&EPA
United States
Environmental Protection
Agency
Risk Reduction Engineering EPA/540/5-89/004a
Laboratory June 1989
Cincinnati, OH 45268
Superfund
Technology Evaluation
Report:
SITE Program
Demonstration Test
International Waste
Technologies In Situ
Stabilization/Solidification
Hialeah, Florida
Volume I
SUPERFUND INNOVATIVE
TECHNOLOGY EVALUATION
-------
-------
EPA/540/5-89/004a
June 1989
Technology Evaluation Report:
SITE Program Demonstration Test
International Waste Technologies In Situ
Stabilization/Solidification
Hialeah, Florida
Volume I
Risk Reduction Engineering Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
-------
NOTICE
The information in this document has been funded by the U.S.
Environmental Protection Agency under Contract No. 68-03-3255
and the Superfund Innovative Technology Evaluation (SITE)
Program. It has been subjected to the Agency's peer review and
administrative review and it has been approved for publication
as a USEPA document. Mention of trade names or commercial
products does not constitute an endorsement or recommendation
for use.
11
-------
FOREWORD
ve Technology Evaluation (SITE)
the 1986 Superfund amendments. The
between EPA's Office of Research and
Solid Waste and Emergency Response.
is to assist the development of
technologies necessary to implement
ch require greater reliance on
is accomplished through technology
provide engineering and cost data on
The Superfund Innovati
program was authorized in
program is a joint effort
Development and Office of
The purpose of the program
hazardous- waste treatment
new cleanup standards, whi
permanent remedies. This
demonstrations designed to
selected technologies.
This project consists of an analysis of the International
Waste Technologies proprietary in situ
stabilization/solidification process and represents the sixth
field demonstration in the SITE program. The technology
demonstration took place at a former electric service shop owned
by General Electric Company in Hialeah, Florida. The
demonstration effort was directed at obtaining information on
the performance and cost of the process for use in assessments
at other sites. Documentation will consist of two reports. The
Demonstration Test Report describes the field activities and
laboratory results and has been previously issued. This
Application Analysis Report provides an interpretation of the
available data and presents conclusions on the results and
potential applicability of the technology.
Additional copies of
charge from EPA's Center
this report may be obtained at no
for Environmental Research Information,
26 West Martin Luther King Drive, Cincinnati, Ohio, 45268, using
the EPA document number found on the report's front cover. Once
this supply is exhausted, copies can be purchased from the
National Technical Information Service, Ravensworth Building,
Springfield, Virginia, 22161, (702) 487-4600. Reference copies
will be available at EPA libraries in their Hazardous Waste
Collection. You can also call the SITE Clearinghouse hotline at
1-800-424-9346 or (202) 382-3000 in Washington, D.C. to inquire
about the availability of other reports.
-------
ABSTRACT
A demonstration of the International Waste Technologies
(IWT) process, utilizing the Geo-Con, Inc. deep- soil-mixing
equipment has been performed under the Superfund Innovative
Technology Evaluation (SITE) Program. This was the first
field demonstration of an in situ stabilization/
solidification process. The demonstration occurred in April
1988 at the site of a General Electric Co. electric service
shop in Hialeah, Fla. where the soil contained polychlorinated
biphenyls (PCBs) and localized concentrations of volatile
organics and heavy metal contaminants. The demonstrated
process mixed in situ the contaminated soil with a mixture of
a proprietary additive, called HWT-20, and water.
The technical criteria used to evaluate the effectiveness
of the IWT process were contaminant mobility, based on
leaching and permeability tests; and the potential integrity
of solidified soils, based on measurements of physical and
microstructural properties. The performance of the Geo-Con
deep- soil-mixing equipment was also evaluated.
The process did appear to immobilize PCBs. However,
because of very low PCB concentrations in the leachates,
caused in part by the low concentrations of PCBs in the
untreated and treated soils, absolute confirmation of PCB
immobilization in this SITE project was not possible.
Physical properties were satisfactory except for the
freeze/thaw weathering test, where considerable degradation of
the test specimens occurred. The microstructural analyses
showed the process produced a dense homogeneous mass with low
porosity.
The Geo-Con deep-soil-mixing equipment performed well,
with only minor difficulties encountered, which can be easily
corrected. The HWT-20 additive was well dispersed into the
soil, as evidenced by the relatively uniform change in
chemical and physical characteristics of treated soil versus
untreated soil .
The cost per ton of treating contaminated soil under the
demonstration conditions was determined to be approximately
$194.
-------
VOLUME I
CONTENTS."
Foreword . . i i i
Abstract iv
Figures vi
Tables vii
Abbreviations and Symbols viii
Conversions ix
Acknowledgement x
1. Executive Summary 1
1.1 Introduction 1
1.2 Process description 3
1.3 Sampling and analysis program 4
1.4 Results and discussion 5
2. Introduction 9
2.1 Background 9
2.2 Program objectives 10
2.3 Technology evaluation criteria 11
2.4 Description of operations 12
2.5 Project organization 13
3. Summary of Performance Data and Evaluation 14
4. Process Description 18
4.1 Reaction mechanisms 18
4.2 Equipment specifications 19
5. Site Characteristics 29
5.1 Sitedescription 29
5.2 Waste characteristics 31
6. Demonstration Procedures 32
6.1 Site and waste preparation 32
6.2 Operational plan 32
6.3 Sampling and analysis activities ........ 33
6.4 Screening analysis 43
6.5 Physical tests 44
6.6 Chemical tests 46
7. Field Activities 51
7.1 Operational history 51
7.2 Unit problems and deviations from
demonstration plan 54
8. Performance Data and Evaluation 58
8.1 Physical tests 58
8.2 Chemical tests 62
8.3 Microstructural studies 66
8.4 Material balance 68
8.5 Data quality assurance 68
9. Economics 84
9.1 Introduction 84
9.2 Cost elements 85
9.3 Overall cost evaluation 89
* Volume II contains eight appendices: [A] EPA operating log
data; [B] LE operating log data; [C] Operations report of
Geo-Con; [D] Microstructural and phase identification study of
pretreatment and treated soil samples from Hialeah, Fla.; [E]
Laboratory report: Pretreatment results; [F] Laboratory
report: Posttreatment results; [G] Sampling log sheets; and
[H] PCB site profiles.
