EPA/600/2-91/023
June 1991
EVALUATION OF SOIL WASHING TECHNOLOGY:
Results of Bench-Scale Experiments
on Petroleum-Fuels Contaminated Soils
by
Mary E. Loden
Camp Dresser & McKee Inc.
Cambridge, Massachusetts 02142
Contract No. 68-03-3409
Project Officers
Richard P. Traver
Chl-Yuan Fan
Superfund Technology Demonstration Division
Risk Reduction Engineering Laboratory
Edison, New Jersey 08837
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45286
-------
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before co'n/ilch
1. REPORT NO. 2
EPA/600/2-91/023
3
A. title and subtitle
EVALUATION OF SOIL WASHING TECHNOLOGY: Results of Bench
Scale Experiments on Petroleum Fuels Contaminated Soils
5. REPORT DATE
June 1991
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
Mary E. Loden
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Camp, Dresser & McKe.e
Cambridge, MA 02142
10 PROGRAM ELEMENT NO.
CBRD1A
11. CONTRACT/GRANT NO.
68-03-3409
<2. SPONSORING AGENCY NAME AND ADDRESS
Risk Reduction Engineering Laboratory--Cin., OH
Office of Research and Development
U.S. Environmental Protection Agency
ninrinnaM DH
13. TYPE OF REPORT AND PERIOD COVERED
1 A, SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
Project Officers: Richard P. Traver/Chi-Yuan Fan Comml: (201) 321-6924
pre- o 2^1
tfie^T?*0Environmental Protection Agency through its Risk Reduction Engineering
Laboratory's Releases Control Branch has undertaken research and development efforts
to address the problem of leaking underground storage tanks (USTs). Under this
effort, EPA is currently evaluating soil washing technology for cleaning up soil
contaminated by the release of petroleum products from leaking underground storage
tanks. Soil washing is a dynamic physical process which remediates contaminated
soil via two mechanisms --particle separation and dissolution of the contaminants
into the washwater. As a result of the washing process, a significant fraction of
the contaminated soil is cleaned and can be returned into the original excavation
or used as cleaned "secondary" fill or aggregate material. Since the contaminants
are more concentrated in the fine soil fractions, their separation and removal from
the bulk soil increases the overall effectiveness of the process. Subsequent treat-
nent will be required for the spent washwaters and the fine soil fractions.
The soil washing program evaluated the effectiveness of soil washing technology in
removing petroleum products (unleaded gasoline, diesel/home heating fuel, and waste
crankcase oil) from an EPA-developed Synthetic Soil Matrix (SSM) and from actual
site soils. Operating parameters such as contact time, washwater volume, rinsewater
volume, washwater temperature, and effectiveness of additives were investigated.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Jnderground storage, Storage tanks,
Leakage, Cleaning--washing, Soils,
Gasoline, Fuel oil, Diesel fuels
Volume reduction, Soil
washing, Particle separa-
tion, Underground storage
tanks, Soil contamination
Spill cleanup, Synthetic
Soil Matrix, Waste crank-
case oil
I
18. DISTRIBUTION STATEMENT
IELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. ^JJAGES
20 SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Fetm 2220-1 (R»». 4-77) previous edition uobsolete
i
-------
-------
DISCLAIMER
The information in this document has been funded by the U.S.
Environmental Protection Agency under Contract 68-03-3409 to CDM Federal
Programs Corps. It has been subjected to the Agency's peer and
administrative review, and has been approved for publication. 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 vith them the increased generation
of materials that, if improperly dealt vith, may threaten both human health
and the environment. The U.S. Environmental Protection Agency (EPA) is
charged by Congress vith protecting the nation's land, air, and vater
resources. Under a mandate of national environmental lavs, the agency
strives to formulate and implement actions leading to a compatible balance
betveen human activities and the ability of natural resources to support
and nurture life. These lavs direct the EPA to perform research to define
our environmental problems, measure the impacts, and search for solutions.
The Risk Reduction Engineering Laboratory is responsible for planning,
implementing and managing research, development, and demonstration programs
to provide an authoritative, defensible engineering basis in support of the
policies, programs and regulations of the EPA vith respect to drinking
vater, vastevater, pesticides, toxic substances, solid and hazardous
vastes, and Superfund-related activities. This publication presents
information on current research efforts and provides a vital communication
link betveen the researcher and the user community.
An area of major concern to the Risk Reduction Engineering Laboratory
is the impacts associated vith uncontrolled releases of petroleum
hydrocarbons from underground storage tanks. This document provides
information on soil vashing technology for cleaning up soils contaminated
vith petroleum products.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
ii i
-------
-------
ABSTRACT
The U.S. Environmental Protection Agency through its Risk Reduction
Engineering Laboratory's Release Control Branch has undertaken research and
development efforts to address the problem of leading underground storage
tanks (USTs). Under this effort, EPA is currently evaluating soil washing
technology for cleaning up soil contaminated by the release of petroleum
products from leaking underground storage tanks. Soil washing is a dynamic
physical process which remediates contaminated soil via two mechanisms -
particle separation and dissolution of the contaminant into the washwater.
As a result of the washing process, a significant fraction of the
contaminated soil is cleaned, and can be returned into the original
excavation or used as cleaned "secondary" fill or aggregate material.
Since the contaminants are more concentrated in the fine soil fractions,
their separation and removal from the bulk soil increases the overall
effectiveness of the process. Subsequent treatment will be required for
the spent washwaters and the fine soil fractions.
The soil washing program evaluated the evaluated the effectiveness of
soil washing technology in removing petroleum products (unleaded gasoline,
diesel/home heating fuel, and waste crankcase oil) from an EPA-developed
Synthetic Soil Matrix (SSM) and from actual site soils. Operating
parameters such as contact time, washwater volume, rinsewater volume,
washwater temperature and effectiveness of additives were investigated.
The additives investigated were CitriKleen (a biodegradable degreasing
agent) and an organic surfactant. Actual soils from UST sites in Ohio and
New Jersey were washed using the optimum parameters derived from the SSM.
This report was submitted in fulfillment of Contract No. 68-03-3409 by
CDM Federal Programs Corporation under the sponsorship of the U.S.
Environmental Protection Agency. This report covers a period from May 1989
to August 1990, and work was completed as of August 1990.
iv
-------
-------
CONTENTS
Foreword iii
Abstract iv
Figures vi
Tables vii
Abbreviations and Symbols viii
Acknowledgement x
1. Introduction 1
2. Conclusions 7
3. Recommendations 10
4. Materials and Methods 11
Pure Product Description 11
SSM Characterization 11
SSM Blending Procedures 12
5. Experimental Procedures 16
Bench Scale Soil Vashing Standard Operating Procedure.. 16
Field Anaytical Program 21
Field COD and TSS Procedures 25
6. Results and Discussion 27
Synthetic Soil Matrix Blending 27
Phase I - Determination of Operational Parameters 38
Phase II - Full Matrix Soils Testing 44
Field Analytical Program Results 48
TPH Analytical Study 54
References 65
Bibliography 66
Appendices 70
A. Sensitivity Analysis Curves for Operational Parameters on
Gasoline and-Diesel Contaminated SSM-Phase 1 71
B. Particle Size Distribution Curves for Bench Scale SSM Soil 82
v
-------
-------
FIGURES
Number Page
1 Soil Vashing Process Diagram 2
2 Bench Scale Soil Washing Experiment Schematic 17
3 Soil Washing Standard Operating Procedure 20
4 Soil BTEX Concentration Equation 23
5 TSS Concentration Equation 25
6 SSH Particle Size Distribution 28
7 Moisture Retention Curve 28
8 Liquid Limit Determination 29
9 Dose/Response Curve for Gasoline 35
10 Dose/Response Curve for Diesel Fuel 35
11 TPH or BTEX Percent Removal Equation 40
12 Rinsevater to Washvater Ratio Effect on Percent BTEX
Removal 42
13 Temperature Effect on Percent TPH Removal with 0.132
CitriKleen 42
14 Contact Time Effect on Particle Size Distribution 43
15 Comparison of Field and Lab Analytical Data 52
16 Test T-Value Equation 57
17 Pooled Estimate of the Population Standard Deviation
Equation 57
18 Pooled Estimate of the Population Standard Deviation
Calculation for TPH Results for Diesel Spiked Soil 57
19 T-Value Calculation for TPH Results for Diesel Spiked Soil.. 57
20 Pooled Estimate of the Population Standard Deviation
for TPH Results for Waste Oil Spiked SSH Using Soxhlet
Extraction Method 59
21 Test T-Value Calculation for TPH Results for Waste Oil
Spiked SSM Using Soxhlet Extraction 59
vi
-------
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Page
7
12
13
22
27
30
31
32
32
34
36
37
38
39
41
44
45
46
46
47
48
49
56
58
60
62
63
TABLES
Relative Effect of Tested Parameters on Soil Vashing
Efficiency of SSM
Investigated Characteristics of Synthethic Soil Matrix
Pure Product Doses During The SSM Blending Operations
GC Retention Times for BTEX Compounds
Synthetic Soil Matrix Particle Size Distribution
Synthetic Soil Matrix Chemical Characteristics
SSM X-Ray Fluorescense Analysis - Major Constituents
SSM X-Ray Fluorescense Analysis - Trace Elements
SSM X-Ray Diffraction Clay Mineralogy
Soil Vashing Dose/Response Bench Scale Tests
Initial Concentation of Nine SSM Blends
Comparison of Bench Scale to Actual Dose/Response Tests
Homogeneity Test for Blending of Medium Diesel SSM
Parameters Investigated in Soil Vashing Experiments
Removal Efficiencies for SSM Soil Vashing Experiments
Phase I Experiment Mass Balance Results
Results of Phase II Experiments
UST Site Soils Used for Phase II Experiments
Soil Vashing Results for UST Site Soils
Particle Size Distribution of Soils
Total Suspended Solids Levels in Vashvaters and Rinsevaters.
Comparison of Field and Lab BTEX Data
TPH Results for Diesel-Spiked SSM
TPH Results for Vaste-Oil Spiked SSM Using Soxhlet
Extraction Method
TPH Results for Vaste-Oil Spiked SSM
Cost Summary of Soxhlet Method of TPH Analysis
Cost Summary of Sonication Method of TPH Analysis
vii
-------
-------
ABBREVIATIONS AND SYMBOLS
ABBREVIATIONS
AA
Atomic Absorption
ASTM
American Society for Testing and Materials
BDAT
Best Demonstrated and Available Technology
BTEX
Benzene, Toluene, Ethylbenzene, Xylenes
CEC
Cation Exchange Capacity
COD
Chemical Oxygen Demand
EPA
U.S. Environmental Protection Agency
G
Grams
GC
Gas Chromotograph
GP
Supelco Trade Name
IR
Infrared
MEQ
Mi Hi equivalents
MSVS
Mobile Soil Washing System
PAH
Polycylic Aromatic Hydrocarbons
PI
Plasticity Index
PID
Photoionization Detector
PL
Plastic Limit
PPb
Parts per Billion
ppm
Parts per Million
PSD
Particle Size Distribution
USDA
U.S. Department of Agriculture
UST
Underground Storage Tank
RREL
Risk Reduction Engineering Laboratory
SP
Supelco Trade Name
SSM
Synthetic Soil Matrix
TOC
Total Organic Carbon
TPH
Total Petroleum Hydrocarbons
TSS
Total Suspended Solids
XRD
X-Ray Diffraction
XRF
X-Ray Fluorescence
SYMBOLS
cm — Centimeter
ft3 — Cubic Foot
gal — Gallon
Hz — Hertz
kg — Kilogram
kV — Kilowatt
(Continued)
viii
-------
ABBREVIATIONS AND SYMBOLS (Continued)
SYMBOLS (Continued)
lb — Pound
ml — Milliliters
ml/s — Milliliters per Second
mm — Millimeter
S — Second
0 — Degree
X — Percent
ix
-------
ACKNOWLEDGMENTS
This document was prepared for the U.S. Environmental Protection
Agency Office of Research and Development (ORD) by Camp Dresser & McKee
Inc. (CDK Inc.) under the direction of Ms. Mary Tabak. This report is a
work product associated with Work Assignment No. 3-21 under EPA Contract
No. 68-03-3409 with CDM Federal Programs Corporation (CDM FPC).
CDM would like to acknowledge the guidance and assistance provided by
Mr. Anthony N. Tafuri, ORD's Project Officer, and Mr. Richard P. Traver and
Mr. Chi-Yuan Fan, ORD's Technical Project Monitors for this Work
Assignment. The work was initiated under Mr. Richard Traver, and concluded
under the guidance of Mr. Chi-Yuan Fan.
This document is based on research and scientific evaluation developed
by Ms. Mary Tabak, Mr. William Glynn and Ms. Carole Kaslick of CDM Inc.,
and Mr. Michael Borst and Mr. John Mazza of CDM FPC. The primary author of
this report was Ms. Mary Tabak with additional material provided by Mr.
John Mazza, Ms. Carole Kaslick, and Mr. Michael Borst.
Special thanks is extended to Mr. Sy Rosenthal of Foster Wheeler
Enviresponse, Inc. for support services and supplying the synthetic soil
matrix, to Ms. Pat Esposito of Bruck, Hartman & Esposito, Inc. for
technical guidance and review of experimental procedures, and to Ms. Joan
Knapp of CDM FPC for program guidance and support.
This report was prepared in its final format based on a technical edit
by Mr. William Glynn, Dr. Warren Lyman and Ms. Christian Plasse. Ms.
Gloria V£liz prepared the final manuscript and Ms. Christian Plasse
prepared the final graphics.
x
-------
-------
SECTION 1
INTRODUCTION
Based on the Hazardous and Solid Vaste Amendments of 1984 and its Land
Ban Regulations, EPA has discouraged the excavation and landfill disposal
practices of the past for contaminated soils resulting from leaking
underground storage tanks (USTs). EPA has encouraged the use of on-site
treatment technologies, however, problems have plagued the development of
on-site treatment technologies for petroleum contaminated soils. Technical
support is needed to develop effective, long-term corrective actions at
leaking UST sites, design cleanup program guidance, and help implement
state programs.
The U.S. Environmental Protection Agency, through its Risk Reduction
Engineering Laboratory's Release Control Branch, has undertaken research
and development efforts to address the problem of leaking USTs. The
remedial options available for the treatment of contaminated soils from UST
sites are broadly segregated into two main categories, namely those which
remove the contaminants without excavation (in situ techniques) and those
which require excavation of the soil and subsequent cleaning on-site. The
former group of remedial options have not yet been demonstrated for high
efficiency removal of contaminants from the subsurface. These techniques
are plagued by the uncertainty of soil contamination levels in the
subsurface after treatment. Soil excavation followed by extensive cleaning
of the soil will ensure a more complete removal of contaminants over in
situ techniques.
On-site soil washing of excavated soils is a viable alternative to in
situ techniques and has been shown to be effective for the cleanup of
hazardous waste contaminated soils (EPA, 1988). The goal of this effort is
to determine the feasibility of soil washing for cleanup of petroleum
contaminated soils.
SOIL VASHING THEORY
Soil washing is a physical process in which excavated soils are
contacted with a liquid medium, usually water, as shown in Figure 1. The
soils undergo intimate contact with washing and rinsing solutions to
promote contaminant transport from the solid to the liquid phase which
result in a treated soil and a washwater, the latter requiring further
treatment. The washwater treatment involves the removal and residuals
management of the fine particulate material that is carried over into the
washwater.
1
-------
Contaminated Soil
Washing
Treated Soils
' and .
Rinsing
Washwater
Washwater with Fines
j
i
Washwater
Treatment
~T~
Res iduals
Management
of
Fines
Figure 1. Soil Washing Process Diagram.
-------
The two principle cleaning mechanisms include the dissolution of the
contaminants into the extractive agent and/or the dispersion of the
contaminants into the extraction phase in the form of particles (suspended
or colloidal). The separation of the highly contaminated fine soil
particles (silts, clay and colloidal) from the bulk of the soil matrix can
result in volume reduction of the contaminated soil. As a result, a
significant fraction of the contaminated soil is cleaned and can be put
back into the original excavation. Since the contaminants are more
concentrated in the fine soil fractions, their removal from the bulk soil
increases the overall effectiveness of the soil washing process.
Subsequent treatment will be required for the spent vashvaters and the
fine soil fractions. The vashvaters will contain suspended solids which
can be removed by using technologies such as coagulation, precipitation,
sedimentation, and cyclones. A resultant sludge will be produced which is
ultimately treated or disposed with incineration or landfilling. In
addition to removal of the particulates from the washwater, solubilized
contaminants must be removed before discharge. Viable technologies for
this application include air stripping and carbon adsorption.
LITERATURE REVIEW
A literature review was completed to obtain information on prior work
done with soil washing processes. The information obtained included
references on:
full scale soil washing,
pilot scale soil washing tests,
soil washing selection at Superfund sites,
liquid/solid extraction (leaching operations),
tar sands extraction and oil recovery,
oil/water emulsions/solubility,
liquid/solid separation (greater than 100 um by sedimentation &
less than 100 um by hydroclones/flotation /filtration),
oily wasterwater treatment,
sludge residual management,
clay/silt dewaterability and hydrometallurgy.
The references cited in this literature review are included in the
Bibliography.
The treatment of contaminated soils by extractive methods has been
investigated by researchers/commercial vendors, and several field
applications have been undertaken for the removal of organic and inorganic
contaminants. Soil washing technology has been selected for several
Superfund sites (Koppers Texarkana TX, Zellwood FL, United Scrap Lead OH,
South Cavalcade Street TX, and L.A. Clarke & Sons VA). The technology has
been demonstrated in support of the best demonstrated and available
technology (BDAT) determination for the RCRA land ban rules (Esposito,
1988). Several commercial vendors have developed soil vashing technology
for contaminated soils (Chilcote, Trost). Several European companies have
3
-------
also developed soil washing technology for the cleanup of contaminated
soils (Assink, Heimkard, and Sonnen).
The literature indicates that the application of the soil washing
process at the various sites involves several steps. The soil is
pretreated to remove large objects such as tree roots, boulders and debris,
as well as the breaking down of soil lumps. The soil and washing solution
are then vigorously contacted to either dissolve or disperse the
contaminants into the washing solution. The next stage involves the
separation of the soil and the washing solution. The sand particles can be
separated by sedimentation, while the silt and clay particles may require
froth flotation or hydrocylonic action for removal from the washing
solution. The washing solution is then treated by either chemical/physical
or biological methods for recycling into the extraction process. The clean
sand can be placed back into the excavation, while the silt/clay sludge may
require dewatering and further treatment.
