcohols), aldehydes, ketones, organic acids and toluene
scintillation cocktails. Aqueous solution wastes may be remains
of radiopharmaceutal preparations, rinse solutions from washing
lab ware, and the excreta of patients and animals receiving injec-
tion or ingesting radioisotopes with half-lives longer than 24
hours.
A detailed list of the common low-level radioactive agents in
wastes from medical research and health care institutions is pre-
sented in Table 1.
Table 1
Common Chemical/Radioactive Waste from Medical Research and
Health Care Activities
Element

Hydrogen
Boron
Carbon
Phosphorus
Sulphur
Chlorine
Calcium
Chromium
Cobalt
Iron

Gallium
Arsenic
Selenium
Krypton
Strontium
Rubidium
Strontium
Technetium
Molybdenum
Cadmium
Indium
Tin
Iodine
Xenon
Ytterbium
Ir1d1um
Gold
Thallium
Mercury
Radon
Radium
Regulated
Hazardous
Chemical

Yes
No
No
Yes
Yes
Yes
No
Yes
Yes
No
II
Yes
Yes
Yes
No
No
No
No
No
No
Yes
Yes
No
No
No
No
Yes
No
Yes
Yes
Yes
Yes
Radionuclide

H-3
B-12
C-14
P-32
S-35
Cl-36
Ca-45
Cr-51
Co-57
Co-58
Co-60
Fe-55
Fe-59
Ga-67
As-74
Se-75
Kr-85
Sr-85
Rb-86
Sr-90
Tc-99m
Ho-99
Cd-109
In-Ill
Sm-113
1-123
1-125
1-130
1-131
Xe-133
Yb-169
Ir-192
Au-198
TI-201
Hg-203
Rn-222
RA-226
TREATMENT 135
-------
HOSPITAL WASTE HANDLING
METHODOLOGY

Waste Handling at the Generation Point
A waste classification system was created to help VAMC staff
identify waste items which must be kept separate and deposit each
class of waste in appropriate containers. Four classes were
chosen: biological waste, chemical waste, radioactive waste and
regulated non-hazardous waste. Individual wastes may be a mix-
ture of these classes.
Regulated non-hazardous waste may contain no dangerous
contamination and still be a safety hazard, e.g., discarded need-
les, a cracked carboy of water or a leaking pressurized cylinder
of air. Regulations for handling non-hazardous wastes are based
on the physical characteristics of the waste, while regulations
for handling designated hazardous wastes are based on hazardous
properties and specific constituents.
Biological, chemical and radioactive wastes are subject to reg-
ulations which are both comprehensive and specific to the agents
concerned. Regulated biological and chemical wastes may be
further subdivided into two groups: Trace- or Bulk-Contamina-
tion. There should be an appropriate receptacle in every room to
hold each class of waste produced through normal operations.
To show how this classification system fits with typical VAMC
waste disposal options, a simplified flow chart is presented in
Fig. 1.
HWL
SOLIDS
ID
SAWTARV
LAKDriLL

DISCHARGE
LIQUIDS
TO
SAKITMT
scwr.il

Figure 1
A simplified flow chart showing how the VA waste classification sys-
tem is used to segregate wastes into groups having similar disposal
options.
Regulations on hazardous waste handling allow on-site destruc-
tion of wastes containing "de minimus" quantities of regulated
items in a properly functioning pathological incinerator. Such
contamination must result from normal and complete use of ma-
terials—not an act of waste mixing—and represent less than 3%
by weight of the total quantity of contaminated non-hazardous
waste. All other waste mixtures containing regulated biological,
chemical and radioactive materials should be handled as Bulk-
Generated waste.1
Transport hazardous
wastes as required
by DOT using a 55 gal.
drum having a steel
exterior, removable lid
and polyethylene liner
Polyethylene
liner
Impermeable.
containers
DOT 17H Steel
Drum
Absorbent
and Impact
cushioning
material
DOT 15A Box
outer container
(or acutely
hazardous waste
container inside
Figure 2
An Example of the Lab Pack

Generated throughout every department of a VAMC, the daily
total of individually "de minimus" quantities is appreciable and
requires special handling. For example, the current regulatory
level at which a 55 gal drum and its regulated contents become
subject to hazardous waste handling regulations is when the con-
tents represent about 2 gal or "3% or more by weight to the total
quantity of the material which could be held" by the drum.
Many patient treatment, medical research and testing proced-
ures are conducted using disposable items to catch and retain
regulated materials. These disposable items should be considered
Trace-Contaminated. All discarded items which contain a regu-
lated material should be considered Bulk-Contaminated waits
mixtures subject to regulations for handling designated hazard-
ous waste, unless they can be shown to be Trace-Contaminated
wastes.

Waste Segregation
A waste deposited in a receptable may contaminate the re-
ceptacle and all the items in it, preventing later waste segrega-
tion. This contamination may be avoided by placing the waste in
a container which effectively presents the contaminant from be-
ing released to the surroundings or by placing all wastes having
a similar type of contamination in one lined, covered container.
Radioactive, chemical and infectious wastes must be segre-
gated at the point of generation. Other non-dangerous regulated
and non-regulated wastes may be combined at the point of gen-
eration for later segregation at a central location.
Wastes containing radioactive contamination are to be retained
in a locked, shielded enclosure until they decay to the minimum
background radiation level. Wastes contaminated with radiotc-
tivity should be immediately sealed, labeled and takes to the
shielded enclosure. The containers must be clearly labeled and
tagged to show each radioisotope present, the quantities and types
136 TREATMENT
-------
of material, the date the waste was generated and the name of the
person preparing the label.
At a VAMC, infectious organisms are the principal biological
dangers. All discarded items known or suspected to contain an
infectious organism should be handled as biohazards. Biohaz-
ard waste contaminated with radioactivity should be sealed in an
impervious bag and marked with a biohazard label. The person
discarding a mixture which contains chemicals in addition to in-
fectious agents should destroy the infectious contamination with-
out causing the release of dangerous chemicals. Wastes which
have only Trace-Contamination with regulated chemicals are
allowed to be destroyed by burning in a properly functioning
pathological incinerator.
Receptacles holding material which is only contaminated by
infectious organisms must be clearly marked with the universal
biohazard warning symbol. This material can be sterilized and
disposed of as regular paper trash or reused. Discarded items
can be taken to the pathological incinerator for destruction if
they are free of glass and plastics. Biohazard waste containing
trace-contaminating chemicals may also be discarded in biohaz-
ard receptacles if the waste is destined for disposal on-site in a
properly functioning pathological incinerator.
Exceptions to these procedures may range from the disinfec-
tion or sterilization of contaminated materials to the incinera-
tion of animal carcasses.9'10-11 Discarded mixtures which also con-
tain chemicals must be so marked. Infectious agents in or on
Bulk-Contaminated chemical waste can be destroyed by the use of
disinfectants and, on occasion, by sterilizing the mixture. Follow-
ing an infection destruction procedure, the remaining Bulk-Con-
taminated chemical waste must be sealed in an impervious con-
tainer and labeled. Specific packaging and labeling procedures
are described below.
If the nature of the waste mixture prevents destruction of the
infectious agents, the waste must be sealed in an impervious con-
tainer and labeled. For chemicals present in bulk, it is necessary to
specify their volume or weight. The proper disposal of spent, used
or aged chemicals, discarded or aged products containing chemi-
cals as their active ingredients or the by-products of using these
chemicals or products (secondary chemicals, contaminated paper
waste, contaminated containers, etc.) is critical. Segregation into
specific receptacles for each type of chemical waste as it is gener-
ated is the first step. Determining what chemicals and chemical
constituents are regulated requires: (1) keeping up with updates of
lists of designated chemicals and (2) conducting an evaluation of
every chemical and product with chemical constituents discarded
in the VAMC.
When the person who generates a waste with chemical constit-
uents is unsure of the regulatory situation, the waste should be
handled like other regulated wastes—prepared for shipment off-
site to a licensed disposal facility. The waste must be sealed in an
impervious container and a label applied. Where possible, very
dangerous substances may be converted into less dangerous ones
in the laboratory rather than being placed directly into containers.
Deactivated chemical waste, if regulated substances are absent,
then may be disposed of by pouring them into a chemical sink or
placing them in the trash with other non-dangerous solids. This
disposal method should be followed with extreme caution. For
example, only water-soluble substances which are known to be
bio-degradable in the waste treatment process of the local waste-
water treatment plant should be disposed of in the laboratory
sink. Strong acids and bases should be diluted to a pH range of
3 to 11 before they are poured into the sink at a rate exceeding
the equivalent of 50mL of concentrated substance per minute.14
Liquid and solid chemical wastes should be placed in containers
provided for that purpose. When breakable containers are used,
they should be placed in carriers. The daily accumulation of com-
patible waste mixtures may be temporarily stored in safety cab-
inets. When these waste containers are full, they can be collected
for further processing.
Proper packaging is important. Trace-Contaminated wastes
may be placed in covered receptacles lined to protect the recep-
tacles from contamination without any additional packaging if
these wastes are to be destroyed in a properly functioning path-
ological incinerator. Otherwise, these wastes must be packaged
for handling, using procedures specific to the type of contamina-
tion present. Bulk-Contaminated wastes always must be pack-
aged to contain the individual waste mixture. Since radioactivity
can pass easily through the walls of containers, radioactive con-
tainment procedures should be used first. Since chemicals can
form toxic vapors or eat through containers, procedures for de-
activating or containing corrosive and gas-producing chemicals
should be used for all wastes that are not radioactive. Since bio-
logical wastes can be adequately contained by a sealed leak-proof
bag, biological waste handling procedures may be used for all
other items; however, biological waste which is known to be con-
taminated with infectious agents must be treated to nullify the in-
fection hazard or be sealed in an impervious container.
Dangerous solid, liquid and gaseous wastes not otherwise con-
taminated by biological, chemical or radioactive material should
be placed in sturdy containers such as fiber boxes for solids like
broken glass; impervious and rigid plastic bottles for liquids like
acid and base solutions after neutralizing pH adjustment; and
vented metal safety cans or cabinets for discarded gas cylinders
and aerosol cans. Other solid, liquid and gaseous wastes not fall-
ing into one of the dangerous contamination classes can be placed
in open receptacles.'3
All waste material leaving a room should be sealed and marked
with labels warning of any dangerous contamination. Waste re-
ceptacles should be clearly marked with text and symbols showing
the classes of waste and the level (Trace or Bulk) of contamina-
tion, whether or not each waste mixture in the receptacle has its
own specific label. Every package of Bulk-Contaminated waste
should have attached to it a label identifying each contaminant
present, quantity and types of every material in the container,
date the waste is generated and name of the person preparing the
label.
Wastes may be transported on-site in a variety of ways. They
may be deposited from room waste receptacles into gravity chutes
or be placed in a cart for horizontal transfer to a chute, waste
marshalling area, storage area, on-site disposal area or dump-
ster for off-site transportation to a sanitary landfill. Whatever
system is employed, it should be reviewed relative to the type and
potential hazard of waste being transported.

On-Site Waste Storage
Containment, Labeling and Segregation of Waste
A waste marshalling and storage area is used to prepare wastes
for on-site disposal and temporary storage on-site prior to trans-
port off-site for treatment and disposal. Wastes for off-site treat-
ment and disposal include non-regulated wastes, regulated non-
hazardous wastes and Bulk-Contaminated wastes of regulated
materials.
Radioactive material should be held in a limited-access shielded
enclosure for decay until background levels are reached. Bar-
ring any contamination with regulated chemicals or infectious
materials, these wastes then can be placed in on-site waste treat-
ment systems or sent to the nearest sanitary landfill. Sealed pack-
ages of waste mixtures containing infectious organisms may be
segregated as biohazard wastes if radioactivity no longer is pres-
ent.
TREATMENT
137
-------
The designated hazard classes established by the DOT are the
established basis for the segregation of regulated chemical wastes
being prepared for off-site transport. In a well ventilated mar-
shalling area, these wastes can be segregated in covered, open-
headed, 55-gal steel drums; one drum is to be used for each des-
ignated hazard class. The designated classes also help divide the
storage area into separate units, because they distinguish between
materials with incompatible physical and chemical properties.
Ignitable wastes with a flash point of less than 140T and reac-
tive wastes that can explode or release toxic gases, vapors or
fumes at standard pressure and temperature are especially danger-
ous and are banned from "lab packs" being taken to a land dis-
posal site except as trace contamination. They must be sent to a
treatment facility prior to disposal on land. For reactive wastes,
only Bulk-Contaminated wastes of cyanide and sulfide-bearing
chemicals are allowed to be included in "lab packs".16

Shipping Wastes
The shipping container for regulated dangerous wastes is a steel
drum filled with an inert absorbent material, such as vermiculite
or Fuller's earth. The drum must be packed tightly with the absor-
bent to cushion blows and absorb any liquids that may leak from
inner containers. This "lab pack" generally is used for landfill
disposal. An example of the "lab pack" is presented in Fig. 2.
The transporter's vehicle should be checked for any obvious
contamination and for visually apparent defects. It should be de-
signed for easy loading without spilling, and any compartments
should be completely enclosed and placarded. The vehicle should
be watertight.
All drums or other containers should be re-inspected prior to
loading to assure that they are not damaged. A final check should
be made for appropriate labeling of waste containers. Placards
appropriate for a hauler of regulated material wastes should be
affixed to the exterior of the vehicle. Vehicle drivers should be
informed of the hazards of the waste, and appropriate emergency
response information for the waste should be given to the driver.
Drums that contain liquids should be inspected to assure the
absence of seepage at seams or bungs. The contents of unaccep-
table drums should be transferred, or the drum should be placed
inside a DOT approved recovery drum and packed with absor-
bent.

Design of a Storage Facility for Regulated Waste
Storage and waste marshalling areas should be accessible only
to authorized waste handlers. These areas must be enclosed,
walled where animals and insects must be excluded or just roofed
and fenced where temperature and moisture will not present a
problem. Radioactive materials should be held in a limited-access,
shielded enclosure to give time for radioactivity decay. Where
volatile chemical wastes are kept, walled enclosures should be
ventilated." Areas for storage of potentially infectious patho-
logic wastes and animal carcasses from the medical and research
services should be temperature controlled.''
The storage area should be roofed in a manner that protects
wastes from heat, container corrosion, internal chemical reac-
tions and deterioration of labels, tags and other markings caused
by storm precipitation. Segregation of incompatible materials
and posting of warning signs is essential. It is particularly impor-
tant to segregate igni tables from oxidizers.
Floors of all storage areas should be smooth, dry, impervious
to the waste being stored there and sloped to a floor drain or col-
lecting basin. The collecting basin should be large enough to re-
tain 10V» of the total volume of liquid wastes stored in the areas
draining to it.
Outer metal containers used to store wastes should be raised on
racks to prevent hidden corrosion and to allow waste handling in-
spectors to detect bottom leaks. The inspection should note the
quality of housekeeping in the storage area as well as the integ-
rity of all emergency control and safety equipment.
RCRA requires that a chemical/biological analysis or a com-
plete contents list be attached to every drum of designated haz-
ardous waste. A copy of the daily record of all items packaged,
labeled and placed in the waste storage area should be kept at a
location away from the storage area.
The area should have hot and cold water connections and be
equipped with a hose, fire and smoke alarms and other appro-
priate structural and equipment components. If large volumes of
waste scintillation fluid are stored they should be kept under an
automatic sprinkler system.
There should be an emergency response plan for situations
such as a leaking drum, a flash fire or a splash of a dangerous
material on a person. Personnel must be trained to deal with the
most probable dangerous situations: plugging pinhole leaks in
large containers; absorbing dangerous liquids; and removing in-
compatible items, such as pressurized containers, from a mixture
of wastes.

On-Site Treatment and Disposal
"De minimus" wastes or those of Trace-Contamination can
be treated on-site without special permits or approvals. This in-
cludes destruction of materials in a properly functioning path-
ological waste incinerator. Waste mixtures containing regulated
biological, chemical and/or radioactive materials which do not fit
within the "de minimus" or Trace-Contamination definitions an
Bulk-Contaminated wastes which must be handled according to
the regulations for hazardous wastes.
Infectious organisms in biological material must be destroyed
before they leave the VAMC. On-site incineration or sterilization
is recommended. Other methods available to nullify infectious
agents include: chemical disinfection using ethylene oxide or for-
maldehyde gas/vapor chambers; direct contact with solutions
containing chlorine, iodine, phenols, acidified alcohols, ozone
and various quaternary ammonium salts; and exposure to ioniz-
ing radiation.
Treatment products can include autoclaved tissue culture
fluids, ash and residue from a pathological waste incinerator and
glassware from a chemical disinfection process using a non-haz-
ardous chemical such as iodine. Following nullification of the
dangerous biological contaminants in a waste which is free of
other regulated hazardous waste or contamination from the treat-
ment process, liquids usually can be discharged to the sewer sjn-
tern and solids can be sent to a sanitary landfill.
Regulated chemical waste can only be disposed of with the
prior approval of government waste management authorities.
Dangerous chemical waste may be destroyed or deactivated to
eliminate its dangerous properties. This treatment must be con-
ducted prior to segregation and packaging the wastes for collec-
tion. The National Research Council recently developed a text on
handling waste chemicals which presents the most recent pro-
cedures for deactivation of the most common laboratory chem-
icals." Deactivated chemical waste, if other regulated substance*
are absent, may then be disposed of by pouring it into a chem-
ical sink or placing it in the trash with other non-dangeroui
solids.
When unsure of the safety or regulatory permissibility of pour-
ing the waste into a chemical sink or putting it with the unregu-
lated solids, the waste should be handled like other regulated
waste prepared for shipment off-site to a licensed chemical dis-
posal facility. Likely candidates for such action include Bulk-
Contaminated wastes with one or more chemicals; highly concen-
138 TREATMENT
-------
trated regulated hazardous chemicals; glass, plastic and metal
containers holding hazardous waste in Bulk-Contamination
quantity and discarded chemicals that cannot be treated on-site.
If biological waste materials contain a chemical contaminant,
the selection of an on-site treatment method becomes compli-
cated. For example, autoclaving biological waste contaminated
with a highly volatile chemical could release the chemical and
expose the staff. If a volatile metal, such as mercury or lead,
or an organic chemical that is difficult to destroy by combustion
is present, incineration may emit a toxic air pollutant. Similarly,
if chemical disinfection is used, an assessment must be made of
the interaction between the disinfectants and all chemical species
in the waste to determine if regulated hazardous by-products
will be produced.
Hospital wastes which the U.S. EPA recommends be inciner-
ated in a pathological incinerator include the following: all waste
from patients isolated for severe and communicable diseases;
all discarded living tissue and fluid from patients, laboratory
animals and their corpses; discarded blood, serum and plasma;
all wastes from surgery, autopsy and treatments such as obstetri-
cal and dialysis care; all wastes from hospital laboratories that
have come in contact with the wastes identified above; discarded
vaccines, drugs and other supplies and equipment suspected of
contamination with infectious organisms; autoclaved wastes in-
itially contaminated with biohazardous material and infectious
organisms; and autoclaved sharps which have come in contact
with patients, laboratory animals, their corpses or any other tissue
or fluids.
VAMCs are required to have pathological incinerators capable
of maintaining a continuous minimum temperature of 871 °C.
Temperatures of 600-900 °C commonly are capable of destroying
most dangerous chemicals in Trace-Contamination quantities."
However, Bulk-Contamination with dangerous chemicals might
require residence times of 2 sec at 1200 °C and, therefore, may
require off-site incineration.
Typical steam sterilizers operate at 132°C with saturated steam
at 35 lb/inz pressure within the chamber. Specially designed auto-
clavable bags or steam-porous paper bags are used to contain the
infectious waste collected in biohazard bags. Once loaded, the
sterilizer is locked and programmed for the desired time and
temperature regime.
Dry Heat Sterilization often is used to sterilize infectious wastes
or contaminated equipment. Ethylene oxide gas usually is used
to disinfect supplies. Formaldehyde gas is more commonly used
to sterilize contaminated equipment, ventilation systems and
rooms. Care must be exercised in the conduct of this process be-
cause there is the potential for worker exposure to these carcino-
genic gases because of "degassing" from the waste.
CONCLUSION
The Veterans Administration has a unique hospital waste dis-
posal problem because of the large number of Veterans Admin-
istration Medical Centers (VAMCs) in different locations across
the country as well as the wide variety of activities conducted at
these VAMCs. For that reason, a waste management plan was
developed to help VAMCs comply with waste handling and dis-
posal requirements established under RCRA and its 1984 amend-
ments. In this paper, we have presented summaries of the first
two parts:
I-Hospital Waste Handling Regulations/Characteristics
II-Hospital Waste Management Methodology

ACKNOWLEDGEMENT
We are grateful to the Veterans Administration for its support
of the work described in this paper and for its permission to sum-
marize the results of that work, although the views expressed are
those of the authors and not necessarily those of the Agency.

REFERENCES
1. U.S. EPA, "EPA Guide for Infectious Waste Management," Office
of Solid Waste, Washington, DC, 1986, 530-SW-86-014.
2. Moody, D., "Veterans Administration Medical Center Policies and
Procedures for Handling Infectable Antineoplastic Drugs," Am. J.
Pharm., 41, 1984,916-919.
3. Vaccari, P., Tonat, K., DeChristoforo, R., Galleli, J. and Zim-
merman, P., "Disposal of Antineoplastic Wastes at the National In-
stitutes of Health.'Mm. /. Hasp. Parhm., 41, 1984, 87-93.
4. LeRoy, M., Roberts, M. and Theisen, J., "Procedures for Hand-
ling Antineoplastic Injections in Comprehensive Cancer Centers,"
Am. J. Hasp. Pharm., 40, 1983, 601-3.
5. Stolar, M., Power, L. and Viele, C., "Recommendations for Hand-
ling Cytotoxic Drugs in Hospitals," Am. J. Hasp. Pharm., 40,
1983,1163-71.
6. Bacovsky, R., "Disposal of Hazardous Pharmaceuticals," Can. J.
Hasp. Pharm., 34, 1981, 12-13.
7. EG&G Idaho, Inc., "Improved Low-Level Radioactive Waste Man-
agement Practices for Hospitals and Research Institutions," DOE/
LLW-21T, 1983, U.S. Department of Energy.
8. Anon., "Most Radioactive Wastes Can Be Disposed of On-site,"
Hospitals, May 1984, 66.
9. Center for Disease Control, Hospital Infections Program, "Disposal
of Solid Wastes From Hospitals," second revision, 00-2523, Center
for Infectious Diseases, CDC, Atlanta, GA 30333,1980.
10. Public Health Service, Centers for Disease Control & National Insti-
tutes of Health, "Biosafety in Microbiological and Biomedical Lab-
oratories," HHS Publication No. (CDC) 84-8395, 1984.

11. U.S. EPA, "Draft Manual for Infectious Waste Management,"
Washington, DC Office of Solid Waste, U.S. EPA, Washington,
DC, 1982, SW-957.
12. International Agency for Research on Cancer, "Handling Chemical
Carcinogens in the Laboratory, Problems of Safety," IARC Scien-
tific Publications No. 33,1979, Lyon, France.
13. University of Wisconsin, Univ. Safety Dept., Hazardous Waste
Program, "Laboratory Waste Disposal and Safety Guide," Univer-
sity of Wisconsin, Madison, WI, 1984.
14. Rose, S.L., Clinical Laboratory Safety, J.B. Lippincott Co., Phil-
adelphia, PA, 1984.
15. Bond, R., Michaelsen, G., DeRoos, R., et al., "Environmental
Health and Safety in Health Care Facilities," New York: Macmillan
Publishing Co., 1973.
16. National Research Council, Committee on Hazardous Substances in
the Laboratory, "Prudent Practices for Disposal of Chemicals from
Laboratories," 1983.
TREATMENT
139
-------
Fixation of Metallic Ions in Portland Cement

Muhammad S.Y. Bhatty, Ph.D.
Construction Technology Laboratories, Inc.
Skokie, Illinois
ABSTRACT
This paper reports the preliminary results of the fixation of
metallic ions in hydrating tricalcium silicate. Mixtures of
tricalcium silicate and solutions of known concentration of
metallic salts such as cadmium chloride, nitrate and sulfate;
chromium chloride, nitrate and sulfate; lead nitrate and sulfate;
mercury nitrate and sulfate; and zinc chloride and sulfate were
reacted for 1, 7, 28 and 90 days.
After the desired reaction period, the mixtures were filtered and
the solids (precipitate) were studied by X-ray diffraction and dif-
ferential thermal analysis techniques. The filtrates were analyzed
for cations and anions. The results indicate that metallic ions are
retained (fixed) in the solid. Possible mechanisms of fixation of
metallic ions are discussed, including: (1) fixation by addition
reaction, (2) fixation by substitution reaction, (3) fixation by for-
mation of new compounds and (4) fixation by a combination of
the above mechanisms.

INTRODUCTION
Under the Land Disposal Restrictions Program under RCRA
Subtitle C, wastes to be land disposed must be treated to meet
U.S. EPA standards. Fixation is treating industrial wastes to pre-
vent hazardous constituents from polluting the environment.
Portland cement or cement-based materials can be used to fixate
certain inorganic wastes, especially those containing heavy
metals.1
Tashiro, et al. ,2-* investigated the effect of heavy oxides of Cr,
Cu, Zn, As, Cd, Hg and Pb on the physical properties of cement.
They found that early hardening and strength development are
generally adversely affected due to the interaction of metals with
cement paste.
Stepanovo5 reported that metal chlorides of Mn, Co, Ni, Cu
and Zn interact with silicates and aluminate of cement to form
complexes that can influence the strength development.
Recently, Poon, et al. ,*•' have published a number of papers on
the mechanism of metal fixation in cement-based materials. The
effects of Zn, Hg and Cd were studied using X-ray diffraction,
electron microscopy and permeability techniques. The cement-
based matrices used by these researchers consisted of commercial-
ly available processes such as Chemfix and Sealosafe.
Other studies on fixation have been reported by Glasser, et
a/.,10, and Malone, et a/."
Despite a large number of studies, there still appears to be a
need for a better understanding of the interaction of metallic
compounds with cement. Thus, the purpose of the present study
is to understand the mechanisms of metal fixation in portland ce-
ment using pure tricalcium silicate.
About 70-80^0 of a portland cement is composed of calcium
silicates (tricalcium silicate, C3S; and dicalcium silicate, C2S),
which makes them fundamentally important to the investigation
of the chemical reactions of portland cement. Because the
behavior of cement is similar to the behavior of calcium silicates
to a large extent, tricalcium silicate was used in the present
studies. It is believed that the use of pure C3S will simulate the
complex reaction chemistry of cements without complications
arising from other components of portland cement.

EXPERIMENTAL
Metallic salt solutions of known concentrations used in this
study were cadmium chloride, cadmium nitrate and cadmium
sulfate; chromium chloride, chromium nitrate and chromium sul-
fate; lead nitrate and lead sulfate; mercury nitrate and mercury
sulfate; and zinc chloride and zinc sulfate. All metallic salts were
reagent grade chemicals. Solutions were made using deionized
water.
Mixtures containing about Sg C3S and 25g metallic salt solution
of known concentration were reacted in polypropylene bottles
with continuous agitation at 23 °C. After the desired reaction
time, the mixtures were filtered, and the filtrates were analyzed
for the appropriate cations and anions. The difference between
the original concentration of cations and anions and the concen-
tration of cations and anions in the filtrate gave the amount of ca-
tions and anions retained by the solid (precipitate). Solids were
studied by diffraction and differential thermal analysis tech-
niques.

RESULTS AND DISCUSSION
Table 1 shows the results of the experiments showing the hydra-
tion of C3S in the presence of water and solutions of various
metallic salts. The hydration or reaction of C3S with water or
metallic salt solutions was studied for 1, 7, 28 and 90 days. The
data in each column in Table 1 are explained below.
The metal concentration as a percent of the C3S is given in Col-
umn 3. Column 4 shows the reaction time. Mixtures were hy-
drated for a desired time as shown in Column 4 and filtered. The
concentration of metallic ions in the starting mixture is given in
Column 5. Column 6 shows the concentration of metallic ions In
the solid (precipitate), which was obtained by subtracting the con-
centration of metallic ions found in the filtrate from the concen-
tration in the starting mixture. The concentrations of anions such
as chloride, nitrate and sulfate in the solid were obtained in the
same way for metallic ions (Column 9). However, the role of
anions will not be discussed in this paper. Column 11 shows the
concentration of calcium in the filtrate.

Fixation of Metallic Ions in Hydrating CjS
C3S is the major cementitious phase of most portland cements.
The majority of C3S hydration takes place in about 28 days and
140
TREATMENT
-------
Table 1
Amount of Metal Retained in Solid Obtained by the Reaction of CjS and Metal Solutions at Various Reaction Times
1
Mixture
No.

2
3
4

Mixture
No.
6

7
8
9

2
Mixture
Composition

5g. C S+
25g. H20

Mixture
Composition
5g. C-S+
25g. CdCl2

Cadmium
pur lOOg
of C3S
E.
6.3

6.3
6.3
6.3

4
Reaction
Time

7
28
90

Cadmium Concentration,
Reaction
Time
1

7
28
90

Starting
Mixture
12500

12500
12500
12500

me/1
Solid Filtrate

12500

12500
12500
12500

Cadmium Concentration,

11
12
13
14

5g. C-S+
25g. Cd(NO,),
o &
„
„

7.8
7.7
7.8
7.8

1
7
28
90

Starting
Mixture
15300
15300
15300
15300

<0.1

<0.1
<0.1
<0.2

mg/1
Solid Filtrate

15300
15300
15300
15300

Cadmium Concentration,

17
18
19

Mixture
No.
21

22
23
24

27
28
29

5g. C-S+
25g. CdS04

„

Mixture
Composition
5g. C S+
25g. CrCl3

5g. C S+
25g. Cr(NO )

13.7

13.7
13.8
13.7
Chromium
per lOOg
Of CgS
g.
2.2

2.1
2.1
2.2

2.4

2.4
2.4
2.5

7
28
90

Reaction
Time
1

7
28
90

7
28
90
Starting
Mixture
26500

26500
26500
26500

Chromium
Starting
Mixture
4300

4300
4300
4300

Chromium
Starting
Mixture
4800

4800
4800
4800

<0.1
<0.1
<0.1
<0.2

me/1
Solid Filtrate

26500

26500
26500
26500

Concentration

<0.1

<0.l'
<0.1
<0.2

, mR/1
Solid Filtrate

4300

4300
4300
4300

Concentration

<0.3

<0.3
<0.3
<0.3

mg/1
Solid Filtrate

4800

4800
4800
4800

<0.3

<0.3
<0.3
<0.3
8

Chloride
Starting
Mixture
7900

7900
7900
7900

Nitrate
Starting
Mixture
16800
16800
16800
16800

Sulfate
Starting
Mixture
22600

22600
22600
22600

Chloride
Starting
Mixture
8700

8700
8700
8700

Nitrate
Starting
Mixture
17400

17400
17400
17400
9

Concentration, mg/1
Solid

300

-300
0
450

Concentrs
Solid

4800
4800
4700
2980

Filtrate

7600

8200
7900
7450

ition, mE/1
Filtrate

12000
12000
12100
13820

Concentration^ mg/1
Solid

21400

21400
21300
21010

Filtrate

1200

1200
1300
1590

Concentration, me/1
Solid

2800

2400
3300
4030

Concentre
Solid

-400

900
3300
6210
Filtrate

5900

6300
5400
4670

itlon, mE/1
Filtrate

17800

16500
14100
11190
11
Calcium
Concentra-
tion in
Filtrate
900~

900
1200
520
Calcium
Concentra-
tion in
Filtrate
5650

5400
5650
2260
Calcium
Concentra-
tion in
Filtrate
3400
3200
3250
1730
Calcium
Concentra-
tion in
Filtrate
1400

1050
1100
720
Calcium
Concentra-
tion in
Filtrate
4400

4200
3400
1800
Calcium
Concentra-
tion in
Filtrate
3400

3300
3000
1370
TREATMENT
141
-------
Table 1 (continued)
1 2

31 5g. C,S»
25g. Cr (SO )
32 ...
33 ...
34 ...

Mixture Mixture
No. Composition
36 5g. C,S+
25g. Pb(NO,),
3 i
37 ...
38 ...
39 ...

1.5
3
1.6
1.5
1.6
Lead
per lOOg
of C3S
g.
11.3
1
11.5
11.5
11.5

7
28
90

Chromium
Starting
Mixture
3100

3100
3100
3100

Concentration, mg/1
Solid

3100

3100
3100
3100

Filtrate

<0.3

<0.3'
<0.3
<0.3

Lead Concentration, mg/1
Reaction
Time
1

7
28
90

Starting
Mixture
22300

22300
22300
22300

Solid

20950

22230
22240
22260

Filtrate

1350

70
60
40

Lead Concentration, mg/1

46 5g. C-S«
25g. PbS04
47 ...
48 ...
49 ...

Mixture Mixture
No. Composition
56 5g. C S*
25g. Hg (NO )
57 ...
58 ...
59 ...

0.1

0.1
0.1
0.1
Mercury
per lOOg
of C3S

12.7

12.6
12.6
12.6

7
28
90

Starting
Mixture
200

200
200
200

Solid

200

200
200
200

Filtrate

<0.3

<0.3
<0.3
<0.3

Mercury Concentration, mg/1
Reaction
Time
1

7
28
90

Starting
Mixture
24500

24500
24500
24500

Solid

24500

24500
24500
23230

Filtrate

<0.5

<0.5'
<0.5
1270

Mercury Concentration, mg/1

51 5g. C,S+
25g. ngSO
52 ...
53 ...
54 ...

Mixture Mixture
No. Composition
61 5g. C,S»
25g. 2nCl2
62 ...
63 ...
64 ...

0.8

0.8
0.8
0.8
Zinc
per lOOg
of C S
g.
4.0

4.0
4.0
3.9

7
28
90

Reaction
Time
1

7
28
90
Starting
Mixture
1700

1700
1700
1700

Zinc
Starting
Mixture
7900

7900
7900
7900
Solid

1700

1700
1700
1600

Filtrate

<0.5

<0.5
-------
Table 1 (continued)
5
9
10
11
41

44
5g. C S+
25g. ZnSO
9.2

9.1

9.2
7

90
Zinc Concentration, mg/1
Starting Solid Filtrate
Mixture
17600
17600
Calcium
Sulfate Concentration, mg/1 Concentra-
Starting Solid Filtrate tlon in
Mixture Filtrate
0.4
17600 17600 0.3:

17600 17600 1.0

17600 17600 2.1
25900
24600
1300
25900 24620 12BO

25900 24640 1260

25900 24310 1590
.600

600

700

390
may be complete in about 1 yr.12 The hydration reaction usually is
represented by the following chemical equation:

2Ca3SiO5 + 6H2O — Ca3Si2O7 • 3H2O + 3Ca(OH)2 (1)

or in cement chemist's nomenclature:
2C3S + 6H —• C3S2H3 + 3CH.
(2)
Thus, the products formed are calcium silicate hydrate (CSH)
and Ca(OH)2. Calcium silicate hydrate constitutes approximately
70% of the total amount of cement hydrated phases. It is believed
that CSH plays a major role in the fixation of metallic ions.
It has been shown that a metallic ion such as sodium can be re-
tained in the structure of calcium silicate hydrates.13- 14 The
amount of metallic ions retained depends upon the calcium oxide
to silica (C/S) mole ratio of calcium silicate hydrate. For example,
more sodium can be retained in a low C/S mole ratio calcium
silicate hydrate than a high C/S mole ratio calcium silicate
hydrate.
It also has been shown that metallic ions such as lithium,
sodium, potassium, cesium and rubidium can be retained in the
hydrating C3S.13- 15 However, the effect of C/S mole ratio on
retention of these metallic ions, other than sodium, has not been
investigated.

Calcium in Filtrate
As found in earlier studies,16 calcium was present in the filtrate
during the hydration of C^S with water at 1, 7, 28 and 90 days,
and the concentration of calcium ions generally increased up to 28
days but then decreased at 90 days (Table 1). Menetrier, et a/.,17
reported that the solution reached the saturation level between 6
and 7 hr with respect to calcium ions during the hydration of C3S.
During hydration of C3S in the presence of salt solutions,
calcium concentration in the filtrate generally decreased as the
hydration time increased (Table 1). The magnitude of decrease
was more pronounced between a reaction time of 28 days and 90
days. In some instances, the concentration decreased significantly
at 90 days compared to 1, 7 and 28 days.
The concentration of calcium ions in the filtrate was affected
by hydration trine as well as by the type of salt solution used. The
calcium ion concentration in the filtrate was highest when metallic
chlorides were used and somewhat lower when metallic nitrates
and metallic sulfates were used. For metallic sulfates, the decrease
can be attributed to decreased solubility of sulfate salts compared
to chloride or nitrate salts which are soluble in water.

Calcium Hydroxide in Solids
Calcium hydroxide was determined using differential thermal
analysis. Table 2 shows the amount of calcium hydroxide pro-
duced at 90 days from hydration of C3S in the presence of water
(used as control) and metallic salt solutions. The larger amount of
Ca(OH)2 formed during the hydration of C3S, the greater the ex-
tent of hydration. All metallic salts except CdSO4 and Cr(NO3)3
resulted in the retardation of C3S hydration.

Table 2
Amount of Ca(OH)2 Obtained from the Hydration of C3S
at 90 Days in the Presence of Water and Salt Solutions
Salt Used

H.O (control)
CdCl,
4

44
CrCl3
Cr(N03)3
Cr2(S04)3
Pb(H03)2
HOSO
34.3

33.7

29.0

30.0

19.

39.

29.

28.

34.

27.

33.6

0.0
C3S did not hydrate in the presence of zinc salts, consequently
no Ca(OH)2 was obtained. These results are consistent with the
findings of Poon, et a/.,8 and others.
The amount of Ca(OH)2 formed also depended on the type of
anion (chloride, nitrate or sulfate) associated with the metal. For
example, in the presence of CrCl3, only 19.3% Ca(OH)2 was pro-
duced; for Cr(NO3)3 and Cr2(SO4)3, the amounts of Ca(OH)2
were 39.6% and 29.8%, respectively.

