PB88-L46808
FIELD STUDIES OF IN SITU SOIL WASHING
Mason & Hanger, Silas-Mason Company,
Incorporated, Leonardo, NJ
Dec 87
-------
TECHNICAL REPORT DATA
(fleaie read Jatouctioru OH the revene before comf
I. REPORT NO. 2.
EPA/600/2-87/110
4. TITLE AND SUBTITLE
Field Studies of In Situ Soil Washing
7. AUTHOR(S)
James H. Nash
9. PERFORMING ORGANIZATION NAME AND AOORESS
Mason & Hanger, Silas-Mason Co., Inc.
Post Office Box 117
Leonardo, New Jersey 07737
12. SPONSORING AGENCY NAME AND AOORESS
Hazardous Waste Engineering Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 4S268
16. SUPPLEMENTARY NOTES
tiering)
PB88-146808
B. REPORT OATF
December 1987
B. PERFORMING ORGANIZATION CODE
•
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
Contract No. 68-03-3203
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA-600/12
6. ABSTRACT
The EPA and US Air Force conducted a research test program to demonstrate the
removal of hydrocarbons and chlorinated hydrocarbons from a sandy soil by in situ
soil washing using surfactants.
Contaminated soil from the fire training area of Volk Air National Guard Base,
WI, was first taken to a laboratory for characterization. At the laboratory, the
soil was reconpacted into glass columns creating a simulated in situ environment.
Under gravity flow, 12 pore volumes of aqueous surfactant solutions were passed
through each of the columns. Gas chromatograph (GC) analyses were used on the wash-
ing effluent and soil to determine removal efficiency (RE). The results of these
tests were highly encouraging. RE's of field tests run at the fire training area
were evaluated by GC, total organic carbon (TOC) and oil and grease data. Ten one-
foot deep holes were dug in the surface of the fire pit. Surfactant solutions were
applied to each hole at a rate of 1.9 gal per sq ft per day. Soil samples, taken
from the undisturbed layers beneath each hole, were analyzed for residual contamina-
tion. Samples experiencing a flow-through of 9 to 14 pore volumes of surfactant solu-
tion still had contaminant levels comparable to 5,000-10,000 ppm prewash conditions.
The field study also included the development of a groundwater treatment process.
Measurements of TOC, VOA, and biochemical oxygen demand (8005) were decreased by 50*,
99Z, and 50Z, respectively. Treated effluent was discharged directly to~~the on-base
aerobic treatment lagoons.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
.
IB. DISTRIBUTION STATEMENT
Release to Public
b. IDENTIFIERS/OPEN ENDED TERMS
19. SECURITY CLASS (TUi Report)
UNCLASSIFIED
20. SECURITY CLASS fUtil p«|tj
UNCLASSIFIED
c. COSATi Field/Group
21. NO. OF PAGES
67
22. PRICE
EPA Form 1220-1
-------
NOTICE
THIS DOCUMENT HAS BEEN REPRODUCED PROM „
THE BEST COPY FURNISHED US BY THE SPONSORING
AGENCY. ALTHOUGH IT IS RECOGNIZED THAT CER-
TAIN PORTIONS ARE ILLEGIBLE, IT IS BEING RE-
LEASED IN THE INTEREST OF MAKING AVAILABLE
AS MUCH INFORMATION AS POSSIBLE.
-------
EPA/600/2-87/110
December 1987
FIELD STUDIES PB88-1 46808
OF IN SITU SOIL WASHING
James H. Hash
Mason & Hanger-Silas Mason Co., Inc.
P.O. Box 117
Leonardo, Rev Jersey 07737
Contract No. 68-03-3203
Project Officer
Richard P. Traver P.E.
Releases Control Branch
Land Pollution Control Division
Edison, Nev Jersey 08837
HAZARDOUS WASTE ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 1*5268
\-b
-------
NOTICE
The information in this document has been funded by the U.S. Environmen-
tal Protection Agency and the U.S. Air Force under Contract No. 68-03-3203 to
Mason & Hanger-Silas Mason Co., Inc. It has been subjected to the Agency's
peer and administrative review, and it has been approved for publication as a
USEPA document. The mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
ii
-------
FOREWORD
Today's rapidly developing and changing technologies and industrial
products and practices frequently carry vita them the increased generation
of solid and hazardous wastes. These materials, if improperly dealt with,
can threaten both public health and the environment. Abandoned waste sites
and accidental releases of toxic and hazardous substances to the environ-
ment also have important environmental and public health implications. The
Hazardous Waste Engineering Research Laboratory assists in providing an
authoritative and defensible engineering basis for assessing and solving
these problems. Its products support the policies, programs, and regula-
tions of the Environmental Protection Agency; the permitting and other
responsibilities of State and local governments; and the needs of both
large and small businesses in handling their wastes responsibly and
economically.
This report describes field activities undertaken to evaluate at
pilot-scale, techniques for surfactant-enhanced in situ soil washing. The
information in this report is useful to those who develop, select, or
evaluate equipment for cleanup of spills or waste sites or for the protec-
tion of response personnel and equipment.
For further information, please contact the Land Pollution Control
Division of the Hazardous Waste Engineering Research Laboratory.
Thomas R. Hauser, Director
Hazardous Waste Engineering Research Laboratory
ill
-------
ABSTRACT
The U.S. Environmental Protection Agency Releases Control Branch and
the U.S. Air Force Engineering and Services Center engaged in a joint
project focused on in situ washing of a fire training pit at Volk Air
National Guard (ANG) Base, Camp Douglas, Wisconsin. The washing fluids
were solutions of commercially available surfactants in water. Of partic-
ular Interest was a blend of Adsee 799 and Hyonic PE90. This blend had
previously proved successful in laboratory studies involving the cleaning
of organic contaminants from soil. A second objective was to treat contam-
inated groundwater underlying the test site.
The fire training pit had served as a site for flrefighting training
as early as World War II up until deactivation in 1979. The subsurface
soil was determined to be 85-95Z sand and 5-15Z fines. The contamination
was principally a medium weight oil (2,000-25,000 mg/kg) with some vola-
tiles (VOA analysis 5-10 mg/kg). The unconfined aquifer at 12 feet depth
is reported to be continuous to 700 feet. The same aquifer serves as the
water supply for the Camp Douglas. Mo contamination has been detected in
the wells supplying the base nor private wells adjacent to the base.
However, organic carbon levels in the groundwater under, and adjacent to,
the pit were measured as high as 700 nig/liter.
Small areas of the pit (ten squares that were one or two feet on a
side) were Isolated and surfactant solutions applied at a rate of 77 L/B?
per day for seven days. Cleaning efficiencies were determined based on
before and after oil and grease measurements. Full scale air stripping and
pilot flushing operations reduced the total organic carbon by as much as
60Z. Volatlles in the groundwater were reduced by 99Z.
iv
-------
CONTEHTS
Foreword ill
Abstract iv
Figures vi
Tables rill
Abbreviations and Symbols ix
1. Introduction 1
2. Conclusions 6
3. Recommendations 1
1». Site Characteristics 8
Soil characteristics 8
Rainfall 8
Eydrologic properties 8
Determination of soil contamination 11
Contaminants at the site Ik
Electromagnetic survey of the fire pit area . . 18
Sanitary vastevater treatment at Volk Field . . 21
5. In Situ Washing 22
Establishing a pre-test baseline 22
Test Cell Layout 22
Wash solutions 23
Washing procedures 23
Sampling and analysis 25
Discussion 27
6. Groundvater Control and Treatment 30
Requirement for groundvater treatment 30
Well field specification & performance 30
Groundvater treatment system 35
7. Analytical Methods and QA/QC Report 1»5
8. References 50
Appendix 52
-------
FIGURES
Number Page
1. Location of Yolk Field ................... 3
2. Site map shoving the fire training pit and shallov
monitoring veils installed in 1981 (see Table l) ...... 1*
3. Spring 1985 Site Study Map ................. 9
U. Soil particle size distribution of soil taken at four
depths (under the fire pit) ................ 10
5. Groundvater equipotential lines around the fire
pit area .......................... 12
6. Oil and grease values were highest near the fire pit
surface .......................... 13
7. Gas Chromatogram of hydrocarbons from a composite of Volk
Field fire training pit soil ............... 15
8. Graphs of split spoon consolidation (blows per foot)
and vapor analysis (peak height counts) during
monitoring veil drilling .................. 17
9. The contaminant plume as determined by TOO
measurements of vater samples at the vater table ...... 19
10. EM Surrey Plot shoving equi-conductivity lines
(units are millimhos /meter) ................ 20
11. Volk Field Test Site for in situ soil vashing and
groundvater treatment ................... 2k
12. SoU Wash O&G Data ..................... 26
13. Proposed model of preferential path development in
organic oil spill ..................... 28
lU. Well field layout for groundvater discharge to the
treatment system ...................... 31
vi
-------
FIGURES (continued)
Number Page
15. Equipotential lines during pumping 3k
16. Yolk Field pilot treatment for vater 36
17. EPA1s Mobile Independent Chemical/Physical Treatment
plant 38
18. Air Stripping Tower 1(0
19. Five sets of data show the reduction in volatlles
brought about by the vater treatment process Ill
20. The measured value for total organic carbon from
each of the six production veils and total veil flow. ... 1*3
21. Effect of vater treatment on TOO for four data sets kk
vli
-------
TABLES
Number Page
1. Chemicals Found in Shallov Wells ............... 5
2. Preliminary Laboratory VGA Characterization of Volk
AFB Site of Opportunity .................. 16
3. Volatile Hydrocarbon Characterization of Volk AFB
Site of Opportunity Soils ................. 16
b. Oil and Grease Measurements of Samples Taken
at 0.9 m (3 ft) Depth and 0.9 m Spacing ........... 22
5. Wash Solution Volumes and Concentrations for Volk
Field Soil Wash Pilot Study - September 198? ........ 25
6. Pumping Test of Production Well ............... 33
7. Analytical Tests and Sampling Points Table .......... 37
8. QA Summary .......................... U7
9. Comparison of Coefficients of Variation for API and
Volk Field Collocated Samples ............... 1(9
A-l Hazardous Parameters of Hydrophobic Organics ......... 51*
A-2 Hazardous Parameters of Hydrophilic Organics ......... 55
A-3 Hazardous Parameters of Hydrophilic Organics ......... 56
Till
-------
LIST OF CONVERSIONS
METRIC TO ENGLISH
To convert from
Celsius
Joule
Joule
kilogram
meter
meter
meter2
meter2
meter9
meter3
meter/second
meter/second
meter9/second
meterVsecond
meter'/second
nevton
vatt
ENGLISH TO METRIC
centistoke
degree Fahrenheit
erg
foot
foot2
foot/minute
foot'/minute
foot-pound-force
gallon (U.S. liquid)
gallon (U.S. liquid)/minute
horsepover (550 ft Ibf/s)
inch
inch2
knot (international)
litre
pound force (ibf avoir)
pound-mass (ibm avoir)
pound/foot2
to
degree Fahrenheit
erg
foot-pound-fore e
pound-mass (ibm avoir)
foot
inch
foot2
inch2
gallon (U.S. liquid)
litre
foot/minute
knot
centistoke
foot9/minute
gallon (U.S. liquid)/minute
pound-force (ibf avoir)
horsepover (550 ft Ibf/s)
meter2/second
Celsius
Joule
meter
meter2
meter/second
meterVsecond
Joule
meter9
meter9/second
vatt
meter
meter2
meter/second
meter9
nevton
kilogram
pascal
Multiply hy
1.8 T +32
1,000 E+07
E+OOt
1.000 E+06
2.119 E+03
1.587 E+OU
2.2U8 E-01
E-03
E-01
2.205 E+00
3.281 E+00
3.937 E+01
1.076 E+01
1.51*9 E+03
2.61*2 E+02
1.000 E+03
1.969 E+02
1.000 E-06
(Tp-32)/1.8
1.000 E-07
3.0^8 E-01
9.290 E-02
5.080 E-03
U.719 E-OU
1.356 E+00
3.785 E-03
6.309 E-05
7.1*57 E+02
2.5*0 E-02
6.5*i2 E-OU
5.11*1* E-01
1.000 E-03
U.UU8 E+00
U.535 E-01
U.788 E+01
ix
-------
SECTION 1
INTRODUCTION
Within reasonable economic limits, pollutants that adsorb veil to
soil are difficult to remove from the soil. Paradoxically, these pol-
lutants can leach into groundvater at concentrations above drinking vater
limits. Trichlorophenol, polychlorinated blphenyls (FOB), and polynuclear
aromatic3 (PNA) are among the presently infamous pollutants that have rela-
tively high soil adsorption characteristics. These also have EPA vater
quality criteria less than parts per billion. (See Appendix A.)