Biotrol (Chilcote) has completed an on-site demonstration of their
pilot scale soil washing system at a Minnesota Superfund site contaminated
with pentachlorophenol and creosote. The pilot study demonstrated
pentachlorophenol and creosote removal of 85 to 99 percent with a total
soil recovery of 73 to 83 percent of the contaminated soil. The process
water was successfully treated and recycled back to the soil washing
process. Similar treatment performance has been experienced at several
sites in Europe.
Several documents were reviewed dealing with the subject of oil spill
cleanup technology. "An Overview of Spill Cleanup Technology" (Griffiths)
contains a review of equipment and techniques for responding to spills of
dangerous cargoes from ocean vessels. These techniques are more concerned
with the emergency response cleanups than long term corrective actions of
oil in soils. The "Field Guide to Inland Oil Spill Cleanup Techniques"
(Beynon) outlined several methods for cleaning up oil contaminated soils.
These included soft surface cleaning, hard surface cleaning,
solidification, sludge farming, and incineration. Soft surface cleaning
consists of utilizing farm machinery to aerate the oily soils for increased
biodegradation. Hard surface cleaning required a hot water or steam jet to
wash down pavement or concrete areas. Beach cleaning trials at Pendine
Sands (Warren Spring Lab) described techniques for scraping heavy oil
emulsions into narrow trenches for removal via vacuum tank trucks.
The known reserves of tar sands (also known as oil sands and
bituminous sands) throughout the world has led to the development of
several techniques for the extraction and recovery of the trapped oil in
the sand matrix. The recovery of oil from tar sands is economically viable
through the raining of the tar sands followed by above ground oil extraction
utilizing hot and/or cold water with additives (Camp, Deynoyers, Margeson).
Cold water separation also termed sand reduction process has been
accomplished with water alone and with the addition of kerosene and soda
ash. The hot water process consists of a water/NaOH mixture at temperature
of 180-200°F. Oil recoveries of up to 15% can be accomplished with the hot
water process.
A
-------
The extraction of metals from natural ores has been accomplished since
the early days of civilization. The principles of hydrometallurgy (also
known as solids leaching) may help us develop appropriate techniques for
removal of contaminants from soil (Canterford, Coyle, Schmidt). The
extraction of zinc, copper, and nickel have been accomplished with sulfuric
acid and water solutions. Uranium mine tailings have also been treated by
acid leaching to remove residual contaminants (Haque).
SOIL WASHING RESEARCH PROGRAM
The research program reported in this document consisted of several
sequential steps which built upon the work performed in each previous step.
This section outlines the various phases of work which were performed and
are reported in this document.
SSM Preparation
The program required the preparation of the contaminated synthetic
soil matrices (SSM), which were the soils to be tested. Prior to preparing
the quantities of SSM needed for the bench scale tests, bench scale
experiments were performed to develop a dose/response relationship between
the quantity of petroleum added to the soil matrix and the analysis
quantification. The petroleum products evaluated during this study
included unleaded gasoline, diesel oil and waste crankcase oil. The
dose/response tests investigated the relationship between the concentration
of gasoline or diesel added to the soil and the analytical BTEX (benzene,
toluene, ethylbenzene and xylenes) and TPH (total petroleum hydrocarbon)
results, respectively.
The results of these tests were used to prepare the full scale SSM
batches at predetermined concentration levels. The soils were prepared at
the EPA Soil Blending Facility in Edison, NJ. TPH analysis was performed
to verify the concentration levels for diesel oil and waste oil, and BTEX
analysis was performed to verify the concentration levels for gasoline.
SSM was prepared in 23 kilograms (50 pound) batches and stored in sealed 19
liter (5-gallon) metal pails until used.
The bench scale experiments were designed to simulate the
EPA-developed pilot-scale Mobile Soil Washing System (MSWS). Specifically,
the bench scale experiments were designed to simulate the drum-screen
washer which separates the >2-mm soil fraction (coarse material) from the
<2-mm soil fraction (fines) by use of a rotary drum screen. A high
pressure water knife operates at the head of the system to break up soil
lumps and strip the contaminants from the soil particles.
Phase I Bench Scale Experiments
The bench scale soil washing experiments consisted of several tasks
with various objectives. The first task was to determine the sensitivity
of various washing parameters on the removal effectiveness of contaminated
SSM. Operating parameters such as contact time, washwater mass, rinsewater
mass, washwater temperature and effectiveness of washwater additives
5
-------
(surfactant, and CitriKleen - a biodegradable degreasing agent) were
investigated.
It was expected that gasoline contaminated soils would have different
optimal washing parameters than diesel contaminated soils. Therefore,
identical sets of experiments were conducted on soils containing high
levels of gasoline and diesel contamination. The experiments consisted of
sets of experiments where all parameters but one were kept constant and the
effect of the varied parameter observed. The result of these experiments
was the selection of a set of "optimal" washing parameters for SSH
contaminated with gasoline and diesel.
Phase II Bench Scale Experiments
The next phase involved the use of the "optimal" parameters to
investigate the removal effectiveness of soil washing on soils at varying
levels of contamination. The "optimal" washing parameters were applied to
all matrices of SSH prepared including the waste oil contaminated matrices.
The "optimal" conditions were applied to soils from actual UST sites
from Ohio and New Jersey. Four site soils were used and the results
compared to those obtained with the SSM.
Field Analytical Program
In support of the soil washing experiments, a field analytical program
was developed to analyze soils for BTEX constituents. The description,
results and conclusions of this program are included in this report.
TPH Procedure Investigation
Finally, an investigation of the laboratory procedures used to analyze
TPH in soil was conducted. The method currently used by most laboratories
is the Soxhlet extraction method. However, the Sonication extraction
method has been found to be quicker, easier, and equally effective in
extracting petroleum constituents from soil. The two methods were compared
via multiple analysis of the same soil using both methods and statistical
evaluation of the results are included in this report.
6
-------
SECTION 2
CONCLUSIONS
1. The Phase I experiments indicated that very high removals of BTEX and
TPH (>90%) were obtained vhen treating the SSM with soil washing.
Various operating conditions were tested for their effect on the soil
washing results. Table 1 summarizes the relative effect of the tested
parameters on the SSM soil washing efficiency.
TABLE 1. RELATIVE EFFECT OF TESTED PARAMETERS ON SOIL WASHING
EFFICIENCY OF SSM
Varied Parameter
Range
Effect On
Contaminant
Removal*
Effect On
Particle
Separation
Contact Time
10-30 min
0/+
+
Soaking Time
15-30 min
0
0
Soil/Vashwater Ratio by mass
0.5-2
+
0
Rinsevater/Washvater ratio
3-10
0
0/+
Citrikleen Addition
0-0.7%
-
0
Temperature
13-82°C
0
0/+
0.13% CitriKleen/Temp.
13-82°C
+
0
0.5% Surfactant/Temp.
24-49°C
0/+
+
* (+) = improvement, (0) ¦ no effect, (-) = negative effect
NOTE - Rating system based on an increasing value of the parameter.
The high removals were, for the most part, independent of the washing
conditions evaluated. It was also observed that most parameters,
except for soil/washwater ratio and contact time, did not change the
effectiveness of the separation of the soil particles into the various
sieve fractions.
7
-------
The results also indicated that the addition of CitriKleen actually
decreased the contaminant removal efficiency of the soil washing
process. This may be attributed to enhanced adsorption of contaminants
onto the soil matrix with increased CitriKleen concentration.
2. The "optimal" parameters for washing the SSM, as determined by the
Phase I experiments, were as follows:
contact time 30 minutes
soil/washwater mass ratio 1:1
rinsewater/washwater mass ratio 3:1
washwater additives none
washwater temperature ambient
These parameters were the same for both gasoline and diesel
contaminated SSM. The parameters were selected based on achieving high
removals as cost-effectively as possible.
3. The Phase II SSM experiments indicated that the high removals of
contaminants were independent of the initial concentration of petroleum
contaminants in the SSM. Greater than 90% removals of BTEX and TPH
were obtained for SSM containing gasoline, diesel, or waste oil at all
concentrations.
4. The removals achieved using actual soils (Phase II) were significantly
lower (50 to 90%) than the removals from the SSM (>90%) using the same
washing conditions. The particle size distributions of the actual site
soils show that these soils contained less fine material than the SSM.
The washwater Total Suspended Solids (TSS) levels in SSM experiments
were much higher (>23% or 230,000 mg/1) than the actual site soils (3
to 4% or 30,000 to 40,000 mg/1).
These results indicate that a major mechanism responsible for
contaminant removal during washing of the SSM is particle separation.
The majority of the contaminants seem to be adhered to the finer
particles, and since these particles make up approximately 44% of the
SSM matrix, removal of these particles assists the high removal of
petroleum contamination. This would explain why the washing results
were independent of most washing conditions which serve to enhance
solubilization of the contaminants into the washwater. As long as the
fines were effectively removed from the bulk matrix, high removals were
obtained.
The removals obtained from the site soils were much lower than those
obtained with the SSM. This may be due to the decreased amount of
fines present in these soils. Since particle separation was a major
contaminant removal mechanism during the washing process, high removals
similar to the SSM matrix were not expected or obtained from these
soils.
8
-------
5. Comparison of the Soxhlet and Sonication extraction methods for TPH
analysis of soils indicated that the results are statistically
equivalent using either method. Since there are several advantages to
using the Sonication method, including lower cost and quicker
turn-around time, it may be desirable to substitute the Sonication
method for the Soxhlet method.
6. The field program developed for analysis of benzene, toluene, ethyl-
benzene, and xylenes proved to be a useful tool for screening of soils
and for quantification of contaminant concentration within an order of
magnitude. However, the high detection limits obtained with the exist-
ing analytical equipment limited the potential use of this equipment to
fully quantify soil concentrations in lieu of an off-site lab.
9
-------
-------
SECTION 3
RECOMMENDATIONS
The particle size distributions of the SSM allowed a full
understanding of the effect of the particle removal mechanism during the
washing process. The use of a synthetic matrix also provided a consistent
scientific basis for experimental testing with highly reproducible results.
However, to further characterize the other parameters which affect the
solubility of the contaminants in the washwater, such as temperature and
chemical additives, it is recommended that future soil washing work be
conducted on soils with less fines than the SSM. This could incorporate
the development of another synthetic matrix containing less fines, or
actual sandy soils could be used.
A great asset of the SSM was the homogeneity of the matrix which
produced good reproducibility of the experiments. Actual soils are
inherently less homogeneous that a synthetic matrix. It is recommended
that for future work conducted on site soils, an attempt should be made to
homogenize the soils prior to testing. This may be accomplished by mixing
contaminated soils in the EPA Soils Blending facility prior to testing.
The experimental procedures and analytical protocols used were
effective for the bench scale determination of soil washing effectiveness.
It is recommended that pilot scale experiments be conducted and compared to
the bench scale results to assess hov representative the bench scale tests
are for predicting larger scale results.
10
-------
-------
SECTION 4
MATERIALS AND METHODS
PURE PRODUCT DESCRIPTION
The gasoline, diesel, and waste crankcase oil used for the preparation
of the respective SSM blends were collected from various sources within the
vicinity of Edison, NJ. Gasoline and diesel were collected from five
sources each. The waste crankcase oil was collected from three sources.
Each product was stored in individual plastic gas cans for later use.
The sources of product were as follows:
Regular Unleaded Gasoline - Shell, Citgo (Route 1), Sunoco, Amoco and
McGas
Diesel - Citgo (Route 1), Citgo (Route 27), McGas,
Exxon and Hess
Vaste Crankcase Oil - Rollins Truck Leasing, Lease Line Truck
Rentals and UPS
Equal volumes of product from the various sources were blended to
obtain a total of 5 liters for each of the three petroleum product blends.
The blending occurred immediately prior to the preparation of the
respective SSM blends.
SYNTHETIC SOIL MATRIX CHARACTERIZATION
The basic formula for the SSM was determined by others (Traver, 1989)
from an extensive review of Superfund sites and a review of the composition
of eastern U.S. soils. The samples of SSM were collected from the mixture
of clay, silt, sand, top soil and gravel prepared by others in two 6,800
kilograms (15,000 pound) batches.
A review of the existing soil characteristics was made and additional
tests were conducted to further delineate the physical and chemical
properties of the SSM. The tests, listed in Table 2, included particle
size distribution, moisture retention curve, Atterberg limits, cation
exchange capacity, base saturation, organic matter, chemical constituents,
and mineralogy.
11
-------
TABLE 2. INVESTIGATED CHARACTERISTICS OF SYNTHETIC SOIL MATRIX
0
Particle Size Distribution
0
Moisture Retention Curve
0
Atterberg Limits
0
Cation Exchange Capacity
0
Base Saturation
0
Organic Matter
0
Chemical Constituents
0
Mineralogy
Quantification and assessment of these specific properties will assist
the technical community to understand the differences that may be observed
between the performance of soil washing technology on the SSM and on actual
UST site soils.
The particle size distribution (PSD) of the SSM was determined from
mechanical and hydrometer analysis using the U.S. Department of Agriculture
(USDA) classification system. The sand fraction was determined from sieve
Nos. 10, 18, 35, 60, and 200. Silt and clay fractions were determined by
hydrometer analysis. The Moisture-retention curve of the SSM sample was
developed from determining moisture content at matric potentials of 0, 0.1,
0.3, 1, 3, and 1,500 kilopascals using a pressure plate apparatus. The
moisture content was calculated from triplicate samples of oven dried soil
(105°C) and soil at the various matrix potentials. Atterberg limits
testing was conducted in accordance with the method outlined by the ASTM
standards (ASTM D 4318-84). Phosphorous concentration of the SSM was
determined ^ Bray Pj and Bray P22^ethods. Exchangeable potassium (K ),
calcium (Ca ), and magnesium (Mg ) were determined by extraction with
neutral 1.0 N ammonium acetate and atomic absorption (AA) spectrometry.
The pH was determined from a mass ratio of 1:1 soil to water solution.
Base saturation was calculated from the cation exchange capacity (CEC) and
AA analysis of individual cations. Organic matter content was determined
by wet combustion using chromic acid digestion. The mineralogy of the SSM
was evaluated with x-ray diffraction (XRD) analysis.
SYNTHETIC SOIL MATRIX BLENDING PROCEDURES
The specific blends of SSM required for the bench scale tests were
prepared at the EPA Soil Blending Facility located in Edison, New Jersey.
The soil was blended in a Marion Tilt-Tub batch mixer Model BPS 2436. The
mixer was made of 316 Stainless Steel and had a 280 liter (10 cubic foot)
mixing capacity. The 4 mixing blades were spaced along a rotating shaft
situated along the cylindrical axis of the mixer. This shaft was driven by
a 7.5 kilowatt (10 HP) explosion-proof electric motor. The mixer cavity had
12
-------
a removable stainless steel cover with a fitting to allow for hook-up to a
nitrogen source to provide a nitrogen blanket within the cavity.
The procedures used to prepare the SSH blends were essentially the
same for each of the three petroleum products (gasoline, diesel, and waste
oil). First, 91 kilograms (200 pounds) of dry SSM was mixed with 15 liters
(4 gallons) of water to prepare a base SSH mix containing approximately 17
percent water content by mass. This moisture content was selected because
past experience has shown that this is the optimal moisture content for
physical handling of the SSH. Low, medium and high concentration SSM
blends for each product were prepared by adding the product to the SSH in
pre-measured doses as determined by the dose response tests (discussed in
Section 6).
The low concentration was prepared first by adding the product to the
wetted soil (91 kilograms of soil, dry weight) and allowing to mix for 15
minutes with a nitrogen blanket applied. At the completion of mixing, 23
kilograms (50 pounds) of soil (wet weight) was removed, placed into a 19
liter (5 gallon) pail and sealed. The second dose was added to the
remaining soil (approximately 68 kilograms of soil, dry weight) and mixed
for 15 minutes with a nitrogen blanket applied. Again, 23 kilograms (50
pounds) of soil was removed, placed in 19 liter (5 gallon) pails and
sealed. The final dose was added to the 45 kilograms remaining (100
pounds) of soil and mixing commenced as before. All remaining soil was
placed in pails and sealed. The individual doses of each product are
provided in Table 3.
TABLE 3. PURE PRODUCT D0SBS DURING THE SSH BLENDING OPERATIONS
Blend No.
(EPA SSM No.)
Product
Approximate
Soil Weight
(kg dry wt.)
Product
Added in
Sequence
Quantity of
SSM Blend
Removed (kg)
1 (7)
Gasoline
91
120
ml
23
2 (8)
ft
68
1290
ml
23
3 (9)
fl
45
2150
ml
41
4 (10)
Diesel
91
800
ml
23
5 (11)
M
68
810
ml
23
6 (12)
M
45
1810
ml
41
7 (13)
Waste Oil
91
635
g
23
8 (14)
ii
68
635
g
23
9 (15)
ii
45
1451.5
g
45
13
-------
The soils vere prepared in this manner to minimize preparation of
excess soils (which would later require off-site disposal). It was
determined that about 23 kilograms (50 pounds) of each soil blend would be
required, but the minimum efficient capacity of the blending facility was
45 kilograms (100 pounds) of soil for any one batch. By starting with
45 kilograms (200 pounds) of soil and preparing the batches of soils
sequentially, a minimum of excess soil was prepared while always containing
at least 95 kilograms (100 pounds) of soil in the blender.
Vhile placing the soil blends into pails, samples were collected from
each blend in duplicate for laboratory analysis. Samples were also
collected from the clean wet SSM in duplicate. The analyses of the soils
was conducted by the CDM Laboratory and the parameters analyzed included:
o total petroleum hydrocarbons (TPH)
o polyaromatic hydrocarbons (PAHs)
o benzene, toluene, ethylbenzene and xylenes (BTEX)
o metals (sum of lead, cadmium and zinc)
o total organic carbon (TOC)
Analysis for TPH, PAH and BTEX were conducted using methods 907/418.1,
625/8270 and 624/8240 respectively. Although the analyses for benzene,
toluene, ethylbenzene and the three xylene isomers, ortho-, meta-, and
para-, were conducted separately, the results were summed and presented as
a single parameter, which throughout this report, are referred to as BTEX.
Metals were also analyzed separately and summed as a single parameter.
Cadmium and zinc analyses were performed using method 6010. The lead
analysis was performed using method 7421. TOC analyses was performed using
two different methods depending on the phase of the sample. Liquid samples
were analyzed using method 9060. This method, however, is not directly
applicable to soil samples. The method used for this analysis is adapted
from the determination of organic carbon in soils as presented in the
report "Evaluation of Small Scale Retardation Tests for BTX in Groundwater
PACE Report No. 87-3, prepared for Environment (PACE), the American
Petroleum Institute (API) and University of Waterloo. The method consists
of treatment of the soil with sulfuric acid to eliminate carbonates. The
treated soil is rinsed with reagent water and an aliquot of the prepared
soil sample is placed in a combustion furnace with an oxygen purge. The
carbon present in the soil is converted to carbon dioxide which is detected
by a non-dispersive infrared detector. The signal is compared to that of
TOC standards based on Potassium Hydrogen Phthalate. The instrument used
for this analysis was a Xertex/Dohrmann Total Organic Carbon system.