X-Ray Diffraction Analysis of Solids
Fig. 1 shows the X-ray diffraction (XRD) results for the solids
obtained after 90 days hydration. These results have confirmed
the findings of the differential thermal analysis with respect to
Ca(OH)2. All the mixtures except those containing zinc com-
pounds produced a peak as a result of the presence of calcium
silicate hydrate around 29° 20. In most cases, this peak was
broader than the peak obtained from the mixture containing
water, which was used as a control.
Compared to the C3S-H2O system, some additional lines were
obtained for some mixtures. These lines appear to belong to some
complex compounds which have not been identified in this study.
They, however, do not belong to the salt used in the individual
mixture.
In mixtures containing zinc salts, virtually no hydration of C3S
took place. Additional XRD lines found in C3S-Zn compound
systems did not correspond to the compound Ca[Zn(OH)3.
Mechanisms of Fixation of Metallic Ions
As indicated earlier, different metallic salt solutions were used in
this study. The metallic ions were cadmium, chromium, lead, mer-
cury and zinc. Test results show that in almost all cases all metallic
TREATMENT
143
-------
Woier

COCI2
<-»*

CdS04

CrClj

CrINOjIj

Cr2(S04)3

Pb(NOj>2
PbSO,

Hg{NOj)2
HgSO,
ZnClj

ZnSO,

\
\
I
ll

f
I

f
1

i
it
ii

i
ff
f?
il i
i ilrl

i
A

I ,
1,
, 1 III

4
! |
f? 1

fit

1
f
V

i|
i

i
,,,ii

c
1
i

1
1
t

11 II
A

,1
Ii
M ill,,
1 II 1
Hi

1 1 1
1
1 III
1 Itlllll
0 10 20 30 40 50 60
28
Key
X - CjS
O = CSH
a = ColOHlj
Figure 1
Phases Found in the X-Ray Diffraction Patterns of Solids
Obtained from the Hydration of C3S at 90 Days in
the Presence of Water and Metallic Salt Solutions
ions present in the mixture were retained in the solid irrespective of
time of hydration.
It appears from these results that the following types of
mechanisms probably are involved in the fixation of metallic ions in
the solids:

• Fixation by addition reaction
• Fixation by substitution reaction
• Fixation by formation of a new compound or compounds
• Fixation by combination of the above mechanisms

Fixation by Addition Reaction
In this mechanism, metallic ion is taken up in the structure of
calcium silicate hydrate by an addition reaction as follows:
CSH + M
Calcium Silicate Metallic
Hydrate Ion
MCSH
Metallic Calcium

Silicate Hydrate
(3)
The amount of metal retained probably depends on the type of
calcium silicate hydrate formed during hydration. Calcium silicate

144 TREATMENT
hydrates containing low C/S mole ratio are more likely to favor fix-
ation of metallic ions by this mechanism.18 This is probably one of
the mechanisms by which metallic ions such as Cd, Cr and Pb are
being retained in the portland cement.
It has been previously postulated that low C/S mole ratio CSH
can be produced by a partial replacement of cement with a poz-
zolan such as ASTM Class F fly ash.18

Fixation by Substitution Reaction
This mechanism involves substitution of metallic ion for caltium
in the structure of calcium silicate hydrate as follows:

CSH + M —, MSCH + Ca++ (4)
Calcium Silicate Metallic Metallic Calcium Calcium
Hydrate Ion Silicate Hydrate Ions

Replacement of calcium ion from the structure of calcium
silicate hydrate generally is found to occur in high C/S mole ratio
calcium silicate hydrates.18 It is not possible to predict from the
present data the extent of fixation of metallic ion in calcium
silicate hydrate by this mechanism. Further studies are needed to
answer this question using calcium silicate hydrate of high and
low C/S mole ratios and different concentration of metallic salt
solutions.
Also, since data indicate (Table 1) that the concentration of
calcium ion in the filtrate of an individual mixture appears to be
affected by the type of metallic salt present, it may be assumed
that as the calcium concentration in the filtrate increases, more
metallic ions may be substituted in calcium silicate hydrate. Tim
assumption, of course, does not take into account the changing
equilibrium of the system under consideration.
Fixation of metallic ions by this mechanism in hydrated cement
probably is limited due to the number of calcium ions that can be
replaced from the structure of CSH.

Fixation by Formation of New Compounds
In X-ray diffraction patterns of some solids, new lines were
found other than those that can be assigned to C3S or its hydra-
tion products or metallic salts used. For example, the solid ob-
tained from the mixture containing ZnSO4 contains new lines.
These lines could not be assigned to a known compound.
It is believed that new compounds containing metallic ions may
have been formed which are insoluble in a particular system such
as C3S-ZnSO4. If this is the case, then the knowledge of the com-
position of such compounds, whether simple or complex, may
lead to deliberate formation of such compounds for fixation of
certain metallic ions.

Fixation by Multiple Mechanisms
In a complex system such as cement and metallic waste, the fh-
ation of metallic ions in cement paste may take place in several
ways as described earlier (i.e., multiple mechanisms may be
operating at any one time). In view of the complexity of the
systems under consideration, it may be possible to separate the ef-
fects of different mechanisms with additional work. Work in ihu
area is planned.

CONCLUSIONS
Present studies on hydration of tricalcium silicate in the
presence of various salts have shown that there are several fO»
ble mechanisms for the fixation of metallic ions using portland ce-
ment: (1) fixation by an addition reaction between a metallic fa"
and calcium silicate hydrate formed by the hydration of cement,
(2) fixation by substitution reaction in which calcium ions ftoo
-------
calcium silicate hydrates are replaced by metallic ions, (3) fixation
of metallic ions by the formation of a new compound or com-
pounds, and (4) fixation by a combination of the above mechan-
isms.
ACKNOWLEDGEMENTS
The author expresses his appreciation to S.W. Tresouthick for
the valuable discussions, to K.A. Tylor for experimental work,
and S.L. Ito for preparing the final copy of the manuscript.

REFERENCES
1. Tittlebaum, M.E., Seals, R.K., Cartledge, F.K. and Engels, S.,
"State of the Art on Stabilization of Hazardous Organic Liquid
Wastes and Sludges," CRC Critical Reviews in Environ. Control,
15, 2, 1985, 179-210.
2. Tashiro, C., Takahashi, H., Kanaya, M., Hirakida, I. and Yoshida,
R., "Hardening Properties of Cement Mortar Adding Heavy Metal
Compound and Solubility of Heavy Metal from Hardened Mortar,"
Cem. Concr, Res., 1, 1977, 283-290.
3. Tashiro, C., Obs, J. and Akawa, K., "The Effects of Several Heavy
Metal Oxides in the Formation of Ettringite and the Microstructure
of Hardened Ettringite," Cem. Concr. Res., 9, 1979, 303-309.
4. Tashiro, C. and Oba, J., "The Effect of Cu(OH)2 on the Hydration
of CjA," Proc. Seventh International Congress on Chemistry of
Cement, Paris, France, 2, II, 1980, 58-63.
5. Stepanovo, I.N., "Hardening of Cement Pastes in Presence of
Chloride of 3d Elements," J. Appl. Chem., 54, 1981, 885-888.
6. Poon, C.S., Peters, C.J. and Perry, R., "Use of Stabilization Pro-
cesses in the Control of Toxic Waste," Effl. Wat. Tr. J., 23, 1983,
451-459.
7. Poon, C.S., Clark, A.I., Peters, C.J. and Perry, R., "Mechanisms
of Metal Fixation and Leaching by Cement-Based Processes," Waste
Man. Res., 3, 1985, 127-142.
8. Poon, C.S., Peters, C.J., Perry, R., Barnes, P. and Barker, A.P.,
"Mechanism of Metal Stabilization by Cement-Based Fixation Pro-
cesses," Sci. of the Total Environ., 41, 1985, 55-71.
9. Poon, C.S., Clark, A.I. and Perry, R., "Permeability Study of the
Cement-Based Solidification Process for the Disposal of Hazardous
Wastes," Cem. Concr. Res., 16, 1986, 162-172.
10. Glasser, P.P., Rahman, A.A., Craford, R.W., McCullich, C.E. and
Angus, M.J., "Immobilization and Leaching Mechanisms of Rad-
waste in Cement-Based Matrices," DOE Report No. DOE/RW/
83.093, 1, Department of the Environment, London, 1983.
11. Malone, P.O., Jones, L.W. and Burker, J.P., "Application of
Solidification/Stabilization Technology to Electroplating Wastes,"
Land Disposal of Hazardous Wastes, Proc. of the Ninth Annual
Research Symposium, Fort Mitchell, KY, PB-84-118777, U.S. EPA,
1984, 247-260.
12. Kantro, D.L., Brunauer, S., and Wiese, C.H., "Development of
Surface in the Hydration of Calcium Silicates. 11-Extraction of In-
vestigations of Earlier and Later Stages of Hydration," J. Phys.
Chem., 151, 66, 1962, 1804-1809.
13. Bhatty, M.S.Y. and Greening, N.R., "Interaction of Alkalies with
Hydration and Hydrated Calcium Silicates," Proc. of the Fourth
International Conference on Effects of Alkalies in Cement and
Concrete, Purdue University, Lafayette, IN, June 1978, 87-111.
14. Kalousek, G.L., "Studies of Portion of Quaternary System Soda-
Lime-Silica-Water at 25°C," J. of Res., National Bureau of Stan-
dards, 32, 1944, 285-302.
15. Bhatty, M.S.Y., Unpublished Data.
16. Taylor, H.R.W., "Hydrated Calcium Silicates, Compound Forma-
tion at Ordinary Temperatures," J. Chem. Soc. (London), 1950,
3682-3690.
17. Menetrier, D., Jawad, I. and Skalny, J., "Effect of Gypsum on
CsS Hydration," Cem. Concr. Res., 10, 1980, 697-701.
18. Bhatty, M.S.Y., "Mechanism of Pozzolanic Reactions and Control
of Alkali-Aggregate Expansion," Cement, Concrete and Aggre-
gates, CCAGDP, 7, No. 2, 1985, 69-77.
TREATMENT 145
-------
Mineralization of Recalcitrant Environmental
Pollutants by a White Rot Fungus

John A. Bumpus
Steven D. Aust
Department of Biochemistry
Michigan State University
East Lansing, Michigan
ABSTRACT
The white rot fungus Phanerochaete chrysosporium is able to
degrade lignin, a structurally complex, naturally occurring and
environmentally persistent, non-repeating heteropolymer.
Previous studies have shown that this fungus is also able to de-
grade a wide variety of synthetic organopollulants and that biode-
gradation is dependent, at least in part, on the lignin degrading
system. Examples of recalcitrant chemicals that are degraded to
carbon dioxide by this fungus include 3,4,3',4'-tetrachlorobi-
phenyl, 2,4,5,2',4',5'-hexachlorobiphenyl, 2,3,7,8-tetrachlorodi-
benzo[p]dioxin (TCDD), 1,1,1 -trichloro-2,2-bis(4-chlorophenyl)-
ethane (DDT), 1,2,3,4,5,6-hexachlorocyclohexane (Lindane) and
benzo[a]pyrene.
In the present study, we have shown that Chlordane, 2,2,2-
trichloro-l,l-bis(4-chlorophenyl)ethanol (dicofol), pentachloro-
phenol, Aroclor-1254, phenanthrene, biphenyl, p-cresol and
2-methylnaphthalene also are degraded to carbon dioxide by this
fungus. A number of these compounds were selected for further
study to more thoroughly document biodegradation. Using
Chlordane and pentachlorophenol it was shown that, like lignin,
mineralization of these two environmentally persistent xenobio-
tics was promoted in nutrient nitrogen deficient cultures while
mineralization was suppressed in nutrient nitrogen sufficient cul-
tures.
Mass balance and metabolite formation studies were performed
on cultures of P. chrysosporium that had been incubated with
i4C-Chlordane, nC-phenanthrene or i^C-dicofol. In all cases,
these compounds were metabolized to more polar metabolites.

INTRODUCTION
Fungi, along with bacteria, are the primary decomposers
responsible for the recycling of organic matter in the biosphere.
The biodegradative ability of bacteria as a group has been exten-
sively studied and exploited for use in waste treatment systems.1'2
In contrast, the use of fungi in waste treatment systems has re-
ceived less attention in spite of the fact that a number of studies
clearly demonstrate that fungi as a group have remarkable
biodegradative abilities. For example, Amorphotheca resinae is
able to grow on kerosene or cresote3-4 and Cunninghamella
bainieri is known to oxidize benzo[a]pyrene.5 Similarly, a number
of fungi have been implicated in the biodegradation of various
synthetic organopollutants, i.e., pesticides.6-*
Our studies focus on the ability of the white rot fungus Phaner-
ochaete chyrsosporium to degrade xenobiotics.*-'5 Research in
our laboratory and by others16-'9 has demonstrated that, under
nutrient nitrogen-limiting conditions, this microorganism is able
to degrade a broad spectrum of structurally diverse environmental
pollutants to carbon dioxide. Direct13-'8 and indirect*-" evidence
has demonstrated that the lignin degrading system of this fungus,
which is expressed at the onset of idophasic (secondary) metabol-
ism, is responsible, at least in part, for its remarkable biodegrada-
tive abilities.
In the present study, we have examined the ability of P. chryso-
sporium to mineralize a number of representative organohalides
(dicofol, Chlordane, pentachlorophenol and Arclor-1254). Addi-
tionally, we have examined the ability of this fungus to mineralize
four constituents found in coal tar (2-methylnaphthalene, phen-
anthrene, p-cresol and biphenyl).

MATERIALS AND METHODS
Fungus
Phanerochaete chrysosporium (BKM-F-1767) was obtained
from the United States Department of Agriculture, Forest Pro-
ducts Laboratory, Madison, Wisconsin. The fungus was main-
tained on malt agar slants at room temperature until used. Sub-
cultures were routinely made every 30-60 days.

Chemicals
Carbon-14 labeled Chlordane (5.91 mCi/mmole), phenan-
threne (10 mCi/mmole), 2-methylnaphthalene (8.57 mCi/
mmole), p-cresol (10.33 mCi/mmole), pentachlorophenol (PCP)
(10.57 mCi/mmole) and biphenyl (15.91 mCi/mmole) were pur-
chased from Pathfinder Laboratories, Inc. (St. Louis, Missouri).
Carbon-14 labeled Aroclor 1254 (32 mCi/mmole) was purchased
from Amersham (Arlington Heights, Illinois). Carbon-14 labeled
dicofol (9.73 mCi/mmole) was a gift from the Rohm and Haas
Co. (Springhouse, Pennsylvania).
Chlordane, Arclor 1254, pentachlorophenol, p-cresol and bi-
phenyl were uniformly labeled. Dicofol was uniformly labeled only
in the aromatic rings, while phenanthrene and 2-methylnaphtbfr
lene were labeled hi the 9 and 8 positions, respectively. The radio-
chemical purity of chemicals used in this study was 98% or
greater. In the case of i4C-Aroclor 1254, a polychlorinated
biphenyl (PCB) mixture, it was shown that approximately 98* of
the radioactivity in this mixture comigrated with authentic,
unlabeled Aroclor 1254 during TLC in hexane. Comparison of
i«C-Aroclor 1254 with unlabeled Aroclor 1254 by GLC rev*!*
that they are very similar in composition, although minor differ-
146
TREATMENT
-------
ences in their GLC elution patterns could be detected.

Culture Conditions and Mineralization Studies
P. chrysosporium was incubated at 37-39 °C in 10 ml of the liq-
uid culture media described by Kirk et a/.20 This medium con-
sisted of 56 mM glucose, 1.2 mM ammonium tartrate, trace ele-
ments and thiamine (1 mg/1) in 20 mM dimethyl succinate buffer
(pH 4.2). Cultures were established by innoculating this media
with spores as described.20 For mineralization studies, i4C-labeled
chemical in a minimal «"30 /il) volume of acetone was added at
this time.
During the first 3 days of incubation, cultures were allowed to
grow under an atmosphere of air in culture bottles equipped with
a gas exchange manifold. After 3 days, and at 3-day intervals
thereafter, cultures were gently flushed with oxygen. The at-
mosphere from each culture was forced through 10 ml of an
ethanolamine-containing scintillation cocktail which served as a
CO2 trap. Radioactivity in the CO2 was assayed by liquid scintilla-
tion spectrometry. Details of this mineralization assay and culture
conditions have been previously reported.9'20

Metabolite Analyses
HPLC of i*C-dicofol metabolites was performed using a
system equipped with an Alltex pump (Model 110A), a Rheodyne
injector, an Alltech R-Sil C-18 reverse phase column (4.6 x
250 mm) and a Schoeffel detector (Model 770). Isocratic elution
was performed using 85% methanol. The retention time of
dicofol and dicofol metabolites was established by monitoring the
elution of authentic standards at 238 nm.
Authentic dicofol and FW-152 were gifts from the Rohm and
Haas Co. Other standards were purchased from the Aldrich
Chemical Co. (Milwaukee, Wisconsin).
To establish the presence of i4C-dicofol metabolites, four
cultures (10 ml) of P. chrysosporium which had been incubated
with nC-dicofol (5-0 nmoles/culture) were pooled. Twenty-
five ml of acetonitrile then were added and the mixture was
homogenized in a Potter Elvehjem hand homogenizer equipped
with a Teflon pestle. The homogenized material then was ex-
tracted with two 50 ml portions of hexane which were pooled. The
extract then was dried over magnesium sulfate and filtered. Hex-
ane subsequently was removed by evaporation under Argon.
Following hexane extraction, the aqueous phase was acidified
to pH 2.0 with concentrated HC1 and extracted with two 50 ml
portions of methylene chloride which were pooled. The methylene
chloride extract then was dried over magnesium sulfate and
filtered. Methylene chloride subsequently was removed by
evaporation under Argon.
The hexane extracts then were dissolved in hexane (— 1 ml)
while the methylene choride extracts were dissolved in methanol
(-1 ml). Ten to 20 /*! of these were used for HPLC analyses. Pre-
cipitates which were sometimes formed upon concentration were
removed by filtration through glass wool or by centrifugation.
Following sample injection, 1 ml aliquots were collected in scin-
tillation vials. Ten ml of Safety Solve (Research Products Inter-
national Corp., Mount Prospect, Illinois) then were added to
each vial and radioactivity was determined by liquid scintillation
spectrometry.
Cultures of P. chrysosporium incubated with i4C-phen-
anthrene were extracted using the same procedures for mass bal-
ance experiments. For HPLC analysis of phenanthrene metabo-
lites, cultures were acidified and extracted with methylene
chloride. The methylene chloride extract (100 ml) was concen-
trated to 2 ml under a gentle flow of Argon in a centrifuge tube.
The extract then was allowed to sit uncovered in a fume hood
overnight. During this time, the remaining methylene chloride
evaporated.
The residue was redissolved in methanol and subjected to
HPLC analysis. Thin layer chromatography was performed on
hexane and methylene chloride extracts obtained from cultures of
P. chrysosporium that had been incubated with '4C-ChIordane.
Extraction of these cultures and concentration of hexane and
methylene chloride extracts were performed as described above
for extraction of '^C-dicofol and i4C-phenanthrene.
Aliquots of the hexane and methylene chloride extracts were
chromatographed on pre-coated, aluminum-backed thin layer
plates (5 x 20 cm, Silica Gel 60 F254, EM Science, Cherry Hill,
New Jersey). The solvent system for chromatography was methy-
lene chloride. Following chromatography, 1 cm fractions were
scraped from the thin layer plate and placed into scintillation
vials. Ten ml of Safety Solve then were added to each vial and
radioactivity was determined by liquid scintillation spectrometry.

Mass Balance Experiments
Following incubation of i^C-labeled chemicals in nutrient ni-
trogen deficient cultures of P. chrysosporium, cultures were ex-
tracted sequentially with hexane and methylene chloride as de-
scribed above. Following extraction, mycelium was separated
from the aqueous fraction by filtration. Ten ml of Safety Solve
then were added to the recovered mycelium in a scintillation vial.
Ten ml of Safety Solve also were added to 1 ml aliquots of the
hexane, methylene chloride and aqueous fractions. The radioac-
tivity of all fractions was determined by liquid scintillation spec-
trometry.

RESULTS
Table 1 shows that a wide variety of structurally diverse organ-
ohalides and polycyclic aromatic hydrocarbons are mineralized by
P. chrysosporium under nutrient nitrogen-limiting conditions.
The structures of the chemicals used in this study are presented in
Fig. 1. In general, very little or no mineralization occurred during
the first 3 days of incubation in spite of the fact that abundant
growth, as evidenced by the appearance of a mycelial mat, oc-
curred during this time. However, substantial mineralization typi-
cally began between day 3 and day 6 of the incubation and usually
was maximal between day 3 and days 12 to 18. Although the rate
of mineralization generally declined during the last one-half of the
30-day incubation period, in no case did mineralization cease.
Furthermore, when supplemental glucose was added after 30 days
of incubation, the rate of mineralization of i4C-labeled com-
pounds increased in all cases.
A lag period before the onset of mineralization is consistent
with our hypothesis that the lignin degrading system of this
fungus is, at least in part, responsible for the biodegradative
abilities of this fungus because a similar lag period is seen before
mineralization of '^C-lignin.
Table 1
Mineralization of 14-C-Radiolabeled Compounds by P. chrysosporium
Initial (
of "C-l«l
111 Concentration
•labeled Conpound
(naolea/culture)
Radlolebeled Subatrate $ of Radlolabeled
Evolved » "C02 Subalratee Evolved I
(piolea i SO) "CO; In 60 [leva
30 daya 60 daya
P«ntiohlorophtnol

Chlordine

Aroolw 1251

Dleofol

Phtninthrane
1.25

5.0

1.25

5.0

1.25

5.0
521.0 *_ 97.2

508.1) l 52.2

179.3 1 17.1

800.6 *_ 200

621.0 *_ 61.5

366.1 *. 86.4

366.B * 91.5

2086.8 *_ 175.0
570.2 * 105.7

608.0 1 56.8

22-1.» «_ 57.0

J26S.2 ^ 336.1

738.1 •_ 62.8

136.6 *, 119.1

155.8 ^ 103.4

2378.1 *, 215.5
15.6 * 8.5

: *'*
i *.6

> 6.7
*,8.3

i ".3
TREATMENT 147
-------
II
III
VI
In addition to mineralization studies, mass balance studies wen
performed using cultures of P. chrysosporium incubated with
i4C-Chlordane, 14C-dicofol and uC-phenanthrene, representing
a polychlorinated aliphatic, a polychlorinated aromatic and a
polycyclicaromatic hydrocarbon, respectively. In all cases, the
formation of substantial quantities of polar and water soluble
metabolites was demonstrated (Table 3). Total mass recovery in
these experiments was relatively low: 40%, 75% and 52% for
Chlordane, dicofol and phenanthrene, respectively. The reason
for this low recovery is unknown. The possibility that abiotic pro-
cesses, in addition to microbial degradation, may account for
some loss of the parent compound cannot be excluded at this
time.
Hexane extracts and methylene chloride extracts were further
examined by TLC or HPLC for the presence of metabolites. Only
unmetabolized parent compound was found in the hexane extract
obtained from cultures of P. chrysosporium incubated with HC-
Chlordane for 60 days. However, in the methylene chloride ex-
tract (Fig. 2), two very polar metabolites of ^C-Chlordane were
found during TLC analysis. Analysis of hexane extracts of HC-
phenanthrene was complicated by the fact that low recoveries of
i4C-labeled material were obtained following concentration of the
extract under Argon.
VII
nil
Figure 1
Structures of Compounds Degraded to CO2 by P. chrysosporium. I. di-
cofol (2,2,2-trichloro-l,l-bis[4-chlorophenyl]ethanol); II. 2-methylnaph-
thalene; III. Chlordane (l,2,4,5,6,7,8,8-octachloro-2,3,3a,4,7,7a-hexa-
hydro-4,7-methano-lH-indene); IV. PCP (pentachlorophenol); V. Aro-
clor 1254; VI. biphenyl; VII. phenanthrene; VIII. p-cresol.

Similar results have been reported by us9-is and by others16-'9
for the mineralization of a number of other persistent environ-
mental pollutants. Eaton16 also studied mineralization of 14C-
Aroclor-1254 by P. chrysosporium (BKM-F-1767) and has
reported similar findings.
Other evidence suggesting that the lignin degrading system is re-
sponsible for the unique biodegradative abilities of this fungus is
presented in Table 2. We have previously shown that, like '*C-
lignin mineralization, mineralization of 14C-DDT is promoted in
nutrient nitrogen limiting cultures whereas mineralization of both
compounds is suppressed in nutrient nitrogen sufficient cultures.9
Table 2 shows that similar results were obtained when '^C-Chlor-
dane and |4C-PCP were mineralized by this fungus. Thus, these
data also suggest that the lignin-degrading system may be respon-
sible for the mineralization of these compounds.

Table 2
Recovery of Radioactivity Following Incubation of l4C-ChIordane,
14C-Phenanthrcnt and 14C-Dicofol with Nutrient Nitrogen Deficient
Cultures of P. chrysosporium
1 *C-OUorteM
• The imlial concentration of '*C labeled chemical was 50 nmolo/cullure.

148 TREATMENT
Table 3
Effect of Nutrient Nitrogen Concentration on Mineralization
of 14C-Chlordane and MC-PCP by P. chrysosporium

Amount of Carbon-11 Labeled
Coapound Mineralized (nanoooles • SO)
Conpound Mineralized
1 l|C-ChlOPdaneb
'""c-pcpc
Aomonlui
1.2 DM
2.83 « 0.26
3.55 i 1.75
a Tartrate Concentration
12.0 aH
0.15 • 0.21
0.96 • 0.66
a. Cultures were incubated for 30 days in the presence of '^C-labeled chemical.
b. Initial concentration = 30 nmole/culture.
k. Initial concentration = 17.6 nmole/culture.
3000
2000
1000
0 2 4 6 8 10 12 14 16 18 20
FRACTION (CM)
Figure 2
TLC elution profile of the methylene chloride extract of culture! of f.
chrysosporium which had been incubated with l4C-Chtordane ft" *"
days. The arrow represents the position to which undegraded MC-CMor-
dane migrated in controls.
-------
In an attempt to circumvent this problem, the hexane extrac-
tion step was omitted and extracts of cultures of P. chrysospor-
ium which had been incubated with i4C-phenanthrene were
acidified and extracted with methylene chloride. Upon gentle con-
centration under Argon, nearly quantitative amounts of the ex-
tracted radioactivity were recovered. Subsequent analysis of this
material via HPLC (Fig. 3) did not detect the presence of "C-
phenanthrene. However, a considerable amount of i4C-labeled
polar material was present. The i4C-labeled metabolites formed
from i4C-Chlordane and nC-phenanthrene were not further
characterized. Hexane extracts of i^C-dicofol which had been in-
cubated with P. chrysosporium for 60 days were examined by
HPLC (Fig. 4).
1500
1000
500
0 2 4 6 8 10 12 14 16 18 20
FRACTION (ML)

Figure 3
HPLC elution profile of the methylene chloride extract of cultures of P.
chrysosporium which had been incubated with 14C-phenanthrene for 60
days. The arrow represents the point at which undegraded 14C-phenan-
threne eluted in controls.

These studies showed that at least four metabolites of 14C-
dicofol were formed by this fungus. Two of these metabolites
were shown to comigrate with FW-152 (2,2-dichloro-l,l-bis[4-
chloro-phenyl]ethanol) and DPB (4,4'-dichlorobenzophenone).

DISCUSSION
Mineralization of Xenobiotics
These studies and others9'19 clearly demonstrate that the white
rot fungus P. chrysosporium has the ability to degrade a broad
spectrum of organic compounds, many of which are normally
regarded as being "hard-to-degrade" or "persistent." A substan-
tial portion of our work focused on the mineralization of 14C-
labeled chemicals. The greatest value of mineralization studies lies
in the fact that positive results unequivocally document the ability
of the microorganism to degrade the i4C-labeled compound in
question to 14CO2, the final oxidation product to aerobic systems.
Unfortunately, however, mineralization provides only a
minimal indication of the amount of biodegradation that has ac-
7000
eooo
sooo
4000
3000
2000
1000
-FK-152
-DICOFOL
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
FRACTION (ML)
Figure 4
HPLC elution profile of the hexane extract of cultures of P. chrysospor-
ium which had been incubated with 14C-dicofol for 60 days.

tually occurred. For example, in our earlier studies,9 we reported
that approximately 4% of the 14C-DDT originally present in
nutrient nitrogen starved cultures of P. chrysosporium was miner-
alized during a 30 day incubation period. However, during the
same period of time, approximately 50% of the DDT originally
present had disappeared.
Mass balance studies9 showed that the 14C-DDT had been
metabolized to more polar and water soluble metabolites that are
intermediates in the pathway between 14C-DDT and 14CO2.
Metabolites identified by GC-MS included ODD (1,1-dichloro-
2,2 bis[4-chlorophenyl]ethane), dicofol, FS-152 and DBF.9.15 In
the present study we have shown that, not unexpectedly, i4C-di-
cofol, an intermediate in 14C-DDT degradation, is also mineral-
ized and that the pattern of metabolite formation is very similar to
that observed for 14C-DDT degradation.
Mass balance studies also showed that, like 14C-dicofol, a
representative polychlorinated aromatic used in this study, the
polycyclicaromatic compound i4C-phenanthrene and the alkyl
halide i4C-Chlordane were converted to polar and water soluble
metabolites by this fungus. Thus these studies clearly document
the extensive biodegradation of these compounds.

The Lignin Degrading System
and Xenobiotic Degradation
The ability of this fungus to degrade such a wide variety of
structurally different organopollutants apparently is due to the
lignin-degrading system of this fungus.9-19 Lignin is a naturally oc-
curring, complex and hard-to-degrade heteropolymer.22 Lignin is
resistant to microbial biodegradation because, unlike homopoly-
mers, the many structural sub-units in lignin do not appear at
regular intervals in the polymer.21 Furthermore, there are at least
12 different types of carbon-carbon and carbon-oxygen bonds
which link the various heterogeneous sub-units.22 The problem is
further complicated by the fact that chiral carbons in the lignin
molecule lack stereoregularity; that is, they occur equally in the D
and L configuration.21
The study of lignin degradation has been the topic of con-
siderable study and speculation for the past 20 yr. Studies showed
that fungi, particularly white rot fungi, were the primary decom-
posers of lignin.21 Many of these studies were performed using the
white rot fungus P. chrysosporium (or its anomorph, Sporotri-
chum pulverulentum) because of its relatively pronounced lignin-
TREATMENT 149
-------
degrading ability, rapid growth and ease of handling. It has only
been within the last 4 yr that substantial progress has been made
in the elucidation of the mechanism by which this fungus
degrades lignin.
Studies by Kirk and his associates showed that under nutrient
limiting conditions this fungus secretes a powerful oxidant that is
able to degrade lignin.20-23 A number of studies suggested that ac-
tive oxygen species such as hydroxyl [• OH] or superoxide [O2~]
free radicals might be the oxidants responsible for the ability of
this fungus to degrade lignin (24-26). A free radical mechanism
would have the advantage of being both non-specific and non-
stereoselective, characteristics that would be required to degrade
the lignin polymer.
Recent studies, however, have shown that under nutrient
nitrogen-limiting conditions, this fungus secretes a family of
hydrogen peroxide-requiring heme proteins, collectively known as
ligninases, that are able to catalyze one electron oxidations of
lignin as well as many other organic compounds which resemble
selected substructures of the lignin polymer and which often are
used to study lignin biodegradation.27-32 The radicals so generated
then decay, according to their structural composition, resulting in
bond cleavage and lignin depolymerization. The lignin-derived
products so formed then may be further modified by the
ligninases or other, more conventional enzyme systems. Eventual-
ly they are converted to Krebs cycle intermediates which then, in
turn, are converted to CO2 to complete the mineralization pro-
cess.
The fact that this biodegradation system was non-specific and
non-stereoselective suggested to us that it might also be effective
in the biodegradation of hard-to-degrade synthetic compounds.9
Furthermore, a free radical mechanism would be expected to be
able to cleave carbon-chlorine bonds. The fact that mineralization
of xenobiotics, like the mineralization of lignin, is promoted
under nutrient nitrogen-limiting conditions and suppressed under
nutrient nitrogen-sufficient conditions is indirect evidence that
the lignin degrading system is mediating biodegradation of these
chemicals.
Direct evidence for involvement of the lignin degrading system
in the biodegradation of xenobiotics also is available. For exam-
ple, it has been shown that purified ligninases from P. chryso-
sporium are able to oxidize benzo[a]pyrene18 and crystal violet.13

Characterization of Xenobiotic
Degradation in P. chrysosporium
Biodegradation of xenobiotics and lignin by P. chrysosporium
appears to require the presence of another organic compound
(glucose, for example) which serves as a growth substrate. Thus,
biodegradation of xenobiotics in this species occurs via co-
metabolism.33 In the more restricted nomenclature proposed by
Matsumura, this type of biodegradation may alternatively be clas-
sified as incidental metabolism "due to generally present broad-
spectrum enzymes."33 However, because these enzymes are only
present during idiophasic metabolism, this definition should be
even further modified to reflect this fact.

Use of P. chrysosporium in Practical
Waste Treatment Systems
Because of its unique biodegradative abilities, considerable at-
tention has been given to the possible use of this fungus in waste
treatment systems. Kirk and his associates have shown that
nutrient nitrogen-deficient cultures of this fungus can be used in
rotating biological contactors to degrade chlorinated phenols and
related compounds that are by-products of the bleaching phase
that often is used with the Kraft paper pulping process.34
Less attention has been given to the use of this microorganism
for the decontamination of soils, sediments and sludges because.
in part, more technical problems are likely to be encountered in
such complex systems. The use of P. chrysosporium to degrade
organopollutants in these systems currently is under study in our
laboratory.

CONCLUSIONS
The white rot fungus Phanerochaete chrysosporium degrades a
wide variety of structurally diverse organopollutants to carbon
dioxide. In this study, Chlordane, dicofol, PCP, Aroclor-1254^
phenanthrene, biphenyl, p-cresol and 2-methylnaphthalene were
shown to be mineralized by this fungus in nutrient nitrogen-
deficient cultures. Direct and indirect evidence indicates that the
lignin-degrading system of this fungus is responsible, at least in
part, for its remarkable biodegradative abilities.

ACKNOWLEDGEMENTS
This work was supported by Cooperative Agreement
#CR813369, U.S. EPA, Office of Research and Development,
Hazardous Waste Engineering Research Laboratory, Cincinnati,
Ohio, P.R. Sferra, Project Officer.

REFERENCES
1. Flint, K.P., "Microbial Ecology of Domestic Wastes," in Experi-
mental MicrobialEcology, R.G. Burns and J.H. Slater, Eds., Black-
well Scientific Publishers, London, 1982, 575-590.
2. Gasner, L.L., "Microorganisms for Waste Treatment," in Microbial
Technology, H.J. Peppier and D. Perlman, Eds., Academic Press,
Inc., New York, NY, 1979, 211-222.
3. Sheridan, J.E., "Studies on the Kerosene Fungus Amorphothecc
resinae," New Zealand Int. Biodetn. Bull. 8, 1972, 65-74.
4. Marsden, D.H., "Studies of the Creosote Fungus Hormodendnm
resinae," Mycologia 46, 1954, 161-183.
5. Ferris, J.B., MacDonald, M.A., Patrie, M.A. and Martin, M.A.,
"Arylhydrocarbon hydroxylase Activity in the Fungus Cunning-
hamella bainieri: Evidence for the Presence of Cytochrome P-450,"
Arch. Biochem. Biophys., 175, 1976, 443-452.
6. Subba-Rao, R.V. and Alexander, M., "Bacterial and Fungal Co-
metabolism of l,l,l-trichloro-2,2-bis(4-Chlorophenyl) Ethane
(DDT) and its Breakdown Products," Appl. Environ. MicrobM.,
49, 1985, 509-516.
7. Schilling, R., Engelhardt, G. and Wallnofer, P.R., "Degradationof
the Herbicide Metribuzin (Sencor) by Pure Cultures of Cunning-
hamella echinulata Thaxter ATCC 38447," Chemosphere, 14,1985,
267-270.
8. Baarschers, W.H. and Heitland, H.S., "Biodegradation of Feni-
trothion and Fenitrooxon by the Fungus Trichoderma viride," 1.
Agric. Food Chem., 34, 1986, 707-709.
9. Bumpus, J.A., Tien, M., Wright, D. and Aust, S.D., "Oxidationof
Persistent Environmental Pollutants by a White Rot Fungus," Sri-
ence, 228, 1985, 1434-1436.
10. Bumpus, J.A. and Aust, S.D., "Studies on the Biodegradation of
Organopollutants by a White Rot Fungus," in Proc. International
Conference on New Frontiers for Hazardous Waste Management,
Pittsburgh, PA, U.S. EPA, EPA/600/9-85/025, 1985, 404-410.
11. Bumpus, J.A., Tien, M., Wright, D. and Aust, S.D., "Biodegrada-
tion of Environmental Pollutants by the White Rot Fungus PHantro-
chaele chrysosporium," in Incineration and Treatment of Hapd-
ous Wastes: Proc. of the Eleventh Annual Research Symptutem>
Cincinnati, OH, 1985, U.S. EPA, EPA/600/9-85/028, 1985, !»•
126.

12. Bumpus, J.A. and Aust, S.D., "Biodegradation of Environmental
Pollutants by the White Rot Fungus Phanerochaete chryso^orhOK
Involvement of the Lignin Degrading System," BioEssays (InPn»)-
13. Bumpus, J.A. and Aust, S.D., "Biological Oxidations by Enzyn*
from a White Rot Fungus," Paper presented at American
150
TREATMENT
-------
of Chemical Engineers 1986 Summer Meeting, Boston, MA, Aug.
1986.
14. Bumpus, J.A. and Aust, S.D., "Biodegradation of Chlorinated
Organic Compounds by Phanerochaete chrysosporium, A Wood
Rotting Fungus, in Solving Hazardous Waste Problems, J.H. Ex-
ner, Ed., ACS Books, Washington, DC (In Press).
15. Bumpus, J.A. and Aust, S.D., "Biodegradation of DDT (1,1,1-
trichloro-2,2-bis[4-chlorophenyi]ethane by the White Rot Fungus
Phanerochaete chrysosporium," (Submitted for publication).
16. Eaton, D.C., "Mineralization of Polychlorinated Biphenyls by
Phanerochaete chrysosporium: A Ligninolytic Fungus," Enzyme
Microb. Technol., 7, 1985, 194-196.
17. Sanglard, D., Leisola, M.S.A. and Fiechter, A., "Role of Extra-
cellular Ligninases in Biodegradation of Benzo[a]pyrene by Phanero-
chaete chrysosporium," Ensyme Microbiol. Technol., 8, 1986,
209-212.
18. Hacmmerli, S.D., Leisola, M.S.A., Sanglard, D. and Fiechter, A.,
"Oxidation of Benzo[a]pyrene by Extracellular Ligninases of Phan-
erochaete chrysosporium: Veratryl Alcohol and Stability of Lig-
ninases," J. Biol. Chem., 261, 1986, 6900-6903.
19. Arjmand, M. and Sandermann, H., "Mineralization of Chloro-
aniline/Lignin Conjugates and of Free Chloroanilines by the White
Rot Fungus Phanerochaete chrysosporium" J. Agric. Food Chem.,
33, 1985, 1055-1060.
20. Kirk, T.K., Schultz, E., Connors, W.J., Lorenz, L.F. and Zeikus,
J.G., "Influence of Culture Parameters on Lignin Metabolism by
Phanerochaete chrysosporium," Arch. Microbiol., 117, 1978, 277-
285.
21. Crawford, R, Lignin Biodegradation and Transformation, John
Wiley and Sons, New York, NY, 1981, 154.
22. Adler, E., "Lignin Chemistry—Past, Present and Future," Wood
Sci. Technol., 11, 1977, 169-218.
23. Keyser, P., Kirk, T.K. and Zeikus, J.G., "Ligninolytic Enzyme
System of Phanerochaete chrysosporium: Synthesized in the Ab-
sence of Lignin in Response to Nitrogen Starvation," J. Bacterial.,
135, 1978, 790-797.
24. Hall, P.L., "Enzymatic Transformation of Lignin," Enzyme Mi-
crobiol. Technol., 2, 1980, 170-176.
25. Forney, L.J., Reddy, C.A., Tien, M. and Aust, S.D., "The Involve-
ment of Hydroxyl Radical Derived from Hydrogen Peroxide in
Lignin Degradation by the White Rot Fungus Phanerochaete
chrysosporium," J. Biol. Chem., 257, 1982, 11455-11462.
26. Kutsuki, H. and Gold, M.H., "Generation of Hydroxyl Radical and
Its Involvement in Lignin Degradation by Phanerochaete chryso-
sporium," Biochem. Biophys. Res. Commun. 109, 1982, 320-327.
27. Tien, M. and Kirk, T.K., "Lignin-Degrading Enzyme from the
Hymenomycete Phanerochaete chrysosporium Burds," Science, 221,
1983, 661-663.
28. Tien, M. and Kirk, T.K., "Lignin-Degrading Enzyme from Phanero-
chaete chrysosporium: Purification, Characterization, and Cata-
lytic Properties of a Unique HaOa-Requiring Oxygenase," Proc.
Natl. Acad. Sci., 81, Washington, DC, 1985, 2280-2284.
29. Renganathan, V., Miki, K. and Gold, M.H., "Multiple Molecular
Forms of Diarylpropane Oxygenase, A H2O2-Requiring, Lignin
Degrading Enzyme from Phanerochaete chrysosporium," Arch.
Biochem. Biophys., 241, 1985, 304-314.
30. Paszczynski, A., Huynh, V.B. and Crawford, R.L., "Comparison
of Ligninase-I and Peroxidase-M2 from the White Rot Fungus
Phanerochaete chrysosporium," J. (unidentified) 224, 1985, 750-
765.