The U.S. Environmental Protection Agency (EPA) is researching
remedial methods to remove pollutants from soil. Accelerating the natural
leaching process by flushing contaminated soil in situ vith an aqueous sur-
factant solution and recovering the vash effluent from the aquifer is one
method being investigated. The soil adsorption constant (K) is a measure
of a pollutant's tendency to adsorb and stay on soil (see Appendix A). A
value of 2,000 for PCB's indicates a two hundredfold greater adsorption
(holding power) than benzene at K = 10. Benzo(a)pyrene, a toxic substance,
and oil have similar values—K • 30,000-UO,000. Grouping contaminants ac-
cording to a K value and evaluating removal efficiencies (RE) gives order
to an otherwise complex collection of chemical classes. Through this and
other ongoing research new, better and more economic remedial methods are
being pursued.
This is a report of the EPA'a and the U.S. Air Force's field evalua-
tion of in situ soil vashlng of compounds having K values between 101 and
10. From 1982 to 1985 the EPA developed soil washing technology using
surfactants. The work was conducted in laboratory studies.1 Although con-
sidered in situ vashlng of soil, the technique was not used on undisturbed
contaminated soil in the field. The Air Force was seeking processes to
clean-up 128 fire training pits at Air Force installations. This mutual
interest led to a pilot field test of In situ soil washing using surfac-
tants.
The primary objective of this Joint project was to evaluate in situ
soil washing using surfactant solutions. A secondary objective was to
provide information to the Air Force that would help develop a comprehen-
sive decontamination strategy for fire training areas of all Department of
Defense (DoD) installations.
In October of 198U the Air National Guard (ANG) Bureau in
Washington, DC and the Base Civil Engineer at Volk Field, ANG Base Camp
-------
Douglas, Wisconsin were contacted concerning the possible use of the fire
training area at Volk for the demonstration of either in situ surfactant
flushing or soil vashing (see Figure l). The enthusiastic responses led to
a November 198b meeting vith the Wisconsin Department of National Resources
(WDNR) in vhich WDNR also indicated strong support for the research
project. In May 198?, the WDNR received a more in-depth project briefing
and continued to shov their cooperation and full support for the project.
With this assurance a detailed site investigation was initiated in early
June. In September 1985 two pilot studies were carried out to determine
the effectiveness of treating contaminated groundwater and the effective-
ness of in situ vashing vith surfactants.
Historical data indicate that the fire training area vas established
in World War II and routinely received waste solvents, used lubrication
oil, and JP-1* fuel (see Figure 2). The total liquid waste deposited at the
site vas as much as 260,000 gallons. An estimated 80 percent of these
wastes burned in fire training exercises, leaving approximately 52,000 gal-
lons to leach into the soil.
In 1981, because of concerns over the pollution potential of this
site, ANG engineers conducted an exploratory site survey and sampling
project. Twelve shallow veil samples were analyzed for purgable organics
using EPA Methods 601 and 602. Table 1 summarizes the 1981 findings. The
average water table depth is 12 feet belov grade. Both chlorinated sol-
vents and fuel components entered the shallow groundvater. Soils beneath
the site contain similar contamination.
This report is in eight sections including the Introduction. Before
undertaking the field operation, laboratory tests were conducted by SAIC
Inc., La Jolla, California. Section U, Site Characteristics, is based on
information obtained during the laboratory study as well as two field
studies and a literature review. Knowing the site characteristics is Im-
portant to understand the setting for this specific work. The reader
should be able to contrast or compare this work to other sites. The in
situ vashing field work is described and discussed In Section 5. At this
particular site (and most likely any site having significant vadose zone
contamination above an unconfined aquifer) groundvater control and treat-
ment is an integral part of In situ soil vashing. The development of a
treatment process from bench scale testing at the site through operation of
the treatment system is given in Section 6. The quality evaluation of the
data obtained during the field study is provided Section 7. Conclusions
and recommendations are in Sections 2 and 3. These are based on the bench
and full scale work done in the field to determine a suitable groundvater
treatment process and the in situ soil vashing. References are given in
Section 8.
-------
N
\
Not to Scale
Figure 1. Locacion of Volk Field
-------
N
O
E-5
o
Q-17
Unimproved
Access
Road
O
N-14
O
CMS
Scale In Feet
so
75
0
0
Legend
Boundary of Training
Area
Bore Hole Location
Groundwater Flow
Direction
Figure 2. Site nap showing the fire training pit and shallow monitoring
wells installed in 1981 (see Table 1).
-------
TABLE 1. CHEMICALS FOUND IN SHALLOW WELLS
(A-l to P-16) Volk Field 1981
(in micrograms per liter)
EPA Method 601
I.D.
Number
A-l
B-2
D-4
F-6
G-7
H-8
J-10
K-ll
L-12
N-lU
0-15
P-16
Chloro-
form
2.3
2.3
1.5
1.1
59.0
130.0
< 1.0
1.3
< 1.0
50.0
< 1.0
120.0
TCA
< 1.0
< 1.0
7.8
39.0
36.0
< 1.0
< i.o
< 1.0
< 1.0
< 1.0
< 1.0
< 10.0
bTCE
< 1.0
< 1.0
22.0
100.0
U2.0
< 1.0
< 1.0
< 1.0
< 1.0
< 1.0
< 1.0
< 10.0
EPA Method 602
Benzene
U 500.0
10.0
570.0
lUOOO.O
31000.0
1900.0
< 1.0
< 1.0
< 1.0
8.5
< 1.0
1*000.0
Toluene
2700.0
100.0
2100.0
8000.0
36000.0
5700.0
< 1.0
U.6
< 1.0
< 1.0
< 1.0
< 50.0
Ethyl
Benzene
270.0
10.0
190.0
950.0
6800.0
200.0
< i.o
< 1.0
< 1.0
2.9
< 1.0
1000.0
(a) trichloroethane
trichloroethylene
-------
SECTION 2
CONCLUSIONS
1. In situ soil washing of the Volk Field fire training pit vith aqueous
surfactant solutions is not measurably effective. It is likely that
this same ineffectiveness vould occur at other chronic spill sites
that have contaminants vith high soll-sorption values (K >10^).
2. In situ soil washing requires groundwater treatment and washing ef-
fluent treatment. Groundwater treatment at this site was successful
with the simple addition of lime. Air stripping removes the volatile
organics. Advantages at this site that facilitated groundwater
treatment operations were its remoteness for workable air emission
limits and a local sewage treatment system (aerobic lagoons) owned by
the responsible party. TOC levels were reduced to one half the Ini-
tial values by precipitation with lime to allow direct discharge to
the aerobic treatment lagoons. These favorable conditions are not
expected at all sites.
-------
SECTION 3
RECOMMENDATIONS
Based on the findings of this study the following recommendations are
made:
1. Regarding the technology of surfactant-enhanced in situ soil
washing, a study should toe conducted to identify the reasons
why surfactant washing of undisturbed Yolk field soil failed.
(A possible reason is presented in Section 5 and Figure 13 of
this report.)
2. Regarding the cleanup of the specific Volk Field Fire Training
Pit,
a. The present well field should be pumped to remove the
contamination in the aquifer. The water should be pumped
directly to the existing treatment lagoons on the base.
b. To remove contamination from the Volk Field fire pit
soil, the pit should be bermed and a fluid distribution *
system should be placed over the pit for recharge. The
natural leaching of the soil should be accelerated by
recycling a portion of the groundwater. additional
withdrawal wells should be drilled further down gradient
of the pit. They should be drilled deeper than 35 feet
to give the screened sections of the wells more exposure
in the plume.
c. The excavation and washing of the soil in a system such
as the EPA's Mobile Soil Washing System should be
evaluated. The basic effect would be to isolate the con-
tamination predominantly associated with the fines of the
soil.