Four samples were also collected from the medium diesel batch from
within the mixer at different locations to try to determine the extent of
homogenization during the blending process.
The level of worker health and safety protection used was level C for
the gasoline blends and level D for the diesel and waste oil blends. The
blending facility is equipped with an overhead ventilation fan which was
used for the gasoline and diesel blends, but not required for the waste oil
14
-------
blends. An HNu photoionization meter vas used during the blending
activities. Readings were as follows:
gasoline 0-10 ppm in general blending area
100 ppm at open mixer
diesel 0-25 ppm at open mixer
waste oil 0-2 ppm at open mixer
The blending facility was originally designed to prepare large
batches of SSM. Certain problems were encountered in preparing the small
batches required for the bench scale tests. Mixing efficiency decreased
during the blending of the high concentration gasoline and diesel blends.
Soil built up on the sides of the mixer and on the blades with a small
quantity of free product found at the bottom of the mixer. After about
eight minutes, the mixer was stopped and the soil was manually scraped
from the sides and blades. Mixing was then resumed for the remainder of
the 15 minute mixing period. For all of the waste oil blends, mixing
efficiency was decreased due to the high viscosity of the oil. The mixer
was periodically stopped to scrape the sides and paddles to redistribute
the soil for more efficient mixing.
Following the blending of the gasoline and diesel high concentration
blends, the mixer was decontaminated by scraping the remaining soil from
the sides of the mixer followed by a high pressure water spray and towel
wiping. Decontamination following the waste oil high concentration
blends consisted of cleaning the mixer by scraping the sides and washing
with detergent, which was then followed by a high pressure water wash.
15
-------
-------
SECTION 5
EXPERIMENTAL PROCEDURES
BENCH SCALE SOIL WASHING STANDARD OPERATING PROCEDURES
This section describes the Bench Scale Soil Washing Standard
Operating Procedure used for all bench scale experiments. The procedures
were field developed specifically for this project. A general
description of the procedures and equipment used is provided followed by
a step by step protocol.
The soil washing process used for the bench scale studies, as shown
in Figure 2, consisted of initially contacting the contaminated soil with
a washing solution for a set period of time, and then separating out the
soil particles while rinsing with water. Low frequency vibration was
used to assist with particle separation. Soil was collected on three
sieves of decreasing pore size (U.S. Standard Sieve Numbers 10, 60 and
140) with the soil fines passing with the wash solution and rinse water.
Samples were collected of the soil contained on the individual sieves,
the wash water and the rinse water.
Several experiments were conducted to determine the optimum process
conditions for the bench scale studies by varying several operation
parameters. The objectives for selection of the operation parameters
were to minimize the quantities of washwater and rinsewater used while
obtaining maximum contaminant removal. Once the optimum conditions were
selected, they were used for the remaining experiments. To assure data
quality, several quality control measures were performed including
duplicates, field blanks, and experiment replicates.
The following parameters were varied to determine optimum operating
conditions:
- Soil-Washwater contact time
- Washwater additive
- Washwater additive concentration
- Washwater temperature
- Soil/Washwater mass ratio
- Washwater/Rinsewater mass ratio
The mass of soil used for each experiment was 1.4 kilograms. This mass
was chosen to allow for enough soil volume to remain on the individual
sieves for subsequent sampling.
16
-------
Contaminated
Soil
(Mass, TPH, BTEX)
Wash
{Contact Time)
Washwate r
(Volume, Temp)
Rinsewater
(Volume, Temp)
1
#10 Sieve Soils
(Mass, TPH, BTEX)
Rinse
and
Vibration
Partical
Seperation
#60 Sieve Soils
(Mass, TPH, BTEX)
#140 Sieve Soils
(Mass, TPH, BTEX)
Washwater
(Volume, BTEX, TPH,
TSS)
Rinsewater
(Volume, Btex, Tph,
Tss)
Note: Items In Italics Are Parameters Which Were Measured For Each Experiment.
Figure 2. Bench scale soil washing experiment schematic.
-------
The soil washing step took place on a shaker table. The soil and
vashvater were combined in a 9 liters (2-gallon) glass container, sealed,
strapped to the shaker table vith bungi cords, and allowed to shake for
the predetermined amount of time. The stroke and frequency of the shaker
table vas 4 centimeters (1.6 inches) and 4 Hertz respectively.
Once the vashing step vas completed, the soil-vashvater mixture vas
transferred to a vet sieve unit for the particle separation and rinsing
steps. The unit used for these experiments vas the Gilson Vet-Vac Model
W-l. The vet sieve unit is equipped vith three 31 centimeter (12 inch)
diameter sieves stacked vertically and has a cover spray system utilizing
a rotating bar vith six spray nozzles to rinse the material on the
sieves. The unit has an adjustable electromagnetic vibration option,
vhich vas utilized during the experiments. Vith the variable transformer
(rheostat) set at 1002, the frequency and stroke of vibration vas 60
hertz and approximately 0.5 millimeters (0.02 inch) respectively (based
on correspondence vith Gilson technical representative).
The unit is equipped vith a vacuum vater recycle system, hovever
this vas not utilized for the vork performed here since the experiments
vere designed to simulate full scale operation. The vacuum vas
disconnected by simply unplugging the vacuum unit.
To allov for better control of the influent and effluent streams to
and from the sieves, the recycle holding tank vas disconnected from the
system. This vas accomplished by connecting an influent hose directly to
the influent side of the vater pump and also disconnecting the discharge
hose from the bottom of the sieve veil and attaching a separate effluent
hose to allov drainage to be collected in buckets. This alloved for more
accurate measurement of the respective streams. The pumping rate vas
measured to be approximately 110 milliliters per second.
Upon completion of the vashing and rinsing steps, five distinct
fractions remained vhich vere sampled for analysis: #10 sieve soils (>2
mm), #60 sieve soils (0.25-2mm), #140 sieve soils (0.106-0.25 mm),
vashvater, and rinsevater. Samples of #10 and #60 soils and vashvater
vere collected for analysis from each experiment. Occasional samples of
#140 soil and rinsevater vere also collected for analysis.
Samples vere placed in appropriately labeled glass or plastic jars,
vith preservative if required, and submitted to the CDM Laboratory for
analysis. Laboratory analyses conducted include TPH, BTEX, metals (sum
of lead, cadmium and zinc), T0C, PAHs and COD. Holding times and
sampling procedures as outlined in the QAPP, vere strictly adhered to.
Folloving collection of the vashvater and rinsevater for CDM Lab
analyses, additional samples vere collected for Total Suspended Solids
(TSS) and Chemical Oxygen Demand (COD) analyses in the field on the
respective vaters. TSS analysis vas performed using a Millipore Pressure
Filter vith 0.45 micron filters. A pressure filter vas used (instead of
a vacuum filter) to minimize loss of volatiles in the filtrate. The
filtrate vas used for the COD test. HACH Co. COD digestion vials vere
18
-------
used for the COD analysis vith a Spectronic 20 analyzer for absorbance
determination.
Standard Operating Procedure
Belov is a step-by-step standard operating procedure for the bench
scale tests. Figure 3 is a schematic of the procedure, and the numbered
steps in the figure correspond to the steps belov.
1. Decontaminate the following equipment: sieve trays, stainless steel
bovls and other sampling implements, graduated cylinders,
vashvater/rinsevater collection buckets, and pressure filter apparatus
by rinsing vith vater, followed by a methanol rinse, then a final
vater rinse.
2. Bring vashvater to the temperature required for the experiment.
3. Prepare the appropriate quantities of vashvater and additive in a
graduated cylinder.
4. Veigh out 1.4 kg of contaminated soil in a glass container.
5. Combine soil and vashvater in the glass container, cover, place on
shaker table and secure vith bungi cords.
6. Set timer to 30 minutes or other predetermined time period and begin
shaking (vashing).
7. Bring rinsevater to the temperature required for the experiment.
8. Measure appropriate quantity of rinsevater in a bucket or a large
graduated cylinder.
9. Veigh the three decontaminated sieves (No. 10, 60, and 140) vhile they
are still vet to obtain a vet tare weight for each.
10. Situate the three sieves on the Vet Sieve unit vith the cover spray
piece in the up position. Fill influent hose vith vater to prime the
pump (approximately 200 milliliters) and place in rinsevater
container. Ready tvo decontaminated graduated buckets for vashvater
and rinsevater collection. Drain effluent hose.
11. Prepare sample jars and sampling equipment.
12. At end of the shaking period, remove the glass container from the
shaker table. Measure temperature of the vashvater.
13. Svirl container to resuspend solids and pour over sieves (make sure
that effluent hose is secured in vashvater collection bucket to avoid
spillage). Rinse container vith 100 milliliters of vater and pour
over sieves. Manually break up any mud balls remaining from the
shaking step vith a decontaminated implement such as a spoon and
19
-------
ro
o
Wash-
Water
Solutio
Washln
Process
110 §
Soil With
Petroleum
Scale ^
ontamination
Sample
Wash
Sieve
Shaker ^
Rins
Samples
Sample
Rinse
Rinsing
Process
Legend
f33] Liquid Sample
Soil/Sludge Sample
Occasional Sample
r-
Note: Numbers correspond to steps in the standard operating procedure
Figure 3. Soil washing standard operating procedure.
-------
distribute materials over the surface of the sieve. Close the vet
sieve unit by positioning the cover spray over the sieves and locking
in place. Note: Perform these actions as quickly as possible to
avoid unnecessary loss of volatiles.
14. Start vibration at 100 percent. Collect vashvater in the vashvater
bucket until no more water can be recovered. Place effluent hose into
rinsevater collection bucket.
15. Start the rinse step by turning the water pump on while leaving on the
vibration. Continue until water has reached the bottom of the
container and immediately turn the water pump off. Do not run the
pump dry.
16. Leave vibration on for one additional minute. Turn vibration off and
allow to drain for another 4 minutes.
17. Measure vashvater and rinsevater volumes. Collect the respective
vashvater and rinsevater samples. Also collect an additional 250
milliliters of each vater fraction for TSS and COD field analysis.
18. After the sieves have drained for the 4 minute period, remove each
sieve and veigh. After veighing, collect the soil samples from the
respective sieves for analysis of BTEX, TPH, etc.
19. Properly dispose of remaining soil and vater fractions.
20. Perform TSS and COD analyses on the vashvater and rinsevater.
FIELD ANALYTICAL PROGRAM
As part of this vork assignment, a field analytical program was
developed to analyze soils for benzene (B), toluene (T), ethylbenzene (E),
and ortho-, meta-, and para-xylenes (X). This section contains a
description of the field analytical program employed during the soil
vashing project at the EPA test site in Edison, N.J.
The analytical equipment used vas supplied by EPA RREL and consisted
of an HNu 321 gas chromatograph vith a photoionization detector and a
Spectra-Physics SP-4270 integrator.
The field analysis of soil samples vas based on the EPA method 8240.
This method requires methanol extraction of the soil to be analyzed and GC
(gas chromatograph) analysis of the extract. All extracts vere analyzed
through injection into the GC equipped vith a 8' x 1/8" stainless steel
column packed vith GP 5X SP-1200/1.75X Bentone 34 on 100/120 Supelcoport at
an oven temperature of 105°C and a photoionization detector (PID) operating
at 202°C.
Each of the BTEX compounds vere identified through comparison of peak
retention times on the chromatographs to known standards. The retention
times for each compound and the GC conditions are shovn in Table 4.
21
-------
TABLE 4. GC RETENTION TIMES FOR BTEX COMPOUNDS
Retention Time (min)
Compound
Oven Temp = 105°C
Oven Temp = 80°C
Benzene
3.2
5.0
Toluene
5.2
9.8
Ethylbenzene
8.6
18.4
p-Xylene
9.9
20.6
m-Xylene
9.9
22.1
o-Xylene
11.3
25.5
NOTES: Operating conditions on HNu 321 include use of 8'xl/8"
stainless steel column packed with GP 5% SP 1200/1.752
Bentone 34 on 100/120 Supelcoport. PID operating
temperature vas 202°C.
Total run time for all gasoline constituents to elute
through the column vas about 1 hour at 105°C oven
temperature and 2 hours at 80°C.
A stock standard solution (supplied by Supelco) of 200 micrograms per
milliliter of each of the six BTEX compounds vas prepared in high grade
methanol to create a series of standard concentration solutions. These
standards included a 2 ppm, 20 ppm, 50 ppm, 100 ppm, and ISO ppm BTEX
concentration. These various standard solutions vere injected into the HNu
321 in duplicate and the calibration curve vas developed from the resulting
chromatographs. A response factor for each analyte vas obtained from the
ratio of the known standard concentration injected, to the area under the
peak as recorded by the Spectra-Physics SP-4270 integrator. These response
factors vere then used to determine the BTEX concentrations for all
samples.
It should be noted that para- and meta-xylene, at higher
concentrations (i.e., >50 ppm), despite eluting as individual peaks, could
not be resolved individually vith the integrator. In developing the
calibration curve, para- and meta-xylene vere identified as one peak and
one response factor vas calculated. Where the isomeric peaks vere resolved
individually, the concentration vas calculated by multiplying the response
factor by the summation of the tvo reported peak areas.
A review of the initial calibration curve determined that the
chromatograms for the 2 ppm standard did not depict adequate resolution of
the BTEX peaks. Therefore standards of 5 ppm and 10 ppm vere created to
establish a lover concentration range and detection limits for the field
analytical program. Standard injections vere made at the beginning of each
day to serve as a baseline for sample injections. If the calculated
22
-------
concentrations for benzene, toluene, ethylbenzene, m,p-xylene, and o-xylene
were not within +/- 15 percent of the injected concentration then a second
injection was made. If these concentrations were not within the criteria
specified then nev standard solutions were prepared and the above process
was repeated.
The HNu 321 was set at the following conditions:
o Carrier gas (Grade 5 Nitrogen) - flow A - 12 milliliters per minute
at 124 kilopascals
o Input - Detector A - low
Output - Detector A - 1024
o Nitrogen tank regulator - 276 kilopascals
The Spectra-Physics SP-4270 was set at the following conditions:
o Attenuation - 1024
o Chart speed - 0.5 centimeters per minute
All soil samples were refrigerated until ready for analysis. Sample
preparation for field analysis consisted of weighing approximately two
grams of soil, placing the soil into a 15-milliliter vial, and adding 5
milliliters of high grade methanol as the extraction solvent. The vial was
mixed for five minutes using a mechanical shaker to ensure sufficient
contact of the soil sample with the methanol. The vial was subsequently
centrifuged to separate the liquid and solid phases. An aliquot of 1 to 4
microliters of liquid sample, dependent on the concentration expected, was
injected into the GC.
Due to problems with the integrator, the concentrations of BTEX
compounds were hand-calculated from the areas identified by the integrator
with the respective response factors determined from the calibration
curves. The concentration of BTEX compounds calculated represented the
concentration in the liquid extract phase. To determine the concentration
in the soil, the liquid phase concentration was multiplied by the volume of
the methanol used as the extraction solvent divided by the mass of soil
sample, as shown in Figure 4.
Cone, in = Cone, in x Amount of x 1
Soil (mg/kg) Extract Solvent in Mass of Sample (kg)
(mg/1) Extract (1) (usually 0.002 kg)
(usually 0.005 1)
Figure 4. Soil BTEX Concentration Equation.
23
-------
Standard Operating Procedure
Belov is a step-by-step standard operating procedure used for the field
analysis of the BTEX compounds in soil.
1. Change the septum on injection port A of the HNu 321 each day.
2. Lower the temperature of the oven, which is increased to 150°C after
each day of analysis to purge the column of heavier gasoline
constituents, to 105°C and turn the PID lamp on.
3. All syringes prior to and after each sample injection were
decontaminated with methanol. This was performed by extracting 5
microliters of the methanol and ejecting it into a waste container and
then flushing the barrel of the syringe several times with additional
methanol. Typically, methanol blanks were performed on a daily basis
or when deemed appropriate following the analysis of a highly
contaminated sample.
4. Inject 1 microliter of a BTEX standard (standard runs took
approximately 15 minutes to completely elute). If the calculated
concentrations exceeded the -15 percent factor then a second BTEX
standard was injected.
5. A 5 microliter liquid tight syringe was utilized to extract 5
microliters of sample into the syringe and eject into a waste
container. This procedure was then repeated.
6. Depending on the expected concentration of the sample, 1 to 4
microliters of liquid sample was extracted and then injected into port
A. The syringe was pushed into the injection port through the septum
and the sample was injected.
7. The syringe was immediately removed from the injection port and the
"INJA" button was depressed on the integrator keyboard to begin the
run. The syringe was then decontaminated with methanol as discussed
in step 3.
8. Following the completion of a sampling run (approximately 60 minutes
for soil containing gasoline contamination), a hard copy printout of
the chromatogram was obtained from the SP-4270 integrator.
9. All samples injected were logged into a notebook which also contained
the chromatograms. Daily observations were recorded in a field log
book.
10. At the end of each sampling day the oven temperature was increased to
150°C and the detector was turned off.
11. All samples prior to and after analysis were stored in the
refrigerator on site.
24
-------
FIELD TSS & COD PROCEDURES
Pield analyses for TSS and COD were performed on the vashvater and
rinsevater streams collected from the bench scale tests. This section
contains a summary of the procedures used. The TSS analysis was required
to characterize the vashvater and to calculate a soil material balance
during the experiment. The COD vas used to obtain a quick field-determined
characterization of the level of organics transferred into the vashvater
and rinsevater. The TSS and COD data are reported in Appendices C and D.
Total Suspended Solids Procedure
The total suspended solids analyses were conducted using a pressure
filter to minimize loss of volatiles during the filtration process. This
vas an important requirement because the filtrate from the filtration
process vas used for COD analysis, vhich is further described in the
folloving section.
The sample vas first shaken to re-suspend any solids that may have
settled during storage. The sample vas then alloved to sit for 15 minutes
to allow the heavier soil fines to settle. During this period, a piece of
142 millimeter diameter 0.45 micron pore size filter paper (Millipore No.
HAVP 142) vas labeled and veighed. The filter apparatus, Millipore
Hazardous Waste Filtration System Model YT 30 142 HV, vas then assembled
vith the filter in place and a sample vial prepared for the filtrate.