-------
FIGURES
Page
1. Overlapping column arrangement . 22
2. Batch mixing plant 23
3 Soil mixing auger 24
4. Mixing auger on downstroke 25
5. Batch mixing plant additive storage 26
6. Slurry feed system 27
7. Overall view of auger assembly 28
8. Miami area regional geological cross-section 30
9. Sampling sector locations 35
10. Sampling locations - Sector B 36
11. Sampling locations - Sector C 37
12. Location of soil columns in Sector B . 52
VI
-------
TABLES
Page
1. Equipment specifications 21
2. Pretreatment analyses 40
3. Posttreatment analyses 41
4. Screening samples results .... 43
5. Physical properties of untreated soils - Sector B . . 74
6. Physical properties of untreated soils - Sector C . . 75
7. Physical properties of treated soils - Sector B ... 76
8. Physical properties of treated soils - Sector C ... 77
9. Results of formulation studies 78
10. PCBs in soils and leachates - Sector B . . . 79
11. PCBs in soils and leachates - Sector C 80
12. Total volatile organics in soils and leachates .... 81
13. Total of four priority pollutant metals
in soils and leachates 82
14. Material balance 83
15. Cost element breakdown 86
16. Estimated cost . . . ; 90
VI 1
-------
ABBREVIATIONS AND SYMBOLS
AA
ANS 16.1
ASTH
cra/s
DSM
EPA
ft
9
gal
GC/ECD
GC/HS
gpm
GE
h
HWT-20
IDLH
IWT
kg
Ib
MCC-1P
mg
ml
NIOSH
O&G
ORD
OSWER
PCB
ppm
psi
rpm
SEM
SITE
SWS
TCLP
TOC
M
UCS
VOC
XRD
atomic absorption - flame or furnace — for metals
detection
Leach test used by the American Nuclear Society
American Society for Testing and Materials
centimeters per second - permeability coefficient
units
deep soil mixing
Environmental Protection Agency
foot
gram
gallon
gas chromatography/electron capture detector
gas chromatography/mass spectrometry
gallons per minute
General Electric Co.
hour
proprietary additive of International Waste
Technologies
immediately dangerous to life or health
International Waste Technologies
kilogram
1 iter
pound
Leach test developed by the Materials
Characterization Center
milligram
mi 11i1i ter
National Institute for Occupational Safety and
Health
oil and grease
EPA Office of Research and Development
EPA Office of Solid Waste and Emergency Response
polychlorinated biphenyl
parts per million
pounds per square inch
revolutions per minute
scanning electron microscope
Superfund Innovative Technology Evaluation
Scientific Waste Strategies, Inc.
Toxicity Characteristic Leaching Procedure
total organic carbon
micron
unconfined compressive strength
volatile organic compound
X-ray diffraction
vi 11
-------
Area:
Flow rate
Length:
Mass:
Volume
CONVERSIONS
English (US) Metric (SI)
1 ft2
9.2903 x 10"3 m2
1 in
1 gal/min
1 gal/min
1 Mgal/d
1 Mgal/d
1 Mgal/d
1 ft
1 in
1 yd
1 Ib
1 Ib
1 ft3
1 ft3
1 gal
1 gal
6.4516 cm2
6.3090 x 10"5 m3/s
6.3090 x 10'2 L/s
43.8126 L/s
3.7854 x 103 m3/d
4.3813 x 10'2 m3/s
0.3048 m
2.54 cm
0.9144 m
4.5359 x 102 g
0.4536 kg
28.3168 L
2.8317 x 10"2 m3
3.7854 L
3.7854 x 10
-3 m3
ft = foot, ft2 = square foot, ft3 = cubic foot
in = inch, in2 = square inch
yd = yard
Ib = pound
gal = gallon
gal/min = gallons per minute
Mgal/d = million gallons per day
m = meter, nr = square meter, m3 = cubic meter
cm = centimeter, cm* = square centimeter
L = 1iter
g = gram
kg = kilogram
m /s = cubic meters per second
L/s = liters/second
o
m°/d = cubic meters per day
-------
ACKNOWLEDGEMENT
This report was prepared under the direction and coordination
of Mary Stinson, EPA SITE Program Manager in the Risk Reduction
Engineering Laboratory, Cincinnati, Ohio. Contributors and
reviewers for this report were the USEPA's Office of Research
and Development; Richard Valentinetti of the Office of
Environmental Engineering and Technology Demonstration; the
USEPA's Office of Solid Waste and Emergency Response; Linda
Galer of the Office of Program Management and Technology;
Jeffrey Newton of International Waste Technologies; Brian
Jasperse of Geo-Con, Inc.; John Harrsen of General Electric
Company; Walter Sumansky of NUS Corp.; and Frank Cartledge,
Harvill Eaton, and Marty Tittlebaum of Scientific Waste
Strategies, Inc.
This report was prepared for EPA's Superfund Innovative
Technology Evaluation (SITE) Program by Stephen Sawyer of
Foster Wheeler Enviresponse, Inc. for the U.S. Environmental
•Protection Agency under Contract No. 68-03-3255.
-------
SECTION 1
EXECUTIVE SUMMARY
1.1 INTRODUCTION
In response to the Superfund Amendments and Reauthorization
Act of 1986 (SARA), the Environmental Protection Agency's Office
of Research and Development (ORD) and Office of Solid Waste and
Emergency Response (OSWER) have established a formal program to
accelerate the development, demonstration, and use of new or
innovative technologies as alternatives to current containment
systems for hazardous wastes. This program is called Superfund
Innovative Technology Evaluation, or SITE.
The major objectives of the SITE Program are to develop
reliable cost and performance information. One process, which
was demonstrated in April 1988 at a General Electric Co. (GE)
electric service shop in Hialeah, Fla., as part of the SITE
Program, was the International Waste Technologies (IWT) in situ
stabilization/solidification process, using Geo-Con, Inc. deep-
soil-mixing equipment. This was the first field demon'strati on of
an in situ stabilization/solidification process. The
demonstration was performed to meet the goals of the SITE program
along with those of GE, which were significantly different. GE's
goals were to meet the requirements of the Metropolitan Dade
County Environmental Resources Management (MDCERM) for the
immobilization of PCBs. The SITE project proposed to determine
the technological and economic viability of the in situ
stabilization/solidification process, as defined in the
Demonstration Plan, and involved a more expansive testing program
than that required of GE by MDCERM.