31. Harvey, P .J., Schoemaker, H.E., Bowen, R.M. and Palmer, J.M.,
"Single-electron Transfer Processes and the Reaction Mechanism of
Enzymic Degradation of Lignin," REBS Lett., 183, 1985, 13-16.

32. Schoemaker, H.E., Harvey, P.J., Bowen, R.M. and Palmer, J.M.,
"On the Mechanism of Enzymatic Lignin Breakdown," FEBS Lett.,
183, 1985, 7-12.

33. Matsumura, F., "Degradation of Pesticides in the Environment by
Microorganisms and Sunlight," in Biodegradation of Pesticides,
F. Matsumura and C.R.K. Murti, Eds., Plenum Press, New York,
NY, 1982, 67-87.

34. Huynh, V.B. , Chang, H.-m., Joyce, T.W. and Kirk, T.K., "De-
chlorination of Chloro-organics by a White Rot Fungus, TAPPI,
68:(tfl), 1985, 98-102.
TREATMENT 151
-------
The Effect of Volatile Organic Compounds on the
Ability of Solidification/Stabilization Technologies
To Attenuate Mobile Pollutants

James H. Kyles
Kenneth C. Malinowski, Ph.D.
CECOS International, Inc.
Department of Research and Development
Buffalo, New York
Judith S. Leithner
Thomas F. Stanczyk
Recra Environmental, Inc.
Amherst, New York
ABSTRACT
Volatile organic compounds (VOCs) are a subset of organic
contaminants that, in a majority of cases, have a negative impact
on conventional pozzolanic solidification of hazardous waste. In
studies performed by CECOS International, Inc. and Recra En-
vironmental, Inc., RCRA hazardous solids, semisolids and pump-
able sludges containing controlled quantities of VOCs were
studied to determine the effect of the VOCs on hazardous waste
solidification.
Throughout the research program, it was clear that, without
prior removal, VOCs antagonistically impacted conventional
solidification processes. Under the alkaline conditions typically
present during pozzolanic solidification, organic constituents
continued to leach from the product after 72 hr of curing. During
investigations evaluating alternative stabilization techniques such
as polymeric encapsulation, VOCs and solvents not only hindered
the curing of the product, but appeared to enhance the leachabil-
ity of pollutants.
The research presented herein represents a portion of on-going
investigations being conducted by CECOS International and
Recra Environmental to produce a technically and economically
attractive alternative to incineration for hazardous wastes con-
taining organics. The targeted market for this alternative includes
those wastes containing high ash, sulfur, metals and nitrogen con-
tents, and/or other problem parameters associated with incinera-
tion destruction.

INTRODUCTION
The Ten Year Technology Plan, developed by CECOS Interna-
tional, Inc. in 1981, presents a comprehensive program leading to
the implementation of appropriate technologies for the treat-
ment, detoxification, recovery, volume reduction and reuse of
hazardous waste. As part of the plan, CECOS has, through its
contracted consultant Recra Environmental, initiated extensive
research on the stabilization/solidification (S/S) of hazardous
waste as pretreatment prior to ultimate secure landfill disposal.
During 1982-1985, CECOS conducted bench- and pilot-scale
research on the solidification of inorganic waste materials to im-
pact both physical and chemical stability to the wastt. Starting in
1984 similar S/S research was initiated for hazardous organic
wastes as part of the Durachem™ II research program.
During the course of Durachem™ II research conducted during
1986, il became readily apparent that volatile organic compounds
(VOCs) had a severe negative impact on both conventional poz-
zolanic S/S processes and novel operations.' Because many regu-
latory agencies, waste management firms and generators an-
ticipate utilizing such S/S techniques as a pretreatment step prior
to landfill disposal, care regarding VOC content in the waste
matrix must be exercised.
In this paper, CECOS International, Inc. and Recra Environ-
mental, Inc. provide background information on conventional
and novel S/S technologies as well as results on the effect of
VOCs on stabilized products. A variety of waste:reagent mixtures
was investigated, and S/S products were evaluated on the basis of
unconfined compressive strength and leachable total organic car-
bon (L-TOC).

BACKGROUND
Previous studies simulating S/S of organic wastes during Dura-
chem™ II feasibility studies generally involved synthetic mixtures
of organic materials spiked with one or more volatile and soluble
organic chemicals. Correlating the results of these simulations to
applications involving actual hazardous wastes was subject to a
number of variables influencing material handling and feasibility
of technology usage. Additionally, wastes directly incinerable
would likely be excluded from landfill disposal. Factors distin-
guishing waste acceptance include viscosity, chemical and ash
content, heat value, mode of generation and competitive costs.
To compensate for these concerns, actual waste samples were
collected for 1985/1986 Durachem™ II research. In order to
evaluate the effects of each treatment scenario on chemical
mobility, the samples selected were spiked with known concentra-
tions of selected organic solvents. In each case, S/S tests were de-
signed to evaluate: (1) reagent:waste mix; (2) material handling;
and (3) product leachability as compared to New York State Land
Burial Certification Limitations.

Waste Characteristics
The wastestreams selected for study were coded and descrip-
tively summarized below.
Waste C
Waste C was a filter cake generated from a major petroleum
refining industry. The residue characteristically displayed a typi-
cal waste type representing filter media (clayish in appearance)
saturated with oils, grease and select organics. In order to evalu-
ate the effectiveness of various additives in attenuating organic
pollutants, the waste was spiked with solvents—1% Jtylene, 1*
toluene and 0.5% benzene by weight.
152
TREATMENT
-------
Table 1
Waste Characteristics
Waste
Waste E
Waste C
Waste L
Waste P
Total
Appearance Solids (I)
Black Liquid 3?. 7
with Approximately
501 Solids
Black Soft 57.4
Liquid
Red Pasty Solid 39.6
Ammonia Odor
Het Tarry 55.0
Sludge With
Free Liquid
TVO Scan
Flash Heat % 011 and (mq/kq
Ash Pp.'"* Combustion Organic UCS ., Phenol Greasp as
WHqht (°F) (BTU/lb) Halogen (Tons/Ft^) (ug/g) (uq/q) Carbon)
13.5 >200 6760 0.59 Below 74 59.000 1.900
Readable
Limits
37.1 91/85 3745 0.03 0.02 53 148,000 11,000
34.9 >200 527 3.46 0.09 2.14 879 900
36.3 64 5680 0.25 Below 0.54 109,00 16,000
Readable
Limits
Metallics
Cr
19 ppm
Pb
<2 ppm
Cu
2400
ug/g

TVHO Scan"
(mg/kg
as
Chlorine)
9.7
°
1000
540
Comments:
• Analyses were performed in accordance with U.S. EPA methodologies, where applicable.
• Values reported as "less than" ( ) indicate the work detection limit for the particular sample and/or parameter.
• Organic scan procedures are used for screening purposes only.
• UCS — unconfined compressive strength was determined using a Soiltest Pocket Penetrometer.
• Phenol content was determined by distillation with 4-aminoantipyrine.
Waste E
The waste sample represented by Code E was collected from an
equalization basin containing various emulsified and insoluble
solids consisting of various resins, pigments and dyes. The waste
represents various individual wastestreams generated by the
organic chemical industry which manufactures an array of dyes,
pigments and resins. With each waste type, volatile organics are
of prime concern hi addition to an array of other prevalent prior-
ity pollutants. In addition to its high ash and water content, this
sludge contains various solids which are soluble under alkaline
conditions.

Waste P
The waste residue representing Code P was an actual waste col-
lected from an acid treatment process involved with a petroleum
re-refining industry. The material composite was collected from a
number of large tar pits. The waste was a highly viscous, acidic
tar. Besides the apparent concentration of entrapped oils and
grease, this residue was spiked with 5% toluene.

Waste L
The sample identified by Code L represents an inorganic filter
cake spiked with an array of chlorinated solvents. As one of the
major industrial sources of waste tonnage and discharge sources,
the U.S. EPA acknowledges that the electroplating and metal
finishing industries will be impacted in terms of waste manage-
ment requirements. Of considerable impact will be sludge with
residual concentrations of chlorinated and non-chlorinated
solvents in addition to variable concentrations of residual oils and
grease.

Each material was analytically characterized for key parameters
indicative of physical properties and chemical content. The
analyses are reported in Table 1. The parameters were individual-
ly monitored during the course of this study for apparent changes
in physical properties and factors influencing both chemical pro-
perties and pollutant mobility.

Reagent Characterization
The following information is pertinent to the reagents selected
for evaluation and chemical use.
Activated Carbon (AC)
Carbon was utilized in combination with inorganic and organic
additives to assess its impacts on adsorption of selected organics.
Carbon previously had been reported as an adsorptive filler with
certain polymer technologies.

Silicarb
Obtained through the Northeast Industrial Waste Exchange,
samples representative of Silicarb were considered in a multitude
of evaluations as a filler. Silicarb is a by-product of rice hulls
which are burned as an energy source by a unique patented system
leaving a whole hull ash. The main constituent is inert SiO2, rang-
ing from 85-90%.

Rubber Dust
A combined mixture of approximately 70% clay to approx-
imately 20% rubber and residual carbon, this material, made
available through the Northeast Industrial Waste Exchange, was
considered as a filler.

Bentonites
Three types of bentonite were obtained for experimental
evaluation of material usage. Sources included regular commer-
cial grade bentonite, polymer enhanced bentonite (PEB) and an
organophylic bentonite. The regular bentonite was composed of
montmorillonite clay containing sodium cations. The PEB was a
polymer enhanced bentonite that has been used in sanitary and
toxic waste landfills as a fluid barrier. Organophylic bentonites
have various substitutions in place of the sodium cation. Organo-
phylic bentonites can be formulated to retain various con-
taminants by changing the aliphatic ammonium salt. When differ-
ent ammonium salts are used, the spacing in the clay structure is
changed, thereby changing the preferential absorbance of the
bentonite. For high organic concentrations, it was suggested that
a dodeclyammonium bentonite be used because of its ability to
expand the clay layers and expose the silicate surfaces for greater
adsorption.

Other Fillers
Various other fillers were individually evaluated for use in the
Durachemsm II process. Materials given consideration included:
TREATMENT 153
-------
• Fly Ash (FA); sample used after drying
• CaO; reagent grade powder
• Portland Cement (PC)
• Calcium Stearate, reagent grade
• PVC Filings—waste identified for potential preferred absorp-
tion

Polymers
Four polymers were chosen for assessment. These polymers
include:
• Acrylic coating. The acrylic coating obtained for experimental
evaluation was a copolymer of acrylic acid and methacrylic
acid monomers:
CH = C
acrylic acid
.H
COOH
CH2 = c

methacrylic acid
These monomers were polymerized by "Emulsion Polymeriza-
tion." This acrylic coating was chosen because of its ability to
combine with water in the sample and possibly, more effici-
ently, decrease the leachable organic mobility by a greater
ability to encapsulate the waste matrix.
• Water base epoxy. The water base epoxy obtained for experi-
mental evaluation was a polyamide resin. The general struc-
ture of polyamide resins is an alternation of hydrocarbon and
amino groups.
•
,R
CO'
The water base epoxy was chosen for the same reasons as the
acrylic.
• Vinyl ester. The vinyl ester coating obtained for experimental
evaluation was a vinyl ester formation. The vinyl ester was
chosen on the basis of its reported chemical resistivity.
• Polymer cement. The polymer cement obtained for experi-
mental evaluation was an epoxy-aggregate-hardener mixture
that is very resistant to abrasion, impact and chemicals. The
polymer cement was chosen because of this resistance and be-
cause of its own structural integrity.

EXPERIMENTAL DESIGN
Using the waste types described previously, experimental evalu-
ations were designed and implemented with the primary intent of
assessing immobilization performance as a result of direct
chemical usage. Operating variables were incorporated in the pro-
gram in a manner that allowed generation of data pertinent to:

• Direct chemical addition and its impact on physical stability
• Direct chemical addition and its impact on chemical stability
• Polymer application, mode of addition, reagent requirements,
curing requirements and polymer impact on leachable organics
Experimental evaluations were conducted in a sequential order
outlined as follows.

Control Samples—No Treatment
Wastes alone were left untreated and allowed to cure at com-
parative conditions of room temperature (approximately 22 °Q
and static heat (103 °C). Leachable total organic carbon (L-TOQ
and physical characteristics were observed at 0 hr, 24 hr and 168
hr on selected samples, covered and uncovered.

Cementitious Stabilization
Wastes were treated with cementitious reagents and cured at
room temperature (approximately 22 °C). L-TOC and physical
characteristics were observed at 9 hr, 24 hr and 168 hr on selected
samples.

Filler Addition
Wastes were treated with additives and/or filler and cured at
room temperature (approximately 22 °C). L-TOC and physical
characteristics were observed at 0 hr, 24 hr and 168 hr.

Cement /Filler Mixtures
Waste were treated with mixtures of cementitious reagents and
inorganic fillers and cured at room temperature (approximately
22 °C). L-TOC and physical characteristics were observed at 0 hr,
24 hr and 168 hr.

Cement/Filler — Filler/Polymer
Wastes were treated with either cementitious reagent, inorganic
fillers or a mixture of the two and further polymer treated.
L-TOC and physical characteristics were observed at 9 hr, 24 hr
and 168 hr.

Covered vs. Uncovered Curing Conditions
Selected waste treatment scenarios were tested under covered
and open curing conditions. The results were compared to tests
performed at elevated temperatures. L-TOC and physical char-
acteristics were observed at 0 hr, 24 hr and 168 hr.

EXPERIMENTAL RESULTS AND DISCUSSION
The effect of volatile organic compounds (VOCs) on the solidi-
fication/stabilization of hazardous waste can be identified
through an examination of product physical factors affected by
VOCs. These factors include appearance; volume change (densi-
ty); flashpoint; free liquid generation; and load bearing capacity.
VOCs affect chemical factors such as leachable total organic car-
bon (L-TOC) content.

Effects of VOCs on Physical Factors
The effect of VOCs on a S/S product's physical factors are
easily measured over time. The loss of volatiles via evaporation
must be considered when determining an S/S process' effective-
ness. As an example, raw waste samples identified as C, L and P
were evaluated for weight loss as a function of curing under am-
bient, laboratory (22 °Q and elevated conditions (103 "Q. The
results of this analysis are presented in Figs. 1, 2 and 3.
I
I
72 « 1M
Tim* (hm)
Figure 1
Evaporation Rates of P Waste Type
154 TREATMENT
-------
I
72 »e
Time (hra)
Figure 2
Evaporation Rates of C Waste Type
!
O 24 48 72 »« 120 144 1M
Time (hr»)

Figure 3
Evaporation Rates of L Waste Type

The effect of volatiles loss on product strength in the form of
unconfined compressive strength was studied. Raw waste samples
were tested using a pocket penetrometer after volatiles were re-
moved at ambient (22 °C) and elevated (103 °C) temperatures. The
results are summarized in Table 2.

Table 2
Effects of Evaporation on Physical Stability
Temperature! C)
Unconflned CoMOresslve Strength
(ton/ft1)
0 hr 24 hrs 168 hrs
22
22
22
22
22
103
22
103
22
103
22
22
22
22
Air-cured
Air-cured
A1 r-cured
Air-cured
Air-cured
Oven-cured
A1 r-cured
Oven-cured
Air-cured
Oven-cured
Air-cured
Closed to stir.
Air-cured
Closed to a tin.
0.03
0.0?
0.05
O.OZ
0.03
0.03
0.05
0.05
0.02
0.02
0.05
0.05
0.02
0.02
0
1
3
0
0
0
3
4
0
4
3
0
0
0
4
0
0
06
04
04
0
5
oe
s
0
02
08
02
).4
!.0
.5
1.5
).04
1.04
.5
.5
.5
.5
.5
1.02
1.5
5.02
C-103
E-107
L-1H
P-115
C-308
COOS
L-309
1-309
P-307
P-307
L-252
L-252
P-243
P-243
The evaporation of VOCs from a raw waste also has an effect
on the flashpoint of the material. Generally, as VOCs are re-
leased, there is a concurrent increase in flashpoint (Fig. 4).
The use of pozzolans and other additives to increase the
physical stability of a hazardous waste is a common technique.1
Typically, low product strengths are observed immediately after
reagent addition/mixing, with a subsequent increase in strength as
a result of CaO hydration and curing. The effect of evaporating
VOCs as the mechanism for increasing unconfined compressive
strength compared to curing the reagents can be questioned. A'
variety of S/S additives were studied under uncovered (22 °C) and
24 48 72 96 120 144 163
Time (hrs)

Figure 4
Effect of Devolatilization on Flashpoint for P Waste Type
covered 22 °C) conditions to determine if VOC loss (drying) was a
major mechanism for S/S. The results of this research may be
found in Table 3.

Table 3
The Effect of Evaporation on Conventional S/S
Processes Unconfined Compressive Strength
Uiste 10
P-168
P-169
P-170
L-172
L-173
1-174
L-213
L-214
P-223
P-163
»rt(«
60
60
ec
60
60
60
60
60
60
60
S/S Additive
PC/Bent
PC/Bent
PC/Bent
PC/Bent
PC/Bent
PC/Bent
FA
FA
FA
PC
. «Jal
5/1
5/2.5
1/5
1/2
1/3
1/5
10
20
10
5
Unconflned Compressive Strength
ten/ft'
Conditions 0 hr 24 hrs 168 hrs
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
0.11
0.11
0.06
0.06
0.12
0.12
0.06
0.06
0.09
0.09
0.17
0.17
0.19
0.19
0.75
0.75
0.13
0.13
0.08
0.08
0.40
0.13
0.35
0.09
0.45
0.13
3.5
0.11
4.5
0.50
4.5
0.65
4.5
0.19
4.5
1.0
0.40
0.19
0.75
0.13
1.30
0.25
1.50
0.13
1.75
0.13
4.5
0.50
4.5
0.40
4.5
1.25
4.5
0.25
4.5
1.25
1.50
0.30
1.75
0.13
Similar results were seen in tests on novel S/S processes, as
identified in Table 4.
Effects of VOCs on Chemical Factors
The effects of VOCs on pollutant mobility were examined for a
variety of wastes and reagents. Specifically, raw wastes and
wastes solidified with conventional S/S reagents were studied at
ambient temperature both covered and uncovered. Table 5 pro-
vides the results of a comparative study for leachable TOC
(L-TOC).
The supplementation of conventional S/S reagent with poly-
mers (acrylic, water base epoxy, vinyl ester and polymer cement)
also was studied. It was hoped that the network structure of poly-
meric coatings would assist in the attenuation of pollutants. In
reality, the tests results showed that the polymers contained either
residual concentrations of solvent or unreacted monomers which
appeared to contribute to high L-TOC values even after suggested
full cure times. This result is consistent with previously published
TREATMENT
155
-------
Table 4
The Effect of Evaporation on Novel S/S Processes Unconfined Compressive Strength
Waste ID
L-245
L-245
1-246
1-247
L-248
L-249
L-250
1-251
P-253
P-254
P-255
wt(g)
100
100
100
100
100
100
100
100
100
100
100
S/S Reagent
FA
FA
FA
FA
FA
FA
FA
FA
CaO/FA
CaO/FA
CaO/FA
S/S Reagent wt(g) Additive wt(g)

20 Acrylic 15
Conditions
Unconfined Compressive Strength
ton/ft?
0 hr. 24 hr. 168 hr.
20
30
30
Acrylic 30
Acrylic 50
Acrylic 70
20 Polymer Cement 15
?0 Polymer Cement 30
20 Polymer Cement 50
20 Polymer Cement 70
ACD
ACD
Table 5
The Effect of Evaporation on Pollutant Mobility
for Conventional S/S Techniques
«rt(nl Conditions
L-TOC
0 hr. 24 hr.
t-103
E-107
L-lll
P-115
1-252
P-243
1-112
1-213
1-214
C-156
C-157
C-1M
60
60
60
60
50
50
60
60
60
60
60
60

-
-
PC

PC/0-Bent

Ft/0- Sent

PC/0-Bent

-
-
-

-
1

1/1

1/3

1/5

Uncovered
Uncovered
Uncovered
Uncovered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
7,500
15,000
3.000
3.500
N/A
N/A
It/A
N/A
N/A
N/A
H/A
N/A
N/A
N/A
N/A
N/A
»/A
»/A
N/A
N/A
2,100
9,000
1,500
2,750
3,000
5.000
5.000
2,750
1,000
1,700
700
1,700
670
1,200
517
1,705
262
1,575
616
1,540
650
1.000
750
11,500
1,750
4,700
4,700
2,000
500
1,950
700
1,400
780
1,300
475
1,261
448
1.227
330
1,210
Table 6
Polymer Control Data at 22 °C
rolyner
Acrylic
Vinyl Ester
Poljner Ccacnt
Utter Itsed Epoiy
Conditions
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
0 hr
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
L-TOC (.g/ks)
74 hr 168 hr
175,000 N/A
430,000 250,000
250 750
400 210
100 700
500 700
215.000 275.000
280,000 N/A
3.3
8.3
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
0.02
0.02
0.02
0.02
0.02
0.02
0.03
0.03
0.05
0.05
0.05
0.05
0.02
0.02
0.02
0.02
0.08
0.08
0.08
0.08
0.04
0.04
4.5
0.06
4.0
0.05
3.25
0.08
1.75
0.09
4.25
0.13
4.5
0.13
3.25
0.19
3.5
0.28
0.15
0.17
0.09
0.22
0.6
0.19
4.5
0.11
4.5
0.09
4.5
0.09
4.0
0.13
4.5
0.25
4.5
0.25
3.25
0.25
3.5
0.35
2.25
0.20
2.25
0.22
2.30
0.22
research.1 Control data for the polymers used are presented in
Table 6, while Table 7 provides L-TOC data for wastes stabilized
with novel S/S techniques.
It was readily apparent that, at 22 °C, the addition of polymers
to the S/S reagent mix offered no apparent attenuation of
leachable organics and, in fact, added considerably to the L-TOC
content of the waste. Any effect of VOCs on the L-TOC content
was lost in the contribution seen from the polymers.

CONCLUSIONS
It was evident from the tests performed that volatile organic
compounds adversely impact the stabilization/solidification of
hazardous waste. Both physical and chemical factors showed a
degradation in various S/S processes tested when VOCs were not
allowed to evaporate from the product. To produce a consistent
quality product suitable for secure chemical landfill disposal, it is
necessary to remove the solvent fraction from the waste matrix
prior to, or concurrent with, solidification/stabilization reagent
addition.
The addition of polymers as an aid to reduce pollutant mobility
must be more carefully studied. The leachable organics supplied
by residual solvents and/or unreacted monomers clearly obscured
the identification of any benefit realized. Detailed GC/MS data
for specific solvents must be generated to clearly identify if raw
waste pollutants are leached or if the L-TOC content is emanating
from the polymers themselves.

REFERENCES
1. Durachem II, Report Summarizing Feasibility of Technology Usafl,
Recra Environmental, Inc., Report No. 5C002480, 1986, 26.
156
TREATMENT
-------
Table 7
The Effect of Evaporation on Pollutant Mobility for Novel S/S Techniques
Conditions
Waste ID
P-235
P-236
P-239
P-240
L-244
L-245
L-246
L-247
vrt(g)
100
100
100
100
100
100
100
100
S/S Reagent
FA
FA
FA
FA
FA
FA
FA
FA
wt(g)
20
20
40
50
20
20
20
20
Additive
Acrylic
Acrylic
Polymer Cement
Polymer Cement
Acrylic
Acrylic
Acrylic
Acrylic
Wt(l
15
30
15
30
15
30
50
70
L-TOC (mg/kg)
0 hr. 24 hr. 168 hr.
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
Uncovered
Covered
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
45,255
48,600
52,500
75,000
18,600
33,325
74,700
2,340
28,350
25,650
28,500
32,250
59,400
57,600
92,400
88,200
87,750
31,050
112,500
135,000
11,625
13,950
8,820
10,260
19,575
19,575
16,125
30,000
49,500
47,700
84,000
68,250
2. Chalasoni, D., et al., "The Effects of Ethylene Glycol on a Cement-
Based Solidification Process," Hazardous Waste and Hazardous
Materials, 3, 1986, 168.
3. Conner, J.R., "Fixation and Solidification of Wastes," Chem. Eng.,
93, No. 21, 1986, 80.
TREATMENT 157
-------
Treatment Technologies for Dioxin Wastes
M. Pat Esposito
PEI Associates, Inc.
Cincinnati, Ohio
ABSTRACT
The results of research conducted over the past 10 years to iden-
tify potential treatment methods for dioxins and other persistent
organochlorine compounds are discussed. Both thermal and non-
thermal technologies are reviewed, including: stationary source
and mobile rotary kiln incineration, liquid injection incinera-
tion, fluidized bed incineration, infrared incineration, high
temperature fluid wall destruction, plasma arc pyrolysis, molten
salt destruction, in situ vitrification, super-critical water oxida-
tion, chemical degradation, ultraviolet photolysis, solvent extrac-
tion, biodegradation, stabilization/fixation, gamma ray radioly-
sis, carbon adsorption and land treatment.
Several of the thermal treatment technologies have demon-
strated 99.9999"% destruction and removal efficiency (DRE) on
chlorinated compounds of various types such as PCBs and car-
bon tetrachloride, but reports have been published with data
fully demonstrating this level of performance on dioxins for only
three systems: mobile rotary kiln incineration, high temperature
fluid-wall destruction and infrared incineration. Supercritical
water oxidation also has been reported to have achieved
99.9999"% DRE on dioxins, but data supporting this claim have
not yet been released. Tests with dioxin wastes at the U.S. EPA
stationary rotary kiln incinerator in Arkansas came very close to
demonstrating 99.9997 DRE, but fell short due to the low con-
centration of dioxin in the waste feed (37 ppb) and the analytical
detection limits of the analytical methods; both of these factors
worked together to prevent the calculation of the DRE. Of the
non-thermal methods, only two—thermal desorption followed by
UV photolysis, and chemical dechlorination—have been capable
of achieving residual levels of 1 ppb or less dioxin in the treated
waste or soil.
Technologies considered effective or which show strong poten-
tial, include:

Thermal Technologies
• Stationary rotary kiln incineration
• Mobile rotary kun incineration
• Liquid injection incineration
• Infrared incineration
• Fluidized-/circulating-bed incineration
• High-temperature fluid wall destruction
• Plasma arc pyrolysis
• Molten salt destruction
• Supercritical water oxidation
• In situ vitrification

Non-thermal technologies
• Chemical degradation
• Ultraviolet (UV) photolysis
• Solvent extraction
• Biodegradation
• Stabilization/fixation
• Gamma ray radiolysis
• Carbon adsorption
• Land treatment

BACKGROUND
Dibenzo-p-dioxins are a family of aromatic compounds com-
monly referred to in the literature as dioxins.1 Each dioxin com-
pound has as a nucleus a triple-ring structure consisting of two
benzene rings interconnected to each other through a pair of
oxygen atoms. Fig. 1 shows the structural formula for the baac
dioxin nucleus.
INTRODUCTION
Research conducted in recent years in search of treatment tech-
nologies for destroying or detoxifying dioxins and other chem-
ically similar compounds such as PCBs or chlorinated benzenes
has been considerable. The results of this research are reviewed
in this paper. In some instances where relatively new technolo-
gies are involved, basic process engineering and design principles
are also presented.
Both thermal and nonthermal technologies for treating dioxin
wastes have been studied, and several are promising. The ability
of a thermal treatment method to achieve a destruction and re-
moval efficiency (DRE) of 99.9999% (often referred to as "six
nines") is the main criterion used to evaluate these technologies.
For non-thermal methods, a reduction of the dioxin content in
the waste to 1 ppb or less usually is required to consider the waste
residual safe and to consider the method effective.
Figure 1
The Dioxin Nucleus
Many types of dioxin compounds are possible. The eight outer
ring positions can be occupied by hydrogen ions (the unaubiu-
tuted form), or the molecule can be partially or fully substituted
wherein one or more of the hydrogens are replaced by such con-
stituents as chlorine or other halogens, amine groups, hyuroxjl
158
TREATMENT
-------
groups, methyl groups or any combination of these. Although
considerable knowledge has been generated relative to the phys-
ical, chemical and toxicologic properties of the chlorinated
dioxins, little is known about the other forms.
Dioxin compounds apparently are formed as byproducts of cer-
tain chemical manufacturing and combustion processes involving
precursor compounds and heat. They usually are formed in very
low yield (i.e., in the ppt to ppm range) but, once formed, they are
quite stable. For example, in the 1960s and 1970s, chlorinated di-
oxins were inadvertently generated as byproducts of the mass pro-
duction of trichlorophenol and its germane and pesticide deriva-
tives (e.g., hexachlorophene, 2,4,5-T, silver and erbon). Dioxin
residuals from this period can be found in the environment today,
contaminating soils and process wastes associated with the pro-
duction and widespread use of those chemicals.
The tetra-, penta- and hexachlorinated dioxins have been the
subject of most research to date because of concerns over their
potential health effects. The tetrachlorinated isomer, 2,3,7,8-
tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) has been of particu-
lar concern worldwide because it is one of the most toxic man-
made compounds known. The pure compound is a colorless
crystalline solid at room temperature. It is soluble in the 10-100
mg/1 range hi most organic solvents, but its solubility in water is
only 2 jig/1. Its natural degradation rate appears to be very slow
in comparison with almost any other organic compound. It has a
very low vapor pressure and does not readily evaporate or volatil-
ize to the atmosphere at ambient temperatures. It can be photo-
lytically dechlorinated when exposed directly to ultraviolet radia-
tion. Recent research has shown that the compound can be either
formed or destroyed (or some combination of both) in combus-
tion devices such as incinerators, depending on the temperature
and other operating conditions (e.g., waste feed composition,
feed rate and amount of excess air). Once released into the en-
vironment, the compound adheres tightly to soil and other partic-
ulate matter from which it does not migrate or leach into ground
or surface waters over time unless the contaminated particles
themselves migrate via erosion processes. Thus, wastes and soils
contaminated with 2,3,7,8-TCDD and other dioxins tend to be
persistent environmental problems for which there are no simple
remedies.

THERMAL TREATMENT TECHNOLOGIES
Stationary Rotary Kiln Incineration
Three commercial rotary kiln incinerators have demonstrated
99.9999% DRE for PCB-containing wastes and have been per-
mitted to burn these wastes; as such, these units have the poten-
tial to effectively burn dioxin wastes. These units are the Rollins
incinerator in Deer Park, Texas, the SCA incinerator in Chicago,
Illinois and the ENSCO incinerator in El Dorado, Arkansas.
None of these units, however, has been tested using dioxin wastes.
Only the U.S. EPA Combustion Research Facility (CRF) in
Jefferson, Arkansas, which operates a stationary rotary kiln in-
cinerator for research purposes, has conducted test burns of
dioxin wastes.2-3'4-5
Four trial burns were performed in 1985 at the CRF incinera-
tor using dioxin-contaminated toluene still bottoms from the
manufacture of chlorophenols and phenoxy herbicides. In all,
about 600 Ib of wastes were incinerated. The burns included a
blank burn to establish background emission levels, a short-
duration (4-hr) burn to establish feed capabilities and to test the
sampling protocol and two full waste burns of 6 and 12 hr dura-
tion to establish DREs for the dioxin. The kiln was operated at
1796F (980 °C) to 1814F (990 °C) with a nominal residence time of
4.9 to 6.0 sec, and the afterburner was operated at 3686F
(2030 °C) with a residence time of 1.8 to 2.3 sec. The dioxin con-
centration in the waste feed was only 37 ppb; feed rates varied
from22to391b/hr.
Problems in the waste feed system and in the emission moni-
toring program were encountered which tend to compromise the
test results to a certain extent. Nevertheless, the results of the two
full waste burns show that the 2,3,7,8-TCDD DRE was greater
than 99.9997%. Unfortunately, six nines DRE could not be firm-
ly established because the detection limits of the sampling and
analysis protocols used were not low enough to make the calcula-
tion. If the waste feed dioxin concentration had been higher, the
calculation and demonstration of six nines DRE very likely would
have been accomplished. Other trial burn tests have shown that it
is very difficult to show even four nines DRE (99.99%) for com-
pounds present in the waste feed in concentrations less than 100
ppm. With this in mind, the investigators concluded that rotary
kiln incineration under the conditions existing during the CRF
trial burns is capable of achieving 99.9999% dioxin DRE.
The concentrations of 2,3,7,8-TCDD and other dioxins were
also measured in the scrubber blowdown water and kiln ash. The
maximum concentration of 2,3,7,8-TCDD found in the blow-
down water was 0.1 pg/1; octochlorodioxin was found at 0.8 pg/1.
No dioxins were detectable in the kiln ash at detection limits rang-
ing from 1 to 37 ppt. These results support the conclusion that a
high degree of dioxin destruction occurred in the incinerator.

Mobile Rotary Kiln Incineration
Two mobile rotary kiln incinerators—one built by U.S. EPA
contractors and the other built and owned by ENSCO—cur-
rently are available to destroy hazardous wastes. The U.S. EPA
mobile incinerator was recently tested using dioxin-contaminated
soils and liquid wastes.6-7 Average operating parameters during
the trial burn were as follows:
Kiln temperature
Secondary combustion chamber
Secondary combustion chamber
gas flow rate
Secondary combustion chamber
residence time
Waste feed rate (solids)
Waste feed rate (liquids)
Auxiliary fuel
Kiln
Secondary combustion chamber
1596°F(869°C)
2167°F(1,186°C)

13,500 actual ft'/min

2.6 sec
l,3001b/hr
2401b/hr

5 to 6 x 106 Btu/hr
4 to 5 X 10« Btu/hr
As Table 1 shows, the trial burn was successful; DREs exceeded
99.9999% at feed rates of up to 2,000 Ib/hr solids and 250 Ib/
hr liquids. During the tests, 3.84 Ib of 2,3,7,8-TCDD contained
in 1,750 gal of liquids and more than 40 tons of soils were de-
stroyed. Particulate emission permit limitations (less than 180
mg/nm3 at 7% ©2) were achieved in three of four test runs. Dur-
ing the fourth run, particulate emissions slightly exceeded the pre-
scribed limit, possibly due to the accumulation of submicron-
sized particles in the air pollution control system. The observed
CO emission values (1.3 to 7.7 ppm) are equivalent to those from
the best available incineration technologies and are indicative of
very complete combustion (C.E. = 99.993 to 99.9999%).'
The treated soil (ash) and process wastewater from the trial
burn were analyzed for a series of specific constituents considered
as likely contaminants. The results of these analyses were used
to support an application to "delist" residues from a planned
larger-scale burn of similar wastes.
The ENSCO mobile rotary kiln incineration system is similar
in design to the U.S. EPA mobile incinerator.8-9 Four mobile
units of varying capacities have been constructed. One of the
TREATMENT
159
-------
earlier and smaller units successfully demonstrated six nines DRE
destruction of PCBs in 1983.'" One of the larger units recently
completed a series of trial burns in Arkansas using dioxin-con-
taining wastes. The results of these tests have not yet been re-
leased by ENSCO.
Table 1
Results of Dloxln Trial Burns In Missouri6-7

*rn!
tot. ttaa (we)
0* t«l. it
2,ta ii
C.t. (1)
•• ^*l»)
U*t« MIU ft**
T*Ul flM (*/•*•)
rcn (p^)

TtUl np> (»/hr)
TCOO JM»)
luck
t-l"i« »•«:
IdD^ (N|/4IJ>)
(•I^B'tTI 0>)
onTi)
"SSI i-r-
TIO-1MO MS
1170*1315 US!
2.S
»4 7.M
• 100 (* •*») T.7
10. »
99. Ml
0.1
1»
Z13
*400 ?*»

.»oo lisa
101
o.os

•0.1M
1W 1M.3
TT~
•71
1IM
I.I
*.*!
1.3
11.3
W.W
O.S
IS
ZM
1ST
M.I
110
982
0.24

•0.127
14T.3
1-4
KB
11*4
2.S
C.44
2.3
11.4
O.S
m
234
2«4
n.o
20U
1010
O.K

(O.O1
m,«
2-S 3-1.-
»ia a
in i
.t
.42 5
.5 1
1.1
0.9
1H
247
22S
K.t
im 9
770
0.4C

4.0M
201. S
.-J*
W
87

TCDO - 2,3,7.8-TCDD; C.E. = combustion efficiency; DRE = destruction and removal
efficiency; TH - total hydrocarbons.
1 Test 2 liquid feed: dioxin-contaminaled TCP still bottoms
solids feed: TCP still bottoms and contaminated soil
b Tell 3 liquid feed: fuel oil; solids feed: dioxin-contaminaled/brominated naphthalene-
conuuninated lagoon sludge; combined results for three one-half hour runs.
c No TCDD detected at detection limits in effect; hence, emission rates denoted are detection
limits and nol measured quantities.

Liquid Injection Incineration
Liquid injection is a technology used daily throughout the
U.S., both at industrial locations and at central treatment facil-
ities. It is the most commonly used incinerator for hazardous
waste destruction, comprising more than 60% of the incinerators
used today.'' These facilities have been used to destroy a variety
of wastes including phenols, CPBs, still and reactor bottoms,
solvents, polymer wastes, herbicides and pesticides. They are not
recommended for burning heavy metals, high-moisture-content
wastes or materials with high inorganic content.
The incinerators aboard the M/T Vulcanus I and II, incinera-
tor ships designed to burn liquid chemical wastes in the open
ocean, are liquid injection units. Each incinerator is capable of
burning 1,650 gal/hr of liquid wastes.1J In July-September 1977
the liquid injection incinerators aboard the Vulcanus I were used
to destroy more than 11,000 tons of Herbicide Orange, a liquid
herbicide that contained several chlorinated substances includ-
ing TCDDs. During these tests, the flame temperature was
2506 T (1375 "Q to 2930 °F (1610 °C), the furnace wall tempera-
ture was 2012°F (1100°C) to 2192 °F (1200 °Q, and the residence
time was 1 to 2 sec.
The reported destruction efficiency for the TCDD was greater
than 99.93%, which was the highest that could be measured given
the detection limits of the analytical method used at that time to
measure the TCDD concentration in the exhaust gas. More re-
cent test burns in similar liquid injection units on the sister ship
Vulcanus II, demonstrated greater than 99.998% destruction
efficiency for the organic chlorine compounds CC14, CMC 13
and 1.1,2-TCE." The U.S. EPA concluded from its evaluation
of these test burns that ocean incineration has the potential to be
an environmentally acceptable alternative to other means of dis-
posal.
Although incinerator ships currently are available for use, they
are neither TSCA- nor RCRA-permitted in the United States.
Further testing and specific permits will be required before any
operation off the shores of this country can occur.