Folloving the 15 minute vait period, 95 to 100 millimeters of sample vas
poured into a graduate cylinder and the volume recorded. This sample
volume vas then poured into the top of the pressure filter, sealed, and
pressurized to a maximum of 483 kilopascals (70 pounds per square inch).
The first 5 to 10 milliliters of filtrate vas discarded, then approximately
40 milliliters vas collected in a sample vial with the remainder of the
filtrate discarded.
The apparatus vas then disassembled and the filter paper vith sludge
removed and placed into an oven to dry. The filter paper and sludge vas
alloved to dry for a minimum of 2 hours. After the material vas dried, the
filter vas removed and alloved to cool, and then veighed. TSS vas
calculated by subtracting the mass of the tared filter paper from the mass
of the dried solids and filter paper then dividing by the initial volume of
sample.
TSS (mg/1) = (mass of filter + sludge) - tared filter mass (mg)
volume of sample (liters)
Figure 5. TSS Concentration Equation.
Chemical Oxygen Demand Procedure
COD digestion vas performed on the filtrates using HACH Co. COD vials
and reactor. Tvo ranges of COD vials vere available, 0-1500 ppm and 0-150
25
-------
ppm. Initially, the 0-1500 ppm range vas used for the vashvater and the
0-150 ppm range vas used for the rinsevater. Since the COD analysis method
used was based on light transmittance, the samples required filtration
prior to analysis. The filtrate from the TSS analyses were used for these
analyses.
The samples vere digested (according to the directions provided by
HACH) by pipetting 2 milliliters of sample into the vial, shaken to mix the
reactants with the sample, and placed in the COD reactor. The HACH COD
Reactor is factory preset to 150°C. The vials vere kept in the reactor for
a minimum of 2 hours to allov the reaction to reach completion and vere
then removed and allowed to cool.
The vial vas then placed in a Baush & Lomb Spectronic 20 spectrometer
to measure the transmittance of light at the wavelength required by the
method. The vavelengths used vere 620 nanometers for the 0-1500 ppm vials
and 420 nm for the 0-150 ppm vials. The percent transmittance vas compared
to a chart supplied by Bausch and Lomb to convert to ppm COD. Vhen the
particular concentration vas out of range, the sample vas diluted and
re-digested or the next higher (or lover) concentration range vial vas used
for re-digestion of the sample to obtain a more accurate reading depending
upon the particular circumstance.
26
-------
-------
SECTION 6
RESULTS AND DISCUSSION
SYNTHETIC SOIL MATRIX (SSM) BLENDING
SSM Characterization Test Results
The SSM vas characterized in accordance vith the test procedures
outlined in Section 4. The tests conducted include:
o particle size distribution
o moisture content
o Atterberg Limits
o cation exchange capacity
o pH
o base saturation
o x-ray fluorescence
o x-ray diffraction
The test results indicate that the SSM is composed of 60 percent sand,
19 percent silt and 21 percent clay as determined by particle size
distribution analysis (Table 5 and Figure 6).
TABLE 5. SYNTHETIC SOIL MATRIX PARTICLE SIZE DISTRIBUTION (USDA)
Percentage
Distribution
Soil Fraction
USDA Method
USCS Method
Gravel
Sand (Total)
60.0
58
Very Coarse (1 to 2 mm)
16.0
Coarse (0.5 to 1 mm)
8.8
Medium (0.25 to 0.5 mm)
11.7
Fine (0.1 to 0.25 mm)
23.5
Silt
19.0
15.2
Clay
21.0
26.8
27
-------
100
7
.001
«r
.01 .1 1
Grain Size (mm)
Figure 6. SSM particle size distribution plot - USDA method.
O1 30
.01 .1 1 10 100
Matric Potential (bars)
Figure 7. SSM moisture retention plot.
28
-------
22
21
4J
c
0)
M
3
iJ
CO
•H
o
z
17
15
10
100
No. of Blows
Figure 8. Liquid limit determination.
Based on this composition the SSM vould be classified (USDA) as having
a sandy clay loam texture. Particle size distribution data may be used to
estimate hydraulic properties (Mishra et al., 1989), residual saturation
(Hoag and Marley, 1986), capillary movement, bulk density, and surface area
of the soil prior to more extensive analyses.
The moisture content of the SSM ranged from 33.1 percent at saturation
(0 pacals) to 8.7 percent at the permanent vilting point (1,500
kilopascals). The moisture content at field capacity (10 pascals) vas 21.0
percent. The moisture-retention curve (Figure 7), vhich is a measurement
of the degree to vhich moisture is held vithin a soil matrix and vas
developed from the moisture content data, vas indicative of a finer
textured soil. The moisture retention curve is a plot of the matric
potential (vhich is the applied pressure force) versus the vater content
remaining in the soil. The moisture content data, can be used to evaluate
moisture and chemical characteristics of the SSM. For example, the amount
of soil vater that can be extracted from the SSM under typical
environmental conditions (0 to 1,500 kilopascals) vill be 24.4 percent.
The remaining soil vater is considered as "unavailable", vhich can be
removed by artificially induced vacuums or pressures. Some similarities
exist betveen the moisture content and residual saturation of petroleum
hydrocarbons in the soil. Generally, stronger competitive absorption of
vater for soil occurs and displaces nonionic organic chemicals that are
present in petroleum in petroleum hydrocarbons (Chiou et al, 1989).
Residual saturation vill be dependent on moisture content and decrease vlth
Increasing moisture.
29
-------
The SSM has a plastic limit (PL) of 15 percent, liquid limit of 19
percent and plasticity index (PI) of 4. A plot of these results on the
Plasticity Chart (Figure 8) indicated that the SSM is a CL-ML material
(inorganic silt or clay) using the Unified Soil Classification System
(USCS). The activity of the clay material, which was calculated from
Atterberg Limit tests and a clay content of 26.8, was 0.6 using the
formula: A = PL/XClay
Atterberg limits testing indicated that the clay fraction of the SSM was
inactive, which means that it will swell very little volumetrically with
addition of water. The estimated swell or heave of the SSM was only 1
percent of the mineral's original volume. This data may indicate the
relative absence of swelling clays such as montmorillinite in the SSM.
Analysis of the SSM for concentrations of various exchangeable ions
indicated that phosphorus was moderate, potassium was low, and both
magnesium and calcium were very high (Table 6).
TABLE 6. SYNTHETIC SOIL MATRIX CHEMICAL CHARACTERISTICS
PARAMETER
UNITS
VALUE
Organic Matter
%
1.3
PH
-
8
Cation Exchange Capacity (CEC)
meq/lOOg
21.7
Base Saturation
X
99
Ca
86.2
Mg
12.4
K
1.3
H
0
Available Phosphorus
PPm
Weak Bray
20
NaHC03
31
Potassium
ppm
112
Magnesium
PPm
324
Calcium
ppm
3740
The pH of the SSM was 8.0. £P"S value is generally the result of
the presence of bases such as Ca + and Mg + ions. Such bases may be read-
il^+removed2with addition of water and/or other ions which may displace the
Ca + and Mg + ions and thus lower the pH. The pH of the SSM will in part
affect the CEC of the pH dependent fraction of the soil (primarily organic
matter), and the adsorption of metals. Mobility of most metals such as
lead (Pb) will be minimal as long as the SSM pH exceeds a value of 6.5.
30
-------
Cations exchange capacity of the SSM vas 21.7 milliequivalents per 100
grams. This CEC value is somewhat typical of soil with a texture finer
than a sandy loam or with elevated organic carbon content. Determination
of CEC is essential in the evaluation of the fate and transport of charged
ionic species, but will have little influence on the non-ionic organic
compounds present in petroleum hydrocarbons.
The base saturation of the SSM vas 99.9 percent, and vas dominated by
the Ca + ion (86.2 percent). Addition of vater or any leaching solution
should considerably reduce the base saturation of the SSM as the Ca + is
replaced by H+ and Al + ions.
The organic matter content of the SSM vas 1.3 percent, vhich is
equivalent to an Organic Carbon (C) content of 0.76 percent (using a
conversion factor of 1.724; Nelson and Sommers, 1982). Organic matter
content determination for soils high in inorganic C (carbonates) requires
their removal prior to analysis. The organic matter fraction of the SSM
vill contribute to the CEC, but is, more importantly, the primary means by
vhich nonionic organic compounds are bound to the soil media. A
partitioning occurs In vhich the nonionic organic compounds are taken up
into soil organic matter from the vater (Chiou et al, 1989).
2+ X-ray fluorescence (XRF) analysis of the SSM indicated that elevated
Ca and Mg concentrations vere present vhen compared to typical U.S.
soils (Lindsay, 1979). Heavy metals such as cobalt (Co), copper (Cu), lead
(Pb), nickel (Ni), and zinc (Zn) vere vithin the range of common soils
(Tables 7 and 8).
TABLE 7. SYNTHETIC SOIL MATRIX X-RAY FLUORESCENCE ANALYSIS -
MAJOR CONSTITUENTS
PARAMETER
SAMPLE #1
SSM SAMPLE #2
COMMON RANGE*
(PPM)
(PPM)
(PPM)
37.9
39.6
49-75
21.5
22.3
1-70
7.85
8.1
2-57
4.03
4.25
0.1-1.0
3.64
3.78
1-79
1.38
1.44
0-3.6
0.44
0.45
0.2-1.7
0.34
0.37
0-1.0
0.31
0.33
0-1.2
0.15
0.15
0-4.0
0.03
0.04
0-0.3
0.02
0.02
0-0.0009
<0.05
<0.05
0-1.0
Silica Dioxide (Si02)
Calcium Oxide (CaO)
Aluminum Oxide (A1_03)
Magnesium Oxide (MgO)
Magnetite (Fe202)
Potassium Monoxide (K^O)
Titanium Dioxide (TiO^)
Sodium Monoxide (Na„0;
Phosphoric Acid
Manganese Oxide (MnO;
Barium Oxide (BaO)
Chlorine (Cl)
Sulfur
* Lindsay (1979)
31
-------
TABLE 8. SYNTHETIC SOIL MATRIX X-RAY FLUORESCENCE ANALYSIS - TRACE ELEMENTS
PARAMETER
SSM SAMPLE #1
(PPM)
SSM SAMPLE #2
(PPM)
COMMON RANGE*
(PPM)
Strontium
(Sr)
369
390
50 -
1000
Zirconium
(Zr)
349
370
60 -
2000
Chromium
(Cr)
131
135
1 -
1000
Vanadium
(V)
57
58
20
- 500
Zinc
(Zn)
46
46
10
- 300
Rubidium
(Rb)
41
46
50
- 500
Ytterbium
(Y)
43
40
25
- 250
Wolfram
(V)
28
32
—
Lead
(Pb)
22
25
2
- 200
Nickel
(Ni)
14
15
5
- 500
Copper
(Cu)
12
14
2
- 100
Tin
(Sn)
<50
<50
2
- 200
Arsenic
(As)
<20
<20
1
- 50
Thallium
(Th)
<10
12
—
Molybdinum
(Mo)
<10
<10
0.
2-5
Uranium
(U)
<10
<10
—
Niobium
(Nb)
<10
<10
Cobalt
(Co)
<10
<10
1
- 40
* Source: Based on Lindsay (1979).
X-ray diffraction (XRD) analysis indicated that the SSM was composed
of 27 percent calcite, 27 percent quartz, 18 percent dolomite, 8 percent
plagioclase feldspar, 8 percent potassium feldspar and a small percent of
clays (Table 9).
TABLE 9. SYNTHETIC SOIL MATRIX X-RAY DIFFRACTION CLAY MINERALOGY
MINERAL X BY WEIGHT*
Quartz 27
Calcite 27
Dolomite 18
Plagioclase feldspar 8
K-feldspar 8
Smectite/vermiculite 5*
Kaolinite <5*
Mica/illite
Chlorite
Polygorskite/attapulgit
"Unidentified" <5
Estimated concentration. Quantification not
possible due to lov concentration of mineral.
32
-------
The calcite and dolomite (45 percent), as determined by XRF, indicate the
presence of a considerable amount of dolmitic limestone in the SSM.
Comparison of the XRD data to the PSD data can be useful in evaluating the
SSM. Sand content as determined by PSD was 60 percent of the SSM by
weight, while quartz content was only 27 percent according to XRD analysis.
The remainder of the sand fraction (33 percent) may include both of the
feldspar minerals (16 percent) and the coarser fraction of the limestone
(17 percent). The silt fraction (19 percent PSD) appears to be primarily
limestone. The clay fraction (21 percent) is made up of some montmorilli-
nitic and kaolinitic clays along with the finer limestone fraction. The
XRD analysis only tentatively identified both montmorillinite and
kaolinite.
Bench Scale Dose/Response Test
The purpose of the dose/response test was to determine the appropriate
amounts of gasoline or diesel fuel which should be added to the SSM to
obtain the desired concentration mixes. Various mixtures of SSM and
gasoline or diesel were prepared and the soil analyzed for benzene, toluene
(T), ethylbenzene (E) and xylenes (X) (the concentrations were summed and
are henceforth referred to as BTEX) and total petroleum hydrocarbons (TPH).
The procedure used was to place approximately 500 grams of SSM into
each of eight 500 milliliter glass wide mouth jars. Water was added to the
jars to make a SSM mixture containing 20 percent water by mass. The water
and soil were mixed by hand with metal laboratory spatulas until the soil
was of even consistency throughout the jar. The gasoline or diesel was
measured out volumetrically with pipets and poured into the jars. The jars
were then mixed by hand for a period of five minutes. The jars were then
capped and submitted to the lab for analysis of BTEX and TPH. The density
of the gasoline and diesel were determined by weighing the contents of a
100 milliliter volumetric flask of both gasoline and diesel.
The result of the dose/response tests are shown in Table 10. The lab
tests indicate that the soils reach a level of liquid saturation at about
23 percent liquid (both water and gas or diesel). The tests were conducted
such that the soils were all prepared to a 20 percent water level.
However, this limited the amount of gas or diesel which could be mixed into
the soil mixture.
At 20 percent water, the highest achievable BTEX concentration was
about 3,000 milligrams per kilogram. For diesel, at a water content of 20
percent, the highest TPH concentration was 60,000 milligrams per kilogram.
The dose/response curves are plotted in Figures 10 and 11. A linear
regression of the data yielded the following relationships for
dose/responses: G«13.33(B)-375 and D-0.675 (T) + 6268 with correlation
coefficients of 0.998 and 0.995 respectively. Where "G" is the gasoline
concentration in milligrams per kilograms, nB" is the BTEX concentration in
milligrams per kilogram, "D" is the diesel fuel concentration in milligrams
per kilogram and nT" is the TPH concentration in milligrams per kilogram.
33
-------
TABLE 10. SOIL UAfflDG DOQK/RKSU® EBCB SCALE HgTS
l^b Detennined l^b Determined
Gasoline
BTEX
TFfl
Jar
Soil
Water
Gasoline
Gasoline
Concentration
Concentration
Concentration
Dose/Response
No.
(g)
(ml)
(ml)
(TO)
(iRg/kg)
X Water
(rig/kg)
(mg/kg)
Ratio*
1
486
100
4
2,964
6,100
21
406
NT
15.0
2
491
100
20
14,820
30,200
20
2,200
NT
13.7
3
635
100
50
37,050
58,300
16
4,420
390
13.2
4
506
100
10
7,410
14,600
20
1180 (1380)
NT
11.4
9
500
50
58
42,987
86,000
10
4,670
NT
18.4
l^b Determined
Lab Determined
Diesel
BTEX
TPH
Jar
Soil
Water
Diesel
Diesel
Concentration
Concentration
Concentration
Dose/Response
No.
(g)
(ml)
(ml)
(TO)
(ng/kg)
X Water
(mg/kg)
(UK/kg)
Ratio
5
511
100
4
6,200
6,200
20
NT
1,950
3.2
6
4%
100
10
16,000
16,000
20
NT
13,300
1.2
7
510
100
20
31,000
31,000
20
NT
33,500
0.9
8
498
100
30
47,700
47,700
20
98.3
59,200 (67,700)
0.8
NCflES:
() - denotes duplicate analysis
NT - Not Tested
* - Dose/response ratio is the ratio of the petroleun additive concentration to the analyte concent rat ion (gasoline
concentration to lab determined ETTEX concentration or diesel concentration to lab determined TFfl concentration).
-------
5000
4000
u 2000
1000
20000 40000 60000 80000 100000
0
Gasoline Added (mg/kg)
Figure 9. Dose/Response curve for gasoline
60000
50000
40000
o 30000
£ 20000
10000
0
10000 20000 30000 40000 50000
Diesel Added (mg/kg)
Figure 10. Dose/Resonse curve for diesel fuel
35
-------
An additional experiment was conducted to determine if a higher level
of BTEX could be achieved if a lover moisture content was used. In this
experiment, a sample of soil was moistened to a 10 percent water content
and then saturated with gasoline. The results, shovn in Table 10 as "Jar
9", indicated that the BTEX content was increased to 4670 milligrams per
kilogram.
Based on the previous gasoline concentration and diesel concentration
calculations, estimates were made to determine the amount of gasoline and
diesel fuel to add to the SSM to obtain the desired concentrations of BTEX
and TPH for the bench scale experiments. The SSH blends were prepared in
the EPA SSH Blending Facility in Edison, NJ in 23 kilogram (50 pound)
batches for use in the bench scale soil washing experiments.
Full Scale SSM Blending Results
Based on the results of the bench scale dose/response tests, the
appropriate "formulas" were determined for preparation of the full scale
SSM blends. The blends were prepared in the EPA Soil Blending Facility in
Edison, NJ.
All nine blends were sampled immediately upon preparation, and
additionally prior to use. In total, all blends were sampled at least
three times and analyzed for BTEX (for gasoline blends) and TPH (for diesel
and waste oil blends). The average of the sampling results are presented
in Table 11. The values presented in this table are the initial soil
concentrations subsequently used to calculate the percent removals from the
soils after the washing experiments.
TABLE 11. INITIAL CONCENTRATION OF NINE SSN BLENDS
CONCENTRATION
SSM BLEND (mg/kg)
Low Gasoline
21
(BTEX)
Medium Gasoline
766
High Gasoline
2067
Low Diesel
917
(TPH)
Medium Diesel
1839
High Diesel
18,425
Low Vaste Oil
1117
(TPH)
Medium Vaste Oil
9327
High Vaste Oil
28,533
NOTE: Above results are averages of several samples
taken for each blend.
36
-------
After the blending results were obtained, a comparison vas made of the
actual blending results to the predicted results from the bench scale
dose/response tests. These results are presented in Table 12.