IWT, the stabilization/solidification technology developer,
and Geo-Con, provider of the specialized drilling and mixing
equipment, were participants in both the SITE and GE programs.
Under the latter program, IWT and Geo-Con served as contractors
to GE for the mandated test before the site remediation. In
addition, under a cooperative agreement with EPA, IWT was
designated as the SITE technology developer for the demonstration
test; Geo-Con verbally agreed that its in situ procedures were to
be evaluated.
The IWT process involved the in situ mixing of the service
shop soil contaminated with polychlorinated biphenyls (PCBs) with
a cement-organo clay mix referred to as HWT-20. Two 10 x 20 ft
test sectors (designated Sectors B and C), relatively high in
PCBs for the site were treated. Sector B was treated to a depth
of 18 ft and Sector C to a depth of 14 ft. These depths were
defined by GE to treat all the soil containing at least 1.0 mg/kg
-------
of PCBs. The developer claimed the wastes would be immobilized
and bound into a hardened, leach-resistant, concrete-like
solidifi ed mass.
The major objectives of this SITE project were to determine
the following:
1. Ability of the stabilization/solidification technology to
immobilize PCBs. (If detected in the untreated soi1,
immobilization of volatile organics and heavy metals were to
be measured.)
2. Effectiveness, performance, and reliability of the Geo-Con
deep-soil-mixing equipment used for the in situ
solidification (including continuity of operation, uniformity
of mixing, and accuracy of column overlap).
3. Degree of soil consolidation (solidification) caused by the
chemical additives.
4. Probable long-term stability and integrity of the solidified
soil .
5. Costs for commercial-seale applications.
The following technical criteria were used to evaluate the
effectiveness of the in situ stabilization/solidification
process:
1. Mobility of the contaminants. Areas of high PCBs and VOCs
were heavily sampled, with the analytical emphasis on
leaching characteristics. Three leach tests were performed:
the Toxicity Characteristic Leaching Procedure (TCLP) and two
other leach tests, MCC-1P and ANS 16.1, which evaluate
performance in solidified blocks. Only the effectiveness on
PCBs was evaluated, as the additive was not designed to
immobilize other contaminants. Permeabilities also were
measured before and after soil treatment. These values
indicate the degree to which the material permits or
prohibits the passage of water through the soi.l mass, and
thus the degree of water contact with the contaminants.
2. Durability of the solidified soil mass. Core sections from
the solidified mass were analyzed to determine uniformity and
long-term endurance potential. However, if a chemical bond
forms between HWT-20 and the PCBs, as claimed by IWT, then
maintaining durability of the solidified mass to prevent the
mobility of the contaminant becomes less important. The
analyses obtained information on the following:
o Integrity of the remediated soil.
o Unconfined compressive strengths, which provided an
-------
1.2
indication of long-term durability.
Microstructural characteristics, which provided information
on treated soil porosity, crystalline structure, and degree
of mixing. This provided information on the potential for
long-term durability of the hardened mass.
Wet/dry and freeze/thaw weathering tests, which provided
information on weight loss. Permeability and unconfined
compressive strength tests of the weathered samples also were
performed. These tests provided a measure of short-term
durabi1i ty.
PROCESS DESCRIPTION
IWT claims that their HWT-20 additive generates a complex
crystalline connective-network of inorganic polymers. Structural
bonding in the polymer is mainly covalent. There is a two-phased
reaction in which the contaminants are complexed, first in a
fast-acting reaction, and then permanently complexed further in
the building of macro-molecules, which continue to generate over
a long period of time. The bonding characteristics and
durability of the structure are adapted by varying the
composition of the HWT additive to suit a particular situation.
The Geo-Con/DSM deep soil mixing system of mechanical mixing
and injection consisted of one set of cutting blades and two sets
of mixing blades attached to a vertical drive auger, which
rotated at approximately 15 rpm. Two conduits in the auger
allowed for the injection of the additive slurry and supplemental
water. HWT-20 additive injection was on the downstroke, with
further mixing occurring on auger withdrawal. The treated
36-in.-diameter soil columns were positioned in an overlapping
pattern (see Figure 1). In each sector, alternating primary and
secondary soil columns existed, with all the primary columns
prepared before the secondary columns were augered. Thus, the
secondary soil columns were drilled between treated soil columns
and represented only 75% new area relative to the primary
columns. Geo-Con indicates that this is a more efficient method
of soil treatment.
A batch mixing system processed the feed additives. HWT-20
powder was conveyed by air from a supply truck to a storage
silo. To treat three or four soil columns, a measured amount of
water was fed to a 1,000-gallon mixing tank. The HWT-20 was fed
to the tank at a weight ratio to water of 4:3. A screw pump then
pumped the slurry to the auger. Water was fed separately to the
drill rig on a ratio basis to the additive slurry. Sufficient
water was provided to produce a final soil product containing
1.6-1.7 Ib of water/lb of HWT-20. For the Demonstration Test,
sodium silicate was added to the bottom 3-4 ft of each column to
provide a quick-setting boundary layer. The IWT technology does
-------
not consider this addition of sodium silicate as part of their
process.
1.3
SAMPLING AND ANALYSIS PROGRAM
Soil sampling, provided by EPA, was performed two weeks
before, and five weeks after, the remediation of the test
sectors. Sampling was carried out at soil column centers, at
column interfaces, and at five locations around one anticipated
hot spot in each sector. Samples were taken at three or four
depths, from the top layer of unconsolidated sand, the limestone
layer, and the lower unconsol idated sand layer. Sampling in
Sector B also was done in the sodium silicate boundary layer.