-------
SECTION k
SITE CHARACTERISTICS
Before the pilot scale treatment studies that are reported here
Mason & Banger made a field study for the Air Force in May of 1983. The
study determined the extent and type of contamination at the fire training
pit (see Figure 3). The work performed at that time included: drilling
and sampling in shallow bore holes, installing seven monitoring veils to
UO feet (12.2 meters) depth, the determining of the vater table height and
gradient, determination of permeability, and sampling and analyzing soil
and vater samples for volatileB and total organic carbon.3*
SOIL CHARACTERISTICS
The soil beneath the fire training pit is contaminated vith vaste
oils, JF-1» Jet fuel, and solvents used in maintenance around the base.
The effect of such contamination on the soil is obvious vhen compared vith
adjacent clean soil. The most obvious difference is color and lack of
vegetation. The surface and near-surface soil of the pit is black, cohe-
sive, and free of any grass except at the edges. The pit emits an odor of
fuel oil, and the soil has enough residual contamination to feel oily.
The local soil has a thin natural organic layer that supports a grass
cover. It is sandy, non-cohesive, and light brown in color belov the top
soil. The grain size distribution of the soil in the vadose zone under
the pit Is 95 percent sand vith 5 percent by veight finer than sand. The
local soil Is also sand but vith 10 to 15 percent finer particles (see
Figure U). Mineralogically both soils are at least 98 percent alpha
quartz and have no clay as determined by x-ray diffraction. The top of
the fire pit vaa covered vith a U-inch layer of 60AO gravel/sand. Under-
lying oil and vapors have infiltrated upward and contaminated the cover.
RAIHFALL
The average annual rainfall over the last 29 years Is 29.08 inches.
The highest 2fc-hour rainfall was k inches. This was in 1976. Fluctua-
tions in the vater table from 3.5 feet to 10.5 feet were measured in a
nearby upgradlent drinking vater veil betveen 1950 and 1966.
HYDROLOGIC PROPERTIES
The soil type at Volk Field is Boone fine sand. According to the
Soil Conservation Service Engineering Field Manual, this is in hydrologlc
soil group A. Group A has high infiltration rates and lov runoff poten-
8
-------
'--^ «V
• original wells
o monitoring
o bore hole
(rf°
r
\
gravel pile
f I ."
\ «\
^ FIRE TRAINING
AREA
>c^.
\
X
V— ------
Drain hole In berm
80
feet
roadway
N
100
_JI
Figure 3. Spring 1985 Site Study Map.
-------
100
SPLIT SPOON SAMPLES FROM ET-5
K
bl
-O.fl -0.6 -0.4.
-1.2
LOG of PARTICLE DIAMETER (mm) „ . „.
2.5* -f 10* o 20* A 25* Sample Depth
Figure 4. Soil particle size distribution of soil taken at four depths
under the fire pit.
10
-------
tial. Standing-water after a rainfall demonstrates the fire pit has a low
infiltration rate. Runoff through a drain hole in the term has spread
contaminants to surface soil adjacent to the pit.
The water table is in a highly weathered sandstone, and the aquifer
extends to a depth of TOO feet. In terms of permeability the soil of the
unsaturated zone below the pit has a laboratory measured value of U x 10"^
to 5 x 10 cm/sec. The permeability of the unconfined aquifer is 5 x 10~
2 cm/sec.' According to measurements of water table elevations made at
the site, natural groundwater flow increases in speed compared to the
background flow as it passes under the pit (see Figure 5). This is con-
sistent with the measured lack of fines at the water table.
The mobilized contamination leaves the site via the groundwater in
an initially easterly direction and then turns to travel northeast. The
volume of soil and groundwater directly involved in this study was ap-
proximately U,600 cubic yards. The pores of the soil contain ap-
proximately 230,000 gallons of contaminated water. Indirect measurements
using an electromagnetic (EM) survey technique indicate a large plume
leaving the site. In addition, the total volume of contaminated water
pumped from beneath the pit between September 7 and November lb, 1985 was
U6b,000 gallons. This is more than twice the volume of water contained in
the study volume. Analytical data shows the contamination levels from the
well field were not significantly lower after pumping.
DETERMINATION OF SOIL CONTAMINATION
To determine the concentration of non volatile contamination, oil
and grease (O&G) tests were run on 36 soil samples taken at various depths
and locations over the area of the pit. The oil and grease test requires
the soil sample be air dried for 24 to 36 hours before extraction with
carbon tetrachloride (Cdj^). Volatiles in the soil were therefore not
contributing to the mass of extract obtained. The quantity of oil and
grease extracted was measured two ways. The first was by infrared absor-
bance at a wave number of 2910 cm'1. This is equivalent to a wavelength
of 3.1*36 microns. Because the O&G values were so high for most of the
samples it was possible to evaporate the carbon tetrachloride on a steam
bath and weigh the residue in a beaker. Agreement between these two
methods was quite good. As expected, the concentrations determined after
evaporation from the steam bath were slightly less than those calculated
from the infrared method. Figure 6 shows the overall distribution of CClj^
extractable oil and grease as a function of depth. Oil and grease values
were highest near the fire pit surface, decreased In soil that was deeper
and then increased in soil that was slightly below the water table. Work
conducted in November of 198U measured chromatagraphable allphatlcs,
aromatics and "unresolved" compounds. These values were an order of mag-
nitude lower than the O&G measurements.
The total amount of extractable material in the pit soil was calcu-
lated by segmenting the soil column below the pit into 10 equal
thicknesses; determine the average concentration of the 10 imaginary slabs
11
-------
AIR
STRIPPER
10 » 10 «o 10
Figure 5. Groundwater equlpotentlal lines around the fire pit area from
September treatment study before pumping.
-------
o>
.X
< a
LJ0
14
DEPTH IN FEET
Figure 6. Oil and grease values were highest near the fire pit surface.
-------
of soil; and then, multiply the average by the weight of the soil. Using
this approach and referring to Figure 6 for the concentrations* the total
ertractable hydrocarbons is 1,700 gallons.
CONTAMINANTS AT THE SITE
Much of the vadose zone contamination at the fire pit is lubrica-
tion oil (see Figure 7). Comparatively smaller quantities of volatile or-
ganic s and oxidized hydrocarbons are present. Early in the project the
contamination of the aquifer was thought to be a floating layer. This is
not the case. The contaminants in the aquifer are water soluble and have
penetrated the aquifer. A possible explanation for this is intense
biological activity in the soil that could have been brought about by the
firefighting foam used in the training exercises.
The non-volatile chemicals, principally the oil and grease on the
soil, comprise the majority of the contamination in the unsaturated zone
(5,000 to 20,000 mg/kg). The oil and grease, a carbon tetrachloride
(CCljj) extraction of the soil, contains oxidized oils and greases indicat-
ing weathering. The oxidized forms are more water soluble than the non
oxidized forms and are in greater abundance deeper in the water table.
Not all the hydrocarbon contamination is extractable with CCL^. The total
organic carbon measurement on a contaminated water sample was 760
mg/liter. Oil and grease on the same sample was only 20 mg/liter. The
depth of contamination into the aquifer is not known. However, monitoring
well (ET-6) was drilled to 1*0 feet (30 feet into the aquifer). Samples
from the bottom of this well were contaminated (see Figure 5).
The chemical species vary with depth and distance from the pit.
Chlorinated volatile organic compounds are low in concentration at the pit
surface. As depth increases, the measured level of volatiles increases
(600-3500 ppb). Chlorinated volatiles detected in the soils are:
dichloromethane, chloroform, 1,1,1, trichloroethane, trichloroethylene.
Chemicals indicative of long term weathering, isoprenoid compounds, were
measured. Non-chlorinated chemicals in the groundwater include benzene,
toluene, xylene, and ethylbenzene which are all principal components of
Jet fuel (see Tables 2 and 3). The groundwater has a soluble organic con-
tent of up to 760 mg/liter carbon. In the absence of oxygen within the
aquifer, the organic material remains soluble. Along with a high quantity
of iron in the water the organic material is partially flocculated into an
organic-iron complex when exposed to air. Also on exposure to air the
volatile organics will begin to volatilize from 10 or 20 mg/liter to 2 or
3 mg/liter in a few hours. pH for the well field effluent is 5.5 to 6.0.
After a few hours of exposure to the atmosphere pH will rise to 7.5 or
8.0.
During the installation of monitoring wells the driller's boring-
log was supplemented with gas chromatography (GC) measurements of the
"head space" of soil samples. Given in units of total counts of peak
height, the GC values represent a rough estimate of the distribution of
volatiles. Figure 8 shows four such logs. The first log is from an
14
-------
R«produc«d from
b.tl .v.llabl* copy
Figure 7. Gas Chromotogram of hydrocarbons from a composite of Volk Field
fire training pit soil.
-------
TABLE 2. PRELIMINARY VGA CHARACTERIZATION OP VOLK APB SITE OP OPPORTUNITY BOILS
Sample/Depth
Conpound Concentration (ppb)
DlcbloroBetbane Chloroform 1.1.1-Trlealoroetbane Trlebloroethylene
Total Chlorinated
Solri ota (ppb)
Pit fl/Surfaee (l)
Pit Il/Surface (2)
Pit fl/1.5 feet
Pit 11/3 feet
Pit *l/5 feet
Pit fl/Burfaee
Pit 12/2.5 feet (1)
Pit 12/2.5 feet (2)
HD»
9.U8
HD
321.