TABLE 12. COMPARISON OF BENCH SCALE TO ACTUAL DOES/RESPONSE TESTS
Blend
Concentration
Goal*
(mg/kg)
Actual
Concentration
(mg/kg)
Actual
Concentration
As Percentage
of Goal
Lov Gasoline
Hed. Gasoline
High Gasoline
100 mg/kg BTEX
1150
3800
21 mg/kg
766
2067
21
67
54
Lov Diesel
Hed. Diesel
High Diesel
1000 mg/kg TPH
15,000
62,000
917 mg/kg
1,839
18,425
92
12
30
* As predicted by bench-scale dose/response tests.
As the table shows, the actual results were much lover that the
predicted results. The full scale BTEX concentration vas as lov as only 21
percent of that predicted in the bench scale tests and TPH vas as lov as 12
percent of the predicted value. In other vords, for any given amount of
gasoline or diesel added to the soil, the actual BTEX or TPH results
obtained in the soil blend vas lover than that predicted by the
dose/response tests.
These results may be explained by the fact that the mixing vas much
more vigorous in the full scale blending process than in the bench scale
tests. Therefore, more volatiles may have been lost during the full scale
blending process thus lovering the organic content in the SSM blends.
A test of the homogeneity of the SSM blending process vas conducted by
sampling the medium diesel blend six times and analyzing for TPH. The
results, presented in Table 13, shov that the mean TPH result vas 2054
milligrams per kilograms vith a standard deviation of 1024 milligrams per
kilogram.
37
-------
TABLE 13. HOMOGENEITY TEST FOR BLENDING OF MEDIUM DIESEL SSM
DUPLICATE NO.
TPH CONCENTRATION
(mg/kg)
1
2120
2
1160
3
3300
4
570
5
2840
6
2320
Average
2052
Stand. Dev.
1024
Confidence Level* Range
50 percent 1750-2350
90 percent 1200-2900
* Calculation based on Student-t test evaluation of data
The standard deviation of the samples vas about 50 percent of the sample
mean which indicates that the sampling results fell vithin a vide range.
For example, according to these results, for any sample of the mixture
taken, there is a 60 percent confidence level that the results will fall
vithin the range of 1182 - 2922 milligrams per kilograms TPH and a 90
percent confidence level that the results fall vithin 362 - 3942 milligrams
per kilograms.
PHASE I - DETERMINATION OF OPERATIONAL PARAMETERS
The objective of this phase of experimental vork vas to determine the
optimal vashing conditions for the high gasoline and high diesel SSM
blends. The parameters vhich vere varied and investigated include:
contact time, soaking time, soil/vashvater mass ratio, rinsevater/vashvater
mass ratio, addition of surfactant to vashvater, addition of CitriKleen to
vashvater, and temperature of vashvater.
Table 14 summarizes the ranges vithin the above parameters vere
varied.
38
-------
TABLE 14. PARAMETERS INVESTIGATED IN SOIL VASHING EXPERIMENTS
Contact Time
Soaking Time
Soil/Vashvater Mass Ratio
Rinsevater/Vashvater Mass Ratio
Citrikleen Addition
Vashvater Temperature
0.13X CitriKleen/Temp.
0.5% Surfactant/Temp.
10-30 min.
15-30 min.
0.5-2
3-10
0-0.IX
55-180
55-180
75-120
The selection of these experimental conditions were based on results of
previous soil washing work (PEI Associates, Inc., 1988). All experiments
were conducted for both diesel and gasoline SSMs and duplicates of each
experiment were performed with results averaged.
Description of Varied Parameters
The contact time is the amount of time that the soil and vashvater
vere allowed to shake together in the mixing jar as described as Step 6 of
the Standard Operating Procedure. The contact time vas varied vithin the
range of 10 to 30 minutes. This parameter is important because in full
scale operation, the larger the contact time required, the lover the
possible hourly throughput. The goal of varying this parameter vas to
determine the minimum required contact time.
The soaking time is the amount of time that the soil and vashvater
vere in contact together in the jar but not actively shaking. Active
shaking (or mixing) in a full scale system vould be energy intensive.
Soaking time vas investigated to determine the effectiveness of contact
between the soil and vashvater vithout actively shaking to promote mass
transfer from the soil to the vashvater.
The soil/vashvater ratio is the mass ratio of the soil and vashvater.
A higher soil/vashvater ratio vould result in a lover volume of vashvater
to be treated. The goal of varying this parameter vas to determine the
highest ratio vhich vould result in adequate removals. Similarly, the
lover the rinsevater/vashvater ratio, the less rinsevater in generated
requiring further treatment.
CitriKleen, vhich is a biodegradable solvent, vas used as a vashvater
additive to determine if it vould enhance the removal of petroleum
contamination. The additive vas used in the vashvater solution up to 0.7
percent by mass. The goal of varying this parameter vas to see hov much,
if any, benefit vas obtained by using this additive.
39
-------
In general, petroleum compounds are more soluble in hot water than
cold water. The washwater temperature was varied between 55 and 180°F to
determine if enhanced removal was obtained with increased washwater
temperature. The effect of temperature variation was also investigated in
experiments using CitriKleen and surfactant additives.
Results of Experiments
A total of 67 experiments were performed on the gasoline and diesel
spiked SSM, including duplicates and field blanks. Each experiment
involved several analyses and/or measurements of the five resultant
fractions (three soil fractions, washwater, and rinsewater). These
measurements include: temperature, volume, mass, BTEX, TPH, TSS, and COD.
The results of all measurements and analyses of Phase I experiments are
shown in Appendix C.
Since a major mechanism of soil washing is separation of the fine
particles from the bulk soil, the particle size distribution (which is an
indicator of how well the particles were separated) was observed for each
experiment. The SSM characterization test results in Section 6 describe
the dry SSM particle size distribution. The closer the experimental
distribution reached the dry distribution, the better the particle
separation during the experiment.
For each set of experiments, a sensitivity analysis curve was
generated to show the effect of varying the parameter while keeping all
others constant. These curves for all experiments are found in Appendix A.
Each sensitivity analysis curve has a corresponding curve which shows the
effect of the varied parameter on the particle size distribution. These
curves are found in Appendix B.
The effectiveness of soil washing on each experiment was evaluated by
the percent removal of TPH or BTEX (Figure 11).
Percent Removal = (Bulk soil conc.) - (Sieve Soil conc.) x 100
(Bulk soil conc.)
Figure 11. TPH or BTEX Percent Removal Equation.
The percent removal equation in Figure 11 uses the initial bulk soil
concentration, rather than the initial concentrations on the individual
sieves, as a basis for percent removal. This is done because it is not
possible to separate wet contaminated soil for analysis by dry sieve
methods. Using the equation in Figure 11, a distinct removal was
calculated for each of the three soil fractions.
Three of the curves are shown in Figures 12, 13 and 14. Figure 12
shows the percent BTEX removal on the high gasoline SSM when varying the
rinsewater to washwater ratio. Figure 13 shows the percent TPH removal as
40
-------
a function of vashvater temperature on the high diesel SSM where the
vashvater contains 0.13 percent CitriKleen. Figure 14 shows the particle
size distribution as a function of contact time for the high diesel SSM.
The ordinate on the particle size distribution curves is percent passing
each sieve. To calculate the percent retained on a sieve, the value of the
percent passing that sieve should be subtracted from the percent passing
the sieve above it. For example, the percent retained on the No. 60 sieve
is the percent passing the No. 10 sieve minus the percent passing the No.
60 sieve.
The percent removals for several experiments are shown in Table 15.
TABLE 15. REMOVAL EFFICIENCIES FOR SSM SOIL VASHING EXPERIMENTS
Percent Removal
Soil
Contaminant
Initial
Concentration
Vashing
Solution
No. 10
Sieve
No. 60
Sieve
No. 140
Sieve
Gasoline
2,100 mg/kg BTEX
77°F Water
99
99
99
Gasoline
2,100 mg/kg BTEX
77°F 0.52
Surfactant
93
98
94
Diesel
18,000 mg/kg TPH
77°F Water
99
98
98
Diesel
18,000 mg/kg TPH
77°F 0.672
CitriKleen
99
97
—
Diesel
18,000 mg/kg TPH
178°F Water
99
99
99
These experimental results are a sampling of the experiments conducted, and
the results for all experiments can be found in Appendix C. As the table
indicates, the removals obtained were very high, all greater than 90
percent for the No. 10, 60 and 140 sieves. The experiments shown in this
table represent a wide variety of operating conditions, and yet they all
obtained very high removals. These high removals were demonstrated across
the board for gasoline and diesel experiments.
Since the removals were very high, improvements in removal efficiency
based on varying operating conditions, vere difficult to detect.
Regardless of the operating conditions used, the removals for both BTEX and
TPH were usually greater than 90 percent.
Some parameters affected the particle size distribution. A contact
time of 30 minutes was found to provide the best particle separation
because the longer mixing time allowed the soil "clumps" to brake up more
readily. Although mixing times greater than 30 minutes were not tested,
41
-------
100
95
(0
>
i
0)
cc
X
u
H
CO
90
85
80
° No. 10 Sieve
° No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater (mass)- 1:1
Rinsewater : Washwater - 3:1
Additive - None
Contact Time - 30 min.
2 4 6 8 10 12
Rinsewater to Washwater Volume Ratio
Figure 12. Rinsewater to washwater ratio effect on percent BTEX removal.
100
° No. 10 Sieve
° No. 60 sieve
* No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater (mass)- 1:1
Rinsewater : Washwater - 3:1
Additive - 0.13% CitriKleen
Contact Time - 30 min.
20 30 40 50 60 70 80
Temperature C
Figure 13. Temperature effect on percent TPH removal with 0.13% CitriKleen
42
-------
100
w
0)
>
0)
•H
CO
&
c
Type of SSM: High Diesel
•H
to
to
IB
&
Constant Parameters:
60
dc
Soil : Washwater (mass)- 2:1
Rinsewater:Washwater(mass)- 3:1
Addit ive
Wash Temperature 25-29 C
Contact Time - 30 nin.
50
None
40
30
35
20
25
10
15
5
Contact Time (min.)
Figure 14. Contact time effect on partical size distribution
for diesel contaminated soil.
the data in Figure 14 indicates that the predicted particle separation for
mixing times greater than 30 minutes is not expected to significantly
improve. The soil/vashvater mass ratio also affected the particle
separation. Ideally, the largest soil to vashvater mass ratio which
results in good removal should be used because a minimum quantity of
vashvater vould be generated. A 1:1 mass ratio vas the highest
soil/vashvater ratio vhich gave effective particle separation. The
remainder of the studied parameters did not significantly affect the
particle separation.
Selection of "Optimal" Conditions
Based on the Phase I findings, an "optimum" set of operating
conditions vas selected. The Phase I results did not shov a significant
difference betveen the gasoline and diesel experiments. Therefore, the same
"optimal" conditions vere selected for both gasoline and diesel. The
conditions are as follovs: a contact time of 30 minutes; a soil to
vashvater ratio of 1 to 1, a rinsevater to vashvater ratio of 3 to 1;
ambient temperature; and no additives.
These conditions vere selected such that upon scale up to a full scale
system, they vould provide the most effective contaminant removals at
minimum operational cost. This set of operating conditions vas used for
the Phase II experiments.
43
-------
Hass Balance Results
For several experiments, sufficient data vas taken to perform a
material balance of the system during the bench scale soil washing test. A
mass balance of the soil vas done by comparing the dry mass of soil used in
the test to the dry mass of soil recovered in the three soil fractions and
the suspended solids content of the vashvater and rinsevater. A mass
balance of the contaminants vas also conducted in a similar fashion by
calculating the mass of contaminant (BTEX of TPH) recovered in each
fraction.
The mass balance results are shovn in Table 16, and indicate that, on
average, 83 percent of the soil vas accounted for.
TABLE 16. PHASE I - EXPERIMENT MASS BALANCE RESULTS
Experiment
Number
Percent Soil
Accounted For
Percent Contaminant
Accounted For
5
76
66
(BTEX)
6A
94
52
(BTEX)
15
90
26
(BTEX)
19
96
35
(TPH)
22
89
99
(TPH)
23
56
50
(TPH)
33A
80
37
(BTEX)
Average = 83
Stand. Dev. = 14
The percent recovery for TPH and BTEX experiments ranged from 26 to 99
percent. These fluctuations in the recoveries may be due to measurement
errors, inherent analytical variations (as shovn in Appendix E),
volatilization of the contaminants, and soils adhering to the equipment.
The mass balance required the use of the total suspended solids
measurement of the vashvater and rinsevater. This analysis is performed
vith a glass fiber filter. Any colloidal particulates smaller than the
filter paper pore size vere not accounted for in the mass balance.
PHASE II - FULL MATRIX SOILS TESTING
The soils used in the Phase I experiments vere the high gasoline and
high diesel SSM blends, and the results indicated that the "optimal"
44
-------
washing conditions were the same for both gasoline and diesel. The
experiments in Phase II incorporated the "optimal" washing conditions
determined in Phase I on all the other remaining SSM matrices (the low and
medium blends of gasoline and diesel SSM, and all the waste soil SSM
blends). The goal was to determine if the same levels of removals could be
achieved with varying soil contaminant concentration.
In addition, actual soils were obtained from sites in Ohio and New
Jersey where leaking underground storage tanks resulted in petroleum
contaminated soils. These soils were washed using the "optimum" conditions
to compare the SSM washing results with those of real soils.
Full Matrix Results
A total of 15 experiments were conducted for completion of experiments
on all SSM matrices. The same procedures were used for these experiments
as for the Phase I experiments. All measurements and analysis results are
presented in Appendix D.
The removals of contaminants in this set of experiments were similar
to the results obtained in Phase I. The effectiveness of the washing
process was greater
than 90 percent for most
of the experiments as
shown in
Table 17.
TABLE 17. RESULTS OF PHASE
II EXPERIMENTS
PERCENT REMOVAL
NO. 10
NO. 60
NO. 140
SIEVE
SIEVE
SIEVE
SOIL
INITIAL
(<2 mm)
(0.25-2mm)
(0.106-
CONTAMINANT
CONCENTRATION
0.25mm)
Gasoline (low)
21 mg/kg BTEX
99
99
99
Gasoline (med)
766 mg/kg BTEX
99
97
97
Diesel (low)
917 mg/kg TPH
98
98
98
Diesel (med)
1,839 mg/kg TPH
97
97
98
Vaste Oil (low)
1,117 mg/kg TPH
97
98
85
Vaste Oil (med)
9,327 mg/kg TPH
96
98
92
Vaste Oil (high)
28,533 mg/kg TPH
88
98
85
NOTE: Results presented in this table are averages of duplicate
experiments.
These results indicate that the effectiveness of the soil washing process
on the SSM was independent of the concentration or type of contaminant in
the soil blend.
45
-------
Actual Site Soil Results
Soils were collected from four sites with contaminants as shovn in
Table 18.
TABLE 18. UST SITE SOILS USED FOR PHASE II EXPERIMENTS
SITE
SOIL
SOIL
LOCATION
CONTAMINANT
CONCENTRATION
Grove City, OH
Gasoline
243 mg/kg BTEX
Hahwah, NJ
Gasoline
23.8 mg/kg BTEX
Princeton, NJ
Home Heating Fuel
1,375 mg/kg TPH
Holmdel, NJ
Kerosine
215 mg/kg TPH
Optimal soil washing parameters vary for each type of soil and
contaminant condition. Hovever, for the sake of comparison, each site soil
was washed using the "optimal" conditions obtained from Phase I. The
purpose was not to optimize the removal of contaminants from the site
soils, but to compare the removals with those of the SSM.
Each experiment was done in duplicate. The results of these
experiments are shown in Table 19.
TABLE 19. SOIL VASHING RESULTS FOR UST SITE SOIL
SITE
LOCATION
NO. 10
SIEVE
PERCENT REMOVAL
NO. 60 NO. 140
SIEVE SIEVE
Grove City, OH
10
49
49
(gasoline)
67
50
Hahwah, NJ
16
80
14
(gasoline)
77
89
Princeton, NJ
82
72
83
(home heating fuel)
86
-0
Holmdel, NJ
91
54
48
(kerosine)
97
91
70
NOTE: Experiments performed in duplicate. Results of each experiment are
shown in this table.
46
-------
The removals from these soils were significantly lover than those obtained
from the SSM experiments. Although most removals were in the range of 50
to 90 percent, there were some results which were substantially lower.
It should be noted that the experiments using site soils were less
reproducible than the SSM experiments. Table 19 shows the duplicate
results for each experiment which range from being very close together to
varying widely. This is markedly different from the SSM experimental
duplicates which were generally within a few percentage points of each
other.
Another major difference between the SSM and the site soils was the
resultant particle size distribution. Table 20 shows the particle size
distribution for the various soils compared with the SSM.
TABLE 20. PARTICLE SIZE DISTRIBUTION OF SOILS
PERCENT IN EACH FRACTION
NO. 10 NO. 60 NO. 140 VASHVATER &
SIEVE SIEVE SIEVE RINSEWATER
SOIL (>2 mm) (0.25-2mm) (0.106-0.25mm) (<0.25 mm)
SSM
13
31
11
45
Grove City, OH
30
50
4
16
Mahwah, NJ
26
43
21
10
Princeton, NJ
23
12
8
43
Holmdel, NJ
14
67
12
7
The particle size distribution in the table is presented as percent by
weight of soil which remained in the No. 10 sieve, No. 60 sieve, No. 140
sieve, and passed into washwater and rinsewater. As the table shows, 45
percent of the original SSM mass becomes part of the washwater and
rinsewater, indicating that at least 45 percent of the soil mass is smaller
than the No. 140 sieve opening size. On the other hand, the site soils had
much lower percentages of soil mass in the washwater and rinsewater. These
soils had less fine material in the bulk matrix than the SSM.
47
-------
These results are also evident in the comparison of SSM and site soil
vashvater and rinsevater total suspended solids (TSS) concentration
presented in Table 21.
TABLE 21. TOTAL SUSPENDED SOLIDS (TSS) LEVELS IN VASHVATERS AND
RINSEVATERS
TSS (mg/1)
SOIL VASHVATER RINSEVATER
SSM 228,000 6,070
Grove City, OH 33,000 3,100
Mahvah, NJ 40,600 3,060
Princeton, NJ 110,000 6,180
Holmdel, NJ 40,600 6,820
The SSM vashvater contained about 23 percent solids by veight. The other
soils had vashvaters vith TSS levels in the order of 3 to 11 percent
solids.
FIELD ANALYTICAL PROGRAM RESULTS
BTEX analyses on soil samples vere performed in the field using a
portable gas chromatograph in addition to samples being submitted to an
off-site laboratory. The results obtained from the field analyses vere
compared to the results of the laboratory analyses in an effort to
correlate the field data to the laboratory data. The intent of this
correlation vas to minimize the need for sending samples from future vork.
(e.g., pilot and full scale demonstrations) to an off-site laboratory.