Chemical analyses were performed to identify and quantify soil
contaminants in both the untreated and treated soil, as well as
oil and grease and total organic carbon in the untreated soi1. In
addition, three different leaching tests were performed:
o TCLP - a commonly accepted procedure for measuring
Teachability of both organics and inorganics.
o ANS 16.1 - simulates leaching from the intact solidified core,
which models a condition of percolating water flow
sufficiently rapid to prevent saturation.
o MCC-1P - simulates leaching from the intact solidified core in
relatively stagnant groundwater (saturated) regimes.
These latter two tests were drawn from the nuclear industry
and modified to suit hazardous waste analysis.
Samples of untreated and treated soil were taken for the
following physical property measurements:
Moisture content
Buik density
Permeabi1ity
pH (untreated soil only)
Unconfined compressive strength (treated soil samples)
Weathering - wet/dry and freeze/thaw (for treated soil
samp!es) .
In order to obtain additional information on potential
long-term integrity, microstructural studies were performed on the
untreated and treated soils. These analyses included:
o X-ray diffractometry - to identify crystalline structures.
o Microscopy - use of scanning electron microscopy and optical
microscopy to characterize porosity, hydration products, and
fractures.
-------
1.4 RESULTS AND DISCUSSION
The following observations were made:
o The chemical analyses of the untreated soils showed the
highest PCS concentration (Aroclor 1260 - a set of known
rat os of PCB congeners) which in Sector B was 950 mg/kg,
while the maximum in Sector C was 150 mg/kg. The maximum
Va Ue f?^the treated soil was 170 mg/kg, wth an other
values 110 mg/kg or less. The untreated soil at aSple
locations B-6, B-7, and B-8 also contained large quantities of
volatile organic compounds (xylenes, chlorobenzene, and
ethy benzene) from 160 to 1,485 mg/kg total, and some heavy
!!/UV; 1 I ?°PPei:' ch™mium, and zinc -- up to 5,000
mg/kg total metals. In the treated soil, the total VOCs
ranged from 2 to 41 mg/kg and the total metals, 80 to 279
mg/kg. The large observed changes were likely due to a
combination of factors. The largest was probably the vertical
'1
SOllJCLP leachates showed PCB -concentrations
«« u'Vn,13 /ig/L' Leachates of all untreated soil
samples which had PCB concentrations below 63 mg/kg, were
below the PCB detection limit of 1.0 jig/L, and all sin
samples with PCB concentrations above 300 mg/kg showed
detectable PCB concentrations in the leachate. For the soil
samples with PCB concentrations between 63 and 300 mg/kg, some
leachate samples had detectable quantities, but others did
?iJ/.' fl }ea.chf*es of treated soil samples were below 1.0
™n io56h ? 10" 11miU US6d for a11 samPles- Seven treated
??i-t S "ates were analyzed a second time with the detection
limit reduced to 0.1 Mg/L, and four of the samples were also
?™ T?-1S Sroecti°n limit- Thus' the IWT additive appears to
immobilize PCBs, but because of the very low values of PCBs
being measured, it cannot be confirmed by this project.
The VOC concentrations in the untreated-soil TCLP leachates
ranged from 2,490 to 7,890 ,g/i. The VOC concentraUons in
the treated soil leachates ranged from 325 to 605 UQ/L This
reduction in VOC concentrations may have been due to a'
combination of factors, the main one being the reduction of
concentration due to the horizontal and vertical mixing.
The total heavy metal concentrations in the TCLP leachates
ranged for the untreated soil from 320 to 2,650 Mg/L, and for
treated soil from 120 to 210 /ig/L. As with the VOCs this
reduction in metals concentration in the leachate may have
been a result of the reduction of metals concentration in the
soils caused by the Geo-Con mixing operation.
trt
treated
sP^ial leach
soil samples,
tests, ANS 16
PCBs and VOCs
1 and MCC-1P, performed on
were not detected in any
-------
of the leachates, even though the maximum contact time for the
MCC-1P was 28 days.
The oil-and-grease and total-organic-carbon contents of the
untreated soil were each approximately 0.1 wt%, except at
sample locations B-6, B-7, and B-8, where values up to 1.5% by
wt. were measured. These results show a soil of very low
organic content, which should not interfere with the cement
hydration reactions.
The bulk density of the soil increased 21% after treatment.
The volume increase of 8.5% was small and was equivalent to a
ground rise of approximately 18.4 inches in Sector B and 14.3
inches in Sector C. This agreed with the general observations
made by the test observers.
The unconfined compressive strength (UCS) measured in both
sectors was quite satisfactory, easily meeting the EPA
guideline minimum of 50 psi. In Sector B the range was from
was from 247 to 866
higher average UCS of
average UCS of 288 psi
to the UCS difference
75 to 579 psi. In Sector C the range
psi. Overall, Sector C samples had a
536 psi compared to Sector B, with an
Two factors that may have contributed
are: the higher additive-injection rate in Sector C, and some
areas that were poorly treated because of insufficient column
overlap near sample points in Sector B. However, because the
degree of overlap of the treated soil columns in Sector C was
not measured, the second factor may not be valid.
In addition, for both sectors, the UCS appeared to increase
with depth. In Sector B, samples taken from the center of
primary columns gave the highest UCS. In Sector C, samples
from the center of the primary columns and column interface
areas gave approximately equal values of UCS, while secondary
column centers gave higher values.
The average permeability of the untreated soils was 1.8 x
10~^ cm/s, and ranged from 0.1 x 10"^ to 12 x 10"2
cm/s. Initial results obtained for the treated soil were
10"° to 10"' cm/s. Most of these values meet the EPA
guideline of 10~' cm/s for the maximum allowable value for
hazardous-waste landfill liners. However, the four-orders-of-
magnitude decrease in permeability achieved by this treatment
will cause the groundwater to flow around, not through, the
treated block.
The wet/dry weathering test results were satisfactory. They
showed very low weight losses -- 0.25 to 0.50% for the
twelve-cycle tests. The relative weight losses of test
specimens to controls were very small, averaging about 0.1%.
The freeze/thaw tests showed large losses -- up to 30.70 wt%.