301
8.08
285
3W
HD
HD
HD
167
189
HD
158
2kk
13.6
Ik.k
fcl.2
HD
2.960
8.68
53.0
HD
HD
10.6
136
HD
HD
HD
HD
HD
13.6
3k. 5
177
fc91
3.1.50
1T.3
U96
590
a - HD Indicates "Hot Detected"
TABLE 3. VOLATILE HYDROCARBON CHARACTERIZATION OP VOLK AFB SITE OF OPPORTUNITY SOILS
Sagple/Depth
1 Pit fl/8urfaee (l)
• Pit fl/Burfaee (2)
Pit 11/1.5 feet
Pit fl/3 feet
Pit 11/5 feet
Reaolved Compounds (ug/g)
Allpbatle Fraction Aromatic Fraction
132
3*1
138
218
622
30.2
1.2.1
33.3
82.li
77.U
Polar Fraction
8.75
58.6
8.29
lh.5
1.59
Total UCM (ug/g)
1.160
1.590
b98
206
1.58
Pit 12/Surface 71. U
Pit /2/2.5 feet (l) 65.0
Pit 12/2.5 feet (2) 8.33
66.2
lt.0
8.31
3.97
0.56li
0.1.85
•25T-
87.0
15.8
-------
ET-5
NXII. COUNTS
100 300
1OO
1 I '
300
4 6
LOO 10
ET-4
NXfl. COUNTS
1 I ' I ' I '
100 400
LOO fO
LOG tO
•
Figure B. Graphs of split spoon consolidation (blows per foot) and vapor
analysis (peak height counts) during monitoring well drilling.
-------
upgradlent veil, the rest, from highly contaminated veils. Note that the
Counts scale is a log scale.
Before installing monitoring veils, bore holes were made in and
around the fire pit. These are half darkened circles in Figure 3. Water
samples, taken from each bore hole at the water table, vere analyzed for
total organic carbon (TOG). Figure 9 is a plot of the plume at the top of
the water table based on these TOC measurements.
ELECTROMAGNETIC SURVEY OF THE FIRE PIT AREA
Because of vork reported by the New Jersey Geological Survey it was
felt that an electromagnetic (EH) survey had a potential for success at
Vblk Field.' By using an induced electromagnetic field in soil or rock
structure it is possible to measure differences in the conductivity of the
soil or rock. More precisely the difference arise more from conductive
solutions in the pore spaces. In the case of the vork done at a Naval air
station in Nev Jersey, residual fuel, left over from fire training, had
entered an unconfined sandy aquifer. The plume was mapped by the EM sur-
vey. Since organic contaminants seldom alter the conductivity of
groundwater it was a surprise when measurable differences in conductivity
in the aquifer mapped out in the form of a reasonable plume. The reason
for the conductivity was attributed to the fire fighting foam "AFFF."
An electromagnetic survey was conducted around the Volk pit area.
The survey included. The instrument used was an EM-3b manufactured by
Geonlc Ltd. Mlssissaugua Ontario, Canada. The EM-31* consists of a 2-foot
diameter coil of wire that transmits a burst of electromagnetic energy at
a low frequency. This induces electromagnetic excitation in conductive or
semiconductive material. A second coll spaced at 10, 20 or 10 meters from
the transmitter receives the initial burst from the transmitter and the
induced signal from the ground. These received signals are electronically
transformed into a conductivity value for the "half space" betveen the
coils. This technique maps large soil structure and not small targets
such as 55 gallon drums. By moving the coils over an area of land in a
grid pattern conductivities of half spaces are measured and plotted on a
map. The coil spacing used on the survey that produced Figure 10 was 20
meters. This results In a 30-meter depth of penetration of the induced
signal. For an explanation of hov the coll spacing affects the depth see
Reference 8.
Figure 10 is the resulting conductivity map near the pit. The ini-
tial easterly path of the plume is different from local groundvater motion
which is to the northeast. Examination of the drilling logs reveals a
less consolidated sandstone (fever blows per foot) in that area, affording
the plume an easier route to the east. A turn to the north is required to
get the overall path of the plume on the northeast course. A piece of
data to help support the possibility of the plume reaching the point
marked with an "S" in the figure is analytical data on soil taken from the
drip line of an environmentally stressed (dying) tree. The sample was
taken from a depth of 12 feet (3.7 meters) using a 2-inch diameter hand
18
-------
Figure 9. The contaminant plume as determined by TOC measurements (shown In
mg/1 on the lines) of water samples at the water table.
-------
Figure 10. EM Survey Plot showing equi-conductivity lines
(units are millihos/meter).
20
-------
auger. The analysis shows an oil and grease content in the 100 mg/kg
range. An Infrared spectrum trace Indicated the presence of slightly
oxidized oil (like the oil and grease found in the veil field). No AFFF
vas analytically identified in the plume.'* Iron is up to 150 times
more concentrated in plume water than background water. The conductivity
differences are likely due to dissolved iron in the plume.
SANITARY WASTEWATER TREATMENT AT VOLK FIELD
Treated water from the fire pit was discharged to the "base sewage
treatment system. The Wisconsin Air National Guard at Volk Field under
the control of the Base Civil Engineer maintains two aerobic lagoons con-
taining a combined 20 million gallons to treat domestic sewage generated
at the base. In addition, the town of Camp Douglas has its sewage treated
at the base. Twice a year, the second of the two lagoons is discharged
Into a tributary of the Lemonwler River.
The VDNR discharge permit requires the following effluent
limitations:
BOD5 17 mg/1
Suspended Solids 17 mg/1
pH 6.0-9.0
Ammonia Nitrogen 3.0 mg/1
Dissolved Oxygen 6.0 mg/1 minimum
The use of Volk Field as a training school is seasonal. The heaviest
usage is in the summer when "The Guardsmen" are doing their two weeks of ac-
tive duty. Sewage pumped from the first to the second lagoons is recorded
daily and can be as much as 300,000 gallons per day (GPD) at a BODe of up to
120 rag/liter. During periods of low activity pumping is around 80,000 GPD.
The VDNR placed a limit on the effluent sent from the fire pit treatment
to the base lagoons. The limit was 60 pounds per day BODe. Since BODe
requires 5 days to determine it was necessary to anticipate what will happen
in 5 days to maintain continuous pumping. An attempt was made to correlate
total organic carbon measurements with BODe.. There is a positive correlation
but a specific relationship between the two could not be demonstrated.
21
-------
SECTION 5
IN SITU WASHING TESTS
ESTABLISHING A PRE-TEST BASELINE
It was important for the test that the soil vashed in situ be undis-
turbed. The wash fluid's path could not be influenced by pre-test sam-
pling methods that created preferred hydraulic paths. To establish the
level of contamination for the specific test cells, six samples were taken
adjacent to the test cells to establish pre-vash levels. The measured O&G
values appear in Table U. These concentrations varied as much as -73? and
+50$ from the average. Since coefficients of variation for replicate O&G
measurements average 12%, this variability, -73!? to 50$, for pre-vash
levels is significantly troublesome when analyzing the post-wash data.
This is all within a 10-foot square section of the pit (2.5% of the pit
area).
TABLE It. OIL AND GREASE MEASUREMENTS OF SAMPLES TAKEN
AT 3 IT DEPTH AND 3 FT SPACING
Sample No. 1*055 ^56 1*057 ^058 1+059 !»060
O&G mg/kg 5l»00 1850 5800 5050 1050 1*060
TEST CELL LAYOUT
A photograph of the test site is in Figure 11. The well field to
withdraw contaminated groundwater is in the foreground of Figure 11. In
the right center of the picture are the test cells used in the soil wash-
ing evaluation. The EPA's Mobile Independent Chemical/Physical Treatment
Unit is in the background. The Air Force's Air Stripping Tower is to the
left. Both are being used to treat contaminated groundwater.
Test holes were dug in the fire pit to determine the soil washing
ability of a number of surfactant solutions. The locations of the holes
were chosen to provide as near a uniform contamination level as could be
predicted from oil and grease measurements and elevation observations of
the surface of the pit. Ten holes were dug. Five of the holes measured 2
foot x 2 foot x 1 foot deep and five of the holes were 1 foot x 1 foot x 1
22
-------
foot (see Figure 11). The depth of the holes matched the planned depth of
soil that would be scraped off the surface of the pit prior to full scale
remedial washing. One foot is the depth below which the carbonized oil
layer is located and at a depth where suitable percolation rates were
measured.
WASH SOLUTIONS
Two terms describe the surfactants used, "synthetic," and "natural."
The synthetic surfactants are those that have been man-made by chemical
processes and are available commercially. The natural surfactants are
those that have their origin at the fire training pit itself, and are by-
products of biological activity. The synthetic surfactants used for the
pilot treatment study were:
1. Surfactant 1 (Si). A mixture of ethoxylated fatty acids sold
by Vitco Chemical Corporation. Used in agriculture as a soil
penetrant.
2. Surfactant 2 (S3). An ethoxylated alkyl phenol sold by Diamond
Shamrock.
3. Surfactant 3 (S3). An anionic sulfonated alkyl ester sold by
Diamond Shamrock.
The natural surfactants are described as:
1. Old natural surfactant - from the first 10,000 gallons pumped
from the well field.
2. New natural surfactant - from the well field after pumping
20,000 gallons, lower in suspended solid than the old natural
surfactant.
3. Clarifier effluent - lime treated well field effluent lower in
iron and therefore less likely to plug the soil pores with
precipitates.
WASHING PROCEDURES
The rate of addition of wash solution was 3 inches per day. This
corresponds to 1.8? gallon per day for the one square foot test holes and
T.U8 gallon per day for the four square foot test holes. Wash solution
was added four times a day for either U or 6 days depending on the
availability of the solution. Percolation in two test holes stopped after
the first day. Tests in these holes were abandoned. A third hole was
abandoned two days later for the same'reason. The remaining seven holes
were then rinsed with only two or seven gallons of clean, upgradient, well
water after the washing. Before finishing the rinse period, U Inches of
rain fell over a 3-day period. The test holes offered significantly less
resistance to percolation of the runoff than the rest of the fire pit.
23
-------
Figure 11. Volk Field Test Site for in situ soil washing, and groundwater treatment,
-------
Rainwater penetrated the surface layer of gravel fill and floved laterally
to the test holes. Attempts to keep the holes from filling up vith runoff
using berms at the surface vere ineffective. Approximately 1100 cubic
feet of rainwater fell on the fire pit during the 3-day rain. Not all of
it vent through the test holes, but some fraction of it vas seen floving
down the vails of the holes from above the black layer of oil. If 5 per-
cent of the rain that fell on the pit floved into the test holes that
vould be equivalent to 33 inches of rain in each of the holes.