Infrared Incineration"
Shirco Infrared Systems, Inc., has developed a system to ther-
mally treat wastes by metering the waste streams onto a woven
metal alloy conveyor belt and passing the waste under infrared
heating elements and then burning combustible off-gases in a
secondary chamber. The unit currently being tested by Shirco is a
portable pilot-scale system that is contained in a 45-ft trailer.
Shirco has reported successful use of the system to treat mu-
nicipal and industrial sludges and a simulated creosote pit waste.
During the week of July 8, 1985, the unit was successfully used to
obtain "> 99.9999% DREs from dioxin-contaminated soil at
Times Beach, Missouri. A summary of the soil decontamination
test results is shown in Table 2.

Fluidized-/Circulating-Bed Incineration
Fluidized-/circulating-bed incineration combustors have been
used to treat municipal wastewater treatment plant sludges, oil
refinery wastes, pulp and paper mill wastes, pharmaceutical
wastes, phenolic wastes and methyl methacrylate. Currently,
there are more than 25 units operating in the U.S. and Europe,
although none are operating commercially as hazardous waste
incinerators.
Dioxin-contaminated wastes have not been tested in a fluidized-
or circulating-bed incinerator. However, results with other chlor-
inated organics, including PCBs, suggest that the unit has good
potential to destroy dioxins.
A low-temperature (1562°F or 850 °C) fluidized-bed combus-
tor designed by Waste Tech Services, Inc., was used to conduct
test bums in 1985 on soil contaminated with carbon tetra-
chloride and dichloroethane." Results of the tests, as measured
by the VOST method, indicate greater than 99.99% destruction
of both contaminants.
GA Technologies has conducted test burns on its pilot-Kale
circulating-bed combustor using various chlorinated organic
liquid wastes. At 1540°F (838°C) to 1600°F (870X5, greater
than 99.99% DRE was achieved for ethyl benzene, vinyl chloride,
toluene, benzene and several chlorinated ethanes and ethyl-
enes.
15,16
Table 2
Dloxln Contaminated Soil Decontamination Tesl Result!—
Shirco System13

Solid phase residence tine (*1n)
TCDD 1n feed (ng/g)
tolsslons sanpllng duration (h)
Paniculate* at 7% 0, (gr/dscf)
Gas phase DRE of 2,3,7,8 TCOO
Detection Halt (plcograns)
Ash analysis for 2,3.7,8 TCDD
Detection Hilt (ppb)
Test 1
30
227
7
0.0010
>99.999996b
14
0.038
Tfitt
IS
1S6
2.5
0.0002
£.*»
IB
0.033
* Paniculate filter only, without train rinse.
b b > indk-mo DRE calculations at detection limit.
c NO indicates nondelecublt.
160 TREATMENT
-------
The GA pilot-scale unit was also recently used to conduct a test
burn on soil contaminated with more than 10,000 ppm PCB. At
an average operating temperature of 1800 °F (982 °C), greater than
six nines DRE performance was achieved in all three tests. Results
are summarized in Table 3. Although dioxins were present in ppb
quantities in the waste feed and are possible products of incom-
plete combustion of PCBs, none were detected in the incinerator
ash.
Table 3
PCB Trial Burn Results—GA Technologies
Circulating Combustor15-16
Parameter
Soil feed rate, Ib/h
Total soil fed. Ib
PCB concentration 1n feed, ppm
Combustion efficiency, X
Residence tin, sec
Destruction temperature, *F
DRE, %
Participate concentration
Dry, gr/dscf
Wet, gr/acf
Test 1
327.5
1,310
11.000
99.94
1.18
1,805
>99.9999
N/R
N/R
Test 2
411.5
1,646
12.000
99.95
1.18
1.805
>99.9999
0.0425
0.0227
Test 3
323.8
1.295
9.800
99.97
1.22
1.795
>99.9999
0.0024
0.0013
Table 4
Plasma Arc Test Results Using Liquid PCB Wastes20
Run 1
Run 2
Run 3
Reactor Operating Temperature
PCB Concentration In Haste Feed
Stack Gas Parameters
Total PCB,"
g/dso»D
Total dioxins,
g/dson
Total furans,
g/dscn
Total BaP,
g/dscn
1136°C
14.11
0.013
0.013
0.076C
0.26
0.18
-
0.46
0.32
0.43
1.66
0.45
-
3.0
0.011
0.13
0.30
2.8
Scrubber Effluent Parameters

Total PCB, ppb a 1.56 2.15 9.4
b 0.06 4.7 0.01

Total dioxins, ppt 5.8 259 1.35

Total furans, ppt 1.5 399 1.35

Total BaP, mg/L 0.04 0.92 2.0

Destruction Removal Efficiency

PCB. Percent DRE
a >99.99999 99.99994 >99.9999
b 99.999999 99.99997 99.999999

a These values are based upon mono-decachlorobiophenyl.
° These values are based upon tri-decachlorobipheny!.
c No tetra- or penta-chlorinated dioxins were detected at 0.05 ng on a GL column, except for
run #1 where 0.06 ng tegta-chlorinated dioxin was reported.

High-Temperature Fluid Wall Destruction-
Advanced Electric Reactor
The Advanced Electric Reactor (AER) is a process designed
specifically for on-site detoxification of soils. It is owned by the
J.M. Huber Company of Borger, Texas. Among the potential
advantages of this process are transportability, extremely high
treatment efficiencies (because of the high process temperatures
and long residence times), intrinsic safety features (such as the
activated carbon beds, electrically driven solids feeder and a large
amount of thermal inertia in the reactor) and the ability to de-
toxify wastes in a pyrolytic atmosphere, thereby minimizing par-
ticulate emissions.
Huber has reported several relevant tests that have been carried
out over the past 2 years. A test in the spring of 1984 to deter-
mine the destruction efficiency and DRE for the process with
carbon tetrachloride over a wide range of operating conditions
demonstrated greater than six nines DRE.14 A September 1983
test on PCB-contaminated solids demonstrated better than seven
nines DRE and resulted in certification of Huber's facility at
Borger, Texas, to destroy PCB-contaminated solids."
In November 1984 a mobile version of the Huber unit was used
to treat soil containing 2,3,7,8-TCDD in Times Beach, Missouri.
Reports of this test state that no 2,3,7,8-TCDD was detected in
the treated soil at detection limits of 0.11 ppb and that no
2,3,7,8-TCDD was detected in the stack gas at detection limits of
0.55 ng/m3.18 Greater than 99.999% DRE was demonstrated,
but six nines DRE could not be confirmed due to the inability
of instrument detection limits to compensate for the law quantity
of TCDD in the soil (79 ppb).
The process' most recent test was at a U.S. Navy facility in
Mississippi in which 1,000 Ib of soil contaminated with Agent
Orange were pyrolyzed." Although testing was completed in
June 1985, results are not yet available.

Plasma Arc Pyrolysis
In this process, waste molecules are destroyed by the action of
a thermal plasma field generated by passing an electric charge
through a low-pressure air stream, thereby ionizing the gas mole-
cules and generating temperatures of up to 18,032 °F (10,000°C).
The system is mobile and can treat liquids, but is not suitable for
solids and sludges. The dioxin wastes for which this technology
has potential include nonaqueous-phase leachate such as that gen-
erated at the Love Canal and Hyde Park Landfills, unused liquid
herbicide solutions and, possibly, toluene stillbottoms from herb-
icide production.
The system has been tested on carbon tetrachloride and PCBs
and in both cases achieved DREs of six nines or better. The sys-
tem has not been tested on wastes contaminated with dioxins;
however, because of the structural similarity between PCBs and
dioxins, the data suggest this method may have potential to de-
stroy dioxins at a DRE of 99.9999%. Table 4 summarizes data
from the PCB tests.

Molten Salt Destruction
The molten salt process has been tested at bench and pilot
scales for its ability to treat and destroy a wide variety of wastes.
Bench-scale units have been used to destroy chemical warfare
agents at efficiencies of 99.9999 to 99.999999%. Other chemicals
that have been destroyed using the bench- or pilot-scale process
include chlordane, malathion, sevin, DDT, 2,4-D, tar, chloro-
form, hexachlorobenzene and PCBs.21'22'23 In the PCB tests,
99.99995% destruction was achieved at 1300°F (704°C). Hex-
achlorobenzene destruction efficiencies ranged from nine nines
to eleven nines DRE and chlordane from seven nines to eight
nines DRE in a pilot-scale system at temperatures from 1685 °F
(918 °C) to 1805 T (985 °C) and 2 seconds residence. In smaller-
scale experiments, simulated pesticide wastes containing 0.1 to
10% 2,3,7,8-TCDD by weight were treated; the results only
TREATMENT
161
-------
demonstrated destruction efficiencies of 99.96% at 1470 °F
(799°C) and 99.98% at 2190°F (1199°C).24
Because molten salt combustion has not been developed to full
scale, it cannot be considered a practical method for waste treat-
ment at time. To date, only bench-scale combustors with waste
feed capacity ratings of up to 10 Ib/hr and a pilot-plant-scale
unit capable of processing 250 Ib/hr of wastes have been con-
structed by Rockwell, and a mobile unit referred to as MOSED is
under development by the State of New Jersey and Questex Cor-
poration." Although the process has potential to treat dioxin and
other toxic wastes, additional testing and development are re-
quired.

Supercritical Water Oxidation
This process utilizes the properties of water at pressures greater
than 218 atmospheres combined with temperatures greater than
705 °F (374°C) to effect oxidation of organics such as TCDD.
Bench- and pilot-scale supercritical water reactors have been used
by Modar, Inc., to treat dioxin-containing wastes; both report-
edly have achieved greater than six nines DRE." Detailed results
of these studies have not yet been released by Modar.

In Situ Vitrification
The basic principle of the in situ vitrification process is the
passage of an electrical current through contaminated soil to
create temperatures of 2462°F (1350°C) or more, thereby melt-
ing the soil and forming a stabilized/immobilized molten glass or
crystalline mass. Organics in the soil tend to either pyrolize or
rise to the surface of the molten mass where they combust upon
contact with the air. Noncombustibles such as metals tend to re-
main in the molten mass, becoming trapped or immobilized in
the crystalline structure as it cools.
PCB-contaminated soils were tested recently by Battelle for
EPR1 using this process, and preliminary results suggest achieve-
ment of six to nine nines DRE." Although no tests of dioxin
wastes have been performed, it appears that this process has po-
tential for both the destruction and long-term immobilization of
dioxins in contaminated soils.

NONTHERMAL TECHNOLOGIES
Chemical Degradation
Chemical degradation processes use special chemical reagents
such as metallic sodium, alkaline polyethylene glycols, chloro-
iodides and strong oxidizing agents to alter the basic structures of
problem organochlorine compounds. Some reagents strip one or
more chlorine atoms from the target compound, leaving the basic
molecular ring structure intact, whereas others cleave carbon-
carbon (C-C) or ether (C-O-C) linkages.
Metallic sodium used alone or in conjunction with other pro-
prietary reagents is most often used to dechlorinate chemicals.
Although most dechlorination research has been aimed at the de-
toxification of PCBs, this research is applicable to many chlor-
inated organic molecules, including 2,3,7,8-TCDD. Processes
utilizing metallic sodium are somewhat limited in applications be-
cause they can only be used to detoxify organochlorine com-
pounds in nonaqueous wastes.
Several companies and research institutions have sodium-based
dechlorination processes. Three examples are the Acurex, PPM
and Sunohio PCBX processes. These processes have effectively
reduced the toxicity of PCBs and other chlorinated organics, in-
cluding dioxins. For example, tests conducted in 1980 proved that
the Acurex technology could reduce the PCB content of waste oil
from 1,000-10,000 ppm to less than 1 ppm.11 More recent tests
show that the 2,3,7,8-TCDD content of transformer oil can be re-
duced from 200-400 ppt to 40 ±20 ppt using the Acurex pro-
cess." The Sunohio PCBX process has reduced the PCB content
of contaminated transformer oil from 225 ppm to 1 ppm in one
pass through the system. It is believed that passing the oil through
the system three times can reduce PCBs from 3000 ppm to less
than 2 ppm. This process was used at a number of sites, includ-
ing: Maxwell Laboratories in San Diego, where 163,000 gal of
PCB-contaminated oil were processed; and Chevron in El Segun-
do, California, where a total of 5250 gal of oil initially containing
420 to 1500 ppm PCBs were processed.29'30 The PPM process has
reduced the PCB concentration in contaminated oil from 200
ppm to below detection levels.29
The U.S. EPA's Hazardous Waste Engineering Research
Laboratory (HWERL) has supported research on another type of
chemical dechlorination system using alkaline polyethylene gly-
cols (APEGs) to destroy dioxins.32 Laboratory and field experi-
ments using TCDD have demonstrated the ability of APEG re-
agents to chemically reduce concentrations of that isomer in soils
and wastes.33
Ruthenium tetroxide is a powerful oxidizing agent that is more
effective than either hypochlorite or permanganate in achieving
aromatic ring cleavage. Several reports of its usefulness in treat-
ing dioxins have been published.34'35'36 In studies of soils con-
taminated with Herbicide Orange and containing 70 ppb of
TCDD, application of ruthenium oxide to CC14 extracts of
dioxin from the soil showed rapid degradation of the dioxin in
1 hr at 171T (76 °C). Another experiment in which the CC14 and
ruthenium oxide were added to the soil showed similar results.
This method could potentially be used for decontaminating glass-
ware or industrial process equipment.
Experiments illustrating the cleavage of either linkages (C-O-Q
through the use of surfactants containing chloroiodides have been
performed on substances such as xanthene, benzofuran and
2,3,7,8-TCDD." In all cases, chloroiodides aided in the decom-
position of the test substances. For example, in one study, solu-
tions containing 2,3,7,8-TCDD in benzene were vacuum evap-
orated and the residues were treated with aqueous surfactant
solutions. Two chloroiodide derivatives were used in the surfac-
tant solutions: benzalkonium chloroiodide and cetylpyridinium
chloroiodide. When benzalkonium chloroiodide was used, a 71%
decomposition of 2,3,7,8-TCDD was observed. When cetylpyrid-
inum chloroiodide was used, a 92% decomposition of 2,3,7,8-
TCDD was achieved. Reaction products included chlorophenols,
phenols and 2-phenoxychlorophenols. The results were obtained
under ideal conditions, so extrapolations to actual decontamina-
tion efforts should be made with great care.
Contaminated soil samples from Seveso, Italy, also were
treated by this method. Samples were prepared by treating the soil
with solutions containing surfactant micelles with and without
chloroiodides. A benzalkonium chloride micellar solution showed
approximately 1 14% decomposition of 2,3,7,8-TCDD. A solu-
tion containing benzalkonium chloroiodide in a micellar solu-
tion showed a decomposition of 52% of 2,3,7,8-TCDD. Thus, the
addition of chloroiodides to micellar surfactant solutions greatly
enhances the decomposition of 2,3,7,8-TCDD.
The potential for the use of micellar surfactant solutions
(coupled with a UV source to enhance decomposition) may have
application in decontaminating surfaces of buildings, furniture
and other personal belongings. The addition of chloroiodides
may improve the application of surfactant micellar solutions
for this type of decontamination. The decomposition of 2,3,7,8-
TCDD by chloroiodides (without a light source) has been proven
in laboratory experiments. However, it has not been demon-
strated that TCDD levels can be decreased to less than 1 ppb to
soils. Additional bench-scale testing is needed for further
mization of these processes, perhaps including the ]
162
TREATMENT
-------
in situ decontamination of contaminated soil. However, in situ
decontamination using solubilizing agents may not be feasible be-
cause it raises the possibility of causing the transport of 2,3,7,8-
TCDD from soils into ground and surface waters.

Photolysis
Photolysis is a process in which chemical bonds are broken by
light energy. Several research studies have shown that chlorinated
dioxins may be photolytically degraded in the environment by
the ultraviolet (UV) wavelengths of sunlight, and that photolysis
occurs most rapidly in the presence of hydrogen donors such as
alcohols, ethers, hydrocarbons and natural oils and waves. The
photolytic degradation process appears to occur as a stepwise loss
of one chlorine atom at a time.
Syntex Agribusiness and IT Enviroscience have jointly devel-
oped a solvent extraction/photolysis process that was applied to
dioxin-laden sludge in Verona, Missouri, in 1980." In this pro-
cess, the dioxin was first extracted from a sludge using hexane as
the solvent. The resulting extract then was passed under eight
10-kW ultraviolet lamps to effect the desired reduction. In all,
4300 gal of the sludge were treated. Sawyer's report of the work
states that "160 gal batches of the waste were run through the
destruction process in five 2-week operating periods. Six extrac-
tions of the waste reduced the dioxin content from 34 mg/1 to
0.2 mg/1. Generally, photolysis reduced the dioxin concentration
(of the extract) to less than 0.1 mg/1 with optimum photolysis
times of about 20 hr. Air sampling at the single process vent con-
firmed that no dioxin (less than 0.01 ug/m3) was emitted to the
atmosphere during operation."
More recently, laboratory- and pilot-scale tests of a thermal
desorber/UV photolysis system were conducted.3' The system
consists of a thermal desorber followed by a solvent-based scrub-
bing system and a UV photolytic unit (Fig. 2). The system was
tested by U.S. EPA contractors on soil contaminated with
TCDD. The laboratory-scale tests indicated a TCDD removal
efficiency of more than 99%. The pilot-scale tests reduced the
TCDD concentration from 250 ppb to less than 1 ppb in 6 hr at
desorber temperatures of 860 °F (460 °C) to 932 °F (500 °C) and
feed rates of 30 to 100 Ib/hr. The study also showed similar reduc-
tions (85 to 95%) in the concentrations of 2,4,5-T and 2,4-D in
the soils.

Solvent Extraction
Only one full-scale process has been developed and used to
extract dioxins from contaminated wastes. The treatment pro-
cess, developed by IT Enviroscience and Syntex Agribusiness,
was used to remove dioxins from toluene still bottoms; the ex-
tract then was treated by photolysis to destroy the dioxin, as
described in the previous section.38 Approximately 15 Ib of dioxin
were removed from about 4600 gal of the waste using hexane as
the solvent. The research showed that hexane performed better
as a dioxin extractant than tetrachloroethylene and o-xylene.
The extraction procedure reduced the 2,3,7,8-TCDD concentra-
tion in the waste from 340 to 0.2 ppm via a series of six hexane
extractions. The extracts then were treated with UV light to de-
stroy the dioxin.
Other laboratory experiments have shown that, in general, 60
to 90% of dioxin can be removed from contaminated soils using
organic and aqueous solvent systems, but residual levels in the
soil tend to remain well above 0.1 ppm.6-40-41 Recent laboratory
research into the use of nonionic surfactant pairs in aqueous solu-
tion for treating contaminated soils has shown that 98% of PCBs
and 93% of hydrocarbons (including chlorophenols) can be re-
moved at an optimum surfactant concentration of 1.5%.43 The
surfactants used were Adsee® 799 (made by Witco Chemical)
and Hyonic® NP-90 (produced by Diamond Shamrock). Al-
though these results are encouraging, further research in the field
will be necessary to determine the full-scale effectiveness of the
process and its applicability to dioxin-contaminated soils.
Purge G*i Makeup
SoJvvnt purge

Figure 2
Thermal Desorption, Solvent Absorption/Scrubbing, UV
Photolysis Process Schematic"

Biodegradation
Early research studies on the biodegradation of 2,3,7,8-TCDD
in soils reported a half-life for the compound on the order of
1 to 2 yr.41'44'45 However, more recent studies have found that
dioxin binds very tightly to soils and does not degrade to any
appreciable extent.46-47'48 For example, one study found that less
than 1% of dioxin applied to soils 12 to 14 yr ago has been de-
graded and 99% remains firmly adhered to the soil.47 It now ap-
pears as though the half-life of the compound in soils is more
likely to be on the order of decades.
It is very likely that the results of early experiments were misin-
terpreted by not recognizing and accounting for the strong ten-
dency of 2,3,7,8-TCDD to bind to soil particles. Strongly-bound
dioxin would not have been detected analytically, and biodegra-
dation would have been assumed incorrectly to be the cause of its
apparent disappearance.
New microbial strains of organisms recovered from the soils
of various waste disposal sites (Love Canal, Eglin Air Force Base
and others) have been genetically engineered which, in pure cul-
ture, are capable of metabolizing 2,4,5-trichlorophenol and the
herbicide 2,4,5-T ,49'50'51'52 Though plans have been made to pur-
sue development of strains to degrade dioxins, work has not yet
begun.
The lignon-degrading white rot fungus, P. Chrysosporium,
recently has been investigated for its potential use as a natural
organism to degrade recalcitrant organopollutants such as
dioxins." The organism secretes an enzyme that can degrade a
wide variety of structurally diverse, recalcitrant compounds such
TREATMENT 163
-------
as lignon. Initial laboratory tests with 2,3,7,8-TCDD have shown
that, in pure culture, the organism can convert the dioxin to CO2-
Though not confirmed, this may represent an overall dioxin de-
gradation efficiency on the order of 99% or more based on sim-
ilar tests with the pesticide DDT.
Other organisms such as Nocardiopsis spp. and Bacillus mega-
terium in pure culture also have metabolized the dioxin." Pseudo-
monas stulzeri has shown its ability to use 3,4-dichlorophenol
as a carbon source, and researchers are hopeful that this organism
may have the potential also to degrade dioxins.5'

Stabilization/Fixation
Two studies have been funded by the U.S. EPA to evaluate the
ability of asphalt and portland cement to stabilize dioxin-contam-
inated soils from three Missouri sites." The soils contained 32 to
770 ppb of TCDD. After cement stabilization, leach tests showed
nanogram quantities of TCDD in the leachate, indicating that the
process may not be effective. Leach test results for the asphalt
process have not been reported.

Gamma Ray Radiolysis
Experiments performed in the 1970s showed that tetra- and
octa-chlorinated dioxins dissolved in solvents could be degraded
by 80 to 90% after a few hours' exposure to gamma radia-
tion.56'57 More recent experiments have shown that the process
can be used successfully to decontaminate laboratory wastes con-
taining small amounts of 2,3,7,8-TCDD." These experiments
accomplished at least 99% decomposition of the dioxin. This re-
cent study also reported 97% decomposition of 2,3,7,8-TCDD in
contaminated soil from the Seveso, Italy accident using the tech-
nique. The major decomposition products found were lower
chlorinated dioxins, indicating that dechlorination is the primary
mode of action.

Carbon Adsorption
One approach to the decontamination of organic and aqueous
liquids contaminated by dioxins is the use of carbon adsorption
techniques to selectively remove the toxic constituents from the
mixtures. Such techniques reduce the volume of material that
must be treated (e.g., incinerated) or securely disposed. One
major disadvantage of the process is that the carbon tightly binds
and concentrates the dioxins and cannot be regenerated.
In at least two instances, quantities of activated carbon heav-
ily contaminated with dioxins are being stored because disposal
methods are not available.' In this country, extensive pilot-plant
studies of carbon adsorption to remove dioxins from stockpiles
of Herbicide Orange were conducted before the Air Force decided
to incinerate the material. Although the herbicide reprocessing
method was technically and environmentally feasible, it was not
possible to demonstrate an acceptable method to safely dispose
of the dioxin-laden carbon. The contaminated carbon now is
stored on an island in the Pacific. Similarly, Union Carbide of
Australia created quantities of dioxin-contaminated carbon in
efforts to detoxify 2,4,5-trichlorophenol after they became aware
of the 2,3,7,8-TCDD contamination problem in 1969. This car-
bon still is stored in steel drums in that country.
In general, carbon adsorption techniques have not been proven
totally acceptable for toxics disposal, even if the carbon and the
contaminants it holds are to be destroyed by incineration or other
methods. After being contaminated with heavy organic chem-
icals, activated carbon usually must be dried and pulverized prior
to incineration to ensure complete destruction. These additional
handling steps provide a chance for fugitive emissions. Neverthe-
less, this method does offer a means to reduce the overall volume
of contaminated material by removing low-levels of contamina-
tion from large volumes of water, wastewater or organic liquids
and concentrating the contaminants in a much smaller and easier
to control volume of solid material.

Land Treatment
Land treatment is a potentially effective method to treat penta-
chlorophenol (PCP) wood preserving wastes and other chloro-
phenol wastes to degrade, adsorb or otherwise reduce the toritity
of the chlorophenol and other hazardous constituents present in
these wastes. Wastes such as sludges and contaminated soils typi-
cally are produced at hundreds of wood preserving plants using
PCP, and PCP often is contaminated with higher forms of chlor-
inated dioxins, especially octachlorodibenzo-p-dioxin. TCDDs
are encountered less frequently in these wastes.
Recent field studies on the effectiveness of land treatment indi-
cate the process can reduce concentrations of penta-, hexa- and
hepta-chlorodioxins in contaminated soils from wood preserving
by 75 to 95% and octachlorodioxins by 0 to 81% in a 4-month
period." Further research is needed to determine whether these
reductions in measurable dioxin content reflect true microbial de-
gradation or physical attenuation/binding of the dioxin to the
soil matrix.

CONCLUSIONS
Several of the thermal treatment technologies have demon-
strated six nines DRE on chlorinated compounds of various types
such as PCBs and carbon tetrachloride, but data fully demon-
strating this level of performance on dioxins have been published
for only three processes: mobile rotary kiln incineration, high
temperature fluid-wall destruction and infrared incineration.
Supercritical water oxidation also has been reported to achieve
six nines DRE on dioxins, but data supporting this claim have
not yet been released. Tests with dioxin wastes at the EPA sta-
tionary rotary kiln incineration in Arkansas came very close to
demonstrating six nines (> 99.997% DRE), but fell short due to
the low concentration of dioxin in the waste feed (37 ppb) and
the analytical detection limits, both of which worked together to
prevent the calculation of six nines.
Of the non-thermal methods, only two—thermal desorpthra
followed by UV photolysis and chemical dechlorination—have
demonstrated a capability to achieve residual levels of 1 ppb or
less dioxin in the treated waste or soil.

REFERENCES
1. Esposito, M.P., Tiernon, T.O. and Dryden, F.E., "Dioxins," U.S.
EPA 600/2-80-197, Nov. 1980, 263-264.
2. Ross, R.W., II, Backhouse, T.H., Vogue, R.N., Lee, J.W. and
Waterland, L.R., "Pilot-Scale Incineration Test Burn of TCDD-
Contaminated Toluene Stillbottoms from Trichlorophenol Produc-
tion from the Vertac Chemical Company," Prepared for U.S. EPA,
HWERL under U.S. EPA Contract No. 68-03-3267, Work Assign-
ment 0-2. Acurex Technical Report TR-86-100/EE. Acurex Corpora-
tion, Energy and Environmental Division, Jan. 1986.
3. Ross, R.W., II., Whitemore, F.C. and Carnes, R.A., "Evaluation
of the U.S. EPA CRF Incineration as Determined by Hexachloro-
benzene Incineration," Hazardous Waste, 1, 1984, 581-597.
4. Carnes, R.A. and Whitemore, F.C., "Characteristics of the Rotary
Kiln Incinerator System at the U.S. EPA Combustion Research
Facility (CRF)," Hazardous Waste, 1, 1984, 225-236.
5. Carnes, R.A., "U.S. EPA Combustion Research Facility Permit
Compliance Test Burn," Proc. of the Eleventh Annual Research
Symposium on Incineration and Treatment of Hazardous Waste,
Cincinnati, OH, Apr. 1985.
6. IT Corporation, "Interim Summary Report on Evaluation of Soils
Washington and Incineration as On-Site Treatment System! for
164
TREATMENT
-------
Dioxin-Contaminated Materials," Prepared for U.S. EPA, HWERL
under U.S. EPA Contract No. 68-03-3069, June 7, 1985.
7. IT Corporation, "Dioxin Trial Burn Data Package, U.S. EPA Mo-
bile Incineration System at the James Denney Farm Site, Mc-
Dowell, Missouri," Prepared for U.S. EPA, HWERL under U.S.
EPA Contract No. 68-03-3069, June 21, 1985.
8. Pyrotech Systems, Inc., "Mobile Waste Processor: MWP-2000-
ER," 1985.
9. Sickels, T.W., "ENSCO's Modular Incineration System: An Ef-
ficient and Available Destruction Technique for Remedial Action at
Hazardous Waste Sites," Preprinted extended abstract of paper
presented before the Division of Environmental Chemistry, Ameri-
can Chemical Society, 191st National Meeting, New York, NY, 26,
No. 1, April 1986.
10. Freestone, F., Miller, R. and Pfrommer, C., "Evaluation of On-
Site Incineration for Cleanup of Dioxin-Contaminated Materials,"
Proc. of the International Conference on New Frontiers for Haz-
ardous Waste Management, sponsored by U.S. EPA, HWERL,
NUS Corporation, National Science Foundation and American
Academy of Environmental Engineers. Pittsburgh, PA, Sept. 1985,
EPA 600/9-85-025.
11. Keitz, E., "Profile of Existing Hazardous Waste Incineration Facili-
ties and Manufacturers in the United States," EPA-600/2-84-052.
PB84-157072/REB, U.S. EPA, Cincinnati, OH, 1984.
12. U.S. EPA, Office of Planning and Evaluation, "Assessment of In-
cineration as a Treatment Method for Liquid Organic Hazardous
Waste," Background Report 1: "Description of Incineration Tech-
nology," Mar. 1985.
13. Daily, P.L., "Performance Assessment of Portable Infrared Incin-
erator," Presented at HMCRI Conference, Washington, DC, 1985.
14. Freeman, H.M., "Innovative Thermal Processes for the Destruc-
tion of Hazardous Waste," Separation of Heavy Metals and Other
Trace Contaminants, AIChE Symposium Series, 81, No. 243, 1985.
15. Richman, W.S., Holder, N.D. and Young, D.T., "Circulating Bed
Incineration of Hazardous Waste," GA Technologies, Inc., Chem.
Eng. Prog., 81, Mar. 1985.
16. Chang, D.P.Y. and Sorbo, N.W., "Evaluation of a Pilot Scale Cir-
culating Bed Combustor with a Surrogate Hazardous Waste Mix-
ture," Proc. of the Eleventh Annual Research Symposium on Incin-
eration and Treatment of Hazardous Waste, sponsored by U.S.
EPA, HWERL, Cincinnati, OH, Apr. 1985, EPA 600/9-85-028.
17. Lee, K.W. and Lewis, D.S., J.M. Huber PCB Destruction Process
Trial Burn Report, Oct. 1983.
18. Weston, Roy F., Inc., "Field Demonstration of the Destruction of
Dioxin in Contaminated Soil Using the J.M. Huber Corporation
Advanced Electric Reactor," Feb. 1985.
19. Boyd, J., Williams, H.D. and Stoddard, T.L., "Destruction of
Dioxin Contamination by Advanced Electric Reactor," Preprinted
extended abstract of paper presented before the Division of Environ-
mental Chemistry, American Chemical Society, 191st National
Meeting, New York, NY, 26, No. 1, Apr. 1986.
20. Kolak, N.P., Barton, T.G., Lee, C.D. and Peduto, E.F., "Trial
Burns—Plasma Arc Technology," Proc. U.S. EPA Twelfth An-
nual Research Symposium on Land Disposal, Remedial Action, In-
cineration and Treatment of Hazardous Waste, Cincinnati, OH,
Apr. 1986.
21. OCA Technology Division, Inc., "Detailed Review Draft Report:
Identification of Remedial Technologies," Prepared for U.S. EPA,
Office of Waste Programs Enforcement, under EPA Contract No.
68-01-6769, Work Assignment No. 84-120, GCA-TR-84-109-G-(0),
Mar. 1985.
22. Edwards, B.H., Paullin, J.N. and Jordan, K.C., Emerging Tech-
nologies for the Control of Hazardous Wastes, Noyes Data Corp.,
Park Ridge, NJ, 1983.
23. Johanson, J.G., Yosim, S.J., Kellog, L.G. and Sudar, S., "Elimina-
tion of Hazardous Waste by the Molten Salt Destruction Process,"
Incineration and Treatment of Hazardous Waste, Proc. of the
Eighth Annual Research Symposium, EPA-600/9-83-003, Apr.
1983.
24. Bellobono, I.R., Selli, R.E. and Veronese, L., "Destruction of
Dichlor- and Trichlor-Phenoxyacetic Acid Esters Containing 2,3,7,8-
Tetrachlorodibenzene-p-dioxin by Molten Salt Combustion Tech-
nique, Acqua-Aria," Jan. 1982.
25. Leslie, R.H., "Development of Mobile On-Site Earth Decontam-
inator," Proc. of the Hazardous '85 Conference, Baltimore, MD,
May 1985.
26. Killiley, W., Modar, Inc., Telephone conversation with Lisa Farrell,
GCA Technology Division, Inc., Feb. 1986.
27. Buelt, J.L., Battelle Memorial Institute, Pacific Northwest Labora-
tories, Telephone conversation with Michael Jasinski, GCA Tech-
nology Division, Inc., 1986.
28. Weitzman, L., "Treatment and Destruction of Polychlorinated Bi-
phenyls and Polychlorinated Biphenyl-Contaminated Materials,"
in Detoxification of Hazardous Waste, Chapter 8, 1982.
29. Metcalf and Eddy, Inc., "Technologies Applicable to Hazardous
Waste," Briefing prepared for U.S. EPA, HWERL, May 1985.
30. Radimsky, J. and Shah, A., "Evaluation of Emerging Technolo-
gies for the Destruction of Hazardous Waste," U.S. EPA Coopera-
tive Agreement R-808908, Jan. 1985.
31. U.S. EPA, Office of Research and Development, Status of Dioxin
Research in the U.S. EPA, Aug. 1985.
32. Freeman, H.M. and Olexsey, R.A., "A Review of Treatment Al-
ternatives for Dioxin Wastes," JAPCA 36, 1986, 67-75.
33. Klee, A., et al., "Report on the Feasibility of APEG Detoxification
of Dioxin-Contaminated Soils," U.S. EP A, lERL/Office of Re-
search and Development, EPA-600/2-84-071, Mar. 1984.
34. Ayres, D.C., "Destruction of Polychlorodibenzo-p-dioxins," Na-
ture, 190, Mar. 26, 1981, 323.
35. Ayres, D.C., Levy, D.P. and Creaser, C.S., "Destruction of Chlor-
inated Dioxins and Related Compounds by Ruthenium Tetroxide,"
in Chlorinated Dioxins and Dibenzofurans in the Total Environ-
ment II, Butterworth Publishers, Stoneham, MA, 1985.
36. des Rosiers, P., "Remedial Measures for Wastes Containing Poly-
chlorinated Dibenzo-p-dioxins (PCDDs) and Dibenzofurans
(PCDFs): Destruction, Containment or Process Modification,"
Ann. ofOccupat. Hyg., 27, 1983, 66.
37. Botre, C., Memoli, A. and Alhaique, F., "On the Degradation of
2,3,7,8-TCDD by Means of a New Class of Chloroiodides," En-
viron. Sci. Tech., 13, 1979, 228-231.
38. Sawyer, C.J., "Environmental Health and Safety Considerations
for a Dioxin Detoxification Process," in Detoxification of Hazard-
ous Waste, Chapter 18, 1982.
39. Helsel, R., et al., "Technology Demonstration of a Thermal De-
sorption/UV Photolysis Process for Decontaminating Soils Con-
taining Herbicide Orange," Reprint extended abstract. Presented
before the Division of Environmental Chemistry, American Chemi-
cal Society, New York, NY, Apr. 1986.
40. Kearney, P.C. et al., "Persistence and Nietabolism of Chlorodi-
oxins in Soils," Environ. Sci. Tech. 6, 1972, 1017-1019.
41. Ward, C.T. and Matsumura, F., "Fate of 2,3,7,8-Tetrachlorodi-
benzo-p-dioxin (TCDD) in a Model Aquatic Environment," En-
viron. Contam. Toxicol., 7, 1978, 349-357.
42. Vanness, G.F. et al., "Tetrachlorodibenzo-p-dioxins in Chemical
Wastes, Aqueous Effluents and Soils," Chemosphere, 9, 1980,
558-563.
43. Ellis, W.D., Payne, J.R. and McNabb, G.D., "Treatment of Con-
taminated Soils with Aqueous Surfactants," EPA/600/S2-85/129,
Dec. 1985.
44. Young, A.L. et al., "Fate of 2,3,7,8-Tetrachlorodibenzo-p-dioxin
(TCDD) in the Environment," Summary on Decontamination
Recommendations, USA FA-TR-76-18, 1976.
45. Commoner, B. and Scott, R.E., "U.S. Air Force Studies on the
TREATMENT 165
-------
Stability and Ecological Effects of TCDD (Dioxin): An Evaluation
Relative to the Accidental Dissemination of 2,3,7,8-TCDD at Seveso,
Italy," Center for the Biology of Natural Systems, Washington
University, St. Louis, MO, 1976.
46. DiDominico, A., el at., "Accidental Release of 2,3,7,8-Tetrachloro-
dibenzo-p-dioxin (TCDD) at Seveso, Italy," Ecotoxicol. and En-
viron. Safely. 4, 1980, 282-356.
47. Young, A.L., "Long-Term Studies on the Persistence and Move-
ment of TCDD in a Natural Ecosystem," in Human and Environ-
mental Risks of Chlorinated Dioxins and Related Compounds,
R.E. Tucker, el a/., eds.. Plenum Press, New York, NY, 1983, 173-
190.
48. Quensen, J.F. and Matsumura, F., "Oxidative Degradation of
2,3,7,8-Tetrachlorodibenzo-p-dioxin by Microorganisms, Environ.
Toxicol.. 2, 1983, 261-268.
49. Banerjee, S., Duttagupta, S. and Chakrabarty, A.M., "Produc-
tion of Emulsifying Agent During Growth of Pseudomonas Cepacia
with 2,4,5-Trichlorophenoxyacetic Acid," Arch. Microbiol., 135,
1983, 110-114.
50. Karns, J.S., Duttagupta, S. and Chakrabarty, A.M., "Regulation
of 2,4,5-Trichlorophenoxyacetic Acid and Chlorophenol Metabol-
ism in Pseudomonas Cepacia AC1100, App. Environ. Microbiol.,
46, 1983, 1182.
51. Karns, J.S., Kilbane, J.J., Duttagupta, S. and Chakrabarty, A.M.,
"Metabolism of Halophenols by 2,4,5-Trichlorophenoxyacetic Acid-
Degrading Pseudomonas Cepacia," App. Environ. Microbiol., 46,
1983, 1176.
52. Kilbane, J.J., Chatterjee, O.K. and Chakrabarty, A.M., "Detoxifi-
cation of 2,4,5-Trichlorophenoxyacetic Acid from Contaminated
Soil by Pseudomonas Cepacia," App. Environ. Microbiol., 4^
1983, 1697.
53. Bumpus, J.A., Tien, M., Wright, D.A. and Aust, S.D., "Biode-
gradation of Environmental Pollutants by the White Rot Fungus
Phanerocheate Chrysosporium, Presented at U.S. EPA, HWERL
llth Annual Research Symposium, Cincinnati, OH, Apr. 1, 1985.
54. Davies, L., Sybron Corporation, Personal communication to B.
Farino, GCA Technology Division, Inc., Jan. 1984.