However, these apparent large freeze/thaw test-
-------
specimen degradations may not affect the mobility of PCBs if
chemical bonding exists, as claimed by IWT. The weight loss
of the controls was 0.25% to 0.70%. The unconfined
compressive strengths of the wet/dry test specimens after 12
cycles of weathering were the same as for the unweathered
samples. For the freeze/thaw specimens where weight losses
exceeded 3%, the UCS values decreased dramatically,
approaching zero for some samples. Permeabilities performed
on four weathered samples with low-to-moderate weight losses
those for unweathered samples. In addition
that they can adjust the additive mix
resistant to freeze/thaw conditions. This
locations where the climate is much more
are equivalent to
IWT has indicated
formulation to be
would be done for
severe than in Florida.
The microstructural analysis, performed on each sample, showed
that the IWT process produced a dense, homogeneous mass with
low porosity. It also showed that structural variation in the
vertical and horizontal directions were absent, which
indicated that mixing was quite satisfactory. In addition,
were quite common, occurring in greater
in a Portland cement sample of typical
0.4. The presence of large amounts of
sample is a symptom of sulfate
lead to a structural failure
needles of ettringite
amounts than observed
water/cement ratio of
ettri ngi te in
attack, which
a Portland cement
can in some cases
due to expansion. The sulfate at Hialeah comes from the
HWT-20 additive and from gypsum found in the untreated soil.
It is not known whether the ettringite observed in the treated
soils will necessarily lead to expansive failures, as it is
claimed to be a part of the IWT chemical fixation technology.
The presence of ettringite may in fact, according to Professor
Perry in London (see Appendix D), aid the immobilization of
metals.
The addition of
the soil weight
addition was 0.171
dry soil in Sector
Ib/lb dry soil and
dosage of additive
additive, water, and sodium silicate increased
by an average of 32%. The average additive
Ib/lb dry soil in Sector B, and 0.193 Ib/lb
C (compared to the targeted values of 0.131
0.150 Ib, respectively). In Sector B, the
for the secondary columns was reduced
compared to the primary columns, by almost 30%
The demonstration operations lasted six days -- three days on
each sector. Operations were well organized and ran smoothly,
although some minor difficulties were encountered, including
the following:
The locations of the soil columns deviated from the
planned points, and some untreated areas between columns
exist; Geo-Con has indicated that their auger actually
creates a column slightly greater in diameter than 36
in., which would reduce the untreated areas.
-------
Automatic feed control could not be maintained.
A major water leak occurred at the drill head, precluding
the use of supplemental water for the last 21 columns.
To save time, Geo-Con was instructed to continue without
repairing the leak.
Sodium silicate was not fed uniformly and was mixed with
more soil than originally intended.
Since the feed system was designed for a 4-auger commercial
unit and given the experience gained by Geo-Con, these minor
difficulties should be readily eliminated during a large-scale
commercial operation.
The estimated remediation cost with operation of the 1-auger
machine used for the demonstration is $194/ton (SlBO/yd5).
For larger applications, using Geo-Con's 4-auger machine,
costs would be lower.
-------
SECTION 2
INTRODUCTION
2.1 BACKGROUND
Concern by the public and government is growing over using
landfills for the containment of hazardous wastes. In response to
the Superfund Amendments and Reauthorization Act of 1986 (SARA),
the Office of Research and Development (ORD) and the Office of
Solid Waste and Emergency Response (OSWER) of the Environmental
Protection Agency (EPA) have established a formal program to
accelerate the development, demonstration, and use of new or
innovative technologies. This program is called Superfund
Innovative Technology Evaluation, or SITE.
The major objective of the SITE Program is to develop reliable
cost-and-performance information on innovative alternative tech-
nologies, so that they can be adequately considered in Superfund
decisi on making. SITE demonstrations usually will be conducted at
uncontrolled hazardous-waste sites, state sites, sites under the
aegis of federal agencies, developers' sites, and commercial
instal1ations .
The two technologies evaluated were the International Waste
Technologies (IWT) in situ stabilization/solidification process,
and the Geo-Con, Inc. deep-soil-mixing equipment. It is claimed
that these technologies may be used together to create a hardened,
leach-resistant, concrete-like solidified mass. This was the
first field demonstration of an in situ stabilization/
solidification process. The demonstration to evaluate these
technologies was performed in April 1988 at a General Electric
(GE) electric service shop in Hialeah, Fla. on two 10x20-ft test
sectors known to be contaminated with polychlorinated biphenyls
(PCBs). GE is required by the local regulatory authority --
Metropolitan Dade County Environmental Resources Management
(MDCERM) -- to remediate the site for PCBs. However, the
objectives of this SITE project -- as defined in the demonstration
plan and the approved Quality Assurance Project Plan (QAPP) --
were much broader than those of GE to meet their obligations to
MDCERM. This expanded effort included three different leaching
procedures, physical and microstructural tests, and analyses for
volatile organic compounds (VOCs) and heavy metals in leachates if
these contaminants were detected in the untreated soil.
The regulatory authority (MDCERM), responsible for the Hialeah
site, mandated the tests on the two sectors for the remediation of
PCBs before the site cleanup. IWT, the technology developer, who
provided the proprietary additive, and Geo-Con, Inc. who provided
the specialized drilling equipment, were contractors to GE for the
site remediation. In addition, under a cooperative agreement with
EPA, IWT was designated as the SITE project technology developer
for the demonstration; Geo-Con, Inc. verbally agreed that its
-------
in situ procedures were to be evaluated.
GE owns an electric service shop in Hialeah, Fla., which it
operated from 1958 to 1984. The property, located in a light
industrial neighborhood, is approximately 210 x 132 ft. The
service shop is a one-story building approximately 100 x 120 ft in
plan. There is a low 60-ft-wide bay on the eastern side of the
building with masonry-block load-bearing walls. On the western
side a high 40-ft-wide bay area of steel frame was added a few
years later.
PCBs were detected in the near-surface soils in early 1984
when the shop was being closed. A cleanup effort was conducted in
March 1984 and March 1985. This involved removal of approximately
700 yd3 of soil, which eliminated more than half of the PCBs,
and replacing it with clean fill. Observations during the initial
soil-removal work, which extended down nearly 5 ft (the
approximate depth to groundwater at that time), showed staining
and apparently oily materials in the bottom of the excavation
adjacent to the concrete pad on the east side of the building.