Therefore, the rinse phase of the pilot study vas extensive. The
mobility of contamination from other areas of the pit to the test holes is
unknown. The 0.25 inch deposit of fine soil at the bottom of each hole
vas distinctly darker than the soil originally at the bottom. It vas
sharply defined by texture and O&G, and vas easily parted from the soil
directly belov. Table 5 lists the total volume of vash solution used in
each test hole, the identity of the vash solution, and its concentration.
TABLE 5. WASH SOLUTION VOLUMES AND CONCENTRATIONS FOR VOLK FIELD
SOIL VASH PILOT STUDY - SEPTEMBER 1985
Pit #
Wash
Solution
Concentration
(% w)
Total
Volume
Gallons
Pore
Volumes
1
2
3
k
5
6»
7»
8«
9»
10
natural surfactant +53
new natural surfactant
old natural surfactant
new natural surfactant
old natural surfactant
clarlfier effluent
S3
50/50 S1/S2
clarifier effluent
0.025*
0.02U
0.02k
0.02k
0.02k
0.015
0.5
1.5
0.015
81* plugged
122
168
112
112 plugged
28
1*2
k2
28
0 plugged
7
10
lU
9
9
9
1U
lU
9
0
*1 sq. ft cross section holes
SAMPLING AND ANALYSIS
After vashing and rinsing the soil belov each of the test holes,
samples vere taken of: the surface of fine material at the bottom of each
hole, soil from 2-k inches and soil 12-1U inches belov the bottom of the
hole. The samples vere placed in vide mouth glass Jars. The Jars vere
then placed in cartons for transportation to the analytical lab at the
EPA's Oil and Hazardous Materials Simulated Environmental Test Tank
(OHMSETT) facility in Leonardo, Nev Jersey. At OHMSETT the soil samples
vere extracted to determine oil and grease and the extracted fluids vere
analyzed using an infrared spectrophotometer. The bar charts in Figure 12
25
-------
2" TO 4" O&G VALUES
TAKEN ATTCR WASHING
PREWASH CONC1 ITERATIONS
NS
NS CLAM 1186
WASH SOLUTION
SO/SO CLAR
12" TO 14" O&G VALUES
TAKCN AFTER WASHING
dor 1186 SO/50
WASH SOLUTION
Figure 12. Soil wash O&G data Volk Field from pilot field study.
26
-------
aid in the understanding of the data. The line across each bar chart rep-
resents the pre-wash O&G value. The dotted line shovs one standard devia-
tion in the pre-wash data.
Additional soil washing screening was done in the lab which included
a shaker test and infrared scans of extracts of the wash fluid. An
analysis of the natural surfactant before and after mixing with pit soil
in an erlenmeyer flask showed an increase in total organic carbon (TOO) of
83 ing/liter. The increase in the TOO of the fluid represented only 0.7
percent of the oil and grease present in the soil. If additional washings
with the natural surfactant would remove the same quantities of oil and
grease, lU2 such washings would be required to remove all of It. Plain
tap water was twice as effective in this shaker table study as the natural
surfactant.
In previous laboratory work* a column packed with Volk Field soil was
washed with 12 pore volumes of the 50/30 blend solution and rinsed. The
results of these tests were highly encouraging. In the field work up to
lU pore volumes passed through the test soil with no apparent effect on
contaminant removed. In spite of the repeated successes of the engineered
surfactant to clean contaminated soils in laboratory tests, there is no
evidence that the soil was cleaned, in-situ at Volk Field. Within the
statistical limits, there is no significant difference between soil that
had been washed and soil that had not been washed.
DISCUSSION
Using surfactants to wash soils in situ did not measurably work.
There is no obvious cause for the Ineffectiveness of the blended surfac-
tant at Volk Field. However, after working with data from the project for
over a year one plausible explanation can be advanced. Early analytical
data derived from soil samples resulted In what, at the time, was ques-
tionable data. Specifically one set of data pair indicated that soil
samples taken within 3 inches of each other differed in TOO by an order of
magnitude. Oil and grease measurements were already discussed as being
significantly variable from point to point. Another aspect of the soil
under the fire pit is its apparent lack of fines. There is a difference
between contaminated vadose zone soil and non-contaminated vadose zone
soil. Because of the greater surface area available in fine soil an equal
mass of fine soil particles will hold more contaminants than coarse
particles.
The assumption might not be valid that given enough time the oily
contamination will equalize Its distribution for any given horizontal sec-
tion under the pit. The process demonstrated (in two dimensions) in
Figure 13 is an explanation for the poor in situ washing of Volk Field
soil. These drawings show the development of preferred paths for rain-
water to flow through the soil. These paths would be greater in available
interparticle space. This means there are less fines in these pathways
and less contaminant.
27
-------
Reproduced from Jfti
best available copy.
Figure 13. Proposed model of preferred path development in organic oil spill.
28
-------
The surfactant solution followed the paths of least resistance when
applied to the top of the undisturbed soil. These paths have less fines
and less contamination. For purposes of this argument assume the surfac-
tant used at Yolk Field did remove 802 of the contamination of the soil in
the preferred path and 5% from the non-preferred path. With an order of
magnitude difference in contamination between the tvo paths the resulting
measured cleanup would be minimally effective.
29
-------
SECTION 6
GROUNDWATER CONTROL AND TREATMENT
REQUIREMENT FOR GROUNDWATER TREATMENT
The treatment pilot study included the groundvater treatment along
vith the soil washing because of the unexpected high level of contamination
in the groundvater below the fire pit. A number of parameters reflect the
poor condition of TOO measurements averaged above 200 parts per million.
Biochemical oxygen demand was around 50 ing/liter. Iron content was 150
times higher than background levels. The pH was 5.5 to 6.0 and required 60
to ISO ppm lime to bring the pH to 7.0 to 7.5. Verified chlorinated
species Included trlchloroethylene and trichloroethane. Volatile hydrocar-
bons were at a concentration of 10-20 ing/liter.
VEIL FIELD SPECIFICATION & PERFORMANCE
The well field consisted of six production wells. This is in addi-
tion to seven monitoring wells that were installed U months earlier.
Figure Ik shows the location of each well. The wells were drilled using a
5-7/8 inch auger and wash bore method. The wells were cased with U-inch,
schedule-J»0 threaded PVC tube. Screens were also schedule bO FVC with a
slot size of 0.010 inch. Back fill around the screen was flint sand #30.
Each production well had a stainless steel submersible pump with a
low water shut off monitor. A throttle valve was at each well head. The
wells were connected to a common pipe to carry the contaminated water to
either the Chemical/Physical treatment system or the temporary lagoon. To
prevent water from entering a well if it was not pumping, check valves were
installed at each well head.
During the drilling operation split spoon soil samples were taken at
5-foot intervals. At the same time penetration tests were run to determine
the degree of consolidation of the soil. Using a lUO-pound weight the
split spoon sampler is driven into the soil through the end of the hollow
stem auger. The number of blows required to drive the sampler 1 foot. Ac-
cording to penetration values and the drill operator's observations
sandstone is encountered at 11 to Ik foot depth. The water table is at 10
to 12 foot depth. Figure 12 shows the equipotential lines of the water
table near the pit.
The first pumping test was attempted using the production well WW-1..
The yield of that well was so low the test was postponed until VW-2 was
30
-------
u>
lined lagoon
e withdraw! well
o monitor well
roadway o
so
feet
Figure 14. Well field layout for groundwater discharge to the treatment system.
-------
ready. Siace the purpose of the first pimping test vas to get a quick
evaluation of the aquifer directly belov the pit the pumping lasted for
only 8 hours. The rate of pumping vas 12 gpm. Drawdown vas measured in
the production veil itself and in monitoring veils ET-2, ET-5, with ET-3.
In addition, an old monitoring veil J-10 vas cleaned out and used along
with production veils WW-1, and WW-3. Table 6 summarizes the dravdown and
recovery data.
In Table 6 the elevation of the water in each of the veils is in
feet from mean sea level at the time specified to the left of each rov.
The top rov of numbers reports the elevation of each of the veil heads. It
is interesting to compare the data from the veil with the poorest produc-
tion capacity, WW-1, and ET-3. These veils are 12.5 feet from each other
and are in a zone that yielded the highest contamination levels. The draw-
down in the aquifer at monitoring veil ET-3 is 0.2 ft. No such drop is
shown in WW-1. The data from ET-2 and ET-5 vas taken electronically using
pressure transmitters and is reported to three decimal points. The rest of
the numbers were calculated from measurements using a resistance probe on a
pre-measured wire to detect the top surface of the water. Since there was
no rainfall within 5 days prior to the test it is not surprising that ini-
tial and final elevations for most of the veils are equal. The 0.1 foot
drop in WW-2 probably indicates it has not fully recovered from the pump-
ing. Monitoring veil J-10, installed In 1980 could very easily have
plugged up again.
The second pumping test was done to evaluate the effect of all six
production veils operating at once. Using the ball valve at each of the
veil heads, individual veils vere throttled to maintain a constant flov so
there vere no lov water level cutoffs. The maximum production from all six
wells was 29 gpm. Equipotential lines during the pumping of WW-1, 3, U and
5 are shovn in Figure 15.
The veil field can handle recharge through the pit at 3" of fluid
per day. The water table gradient belov the fire pit is small, and in
spite of the high permeability of the soil the lateral water flov through
the area is only 1.7 inches per day. By devatering at 28-30 gallons per
minute, an artificial gradient vas established that reverses the downstream
gradient. The local vater table is depressed by 0.5 feet and the gradient
instead of being 0.001 to the east is 0.006 to the south and vest. By
using the method of Keely and Tsang11 to determine the radius of capture
around a withdrawal veil and considering the demonstration veil field as
"one veil" and only half the aquifer thickness (90 meters instead of 180
meters), the radius of capture is 85 feet (26 mj. If fluid vere being
recharged to the aquifer in the form of a wash solution, the rather safe
margin of an 85 foot radius would be reduced. By pumping 28 gallons per
minute for 2k hours a total of U0,320 gallons would be pumped. The
proposed 3" of fluid per day for in situ washing of the 75 foot diameter
fire pit would require 8,260 gallons. This 20 percent additional flov into
the recovery system should not create a situation where vash fluid would
escape since maximum daily rainfall at Volk Field is not expected to exceed
k inches.