The results of this comparative analysis are tabulated in Table 22 and
presented graphically in Figure 15. This includes graphs of the individual
BTEX compounds and a graph of total BTEX (sum of all individual compounds).
The graphs on the left side of the page represent the results of all the
field and laboratory BTEX results. The graphs to the right present the
same data as the graphs on the left vith the axis scales expanded to focus
on the lover concentration data. The field GC results are plotted on the
ordinate and the laboratory results are plotted on the abscissa. The
closer the data point is to the 45° line dravn onto the graph, the closer
the field value is to the lab-obtained value.
The average percent difference betveen the results obtained in the
field and in the laboratory is about 90 percent. This indicates that the
field GC results can predict the soil concentrations vithin an order of
magnitude.
48
-------
TABLE 22. COMPARISON OF FIELD AMD LAB BTEX DATA FOR SOIL SAMPLES.
EXPERIMENT
BENZENE (PPM)
TOLUENE
(PPM)
FIELD
1AB
AVERAGE
% DJllErt
ETELD
LAB
AVERAGE
% DltlER
EXP
3 SIEVE 10
ND
0.03
_
JO
0.35
_
_
EXP
3 SIEVE 60
1.00
0.24
0.62
123%
3.20
3.40
3.30
-6%
EXP
3 SIEVE 140
1.00
1.10
1.05
-10%
5.50
23.00
14.25
-123%
EXP
5 SIEVE 10
1.30
0.17
0.74
154%
3.80
2.30
3.05
49%
EXP
5 SIEVE 60
1.30
<.4
-
-
12.20
3.20
7.70
117%
EXP
5 SIEVE 140
7.20
19.00
13.10
-90%
162.00
330.00
246.00
-68%
EXP
6 SIEVE 60
ND
0.18
-
-
9.70
2.30
6.00
123%
EXP
6 SIEVE 140
ND
0.01
-
-
JO
0.21
-
-
EXP
6A SIEVE 60
6.10
0.48
3.29
171%
19.30
4.20
11.75
129%
EXP
6A SIEVE 140
JO
<.4
-
-
JO
2.30
-
-
EXP
33 SIEVE 60
6.60
3.00
4.80
75%
8.90
33.00
20.95
-115%
EXP
33 SIEVE 140
36.00
6.90
21.45
136%
60.00
110.00
85.00
-59%
EXP
33A SIEVE 10
M)
0.28
_
-
5.50
2.30
3.90
82%
EXP
33A SIEVE 60
M)
0.79
-
-
6.70
11.00
8.85
-49%
EXP
33A SIEVE 140
JO
1.40
-
-
8.60
59.00
33.80
-149%
EXP
9 SIEVE 10
JD
0.20
-
-
37.00
3.00
20.00
170%
EXP
9 SIEVE 60
JD
1.10
-
-
JO
11.00
-
-
FBI
SIEVE 60
ND
<.001
-
-
ND
<.001
-
-
EXP
7 SIEVE 10
ND
0.62
-
-
7.51
7.80
7.66
-4%
EXP
7 sieve; 60
JO
0.60
-
-
2.00
6.70
4.35
-108%
EXP
7 SIEVE 140
JO
0.47
-
-
2.30
20.00
11.15
-159%
EXP
8 SIEVE 60
JO
0.89
-
-
9.75
9.80
9.78
-1%
EXP
8 SIEVE 140
4.40
-
-
15.07
76.00
45.54
-134%
EXP
10 SIEVE 10
JO
0.46
-
-
41.60
24.00
32.80
54%
EXP
10 SIEVE 60
JO
0.82
-
-
3.90
11.00
7.45
-95%
EXP
10 SIEVE 140
2.50
<.20
-
-
4.50
15.00
9.75
-108%
EXP
10A SIEVE 60
JD
0.88
-
—
9.51
11.00
10.26
-15%
EXP
10A SIEVE 140
6.44
1.40
3.92
129%
15.44
28.00
21.72
-58%
EXP
11 SIEVE 60
JO
0.77
-
2.50
8.70
5.60
-111%
EXP
11 SIEVE 140
6.80
1.50
4.15
128%
18.20
28.00
23.10
-42%
EXP
11A SIEVE 60
JO
0.16
-
-
6.30
1.70
4.00
115%
EXP
12 SIEVE 10
JO
0.60
—
—
2.17
7.50
4.84
-110%
EXP
12 SIEVE 60
JO
0.44
-
-
22.96
5.00
13.98
128%
EXP
12 SIEVE 140
ND
0.01
-
-
1.43
0.04
0.73
190%
EXP
13 SIEVE 10
JO
1.60
—
—
3.30
22.00
12.65
-148%
EXP
13 SIEVE 60
3.00
0.66
1.83
128%
10.80
10.00
10.40
8%
EXP
13 SIEVE 140
JO
0.23
-
-
1.90
10.00
5.95
-136%
(continued)
49
-------
X»wT.g 22. (continued)
ETHYIfiENZENE (PPM) TOTAL XYLENES (PPM)
EXPERIMENT FIELD IAB AVERK2 % DIFFER FIELD LAB AVERAGE % DIFFER
EXP
EXP
EXP
3 SIEVE 10
3 SIEVE 60
3 SIEVE 140
ND
1.20
6.30
0.14
1.30
15.00
1.25
10.65
-e%
-82%
ND
6.60
40.00
0.85
7.10
87.00
6.85
63.50
-7%
-74%
EXP
EXP
EXP
5 SIEVE 10
5 SIEVE 60
5 SIEVE 140
1.00
6.00
90.00
1.00
2.50
160.00
1.00
4.25
125.00
0%
82%
-56%
5.50
32.00
479.00
5.20
16.10
860.00
5.35
24.05
669.50
6%
66%
-57%
EXP
EXP
6 SIEVE 60
6 SIEVE 140
2.10
N3
0.67
0.17
1.39
103%
11.10
ND
3.50
0.97
7.30
104%
EXP
EXP
6A SIEVE 60
6A SIEVE 140
2.60
N3
1.20
2.50
1.90
74%
13.70
2.10
6.10
17.60
9.90
9.85
77%
-157%
EXP
EXP
33 SIEVE 60
33 SIEVE 140
3.70
26.00
15.00
46.00
9.35
36.00
-121%
-56%
20.90
143.00
91.00
242.00
55.95
192.50
-125%
-51%
EXP
EXP
EXP
33A SIEVE 10
33A SIEVE 60
33A SIEVE 140
0.90
3.00
8.80
0.61
4.10
35.00
0.76
3.55
21.90
38%
-31%
-120%
3.70
17.20
56.00
3.14
24.20
176.00
3.42
20.70
116.00
16%
-34%
-103%
EXP
EXP
9 SIEVE 10
9 SIEVE 60
0.40
3.30
1.60
5.70
1.00
4.50
-120%
-53%
2.50
10.20
9.10
34.00
5.80
22.10
-114%
-108%
FBI
SIEVE 60
ND
<.001
-
-
ND
<.001
-
-
EXP
EXP
EXP
7 SIEVE 10
7 SIEVE 60
7 SIEVE 140
2.04
1.34
4.40
3.40
2.80
14.00
2.72
2.07
9.20
-50%
-71%
-104%
6.71
8.63
23.60
19.70
16.90
82.00
13.21
12.77
52.80
-98%
-65%
-111%
EXP
EXP
8 SIEVE 60
8 SIEVE 140
2.27
14.51
4.10
40.00
3.19
27.26
-57%
-94%
11.94
72.82
25.50
199.00
18.72
135.91
-72%
-93%
EXP
EXP
EXP
10 SIEVE 10
10 SIEVE 60
10 SIEVE 140
24.00
1.90
9.80
18.00
4.50
18.00
21.00
3.20
13.90
29%
-81%
-59%
116.70
9.60
50.00
104.00
24.60
100.00
110.35
17.10
75.00
12%
-88%
-67%
EXP
EXP
10A SIEVE 60
10A SIEVE 140
1.92
4.42
5.10
18.00
3.51
11.21
-91%
-121%
12.03
28.52
30.20
99.00
21.12
63.76
-86%
-111%
EXP
EXP
11 SIEVE 60
11 SIEVE 140
1.80
13.20
3.50
18.00
2.65
15.60
-64%
-31%
7.70
65.00
18.20
86.00
12.95
75.50
-81%
-28%
EXP
1LA SIEVE 60
6.80
0.69
3.75
163%
16.90
2.02
9.46
157%
EXP
EXP
EXP
12 SIEVE 10
12 SIEVE 60
12 SIEVE 140
0.27
51.66
0.45
3.00
2.40
0.06
1.64
27.03
0.25
-167%
182%
154%
3.42
36.92
3.26
15.70
14.20
0.49
9.56
25.56
1.88
-128%
89%
148%
EXP
EXP
EXP
13 SIEVE 10
13 SIEVE 60
13 SIEVE 140
1.00
5.80
6.60
9.60
4.40
8.50
5.30
5.10
7.55
-162%
27%
-25%
5.60
30.90
29.70
51.00
24.30
51.00
28.30
27.60
40.35
-160%
24%
-53%
(continued)
50
-------
TABLE 22.(continued)
TOTAL BTEX (PPM)
EXPERIMENT
FIELD
IAB
AVERAGE
EXP
3 SIEVE 10
ND
1.37
_
_
EXP
3 SIEVE 60
12.00
12.04
12.02
0%
EXP
3 SIEVE 140
52.80
126.10
89.45
-82%
EXP
5 SIEVE 10
11.60
8.67
10.14
29%
EXP
5 SIEVE 60
51.50
22.20
36.85
80%
EXP
5 SIEVE 140
738.20
1369.00
1053.60
-€0%
EXP
6 SIEVE 60
22.90
6.65
14.78
110%
EXP
6 SIEVE 140
ND
1.36
-
-
EXP
6A SIEVE 60
41.70
11.98
26.84
111%
EXP
6A SIEVE 140
2.10
22.80
12.45
-166%
EX?
33 SIEVE 60
40.10
142.00
91.05
-112%
EXP
33 SIEVE 140
265.00
404.90
334.95
-42%
EXP
33A SIEVE 10
10.10
6.33
8.22
46%
EXP
33A SIEVE 60
26.90
40.09
33.50
-39%
EXP
33A SIEVE 140
73.40
271.40
172.40
-115%
EX?
9 SIEVE 10
39.90
13.90
26.90
97%
EXP
9 SIEVE 60
13.50
51.80
32.65
-117%
FBI
SIEVE 60
N3
<.001
-
-
EXP
7 SIEVE 10
16.26
31.52
23.89
-64%
EXP
7 SIEVE 60
11.97
27.00
19.49
-77%
EXP
7 SIEVE 140
30.30
116.47
73.39
-117%
EXP
8 SIEVE 6C
23.96
40.29
32.13
-51%
EXP
8 SIEVE 140
102.40
319.40
210.90
-103%
EXP
10 SIEVE 10
182.30
146.46
164.38
22%
EXP
10 SIEVE 60
15.40
40.92
28.16
-91%
EXP
10 SIEVE 140
66.90
133.20
100.05
-66%
EXP
10A SIEVE 60
23.46
47.18
35.32
-67%
EXP
10A SIEVE 140
54.82
146.40
100.61
-91%
EXP
11 SIEVE 60
11.90
31.17
21.54
-89%
EXP
11 SIEVE 140
103.20
133.50
118.35
-26%
EXP
11A SIEVE 60
30.00
4.57
17.29
147%
EXP
12 SIEVE 10
5.86
26.80
16.33
-128%
EXP
12 SIEVE 60
111.54
22.04
66.79
134%
EXP
12 SIEVE 140
5.14
0.60
2.87
158%
EXP
13 SIEVE 10
9.80
84.20
47.00
-158%
EXP
13 SIEVE 60
50.50
39.36
44.93
25%
EXP
13 SIEVE 140
38.20
69.73
53.97
-58%
51
-------
Benzene
Benzene
LAB (mg/kg)
6 7
LAB (mg/kg)
Toluene
Toluene
400
300 -
o>
3>
E
200
100 -
200
o>
JC
O)
e
100 -
100 200 300 400
LAB (mg/kg)
Ethylbenzene
1 oo
LAB (mg/kg)
200
10 20 30 40
LAB (mg/kg)
Ethylbenzene
0 10 20
LAB (mg/kg)
(continued)
Figure 15. Comparison of field and lab analytical data.
52
-------
Total Xylenes
1000
200 -
200 400 600 800
LAB (mg/kg)
'.000
rotal Xylenes
O!
e
a 20
10 20 30 «0
LAB Smg/kgj
Total BTEX Total BTEX
40 -
\H 20 -
u.
10 -
50
30
40
0
1 0
20
1500
o
LU
u.
SCO -
1000
soo
0
LAB (mg/kg) LAB (mg/kg)
Figure 15. (continued)
As shown on the graphs, the concentrations determined in the field
were more often lover than the concentrations reported by the laboratory.
This may be due to the methodology used in the extraction procedure. The
ability to extract the BTEX compounds from the soil to the liquid phase may
not have been as controlled in the field as it is in the laboratory. For
the field data, the detection limit is approximately 5 parts per million
whereas the laboratory reported a detection limit of 1 part per billion.
However, for comparative purposes a reported non-detect for the field data
was assigned a value of 0.05 parts per million. This is particularly
evident in the benzene graphs.
53
-------
Since benzene is the most stringently regulated of the BTEX compounds,
it was important to be able to quantify benzene. Predominantly, however,
the concentration of benzene was not quantified because the benzene peak
eluted out at a similar time as the peak of the extraction solvent.
Different techniques were employed with the instrumentation available
including lowering the oven temperature of the GC from 1058C to 80°C. An
improvement in separation and quantitation of the benzene peak was
observed, however this increased the sample run time to two hours. The
problem encountered was that techniques to separate the early eluting BTEX
peaks caused the run times to increase. The late eluting gasoline
constituents could take up to several hours to pass through the column.
Various time functions were programmed into the integrator to either
ignore or integrate the methanol peak in an attempt to isolate the benzene
peak. This did not resolve the difficulties in quatifying benzene.
Recommendations were made to modify the field program for the next phase of
work including the purchase of a different column and accessing temperature
programming capabilities with the microcontroller unit of the HNu 321.
In summary, the results of the field program indicate that the data
may only be suitable for screening of soils to determine relative
concentrations of BTEX compounds and should only be utilized to supplement
the off-site laboratory results.
TOTAL PETROLEUM HYDROCARBON ANALYTICAL STUDY
Total Petroleum Hydrocarbon (TPH) is a parameter often used to
determine the extent of contamination present in soils contaminated with
petroleum products such as diesel oil and waste oils. The analytical
method to do this analysis involves the extraction of the soil with a
solvent and the subsequent analysis of the extract using infrared
spectroscopy.
There are two methods available to perform the extraction of the
soils: Soxhlet Extraction (EPA Test Method 3540, U.S. EPA Test Methods for
Evaluation Solid Vaste, Third Edition, SV-846, November 1986) and
Sonication Extraction (EPA Test Method 3550). The method currently used by
most laboratories is the Soxhlet method, although the Sonication method is
considered to be a quicker and simpler method.
As part of the soil washing work assignment, a study was conducted to
determine if the extraction efficiencies of the two methods were
statistically equivalent.
Procedure
The soil washing evaluation project required the preparation of
various blends of SSM with gasoline, diesel fuel, or waste crankcase oil.
A 2000 gram sample of the medium level diesel spiked SSM and waste oil
spiked SSMs were submitted to the Camp Dresser & McK.ee laboratory for
analysis of the two soil samples using both extraction methods. Based on
discussion with Lisa Moore, Statistician, EPA RREL, a sample size of 14
54
-------
analyses per method was considered adequate to determine, with a 90 percent
confidence level, if the analytical results for the two methods were within
10 percent of each other.
The sample preparation was conducted by placing the entire sample
submitted to the lab in a stainless steel mixing bovl and thoroughly
homogenizing. Twenty-gram sample aliquots were used for Soxhlet extraction
and ten-gram aliquots were used for Sonication extraction.
Sonication Methodology —
Approximately 10 to 15 grams of sample were accurately weighed into a
250 milliliter beaker. The sample was homogenized prior to aliquots being
taken to ensure uniform subsamples. The aliquot was acidified with
concentrated sulfuric acid (0.5 to 1 milliliters) and mixed. Approximately
30 grams of dried magnesium sulfate monohydrate was added to the sample and
mixed to absorb water. The prepared sample was then extracted with 80
milliliters of freon. The extraction was performed using the Tekmar Sonic
Disruptor (with a 3/4" horn, 3 minute cycle at 50 percent pulsed duty
cycle). The solvent was then decanted off and filtered through phase
separator paper (Whatman). The sample was then extracted two more times
with 80 milliliter freon per extraction with the extracts filtered and
combined. The final extract was brought to 250 milliliter total volume
(freon) in a volumetric flask. The samples were analyzed by EPA method
418.1 (i.e. spectrophotometry). The final results were corrected to dry
weight based on a total solids determination (overnight drying at
103-105°C).
Soxhlet Methodology —
Approximately 20 to 30 grams of sample were accurately weighed into a
250 milliliter beaker. The sample was acidified to a pH of 2 with
concentrated sulfuric acid (0.5 to 1 milliliters). Magnesium sulfate
monohydrate was added to dry the sample. The sample was then transferred
to a cellulose extraction thimble. If necessary, a piece of Whatman 4
filter paper wetted with freon was used to obtain a quantitative transfer
of the material; if used, the filter paper was added to the thimble. The
thimble was then covered with a plug of glass wool to prevent soil from
spilling out of thimble. The thimble was placed in the Soxhlet extraction
apparatus and extracted with 100 milliliter of freon for four hours with
the apparatus operating at a minimum of 4 cycles per hour. After the four
hours of extraction, the apparatus was allowed to cool and the freon
transferred into a 100 milliliter volumetric flask and brought up to volume
with fresh freon. The samples were analyzed by EPA method 418.1 (i.r.
spectrophotometry). The final results were corrected to dry weight based
on a total solids determination (overnight drying at 103 to 105°C).
Experimental Results
The data collected by the Soxhlet and Sonication extraction methods
for medium-level diesel spiked SSM are shown in Table 23.
55
-------
TABLE 23. TPH RESULTS FOR DIESEL-SPIKED SSM
Replicate
TPH Concentration
(mg/kg dry basis)
Number
Sonification
Soxhlet
1
4,980
6,300
2
7,760
5,680
3
4,580
7,500
4
4,190
9,000
5
4,760
8,150
6
6.680
7,540
7
8,100
5,210
8
10,300
7,170
9
5,220
3,570
10
4,750
5,970
11
5,360
4,150
12
4,300
5,690
13
8,530
4,790
14
5,630
5,660
Average Concentration
6,081
6,170
Standard Deviation
1,817
1,489
n
14
14
Maximum
10,300
9,000
Minimum
4,190
3,570
The data represent fourteen separate extractions. Blank samples consisting
of clean vet SSM vere also submitted as a QA check. The TPH concentration
of the blank was less than 10 milligrams per kilogram using both extraction
methods indicating that cross contamination is not a concern in the data
for either extraction process.