The presence of PCB oils (primarily Aroclor 1260) was suspected.
This suspicion resulted in GE's initiating a five-phased study of
the extent of PCBs in the subsurface. This phased exploration
progressively added sampling points to define the zones where data
indicated PCB concentrations above 50 ppm. Overall, 536 soil
samples were analyzed for PCBs. A water sample from each of the
onsite and offsite monitoring wells was analyzed for PCBs and VOCs
that are on the National Pollution Discharge Elimination System
(NPDES) list.
2.2 PROGRAM OBJECTIVES
The major objectives of this SITE project for the in situ
stabilization/solidification of the soils at the GE electric
service shop, using the IWT process and the deep-soil-mixing
equipment of Geo-Con, Inc., were to determine the following:
1. Ability of the stabilization/solidification technology to
immobilize the PCBs. Extensive sampling around two of the
remaining high-concentration areas of PCB contamination was
performed. If VOCs and heavy metals were detected in the
soil, the ability of the process to immobilize these
contaminants was also to be evaluated.
2. Effectiveness, performance, and reliability of the Geo-Con
deep-soil-mixing equipment used for the in situ
solidi ficati on.
10
-------
3. Degree of soil consolidation (solidification) caused by the
chemical additives.
4. Comparative effectiveness of the stabilization/solidification
for unconsolidated sand and limestone beds at the site;
comparative effectiveness above and below the water table (5-7
ft below grade).
5. Continuing integrity of the solidified soil and immobilization
of the organics over a period of five years. (Long-term data
will be col 1ected.)
6. Costs of applying this technology to commercial-size
installations and for use at Superfund sites.
7. Viability of the technology for use at other sites.
2.3 TECHNOLOGY EVALUATION CRITERIA
The following criteria were used to evaluate the effectiveness
of the IWT in situ stabilization/solidification process used at
the 6E site:
1. The effectiveness in immobilizing the PCBs was determined from
leaching and permeability tests. Areas of high PCB
concentrations were extensively sampled, before and after
treatment, with the analysis emphasis placed on leaching
characteristics. Three Teachability tests were performed:
the Toxicity Characteristic Leaching Procedure (TCLP), which
required grinding of the solidified treated soil; and leach
tests MCC-1P and ANS 16.1, which simulated the solidified
condition that exists after soil treatment.
Permeability was measured, to indicate the rate of movement of
water through the soil mass, and thus the quantity of water
contacting the contaminants in the treated and untreated
soil. Typical unconsolidated soils have a permeability of
10~ -10~3 cm/s. A reduction to less than 10"' cm/s, a
value considered satisfactory by EPA for soil barrier liners
in landfills, would indicate the process produced a highly
impermeable soil mass.
2. The durability of the in situ solidified-soi1-mass and its
long-term endurance potential was indicated by analyzing
portions of the solidified mass to determine uniformity at the
time of sampling. However, if chemical bonding occurred, as
claimed by IWT, durability and some of the physical test
results would become less important. Information was gathered
as follows:
(a) Samples were taken at remediation column-overlap areas as
well as at column centers to identify areas of low
integrity due to poor overlap or inability to solidify.
11
-------
(b) Unconfined compressive strength (UCS) was measured. The
results provided an indication of the degree of
uniformity of the mixing of soil and additive. The
laboratory UCS data previously obtained by GE and the EPA
in 1986 and 1987 were used as a baseline.
(c) X-ray diffraction and microscopy studies were performed
to determine microstructural characteristics. These
samples provided information on treated soil porosity,
crystalline structure, and degree of mixing.
(d) Core samples taken during the borings were visually
inspected for cracks and void zones, which may cause
degradation of the monolith over many years.
(e) The weight loss during freeze/thaw and wet/dry weathering
tests provided an indication of treated soil durability.
In addition, the results for the UCS and permeability
tests performed after the 12 weathering cycles provided
additional evidence on potential durability of the
treated soil.
(f) The difference between soils above and below the water
table, and variations in properties of treated soils from
the sand and limestone layers were determined, providing
information on the range of soils that can be processed.
In addition, a long-term monitoring program, designed to
collect treated soil samples over a five-year period, will provide
more information on the durability of the treated soil.
2.4 DESCRIPTION OF OPERATIONS
Two sectors (designated B and C), each approximately 10 x 20
ft, were remediated -- one to a depth of 18 ft, and the other to a
depth of 14 ft. These depths were defined by GE to treat all the
soil containing at least 1.0 mg/kg of PCBs. The IWT additive was
injected into the soil as a slurry at a nominal rate of 0.15 Ib of
dry additive per Ib of dry contaminated soil.
The additive was air-conveyed from a supply truck to a storage
silo. It then was slurried with water at a ratio of 4:3 solids to
water in a 1,000-gal mixing tank. The slurry then was pumped to
the drill rig. Water also was fed to the drill rig at
ratio to the slurry, depending on water content of the
the bottom 3 ft of the injection column, sodium silicate was added
Ai?u9 With the additive to provide a fast-setting boundary layer.
Although the IWT process does not require any addition of sodium
silicate, this was done at the request of GE.
The Geo-Con/DSM drill rig provided the soil mixing, with the
additives injected through piping at the bottom of the mixing
a constant
soil. For
12
-------
auger. The auger contained one set of cutting blades and two sets
of mixing blades, each 1 ft apart, at the bottom of the shaft.
The additive was injected on the downstroke and mixed into the
soil, with additional mixing occurring on the upstroke.
The DSM machine was tracked into position and the horizontal
and vertical alignments checked. The elevation measurements were
made by using a small tracking wheel attached to a digital
tachometer. Machine location was verified by the use of a
stationary laser.
2.5 PROJECT ORGANIZATION
For the SITE Project demonstration, a Cooperative Agreement
was signed between EPA and IWT. In addition, 6E provided the test
site, and Geo-Con, GE's site remediation contractor, was respon-
sible for injecting the additive, provided by IWT as a dry powder.