32
-------
TABLE 6. PUMPING TEST OF PRODUCTION WELL VW-2
VOLK FIELD FIRE TRAINING PIT
(Elevation in feet mean sea level)
Well Head
Elevation
Time
(min)
919.09 919.36
WW-2 J-10
919.33
ET-2
918.75
ET-5
917. U6
ET-3
917.92
WW-l
917.36
WW-3
0
0.5
1
1.5
2
2.5
3
3.5
U
5
6.5
7.5
10
15
16
2k
29
J»5
60
75
120
150
180
2ltO
300
360
U80
1*95
510
51*0
600
660
720
780
81*0
900
960
1020
1200
1320
902.59 902.61
896.19
895.59
895.09
891*. 79
89U.79
89^.79
89k .69
89U.69
902.56
89^.59
902.56
891* .1*9
89^. 39
89U. 29
902.56
89!*. 09 902.16
902. hi
902.36
89^.09
902.59 902.56
902.63
902.578
902.51*6
902.536
902.536
902.51
902.502
902.1*92
902.U79
902.1*72
902.1*61*
902.1*1*8
902.1*63
902. U73
902. U77
902.1*85
902.1*91*
902.525
902.51*6
902.563
902.577
902.589
902.601
902.621*
902.629
902.55
901.977
901.91*
901.932
901.91*3
901.911*
901.907
901.897
901.882
901.891*
901.851*
901.851*
901.875
901.895
901.902
901.915
901.925
901.953
901.972
901.985
901.997
920. on
902.021
902.0U1
902.01*7
902.1*6
902.U6
902.1*6
902.1*6
902.36
902.36
902.36
902.36
902.26
902.26
902.1*6
903.32
903.32
903.32
903.32
903.32
903.32
903.32
903.32
903.32
903.32
902.61
902.56
902.56
902.51
902.51
902.51
33
-------
Figure 15. Equlpotentlal lines during pumping of the well field at 14-16 gpm.
-------
GROUHDWATER TREATMENT SYSTEM
To keep the contamination in the withdrawn groundvater from "being a
hazard surface treatment of the contaminated water was planned. The ini-
tial concept vas to use an equalization pond followed by air stripping.
However, extensive precipitation of iron-organic floe required a solids
removal pretreatment. Withdrawn fluid was subjected to a number of bench
scale treatments. 2 These treatments were: addition of lime, sulfuric
acid, hydrogen peroxide, alum, ferric chloride, polymers and emulsion
breakers. The pilot scale tests used a simple groundvater treatment
process. The flow diagram from the aquifer to the effluent end of the air
stripper1^ is in Figure 16. The numbered sampling points are along the
flow path. These numbers correspond to the number in the left-hand column
of Table 7.
The first step in the process occurs in the well at the backfill
around the screen. As the contaminated water enters the well casing and
descends to the pump it is exposed to air. In a normal well this would be
of little significance. However, with the particular contamination at Volk
Field this is the first change from a near anaerobic (lack of oxygen) to an
aerobic condition. The water enters the pumps and then flows to the flash
mixer tank where lime is added and a mixer stirs the contents of the 600
gallon tank. Depending on the pumping rate, the residence time in the
flash mixer varied between 20 minutes and 70 minutes. Both the addition of
lime and the addition of oxygen caused the formation of a brown
precipitate. This brown precipitate is visually the same as that found in
the initial pumping done in May of 1985 (^ months earlier). Oxygenation
from rain water percolating down from the surface or the minor amount of
oxygen above the capillary zone is enough to cause this precipitation and
create a blanket of brown fluid around a clear contaminated fluid. The
source of the brown color is iron. Tests run by the Air Force confirm the
presence of high levels of iron in contaminated water—as much as 52
mg/liter.
The lime was added in a slurry through a metering pump. The lime
was added at a rate that would cause the effluent from the flash mixer to
be between T.O and 7.5 pH. Control of the slurry pump was independent of
the well field. For short periods of time values aa low as 6.0 and as high
as 9.7 were measured. Depending on the amount of lime added the reduction
In contamination was between 20 and 60 percent. Interestingly the color of
the effluent from the flash mixer at a pH around 7.0 was a chocolate brown.
At higher pi the color was more intensely orange.
From the flash mixer the water and newly formed precipitate entered
the bottom of a 1*500 gallon clarifier. The clarifier allowed a 3 to U hour
settling time for the precipitate (see Figure 17). Effluent from the over-
flow weir of the clarifier passed to the 16,000 gallon lined aeration
lagoon. Additional oxygenation created more precipitate and a subsequent
lessening of soluble contaminants. With an average retention time of 1*8
hours, volatiles were released to the air.
35
-------
LIME
WELL
•th
VOLATILE*
CLARIFIER
DV PASS
5
9
in
LAGOON
lAIR
g
TO AEROBIC
LAGOON
Figure 16. Volk Field pilot treatment for water.
-------
TABLE 7. ANALYTICAL TESTS AND SAMPLING POINTS FOR THE
WATER TREATMENT PROCESS
Pt
No.
Description
Tests
Performed
Approximate
Values, Average
Or Range
Individual veil head
Well field effluent
Flash mixer effluent
Clarifler effluent
Air stripper feed
Air stripper
effluent
7
8
9
Clarifier
Clarifier bottom
Soil
volatile organic
total organic
chemical oxygen demand
oil and grease
pH
volatile organic
total organic
iron
pH
chemical oxygen demand
flov rate
total organic (dissolved)
suspended solids
pH
flov rate
total organic
suspended solids
pi
flov rate
volatile organic
total organic
temperature
flov rate (water)
oil and grease
volatile organic
total organic
flov rate (air)
oil and grease
biochemical oxygen demand
chemical oxygen demand
suspended solids
suspended solids
oil & grease
10-20 mg/liter
60-760 mg/liter
6-500 mg/liter
0.2-U6 mg/liter
5.1-6.2
10-20 mg/liter
250 ± lUf mg/liter
32 mg/liter
6.0 ± 0.2
Ul mg/liter
t-28 gpm
160 mg/liter
350 mg/liter
6.8-9.7
9-28 gpm
205 ± 1% mg/liter
13.6-10U mg/liter
7.6
9-28 gpm
3.5 - 7.0 mg/liter
151 ± 13* mg/liter
6-15°C
15-20 gpm
3.6 mg/liter
0.5-0.3 mg/liter
lU6 mg/liter
215 cu.ft/min
3.6 mg/liter
2.5 mg/liter
180 mg/liter
U.lt mg/liter
2331 mg/liter
800-16000 mg/kg
37
-------
00
M SHUT-OFF VALVE
CHECK VU.VE
/Figure 17. EPA's Mobile Independent Chemical/Physical Treatment plant.
-------
Volatiles from the groundwater passively entered the atmosphere from
the clarifier and lagoon. Final removal of the remaining volatile organics
occurred in a packed tower air stripper provided by the HQ AFESC Environics
Division (see Figure 18). The tower contained an 8 foot high, 1.5 foot
diameter column of 3/8 inch Pall-Rings and can treat 50 gpm flovs at an air
to water ratio of U0:l. Flow rate through the air stripper averaged 28 gpm
creating an air to water ratio of approximately 80:1. This vas helpful to
the overall performance particularly for less volatile xylenes and toluene.
The air stripper effectively removed an average of 96 percent of the
volatiles entering the stripper and reduced all chlorinated hydrocarbons to
less than 5 ppb (according to Air Force VOA measurements).
To compare data on volatile organics between well field effluent and
air stripper effluent, the time the air stripper effluent entered the
process was calculated. For example, the fluid leaving the air stripper on
the 306th hour entered the process at the 258th hour. To compare the
reduction in volatiles, sets of triplets (well field effluent, air stripper
feed, and effluent) were established by accounting for the residence time
in the treatment system. Five triplets were identified and appear below in
Figure 19. Specific compounds that were identified in the air stripper
feed samples were:
1. Trichloroethane
2. Benzene
3. Trichloroethylene
U. Toluene
5. Ethylbenzene
Total organic carbon measurements were by far the most numerous and
precise of all the analyses run on the groundwater. With coefficients of
variations between 1 and 5 percent it is well worthwhile to use TOO values
as a basis of discussion to characterize the aquifer contamination.
The wen field, of six individual veils, did not yield a fluid of
uniform contamination over time. This is attributed to the wells that had
high contamination. Water velocity through the aquifer is highest at the
well therefore fine particles put into suspension during drilling are
cleared rapidly during development. Farther away from a properly operating
well, fine particles are less likely to be in suspension from the drilling
and less likely to become suspended since the flow velocity is not sig-
nificant. A properly developed well should not plug and should have a
yield consistent with the aquifers potential to supply the water. In this
well field the wells producing water with the higher contamination levels
are doing so at lower production rates. Particulate matter, fine sand and
organic-iron complexes associated with the contamination, are restricting
flow. Contaminated wells Intermittently "cut off" because of lov water in-
side the well casing. Particularly susceptible to this were wells WW-1 and
WW-5. Production wells WW-1 and WW-5 each contributed only 1.5 to 2 gal-
lons per minute of water. TOO in water for this well was between 600 and
700 mg/llter. Production well WW-5 had TOC's around 500 mg/liter. AS
these wells cut on and off TOCej rose and fell rapidly in well field ef-
39
-------
*-* • •'. ^
Figure 18. Air Stripping Tower.
40
-------
$00 -r
8OO -
1= 700 -
O
^ 600 -
ui
If so°-
JE o 400 -
§ 300 -
u 200
100
0
^^••^MWB
well
1
3 •*. , ^
jn strip //
k %
V '/\
^ ^:
s>v gu^, //^
O
i
^ %
V ^
i
b «
TRIPLET
•
i
^
I
^MMB— V
1
4^
1
|
d •
SET
Figure 19. Five sets of data show the reduction In volatiles brought about
by the groundwater and air stripping treatment process.
-------
fluent samples. Figure 20 is a graph of the TOO values measured for the
six veils in the veil field between September 9 and 27, 1985 and measured
TOO values for the entire veil field. Although the individual veils main-
tained a near constant level of TOC over the 380 hour time span the collec-
tive effect was a lover average TOC from the veil field. This can be ex-
plained if there was a shift in the balance of the veils and some of the
less contaminated veils started yielding more fluid.