The first question of concern is whether the two population averages
are equal. If the averages are not equal, the second question is whether
or not there is a difference of 10 percent or more between the results.
Each population is assumed to be normally distributed and independent. The
null hypothesis is that the two groups have equal averages, and the
alternate hypothesis is that the averages are unequal. The comparison of
means has 26 degrees of freedom. The critical student-t test value, t ,
is 1.706 at the 90 percent level of confidence. The test t-value is
computed using the equations in Figures 16 and 17.
56
-------
S n. n,
sp 1 2
NOTE: n, and n, are the number of data values in each set, and X, and X2
are the respective averages.
Figure 16. Test T-Value Equation.
Where S is the pooled estimate of the population standard deviation
computeapas
sp
(nj-l)Sf+(n2-l)S
n, + n2 - 2
NOTE: S, and S2 are the standard deviations and n, and n2 are the number
of data values in each set.
Figure 17. Pooled Estimate of the Population Standard Deviation Equation.
Table 23 lists the values of the standard deviations and averages along
vith the individual data points. Figures 18 and 19 show the evaluations of
the equations presented in Figures 16 and 17 for diesel spiked soil.
6081 - 6170
t =
1661
= - 0.14
1_ + ]_
14 14
Figure 18. Pooled estimate of the Population Standard Deviation
Calculation for TPH Results for Diesel Spiked Soil.
sp
(13)(1817)' + (13)C1A89)2
26
1661 mg/kg
Figure 19. T-Value Calculation for TPH Results for Diesel Spiked Soil.
57
-------
Since the test value of the t-variable (Figure 16) is within +t , the
null hypothesis can not be rejected. That is, the differences computed are
not significant at the 90 percent level of confidence. This follows the
intuitive feel since the averages were only 89 milligrams per kilogram
different on averages on the order of 6,000 milligrams per kilogram. Since
the averages were found to be statistically equal, the difference between
them is less than 10 percent.
A similar procedure was followed using the SSH spiked with waste oil.
The laboratory Soxhlet apparatus can only do 11 extractions in one batch,
so to obtain the 14 extractions required, two batches were extracted. The
TPS extraction data for the waste-oil matrix using the Soxhlet method are
shown in Table 24.
TABLE 24. TPH RESULTS FOR VASTE-OIL SPIKED SSM USING SOXHLET
EXTRACTION METHOD
TPH Concentration (mg/kg dry basis)
9/19/89 9/21/89 9/25/89
Replicate Number (Extraction Date)
1
3,250
6,770
2,060
2
3,000
14,000
2,580
3
6,850
13,500
3,080
4
12,100
5
8,680
6
5,590
7
6,110
8
6,730
9
9,700
10
6,070
11
9,370
Average Concentration 4,367 8,965 2,573 mg/kg
Standard Deviation 1,750 2,927 416 mg/kg
n 3 14 11 mg/kg
Maximum 6,850 14,000 3,080 mg/kg
Minimum 3,000 5,590 2,060 mg/kg
NOTE: The COM laboratory recommends that the data obtained from the
extraction on September 19 and 25th be excluded from the reported
results.
Three samples were extracted on 9/19 and eleven were extracted on 9/21.
Upon examination of the data of the samples extracted on 9/19, the mean of
58
-------
these samples were significantly lover than the mean of those extracted on
9/21. Three more extractions were done on 9/25, and this data also
appeared to fall out of the range of those analyzed on 9/21. The
determination from the laboratory was that the data obtained on 9/19 and
9/25 should not be used in the statistical evaluation.
The first question is whether the laboratory analysts correctly
recommended eliminating the results from a statistical perspective. The
only physical difference noted in the extraction process was the date of
extraction. The two sets of three extractions can be considered as a
comparison to the original 11 in the same manner as the TPH comparisons
made for the medium diesel oil tests (Figures 20 and 21).
sp
(10)(8965)2<2)(4367)» _ 2?6? ng/kg
Figure 20. Pooled Estimate of the Population Standard Deviation.
8965 - 4367
2767
3.0
11
Figure 21. Test T-Value Calculation For TPH Results for Vaste Oil Spiked
SSM Using Soxhlet Extraction Method.
Vith 12 degrees of freedom, the alternate hypothesis can not be
rejected. There is evidence to support the conclusion that they are not
equivalent. The critical t value is 1.782 for this application.
The second three extractions (9/25/89 extraction date) were conducted
as substitutes for the three originally excluded. The results yield a test
value of 6.0, again lending support for excluding them from the analysis.
Statistically, these extraction results are outliers resulting from unknown
factors on the different days. Statistically, the lab was correct in
excluding these from the presentation of the data.
Table 25 lists the reported results from the laboratory for the
accepted results for the TPH determinations by Soxhlet extraction and
Sonication.
59
-------
TABLE 25. TPH RESULTS FOR VASTE OIL SPIRED SSM
Replicate
TPH Concentration
(mg/kg dry basis)
Number
Sonication
Soxhlet
1
4,640
6,770
2
10,900
14,000
3
9,190
13,500
4
12,600
12,100
5
10,100
8,680
6
7,960
5,590
7
9,930
6,110
8
11,000
6,730
9
7,550
9,700
10
8,010
6,070
11
6,510
9,370
12
7,330
13
8,110
14
9,500
Average Concentration
8,813
8,965
Standard Deviation
1,990
2,927
n
14
11
Haximum
12,600
14,000
Minimum
4,640
5,590
Applying the same procedures to compare the Soxhlet and Sonication
extractions, the null hypothesis is stated that the averages are equal for
the methods. The application of the same procedures listed above yield a
test value to the t variable of -0.16. This is within the range that does
not allow rejection of the null hypothesis. There is no statistical
evidence to believe that there is a difference in the reported results
based on the extraction method.
An argument can be presented that it is impossible, by definition, to
have outliers in this procedure. All data, no matter the numerical values
must be considered. Vhen this arbitrary rule is applied, the test t value
computed is 1.64. This, too, is within the range that does not allow the
rejection of the null hypothesis at the 90 percent confidence level.
In summary,
(1) The data presented for the determination of TPH in the diesel spiked
SSM by the two methods is statistically equivalent at the 90 percent
confidence level.
(2) There is statistical support for the elimination of data outliers at
the 90 percent confidence level.
60
-------
(3) The data found to be acceptable by the laboratory for the analysis of
vaste-oil and diesel spiked SSM by Soxhlet extraction is statistically
equivalent to the Sonication extraction data at the 90 percent
confidence level.
.(A) The waste oil TPH determination by Soxhlet extraction is equivalent to
the TPH determined by Sonication extraction even when all data is
forced to be included in the data sample population at the 90 percent
confidence level.
Comparison of Extraction Methods
Since the analytical results of both methods were found to be
statistically equivalent, it is of interest to compare the two methods to
determine if there is any advantage in using one over the other. In order
to make a comparison, the CDM laboratory personnel were asked to prepare a
summary of the material and labor time required to analyze a batch of
samples using each of the methods. A batch size of 10 samples was chosen
because the Soxhlet apparatus can only process 12 samples at a time, and
each batch must also be accompanied by either one or two QA samples, such
as duplicates or blanks. The Sonication apparatus processes 1 sample at a
time, and this assumption was used in developing the cost of sample
preparation.
Each sample analysis for total petroleum hydrocarbon consists of the
following steps: sample preparation, extraction, analysis by infrared
spectrophotometry, and cleanup of extraction and analytical apparatus.
A summary of the costs and labor required for each extraction method
is presented in Tables 26 and 27. The sample preparation and analysis by
i.r. is similar for both methods. However, a significant difference can be
seen in the amount of time required for the extraction process.
The Soxhlet method requires 1.6 hours per sample of interactive labor
for all activities such as cleanup and reporting of data. This amount of
time per sample is a total amount. Therefore, if two chemists were working
to perform the analysis, one sample could be completely prepared with 0.8
hours of labor for each chemist. However, each batch of samples require an
extraction time of four hours in addition to the 1.6 hours per sample. A
batch of 10 samples would take 16 labor hours plus 4 hours of extraction
time.
The Sonication method requires less time than the Soxhlet method. The
total amount of time required per sample, including time extracting in the
Sonication apparatus, is 1.4 hours per sample.
The capital equipment costs and expendable item cost is comparable for
the two methods, as shown in Tables 26 and 27. The expendable item cost
for the Soxhlet method is 17.2 dollars per sample and the cost for the
Sonication method is 15.4 dollars per sample.
61
-------
TABLE 26. COST SUMMARY OP SOXHLET METHOD OP TPH ANALYSIS
Time Requirement
Activity In Total Labor Hours Expendable Items Capital Equipment Cost
Sample Preparation
Extraction
Analysis by I.R.
Spec t ropho t ome t ry
(To Report Stage)
Cleanup of Extraction
and Analytical Apparatus
0.2 hr/sample
0.3 hrs for sample set
up and breakdown time
0.9 hr/sample
0.2 hr/sample
Acid S0.2/Sample
Magnesium Sulfate $1.8/sample
Acids and Freon $4.00/sample
Beakers, 250 ml
Heavy Duty $3.25/each
Extract Glassware
$160.00/each
IR Spectrophotometer
IR Cells $1100-$1500/pair
Drying Oven
Freon $3.0/sample
Extraction thimble $2.00/sample
Freon $5.0/sample
Silica Gel $1.2/sample
Summary 1.6 hr/sample $17.2/sample
interactive labor
4 hr/batch extraction
time
-------
TABLE 27. COST SUMMARY OP SONICATION METHOD OP TPH ANALYSIS
Activity
Time Requirement
In Total Labor Hours
Expendable Items
Capital Equipment Cost
Sample Preparation
0.2 hr/sample
Acid $0.2/Sample
Magnesium Sulfate $1.8/sample
Beakers, 250 ml
Heavy Duty $3.25/each
Extraction
0.2 hr/sample
Freon $6.2/sample
Sonication Extractor
$4,000
Extraction Beakers
$3.25/each
Analysis
0.9 hr/sample
Freon $5.0/sample
Silica Gel $1.2/sample
IR Spectrophotometer
IR Cells $1100-$1500/pair
Cleanup of Extraction
and Analytical Apparatus
0.1 hr/sample
Acids and Freon $1.00/sample
Drying Oven
Summary 1.4 hr/sample $15.4/sample
-------
Since both methods produce statistically similar results, the
Sonication method is preferable for use for several reasons. Since the
extraction time for the sample is about 10 minutes versus 4 hours for the
Soxhlet method, a sample can be quickly turned-around by the laboratory for
more immediate results. The lab can produce results using the Sonication
method as quickly as two hours after receipt of the sample. Another
advantage to Sonication is that this method can easily handle water samples
with high solids content, whereas the Soxhlet method can be used only for
solid matrix samples. In addition, the Sonication method requires less
transferring of sample from various laboratory glassware thus reducing the
chance of spill, breakage, and loss of sample. The Sonication extraction
unit takes up less lab space than the Soxhlet apparatus, and the unit does
not have to be placed in a hood (although it should be vented through
tubing into a hood) whereas the Soxhlet method should be placed in a hood
to capture any escaping freon vapors.
One disadvantage of using the Sonication extractor is that the unit is
somewhat loud, and noise reduction housing is required for use of the unit.
Summary and Conclusions
A sample of SSM spiked with diesel and waste-oil were submitted to the
CDH laboratory for analysis of total petroleum hydrocarbon using both the
Soxhlet and Sonication method. Both soil matrices were analyzed at least
14 times using each of the two methods and the results were statistically
evaluated. The statistical evaluation indicates that within a 90 percent
confidence level, the sample results from the two methods are within a 10
percent range.
The Sonication method is preferable for use because, although the
material cost is similar, the time required for sample preparation is less.
The Soxhlet method requires that the samples extract for four hours,
whereas the Sonication method requires only 10 minutes per sample of
extraction time. The reduced extraction time provides the benefit of
increased flexibility for lab scheduling, and quicker turnaround of sample
results, if needed.
No drawbacks are currently known regarding the use of the Sonication
method over the Soxhlet method, except for the requirement of noise
reduction housing on the Sonication apparatus for lab personnel safety.
64
-------
REFERENCES
1. Chiou, C.T. 1989. Theoretical considerations of the partition uptake
of nonionic organic compounds by soil organic matter. In B.L. Savhney
and K. Brovn (eds.). Reactions and movement of organic chemicals in
soil. Soil Science Society of America, Inc. American Society of
Agronomy, Inc., Madison, VI.
2. EPA Office of Solid Waste and Emergency Response, 1988. Assessment of
International Technologies for Superfund Applications,
EPA/540/2-88/003.
3. Hoag, G.G. and M.C. Marley. 1986. Gasoline residual saturation in
unsaturated uniform aquifer materials. Journal of Environmental
Engineering, 112 (3):586-604.
4. Lindsay, V.L. 1979. Chemical Equilibria in Soils. John Wiley & Sons.
New York.
5. Mishra, S., J.C. Parker, and N. Singhal. 1989. Estimation of soil
hydraulic properties and their uncertainty from particle size
distribution data. Journal of Hydrology, 108:1-18.
6. Nelson, D.V. and L.E. Sommers. 1982. Total Carbon, and organic
matter. In A.L. Page (ed.) Methods of Soil Analysis Part 2, Agronomy
monography No. 9 (2nd ed.) Soil Science Society of America, Inc.,
Madison, VI.
7. PEI Associates, Inc. 1988. CERCLA BDAT SARM Preparation and Results
of Physical Soil Vashing Experiments, Cincinnati, OH.
8. Traver, R.P. 1989. Development and Use of EPA's Synthetic Soil Matrix
(SSM/SARM). U.S. EPA Releases Control Branch, Risk Reduction
Engineering Laboratory, Edison, NJ.
65
-------
-------
BIBLOGRAPHY
Alfieri, D. J., 1987. "Lesson to be Learned From a Spill", Pollution
Engineering, June 1987, pp. 40-41.
Assink, J. V., V. J. Van Den Brink, 1985. Contaminated Soil. First
International TNO Conference on Contaminated Soils, Utreckt, The
Netherlands, November 1985, pp. 655-667
Beynon, L. R., B. Goden, R. H. Lilie, 1983. Field Guide to Inland Oil
Spill Clean-Up Techniques. Concave's Oil Spill Clean-Up Technology
Special Task Force No. 3.
Bissett, J., P. Vidden, 1972. "An Automatic, Multichamber Soil-Washing
Apparatus For Removing Fungal Spores From Soil". Canadian Journal of
Microbiology. Vol. 18 pp. 1399-1404.
Brovn, L. R., 1987. "Oil Degrading Microorganisms". Chemical Engineering
Progress, Oct. 1987, pp. 35-40.
Camp, F. U., 1969. "Tar Sands", Kirk-Othmer, Encyclopedia of Chemical
Technology, Volume 19, 2nd Edition, Interscience Publishers, Nev York.
Canterford, J. H., 1985. "Hydrometallurgy: Winning Metals vith Water".
Chemical Engineering, Oct. 28, 1985. pp. 41-48.
Chikote, D. D., T. J. Chresand, S. B. Valine, 1989. Conceptual Designs for
In-Situ Bioreclamation of Contamination Sites. BioTrol, Chaska, MI.
Chilcote, D. D., T. J. Chresand, S.B. Baline, 1989. Soil Washing System
for Use at Wood Preserving Sites. Biotrol, Chasha, MI. U.S.
Environmental Protection Agency International Symposium on Hazardous
Waste Treatment.
Coles, E. Y., R. Raghavan, D. Williams. Soil Washing - Removal of
Semivolatile Organics Using Aqueous Surfactant Solutions. Risk
Reduction Engineering Laboratory Office of Research and Development
U.S. Environmental Protection Agency.
Connor, R. J. 1988, "Case Study of Soil Venting". Pollution Engineering,
July 1988, pp. 74-78.
Coyle, B. H. Jr., J. A. Koperski, R. N. Anderson, 1977.
"Hydrometallurgical Separation of The Zinc - Base Fraction of Shredded
Automobiles". Conservation & Recycling, Vol. 1 pp. 189-200.
66
-------
Cracker, V. A., 1982. "Evaluation of Sludge Devatering Alternatives At a
Metallurgical Refinery". Journal VPCF, Vol. 54, No. 10, pp. 1417-1424.
Dick, M. R. 1982. "Ultrafiltration For Oily Wastewater Treatment."
Lubrication Engineering. Vol. 38, No. 4 pp. 219-222.
Ellis, W. D., J. R. Payne, 1984. "The Development of Chemical
Countermeasures for Hazardous Waste Contaminated Soil". Proc. of The
Hazardous Spills Conference Proceedings, pp. 117-124.
Ellis, V. D., J. R. Payne, In Situ Flushing & Soils Washing Technolgoies
for Superfund Sites. Hazardous Waste Engineering Research Laboratory,
U.S. Environmental Protection Agency, RCRA/Superfund Engineering
Technology Transfer Symposium.
Esposito, P., et. al., 1988. Results of Treatment Evaluations of
Contaminated Soils. PEI Associates, Inc. Cincinatti, OH and U.S.
Environmental Protection Agency, Hazardous Waste Engineering Research
Laboratory. Vol. 5 pp 41-79
Fairveather, V., Ed.f 1988. "Bio-Polymer Slurry Promotes Drainage". Civil
Engineering. Mar. 1988, pp. 16
Gibson, S. C., 1980. Experimental Oily Wastewater Separation System.
Rockwell International - Environmental Monitoring & Services Center.
Gilmer, H., F. J. Freestone, 1977. Clean-Up of An Oil and Mixed Chemical
Spill at Dittmer, Missouri, April - May, 1977. U.S. Environmental
Protection Agency, Edison, NJ. pp. 131-134.
Griffiths, R. A., 1982. An Overview of Current Spill Clean-Up Technology.
Municipal Environmental Research Laboratory, Office of Research &
Development U.S. Environmental Protection Agency, Cincinnati, OH
Gunn, D. 1988. Summary of Findings on Evaluation of Aqueous Degreasers
Verses Chlorinated Solvents At The Paducah Gaseous Diffusion Plant.
Padacah Gaseous Diffusion Plant, Martin Marietta Energy Systems, Inc.,
Padacah, KY
Haque, K. E. 1988. Radium (226) Removal From a Contaminated Soil. Mineral
Sciences Laboratories Division - Report MSL 88-143 (OP&J).