13
-------
SECTION 3
SUMMARY OF PERFORMANCE DATA AND EVALUATION
This SITE project obtained a large amount of analytical and
operating data. It showed that PCBs, the primary contaminant at
the site, were probably immobilized and the physical properties
of the solidified mass were satisfactory. Operation of the
equipment was quite satisfactory, with only minimal and
correctable difficulties encountered. Results from the samples
collected showed an homogeneous mix with low porosity. Wide
variations in physical and chemical properties were not
observed within the limitations of the sampling and analysis
program.
The results, summarized below, provide evidence of meeting
most of the program objectives. See Section 8 for a detailed
discussion.
1. The TCLP leaching tests for treated soil samples showed no
. PCBs in the leachate for the samples analyzed to a detection
limit of 1.0 Mg/L. However, seven of the TCLP leachate
analyses were repeated to a detection limit of 0.1 ptg/L, and
four of the values were below the new detection limit. For
the untreated soil, all leachates of soils containing 300
mg/kg PCBs or more showed some PCBs, up to 13 p.g/1 (except
for one at 400 #g/L), as did some soil sample leachates for
soils containing between 63 and 300 mg/kg PCB. The maximum
treated-soil concentration for PCBs was 170 mg/kg (in Sector
B), with all except two samples at 100 mg/kg or less.
Tables 10 and 11 present these results. From the data
available, it appears that the process immobilizes PCBs, but
since the leachate results are so close to the PCB detection
limits, confirmation of PCB immobilization in this SITE
project is difficult.
2. The VOCs (xylenes, chlorobenzene, and ethyl benzene) in the
untreated soil leachates ranged from 2,490 to 7,890 p.g/1.
The total VOCs in the treated soil leachates, where the
treated soil concentration ranged from 2.4 to 41 mg/kg,
ranged from 325 to 605 ng/l. The three VOCs -- xylenes,
chlorobenzene and ethylbenzene -- had leachate
concentrations that were reduced essentially equally, from
untreated to treated soil samples. However, according to
IWT, the composition of HWT-20 used was designed only for
PCBs, not VOCs. The results are summarized in Table 12.
3. The combined metals concentrations in the untreated soil
leachate ranged from 0.32 to 2.65 mg/L. For the treated
soil, the combined leachate concentration was 0.1 to 0.2
mg/L. These results are shown in detail in Table 13.
14
-------
4. PCBs and VOCs were not detected in any of the leachates from
the special leach tests, MCC-1P and ANS 16.1.
6.
8
The oi1-and-grease
soil were very low
sample locations B
wt% were measured.
by wt. would cause
hydration reaction
wt% was too low to
reaction.
and total-organic-carbon contents of the
and were approximately 0.1 wt%, except at
6, B-7, and B-8, where values up to 1.5
An organic content of about 10% or more
significant hindrance of the cement
Thus, the organic content of up to 1.5
interfere with the cement hydration
The volume increase of the treated soil was low -- 8.5%. On
average, the bulk density increased from 1.55 g/mL (96.7
lb/ft3) to 1.88 g/mL (117.3 lb/ft3) for a soil weight
increase of 32%.
The.treated-soil permeability results were in the range of
10"° to 10"' cm/s. These values were slightly higher
than the targeted value for hazardous-waste landfill liners
of 1 x 10"' cm/s, but showed a major improvement over
untreated soils. The untreated soil averaged about 1.8 x
10"* cm/s and ranged from 0.05 x 10"^ to 24 x 10~*
cm/s. See Tables 5 to 8 for more details.
The unconfined compressive strength (UCS) was quite
satisfactory. In Sector B the values ranged from 75 to 579
psi, averaging 288 psi. The values in Sector C ranged from
247 to 866 psi, averaging 536 psi. These values indicate
that sufficient load-bearing strength will occur, since the
usual criterion on minimum strength requirements recommended
by EPA is 50 psi. See Tables 6 and 7 for more details.
In addition, for both sectors, UCS appeared to increase with
depth. In Sector B, samples taken from the center of
primary columns gave the highest UCS. In Sector C, samples
from the center of the primary columns and column interface
areas gave approximately equal values of UCS, while values
at secondary column centers were higher. The HWT-20
injection rate was higher in Sector C -- 0.193 Ib
additive/lb dry soil versus 0.171 Ib in Sector B, but this
difference probably would account for only part of the
difference in UCS seen. Possible additional factors may be
uniformity of the additive injection and lack of
correspondence between sample core locations and treated-
soil column locations.
Wet/dry weathering test results showed very low weight
losses -- 0.25 to 0.50% for the twelve-cycle tests. The
weight losses of test specimens relative to controls were
very smal1.
-- up to 30
being 6.5%.
mobi1ity if
For the freeze/thaw tests, major weight losses
7% -- were encountered, with the overall average
This degradation may not affect contaminant
chemical bonding exists as claimed by IWT. The
15
-------
weight loss of the controls was 0.25% to 0.70%. The UCS
tests performed on many of the weathered samples showed some
loss of UCS when the weight loss was above 3%. IWT has
indicated that they can adjust the additive mix formulation
to be resistant to freeze/thaw conditions. This would be
done for locations where the climate is much more severe
than in Florida.
For the three largest weight-losses (above 27%) the strength
was zero. These results are provided in detail in Tables 7
and 8.
10. The microstructural studies showed the following:
o The treated soil produced a dense, homogeneous mass with
low porosity.
o Ettringite was common to all samples and was often
present in larger amounts than observed in hydrated
Portland cement. It was the intent of IWT, in the
design of their HWT-20 additive, to produce ettringite,
thus explaining the unexpectedly high quantities seen.
o Significant differences in hydration products from
quartz-rich and calcite-rich samples were not observed,
indicating the cement-based reactions were the same in
both soil layers.
11
12
13
The two plant sectors selected for treatment were
approximately 20 x 10 ft and, based on prior sampling, were
expected to be high in PCBs. These areas were selected by
the owner for the evaluation of the remediation technology
and, concurrently, were used for the SITE Program.
Sector B contained PCBs up to 950 mg/kg (in the untreated
soil); one localized area was high in VOCs, with some heavy
metals. The maximum total VOC and total heavy-metals
concentrations measured were 1,485 and 5,000 mg/kg,
respectively. In Sector C, the maximum PCB concentration
measured was 150 mg/kg, and no significant VOCs or heavy
metals were detected.