After vlthdraval from the veil field the groundvater vas mixed vith
lime in a flash mixer. As vas stated earlier in this section and reported
in the bench study, the addition of lime brought about the formation of a
brown precipitate. The precipitate consisted of organic matter and an iron
hydroxide. On September 19 samples were taken from the flash mixer. The
process flow vas Ik gpm. pH in the flash mixer varied betveen 6.3 and 9.1*
by changing the lime dosage. Water coming into the flash mixer had a TOC
value of 270 mg/liter and a pE betveen 5.8 and 6.1. The TOC of the super-
natant (clear water phase) coming out of the flash mixer vas 2UO mg/liter
at a pH of 6.7. A pH of 9.^ vas attained when lime vas overdosed. The TOC
dropped to lb5 mg/liter.
After reconciling the difference in time a slug of fluid passes the
various process stages, four sets of data were prepared in a bar chart to
show the change in TOC. The chart shovs the veil field effluent, the
clarifier effluent, the air stripper feed and finally the air stripper ef-
fluent to the sever (see Figure 21). The second bar in each set is the
clarifier effluent. The organic content in each set decreased betveen the
time the fluid left the clarifier and entered the air stripper (third bar).
It is during this time the fluid is in the 16,000 gallon lagoon.
Precipitation and volatilization continue in the lagoon. It was only a
period of 6 hours at a flow rate of 9 gpm that this pH vas maintained.
Comparing the difference betveen TOC change in the air stripper with the
change in volatiles content from the gas chromatography test shovs a rather
weak correlation.
42
-------
o>
E
^./
u
o
800
700 -
600 -
500 -
400 -
300 -
200 -
100 -
0
•f ww—2
100
200
300
400
LAPSED TIME IN HOURS
ww-3 A ww—4 X ww—5 7 ww-
360
o
o
u
o
o
a
CD
r
o
a.
>^
u
o
340 -
320 -
300 -
280 -
260 -
240 -
220 -
200 -
180 -
160 -
140 -
120 -
100
250 270 290 310 330 350
LAPSED TIME (hours)
370
390
410
Figure 20. The measured value for total organic carbon from each of the six
production wells (top) and the average TOC (bottom).
43
-------
\
SET 1
SET 2
SET 3
SET 4
Figure 21. Effect of water treatment on TOC for four data sets
after four steps in the process.
-------
SECTION 7
ANALYTICAL METHODS AND QA/QC REPORT
The analyses vere directed at common volatiles and the total non-
volatile organic compounds. Identification of the volatile compounds vas
limited to benzene, toluene, ethylbenzene, trichloroethylene, and trich-
loroethanes. There are more unresolved volatiles present.
The methods of evaluation vere:
1. Gas chromatography, to determine the loss of volatile com-
ponents. A ten tube automatic sampler, sample concentrator and
microprocessor-controlled G.C. vere used. The temperature
program called for 50°C for b min. temperature increase to
220°C at 8°C/min. The column used vas 1/8 I.D. x 10* stainless
steel packed with 60/80 Carbopack B/l* SP-1000.
2. Total organic carbon (TOC) to determine treatment effects on
all the organics. Water samples from the vaste stream vere run
on a Dohrmann Model DC-60 Total Organic Carbon (TOC) analyzer
using a lov temperature ultraviolet reactor vhere a pure oxygen
atmosphere converts all the organic carbon into carbon dioxide.
The quantity of carbon dioxide in the effluent gas is measured
in a nondispersive infrared unit and is related to the organic
content of the original water sample.
3. Suspended solids. To determine the physical form of the con-
taminant from Standard Methods for the Examination of Water &
Vastevater Ho. 22U.
b. pH vas measured as a control function using an Orion pH meter
and an Accu pE combination electrode in the EPA mobile
laboratory. pH sensors and transmitters vere placed in the
treatment system. The transmitters vere connected to a data
logger located in the mobile laboratory.
5. Chemical oxygen demand (COD) vas measured as "ultimate"
biological oxygen demand. Standard Methods for the Examina-
tion of Water & Vastevater No. 220.
45
-------
6. Biochemical oxygen demand (BOD,-) was measured as a regulation
requirement by the State of Wisconsin. Maximum allowable BODc
per day was 60 pounds. Standard Methods for the Examination
of Water & Wastevater No. 219.
7. Oil and grease (O&G), which is the material that can be ex-
tracted with carbon tetrachloride, measurements were used to
determine the amount of residual Jet fuel and lubricating oil.
8. Dissolved oxygen (DO) measurements were made on selected
groundwater samples to monitor the oxidizing of the con-
taminants in recovered water. The BOD^ test also requires DO
measurement. Standard Methods for the Examination of Water &
Wastevater Ho. 218A.
9. Drawdowns and pumping rates were measured to monitor the core
of influence and fluid removal volumes respectively. This was
done manually using a continuity tester on a dropline and also
electronically with pressure transmitters connected to a data
logger.
The quality assurance objectives set forth in the Quality Assurance
Plan written for the project were not met. Host notably, volatile organic
analysis with a coefficient (CV) of variation of 31%. Interlab data from the
Environics Lab, Tyndall Air Force Base shows the field measured data to have
an average positive bias of 3l£ in the parts per million range. Collocated
analyses in the hundreds of parts per billion range has a coefficient of
variation of 150 to 210?. The interlab comparison with Tyndall shows a sig-
nificant positive bias of 3^0 percent. The Performance Evaluation Standards
set by the EPA-Edison labs were also measured with a high positive bias. The
standard was in the tens of parts per billion range and the bias was + 2000%.
Although it is unfortunate that these biases are so high in the lover ranges,
the purpose for the VOA measurements is not compromised. The purpose of the
VGA measurements was to be sure the 15 pounds per day volatiles emission was
not exceeded. See Table 8 for a summary of the CV's.
Biochemical oxygen demand measurements showed a negative bias of 9%
compared to collocated samples run at Tyndall. The plan called for ±5%.
Chemical oxygen demand (±5? in the plan) had a CV of 35?. Oil and grease QA
objectives were ± 5% and turned out to be ± 12? on replicate runs and ± 6%
when two methods were compared. The strongest data was from TOG analyses.
Replicate samples had a CV of 2%\ collocated samples, a CV of 12%. Table 8 is
a summary of this QA data.
The QA objectives stated in the plan were over ambitious. In addi-
tion, the attention given to correcting analysis difficulties in the field was
limited. The field laboratory apparatus was mostly purchased or rented for
this work and because of time restraints was delivered directly to Volk. The.,
field team never worked with the equipment before starting to make the
measurements reported.
46
-------
TABLE 8. QA SUMMARY OF COEFFICIENTS OF VARIATION FOR MEASUREMENTS MADE ON VOLK FIELD SAMPLES
Replicate
•G.C. VOA
Performance
Evaluation
Veil Pield
Effluent
Retention
tlaes
Peak
Belgbte
Bensene
Toluene
Etnylbenxeae
. Air Stripper
Peed
Bencene
Toluene
Btbylbeocene
; Effluent
Benzene
Toluene
; Etbylbencene
TOO
Veil Field
Clar trier EFF
Air Stripper
Feed
Air Stripper
Effluent
Oil a Oreaee
Soil
BOO
COD
Are CV
(ppa) (1)
0.22 6$
0.12
31
_ _
_ .
-
_ _
_ .
-
_ .
_ ±1
tm •»
•A 2.0
• • •
_ .
• -
.
HA 12
III ve Gravity
HA 6
-
_ —
Ho. of
TeetB
3
2
2
_ .
.
-
.
_
-
_
_
-
167
^
.
. -
•
9
Method
8
-
_
Ave
(ppm)
—
_
-
3.3
2.8
0.28
•
0.67
0.67
o.ob
0.13
0.13
0
-
251
205
151
1*6
d039
-
29
203
Collocated
CV
(*>
—
1.0
•6
kk
67
72
•
lOb
109
220
210
ISO
0
-
.13
7
13
15
38
e»
51
35
Ho. of
Banplee
—
It
h
21
21
21
11
11
11
9
9
9
-
22
16
15
15
6
-
b
$
Ave CV
(ppa) (f)
0.01 Hot
Reported
_ -
• -
2.5
2.0
0.23
0.56
0.51
O.Ob
0.03
0.03
0.003
-
— _
_
^ ..
•
-
•• ~
32
.
Interlab
Ho. of
Samplea
3
-
•
2
2
2
3
3
3
*
*
*
-
—
.
•
.
-
-
3 '
—
Lab
EPA
-
-
Tyndall
Tyndall
Tyndall
Tyndall
Tyndall
Tyndall
Tyndall
Tyndall
Tyndall
-
—
_
-
-
-
-
Tyndall
_
8 ample
Type
—
-
•
Collocated
Collocated
Collocated
Collocated
Collocated
Collocated
Collocated
Collocated
Collocated
-
—
-
-
•
-
Collocated
.
-------
la claiming that the objectives vere over ambitious reference is made
to the American Petroleum Institute report Refinery Wastevater Priority Pol-
lutant Study - Sample Analysis and Evaluation of Data.Analytical data for
that report was obtained from three laboratories (tvo private and one EPA con-
tract lab). A comparison of the CV's from that study and the Voli Field work
appears in Table 9. In all but one listing — Lab A's VOA - the field data at
Volk Field has a lover coefficient of variation than the other seven listings
48
-------
TABLE 9. COMPARISON OF COEFFICIENTS OF VARIATION FOR API
AND YOLK FIELD COLLOCATED.SAMPLES
API Refinery Study ""Volk Field Study
Lab A Lab B EPA OEMSETT
VOA 2h% 119JK 1W* 72*
TOC 1*5 35 69 13
Oil & Grease 112 - 1*7 38
49
-------
SECTION 8
REFERENCES
1. Ellis, W. D. and J. R. Payne. Treatment of Contaminated Soils With
Aqueous Surfactants (Draft Interim Report) to EPA Releases Control
Branch, September 6, 1985. EPA-600/2-85/129, NTIS PB 86-122
561/REB.