55. Vick, W.H., Denzer, S., Ellis, W., Lambauch J. and Rottunda, N.,
"Evaluation of Physical Stabilization Techniques for Mitigation of
Environmental Pollution from Dioxin-Contaminated Soils," In-
terim Report: Summary of Progress-to-Date. Submitted to U.S.
EPA, HWERL by SAIC/JRB Associates, EPA Contract No. 68-02-
3113, Work Assignment No. 36, June 1985.
56. Buser, H.R., "Preparation of Qualitative Standard Mixtures of
Polychlorinated Dibenzo-p-dioxins and Dibenzofurans by Ultra-
violet and -Irradiation of the Octachloro Compounds," J. Chroma-
log., 129, 1976, 303-307.
57. Fanelli, R., Chiabrando, C., Salmona, M. Garattini, S. and Calders,
P.G., "Degradation of 2,3,7,8-Tetrachlorodibenzo-p-dioxin in Or-
ganic Solvents by Gamma Ray Irradiation, Experientia, 34, Sept. 9,
1978, 1126-1127.
58. Buser, H.R. and Zehnder, H.J., "Decomposition of Toxic and
Environmentally Hazardous 2,3,7,8-Tetrachlorodibenzo-p-dioxin by
Gamma Irradiation, Experientia, 41, 1985, 1082-1084.
59. Ryan, J.R. and Smith, J., "Land Treatment of Wood Preserving
Wastes," in Proc. of the National Conference on Hazardous Wastes
and Hazardous Materials, Atlanta, GA, Mar. 1986, 80-86.
166
TREATMENT
-------
Stabilization/Solidification of Hazardous Wastes

M. John Cullinane, Jr., P.E.
R. Mark Bricka
USAE Waterways Experiment Station
Vicksburg, Mississippi
Larry W. Jones, Ph.D.
University of Tennessee
Knoxville, Tennessee
INTRODUCTION
Stabilization and solidification refer to treatment processes
that are designed to accomplish one or more of the following:
improve the handling and physical characteristics of the waste;
decrease the surface area of the waste mass across which transfer
or loss of contaminants can occur; and/or limit the solubility of
any hazardous constituents of the waste such as by pH adjust-
ment. '
Stabilization techniques generally are those whose beneficial
action is obtained primarily through limiting the solubility or
mobility of the contaminants with or without change or improve-
ment in the physical characteristics of the waste. Both solidifica-
tion and chemical stabilization usually are included in commercial
processes and result in the transformation of liquids or semisolids
into environmentally safer forms. This seminar provides guid-
ance for the evaluation, selection and use of stabilization/solid-
ification technology.

REGULATORY BASIS FOR USING
STABILIZATION/SOLIDIFICATION
Primary authorization for the control and regulation of haz-
ardous wastes by the U.S. EPA is based on either CERCLA, as
amended by the Superfund Amendments and Reauthorization
Act of 1986 (SARA), or RCRA, as amended by the Hazardous
and Solid Waste Amendments of 1984. Numerous implementing
regulations have been proposed under authorization of these
statutes. Stabilization/solidification is recognized as a potential
technology under both SARA and RCRA. The RCRA amend-
ments have a significant impact on the use of stabilization/solid-
ification at RCRA-permitted landfills. Specifically, Section 3004
(c)(l), which took effect on May 8, 1985, addresses the disposal
of bulk liquids in landfill. This provision states:

"Effective 6 months after the date of enactment of the
Hazardous and Solid Waste Amendment of 1984, the
placement of bulk or noncontaminated liquid hazardous
waste or free liquids contained in hazardous waste (whether
or not absorbents have been added) in any landfill is pro-
hibited."

The U.S. EPA2 interprets Section 3004(c)(l) as requiring that
bulk liquid hazardous waste intended for disposal in a landfill
should first be chemically, thermally or biologically treated with-
out the use of sorbents to include both absorbents and adsor-
bents. A sorbent material may, however, be used as one of the
ingredients in a chemical stabilization process if the final product
passes an unconfined compression test.
It is difficult, using current technology, to determine whether a
particular process involves chemical stabilization or merely ab-
sorption. To aid in this determination, the U.S. EPA2 proposes
the unconfined compressive strength test (ASTM D2166-85, Un-
confined Compressive Strength of Cohesive Soils) as an indirect
method to determine the stability of treated waste. A minimum
allowable strength of 50 lb/in.2 is established as the measure of
adequate bonding. This level was selected in an effort to require a
bonding level in excess of that achieved with sorbents.

STABILIZATION/SOLIDIFICATION PROCESSES
Several generic treatment systems have been developed for
waste stabilization/solidification. Most stabilization/solidifica-
tion systems being marketed are proprietary processes involving
the addition of absorbents and solidifying agents to a waste.
Often the marketed process is changed to accommodate specific
wastes. Since it is not possible to discuss completely all possible
modifications to a process, discussions of most processes have to
be related to generic process types. The exact degree of per-
formance observed in a specific system may vary widely from its
generic type, but the general characteristics of a process and its
products can be discussed. Comprehensive general discussions of
waste stabilization/solidification are given by Malone, et al.'
and Cullinane, et al.}
Generic waste stabilization/solidification processes that have
been identified include:' sorption, portland cement, lime-fly ash
pozzolan, portland-cement pozzolan, thermoplastic microencap-
sulation and thermoplastic macroencapsulation. Other technol-
ogies, such as fusing waste into a vitreous mass or using self-
cementing material, are too specialized or not sufficiently field
demonstrated to be used at present.'
Sorption involves adding a solid to soak up any liquid present
in the waste. Sorption generally produces a soil like material.
Typical examples of sorption materials are activated carbon, an-
hydrous sodium silicate, various forms of gypsum, celite, clays,
expanded mica and zeolites. Sorption processes have been severe-
ly restricted by the 1984 amendments to RCRA.
Lime-fly ash pozzolanic processes use finely divided noncrystal-
line silica in fly ash and the calcium in lime to produce low
strength cementation. The waste containment is produced by en-
trapping the waste in the pozzolan concrete matrix.
Pozzolan-portland systems use portland cement and flyash or
other pozzolanic materials to produce a stronger type of waste/
concrete composite. The waste containment is produced by
microencapsulation in the concrete matrix. Soluble silicates also
may be added to accelerate hardening and improve metal con-
tainment.
Thermoplastic microencapsulation involves blending fine par-
ticulate waste with melted asphalt or other matrix. Liquid and
TREATMENT 167
-------
volatile phases associated with the waste are driven off, and the
waste is isolated in a mass of cooled, hardened asphalt. The ma-
terial can be buried with or without secondary containment.
Thermoplastic macroencapsulation systems contain a waste by
isolating large masses of the waste using some type of jacketing
material. Typically, a 55-gal drum or polyethylene jacket is fused
over a monolithic block of solidified waste.

FIELD APPLICATION OF STABILIZATION/
SOLIDIFICATION PROCESSES
The development and selection of the stabilization/solidifica-
tion operations plan for a particular remedial action site are de-
pendent on several factors such as the nature of the waste ma-
terial, the quantity of the waste material, the location of the site,
the physical characteristics of the site and the solidification pro-
cess to be utilized. When the solidification program is being devel-
oped, the primary goal is to create optimum efficiency which is
constrained by both short- and long-term environmental and
public health considerations.
Four alternative scenarios are available for stabilization/solid-
ification of hazardous wastes: in-drum mixing, in-situ mixing,
mobile plant mixing and area mixing.3 The selection of an appro-
priate solidification/stabilization technique is based on an analy-
sis of waste, reagent and site-specific factors. As a result, only
generalized criteria can be developed for any given waste or dis-
posal site.
In-drum mixing is best suited for application to highly toxic
wastes that are present in relatively small quantities. This tech-
nique also may be applicable in cases where the waste is stored in
drums of sufficient integrity to allow rehandling. In-drum mixing
is typically the highest-cost alternative when compared with in
situ, mobile plant and area mixing scenarios. Quality control
also presents serious problems in small batch mixing operations;
complete mixing is difficult to achieve, and variations in the waste
between drums can cause variations in the characteristics of the
final product.
In situ mixing is primarily suitable for closure of liquid or slurry
holding ponds. In situ mixing is most applicable for the addition
of large volumes of low reactivity, solid chemicals. The present
state of technology limits application of in situ mixing to the
treatment of low solids content slurries or sludges. Where applic-
able, in situ mixing usually is the lowest cost alternative. Quality
control associated with in situ mixing is limited with present tech-
nology. A variety of new equipment for in situ application of
stabilization/solidification technology currently is under develop-
ment.
Mobile mixing plants can be adapted for application to liquids,
slurries and solids. This technique is most suitable for applica-
tion at sites with relatively large quantities of waste materials to be
treated. It gives best results in terms of quality control. Mobile
plant mixing is applicable at sites where the waste holding area is
too large to permit effective in situ mixing of the wastes or where
the wastes must be moved to their final disposal area.
Area mixing consists of spreading the waste and treatment re-
agents in alternating layers at the final disposal site and mixing jj
place. This technique is applicable to those sites where slurries
contain high concentrations of solids or where contaminated
soils or solids must be treated. Area mixing requires that the waste
materials be handled by construction equipment (i.e., dump
trucks, backhoes, etc.) and is not applicable to the treatment of
liquids. Area mixing is land-area intensive, as the process requires
relatively large land areas. Area mixing presents the greatest pos-
sibility for fugitive dust, organic vapor and odor generation. Area
mixing ranks below in-drum and plant mixing in terms of quality
control.

RESEARCH NEEDS
To date, most investigations related to the performance of
stabilization/solidification technology have been conducted by
vendors and are empirically based. Acceptance of stabilization/
solidification technology as a legitimate hazardous waste treat-
ment process will require additional research. Five major cate-
gories of research needs have been identified to improve both the
basic understanding of the physical treatment processes and the
acceptability of such methods for treating hazardous wastes.'
These include: development of uniform definitions, terminology
and testing regimes; process development; development of testing
procedures; equipment development; and research into process
cost-effectiveness.

ACKNOWLEDGEMENTS
The development of information presented in this seminar was
primarily funded by the U.S. EPA under Interagency Agreement
No. AD-96-F-2-A145 with the U.S. Army Engineer Waterways
Experiment Station(WES). The materials presented here were
developed by the Environmental Laboratory of the WES. Pro-
ject supervision was provided by the Hazardous Waste Engineer-
ing Research Laboratory. Permission to publish this information
was granted by the U.S. EPA and the Chief of Engineers.
REFERENCES
1. Malone, P.O., Jones, L.W. and Larson, R.J. "Guide to the Disposal
of Chemically Stabilized and Solidified Waste," SW-872, Office of
Water and Waste Management, U.S. EPA, Washington, DC, 1980.
2. U.S. EPA. "Prohibition of the Placement of Bulk Liquid HazardotB
Waste in Landfills—Statutory Interpretative Guidance," OSWER
Policy Directive 9487.00-2A, Office of Solid Waste and Emergency
Response, Washington, DC, 1980.
3. Cullinane, M.J., Jones, L.W. and Malone, P.O. "Handbook for
Stabilization/Solidification of Hazardous Wastes," EPA/540/M6V
001, Hazardous Waste Engineering Research Laboratory, U.S. EPA,
Cincinnati, OH, 1986.
4. Cullinane, M.J., Bishop, P.L., Jones, L.W., Pitt, W. and Tittle-
baum, M.E. "Report of Panel 4: Physical Treatment Panel," Pnc.
of Research Needs Workshop: Hazardous Wastes Treatment ad
Disposal, Sponsored by the National Science Foundation, Droe!
University, June 1986, Philadelphia, PA.
168 TREATMENT
-------
Construction of Multiple-Lined
Hazardous Waste Landfills

Michael G. Ruetten, P.E.
Philip P. Stecker, P.E.
Donohue & Associates, Inc.
Sheboygan, Wisconsin
Wendell W. Lattz
Chemical Waste Management, Inc.
Fort Wayne, Indiana
ABSTRACT
Multiple-liner systems became a requirement for hazardous
waste landfills on Nov. 8, 1984, with passage of the Hazardous
and Solid Waste Amendments of 1984 to RCRA. Draft guide-
lines subsequently published by the U.S. EPA present the min-
imum technology requirements for landfill and surface impound-
ment design, construction and operation.
A comprehensive approach to the design of a multiple-lined
landfill should address site and geological constraints, regulatory
requirements, construction and operation conditions and docu-
mentation standards. This paper discusses the engineering ap-
proach used by Donohue & Associates for multiple-lined haz-
ardous waste landfills. The major elements addressed are design,
construction and documentation.

INTRODUCTION
Landfills are part of a complex, highly integrated waste hand-
ling system of concern to generators, haulers, facility owners,
regulatory agencies and the general public. The current regula-
tions and minimum technology guidelines are the backbone of
hazardous waste landfill projects. This paper addresses design,
construction and documentation concerns critical to the per-
formance of a multiple-lined landfill.
Donohue & Associates has designed several area-fill, multiple-
lined landfills. These landfills have been constructed in stages
under different contracting methods including single-prime con-
tracts and multiple-prime contracts. Our involvement has in-
cluded initial site investigations, data interpretation, design,
preparation of construction drawings and specifications, prepa-
ration of bid documents, monitoring quality assurance and de-
sign reviews. Through this comprehensive involvement, an ap-
proach has been developed to multiple-lined landfill design, con-
struction and documentation based on the site constraints, own-
er's concerns and minimum technology guidelines.

DESIGN
Because many hazardous waste landfills are located at a com-
plex waste handling facility, the primary step in the landfill de-
sign is an evaluation of controlling conditions. Concerns in the
design process are as follows:

• Site and geology
• Regulations
• Construction
• Operations
• Closure

Site and Geology
To provide an effective landfill design, the site waste hand-
ling procedures should be evaluated. The site considerations that
affect landfill design are access roads, power distribution lines,
leachate transmission pipes and storage tanks and current opera-
tions. These considerations are shown on Fig. 1. This type of flow
schematic is typically found within the Part B permit for a facil-
ity or can be generated by the design engineer following a site
visit.
The landfill location and size are dictated by physical con-
straints such as geology and topography and by legal constraints
such as zoning and permitting. Therefore, an integrated approach
that incorporates site constraints with natural and regulatory con-
straints is required for an efficient operation.
Although extensive site data have been available from past
geologic and hydrogeologic studies for the facilities with which we
have been involved, additional detailed landfill location data were
required for design. These data include soil borings, soil testing,
survey data, topographic mapping and locating electrical services
and utility easements. For some sites, preliminary monitoring of
soil or water-quality characteristics also may be required.
At the beginning of a project, an initial design meeting is held
with the owner to cover regulatory constraints, owner concerns
and integration of the landfill into the plant flow process. This
is one method of allowing the owner to customize the design. In-
formation critical to the operator of the facility may include:
• Haul road location
• Transport vehicle types and sizes
• Waste load volumes
• Waste processing time and conveyance (drum handling, de-
canting, stabilization, etc.)
• Leachate handling
• Truck wash
• Contracting methods
• Construction timing
We use computer programming and graphics to achieve an
efficient design. The site topography data are entered into a
computer-aided design and drafting (CADD) file either through
an aerial photography firm or by digitizing the data. The land-
fill size, depth and cell layout are then developed into an alpha-
LAND DISPOSAL
169
-------
numeric computer data file which is entered into the interactive
landfill design program. This program is used to generate CADD
drawings which are compiled with the topographic file for de-
sign review. Computerization of the preliminary design process
allows the project engineer to evaluate the design quickly to max-
imize space utilization and meet applicable regulatory guidelines.
This program also is utilized to prepare final drawings for the
construction documents.

Regulations
The design of the landfill can then proceed to address regula-
tory agency requirements, the backbone of the design considera-
tions. Because much has been written about the minimum tech-
nology requirements under RCRA, this paper will only summar-
ize design requirements briefly.

Secondary Composite Liner System
• Flexible membrane liner (FML) at least 30 mil
• Bedding above the FML
• Soil component 36 in. thick
• Permeability 1 x 10~7cm/secor less
• Subsoil structurally immobile
• Quality assurance plan and specifications

Secondary Leachate System
• Rapid detection of liquids in the layer
• Rapid collection of liquids in the layer with little or no head
buildup on the liner
• Hydraulic conductivity of 1 x 10"2 cm/sec minimum
• Lie directly on FML
• Slope at minimum 2%
• Sump at least 12 in. below drainage layer grade
• All areas between liners covered (i.e., sideslopes and base)
• Monitoring and liquid removal features
• Quality assurance plan and specifications

Primary Liner
• FML at least 30 mil
• FML chemically compatible with waste and leachate
• Bedding appropriate for FML protection
• FML protected and continuous around structures
• Quality assurance plans and specifications

Primary Leachate System
• Thickness at least 12 in.
• Hydraulic conductivity not less than 1 x 10~2 cm/sec
• Filters to protect the layer from clogging
• Base and sideslopes covered
• Monitoring and liquid removal systems capable of handling
flows
• Quality assurance plan and specifications
Often less attention is given to the requirements for leachate
pumping, surface water control, construction contracting and
landfill operation than is given to the items above, but these items
are critical to landfill development and performance. Many of the
principles and approaches used for landfills originate from other
types of civil engineering projects (for example, some of the
leachate pumping methods are similar to those used with waste-
water pump station and force main systems). However, landfills
require various specialized design features because of the nature
of the materials being disposed and the consequences of inade-
quate design.
Leachate handling problems typically arise from both opera-
tional concerns and construction of landfills in phases or mul-
tiple cells. Leachate systems should be designed to address needs
during three distinct stages: construction, operation and closure,
During construction, the leachate system within the landfill
should be capable of handling surface water accumulation. The
design may require temporary power, alternate pumps and an
outlet channel. For landfill operations, the system must be in-
stalled for continuous discharge and power because the system
will operate as a demand system, activating as leachate levels dic-
tate.
Instrumentation for pumping control must be integrated with
downstream storage to prevent spillage and to provide the opera-
tor with an indication of system status. Typical status indicators
include low level alarm, low level off, high level on and high level
alarm. The system also should include power disruption features,
downstream protection which cuts off pump power, and startup
after power disruption or shutdown. A similar system should be
available for primary and secondary leachate systems. This inte-
grated control system should be monitored from a central loca-
tion and system components should be individually recorded.
During closure, the system remains a demand system, but flow
volumes are expected to decrease because infiltration is restricted
by the cap. During closure, the system must continue to operate
as before, but it should require little maintenance and be easy to
monitor since site staff numbers will be fewer than during opera-
tions.
Surface water control measures during construction and opera-
tions require specific design considerations. During construction
surface water control is essential to achieve quality work on the
clay and synthetic components of the liner systems. Management
of surface water includes daily measures to restrict open areas,
grading excavations or fills to drain, maintaining 24-hr pumping
potential and collecting discharged water. Management of water
pumped from the construction area would include discharging to
stormwater structures or ditches, preventing flooding of accessor
operation areas and providing sedimentation control.
Surface water controls during operations include runon and
runoff control. Containing all contact water within the limits of
the landfill and handling it as leachate may affect the leachate
pumping design. Surface ponds and temporary pumps may be
more effective than a permanent pumping system to collect con-
tact water before it moves into the leachate collection layer. Ex-
treme care must be taken when purposely preventing water from
entering the leachate system; otherwise, zones of perched water
may occur during filling and closure. Runon control measures
may include culverts, berms and ditches designed to fit the site
constraints established by operations. For example, runon control
should not cause ponding on access roads or increase peak flows
through existing stormwater structures unless capacities are veri-
fied.

Construction
Design conditions related to construction of multiple-lined
landfills are dependent on landfill cell configuration and con-
struction stages. Generally area-type landfills will be constructed
over several years which is reflected by staged construction draw-
ings. The staged construction will require design of temporary
access, surface water control, leachate handling, protection of
synthetics, access during construction, leachate handling and
power distribution. Stability analyses of the waste placement for
area-fill sites are critical because one side of the landfill nonnW
is not considered to be fully contained because of construrfoB
activities.
Quality assurance plans and construction as related to
are important to all parties involved. The engineer must i
LAND DISPOSAL
-------
quality control measures and construction methods during the de-
sign to ensure a quality project. Construction in stages and with
individual cells presents design concerns relating to surface water
control, access, synthetic liner protection, contamination preven-
tion, soil stockpiling and construction scheduling. Construction
drawings and specifications should provide details for temporary
roads, minimum fill thickness over synthetics, maximum ex-
posure times for clay and synthetics and repair measures for clay.
Specifications should provide measurable limits for items such as
desiccation cracks, water ponding depths, load restrictions on
liners, surface conditions prior to lining and anchor trench main-
tenance. Additional concerns for quality assurance and construc-
tion are included under those sections of this paper.

Operations
Landfill operations provide design challenges as addressed
previously for access roads, surface water control, leachate pump-
ing and cover placement. Additional items critical to the design
which occur during operations are side slope protection, access
road construction within cells, leachate pumping and transporting
and mechanical/electrical system maintenance.

Closure
The design process for closure requires the engineer to incorpo-
rate features necessary for ease in maintenance, access, monitor-
ing and remedial repairs. The requirements on landfill covers are
detailed within regulation and guidance documents. Design items
within the regulations include slope stability, settlement potential,
erosion and runoff control and monitoring. In addition, the de-
sign engineer needs to analyze the performance of the leachate
system components, monitoring devices and cover materials for
an extended service life. The landfill oftentimes is a corrosive or
abrasive environment which creates additional challenges for
long-term performance. Centralized control and monitoring of
the leachate system is a benefit during closure when staff numbers
are lower.

CONSTRUCTION
Construction in phases allows an owner to design an entire
landfill and operate only that area required for projected receipts.
This phased approach minimizes capital construction cost, re-
duces operating costs and could allow landfill upgrading as tech-
nology changes.

Site and Geology
Site constraints will affect construction activities, and provis-
ions should be discussed between the engineer and owner prior to
bidding. Site constraints may include the following:

Equipment parking area
Material storage and staging area
Stockpile areas
Construction equipment routing
Contractor employee parking
Temporary offices and facilities
Temporary electrical service
Temporary surface water control
Local union agreements
Local building permit requirements
Geological conditions will affect construction activities in terms
of access, equipment requirements and documentation. The de-
sign drawings should provide the contractor with an explanation
of soil conditions existing at the landfill site, and the design engi-
neer should prepare specifications on excavation, stockpiling,
soil classification and water control. Foundation documentation
requirements also should be provided. An important considera-
tion in foundation documentation is corrective action. If the con-
ditions encountered are slightly different from the specifications
(i.e., foundation soil not uniform in classification), repair meas-
ures must be prepared quickly and regulatory approval must be
granted to allow construction to proceed on schedule.
When construction proceeds in stages over several years, a crit-
ical detail is the transition zone between the stages. Securing and
protecting multiple-lined synthetic layers is critical to the success-
ful extension of the active area. If the synthetic layers are dam-
aged in the transition area between construction stages, extensive
time delays and additional construction costs can be incurred.
More importantly, the integrity of the containment can be jeopar-
dized. To reduce problems associated with staged construction,
the entire landfill should be designed at one time, and special de-
tails should be prepared to indicate fill limits within the contain-
ment and runoff control areas. Care should be taken to secure the
ends of the synthetic materials to ensure containment. By ending
the synthetic layers on a cell separation berm, recovery can be
achieved without disturbing operations and adjacent construc-
tion. Fig. 2 and 3 show the use of a berm to protect synthetic ma-
terial in the transition zones.

Regulations
Because the regulations require documentation of construction
activities and materials, the construction drawings should be
structured to allow efficient and frequent field checks. For ex-
ample, clay grades should be provided on the drawings with suf-
ficient clarity to allow layout and construction documentation.
This detail may include indicating the frequency of contours,
marking critical control points, indicating corner radii, referring
to details and referring to the relationship to previous layers.
Since the contractor will be required to use a localized grid sys-
tem to lay out the project, providing design control date will pre-
vent later modifications due to misalignments.
The design engineer must provide clear, concise specifications
outlining the responsibilities of the owner, contractor, inspector
and engineer. These specifications provide the basis for the devel-
opment of a construction documentation and quality assurance
plan required by the regulations. The engineer should note that
there is a distinct difference between specifications and a quality
assurance plan. On a multiple-lined facility, the synthetic liner
usually is provided and installed under the site quality control
program. However, the soils, excavation, fabrics, mechanical and
electrical components generally are not. Therefore, the entire con-
struction process must be documented under a comprehensive
quality assurance program. The inspector/engineer must verify
that specifications are followed and that synthetic fabrication/in-
stallation quality control is performed.
Bidding and contracting are becoming increasingly complex as
landfills grow more technologically sophisticated. In the past,
projects typically involved just an owner and an earthwork con-
tractor. Now, multiple-lined hazardous waste landfill projects are
multi-million dollar projects involving multiple contractors (i.e.,
earthwork, synthetic liner, mechanical and electrical contractors).
This multiplicity of firms provides for numerous contracting
methods, including single-prime general contractors, multiple-
prime contractors and multiple subcontractors acting under the
assigned construction manager (a firm or person). These con-
tracts all have legal boundaries too extensive to cover within this
paper. However, some important factors to consider are insur-
ance, bonding and warranties.
When developing construction drawings and specifications for
a landfill being constructed in stages, the design engineer must be
aware of the final bidding procedure to best utilize the referenc-
LAND DISPOSAL 171
-------
ing of details and cross-referencing of the specification. Many
variations can be developed in contracting and bidding docu-
ments. The process is much easier if decided before all the draw-
ings and specifications are completed.
As construction begins the engineer will encounter requests for
field modifications or, at least, design clarifications. The pro-
cedures to handle these items must be fully established within the
specification to prevent delays or regulatory restrictions.
The design engineer may become involved in construction man-
agement because of his knowledge of the project. The design,
specification and quality assurance plan will place the engineer in
the role of verifying, certifying, clarifying or redesigning, all of
which directly affect construction progress. Construction on any
multi-million dollar project should have sufficient management,
scheduling and owner control to provide smooth project imple-
mentation.
Methods available for construction scheduling and methods for
contracting vary. On some projects, a simple bar chart may suf-
fice, while on others a sophisticated critical path method may be
more appropriate. The obvious construction milestones can be
assigned to completion of each landfill layer. However, other
components may alter completion of a layer, such as leachate
system, mechanical system and electrical system installation.
Proper project scheduling will help solve coordination problems
which may arise between subcontractors or inspectors.

Operations
Some aspects of landfill construction will continue during oper-
ations and generally be performed by the owner. Construction
activities include extending leachate risers, raising berms, main-
taining access roads, placing intermediate soil cover and placing
sideslope cover. Details of inspection requirements and construc-
tion specifications should be included as part of the construction
set for all operational construction activities. Photographic docu-
mentation of these activities generally aids closure document
preparation significantly.

Closure
Closure construction can be performed in stages or at one time
for the entire area depending on regulatory requirements. Sep-
arate drawings related to closure construction and the landfill
base liner will help simplify the contracting. Construction ma-
terials, procedures and site concerns are very similar to landfill
base construction. The additional items for closure may include
vegetation, access road removal and recontouring of the sur-
rounding area for aesthetic concerns. Early establishment of veg-
etation will decrease erosion and runoff impacts and help reduce
the potential for infiltration. Construction activities can be ex-
tended to cover a repair and maintenance period for vegetation
establishment if desired by the owner. This measure will reduce
personnel and equipment demands on the owner following land-
fill closure.
Quality Control
The design engineer should develop quality assurance and qual-
ity control (QA/QC) requirements to verify that the materials and
construction meet design requirements. Each site QA/QC pro-
gram needs to be developed as an individual program because of
regulatory and site constraints. The U.S. EPA Technical Guid-
ance Document "Construction Quality Assurance for Hazardous
Waste Land Disposal Facilities," provides an excellent starting
point for developing a site specific plan.
The U.S. EPA document outlines the five major components
of the QA/QC plan as:
• Responsibility and Authority
The responsibility and authority of organizations and key per-
sonnel (by title) involved in permitting, designing and construe-
ting the hazardous waste land disposal facility should be de-
scribed in the QA/QC plan.
• QA /QC Personnel Qualifications
The qualifications of the QA/QC officer and supporting QA/
QC inspection personnel should be presented in the QA/QC
plan in terms of the training and experience necessary to fulfill
their identified responsibilities.
• Inspection Activities
The observations and tests that will be used to ensure that the
construction or installation meets or exceeds all design criteria,
plans and specifications for each hazardous waste land dis-
posal facility component should be described in the QA
/QC plan.
• Sampling Strategies
The sampling activities, sample size, methods for determining
sample locations, frequency of sampling, acceptance and re-
jection criteria and methods for ensuring that corrective meas-
ures are implemented as addressed in the design criteria, plans
and specifications should be presented in the QA/QC plan.
• Documentation
Reporting requirements for QA/QC activities, including such
items as daily summary reports, inspection data sheets, prob-
lem identification and corrective measures reports, block eval-
uation reports, acceptance reports, final documentation, and
provisions for the final storage of all records should be pre-
sented in the QA/QC plan.

CONCLUSION
When the design engineer approaches a multiple-lined haz-
ardous waste landfill project with a comprehensive program, in-
cluding design considerations, construction considerations and
documentation requirements, an effective site development plan
Figure 1
General Operations Flow Sheet
Figure 2
Synthetic Materials Protected on Cell Separation Berm
172 LAND DISPOSAL
-------
Figure 3
Synthetic Materials Continued into Expansion

can be developed. The guidelines presented by regulatory agencies
provide a base upon which the designer must build to meet the
owner and site constraints and achieve successful implementa-
tion.
REFERENCES
1. Koerner, R.M., Designing with Geosynthetics, Prentice-Hall, New
York, NY, 1986.
2. Levy, Sidney M., Project Management in Construction, McGraw-
Hill, New York, NY, 1986.
3. U.S. EPA, "Draft Minimum Technology Guidance Document on
Double Liner Systems for Landfills and Surface Impoundments—
Design, Construction, and Operation." U.S. EPA, Washington,
DC, EPA/530-SW-85-014, 1985.
4. U.S. EPA, "Technical Guidance Document: Construction Quality
Assurance for Hazardous Waste Land Disposal Facilities." U.S.
EPA, Washington, DC, EPA/530-SW-86-031,1986.
LAND DISPOSAL
173
-------
An Improved Protective Cover Design
For Hazardous Waste Landfills
G. Stephen Mason, Jr.
U.S. Pollution Control, Inc.
Oklahoma City, Oklahoma
Robert T. Pyle
Pyle and Associates
Avon, North Carolina
ABSTRACT
The Hazardous and Solid Waste Amendments of 1984 require
that all hazardous waste landfills be constructed with a double
liner system. A protective cover is frequently installed above the
liner system to protect it from mechanical damage. This paper
compares the effectiveness of various protective cover options
and the attributes associated with each option. The technique
utilized to install a stabilized soil cover at an Oklahoma hazardous
waste landfill is presented.

INTRODUCTION
A hazardous waste landfill that was completed at an Oklahoma
commercial facility in 1986 complied with the Hazardous and
Solid Waste Amendments of 1984 (HSWA). The cell liners, in
ascending order, consist of a composite secondary liner system, a
secondary leachate collection system, a primary liner, a primary
leachate collection system and a protective cover (Fig. 1).
The secondary liner system consists of a 3-ft layer of com-
pacted clay and a 60 mil High Density Polyethylene (HDPE)
liner, while the secondary leachate collection system is a Medium
Density Polyethylene (MDPE) drainage net having a transmissiv-
ity of approximately 3 x 10~4 mVsec. The primary liner consists
of an 80 mil HDPE liner, and the primary leachate collection
system consists of a MDPE drainage net and a nonwoven poly-
propylene geotextile.
The cell is triangular and measures approximately 250 ft at the
interior base along each leg; the depth varies between 21 and 25 ft
with a surface area of approximately 133,000 ft2. The design ca-
pacity of the cell is 37,000 yd1.
9 INCHES SCREENED
MATERIAL
9 NCHES SCREENED
MATERIAL
6 IMCHES STABILIZED SOIL
NON-WOVEN
GEOTEXTILE FABRIC
PRIMARY LEACHATE
COLLECTION SYSTEM
C DRAINAGE NET )
PRIMARY LWER
C 80 mi HDPE )
SECONDARY LEACHATE
COLLECTION SYSTEM
( DRAINAGE NET )
SECONDARY
COMPOSITE LINER
t 60 ma HDPE AND
THREE FOOT COMPACTED
CLAY LINER )
Figure 1
Typical Hazardous Waste Cell Cross-Section
Theoretically, a protective cover is installed above the liner sys-
tem in order to protect it from mechanical damage during opera-
tions. Mechanical damage might result from vehicular traffic
directly above the liner or penetration of the liner by material dis-
posed of in the cell. From the final four of the various optiom
considered in designing a protective cover at that site, a soil ad-
mixture was selected as the most effective alternative. This ma-
terial was successfully installed during July 1986.

OPTIONS CONSIDERED FOR A
PROTECTIVE COVER
Recent U.S. EPA guidance recommends some form of cover to
protect the underlying liner system in hazardous waste landfills.'
The initial design for a cell in Oklahoma incorporated a 2-ft pro-
tective soil cover on the cell bottom which also would extend for
5 vertical feet up the slopes. The soil cover would have been ex-
tended up the slopes as the cell was filled. Concerns were raised
about: the stability of such slopes, given the available soils; plac-
ing the protective cover without damaging the liner system; and
leaving portions of the liner system on the slopes exposed to the
elements and other influences. Therefore, additional options were
considered. The final four options were as follows:

A. No cover
B. Soil cover
C. Concrete-filled fabric forms
D. Stabilized soil
Option A, no cover, consisted of not covering the synthetic
liner system and leaving it exposed to the elements and other in-
fluences.
Option B, soil cover, consisted of installing a 2-ft layer of toil
over the entire bottom of the cell and extending it for 5 verticil
feet up the slope. The soil cover would be extended up the slope
as the cell was filled. A soil ramp would initially be installed in I
cell using a conveyor. The soil would then be installed using earth
moving equipment such as dozers, backhoes and dump tnido.
Option B was the original design prior to the considerationi of
these additional operations. This option is described in U.S. EPA
guidance for double-lined cells.'
Option C, concrete-filled fabric forms, consisted of insUffini
a 6-in. thick layer of stabilized soil on the cell bottom. Four*.
thick concrete-filled forms of fabric, similar to canvas, wouWK
installed on the slopes, while 18 in. of screened waste would be*
stalled on the bottom and for 5 vertical feet up the side ttop*
A concrete pump would place the stabilized soil and concrt»i
while the earth moving equipment mentioned above would IB-
stall the two size fractions of screened waste.
Option D, a stabilized soil material, would place a 6-in-W^
of stabilized soil over the entire cell and an 18-in. layer of jcreene"
174
LAND DISPOSAL
-------
Table 1
Comparison of Protective Cover Options

A.
B.
C.

No Cover
Soil Cover
Cement Filled
Forms
Stabilized Soil
Cover
Approximate
Approximate
Cost
(*/sy>
0
1.2

1.8

1.2

Installation
Time
(days)
0
17

Cover
Stability

N/A
Minimal

Substantial

Adequate

Penetration
Protection

NONE
Adequate

Substantial

Increased Cell
Capacity
( cubic yards )
9,900
0

7,*00

7, tOO

Percent of Slope
Initially Covered

0
30

100

Required
Ma intenance

Substantial
Substantial

Minimal

Minimal
sy = square yard
N/A = Not Applicable

waste over the bottom and for 5 vertical feet up the slopes. The
constituents of the stabilized soil cover would be silty sand, fly
ash and water. Mixed in this way, fly ash provides a pozzolanic
effect increasing the resistance of the resulting material to both
mechanical and chemical attack. The mixture would be distrib-
uted with a concrete pump, and the screened waste would be in-
stalled by the above mentioned earth moving equipment.

COMPARISON OF OPTIONS
Each of the four options was evaluated using seven criteria:
cost, time of installation, stability, resistance to penetration,
diminishment of capacity, percent of slope initially covered and
maintenance. Increased cell capacity compares the capacity of the
cell with 2 ft of soil cover versus the proposed options. Normally,
an engineer also would consider requirements for land area and
the maximum interior slope when designing a cell. However, since
the cell was designed prior to consideration of the various altern-
atives, the latter two considerations were unrelated to this pro-
ject. Each criterion and a comparison of the options is shown in
Table 1.
Option A, no cover, would obviously cost nothing and require
no extra installation time. The increased cell capacity would be
approximately 9900 yd3, but the liner would not be protected
from penetration. Maintenance would be substantial. Due to the
U.S. EPA guidance concerning a protective cover over the liner
system, regulatory approval of Option A would be unlikely.
Option B, soil only, would cost approximately $1.2/yd2 and re-
quire approximately 17 days to install. It probably would offer
adequate penetration protection but the soil cover would not be
stable up the entire cell slope. Therefore, only about 30% of the
slope could be covered initially. Wind and water erosion would
require substantial maintenance during operations. Soil cover
has been previously examined for several similar installations.2'3
Option C, concrete-filled fabric forms, would cost approxi-
mately $1.8/yd2 and require approximately 20 days to install. The
stability and resistance to penetration would be substantial, ca-
pacity would be increased approximately 7400 yd3, all of the
slopes could be covered at one time and maintenance would be
minimal. Concrete-filled forms were previously installed to pro-
tect the slopes of an impoundment at the Oklahoma facility.
These forms continue to function as designed 3 yr after their in-
stallation with no maintenance required. Concrete-filled forms
also have been used elsewhere to protect the slopes of rivers and
other impoundments.4'5'6
Option D, stabilized soil, would cost approximately $1.2/yd2
and require approximately 26 days to install. Penetration resis-
tance and stability would be substantial, the capacity of the cell
would be increased approximately 7400 yd3, all of the liner sys-
tems would be protected, and maintenance would be minimal.
The stabilized soil also would effectively provide an additional
relatively impermeable liner.
Option A, no cover, was immediately dropped from considera-
tion because regulatory approval was doubtful and the liner
would not be protected. Therefore, only options B, C and D
were considered as viable alternatives. Option D, protective soil
cover, was selected because:
• It was the least expensive option
• Cover stability was adequate
• Resistance to penetration was substantial
• Its selection would increase capacity by approximately 7400 yd3
• All of the slopes would be initially protected
• Maintenance during operations would be minimal
• An additional relatively impermeable liner would be provided
The significant disadvantage to Option D was that it required
the most time to implement. The larger the project, the more time
such an option would require. This time increase could become
highly significant. Option D also might not be the most advan-
tageous for projects with settlement in excess of 10% since settle-
ment could fracture the stabilized soil.

MIX PREPARATION
A comprehensive strength of 150 lb/in.2 at 28 days was estab-
lished as the design criterion. Typically, an AASHTO-20 truck
has a tire pressure of 110 lb/in.2 This vehicle pressure combined
with a safety factor of 1.35 necessitated the design compressive
strength of 150 lb/in2.
A series of mixes was prepared and tested for 7 and 28 day com-
pressive strengths. Standard proctors (ASTM D-698) were pre-
pared and tested for each mix ratio. Table 2 compares the com-
position of the mixes and the associates compressive strengths;
the mix ratio compares sand to fly ash by weight.