Humenick, M. J., Davis, B. J., 1978. "High Rate Filtration of Refinery
Oily Wastewater Emulsions". Journal WPCF, Aug. 1978, pp. 1953-1964.
Sudan, M. T., C. E. Strubler, S. W. Kao, J. T. Pfeffer, 1983. "Treatment
of Coal Gasification Wastewater With Anaerobic Filter Technology".
Journal WPCF, Vol. 55, No. 10: 1263-1270.
Kelly, J. M., R. C. Strickland, 1987. "Soil Nutrient Leaching In Response
To Simulated Acid Rain Treatment". Water, Air and Soil Pollution Vol.
34 pp. 167-181.
67
-------
Luthy, R. G., M.J. Carter,. "Leaching Characteristics of Coal Gasification
Process Ash and Char". Containments and Sediments, Vol. 2. pp. 137-166.
Mayo, D. V., D. S. Page, J. Cooley, E. Sorenson, F. Bradley, 1978.
"Weathering Characteristics of Petroleum Hydrocarbons Deposited in Fine
Clay Marine Sediments, Searsport, Maine". J. Fisk. Ras. Board Can.
Vol. 35: pp. 552-562.
Neufeld, R. D., S. Uallack, 1984. "Chemical and Toxicity Analysis of
Leachates from Coal Conversion Solid Wastes". Journal UPCF, Vol. 56,
No. 3 pp. 266-273.
Oblinger, F. G., M. Uientraub, E. J. Blais, R. L. Gealer, 1986.
"Automation of An Electrolytic Cell For The Treatment of Oily
Wastewater". Environmental Progress, Vol. 5, No. 4 pp. 283-286.
Perras, J. C., 1973. "A Portable Gas Chromatographic Technique to Measure
Dissolved Hydrocarbons In Sea Water". Nova University, Dania FL.
Schmare, D. W., 1979. "Extraction of Organic Matter In Water - The
Carolina Method". Journal WPCF, Vol. 51, No. 10, pp. 2467-2473.
Stover, E. L, M. F. Torpy, 1982. "Petroleum Processing Wastes". Journal
WPCF, Vol, 54, No. 6, pp. 723-725.
Talenta, M., W. H. Smit, Y. C. Tuang, "Disposal of Oily Petroleum Sludge by
Sludge Farming". Asian Environment, Vol. No. pp. 13-16.
Thorhaug, A., J. Marcus, 1987. "Oil Spill Clean-Up: The Effect of Three
Dispersants on Three Subtrapical/Tropical Seagrasses". Marine
Pollution Bullet in, Vol, 8. No. 3 pp. 124-126.
U.S. Environmental Protection Agency, 1987. Superfund Record of Decision:
Zellvood, Fl.
U.S. Environmental Protection Agency, 1988. Superfund Record of Decision:
L.A., Clarke & Son, VA.
U.S. Environmental Protection Agency, P.A., 1988. Superfund Record of
Decision: United Scrap Lead, OH.
U.S. Environmental Protection Agency, P.A., 1988. Superfund Record of
Decision: Koppers Texarkana, TX.
U.S. Environmental Protection Agency P.A., 1988. Superfund Record of
Decision: South Cavalcade Street, TX.
U.S. Environmental Protection Agency, P.A., 1987. Harbauer Soil Cleaning
System.
Warren Spring Lab., Stevenage England. 1986. "Beach Cleaning Trials:
Pendine Sands". Warren Spring Lab., Stevenage, England.
68
-------
Young, J. C., 1977. "Removal of Oil and Grease by Biological Treatment
Processes". Journal VPCF. Vol. 51, No. 8 pp. 2084-2086.
Zall, J., N. Gail, M. Rehbun, 1987. "Skeleton Builders For Conditioning
Oily Sludge". Journal VPCF, Vol. 59, No. 7, pp. 699-705.
69
-------
APPENDICES
70
-------
-------
APPENDIX A
SENSITIVITY ANALYSIS CURVES POR OPERATIONAL PARAMETERS ON
GASOLINE AND DIBSEL CONTAMINATED SSM - PHASE I
71
-------
-------
100
95
(0
>
o
E
D
DC
X
a.
E-i
90
85
80
I V VIII
10 15 20 25
Contact Time (min.)
30
° No. 10 Sieve
D No. 60 Sieve
* No. 14 0 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Additive - None
Wash Temperature - 25-29 C
Rinsewater : Washwater - 3:1
35
Figure A1. Contact time effect on percent TPH removal.
100
95
90
(0
£ 85
80
x
w
E-
E 15
c*°
10
65
60
0 10 20 30
Contact Time (min.)
: 1
I i
i
f
\
I
« j
: |
j
; !
I I
1 1
i | I
° No. 10 Sieve
n No. 6C Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Additive - None
Wash Temperature - 25-29 C
Rinsewater : Washwater - 3:1
40
Figure A2. Contact time effect on percent BTEX removal.
72
-------
100
80
10
Figure A3.
20 30
Soaking Time (min.)
° No. 10 Sieve
0 No. 60 Sieve
x No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 2:1
Additive - 0.067% CitriKleen
Wash Temperature - 25-29 C
Wash Time - 1 min.
40
Soaking time effect on percent TPH removal
with 0.067% CitriKleen.
100
80 ' '
Figure A4.
5 10 15
Shaking Time (min.)
° No. 10 Sieve
° No. 60 Sieve
x No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 2:1
Additive - 0.067% CitriKleen
Wash Temperature - 25-29 C
Contact Time - 30 min.
Rinsewater : Washwater - 6:1
20
Shaking time effect on percent TPH removal
with 0.067% CitriKleen.
73
-------
1 2
Washwater to SSM Mass Ratio
° No. 10 Sieve
D No. 60 Sieve
x No. 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Rinsewater : Washwater - 3:1
Additive - None
Wash Temperature - 25-29 C
Contact Time - 30 min.
Figure A5 . Washwater to soil ratio effect on percent BTEX removal.
100
90
80
>—1
*J
E 70
a>
cc
£ 60
E-i
OP
50
40
30
1
1
J 1
1
/
tj
I
° No. 10 Sieve
° No. 60 Sieve
x No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Rinsewater : Washwater - 3:1
Additive - None
Wash Temperture - 25-29 C
Contact Time - 30 min.
1 2
Washwater to SSM Mass Ratio
Figure A6. Washwater to soil ratio effect on percent TPH removal.
74
-------
100
95
*
>
o
e
0)
cc
X
Q-
E-<
90
85
80
T
° No. 10 Sieve
° No. 60 Sieve
x No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Rinsewater : Washwater - 3:1
Additive - 0.067% CitriKleen
Contact Time - 30 min.
Wash Temperature - 25-29 C
0 12 3
Washwater to SSM Mass Ratio
Figure A7. Washwater to soil ratio effect on percent TPH removal
with 0.067% CitriKleen.
100
95
>
o
6
0)
X
X
u
E-
B3
90
85
80
o
No .
10 Sieve
~
NO .
60 Sieve
X
NO.
140 Sieve
—I 1—
2 4 6 8 10 12
Rinsewater to Washwater Volume Ratio
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Additive - None
Wash Temperature - 25-29 C
Contact Time - 30 min.
Figure A8. Rinsewater to washwater ratio effect on percent BTEX removal.
75
-------
100
(0
>
o
E
0)
X
X
Cl
Eh
95
90
85
80
r
4 6 8 10 12
Rinsewater to Washwater Volume Ratio
® No. 10 Sieve
° No. 60 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Additive - None
Wash Temperature - 25-29 C
Contact Time - 30 ir.in.
Figure A9. Rinsewater to washwater ratio effect on percent TPH removal.
100
<0
>
o
E
(D
X
X
u
E-
m
80
60
40
20
T"
~
K
° No. 10 Sieve
° No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 2:1
Additive - None
Wash Temperature - 25-29 C
Contact Time - 30 min.
3 4 5 6 7
Rinsewater to Washwater Volume Ratio
Figure A10. Rinsewater to washwater ratio effect on percent BTEX removal.
76
-------
105
80
0 No. 10 Sieve
D No. 60 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 2:1
Additive - None
Temperature - 25-2 9 C
Contact Time - 30 min.
2 3 4 5 6 7
Rinsewater to Washwater Volume Ratio
Figure All. Rinsewater to washwater ratio effects on percent TPH removal.
105
80
° No. 10 Sieve
D No. 60 Sieve
x No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 2:1
Additive - 0.067% CitriKleen
Temperature - 25-29 C
Contact Time - 1 min. wash
30 min. soak
2 3 4 5 6 7
Rinsewater to Washwater Volume Ratio
Figure A12. Rinsewater to washwater ratio effect on percent TPH removal.
7 7
-------
0
!
o i
~
o
~
X
X |
i
i
° No. 10 Sieve
° No. 60 Sieve
x No. 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Wash Temperature - 25-2 9 C
Contact Time - 30 min.
0.0 0.2 0.4 0.6 0.8
% CitriKleen in Washing Solution
Figure A13. Percent CitriKleen effect on percent BTEX removal.
100
95
90
85
80
0.0
Figure A14.
0.2
0.4
0.6
° No. 10 Sieve
° No. 60 Sieve
x No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Wash Temperature - 25-29 C
Contact Time - 30 min.
0.8
% CitriKleen in Washing Solution
Percent CitriKleen effect on percent TPH removal.
78
-------
100
98
o
E
a>
cc
x
u
Eh
CD
96
94
92
1
1
b
1
—V
a
a
0
0
1
1
1
t
>
X
20 40 60
Temperature (C)
80
° No. 10 Sieve
D No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater
Additive - None
Contact Time - 30 min.
- 3:1
Figure A15. Temperature effect on percent BTEX removal
100
90
96
>
0
£
01
06 94
a,
H
92
90
88
20 40 60 80
Temperature (C)
100
° No. 10 Sieve
° No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Additive - None
Contact Time - 30 min.
Figure A16. Temperature effect on percent TPH removal
-------
20 30 40 50 60 "70
Temperature (C)
"81
80
® No. 10 Sieve
D No. 60 Sieve
* No. 14 0 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Additive - 0.13% CitriKleen
Contact Time - 30 min.
Figure A17. Temperature effect on percent BTEX removal
with 0.13% CitriKleen
20 30 40 50 60 70
Temperature (C)
80
® No. 10 Sieve
° No. 60 sieve
* No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater (mass)- 1:1
Rinsewater : Washwater - 3:1
Additive - 0.13% CitriKleen
Contact Time - 30 min.
Figure A18. Temperature effect on percent TPH removal
with 0.13% CitriKleen.
80
-------
.t
l
j
i
¦" (
X
'0 ",,~
20 30 40 50
Temperature (C)
60
° No. 10 Sieve
D No. 60 Sieve
* No. 140 sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater
Additive - 0.5% Tide
Contact Time - 30 min.
- 3:1
Figure A19. Temperature effect on percent BTEX removal with 0.5% Tide.
~|
¦ i
—
—
¦
1
1
O
X
20
30 40 50
Temperature (C)
60
° No. 10 Sieve
° No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Additive -0.5% Tide
Contact Time - 30 min.
Figure A20. Temperature effect on percent TPH removal with 0.5% Tide.
bi
-------
APPENDIX B
PARTICLE SIZE DISTRIBUTION CURVES FOR
BENCH SCALE SSH SOIL VASHING EXPERIMENTS - PHASE I
82
-------
-------
100
to
O) •
>
0)
•H
10
C
c
•H
n
n
<0
cu
90
80
70
60
50
40
¦ i i i
10 15 20 25 30
Contact Time (min.)
35
° No. 10 Sieve
° No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater:Washwater - 3:1
Additive - None
Wash Temperature 25-29 C
Figure B1. Contact time effect on particle size distribution
for diesel contaminated soil.
100
90
in
0)
I 80
w
C
c
to
to
IB
a
70
60
50
40
1
1
<
>
1
1
1
i
!
1
:
E
10 15 20 25 30
Contact Time (min)
35
° No. 10 Sieve
° No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Additive - None
Wash Temperature 25-29 C
Figure B2. Contact time effect on particle size distribution
for gasoline contaminated soil
83
-------
10
<
1
(
»
1
1
1
1
]
t
1
s
20 30
Soaking Time (min.)
40
® No. 10 Sieve
D No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 2:1
Additive - 0.067% CitriKleen
Wash Temperature - 25-29 C
Wash Time - 1 min.
Figure B3. Soaking time effect on particle size distribution
for diesel contaminated soil.
o [
(
>
!
B
i
i
i
e
° No. 10 Sieve
Q No. 60 Sieve
* No. 14 0 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 2:1
Rinsewater : Washwater - 6:1
Wash Temperature - 25-29 C
Additive - 0.067% CitriKleen
Contact Time - 30 min.
0 5 10 15 20
Shaking Time (min.)
Figure B4. Shaking time effect on particle size distribution
for diesel contaminated soil.
84
-------
100
90
m
70
60
50
40
i
o
No .
10 Sieve
~
No .
60 Sieve
X
No.
140 Sieve
Type of SSM: High Gas
Constant Parameters:
Rinsewater : Washwater - 3
Additive - None
Wash Temperature - 25-29 C
Contact Time - 30 min.
0 12 3
Washwater to SSM Mass Ratio
Figure B5. Washwater to soil ratio effect on particle size distribution
for gasoline contaminated soil.
100
90
to
<1)
>
<1)
Oi
C
•H
n
tn
<0
cu
80
70
60
50
40
o
No .
10 Sieve
~
NO .
60 Sieve
X
NO .
140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Rinsewater : Washwater - 3:
Additive - None
Wash Temperature - 25-29 C
Contact Time - 30 min.
0 12 3
Wash Water to SSM Mass Ratio
Figure B6. Washwater to soil ratio effect on particle size distribution
for diesel contaminated soil.
8b
-------
100
90
m
0)
I 80
•H
V)
•5 70
n
n
-------
100
90
00
01
I 80
-H
to
Oi
c 7°
n
n
a
60
50
40
O No. 10 Sieve
D No. 10 Sieve
* No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Additive - None
Wash Temperature - 25-29 C
Contact Time - 30 min.
2 4 6 8 10 12
Rinsewater to Washwater Volume Ratio
Figure B9. Rinsewater to washwater ratio effect on particle size
distribution for diesel contaminated soil.
° No. 10 Sieve
D No. 60 Sieve
x No. 14C Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 2:1
Additive - None
Wash Temperature - 25-29 C
Contact Time - 30 min.
2 3 4 5 6 7
Rinsewater to Washwater Volume Ratio
Figure B10. Rinsewater to washwater ratio effect on particle size
distribution for gasoline contaminated soil.
8/
-------
100
90
80
n
0)
>
50
40
° No. 10 Sieve
° No. 60 Sieve
x No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 2:1
Additive - None
Temperature - 25-2 9 C
Contact Time - 30 min.
2 3 4 5 6 7
Rinsewater to Washwater Volume Ratio
Figure Bll. Rinsewater to washwater ratio effect on particle size distribution
for diesel contaminated soil.
100
90
m
0)
5 80
CO
6
.5 70
in
tn
10
Cb
60
50
40
¦
¦
(
1
(
>
¦
•
¦
1
1
1
1
¦
i
1—
—3
° No. 10 Sieves
D No. 60 Sieves
* No. 140 Sieves
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 2:1
Additive - 0.067% CitriKleen
Wash Temperature - 25-29 C
Contact Time - 1 min. wash
30 min. soak
2 3 4 5 6 7
Rinsewater to Washwater Volume Ratio
Figure B12. Rinsewater to washwater ratio effect on particle size distribution
for diesel contaminated soil.
88
-------
100
90
in
0)
%> 80
•H
w
O1
c io
m
v>
10
Oi
60
50
40
® No. 10 Sieve
° No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3
Wash Temperature - 25-2 9 C
Contact Time - 30 min.
0.0 0.2 0.4 0.6 0.8
% CitriKleen in Washing Solution
Figure B13. Percent CitriKleen effect on particle size distribution
for gasoline contaminated soil.
90 i
0.0 0.2 0.4 0.6 0.8
% CitriKleen in Washing Solution
° No. 10 Sieve
° Nc. 60 Sieve
* No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3
Wash Temperature - 25-29 C
Contact Time - 30 min.
Figure B14.
Percent CitriKleen effect on particle size distribution
for diesel contaminated soil.
-------
1 » 1
o <
(
1
O
a
0
© No. 10 Sieve
° No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater -
Additive - None
Contact Time - 30 min.
o u
70 '
60 "
1
1
~
a
I
bO "
40 '
X
K
3:1
20 40 60
Temperature (C)
80
Figure B15. Temperature effect on particle size distribution
for gasoline contaminated soil.
20 40 60
Temperature (C)
80
100
o
No .
10 Sieve
~
No .
60 Sieve
X
No .
140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Additive - None
Contact Time - 30 min.
Figure B16. Temperature effect on particle size distribution
for diesel contaminated soil.
90
-------
90
80
>
0)
w 70
er-
c
3 60
&
50
40
20 30 40 50 60 TO
Temperature (C)
80
° No, 10 Sieve
O No. 60 Sieve
* No, 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Additive - 0,134 CitriKleen
Contact Time - 30 min.
Figure B17. Temperature effect on particle size disribution
for gasoline contaminated soil with 0.13% CitriKleen.
90
80
m
<2
>
0)
c/i 7 0
CP
c
•H
CO
S 60
Oi
50
40
0
o
0
~
6
c
X
«
X
o
No.
10 Sieve
~
No .
60 Sieve
X
No.
140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Additive - 0,131 CitriKleen
Contact Time - 30 min.
20 30 40 50 60 70 80
Temperature (C)
Figure BIB. Temperature effect on particle size distribution
for diesel contaminated soil with 0.13% CitriKleen.
91
-------
90
80
0)
>
m
•H
t/i
70
-5 60
n
n
«
o<
50
<#>
40
30
20 30 40 50
Temperature (C)
60
° No. 10 Sieve
D No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Gas
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater - 3:1
Additive - 0.5% Tide
Contact Time - 30 min.
Figure B19. Temperature effect on particle size distribution
for gasoline contaminated soil with 0.5% Tide.
90
80
in
<1>
I 70
•H
CO
.5 60
co
co
in
o<
50
40
30
20
Figure B20.
o
I
V)
~
1
b
1
[
X
*'
30 40 50
Temperature (C)
60
° No. 10 Sieve
° No. 60 Sieve
* No. 140 Sieve
Type of SSM: High Diesel
Constant Parameters:
Soil : Washwater - 1:1
Rinsewater : Washwater
Additive - 0.5% Tide
Contact Time - 30 rr.in.
-3:1
Temperature effect on particle size distribution
for diesel contaminated soil with 0.5% Tide.
9Z
------- |