2. Hazardous Materials Technical Center, Installation Restoration
Program Records Search prepared for 820Uth Field Training Site, Wis-
consin Air National Guard, Volk Field, Camp Douglas, Wisconsin,
August 198U.
3. Mason & Hanger-Silas Mason Co., Inc. Field Study of the Fire Train-
ing Area. Volt Field ANG Draft Report, July 17, 1985 to EPA Releases
Control Branch.(Internal Report to EPA)
b. McNabb, G. D. et al., Chemical Counteraeasure Application at Voile
Field Site of Opportunity.EPA report September 19, 1985.
(Internal report to EPA;
5. Bradbury, K. R. and E. R. Rothchild, "A Computerized Technique for
Estimating the Hydraulic Conductivity of Aquifer From Specific
Capacity Data," Ground Water Vol 23, No. 2, March-April 1985.
6. Guire, P. E., et al., "Production and Characterization of Emulsify-
ing Factors From Hydrocarbonoclastic Yeast and Bacteria," Mierobial
Degradation of Oil Pollutants. Pub. No. LSU-SG-73-01 Louisiana
State University, Baton Rouge, LA. 1972
7. Andres, K. G. and R. Crance, Proceedings of the NWWA/API Conference
on Petroleum Hydrocarbons and Organic Chemicals in Ground Water,
"Use of the Electrical Resistivity Technique to Delineate a
Hydrocarbon Spill in the Coastal Plain Deposits of Nev Jersey."
November 5-7, 198U, NWWA & API.
8. Technical Note TN-6 "Electromagnetic Terrain Conductivity Measure-
ment at Lov Induction Numbers." Geonics Limited Mississauga Ontario
Canada. October 1980.
9. Chan, D. B., Analytical Method of Aqueous Film Forming Foam (AFFF),
September 1978. Civil Engineering Laboratory, Naval Construction
Battalion Center, Port Hueneme, California 9301*3, TM No. M-5U-78-08.
50
-------
10. Pink, P. T., AqueouB Film Forming Foam Treatability. 1978 Civil and
Environmental Engineering Development Office (Air Force Systems
Command), Tyndall Air Force Base, Florida 32U03
11. Keely and Tsang. A Handbook for the Use of Mathematical Models for
Subsurface Contaminant Transport Assessment. Earth Sciences Divi-
sion, Lawrence Berkeley Labortory, University of California,
Berkeley, CA 9^720 report to the USEFA, January 1983, TAG IAD89F ZA
175
12. Mason & Hanger-Silas Mason Co., Inc. Volk Field Contaminated
Groundvater Bench Scale Treatability Studies, September 198$. Draft
report.November 1985 to EPA Releases Control Branch.
13. Stallings, R. L., T. N. Rogers, Packed-Tover Aeration Study to
Remove Volatile Organics from Groundvater at Wurtsmlth Air Force
Base. Michigan. June 1985.Engineering & Services Laboratory, Air
Force Engineering & Services Center, Tyndall Air Force Base, Florida
321*03.
lU. Radian Corporation, Refinery Wastevater Priority Pollutant Study -
Sample Analysis and Evaluation of Data.December 1981 to American
Petroleum Institute, Environmental Affairs Department 2101 L Street,
Washington DC 20037
51
-------
APPENDIX A1
The categories of organic compounds in Tables A-l to A-3 vere based on
the logarithm of the octanol/vater partition coefficients (log P) of the com-
pounds, as follows:
Hydropbobic organics: log P 3.00
Slightly hydrophilic organics: log P 1.00, 3.00
Hydrophillc organics: log P 1.00
The log P is a measure of the tendency of a compound to dissolve in hydrocar-
bons, fats, or the organic component of soil rather than in water. For in-
stance, many hydrophobias, some slightly hydrophllics, and no hydrophillcs
vere detected in soil, which contains organic components that tend to adsorb
other organics; only groundwater samples contained any hydrophilics (see
Table U) [table not included]. This does not mean that only hydrophobics and
slightly hydrophilics are found in soil, but they are normally found more than
hydrophllics are.
Not only is the log P a measure of the tendency of a compound to dis-
solve in octanol, fat, or oils, it can also be used to estimate the tendency
of an organic compound to become (or remain) adsorbed in soil. Several re-
searchers have published regression equations relating log F to the soil ab-
sorption constant, KQC or K (Lyman et al. 1982). The partitioning of a com-
pound between the organic components of soil and a water solution is expressed
as follows:
^ _ g adsorbed/g organic carbon
oc ~ g/ml solution
The adsorption tendency is mainly dependent on the weight of organic carbon
(oc) in the soil. If the organic carbon content of a soil is known, then the
soil adsorption constant (K) can be derived from KQC (Lyman et al. 1982):
K _ % organic carbon /.,• \
K 100 (KOC)
£ _ g adsorbed/g soil
g/mL solution
1. Taken from report Chemical Countermeasures for In Situ Treatment of Hazard-
ous Material Releases by JRB Associates, 8UOO Westpark Drive. McLean. Vir-
ginia. EPA Contract No. 68-01-3113, Task No. 29, JBB project No. 2-817-03-
956-29. pp. U-10 to U-lU.
52
-------
Thus, K con be used to estimate what fraction of a compound will be adsorbed
on soil and vhat fraction will remain dissolved in water when the soil and
water are in equilibrium with each other.
The K values for the waste compounds found in soil and groundwater at
Superfund sites are presented in Tables A-l through A-3 [original Tables 6
through 8) for hydrophobic organics, slightly hydrophillc organics, and
hydrophllic organics, respectively. They were obtained from published data
(Lyman et al.t 1982) or calculated from log P values (Eansch and Leo, 1979).
Besides the 17 hydrophobic compounds found in soil, another 7
hydrophobic compounds were found in groundwater near the Superfund sites.
These compounds may have been found in the soil if analyses were made, but
groundwater samples are analyzed more often than soil samples in Field Inves-
tigation Team investigations. The same is true for the 17 slightly hydrophilic
organics and the 10 inorganic contaminants measured in groundwater but not in
soil.
U.I.2 Significant Human Health Hazard
The substances for which countermeasures are most needed are those
likely to cause significant adverse health effects in the exposed population.
Several measures of the human health risk are available, and the EPA Water
Quality Criteria are most appropriate. A large proportion of the chemicals
reported at Superfund sites are carcinogenic or at least acutely toxic. The
EPA Water Quality Criteria for carcinogens are expressed as levels presenting
a known increase in risk, rather than as safe levels. These are presented in
Tables A-l through A-3, [original 6 through 9-9 not included] along with
median acute lethal dose data (LDeg's) for rats, and whenever available,
lowest carcinogenic dose data (TDLo's) for all listed carcinogens. Clearly,
although both are carcinogenic, the carcinogenic potency of PCB's (TDLo: 1220
mg/kg is much less than that of dloxln (TDLo: O.OOllU mg/kg), and the TDLo
values allow one to assess relative carcinogenic hazard.
53
-------
TABLE A-l. HAZARD PARAMETERS OF HYDROPHOBIC ORGANICS
Substance
Soil Adsorption Quality Cri
Constant K1*2 (ppm)
EPA Water
Quality Criteria
Chlordane
Dieldrin
Anthracene
Benzo( a)anthracene
Benzo(a)pyrene
Fluoranthene
Pyrene
DDT
Bis ( 2-ethylhexyl ) phthalate
Di-n-butyl phthalate
o-Dichlorobenzene
PCBs
Dioxin 2
Naphthalene
Oil
Grease (5,
1,2, U-Tr ichlorobenzene
Hexachlorobut adiene
Trichlorophenol
Ethyl benzene
Bis ( 2-ethylhexyl )Adipate
Cyclohexane
Benzo(b)pyrene
1 . 1 ,2-Trichlorotrif luorethane
* Corresponds to an incremental
** Estimated based on n-Cje
*** Estimated based on n-C9Q
200
200
700
60,000
1(0,000
8,000
2,000
10,000
20,000
100
70
2,000
,000,000
600
(30,000) ••
000,000) *»*
200
200
2,000
50
90,000
70
1(0,000
60
Increase in
U.6xlO-T*
-
^ £
2.8x10'°*
2.8xlO-6*
O.OU2 ,
2.8xlO-6*
2.UxlO-8»
15
3U
0.1(
7.9xlO~7*
-
-
-
-
,
l(.5x!0 *
-
1.1(
-
2.8xlO~6»
-
cancer risk of 10
54
-------
TABLE A-2. HAZARD PARAMETERS OF HYDROPHILIC OHGANICS
Water
Soil Adsorption Quality Cri
Substance Constant K1'2 (ppm)
Xylene 30
Phenol 20 3.5
Carbon Tetrachloride 20 O.OOU
Methylene Chloride 5
Perchlorethylene 20 0.008
Toluene 30 lU.3
Trichloroethylene 20 0.0027
Dichlorophenol 30 1.1»
Methyl Chloroform 20
Vinylidene Chloride 10 -
Chloroform 10 1.9xlO~u
Ethyl Chloride 6
Fluorotrichloromethane 20 ,
Ethylene Bichloride 6 9.1»xlO~4«
Methyl Isobutyl Ketone 5
Vinyl Chloride 6 0.002 .
Benzene 10 6.6xlO~4«
1,2-Dichloroethylene • 10 -
1,2-Diphenylhydrazine 20 U.2xlO~5«
Tetrahydropyran U
1,1-Dichloroethane 6
Chlorobenzene 20 O.U9
2-Ethyl-l»-methyl-l(3,-dioxolane 10
Isopropyl Ether 9
* Corresponds to an incremental increase in cancer risk of 10
55
-------
TABLE A-3. HAZARD PARAMETERS OF HYDROPHILIC ORGAJTICS
Sub stance
Acetone
Methyl Ethyl Ketone
Acroleln
Tetrahydro furan
l,li-Dioxane
Acrylonitrile
Isobutanol
2-Propanol
Soil Adsorption
Constant K1'2
0.7
1
0.8
2
1
o.u
2
1
EPA Water
Quality Criteria
(ppm)
_
0.32
5.8xlO~5»
_
* Corresponds to an incremental increase in cancer risk of 10~°
56
------- |