Table 2
Test Mixes
Mix
Ratio

4tol
5tol
Component

sand
fly ash
water
sand
fly ash
water
Quantity
(grams)

180
45
33
150
30
25
7-Day Compressive
Strength (lb/in.1)

160
140
LAND DISPOSAL
175
-------
The 5 (sand) to 1 (fly ash) mix by weight fulfilled the compres-
sive strength specifications of 150 lb/in.! at 28 days. A mix ratio
of 5 to 1 by weight was established as the product specification.
The fly ash was a Type C fly ash obtained from a coal fired
power plant in eastern Oklahoma and the sand was a silty sand
mined from a nearby alluvial floodplain. Table 3 shows a typical
analysis of the fly ash while Table 4 presents a sieve analysis of
the sand.
Table 3
Typical Fly Ash Characteristics
Fineness C325 Mesh)
Fineness Variation
Moisture Content
Specific Gravity
Sp. GR. Variation
Loss on Ignition
Soundness
Lime, Pozz, 7 days
Silica SiO~
Aluminum Oxide
Ferric Oxide Fe-0,.
Total 2 J
~~
Sulphur Trioxide SO.,

Calcium Oxide CaO
ASTM C-618
Type "C"
Requirements

34% Max
5% Max
3% Max

5% Max
6% Max
0.8% Max
BOO PSI MIN
50% Min

5% Max
Measured
Values

8.33%
0.03%
0.08%
2.70
0.45%
0.31%
0.02%
993
28.63%
26.07%
6.58%
61.28%

2.47%

29.75%
Table 4
Typical Sand Characteristics
Sieve Size

#16
#30
#50
#100
#200
Percent Passing

100
96
67
16
7
COVER INSTALLATION PROCEDURES
A cross-section of the protective-cover design is shown in Fig.
1. The base consists of 6 in. of stabilized soil overlain by 9 in.
of Vi-in. screened material and 9 in. of 4-in. screened material.
The stabilized soil cover was produced with a dry-mix truck and
placed by a boom truck with a concrete pump. A three-com-
partment concrete truck was used for mixing. The sand was
stored in the gravel and sand hoppers, the fly ash was stored in the
cement compartment and water was added at the mixing screw.
The mixture was placed above the liner system utilizing a 4-in.
diameter positive displacement concrete pump.
Adequate quantities of material were stockpiled adjacent to the
mix truck to ensure continuous production. Water was provided
from an on-site wastewater treatment plant and was stored in a
10,000 gal mobile tank which was filled prior to the start of daily
production. The water level was maintained during the day using
a 4-in. diameter centrifugal pump.
A storage tank with a pneumatic conveyor which was capable
of holding approximately six truck loads of fly ash was located
adjacent to the water tank and mixer. The fly ash storage tank
was filled prior to the start of production, and the material was
replenished and maintained on a daily basis.
Three thousand cubic yards of sand were transported from the
sand borrow area and stockpiled adjacent to the mixing area prior
to project initiation. The sand was not screened. However, con-
cerns about plugging the pump necessitated the visual inspection
of the sand as it was loaded into the mixer. Any roots and other
debris were removed. A rubber-tired front-end loader was used
to move the sand from the stock piles to the mixer during pro-
duction.
The primary leachate collection system and protective cover
system were installed in conjunction with each other. The primary
leachate collection system was installed within 72 hr prior to
placement of the protective cover to prevent ultraviolet degrada-
tion of the geotextile and clogging of the drainage net by blow-
ing dirt and debris. The primary liner was swept and all debris
was removed prior to drainage net and geotextile placement.
The sand and fly ash were mixed in the mix truck at an appro-
priate mix ratio of 5 parts sand to 1 part fly ash by weight. The
loading of the sand in the mix truck was carefully monitored, and
any debris in the sand was removed. Sufficient water was added
to provide a flow which could be readily pumped. The quantity
of water utilized fluctuated depending on ambient weather con-
ditions and the distance pumped.
After mixing the stabilized soil, the mix truck deposited the ma-
terial in the concrete pump inlet. The pump distributed the mix-
ture through 4-in. diameter hoses and onto the liner system.
When placing material within approximately 100 ft of the boom
truck, the location of the material placement was varied by mov-
ing the boom in a sweeping motion across the cell bottom and
sides. When the location for material placement exceeded 100ft,
personnel placed the material by moving the hoses along the cell
sides and bottom in a fashion similar to installing spray insula-
tion. Rapid set-up times enabled the placement of a continuous
layer of the stabilized soil vertically up the cell slopes. The stabil-
ized soil was installed as a single 6-in. thick lift. The rapid set-up
times also necessitated well-organized procedures in order to pre-
vent the hardening of the mixture in the hoses.
The screened waste material was placed after the installation of
the stabilized soil layer. The waste material chosen had been
stored on an on-site waste pile. The waste material consisted pri-
marily of stabilized sludges and contaminated dirt which were
segregated through 4-in. and Vi-in. screens and stockpiled prior
to placement.
The screened material was installed by progressively dumping
the waste and pushing it onto the stabilized soil. The screened
waste was transferred from the stockpile to a dump truck with a
front-end loader; the dump truck backed to the edge of the prev-
iously deposited waste material and deposited its load on the edge
of previously placed waste. A dozer then pushed the material
onto the exposed stabilized soil layer. A 9-in. thick layer of the
3/<-in. screened material initially was placed on the cell bottom
and for 5 vertical feet up the cell slopes. Subsequently, an addi-
tional 9-in. thick layer of the 4-in. screened material was placed
above the 3/4-in. screened material. Material which did notpasi
the 4 in. screen was rejected. The rejected material was disposed
in other on-site cells or as part of the standard filling of this cell.

RESULTS
The installation of the stabilized soil layer and screened waste
required 18 and 5 days, respectively. During the initial 3 days of
placing the stabilized soil, fluctuations in the mix, clogging of the
pump by debris in the sand and rapid set-up time resulted in the
daily placement of less than 150 yd1 of material. The stabilized
soil set-up in the hoses several times during the initial pha*» of
the project. Between 150 and 250 yd' of stabilized soil were in-
T6 LAND DISPOSAL
-------
stalled during each subsequent 10-hr day; the installation rate
would have increased if two mix trucks rather than one had been
used.
Test cylinders were prepared daily to determine the compres-
sive strength of the stabilized soil. Three 3-in. by 6-in. cylinders
were prepared daily from material taken directly beneath the out-
let of the pipe. These cylinders were filled as the stabilized soil
was pumped onto the liner system. Fig. 2 shows the compressive
strength results: the 28 day average compressive strength was
185 lb/in2 with a standard deviation of 69 lb/in.2; the 84 day
compressive strength was 364 lb/in.2 with a standard deviation
of 178 lb/in.J. The 84 day compressive strength should approxi-
mate the ultimate compressive strengths.
The protective cover system was inspected 6 months after in-
stallation when the cell had been partially filled with hazardous
waste. The stabilized soil was intact on the cell slopes; cracks
600 r
500
O
z
400
2
O
O
O
HI
z
u.
z
O
O
z
300
200
100

(9)-* OF SAMPLES
7 28

AGE (DAYS)
56 84
Figure 2
Stabilized Soil Compressive Strength
were evident, however differential settlement or fracturing was
not apparent; the screened waste material had not eroded or sluf-
fed significantly; and the protective cover system had not required
any maintenance.

CONCLUSION
Various options exist for the protective cover above a liner sys-
tem in a hazardous waste landfill. Stabilized soil and screened
waste material may be the most advantageous protective system.
This option may be the best design due to relatively small cost,
adequate stability, substantial resistance to penetration, increased
cell capacity, total coverage of the slopes and minimal mainten-
ance. The disadvantages associated with such an option include
significant installation time and the possibility that projects with
extreme settlement could produce fracturing.
After considering the positive and negative attributes assoc-
iated with stabilized soils and screened waste, stabilized soil was
selected as the most feasible design. An inspection of the protec-
tive cover system 6 months after installation indicated that it con-
tinued to function as designed. In the future, engineers should
consider a protective cover system of stabilized soil and screened
waste when designing hazardous waste landfills.

ACKNOWLEDGEMENT
The authors wish to acknowledge the support and guidance of
Mr. Ken Jackson and Mr. Steve Fan during the original project
conceptualization and subsequent construction. Both gentlemen
are associated with U.S. Pollution Control, Inc.

REFERENCES
1. U.S. EPA, "Draft-Minimum Technology Guidance on Double Liner
Systems for Landfills and Surface Impoundments...Design, Construc-
tion and Operation," May 24,1985.
2. Giroud, J.P. and Ah-Line, C., "Design of Earth and Concrete Covers
for Geomembranes," Proc. International Conference on Geomem-
branes, June 1984, 487-492.
3. Rankilor, P., "Design Aspects of Flexible Revetmats Constructed on
Impermeable Synthetic Linings," Proc. International Conference on
Geomembranes, June 1984, 347-352.
4. Lamberton, B., "Fabric-Formed Revetmat Technology Opens New
Engineering Applications," Geotechnical Fabrics Report, Summer
1983.
5. Pildysh, M. and Wilson, K., "Cooling Ponds Lined with Fabric-
Formed Concrete," Concrete International, Sept. 1983.
6. Lamberton, B., personal communication, Texicon, Dec. 1, 1986.
LAND DISPOSAL 177
-------
Common Problems Associated with the Design
And Installation of Groundwater Monitoring Wells
David M. Nielsen, C.P.G.
IEP, Inc.
Westerville, Ohio
INTRODUCTION
The installation of groundwater monitoring wells for the pur-
pose of detecting trace (i.e., part-per-billion) levels of both
organic and inorganic contaminants in groundwater systems is a
common practice in many waste disposal and chemical spill
scenarios. It is estimated that tens of thousands of monitoring
wells have been designed and installed over the past five years,
and that thousands more are installed annually, many of these by
consultants and contractors who are not aware of the proper
practices for monitoring well construction. As a result, many ex-
isting and currently installed monitoring wells have critical design
or installation flaws that adversely affect the quality of ground-
water samples taken from them. Because the objective of most
groundwater monitoring programs is to obtain "representative"
groundwater samples, it is imperative that proper groundwater
monitoring well design and installation techniques, that mini-
mize the potential for sample chemical alteration, be employed.
Most groundwater monitoring well design and installation
problems can be traced back to the mistaken belief that a "cook-
book" approach which ignores site-specific hydrogeologic, geo-
graphical and contaminant-related conditions, can be used in all
situations. The fact is that each site at which a monitoring well is
installed is unique. The designer must therefore develop well de-
sign and installation specifications that both take into account
anticipated site-specific conditions and are flexible enough to
accommodate alterations necessitated by unanticipated con-
ditions encountered during drilling.
Other groundwater monitoring well design and installation
problems stem from the fact that few professionals exist who
are adequately trained and experienced in proper monitoring well
construction practices, procedures and the potential effects of
construction methods and materials on analytical results. It must
also be realized that the analytical power of modern laboratories
is now reaching the parts-per-trillion detection level, while our
means of gaining access to the subsurface to obtain groundwater
samples for analysis is crude by comparison. Still, most potential
sources of sample chemical alteration inherent in the monitoring
well construction process can be anticipated and controlled. What
is required is a workable set of flexible guidelines for monitoring
well design and installation, adaptable to a wide variety of
groundwater monitoring situations, that can be employed by con-
sultants and contractors.
The first step toward developing guidelines for monitoring well
design and installation is identifying the areas in which most prob-
This paper has been severely edited to conform to the format and limi-
tations for proceeding manuscripts. The full 90 pages of text are avail-
able as a monograph from HMCRI.
lems arise. Among the most common monitoring well design
flaws and installation problems are the following:

• Use of inappropriate well casing or well screen materials (i.e.,
materials that have not been selected to be compatible with the
hydrogeologic environment, anticipated contaminants, or the
requirements of the groundwater sampling program), resulting
in sample chemical alteration or failure of the well;
• Use of non-standard well screen (i.e., field-slotted or drilled
casing), or use of incorrect screen slot sizing practices, result-
ing in sedimentation of the well and the acquisition of turbid
samples throughout the life of the monitoring program;
• Improper length and placement of the well screen, making the
acquisition of water level or water quality data from discrete
zones impossible;
• Improper selection and placement of filter pack materials,
resulting in sedimentation of the well, plugging of the well
screen, groundwater sample chemical alteration, and poten-
tially failure of the well;
• Improper selection and placement of annular seal material!,
resulting in alteration of sample chemical quality, plugging of
the filter pack and/or well screen, or cross-contamination
from improperly sealed-off geologic units; and
• Inadequate surface protective measures, resulting in surface
water entering the well bore, alteration of sample chemical
quality, or damage to/destruction of the well.
Any one or a combination of these design/installation prob-
lems could result in a determination that a well or a series of welli
is unsuitable for obtaining representative groundwater samples.
In many cases this necessitates the abandonment of the improp-
erly designed or installed well(s) and the installation of replace-
ment wells, which can be very costly and time consuming. The
use of proper groundwater monitoring well design and initaDi-
tion practices is thus essential to ensure time- and cost-effective
acquisition of representative groundwater samples. Fig. 1 uTH-
trates the basic design components of a groundwater monitoring
well.
Proper design and installation of groundwater monitoring wdb
require a thorough review of a variety of site-specific condition!
as well as an up-to-date knowledge of well design and installatipn
practices and procedures. Site-specific design considerations in-
clude:

• The purpose or objective of the groundwater monitoring &**•
gram (i.e., water quality monitoring vs. water level monitor-
ing);
• Surficial conditions, including topography, drainage. dnMfc
seasonal variations in climate and site access;
• Known or anticipated hydrogeologic settings, including typK"
178 LAND DISPOSAL
-------
geology (unconsolidated/consolidated), aquifer physical char-
acteristics (type of porosity, hydraulic conductivity), type of
aquifer (confmed/unconfined); recharge/discharge conditions,
and aquifer interrelationships;
• Characteristics of known or anticipated contaminants (chem-
istry, density, viscosity, reactivity, concentration);
• Anthropogenic influences (man-induced changes in hydraulic
conditions); and
• Any regulatory requirements that must be met.
A unique set of site-specific design considerations will exist for
each site and, in fact, for each individual well installation, requir-
ing that each well be designed as a unique structure.

MONITORING WELL CASING AND/OR
SCREEN MATERIALS
Casing used in monitoring wells could conceivably be made of
nearly any rigid tubular material, though experience dictates that
the choices are limited to only a few materials. Casing materials
typically used in groundwater monitoring wells can be categorized
into four general types:

• Thermoplastic materials, including polyvinylchloride (PVC)
and acrylonitrile butadiene styrene (ABS);
• Fluoropolymer materials including polytetrafluoroethylene
(PTFE), tetrafluoroethylene (TFE), fluorinated ethylene pro-
pylene (FEP), perfluoroalkoxy (PFA), and polyvinylidine
fluoride (PVDF);
• Metallic materials, including carbon steel, low carbon steel,
galvanized steel, and stainless steel 304 and 316; and
• Fiberglass-reinforced materials, including fiberglass-reinforced
epoxy'(FRE) and fiberglass-reinforced plastic (FRP).
Each of these materials possesses a set of physical and chemical
characteristics that influences its use in site-specific hydrogeologic
and contaminant-related groundwater monitoring situations.
Polycinylchloride (PVC), stainless steel, and fluoropolymer cas-
ing materials will be discussed in greater detail in the following
sections; other materials are either not widely used in monitoring
wells or few data exist upon which to base a decision regarding
the selection of those materials for use in monitoring wells.
Historically, the selection of well casing materials focused on
the material's structural strength, durability in long-term ex-
posures to natural subsurface conditions, and ease of handling.
When chemical analyses of water samples taken from monitor-
ing wells began to be made at the micrograms-per-liter level, how-
ever, the focus shifted to the potential impact that casing ma-
terials may have on the chemical integrity, or "representative-
ness" of groundwater samples. It is evident that the selection of
appropriate materials for monitoring well casing must consider all
of these factors.
It is important to note that the ultimate control over casing ma-
terial selection is or should be exercised by a unique set of site-
specific and logistical factors, including the following:
Site-Specific Factors
Geologic Environment
Natural Geochemical
Environment
Anticipated Well Depths
Types of Contaminants
Logistical Factors
Well Drilling Method Used
Ease of Handling

Ease of Cleaning
Cost (materials, shipping, etc.)
There is no single casing material that is applicable over the
wide range and variety of natural and man-induced site-specific
conditions. It is therefore critical that these conditions be closely
evaluated prior to the selection of the material used for monitor-
ing well casings.
Casing/Screen Strength
Monitoring well casing and screen must have the structural
strength to withstand both the forces exerted on it by the sur-
rounding geologic materials and the forces imposed on it during
installation (Fig. 2), and it should be able to retain its structural
integrity for the expected duration of the monitoring program
under both natural and man-induced sursurface conditions.
Three components of casing and screen structural strength—
tensile strength, compressive strength, and collapse strength—
must be evaluated to determine whether a particular casing ma-
terial is suitable for a particular application.
For a monitoring well installation, the casing material selected
should have enough tensile strength to support its own weight
when suspended from the surface in an air-filled borehole.
The compressive strength of a casing or screen material is de-
fined as the load in the direction of pushing the material together
that is required to deform it. The properties of the casing or
screen parent material, specifically the yield strength and stiff-
ness, are more significant to determining compressive strength
than are the dimensional parameters, though casing wall thick-
ness is also important.
The third significant strength-related property of casing or
screen materials is collapse strength, or the capability of a casing
to resist collapse caused by any and all external loads to which it
is subjected both during and after installation. The collapse
strength of a casing material is proportional to the cube of its wall
thickness. Therefore, a small increase in wall thickness proves a
significant increase in collapse strength. A casing is most suscep-
tible to collapse during installation, when it has not yet been con-
fined and restrained by the placement around it of filter pack or
annular seal materials. Once a casing is properly installed and
confined, its resistance to collapse is enhanced to a point that
collapse is no longer a point of concern (NWWA, 1981).

Casing/Screen Chemical Resistance and
Chemical Interference
Materials utilized for monitoring well casings must be durable
enough to withstand potential chemical attack from either natural
chemical constituents or contaminants in groundwater. The prin-
cipal processes to which casing materials should be resistant are
corrosion (galvanic or electrochemical) and chemical degradation.
Typically, metallic casing materials are most subject to corrosion
and plastic casing materials are most subject to chemical degra-
dation.
Because the extent to which chemical attack occurs is primarily
dependent on the presence and concentration of certain chemical
constituents in groundwater, the selection of casing material can
only be carried out with a knowledge of existing or anticipated
groundwater chemistry. Not only may natural or man-induced
groundwater chemistry affect the structural integrity of monitor-
ing well casings, but the product(s) of casing deterioration may
adversely affect the chemistry of water samples taken from moni-
toring wells.
Materials used for monitoring well casings must not exhibit
tendency to either sorb (by adsorption or absorption) or leach
chemical constituents from or into the water taken as a sample
from the well. If the casing material chosen for a monitoring
well were to sorb constituents from water, it would essentially be
removing these constituents from the water that is then sampled,
creating so-called "false negatives." The sample taken is thus not
representative of groundwater quality, as it either completely
lacks the sorbed constitients or has had those constituents reduced
to some level below ambient. Additionally, if groundwater chem-
istry were to change over time, chemical constituents that were
previously sorbed into the casing may begin to desorb or leach
LAND DISPOSAL 179
-------
back into solution, causing slugs of those constituents to appear
in water samples at higher levels than exist in ambient water. If
casing materials leach chemical constituents from them in the
presence of aggressive aqueous solutions, those chemical constit-
uents may show up in a groundwater sample when they do not
exist in ambient water, creating so-called "false positives," or in-
dications of possible groundwater contamination. These indica-
tions do not relate to true groundwater contamination, but rather
to water sample contamination contributed by the well casing ma-
terial.
The selection of a monitoring well casing material must there-
fore consider all potential interactions between the casing material
and the natural and man-induced geochemical environment. Un-
fortunately, very little published information exists regarding
these interactions under actual field subsurface conditions.

Types of Casing Materials
Polyvinylchloride (PVC)
Polyvinylchloride (PVC) used for well casing is composed of a
rigid hardened (unplasticized) polymer formulation (PVC type 1)
LOCKING CASING CAP

INNER CASING CAP

DRAIN
SURFACE SEAL
WELL INTAKE.
BOTTOM PLUG
. FILTER PACK
Figure 1
Basic Design Components of a Groundwatering Monitoring Well
TENSILE (PULL-APART) FORCES
(critical at casing joints)
BOREHOLE -
WELL CASING •
COMPRESSIVE FORCES
(critical at higher casing weights)
I

t
GROUND SURFACE
STATIC WATER LEVEL
COLLAPSE FORCES
(critical at greater depths)
t_
BUOYANT FORCES
- WELL INTAKE
Figure 2
Forces Exerted on a Monitoring Well Casing and
Screen During Installation
i 1 1 1 1 1 1 1
i I . !
COMPRESSION LOAD - POUNDS
Figure 3
Results of Short-Term Static Compression Tests on Teflon Screen
that has high tensile, compressive and collapse strength, and has
good chemical resistance except to high concentrations of low
molecular weight ketones, aldehydes and chlorinated solvents
(Barcelona, etal., 1983).
In comparison to metallic materials, both the tensile strength
and the collapse strength of PVC well casing is relatively low,
but the light weight of PVC offsets the lower tensile strength so
that for most installations of PVC well casing, the axial loading
is not a limiting factor; the collapse strength is sufficient to with-
stand most compressive stresses encountered in a normal moni-
toring well installation. Table 1 illustrates the tensile and collapse
strengths for several common dimensions of PVC and other cas-
ing materials. Used together with the data provided in Table 2 on
weight per unit length of PVC and other well casing materials, it
is possible to calculate the maximum permissible casing string
length for PVC casing material.
Table 1
Comparative Strengths of Well Cuing Material.
CASING TENSILE STRENGTH CASINO COUArS£ STtMCYM
(poundi) (poundf ptr tquirf **ffl_
1" nominal •• nominal
POIT.inrlehlor,a. (PVCI 7M>0 11101

PVC Caiin? Jomit 1100 toSO

Slam]*,, Sl«*l ISSI 17710 91000

SS Catino. JOinu U900 117)0

Pelri«tranuoro«ihrl«n« (PTFEI No Data No Oil*

PTFE CM.ng Jo.nl! t.o 1 1,0
•SI..I citing ..t.r,.i, art ,ci,.dul. i. pla.t.c c»in9 -.l.rn., (PVC. PTFCI •'• leu***
• 'joint* ara all threaded, lluto jeml*
180
LAND DISPOSAL
-------
Table 2
Weight Per Unit Length and Specific Gravity of Well Casing Materials
(2" Nominal)
MATERIAL

PVC

STAINLESS STEEL

PTFE
WEIGHT BY SCHEDULE NUMBER (Ib/ft)
»5 »10
2.06
140
O.GS
3.00

1.21
»80
0.91
5.07

1.90
SPECIFIC
GRAVITY
1.37

N/A

2.20
Table 3
Composition of Stainless Steel Well Casing/Screen Materials
CHEMICAL COMPONENT

Carbon
Manganese
Phosphorous
Sulfur

Chromium
Nickel
Molybdenum
With respect to chemical resistance, PVC well casing is in some
ways superior to metallic materials because it is a non-conductor
and thus immune to electrochemical or galvanic corrosion. In
addition, it is resistant to biological attack, and to chemical attack
by soil, water and other naturally occurring substances present in
the subsurface (NWWA, 1981). PVC is, however, susceptible to
chemical attack by certain organic solvents. These solvents can
produce an effect called solvation, the physical degradation of
the plastic. Solvent cementing of PVC well casing is based on sol-
vation, which occurs in the presence of very high concentrations
of specific organic solvents. If these solvents, which include
tetrahydrofuran (THF), methyl ethyl ketone (MEK), methyl iso-
butyl ketone (MIBK), cyclohexanone, and dimethylformamide,
are present in high enough concentrations in a contaminant mon-
itoring situation, they could be expected to chemically degrade
PVC well casing to some degree, though the extent of the degra-
dation is not known.

LEACHING FROM PVC
Barcelona, etal. (1983) list the groups of chemical compounds
that may cause degradation of the PVC polymer matrix and/or
the release of compounding ingredients which otherwise would
remain in the rigid material. These chemical compounds include
low molecular-weight ketones, aldehydes, amines, chlorinated
alkenes and alkanes. Unfortunately, there is currently a signifi-
cant lack of information regarding critical concentrations of these
chemical compounds at which deterioration of the PVC material
is significant enough to affect either the structural integrity of the
material or groundwater sample chemical quality; there exists
even less published information regarding the performance of
PVC well casing material under actual field conditions.

SORPTION BY PVC
Miller (1982) conducted a laboratory study to determine
whether several plastics, including rigid PVC well casing, exhib-
ited any tendency to sorb potential contaminants from solution.
Under the conditions of his test, Miller found that PVC moder-
ately adsorbed tetrachloroethylene and strongly adsorbed lead,
but did not adsorb trichlorofluoromethane, trichloroethylene,
bromoform, 1,1,1-trichloroethane, 1,1,2-trichloroethane or
chromium. To determine whether or not the tetrachloroethylene
could be desorbed and recovered, only a small fraction of the
tetrachloroethylene was recovered. In a laboratory study of Park-
er and Jenkins (1986), it was found that significant losses of TNT
and HMX from solution occurred in the presence of PVC well
casing. A follow-up study to determine the mechanism for the
losses led them to attribute the losses to increased microbial de-
gradation rather than to adsorption. These studies indicate that
loss mechanisms could be attributed to biodegradation rather
than to either adsorption or absorption. Only additional research
into this question will provide a suitable answer.
It is clear that, with few exceptions, the work that has been
done to determine chemical interference effects of PVC well cas-
ing (whether by leaching from or sorbing to PVC of chemical con-
stituents) has been conducted under laboratory conditions.
Furthermore, in most of the laboratory work the PVC has been
exposed to a solution (usually distilled, deionized, or "organic-
free" water) over prolonged periods of time (several days to sev-
eral months); thus the PVC had an extended period of time in
which to exhibit sorption or leaching effects. While this may be
comparable to a field situation in which groundwater was ex-
posed to the PVC well casing as it may be between sampling
rounds, few studies consider the fact that prior to sampling, the
well casing is usually evacuated of stagnant water residing in the
casing between sampling rounds. Thus, the water that would have
been affected by any sorption or leaching effects (if they were
present at all) would ideally have been removed and replaced with
aquifer-quality water that is eventually obtained as "represen-
tative" of existing groundwater conditions. Because the sample is
generally taken immediately after the purging of stagnant water in
contact with the casing, it will have had a minimum of time with
which to come in contact with casing materials and thus be af-
fected by sorption or leaching effects. Because of this, Barcelona,
et al. (1983) suggest that the potential sample bias effects due to
adsorptive interactions with well casing materials may be dis-
counted. They correctly point out that these effects are far more
critical in sample transfer and storage procedures employed prior
to sample separation or analysis. Because no research to date has
adequately addressed this problem, it is currently unclear what
effect, if any, very short-term exposure of groundwater to PVC
well casing may have on groundwater sample integrity. However,
based on available laboratory-generated data, it seems very likely
that the effects would be subdued if present al all.

Fluoropolymer Materials (PTFE)
Four principal physical properties of polytetrafluorethylene are
(Hamilton, 1985):

• Extreme temperature range—from -400° F (-240° C) tol
+ 550 ° F (+ 287 ° C) in constant service
• Outstanding electrical and thermal insulation
• Lowest coefficient of friction of any solid material
• Almost completely chemically inert, except for some reaction
with halogenated compounds at elevated temperatures and
pressures

In addition, PTFE is flexible without the addition of plasti-
cizers, and is fairly easily machined, molded or extruded. Be-
cause PTFE never melts, molding and extruding the material
pose some difficulties not encountered with melt-processible ma-
terials. Other fluoropolymers have been designed to overcome
this process disadvantage; all are melt-processible. Despite this,
PTFE is by far the most widely used and produced fluoropoly-
mer.
For construction of groundwater monitoring wells, fluoropoly-
mers are considered to be almost completely inert to chemical
LAND DISPOSAL 181
-------
attack, even by extremely aggressive acids (i.e., hydrofluoric,
nitric, sulfuric and hydrochloric) and organic solvents. In addi-
tion, flurorpolymers are thought to neither adsorb nor absorb
chemical constituents from solution and to not leach any ma-
terials from their chemical structure. A recent study by Reynolds
and Gillham (1985) indicates, however, that at least one fluoro-
polymer (PTFE) is prone to either adsorption or absorption of
selected organic compounds, specifically 1,1,1-trichloroethane,
1,1,2,2-tetrachloroethane, hexachloroethane, and tetrachloro-
ethane; a fifth organic compound studied, bromoform, was not
absorbed by PTFE. An observation of particular note made by
Reynolds and Gillham was that tetrachloroethane was strongly
and rapidly absorbed by PTFE such that significant reductions
in concentration occurred within minutes of exposure to a solu-
tion containing the aforementioned organic compounds. These
results indicate that PTFE may not be as chemically inert as prev-
iously thought.
Dablow (1986) points out that several strength-related proper-
ties of fluoropolymers (PTFE in particular) must be taken into
consideration during the well design process, including:

1) Pull-out resistance of flush-joint threaded couplings (tensile
strength)
2) Compressive strength of the intake section; and
3) Flexibility of the casing string.

The tensile strength of fluoropolymer casing joints is the limit-
ing factor affecting the length of casing which can be supported
safely in a dry borehole. According to Dablow (1986), experi-
mental work conducted by DuPont indicates that PTFE threaded
joints will resist a pull-out load of approximately 900 pounds.
With a safety factor of two, two-inch schedule 40 PTFE well cas-
ing, with a weight of approximately 1.2 pounds per foot, should
be able to be installed to a depth of about 37S feet. Barcelona,
el al. (1985) suggest that the recommended hang length not ex-
ceed 320 feet. Data from E.I. DuPont de Nemours Company
(Table 1) suggest that the tensile strength of two-inch nominal
threaded PTFE casing joints, 540 pounds, would support 450
feet of casing in an air-filled borehole. With a safety factor of
two, this is reduced to approximately 225 feet maximum casing
string length. In any event, this is less than one-tenth the tensile
strength of an equivalent sized PVC well casing material. Addi-
tionally, because the specific gravity of PTFE is much higher
than that of other plastics (about 2.2), the buoyant force pro-
duced by water in a borehole is negligible.
Compressive strength of fluoropolymer well casings, and par-
ticularly well intakes, is also a recognized problem area. A low
compressive strength (compared to PVC) may lead to failure of
the fluoropolymer casing at the threaded joints, where the cas-
ing is weakest and the stress is greatest. Additionally, according
to Dablow (1986), the "ductile" behavior to PTFE under com-
pressive stress has resulted in the partial closing of intake open-
ings with a consequent reduction in well efficiency in deep TPFE
wells. He outlines several design considerations that can be em-
ployed to minimize this problem; the first would be to specify a
larger slot size than indicated by sieve analyses. In compressive
strength tests conducted by DuPont to determine the amount of
deformation in TPFE well intakes that would take place under
varying compressive stresses, it was determined that a linear rela-
tionship exists between applied stress and the amount of intake
deformation. This relationship is graphically presented in Fig. 3;
using this, Dablow (1986) suggests that the anticipated intake
opening deformation can be determined and included in intake
design by calculating the load, and adding anticipated intake
opening deformation to the intake opening size determined by
sieve analysis. Unfortunately, this leaves the well designer with
the problem of deciding the precise time at which well develop-
ment should be performed. If it is performed too soon, large
quantities of sediment may enter the well and potentially lodge
in the screen slots, thus reducing the efficiency of the well. If
development is attempted too late, the slots may have closed to
the point at which development becomes very difficult or is in-
effective because of the lack of screen open area.
Another design consideration for fluoropolymer well casing
and screen to minimize compressive stress problems, as noted by
Dablow (1986), is to keep the casing string suspended in the bore-
hole, so that the casing is in tension, and backfill the annulus
around the casing while it remains suspended. The intent here is
to reduce compressive stress by supplying support on the outer
wall of the casing. This could only be accomplished successfully
in relatively shallow wells, in which the long-term tensile strength
of the fluoropolymer casing was sufficient to withstand tensile
stresses imposed on the casing by suspending it in the borehole.
Additionally, continuous suspension of casing in the borehole
would be impossible with hollow-stem auger installations.
The third strength-related area of concern with respect to in-
stallation of fluoropolymer well casing is the extreme flexibility of
the casing string, which causes the casing to become bowed and
non-plumb when a load is placed on it. This may result in diffi-
culties, following well installation, in obtaining samples or accu-
rate water levels from these wells. Dablow (1986) suggests three
means of solving this problem:
1) Suspending the casing string in the borehole during back-
filling (as discussed above);
2) Using casing centralizers; or
3) Inserting a rigid PVC or steel pipe temporarily inside the
fluoropolymer casing during backfilling.

Metallic Materials
Metallic well casing and screen materials available for use in
monitoring wells include carbon steel, low carbon steel, galvan-
ized steel and stainless steel. Well casings made of any of these
metallic materials are generally stronger, more rigid and less
temperature sensitive than PVC or fluoropolymer casing ma-
terials are sufficient to meet virtually any subsurface condition
encountered in a groundwater monitoring situation. However, all
metallic materials are subject to corrosion, a chemical resistance
and chemical interference problem that may also affect casing
strength, in long-term exposures to certain subsurface geochem-
ical environments.
Corrosion of metallic well casings and screens can both limit
the useful life of the monitoring well installation and result in
groundwater sample analytical bias. It is important, therefore, to
select both casing and screen that are fabricated of corrosion-
resistant materials.
To determine the potential for corrosion of metallic materials,
it is first necessary to determine natural geochemical conditions.
The following list of indicators of corrosive conditions can help
recognize potentially corrosive conditions (modified from John-
son Div., 1966):

• Low pH—if groundwater pH is less than 7.0, water is acidic
and corrosive conditions exist
• High dissolved oxygen content—if dissolved oxygen content
exceeds 2 m/1, corrosive water is indicated
• Presence of hydrogen sulfide (H2S)—presence of H2S in quan-
tities as low as 1 m/1 can cause severe corrosion
• Total dissolved solids (TDS)—if TDS is greater than 1000 m/1,
the electrical conductivity of the water is great enough to cause
serious electrolytic corrosion
• Carbon dioxide (COj)—corrosion is likely if the CO2 content
182 LAND DISPOSAL
-------
of the water exceeds 50 m/1
• Chloride ion (Cl -) content—if Cl - content exceeds 500 m/1,
corrision can be expected
Combinations of any of these corrosive agents generally in-
crease the corrosive effect. To date, however, no data exist on the
expected life of steel well casing materials exposed to natural sub-
surface geochemical conditions, perhaps because the range of
subsurface conditions is so wide and unpredictable.
Clearly the presence of corrosion products represents a high
potential for the alteration of groundwater sample chemical quali-
ty; the surfaces on which corrosion occurs also present potential
sites for a variety of chemical reactions and adsorption to occur.
These surface interactions can cause significant changes in dis-
solved metal or organic compounds in groundwater samples
(March and Lloyd, 1980). According to Barcelona et al. (1983),
even flushing the stored water from the well casing prior to sam-
pling may not be sufficient to minimize this source of sample bias
because the effects of the disturbance of surface coatings or ac-
cumulated corrosion products in the bottom of the well would be
difficult, if not impossible, to predict. On the basis of these obser-
vations, the use of carbon steel, low-carbon steel and galvanized
steel in monitoring well construction should be discouraged in
most natural geochemical environments.
On the other hand, stainless steel performs very well in most
corrosive environments, particularly under oxidizing conditions.
In fact, stainless steel requires exposure to oxygen in order to at-
tain its highest corrosion resistance; oxygen combines with part of
the stainless steel alloy to form an invisible protective film on the
surface of the metal. As long as the film remains intact, the corro-
sion resistance of stainless steel is very high.
Several different types of stainless steel alloy are available; the
most common are Type 304 and Type 316. Chemical composition
of both types of stainless steel is presented in Table 3. Type 304
stainless steel is perhaps the most practical material from a corro-
sion resistance and cost standpoint. The chromium and nickel
give the 304 alloy excellent resistance to corrosion; its low carbon
content improves its weldability. Type 316 stainless steel is com-
positionally similar to Type 304 with one exception—a 2-3%
molybdenum content and a higher nickel content (replacing the
equivalent percentage of iron). This compositional difference
gives Type 316 stainless steel an improved resistance to sulfur-
containing species as well as sulfuric acid solutions (Barcelona, et
al., 1983), so it performs better under reducing conditions than
Type 304. According to Barcelona, et al. (1983), Type 316
stainless steel is less susceptible to pitting or pinhole corrosion
caused by organic acids or halide solutions. However, they also
point out that for either formulation of stainless steel, long-term
exposure to very corrosive conditions may result in corrosion and
the subsequent chromium or nickel contamination of samples.

MONITORING WELL ANNULAR SEAL
DESIGN AND PLACEMENT
Any annular space that is produced as the result of the installa-
tion of well casing in a borehole provides a potential channel for
downward movement of water and/or contaminants unless the
annulus is sealed. In any casing/borehole system, there are several
potential pathways for water and contaminants to follow (Fig.
13). One pathway is through the seal material; if the material is
not properly formulated and installed or if it cracks or
deteriorates after emplacement, the permeability in the vertical
direction could be significant. Because casings are relatively
smooth, another potential pathway exists between the casing and
seal material. This pathway could occur because of any of several
reasons, including (1) temperature changes of the casing and seal-
ing material (principally neat cement) during the curing or setting
of the sealing material, (2) swelling and shrinkage of the sealing
material while curing or setting, or (3) poor bonding between the
sealing material and the casing (Kurt and Johnson, 1982). A third
pathway, resulting from improper emplacement of seal materials
(i.e., resulting from bridged annular seal materials) may also ex-
ist. All of these pathways can be anticipated and avoided with
proper annular seal formulation and placement methods.
The annular seal in a monitoring well (i.e., the sealing material
placed above the filter pack in the annulus between the borehole
and the well casing) serves several purposes, namely (1) to provide
protection against infiltration of surface water and potential con-
taminants from the ground surface down the casing/borehole an-
nulus, (2) to seal off discrete sampling zones, both hydraulically
and chemically, and (3) to prohibit vertical migration of water,
which may be of different quality, in the casing/borehole annulus
from one aquifer to another or from zones of high hydraulic head
to zones of low hydraulic head. Such vertical movement can cause
cross contamination, which can greatly influence the represen-
tativeness of groundwater samples, and can cause an anomalous
hydraulic response of the monitored zone, resulting in distorted
maps of potentiometric surfaces. The annular seal around the cas-
ing also increases the life of the casing by protecting it against
exterior corrosion or chemical degradation, and may provide an
element of structural integrity. A satisfactory annular seal should
result in complete filling of the annular space and must envelope
the entire length of the well casing (exclusive of the filter pack) to
ensure that no vertical migration occurs within the borehole.
The annular seal may be comprised of several different types of
permanent, stable, low-permeability materials including bentonite
(pelletized, granular or powdered) and neat cement grout, and
variations of both. The most effective seals are obtained by using
expanding materials that will not shrink away from either the cas-
ing or the borehole after curing or setting. Bentonite, expanding
neat cement, or mixtures of neat cement and bentonite are among
the most effective materials for this purpose (Barcelona, et al.,
1983; 1985). If the casing/borehole annulus is backfired with any
other material (i.e., recompacted drill cuttings, sand or borrow
material), a low permeability seal cannot be ensured, and the
borehole may then act as a conduit for vertical migration of
potentially contaminated groundwater. This is especially true
regarding the use of drill cuttings, because recompacted drill cut-
tings will always have a higher permeability than the natural for-
mation materials from which they were derived.

Surface Completion of a Monitoring Well:
Protective Measures
Two types of surface completions are common for ground-
water monitoring wells: the above-ground completion, which is
preferred wherever practical, and the flush-to-ground-surface
completion, which may be required under some site conditions.
The primary purposes of either type of completion are to prevent
surface runoff from entering and infiltrating down the annulus of
the well, and to protect the well from accidental damage or van-
dalism.
Whichever type of completion is selected for any given well,
there should always be a surface seal of neat cement or concrete
surrounding the well casing and filling the annular space between
the casing and borehole at the surface. The surface seal may be an
extention of the annular seal installed above the filter pack, or it
may be a separate seal emplaced atop the annular seal. Because
the annular space is generally larger and the surface material adja-
cent to the borehole more highly disturbed from drilling at the
surface than at depth, the surface seal will generally extend to at
least one to two feet away from the well casing at the surface; the
LAND DISPOSAL
183
-------
seal will usually taper down to the size of the borehole within a
few feet of the surface. Some well installers prefer to mound the
cement surface seal around the well casing or protective casing, to
account for shrinkage of the cement and to provide a gentle slope
away from the well to discourage surface runoff from entering the
wellbore. The mound, however, should be limited in size and
slope so that access to the well is not impaired. In climates in
which alternating freezing and thawing conditions are expected, it
becomes necessary to install the cement surface seal so that it ex-
tends below the frost depth to prevent potential well damage caused
by frost-heaving. In some states, well installation regulations
developed for water supply wells, but also applied to monitoring
wells, may require that the cement surface seal extend to greater
depths (i.e., 10 ft or more) to ensure sanitary protection of the
well.

REFERENCES
Ahrens, T.P., 1970, "Basic Considerations of Well Design: Part III,"
Water Well Journal. Vol. 24, No. 3, 47-51.
Ahrens, T.P., 1957, "Well Design Criteria: Part One," Water Well
Journal. Vol. 11, No. 5, 13-30.
Anderson, D.C., Brown, J.W. and Green, J.W., 1982, "Effect of Or-
ganic Fluids on the Permeability of Clay Soil Liners;" Land Disposal
of Hazardous Waste: Proc., U.S. EPA Report No. EPA-600/9-82-
002.
Barcelona, M.J., Gibb, J.P., Helfrich, J.A. and Garske, E.E., 1985<>,
Practical Guide for Ground Water Sampling; Illinois Dept. of Energy
and Natural Resources, Water Survey Div., Champaign, IL, SWS
Contract Report No. 374, 95 pp.
Barcelona, M.J., Helfrich, J.A. and Garske, E.E., 1985", "Sample
Tubing Effects of Ground Water Samples;" Analytical Chemistry,
Vol. 5, 460-464.
Barcelona, M.J., 1984, "TOC Determinations in Ground Water,"
Ground Water, Vol. 22, No. 1, 18-24.
Boettner, E.A., Ball, G.L., Hollingsworth, J. and Aquino, R., 1981,
"Organic and Organotin Compounds Leached from PVC and CPVC
Pipe," U.S. EPA Report No. EPA-600-1-81-062, 102 pp.
Brown, K.W., Green, J.W. and Thomas, J.C., 1983, "The Influence of
Selected Organic Liquids on the Permeability of Clay Liners," Land
Disposal of Hazardous Waste: Proceedings, U.S. EPA Report No.
EPA-600-9-31-018, 114-125.
California Department of Health Services (DOHS), 1985, "The Cali-
fornia Site Mitigation Decision Tree;" California DOHS, Toxic Sub-
stances Central Div. Draft Working Document.
Campbell, M.D. and Lehr, J.H., 1973, Water Well Technology; Mc-
Graw-Hill Book Co., New York, NY, 681 pp.
Campbell, M.D. and Lehr, J.H., 1975, "Well Cementing," Water Well
Journal, Vol. 29. No. 7. 39-42.
Curran, C.M. and Tomson, M.B., 1983, "Leaching of Trace Organic:
Into Water From Five Common Plastics;" Ground Water Monitoring
Review, Vol. 3, No. 3,68-71.
Dablow, J., 1986, "Design and Installation of Poly (Tetrafluoroethy-
lene) Resin Monitoring Wells;" Proc. ASTM Symposium on Field
Methods for Ground Water Contamination Studies, ASTM Special
Technical Publication.
Dunbar, D., Tuchfeld, H., Siegel, R. and Sterbentz, R., 1985, "Ground
Water Quality Anomalies Encountered During Well Construction,
Sampling and Analysis in the Environs of a Hazardous Waste Man-
agement Facility;" Ground Water Monitoring Review, Vol. 5, No. 3,
70-74.
Hamilton, Hugh, 1985, "Selection of Materials in Testing and Purifying
Water;" Ultra Pure Water, Jan./Feb. 1985, 3 pp.
Johnson Div., U.O.P., 1966, Ground Water and Wells; Edward E. John-
son, Inc., St. Paul, MN, 440 pp.
Johnson, R.C., Jr., Kurt, C.E. and Dunham, C.F., Jr., 1980, "Well
Grouting and Casing Temperature Increase*;" Ground Water, Vol. 18,
No. 1, 7-13.
Junk, G.A., Svec, J.J.. Vick, R.D. and Avery, M.J., 1974, "Contam-
ination of Water by Synthetic Polymer Tubes;" Environmental Sci-
ence and Technology, Vol. 8. No. 13, 1100-1106.
Kurt, C.E. and Johnson, R.C., Jr., 1982, "Permeability of Grout Seals
Surrounding Thermoplastic Well Casing;" Ground Water, Vol. 20,
No. 4, 415-419.
Lerch, W. and Ford, C.L., 1948, "Long-Time Study of Cement Per-
formance in Concrete, Chapter 3—Chemical and Physical Tests of
the Cements;" J. American Concrete Institute, Vol. 19, No. 8.
Marsh, J.M. and Lloyd, J.W., 1980, "Details of Hydrochemical Varia-
tions in Flowing Wells;" Ground Water, Vol. 18, No. 7, 366-373.
Miller, G.D., 1982, "Uptake and Release of Lead. Chromium and Trace
Level Volatile Organics Exposed to Synthetic Well Casings;" Proc:
Second National Symposium on Aquifer Restoration and Ground
Water Monitoring, National Water Well Association, Worthington,
OH, 236-245.
Moehrl, K.E., 1964, "Well Grouting and Well Protection;" J. American
Water Works Association, Vol. 56, No. 4, 423-431.
Morrison, R.D., 1984, "Ground Water Monitoring Technology: Pro-
cedures, Equipment and Applications;" Timco Mfg., Inc., Prairie
Du Sac, WI. Ill pp.
National Water Well Association and Plastic Pipe Institute, 1981, Man-
ual on the Selection and Installation of Thermoplastic Water WeU
Casing; National Water WeU Association, Worthington, OH, 64.

Parker, L.V. and Jenkins, T.F., 1986, "Suitability of Polyvinyl Chloride
Well Casings for Monitoring Munitions in Ground Water;" Ground
Water Monitoring Review, Vol. 6, No. 3, 92-98.
Ramsey, R.H. and Maddox, G.E., 1982, "Monitoring Ground Water
Contamination in Spokane County, Washington;" Proc.: Second
National Symposium on Aquifer Restoration and Ground Water
Monitoring, National Water WeU Association, Worthington, OH,
198-204.
Reynolds, G.W. and Gillham, R.W., 1985, "Absorption of Halogenated
Organic Compounds by Polymer Materials Commonly Used in Ground
Water Monitoring;" Proc., Second Canadian/American Conference
on Hydrogeology. National Water Well Association, Dublin, OH,
125-132.
Sosebee, J.B., Geiszler. P.C., Winegardner, D.L. and Fisher, C., 1983,
"Contamination of Ground Water Samples With Poly (Vinyl Chloride)
Adhesives and Poly (Vinyl Chloride) Primer From Monitor WeOs;"
Proc., ASTM Second Symposium on Hazardous and Industrial Solid
Waste Testing, ASTM Special Technical Publications #805, 38.
Tomson, M.B., Hutchins, S.R., King, J.M. and Ward, C.H., 1979,
"Trace Organic Contamination of Ground Water: Methods for
Study and Preliminary Results;" Third World Conference on Water
Resources, Mexico City, Mexico, Vol. 8, 3701-3709.
Troxell, G.E.. Davis, H.E. and Kelly, J.W., 1968, Composition and
Properties of Concrete; McGraw-Hill Book Co., New York, NY.
U.S. EPA, 1975. Manual of Water Well Construction Practices; Report
No. EPA-570/9-75-991, 156 p.
Verbeck, G.J. and Foster, C.W., 1950, "Long-Time Study of Cement
Performances in Concrete with Special Reference to Heats of Hydra-
tion;" Proc., American Society for Testing and Materials, Vol. 50.
Villaume, J.F., 1985, "Investigations at Sites Contaminated with Dense
Non-Aqueous Phase Liquids (NAPLS);" Ground Water Monitoring
Review, Vol. 1, 60-74.
Walker, W.H., 1974, "Tube Wells, Open Wells and Optimum Ground
Water Resource Development;" Ground Water, Vol. 12, No. 1,10-15.
Williams, E.B., 1981, "Fundamental Concepts of Well Design;" Crowd
Water, Vol. 19, No. 5, 527-542.
•These references are cited in the total text which will be published separately
monograph by HMCR1.
184 LAND DISPOSAL
-------
Determination of the Capacity of Landfill Site Soils
To Attenuate Leachate Components
Michael L. Grosser
Donohue & Associates, Inc.
Sheboygan, Wisconsin
ABSTRACT
From the late 1960s to the late 1970s, a landfill of approximate-
ly 8 acres located in southern Wisconsin accepted industrial
waste. The landfill is closed in accordance with Wisconsin DNR
regulations, and a groundwater monitoring program has shown
no adverse environmental impact from the site. The purpose of
our investigation was to quantify the capacity of site soils to at-
tenuate components in the landfill leachate.
Our approach was an adaptation of the adsorption isotherm
methods under development by investigators at the Illinois State
Geological Survey and reported in papers by Ainsworth, et al.,
Roy et al. and Griffin, et al. In this method, soil and leachate are
contacted at different ratios for 24 hr. The soil is then removed by
filtration or centrifugation from the leachate, and the concentra-
tions of components remaining in the leachate are measured. The
adsorption isotherm is obtained by plotting the mass of com-
ponent removed per unit mass of soil versus the equilibrium con-
centration of the component in solution. The isotherm curve
allows calculation of removal characteristics throughout the range
of concentrations tested and an estimation of the retardation
coefficient for each component.
This paper presents the approach and experimental procedures
used and the resulting adsorption curves and retardation coeffici-
ents calculated for major leachate components. Discussions of the
advantages and limitations of the procedures and sample handling
considerations also are presented.

INTRODUCTION
The technical literature contains many studies which show that
soil can remove cations, anions and neutral organic compounds
from water solutions. Much of the early work is reported in the
soil science literature and addresses the chemical association of
fertilizer chemicals and pesticides with soil. Papers by Kamprath,
et al.' and Lambert, et al. ,2 are examples of this work. During the
past 15 yr, a number of investigators have studied the ability of
soils to remove components found in leachates from municipal
and industrial landfills. Some studies have shown that the
removal efficiencies can be described by adsorption isotherm
equations usually associated with the removal of organic com-
pounds by activated carbon. A discussion of these isotherm equa-
tions is provided by Barrow.3 Other workers have applied parti-
tion theory by modeling the soil as a chromatographic column.
This model is discussed by Lambert, et al.2
The soil/leachate systems are chemically complicated and dif-
ficult to describe by theoretical models. The following four com-
plications are most obvious:
• The liquid phase, leachate, can contain many chemical com-
ponents that compete for "adsorption sites" on the soil solid
phase.
• Some leachate components are affected by removal mechan-
isms that are difficult to differentiate experimentally from ad-
sorption mechanisms. An example is precipitation.
• The solid phase, soil, is difficult to define chemically.
• Many of the removal mechanisms are not well understood;
therefore, it is difficult to predict the performance of a system
by analogy to another system that has been studied. A number
of the mechanisms have been discussed by Roy, et al.4-5
Despite the practical and theoretical difficulties, progress has
been made in developing methods to study the ability of soils to
attenuate components in leachates. The approach used in this
study is an adaptation of the adsorption isotherm method under
development by investigators at the Illinois State Geological
Survey and reported in papers by Ainsworth, et al.,* Roy, et al.5
and Griffin, et al.1
Soil and leachate were contacted at different ratios for 24 hr.
Following this contact period, the soil was removed from the
leachate, and the concentrations of components remaining in the
leachates were measured. The adsorption isotherm was obtained
by plotting the mass of component removed per unit mass of soil,
versus the concentration of the component in solution.
Often the relationship between the amount of a component
that "adsorbs" to the soil and the concentration remaining in the
leachate can be described by a Freundlich-type equation:
S =
(1)
where S is the weight of component removed per unit weight of
soil, C is the equilibrium concentration in the leachate (the con-
centration after adsorption has occurred) and Kf and N are con-
stants for the system for a specific component. If N is nearly equal
to 1, Kf can be considered a partition coefficient for the compon-
ent between the soil and leachate phases and the equation simpli-
fies to:
S = KDC
(2)
In either case, the values of N and Kf (or KD if N is equal to 1)
can be obtained from the isotherm curves. Retardation coeffici-
ents are then obtained from one of the following two equations:
RF = 1 +
- 1
(3)

(4)
where Q is the soil bulk density, 6 is the effective porosity, Co is
the concentration of the component in the leachate and RD and
RF are retardation coefficients based on the partition model and
LAND DISPOSAL
185
-------
the Freundlich model, respectively.
The velocity of the component can be related to water flow
velocity by the equation:

Vwater (5)
component
where V
component
is the velocity of the component, V waler is the
flow velocity of the groundwater and R is the retardation coeffi-
cient.
In one method under development at the Illinois State Geologi-
cal Survey, the ratio of leachate components to soil is varied by
diluting the leachate and maintaining a constant liquid to soil
ratio. For our initial tests, the ratio was varied by maintaining a
constant volume of leachate and varying the amount of soil. For
the followup tests, the ratio was varied by diluting the leachate to
obtain information at low total organic carbon and chemical oxy-
gen demand concentrations. A discussion of the differences be-
tween these methods is provided by Ainsworth, el at.*
Throughout the study, we attempted to maintain the leachate
under anaerobic conditions to simulate the reducing environment
of the leachate in the landfill.

EXPERIMENTAL PROCEDURE
Sample Collection
Leachate from the landfill leachate monitoring wells was col-
lected in 5-gal glass carboys. Prior to sampling, the carboys were
washed with detergent and rinsed with tap water followed by
distilled water. Immediately prior to filling, each carboy was
purged with a mixture of carbon dioxide and nitrogen gas to
remove oxygen and provide a concentration of carbon dioxide
similar to the composition of landfill gas: 33% CO2 and 67% N2.
Samples were collected using a bailer. The leachate was trans-
ferred from the bailer to the bottle with a minimum of splashing
to reduce contact with oxygen. The leachate composite was grey
to green-grey and turbid.
Soil samples were collected from the base of a proposed landfill
expansion area. The site is underlain by glacial till of the Oak
Creek Formation. Several tills are present in the Oak Creek For-
mation, and it can be expected that more than one till is present
beneath the site. The Oak Creek Formation includes fine textured
glacial till, lacustrine clay, silt, sand and some glaciofiuvial sand
and gravel. The till is strongly calcareous and fine grained, com-
monly containing 80 to 90% silt and clay. The texture ranges from
silty clay to clay loam and silty clay loam to silt loam. Illite is the
dominant clay mineral in the less than 2-micron fraction.8'9 The
first sample was taken 1 ft below the existing grade. A second
sample was collected from a point 3 ft below the existing grade.
The samples contained no topsoil and were predominantly fine
grain material classified CL under the Unified Soil Classification
System.
The leachate samples and soil samples were returned to the
laboratory and stored for 4 days prior to beginning the ex-
periments. The samples were stored with sealed stoppers at 60 °F
to maintain environmental conditions similar to the landfill.

Sample Preparation
Because the purpose of this study was to determine the removal
of soluble components in the leachate, the leachate suspended
solids were allowed to settle in the collection bottles for 4 days.
Following this settling period, the supernate in each carboy was
relatively clear. The leachate supernate was not filtered prior to
use because we wanted to minimize handling of the leachate and,
in particular, minimize the chance that it would contact air.
The soil samples were mixed and crushed by hand to obtain pea
size and smaller particles. A sample of each soil was collected and
analyzed for moisture content to allow later determination of the
amount of dilution of leachate components by soil water.

Initial Adsorption Tests
Each leachate bottle was fitted with a glass and plastic tubing
siphon line to allow removal of the leachate from the bottle. In all
cases, the line in contact with the leachate in the carboy was glass.
A tent was constructed over all four leachate carboys using plastic
sheets, and the air within the tent was purged with the CO2/N2
gas mixture to exclude air. In all experiments, equal amounts of
leachate supernate were siphoned from each bottle to form a com-
posite.
A SOO-ml aliquot of each leachate supernate was siphoned into
a jar. This leachate composite was submitted to the laboratory for
analysis.
Mixtures of leachate and soil were prepared by the following
steps:
• A 1- gal wide mouth glass jar was purged with the CO2/N2gaj
mixture and immediately capped.
• Soil from one of the samples was quickly weighed and placed
into the jar, and the jar was capped again.
• Leachate supernate from each leachate sample carboy mi
siphoned into a graduated cylinder, [hiring the siphoning, the
graduated cylinder was purged with the CO2/N2 gas mixture.
Immediately after obtaining the proper amount of leachate
supernate, the mixture was transferred to the jar containing
soil. The jar was once again purged and capped, and the cap
was sealed with plastic tape.
• Soil blanks were prepared by adding soil and distilled water to
two glass jars.
• Two leachate blanks were prepared. One leachate blank was
prepared with the above procedure to exclude air. No soil was
added to the jar. A second leachate blank was prepared as
above, except the jar was kept open to the atmosphere. The
purpose of this procedure was to determine the impact of oxy-
gen on leachate components.
Table 1 shows the amount of soil, leachate and water added to
Table 1
Initial Adsorption Tests Experimental Data
Jar
No.
I
1
3
4
5
6
7
8
9
10
11
12
13
14
Soil
Sample
No.
1
1
2
1
1
1
2
1
1
1
1
2
None
None
Soil Weight
(wet Might)
qrau
167.0
167.0
167.0
333.2
666.2
666.4
666.2
1667.2
1331.1
2666.0
1667.0
1667.0
0.0
0.0
Soil Weight
(dry weight)
qrau
144.6
144.6
143. 8
288.6
S76.9
S76.9
S71.S
1443.7
1154.5
2308.7
1443.7
1430.6
0.0
0.0
Leachate
Voliwe
ml
2400
2400
2400
2400
2400
2400
2400
2400
1200
1200
0
0
2400
2400
DUtilled
Hater
Voluae
•1
0
0
0
0
0
0
0
0
0
0
2400
2400
0
0
Note: Jar Nos. 1-13 purged with CO2/N2. closed
Jar No. 14, open
186 LAt^D DISPOSAL
-------
each jar. The mixture provided a range of leachate to soil ratios
from 14.4:1 to 0.45:1 on a wet soil weight basis and from 16.6:1 to
0.5:1 on a dry soil weight basis.
All sample bottles were mixed for 24 hr on a rotating mixer
operating at 20 rev/min. The samples were removed from the
mixer and allowed to stand to settle the soil. At the end of the
24-hr mixing period, the samples were homogeneous with no large
soil particles.
The samples were allowed to settle for 2 days, and the superna-
tant from each jar was poured into a Millipore pressure filter tank
and filtered through a combination glass fiber filter and Millipore
0.45 micron membrane filter. The filter system was pressurized
with nitrogen gas.
By the end of the test period, the leachate sample that was open
to the atmosphere had become black and turbid with suspended
solids. There was essentially no color change or apparent increase
in suspended solids in the leachate maintained in anaerobic condi-
tions. A sample of the aerobic leachate was collected for sus-
pended solids analysis prior to filtration. All other analyses were
run on the filtrates.

Followup Tests (TOC and COD Adsorption)
Additional adsorption tests were run to further define removal
efficiencies for COD and to determine the removals of TOC. The
procedures were Identical to those used in the initial tests except
the leachate was diluted with clean well water to allow definition
i6f the adsorption curve at lower concentration. Experimental
data are presented in Table 2.

Table 2
Followup Adsorption Tests Experimental Data
Jar
No.
IF
2F
3F
4F
5F
6F
If
8F
9F
10F
IIP
12F
13F
14F
15P
16F
17F
18F
Soil
Sample
No.
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
None
3
3
Soil Height
(wet weight)
grams
1666.9
1666.9
1331.5
2661.1
2660.0
2660.0
2660.9
2660.6
2660.5
2666.0
2660.9
2660.0
2660.1
2660.4
2660.3
0.
1667.0
1667.1
Soil Weight
(dry weight)
grams
1425.2
1425.2
1138.4
2275.2
2274.3
2274.3
2275.1
2274.9
2274.7
2274.3
2275.1
2274.3
2274.4
2274.6
2274.6
0.
1425.3
1425.3
Leachate
Volume
ml
2400
2400
1200
1200
1200
600
600
300
300
150
150
50
50
10
10
1500
0
0
Clean Nell
Water
Volume
ml
0
0
0
0
0
600
600
900
900
1050
1050
1150
1150
1190
1190
0
2400
2400
RESULTS

Leachate Control Samples
The change in leachate quality throughout the initial adsorp-
tion test period where no soil was added to the jars is shown in
Table 3. The data provide the following information:

• The suspended solids concentration in the anaerobic leachate
supernate after 4 days of settling was 35 mg/1. Visual observa-
tions indicated little change over the period. The leachate com-
posite sample that was allowed to contact the atmosphere in-
creased to 184 mg/1 suspended solids, and the color changed
from a green/gray to black.
• The BOD of the leachate supernate sample decreased by 63
mg/1 over 7 days. This decrease in BOD is probably due to the
removal of the 35 mg/1 of suspended solids. The decrease in
COD over the period was 100 mg/1 and is probably attributable
to the same factor. However, there was essentially no change in
BOD or COD when the leachate was allowed to contact the
atmosphere. The obvious chemical change when the leachate
contacts oxygen apparently is not due to a biological reaction
or to a physical loss in oxidizable components to the atmos-
phere.
The concentration of arsenic, cadmium, chromium, copper,
mercury, selenium, silver, fluoride and sulfate are very low in
the soluble portion of the leachate. Only cadmium and chrom-
ium exceed the NR 140 enforcement standard for groundwater.
If the leachate did migrate, dilution and dispersion would re-
duce the concentration to below the enforcement standards,
therefore, attenuation for these components is less critical.
There was a small removal of most inorganic leachate com-
ponents in the leachate blank. Once again, this removal can be
attributed to the removal of suspended solids during the filter-
ing operation. Barium, iron and zinc maintained significant
concentrations throughout the experiment in the anaerobic
leachate.
There is a significant difference in the composition of the
leachate exposed to the atmosphere and the leachate main-
tained anaerobic throughout the experiment. Specifically,
these differences are the much lower concentrations of iron
and zinc in the aerobic filtrate and the higher leachate pH. The
removal of soluble iron can be explained by the oxidation of
ferrous ions to ferric ions and the precipitation of ferric hy-
droxide.

Table 3
Landfill Leachate Quality
Initial Adsorption Tests

Leachate Supernate Leachate Supernate Leachate Supernate
Composite After Composite After Composite After
4 Days Settling 1 Days and Filtered 7 Days and Filtered
(Anaerobic) (Anaerobic) (Aerobic)
mq/1
mq/1
Aluminum
BOD5
COD
Suspended
Solids
Dissolved
Solids
Phosphorus
Arsenic
Barium
Cadmium
Calcium
Chromium
Copper
Iron
Magnesium
Manganese
Mercury
Nickel
Selenium
Silver
Sodium
Zinc
Alkalinity
Hardness
Chloride
Fluoride
Sulfate
pa
2.
163
930

3830
0.
0.
1.
0.
65.
0.
0.
32.
133
0.
0.
0.
0.
0.
712
5.
2210
930
570
1
<5
7.
30

51
003
42
02
4
06
05
4

26
001
36
001
01

11
1.
100
832

Clear

No Data
0.
0.
1.
0.
80
0.
0.
22.
121
0.
0.
0.
0.
0.
672
4.
2160
898
No Data**
0.
<5
6.
30

27
OOS
21
01

05
05
1

20
0005
27
001
02

8
0
86
827

Clear*

No Data
0
0
0
0
55
0
0
0
125
0
0
0
0
0
672
0
2140
780
No Data**
1
-------
Soil-Leachate Mixture Results
from Initial Adsorption Tests

Odor and Appearance
Observations concerning the odor and appearance of leachate
supernate following contact with soil are presented in Table 4.
Significant odor and color removal were observed at the lower
leachate to soil ratios. At a leachate to soil ratio of 0.5:1, the char-
acteristic leachate odor was nearly gone and only a slight yellow
color remained.
Table 4
Odor and Appearance of Filtered Samples After Treatment
Jar

1
2
3
4
5
6
7
8
10

11
12
13
14
Leachate:
Soil Ratio
Odor
16.61 Strong leachate
16.61 Strong leachate
16.68 Strong leachate
8.32 Strong leachate
4.16 Strong leachate
4.16 Strong leachate
4.19 Strong leachate
1.67 Soil-like slight
leachate
1.04 Soil-like, slight
leachate
0.52 Soil-like, very
slight leachate
Soil blank Soil-like
Soil blank Soil-like
Leachate Strong leachate
blank
(anaerobic)
Leachate Strong leachate
blank
(aerobic)
Appearance
Yellow - strong
Yellow strong
Yellow - strong
Yellow - strong
Yellow lighter than 4
Yellow lighter than 4
Yellow lighter than 4
Slight yellow

Slight yellow

Very slight yellow

Hater white
Water white
Yellow/green intense
Black with solids
Chemical Components
The data and calculations for the leachate components follow-
ing contact with soils are provided in Appendix A (Tables A1 to
A7). The amount of the component removed per unit weight of
soil was calculated using the difference between the initial and
final solution concentration and correcting for soil moisture.

pH Change
The change in pH as a function of leachate:soil ratio is shown
in Fig. 1. There is an increase in pH as the leachate/soil ratio
decreases. At the higher leachate:soil ratios, the pH ranges from
6.9 to 7.1 The soil water blank had a pH of 7.9. The relatively low
pH of the leachate/soil systems suggests that precipitation is not a
significant mechanism for removal of soluble metals, at least
under anaerobic conditions, because the solubility of metals is
relatively high under these conditions.
10 20 30 4O 90 60 70 60 90 100
EQUILIBRIUM CONCENTRATION OF BOO MG/L

Figure 2
Removal Isotherm—BOD

Removal Isotherms
The removal isotherms for BOD, COD, barium, iron, zinc and
TOC are shown in Figs. 2 through 8, and the calculations are
shown in the Appendix tables. Calculated isotherm and partition
O 00 20O SCO 4OO SCO 600 TOO BOO 900

EQUILIBRIUM CONCENTRATION OF COO MG/L

Figure 3
Removal Isotherm—COD
10
i
12
I
14
I
16
I
18
T
20
LEACHATE: SOIL RATIO, WEIGHT BASIS
02 03 04 os oe or o> °>

EQUILIBRIUM CONCENTRATION OF BARIUM MG/L
Figure 1
System pH as a Function of LeachaterSoil Ratio
Figure 4
Removal Isotherm—Barium
188 LAND.DISPOSAL
-------
coefficients and retardation coefficients are given in Table 5.
The correlation coefficients show a good fit to the Freundlich-
type equation for COD (low level experiments), zinc and TOC.
The COD data also fit the linear model quite well. The fit of
barium and iron data was reasonably good for the Freundlich
model. The BOD correlation was not good. In general, the data
fit the Freundlich model better than the linear equation.
Retardation coefficients calculated for COD (low level) and
TOC were similar with values ranging from 4.5 to 6.4. The metals
had higher retardation factors with barium the lowest of the
metals at 115 and zinc the highest at 100,000.
04-
0 02 0.4 0.6 Q8 10 12 14 L6 18 2.0 i2 2.4 2.6 2-8 SO
EQUILIBRIUM CONCENTRATION OF IRON MG/L
Figure 5
Removal Isotherm—Iron
0.1 0.2 0.3 0.4
EQUILIBRIUM CONCENTRATION OF ZINC MG/L

Figure 6
Removal Isotherm—Zinc
0 10 20 30 40 SO 60 70 80 9O 100 110 120130 140 ISO 160170 180 190 200
EQUILIBRIUM CONCENTRATION OF TOC MG/L
Figure 7
Removal Isotherm—Total Organic Carbon
0.3-
0.2-
100 200 300 400 500
EQUILIBRIUM CONCENTRATION OF COO MG/L
600
Figure 8
Removal Isotherm—COD (Low Concentrations)

Table 5
Calculated Isotherm and Retardation Coefficients
Freundlich Equation
Linear Equation
Parameter Kf
BOD
COD
(high
cone.
COD
(low
cone .
Barium
Iron
Zinc
TOC
NOTE:
22
0.037
0.42
9.8
81
1080
0.138
Based on
N
0.22
1.73
0.98
2.08
1.48
2.61
1.26
bulk der
Corr.
Coef.
0.49
0.75
0.97
0.87
0.80
0.97
0.93
isity of
Rf
6.8
48.4
4.5
115
3401
100,000
6.4
1.9 gr/cc
Kr>
0.24
0.88
0.34
9.6
110
190
0.40
Corr.
Coef.
0.40
0.62
0.98
0.63
0.83
0.86
0.72
_RD
3.3
9.4
6.4
92
1050
1834
4.8
and effective
porosity of 0.2.
Note: Based on bulk density of 1.9 gr/cc and effective porosity of 0.2.

CONCLUSIONS
• The components in the leachate in this landfill can be grouped
into three categories. The first category contains components
that are present in the leachate in very low concentrations:
copper, mercury, nickel, selenium, silver, fluoride and sulfate.
These components need not be addressed further for this land-
fill.
The second category of chemicals in the leachate contains
components that have environmental significance that are at-
tenuated by soils present at the site: barium, iron, zinc, BOD,
COD and TOC.
The third category of leachate components includes those
components for which chemical attenuation in soils would be
expected to have little or no impact: total dissolved solids and
chloride. The concentrations of these components in the leach-
ate are higher than background water quality, as expected.
Low transport rates due to low permeability soils and dilution
and diffusion mechanisms would be required to reduce the
concentrations of these components.

• The COD and TOC concentrations showed a steady decline
with decreasing leachate:soil ratios and adsorption continues
even at low COD and TOC concentrations. Because of the
significant depth of till, we can expect a high removal rate of
these components. Furthermore, biological degradation of
LAND DISPOSAL 189
-------
components of the COD and TOC can be expected. The re-
tardation of these components by the soil provides more reac-
tion time for the biological reactions.
• The difference in composition between the leachate that was
maintained anaerobic and the leachate exposed to air strongly
demonstrates the need to conduct experiments such as this
under the expected environmental conditions of the system.
• The data scatter in some of the experiments could be due to
deviation from the standard procedures for drying and pul-
verizing the soil sample. Future work should follow the stan-
dard procedures for handling soil.
• The adsorption isotherm method provides valuable informa-
tion concerning the interaction of leachate components and
soils. The interpretation of the retardation coefficients must be
done with the same caution as for flow velocity calculations.
If unanticipated soil fractures exist or if the effective porosity
of the soil is greater than expected, the retardation will be less
than predicated.

REFERENCES
1. Kamprath, E.J., Nelson, W.L. and Fills, J.W., "The Effect of pH,
Sulfate and Phosphate Concentrations on the Adsorption of Sulfate
by Soils," Soil Sci. Soc. Proc., 1956, 463-466.
2. Lambert, S.M., Porter, P.E. and Schieferstein, R.H., "Movement
and Sorption of Chemicals Applied to Soil," Weeds, 13, 1965, 185-
190.
3. Barrow, N.J., "The Description of Phosphate Adsorption Curves,"
J. Soil. Sci., 29, 1979, 447-462.
4. Roy, W.R., Ainsworth, C.C., Griffin, R.A. and Krapac, I.G., "De-
velopment and Application of Batch Adsorption Procedures for De-
signing Earthen Landfill Liners," Proc. Seventh Annual Madison
Waste Conference, University of Wisconsin, Madison, Sept.
1984, 390-398.
5. Roy, W.R., Griffin, R.A., Chou, S.F.J. and Krapac, I.G., "Stan-
dardized Batch Adsorption Procedures for Soils and Sediments,"
Ninth Quarterly Report of Progress, Illinois Geological Survey,
Champaign, IL, Jan. 1985.
6. Ainsworth, C.C., Griffin, R.A., Krapac, l.G. and Roy, W.R., "Use
of Batch Adsorption Procedures for Designing Earthen Liners for
Landfills," Proc. of Tenth Annual Research Symposium of the Solid
and Hazardous Waste Research Division, Ft. Mitchell, KY, April
1984, EPA-600/9-84-007, 238-249.
7. Griffin, R.A., Sack, W.A., Roy, W.R., Ainsworth, C.C. and Krapac,
I.G., "Batch Type 24-hour Distribution Ratio for Contaminant Ad-
sorption by Soil Materials," Draft for Review by ASTM Committee
D-34 on Waste Disposal, ASTM Symposium on Environmental Test
Method Development, 1985.
8. Mickelson, D.M., Clayton, L., Baker, R.W., Mode, W.N. and
Schneider, A.F., "Pleistocene Stratigraphic Units of Wisconsin,"
Wisconsin Geological and Natural History Survey, 1984.
9. Mickelson, D.M., Clayton, L., Baker, R.W., Mode, W.N. and
Schneider, A.F., "Pleistocene Stratigraphic Units of Wisconsin,
Miscellaneous Paper 84-1," Geological and Natural History Survey,
University of Wisconsin Extention, 1984.

Table Al
Leachate Attenuation by Soil
Chemical Parameter = BODS
Table 2A
LeBcbale Attenuation by Soil
Chemical Parameter — COD
•41 r»
'** **
*M M
M *»
•I 't
iiis
».4MI
• .Nfl
• .Ml'
•-till
• -IM*
Note. No 11 & 12 arc toil blanks; No. IJ U Uachalc blank.

Table A3
Leachate Attenuation by Soil
Chemical Parameter = Barium
Nolc: No 11 & 12 are toil blank*: No U u leachale blank.

Table A4
Leachate Attenuation by Soil
Chemical Parameter = Iron
Note: No. 11 A 12 arc toil blanks: No I) is teachale blank.

Table AS
Leachate Attenuation by Soil
Chemical Parameter = Zinc
Note: No 11 A 12 arc soil blanks: No 13 is leachalc blank.

Table A6
Leachate Attenuation by Soil
Chemical Parameter = TOC
IS
ill
«2 •

11
iSs
!:8
SiS
3:5

1
•s
•.MH
• Wt»
Sss

I
Note: No. 17 & 18 are soil blanks; No. 16 is leachate blank.

Table A7
Leachate Attenuation by Soil
Chemical Parameter = COD-Low Range

t.ut •
MI.«
<«n *» '•** *•
H-S £2
S3 !S:S
Mi,H
-------
Construction Quality Assurance for Hazardous
Waste Land Disposal Facilities with Emphasis on
Soil Barrier Layers and Final Multilayer Cover Systems
Richard C. Warner, Ph.D.
Nathaniel Peters
Department of Agricultural Engineering
University of Kentucky
Lexington, Kentucky
OBJECTIVE
To assist landfill design engineers, operators, contractors,
specifiers and reviewers in the planning, construction and testing
of the complete liner and cover system. Emphasis will be placed
on construction of the soil liner and cover barrier layer with
associated laboratory and field testing.

EXPANDED OUTLINED AND
SHORT COURSE SUMMARY
I. Overview and Purpose of Course
II. Elements of the CQA Plan
A. Administrative Details
1. Organization
2. Personnel Qualification
B. Inspection Activities
1. Foundation
2. Dikes
3. Low Permeability Soil Liners
4. Flexible Membrane Liners
5. Leachate Collection Systems
6. Final Closure Cover
III. Sequencing of Liner Layers
A. Foundation Preparation
B. Groundwater Collection Drains
C. Bedding Layer
D. Flexible Membrane Liner
E. Geotextile Fabric
F. Drainage Layer
G. Witness Collection Drains
H. BeddingJLayer )
I. Flexible Membrane Layer >
J. Geotextile Fabric )
K. Drainage Layer
L. Leachate Collection Drains
M. Selected Waste Material
N. Waste Material
0. Gas Venting Layer
P. Low Permeablity Soil Layer
Q. Bedding Layer
R. Flexible Membrane Liner
S. Geotextile Fabric
T. Drainage Layer
U. Top Soil Layer
or Low Permeability
Soil Layer
or Low Permeability
Soil Layer
V. Vegetation
IV. Fundamental Soil Relationships
A. Moisture Content
B. Unit Weights
C. Void Ratio
D. Porosity
E. Specific Gravity
F. Grain Size
G. Saturated/Unsaturated Properties
H. Field Capacity
I. Wilting Point
V. Low Permeability Soil Barrier
A. Pre-construction QA/QC Activities
1. Material Inspection
a. Field Identification
(1) Fine-grained Soils
(a) Color
(b) Stratification
(c) Appearance and Structure
(d) General Field Behavior
(e) Consistency
(2) Coarse-grained Soils
(a) Color
(b) Grain Size
(c) Grading
(d) Compactness
(e) Appearance
b. Laboratory
(Refer to Section VI)
B. Compaction
1. Compaction Curve
2. Compaction and Permeability
3. Compaction and Molding
C. Permeability
1. Potential Laboratory Problems
2. Potential Field Problems

VI. Soil Testing
A. Gradation
1. Sieve
2. Hydrometer
B. Unified Soil Classification System
C. USDA System
D. Atterberg Limits
LAND DISPOSAL 191
-------
E. Moisture-density Compaction
F. Maximum Clod Size
O. Permeability (Laboratory)
1. Pressure Cell
2. Compaction Permeability
3. Triaxial Cells
H. Permeability (Field)
1. Double Ring Infiltrometer
2. Sealed Double Ring Infiltrometer
VII. Test Fill Construction and Testing
A. Objectives
B. Laboratory vs. Field Hydraulic Conductivity
C. Data to be Recorded
1. Equipment
2. Number of Passes/Speed
3. Clod Breakdown
4. Moisture Content
a. Method
b. Equilibration
5. Lift Thickness
a. Loose
b. Compacted
6. Dimensions
D. Construction Procedures
1. Layer Sequence
a. Foundation
b. Bedding Layer
c. Flexible Membrane Liner
d. Bedding Layer (Geotextile Fabric)
e. Drainage Layer
f. Compacted Lifts (3)
2. Soil Characteristics
a. Samples
3. Clod Size Reduction
a. Method
b. Final Size
4. Soil Moisture Content
a. Incorporation Method
b. Equilibration
c. Testing
5. Incorporation of Soil Amendments
6. Compaction Testing
7. Extraction for Laboratory Sampling
8. Scarification of Lifts
9. Prevention of Desiccation Cracks
10. Ponding of Water
11. Percolation Water Measurement
12. Sealed Double Ring Infiltrometer
13. Tracers
VIII. Final Cover System
A. University of Kentucky Research Experience
B. Water Balance
C. Modeling
D. Cost Trade-offs
192 LAND DISPOSAL
-------
Comparative Analysis of Soil Gas Sampling Techniques
Peter P. Jowise
JeffD.Villnow
Lazar I. Gorelik
Ecology and Environment, Inc.
Seattle, Washington
John M. Ryding
C.C. Johnson and Associates
Denver, Colorado
ABSTRACT
Five soil gas sampling techniques and two analytical methods
were examined during the course of an investigation to delineate
the extent of trichloroethylene contamination in a shallow sand
and gravel aquifer in Western Washington. Two sampling tech-
niques were chosen for extensive examination of plume delinea-
tion, while all five techniques provided comparable results at a
limited number of common sampling locations. The choice of an
appropriate sampling method can be based on an evaluation of
both contaminant and environmental characteristics. Sampling
technique cost differences were minimal. One analytical method
proved superior, as a more sensitive contaminant detection ele-
ment was employed. Analytical variability was a minimal com-
ponent of overall field variability.

INTRODUCTION
Although the movement of volatile organics through the un-
saturated zone has been the subject of many investigations,''2'3 no
consistent or standard sampling methodology has been reported
to address the effects of contaminant or soil characteristics and
the impact of sampling method on measured concentrations in
soil gas. The study outlined below attempts to provide a consis-
tent framework for application of soil gas sampling and analysis
used as a screening technique for the optimal placement of
monitoring wells under various environmental conditions and for
differing physical-chemical contaminant properties.
Mass transport of volatile organic contaminants from the
saturated zone to the soil surface is due primarily to unsteady
molecular diffusion through both gas and non-gas phases. For
most volatile compounds with a Henry's Law Constant greater
than 10-4, gas phase diffusion is considered to be dominant.4 As
an organic compound travels through the soil column in the gas
phase, it is absorbed by soil water and adsorbed by soil particles.
Because unsteady state conditions prevail, the soil gas concentra-
tion profile, as well as the distribution profile between the three
phases, will vary over the soil column with time.
An examination of the parameters that govern compound
distribution between gas, liquid and solid phases is of practical in-
terest in conducting soil gas sampling programs. This distribution
depends primarily on volumetric fractions of soil air and water,
organic carbon content of soil, soil bulk density, compound
aqueous solubility, vapor pressure and affinity to soil. The
following equation describes the influence of these parameters on
soil gas concentration and mass distribution between the three
phases under equilibrium conditions:5

Cs = Ca [0 „ + (0 W/H) + (KdPs/H)] = Ca/3
(1)
where:
Cs = total concentration of compound in soil matrix,
^g/cm3 soil
Ca = concentration in soil air, /ig/cm3 air
0a = soil air content, cm3/cm3
0W = soil water content, cm3/cm3
H = Henry's Law Constant, unitless
Kd = adsorption constant, cm3/g
Ps = soil bulk density, g/cm3

This equation assumes that for dilute concentrations in soil water,
the constants H and Kd are independent of solute concentration.
The equation is based on a simplified representation of mass
distribution, especially with regard to the linear relationship be-
tween solute concentrations of solid and liquid phases. It is
unlikely that equilibrium conditions exist throughout the soil pro-
file across a sample space; however, the equation will provide
reasonable distribution estimates for organic compounds and in-
dicate changes in distribution as soil properties vary.
The collection of terms represented by /3 in the equilibrium
equation includes both environmental and contaminant proper-
ties. By keeping one set of properties constant, the effects of
changing environmental conditions or the effects of differing con-
taminants on phase distribution can be assessed. Assuming cons-
tant field conditions, contaminants with large H and low Kd pro-
vide low values of j3 and relatively high values of Ca. This condi-
tion is satisfied for most volatile organics listed on the U.S. EPA's
Hazardous Substance List. Compounds of interest may be assessed
individually and ranked according to their tendency to be found in
the gas phase. Although the bulk of transport occurs in the gas
phase, a substantial amount of contaminant mass may be
associated with water and soil particles along the transport
pathway. This same evaluation can be conducted for con-
taminants across a range of field conditions.
The variable effects of changing field parameters on phase
distribution can be assessed by accounting for relative soil air/
moisture content, bulk density (porosity) and availability of ad-
sorption sites on solid phase particles.6 By increasing the amount
of water in the soil matrix (due to precipitation, flooding or
changing soil moisture retention characteristics), contaminant
MONITORING & SAMPLING 193
-------
mass will move into the liquid. By increasing the air content of
soil, relative to water, the gas phase contaminant fraction will in-
crease. Increasing bulk density, which is inversely porportional to
porosity, provides a greater particle surface area to pore space
ratio, resulting in proportionally more contaminant binding sites.
An increase in organic carbon content also will provide increased
soil adsorption capacity.
Contaminant mass transport and equilibrium are affected by
the soil conditions described above. Transport by diffusion is
controlled chemically by the concentration gradient and physical-
ly by pore space continuity. By decreasing the pore air volume,
due to increased water or solids content, the cross-sectional area
for flow is decreased and the flow path length is increased. Con-
taminant movement is slowed in the gas phase by physical im-
pediments. If infiltration of water is great enough, contaminant
mass may be carried back toward its source, depleting soil gas
concentrations.
The implications of environmental conditions on soil gas
sampling methodology are two-fold. First, if a substantial
amount of contaminant is likely to associate with water and soil
while moving through the soil air, increased detection sensitivity
may be realized by collecting a soil sample and then rearranging
the equilibrium phase distribution in the laboratory before
analysis. Second, if contaminant transport is suppressed by
physical impediments, sampling depth will be critical to definition
of plume boundaries at low concentrations. High concentrations
always will be found closest to the source. The acceptable com-
bination of drilling depth (and cost) versus collection and
analytical sensitivity can be defined by this preliminary field
assessment.

SITE CHARACTERISTICS
A soil gas sampling study was conducted in Western
Washington, where a water supply well tapping a deep confined
aquifer was potentially threatened by trichloroethylene (TCE)
contamination existing approximately 0.25 mi upgradient in a
shallow aquifer. The area of concern was approximately rec-
tangular in shape; 1 mi long and 0.5 mi wide (Fig. 1).
0 000 1000 tWO MOO
LEGEND
Ltlltary ronrvotlon boundory
N
Figure 1
Soil Gas Sampling Site Map

The hydrogeologic regime of the region was complex, due to
repeated glacial episodes, resulting in a sequence of glacially
related drift deposits interbedded with interglacial sediments. The
various drift deposits were similar in nature and usually consisted
of unconsolidated sands and gravels layered with compact, un-
sorted till. They were heterogeneous, locally discontinuous and of
variable thickness, making regional correlation of individual units
difficult. The interglacial deposits, by comparison, tend to be
finer grained, more homogeneous and more laterally continuous
than drift deposits. Major aquifers of the area occurred primarily
in the various drift deposits. The interglacial sediments were,
typically, regional aquitards.
A shallow, unconfined aquifer occurred in the uppermost drift
deposit. Depth to groundwater ranged between 20 and 30 ft below
land surface. Seasonal fluctuations of up to 10 ft had been
observed in nearby wells. The base of the aquifer was formed by a
locally discontinuous deposit of interglacial sediments occurring at
an average depth of 80 ft and approaching thicknesses of 150 ft.
Groundwater in the shallow aquifer flowed generally to the
northwest. Previous geophysical studies of the vicinity indicated
the possibility of paleochannels as conduits of increased ground-
water flow.7 These conduits typically were oriented with the
regional groundwater flow direction, although localized
divergences existed. No investigation was conducted to identify
paleochannels within the study area, and no estimates were
reported of channel permeability relative to the surrounding for-
mation, i
Documented TCE contamination existed to the southeast of
the study area, although the extent of contamination was not
known at the beginning of this investigation. Soil gas monitoring
was used to identify the extent of TCE contamination in the
shallow aquifer and thereby locate monitoring wells in an efficient
manner. Soil gas samples were collected at 66 stations, including
10 test locations (Fig. 2). The test locations were selected near ex-
isting monitoring wells for comparison of water and gas concen-
tration levels at different depths.
LEGEND
N
Figure 2
Soil Gas Sampling Locations

SAMPLING METHODOLOGY
Five soil gas sampling techniques were utilized in the study:
Direct Injection-Auger (DIA), Direct Injection-Stopper (DIS),
Soil Headspace, 1-Liter and Tenax Tube. The last three tech-
niques utilized sample concentrating steps to increase detection
sensitivity. All five methods required the drilling of shallow bore-
holes. DIS and Tenax Tube required installation of temporary
sampling probes.
Sample probes consisted of 2-in. I.D. galvanized steel pipe cut
into 5-ft sections with threaded ends. The bottom 2 ft of each
probe contained four series of eight 0.25-in. diameter drilled holes.
194 MONITORING & SAMPLING
-------
The probes were placed down the inside of an 8-in. O.D. hollow
stem-auger after drilling to the final depth of 12 to 17 ft. Borehole
walls then were allowed to collapse around the probe as the augers
were pulled. Probe installations were completed by pouring a 2-ft
thick bentonite slurry plug around the top of the hole and sealing
the probe with a stopper and screw cap. The typical sequence of
drilling and sampling activities utilized in the soil gas sampling
program is described in Table 1. Method-specific descriptions of
each sampling technique follow.

Direct Injection-Auger
Samples were collected by lowering a 500-/*1 gas-tight, side port
syringe to the bottom of the hollow stem auger immediately upon
reaching the desired depth. The syringes were filled remotely by
manipulating two attached rods. Elapsed time between hole com-
pletion and sample collection at each location was uniform to
minimize sampling variability. A 100 ^1 aliquot from the syringe
then was injected directly into a field laboratory gas chromato-
graph.

Direct Injection-Stopper
Gas samples were collected from finished probes after a 2-day
equilibration period. Syringes were inserted through the probe
stoppers, and 100 jil of soil gas were drawn for direct GC injec-
tion. After sampling, the stoppers were visually examined for
leaks and replaced, if necessary, to maintain an adequate seal on
the sample probe.

Table 1
Sequence of Drilling and Sampling Activities
Activity Description
Drill 8-Inch O.D. borehole
Collect 100 ul gas sample from
the bottom of augered hole
Collect 1-liter gas sample from
botton of augered hole
Collect soil sample for
headspace analysis
Install sampling probe
Stapling probe equilibration
Average Duration
15-30 minutes
1 minute
1 minute
5 minutes
10 minutes
Minimum 2 days
Sampling
Techniques

DIA
1-Liter
Soil Headspace

Collect 100 ul and tenax
tube samples
Remove sample probe and
backfill hole
1-45 minutes
30 minutes
D1S. Tenax Tube
Soil Headspace
Soil samples were collected by driving an 18-in. split spoon
sampler fitted with three clean brass sleeves into undisturbed soil
beneath the lead auger. After retrieval, the sampler was opened
and the contents of one brass sleeve were immediately pushed to
fill a 236 ml (8 oz) glass jar containing 100 ml of carbon-free
water.
The soil/water samples were transferred to the mobile
laboratory, placed in a sonic bath and agitated for 5 min to
volatilize TCE from the liquid and solid matrix. A 100 jtl air sam-
ple then was drawn for injection into the field laboratory CG.

1-Liter
Samples were collected by lowering a 10-ft length of Teflon
tubing attached to a 1-liter gas-tight syringe to the bottom of the
hollow stem auger (same as DIA method). The Teflon line was
conditioned by drawing a 500 ml air sample, disconnecting the
Teflon line and evacuating the syringe. The Teflon line then was
reattached to the syringe, and the 1-liter sample was withdrawn
from the bottom of the auger. The entire sample then was injected
into a purge and trap unit for concentration followed by analysis
on a GC.

Tenax Tube
The Tenax Tube sampling set-up included tenax packed into
stainless steel desorption tubes connected in line to a Gillian pump
set to deliver a flow rate of 100 ml/min. The Tenax Tubes were
suspended 3 ft down the probe on a section of rubber tubing that
passed through a stopper at the top of the probe before connec-
tion to the pump. Three liters of soil gas were pumped through
each tube. The Tenax Tubes were taken to the field laboratory,
contaminants were thermally driven off the column and a 100 /tl
sample aliquot was injected into a GC for analysis.

ANALYSIS
Analytical testing of soil gas samples was performed in an on-
site mobile laboratory using field screening procedures developed
by Ecology and Environment, Inc. Analyses of 100 /*! samples
were performed using a Shimadzu GC Mini-2 equipped with an
electron capture detector (BCD) connected to a two meter 1%
SP-1000 packed column. The 1-1 samples were analyzed using a
Shimadzu GC Mini-3 equipped with an QIC model 4420 (electro-
lytic conductivity) Hall detector with an in-line Tekmar LSC-1
liquid sample concentrator (purge and trap unit) and a 30-m
SPB-1 megabore capillary column. Both systems were operated
isothermally; both used a Shimadzu C-R3A integrator for quan-
tification. The quantifiable detection limit for all samples was
estimated at 0.2 /tg/1.
The 100 id samples collected from auger and probes were in-
jected directly into the GC. Sample holding time never exceeded
45 min. Sample analysis time was approximately 10 min. Follow-
ing injection, syringes were flushed using a vacuum heated syringe
cleaner. The 1-1 samples were injected directly into the LSC-1 for
preconcentration prior to GC analysis. Sample holding time never
exceeded 45 min. Sample introduction required 15 min. Because
injection was performed by hand, the input rate was variable,
resulting in possible TCE breakthrough in the purge and trap
unit. Total analysis time was approximately 25 min. Empty
syringes were cleaned by multiple flushing with nitrogen gas.
The Tenax Tube sampling method required pre-conditioning
and assembly of tubes prior to field mobilization. Tenax was
chosen as the solid adsorbent, based on its known effectiveness in
adsorbing chlorinated hydrocarbons and for ease and extent of
thermal desorption. Activated carbon, the more commonly used
adsorbent, often is chosen for its superior adsorption properties.
It is less desirable, however, for many compounds when using
thermal desorption techniques, due to practically irreversible
bonding. Thermal desorption from Tenax is more easily accom-
plished, affording more complete recoveries and the convenience
of reuse.
35/60 mesh Alltech Tenax-TA was solvent extracted sequential-
ly for 6 hr with methanol followed by hexane. The Tenax then
was heated in a continuously purged nitrogen atmosphere. Then
the Tenax was placed in a cleaned and hexane-rinsed liquid
chromatography tube, wrapped with heating tape and heated to
325 °C for 17 hr under a stream of ultrapure nitrogen with an ap-
proximate flow of 50 ml/min. After cooling, the Tenax was packed
into Foxboro stainless steel thermal desorption tubes under
nitrogen until full. The tubes were capped and placed in 40-ml
VOA vials, with a small amount of anhydrous sodium sulfate, un-
MONITORING & SAMPLING 195
-------
til used in the field.
Tenax Tubes brought from the field were thermally desorbed at
225 °C with a Foxboro Century PTD 132A programmed thermal
desorber. A 100 n\ sample was withdrawn from the desorber and
injected into the GC. The desorber was cleaned with nitrogen gas
flushing. Tenax Tubes were cleaned for reuse by baking at 275 °C
in the desorber while flushed with ultrapure nitrogen.

Standards
To quantify the OC response, gas standards were prepared us-
ing volume calibrated glass and Teflon gas sampling bulbs. Pure
TCE was injected into a bulb and gently heated. This primary
standard was diluted to prepare working standards by injecting
known volumes from the primary standard bulb into clean bulbs.
Two-point calibrations were performed periodically, with daily
one-point calibration checks to monitor the continuing daily
response factor of the GC. These daily calibrations were used to
quantify sample TCE responses. If daily calibration response fac-
tors differed from the initial curve by more than 15%, a new
reference calibration curve was constructed. Use of this technique
ignored the effect of adsorption on container and syringe surfaces.

Blanks
Numerous daily blanks were run, especially after receiving
samples of moderate to high concentrations of TCE, to insure that
no carry-over or cross-contamination of samples occurred. Blanks
were analyzed from gas sampling bulbs, syringes, the thermal
desorber and Tenax Tubes on a frequent basis. Occasional field
blanks were run using the Tenax Tubes to check for ambient air (at-
mospheric) background contamination.

RESULTS
A primary purpose of this investigation was to define the extent
of TCE contamination in the shallow aquifer system. Five soil gas
sampling methods were developed to define the plume and direct
appropriate placement of monitoring and pump-test wells. Man-
power, cost and time constraints allowed limited sampling and
testing of all techniques. Only the DIA and Tenax Tube methods
were used across the entire sample space. Results from these
methods were used to map the contaminant plume, as indicated
by soil gas. The two methods requiring sample probes (DIS and
Tenax Tube) were utilized to study field variability by repeated
sampling at single locations. The 1-Liter and Soil Headspace tech-
niques were discontinued halfway through the investigation. A
project sampling summary, including the number of field QA
samples, is presented in Table 2.

Table 2
Soil Ga« Sample Numbers by Sample Method
S»ple Method
Soil Gas Saaple Niabers
Stapled
Location! Tottl
Direct Injection-Auger
Direct Injection-Stopper
Soil Headspace
One-Liter Syringe
Tenax Tube
TOTAL
64
36
34
43
59
66
67
128
62
43
128
428
Field QA*
2
48
30
0
70
150
* Quality Assurance duplicates and replicates

Sampling activity encompassed an area of approximately 1
square mile, transecting the groundwater contaminant plume of
interest. Fig. 3 provides a comparison of DIA and Tenax Tube
sampling results to the measured groundwater plume determined
from monitoring well data. Samples collected with Tenax Tubes
yielded a comparatively wider range of values and indicated wide-
spread, low level contamination across the study area. Only three
locations sampled with Tenax were characterized by undetectabk
TCE concentrations. Significantly fewer DIA samples identified
TCE presence across the sample space, indicating that it is a
relatively less sensitive method. The DIA method produced a
relatively narrow range of measured concentrations, requiring
special care in drawing isopleths. Data accuracy must be assumed
adequate to distinguish values within this small range. Accepting
this assumption, both DIA and Tenax Tube results provide con-
sistent contaminant plume mapping in this case.

DISCUSSION OF SAMPLING METHODS
Four criteria have been chosen to compare the five sampling
techniques: field effectiveness (logistical considerations), analyti-
cal limitations, data validity and cost. Because the study focused
on plume definition and was not a soil gas sampling research pro-
ject, not all techniques were investigated to the extent required for
rigorous comparison. The discussion below presents specific
method limitations as well as an overall method comparison.
Table 3 summarizes these considerations.
I) SOL GAS PUME-1GKAX
(C) GROUND WATER HM*
LEGEND
IMOI.UI
untary raoorvaUon boundary
A SampK location
• am location
Figure 3
Comparison of Contaminant Plume Delineations
196 MONITORING & SAMPLING
-------
Table 3
Summary of Soil Gas Sampling Methods
1 «1n/10 ain o Fatt collection ind anilyilt
o Elevated concentration due
to wchanfcal •l«1ng or ioll
o Monitoring hole can be IB-
•edUtelj backfilled
1 •In/10 Bin o Long-tenn sampling
0 SUtlc tangling environment
that reflect! true Mil gai
concentration!
10-20 mln/lS "In 0 Concentrates contaminant

40 Blri/iS •*" 0 Concentrttei contanlntnt
o Long-ter* tJBpltiw
o Tine-average umpllng
o No repeated iivpUng
0 Si«pltnd cn*tronoent '* not
controlled (dyntclc pro-

o Requires probe InitalUtlon

o Ooei not oork xll «UH non-
coheilte lolls
o Ho repeated ladling

line
o No repeated steeling

o Requires probe Instillation
o Bequlrei preconditioning of
tena* t-jbes
Remote ia«vl«r
Gii-lljht syringe

Probe

Sfllti ipoen
SiBpIt container
Son tcitor
Ca»-t1ght sjrtnge
Syringe pu«o '
Cyro;en1c focuiiing device
r-jrgi and trtp ,w\c**nl
Tena» deierpHon ••Jt«(
CaHbrited air puep
Field Effectiveness
The field effectiveness of each technique was evaluated by con-
sidering sample collection and analysis times, ease of sample col-
lection and various logistical considerations. Table 3 provides a
comparison of both sample collection and analysis time for each
sampling technique. The syringe collection techniques required
minimal sample preparation time prior to analysis. Effectiveness
of soil sample collection depended on soil texture. Tenax Tubes
required preliminary equipment setup with lengthy sample pump-
ing, although many locations could be sampled at once with a
rotating schedule. Analysis time reflects an approximate 6-min
GC retention time for all methods and appropriate cleanup of col-
lection equipment (syringes and Tenax Tubes).
Sample collection was comparatively easy for all methods ex-
cept Soil Headspace. The cobbly and non-cohesive nature of soils
in the study area caused loss of sample from the split spoon in
some cases and inefficient sample transfer to containers in others.
Sample collection also was hindered by overhead obstructions
when driving the split spoon in residential areas. Although sample
collection from probes was easily performed, installation of probes
in residential areas at preferred locations was not always possible.
Logistical problems associated with sample analysis that should
be considered are related to laboratory proximity, equipment
power supply and sampling apparatus failure. During this investi-
gation, the laboratory was located within 0.5 mi of any sampling
station, and local city power was employed. A continuous power
supply was required to run analytical equipment for the entire
field study period (approximately 1 mo). Gas-tight syringes were
manufacturer tested but were not subjected to rigorous field
testing. Samples were injected typically within 30 min of sam-
pling. Tenax Tubes occasionally were stored in the freezer over-
night with no apparent loss of sample integrity.

Analytical Limitations
Two analysis techniques were performed concurrently in the
laboratory. Equipment configurations were described in the
methodology section. Direct injection of 100 ;tl to the GC from
small syringes posed no problems, although care was taken to
minimize potential water vapor inclusion from all samples. The
1-Liter injection technique required approximately 15 min of slow
injection to insure that no analyte escaped from the Tenax in the
Purge and trap unit. Because injection was performed by hand, a
constant flow could not be guaranteed and blow-by may have oc-
curred. Cryogenic focusing was performed on a limited number
of 1-Liter samples, which resulted in a marked positive effect on
GCpeak definition. The Hall detector, used in the 1-Liter sample
analysis, was not as sensitive as the ECD used for all other sample
analyses. The overall length of a single 1-Liter sample analysis
made duplicate or replicate analysis untenable in the field.
The Tenax Tube method required an additional step in the ana-
lytical routine by use of the programmed thermal desorber
(PTD). The additional step provided extra potential for equip-
ment failure. High analyte concentrations associated with field
hot spots resulted in residual contamination after routine PTD
cleaning. Additional flushing sometimes was required. Continual
checks were time-consuming.

Data Validity
Data validity was evaluated by first assessing field and
analytical variability of the DIS and Tenax Tube collection
methods. Both sampling techniques utilized long-term monitoring
probes, allowing repeated sampling over the course of the study
period. All five techniques then were evaluated together by com-
paring detected concentrations at 18 common sampling locations.
Finally, all techniques were compared by correlation to each other
as well as to water concentrations at common locations.
Analytical variability of the DIS method was evaluated by con-
currently collecting at least two samples from a single probe on 18
different days and examining the precision of the daily analytical
results. The measure of precision utilized was the relative stan-
dard deviation (RSD) of the daily results (also known as the coef-
ficient of variation). RSD was used instead of relative percent dif-
ference (RPD) for consistent evaluation when both two and
greater than two samples per set were compared. Results of this
analysis (Table 4) indicate the average analytical variability for
DIS sampling at approximately 7%.
Field variability was evaluated by examining relative differences
in TCE concentrations for all DIS samples at the same probe. The
RSD for the 42 samples was 40%. By subtracting analytical
variability as computed in Table 4, field variability can be as-
sumed at approximately 33% RSD. This evaluation includes a
known sampling variability bias, since stoppers occasionally
leaked, resulting in short-term downward trends in detected con-
centrations. For locations with non-repeated sampling, field
variability is expected to decrease below the 33% level.
Variability of data associated with the Tenax Tube sampling
technique was evaluated on three levels: total field variability,
analytical variability and method variability. Field variability was
evaluated by analyzing 10 samples collected from a single location
over a 4-wk period. The RSD for the 10 samples as a group was
approximately 24%. Analytical variability was measured by

Table 4
Analytical Variability of Daily DIS Sample Groupings
Collected from Probe T2

Date
4/10/86
4/14/86
4/14/86
4/17/86
4/18/86
4/21/86
4/Z2/86
4/23/86
4/24/86
4/25/86
4/29/86
4/30/86
5/01/86
5/02/86
5/06/86
5/12/86
5/13/86
5/14/86
5/15/86
DAILY
AVERAGE
AVERAGE
No. of
Samples
4
2
2
3
2
2
2
3
2
2
2
2
2
2
2
2
2
2
2

Average TCE
Concentration
7.48
6.75
3.95
3.10
3.55
7.70
6.40
4.63
3.85
2.30
2.90
2.20
4.65
1.40
5.35
5.80
5.55
6.95
6.35

Standard
Deviation
1.31
0.07
0.35
0.44
0.07
0.28
0.14
0.67
0.49
0
0.42
0.14
0.49
0.28
0.21
0.14
0.21
0.07
0.07

I RSD
17.5
1.0
8.9
14.1
1.9
3.7
2.2
14.4
12.8
0
14.6
6.4
10.6
20.2
3.9
2.4
3.8
1.0
1.1

7.39
40.
MONITORING & SAMPLING 197
-------
analyzing six replicate sample sets drawn from the PTD sample
chamber. The average RSD for the six sample sets was approxi-
mately 2%. Method variability was evaluated by analyzing three
triplicate samples collected from different field locations, ex-
hibiting a wide range of measured concentrations. The RSD
varied inversely to concentration and. as one would expect, a very
high RSD (44"%) was associated with measured concentrations
near the instrument detection limit. The other two RSDs were
calculated at 5% and ICWo. It can be postulated that the dif-
ference in method and analytical variabilities may be attributed to
the thermal desorption process. It is apparent that the variability
attributed to sampling technique and environmental conditions
(field variability) far outweighs analytical considerations, except
at very low concentrations.
All five sampling techniques were conducted at 18 common
field locations. A statistical evaluation of measured concentra-
tions is provided in Table 5. The concentrating effect of the Tenax
Tube method is reflected in an approximately 10-fold increase in
average concentration over the other methods. The 1-Liter
method, also a concentrating technique, should have produced
higher measured values. The relatively low results can be attri-
buted to use of the Hall detector instead of the ECD used for all
other sampling methods. The high RSD values for all sampling
methods reflect the wide range of measured concentrations. The
comparatively high RSD value for the Soil Headspace method can
be attributed to inconsistent sample collection performance due
to non-cohesive soils.

Table 5
Sample Method Statistics for Common Sample Locations
Swple
Method
Direct Injection-Auger
Direct Injection-Stopper
Soil Headspace
One-Liter
Tenax Tube
Average
Concentration (ug/D
1.6
0.71
1.8
0.64
13.6
Standard
Deviation
3.1
1.5
5.1
1.2
31.6

t RSD
194
211
283
188
232
Table 6
Correlations Between SGS Techniques
(r2)
D1A
One-L
HS
015
Tenax
Water (5)'
Water (V)**
1.00
.68
.72
.78
.80
.99
.41

1.00
.80
.87
.90
.99
.99

1.00
.73
.84
.99
.89

1.00
.98
.99
.87

1.00
.99
.80

1.00
1.00
D1A
One-L
HS
D1S
Tenax
* 5 common sampling locations
** Variable number of sampling locations depending on method: DIA (12). One-L (8).
HS(6). DIS(IO). Tenax (13).
The five sampling techniques are compared by correlating them
to each other based on a linear regression (Table 6). It can be seen
that there is generally good correlation between methods, with the
1-Liter and Tenax Tube concentrating methods marginally better
than the others. Table 6 also includes correlations of soil gas sam-
pling results to water concentrations for five locations common to
all sampling techniques. At these locations, all techniques per-
form extremely well. The five locations included water concentra-
tion values of < 5, <; 5, 5, 6 and 117 /tg/1 TCE. In addition, Table
6 includes correlation values for SGS technique results associated
with all water sample locations associated with a particular
method. It is difficult to compare values, due to the inconsistent
sample space represented, except for DIA and Tenax Tube (as-
sociated with the same group of water samples). Here it can be
seen that the DIA method falls far short of Tenax Tube in ac-
curately reflecting water concentrations.

Costs
Cost considerations typically dictate the choice of sampling
technique within a sampling program. Of the five sampling
methods discussed, all require a hole to be drilled. The hole
should at least penetrate beyond surface soils containing the bulk
of organic matter to bypass competing adsorption sinks. Ap-
propriate drilling depth must be addressed either by a short test
hole program or a combination of experience and evaluation of
site specific physical and chemical factors as discussed in the
theory section. During this investigation, a test hole program was
utilized. Sampling depth was chosen to consistently approach the
groundwater surface, due to its shallow occurrence.
Assuming hole drilling to be a constant cost for all sampling
methods, program cost depends on sampling and analytical
method. Reusable sampling apparati included syringes, split
spoons, Tenax Tubes and galvanized pipe (probes). The down-
auger methods were cheapest from both labor and equipment
standpoints. The Tenax Tube method required a fair amount of
preparation, a small amount of extra sampling time (20 min per
sample) and extra analytical extraction and cleanup time (25 min
per sample). The Tenax Tube method also required extra equip-
ment (desorption unit and pumps). Equipment cost on a per-hold
basis is difficult to define because most items can be reused and
amortized. As each sampling technique became efficient, it was
evident that minimal collection time differences existed and that
sample analysis was the major differential cost factor. The time
costs associated with sample analyses were presented in Table 3.

CONCLUSIONS
The five soil gas sampling techniques were previously categor-
ized as contaminant concentrating (Soil Headspace, 1-Liter,
Tenax Tube) and non-concentrating (DIA, DIS). Of the concen-
trating techniques, the Tenax Tube method proved most useful
during this investigation. The Soil Headspace method required
more cohesive soils than encountered in the field, and the 1-Liter
method would have benefitted from a different combination of
analytical equipment (cryogenic focusing and a more sensitive
detector). Of the non-concentrating techniques, the DIA method
produced slightly higher TCE concentrations, on the average,
than the DIS method (due to heat generated during drilling opera-
tions).
The sampling techniques also could be separated into auger ver-
sus probe methods. While the auger methods provided expedient
results and apparent elevated contaminant levels due to heat gen-
eration, probes offered the capability of repeated sampling and
the use of a long-term monitoring network. Due to the slow pro-
cess of molecular diffusion, long-term changes in soil gas concen-
tration will not be readily evident using the techniques outlined in
this report. It would, therefore, be necessary to provide periodic
direct mixing of probe contents, prior to sampling, to facilitate
equilibrium conditions along the entire probe length.
The choice of an appropriate sampling technique is site-
specific. Each technique proposed in this report has been sum-
marily evaluated in Table 3. Judging from the results of this in-
vestigation, the simplest and fastest technique, DIA, was sensitive
enough for TCE plume delineation. This resuh was due to the
high volatility characteristics of the contaminant, sampling depth
close to groundwater and the relatively high concentration gradi-
ents encountered (possibly due to paleochannels). The Tenax
198
MONITORING & SAMPLING
-------
Tube method provided additional information about widespread,
low-level contamination, but provided a no more detailed map
than DIA of the major plume under investigation. The other tech-
niques could have provided information comparable to DIA
across the study area, as method sensitivities appear essentially
equal (Table 2). Under different soil conditions (texture and
moisture content) it can be assumed that Soil Headspace would
provide relatively higher values than the DIA method (as dis-
cussed in theory). The 1-Liter method also should perform better
than DIA when more appropriate analytical equipment is utilized.
The precision analyses performed on probe sampling with the
DIS and Tenax Tube methods indicate low analytical variability
and relatively high field variability. Field variability can be di-
vided into two components: environmental and sampling.
Although sampling variability cannot be quantified as separate
from environmental spatial and temporal variability, it can be
minimized by adhering to effective and consistent technique. It is
hoped that, of the five soil gas sampling techniques presented
here, at least one will prove appropriate for any contaminant/en-
vironmental conditions to be investigated.

ACKNOWLEDGMENT
This material has been funded wholly or in part by the U.S.
EPA under Contract No. 68-01-7347 to Ecology and Environ-
ment, Inc. It has been subject to the Agency's review, and it has
been approved for publication as a U.S. EPA document. Mention
of trade names of commercial products does not constitute en-
dorsement or recomendation for use.

REFERENCES
1. Lappala, E.G. and Thompson, G.M., "Detection of Groundwater
Contamination by Shallow Soil Gas Sampling in the Vadose Zone—
Theory and Applications," Proc. Management of Uncontrolled Haz-
ardous Waste Sites, Washington, DC, Nov. 1984, 20-28.
2. Quinn, K.J., Wittman, S.G. and Lee, R.D., "Use of Soil Gas Sam-
pling Techniques for Assessment of Groundwater Contamination,"
Proc. Management of Uncontrolled Hazardous Waste Sites, Wash-
ington, DC, Nov. 1985, 157.
3. Nadeau, R.J., Lafornara, J.P., Klinger, G.S. and Stone, T., "Meas-
uring Soil Vapors for Defining Subsurface Contaminant Plumes,"
Proc. Management of Uncontrolled Hazardous Waste Sites, Wash-
ington, DC, Nov. 1985, 128.
4. Lyman, W.J., Reehl, W.F. and Rosenblatt, D.H., Handbook of
Chemical Property Estimation Methods, McGraw-Hill, New York,
NY, 1982.
5. Jury, W.A., Grover, R., Spenser, W.F. and Farmer, W.J., "Modeling
Vapor Losses of Soil-Incorporated Triallate," Soil Sci. Soc. Am. J.,
44, 1980, 445-450.
6. Jury, W.A., "Volatilization from Soil," Chapter 7, in Vadose Zone
Modeling of Organic Pollutants, S.C. Hern and S.M. Melancon,
Eds., Lewis Pub. Inc., Chelsea, MI.
7. SAIC, "Installation Restoration Program Phase II," Final Report
For American Lake Garden Tract, Washington, 1985.
MONITORING & SAMPLING 199
-------
Comparison of In Situ Soil Pore Gas Screenings,
Soil Sample Head Space and Laboratory Techniques
John E. Adams
Rizzo Associates, Inc.
Natick, Massachusetts
ABSTRACT
A soil pore gas screening method was developed to provide
rapid cost-effective screening of soil and/or groundwater. First,
a site survey is conducted and areas which exhibit evidence of
chemical spills are identified. Next, the site is divided into equal
size grids. Areas identified as potential spill locations and the grid
centers are sampled utilizing a soil pore gas field screening
method. Using this method, one penetrates the upper soil horizon
with an auger hole and seals it at the surface. The gas contained
in the auger hole is allowed to equilibrate and is sampled with a
flame ionization and photoionization field monitoring equip-
ment. Concurrently, a soil sample is obtained from the auger hole
and sealed in a glass container. The head space in the sample con-
tainer is analyzed with the field monitoring equipment and the re-
sults are compared. This process may be repeated at increasing
radial distances from positively identified locations in an effort to
quantify the area of soil and/or groundwater contamination.
Sampling locations are identified which are characteristic of
the range of values obtained from the field screening program and
soil samples are collected for laboratory analyses. The results of
the field and laboratory analyses are then compared. Areas of soil
contamination may then be quantified based on the correlation of
field and laboratory data.

INTRODUCTION
A cost-effective method of screening large areas for soil and/or
groundwater contamination was developed utilizing readily avail-
able field organic vapor analyzing instruments. The method
makes use of the volatile nature of many contaminants which di-
fuses into the soil pore gases of the vadose zone from contami-
nated soils and/or groundwater. Similar techniques have been
utilized in an effort to delineate areas of contaminated soils
and/or groundwater by Marrin and Thompson,1 Spittler et. al.,]
Lappala and Thompson,1 and others. The method is intended to
be used as a field screening tool which can aid in determination of
monitoring well placement, aerial extent of soil contamination as
well as providing early detection of contaminant migration orig-
inating from disposal facilities.
Monitoring well installation and laboratory analytical cost have
increased tremendously in recent years. To best utilize the funds
available to study hazardous waste sites and disposal facilities, a
method is required to aid in the placement of monitoring wells
and specify analytical sampling locations. The majority of geo-
physical techniques have been shown to be ineffective at delin-
eating low to moderate levels of organic contamination in
groundwater. Therefore, it was necessary to develop a method
with increased sensitivity for delineating areas of potential con-
tamination.
In recent years the detection limits of field organic vapor
analyzing instruments have increased significantly providing a
new and cost-effective tool in the effort to quickly and accur-
ately delineate areas of soil and groundwater contamination.
The method involves the analysis of a representative sample of
soil pore gas with field instrumentation and a comparison rela-
tive to surrounding sample points. An isocon map of soil pore
gas organic apor concentrations can then be constructed utilizing
the data. A sampling plan can be formulated from the isocon map
to skew sample collection to those areas suspected of ground-
water or soil contamination.
Soil pore gas sampling has been shown to be effective in many
geologic and physical settings making this technique an attrac-
tive alternative to random monitoring well placement or geophysi-
cal surveys. However, the technique is not without problems.
Results from different instruments may vary greatly and analyti-
cal results may demonstrate little or no correlation with the field
screening data. Results must be carefully scrutinized with respect
to representation of subsurface chemistry and always substan-
tiated with analytical data.

METHODOLOGY
A field method was developed to quickly screen broad areas
utilizing readily available instrumentation. The method involves
augering a borehole to a depth of 2 to 3 ft. A composite soil sam-
ple is collected in a glass container. The container is filled approx-
imately two-thirds full and tightly sealed with aluminum foil se-
cured with a rubber band. The sample is placed in a controlled
environment at room temperature and allowed to equilibrate for a
period of 15 to 20 min.
Concurrently, the borehole is sealed with polyethylene attached
to a 2-ft square piece of 0.75-in. thick plywood. A hole is cat
through the center of the plywood which permits access to the
borehole. The borehole is permitted to equilibrate for a similar
period.
The instruments tested for use with this method included the
Century 128 Organic Vapor Analyzer (OVA) operated in the total
screening mode, and the HNu Photoionization Analyzer with a
10.2 eV source. These instruments are regularly used for health
and safety monitoring during field investigation activities at haz-
ardous waste sites and, as such, are readily available for soil/pore
gas investigations. The OVA is a flame ionization device while
the HNu utilizes photoionization technology. Both instruments
are capable of detecting numerous organic compounds. How-
However, there are some significant differences between the
compounds which each instrument is capable of detecting. The
most significant compound affecting the data quality obtained
from soil pore gas studies is methane. The OVA is capable of de-
200 MONITORING & SAMPLING
-------
tecting this compound while the HNu cannot. In fact, methane
in high concentrations has been shown to interfere with the de-
tection of other organics by some field instruments. The HNu is
available with an 11.7 eV source which is able to detect a few addi-
tional compounds. However, the lifetime of the higher energy
source is substantially less than the standard 10.2 eV source.
Subsequent to the equilibration period, a vent hole is cut through
the polyethylene seal covering the borehole. The instrument probe
is inserted through the vent and permitted to stabilize. The quan-
tity of organic compounds detected by the instruments is re-
corded in the field note book. Similarly, the instrument probes
are inserted through the aluminum foil covering the composite
soil samples and the head space in the glass container is analyzed.
The grid spacing utilized in this type of field investigation is
particularly important. A finer grid spacing will result in more
sampling locations and a higher data density. However, the in-
creased locations will require substantially more time to sample.
Conversely, if the grid spacing is too large, it is likely that sub-
stantial areas of contamination may be undetected.
The intent of this study was to skew sampling toward detect-
ing areas of potential contamination. A 100-ft grid was overlain
on the site. A site inspection was conducted to identify any sus-
picious areas of stressed vegetation and/or stained soils. Each
grid was sampled either at the locations identified in the walkover
or in the grid center. A total of 53 locations were sampled within
agridded area. These sample locations are identified on Fig. 1.

FIELD SCREENING RESULTS
The results of the field screening program revealed substantial
differences between the data obtained from the OVA and that of
the HNu. A sample data set is presented in Table 1. The data in-
dicates the presence of four areas of soil and/or groundwater
contamination (Fig. 2). The OVA data were utilized as the initial
indicators due to their ability to detect some compounds shown
to be present on the site which were undetectable with the HNu.