Total addition of additive, water, and sodium silicate
increased the soil weight by 32%. The average additive
addition was 0.171 Ib additive/lb dry soil in Sector B,
0.193 Ib in Sector C, compared to the
0.131 Ib and 0.150 Ib, respectively.
of additive for the secondary columns
the primary columns almost 30%.
and
targeted values of
In S,ector B, the dosage
was reduced compared to
The demonstration lasted six days, with the operations
performed in a very satisfactory manner. However, some
difficulties did occur that may have impacted on the
stabilization/solidification process. These were as follows
16
-------
14
o The soil columns deviated from targeted locations, and
areas of untreated soil occurred. However, Geo-Con has
indicated that its auger actually creates a column of a
diameter greater than 36 in., which would reduce the
anticipated untreated area.
o Automatic feed control of additive and wastes could not
be maintained, resulting in lean and rich injection
areas. Manual control was the predominant technique
used. This difficulty was caused in part by trying to
adapt a system designed for the larger 4-auger
commercial unit to the 1-auger unit used in the
demonstration.
o A water leak at the auger head occurred, necessitating
elimination of the supplemental water for the final 21
soil columns. To save time, Geo-Con was instructed by
GE to continue without repairing the leak.
o Sodium silicate was not fed uniformly and was mixed with
more soil than originally intended.
The cost of remediation was $194/ton ($150/yd3), based on
input from IWT and Geo-Con for the 1-auger unit. For larger
applications, using Geo-Con's 4-auger machine, costs would be
lower.
17
-------
SECTION 4
PROCESS DESCRIPTION
The IWT in situ stabilization/solidification process used the
equipment of Geo-Con, Inc. for injecting the HWT-20 additive.
The additive was injected and mixed into the soil to provide a
solidified mass that was intended to immobilize the contaminants
and provide a long-lasting durable mass of low permeability. The
following sections provide descriptions of the chemistry of the
IWT additive and of the equipment of Geo-Con.
4.1 REACTION MECHANISMS
International Waste Technologies (IWT) provided the following
description of the process chemistry from their literature:
The Hazardous Waste Treatment (HWT) set of chemical fixation
or detoxification products generates a complex crystalline,
connective-network of inorganic polymers. These macromolecules
are made up of selected polyvalent inorganic elements that react
to produce branched and cross-linked polymers of sufficient
density to cause some interpenetrating polymer network (IPN)
bonding. These polymers have a high resistance to acids and
other naturally existing deteriorating factors. Structural
bonding in the polymers is mainly covalent. There is a
two-phased reaction, in which the toxic elements and compounds
are complexed first in a fast-acting reaction, and then
permanently complexed further in the building of macromolecules,
which continue to generate over a long period of time.
The first phase of the chemical fixation reaction can be
described as generating irreversible colloidal structures and
chemical reactions with toxic metals and organics via special
intercalation compounds. These are compounds that provide
interactions between organics in the soil and the silicate-based
molecules in the additive mix.
Phase Two -- the generation of the macromolecular framework
--is also a relatively irreversible colloid synthesis. However,
this is a slower-moving reaction, going from solution to gel to
crystalline three-dimensional inorganic polymer. The treated
material should be able to pass the required leaching tests
within 7 to 28 days. Of particular importance in the bonding of
hazardous elements and compounds is the chemical reaction of the
sulpho-ferri-aluminate hydrates. The bonding characteristics and
durability of structure are achieved by varying the composition
of the HWT treatment compound to suit a particular waste
situation and the desired leaching standards.
IWT is performing extensive laboratory studies to prove and
18
-------
improve the capability of various HWT additives to bond to
different contaminants. Various tests, such as Fourier transform
infrared, differential scanning calorimetry, and
thermo-gravimetric analyses are being performed.
4.2 EQUIPMENT SPECIFICATIONS
The Geo-Con/DSM system of mechanical mixing and injection
consisted of a one-auger machine with one set of cutting blades
and two sets of mixing blades attached to a vertical-drive auger.
The blade rotated at approximately 15 rpm. Two conduits were
constructed in the drive rod, and injection ports were provided at
the bottom of the shaft so that the additive slurry and liquid
(water or sodium silicate solution) could be injected into the
zone being agitated by the rotating blades. To create a vertical
column of treated soil, the blade was advanced to the desired
maximum depth of treatment. The HWT-20 additive was injected in a
slurry form, mixed into the soil and limestone as the blade
rotated during entry into the soil, and mixed again on withdrawal
from the ground. As necessary, additional cycles of injection and
mixing were made along the length of the column to provide the
required blending. Column positioning was planned to provide
sufficient overlap to avoid untreated areas (see Fig. 1). The
diameter of the treated soil column was 36 in. For larger
remediations, Geo-Con can provide a 4-auger machine, where primary
and secondary soil columns are prepared in groups of four.
The DSM mach
and verti cal ali
were made by usi
tachometer. Thi
and tracked the
was shown on a d
verified by use
of the suspended
degree.
ine was tracked into position, and the horizontal
gnments were checked. The elevation measurements
ng a small tracking wheel attached to a digital
s fixture was mounted at the top of the auger head
depth of the drill head. The tachometer output
igital display. Machine horizontal location was
of a stationary laser. The vertical positioning
auger was controlled to about one-tenth of a
A batch-m
Fig. 2), to p
specification
conveyed from
three or four
a 1,000-gal m
weight ratio
rated at 120
the nominal s
was recycled
for 15 Ib of
ixing plant was
repare and feed
s are brief.ly de
a supply truck
soi1 columns, a
ixing tank. HWT
of 4:3 to water.
gpm, then pumped
lurry requiremen
to the mix tank.
dry additive per
located in the building high bay (see
the additives. The equipment
scribed in Table 1. HWT-20 was
by air to a storage silo. To treat
measured amount of water was fed to
-20 was added to the mixfng tank at a
A Moyno positive-displacement pump,
the slurry to the drill rig. Since
ts ranged from 10-20 gpm, the excess
The HWT-20 feed rate was designed
100 Ib of dry soil .
Water was fed to the drill rig on a ratio basis to the
slurry. This ratio varied with soil moisture content. At
this was based on whether the additive injection was
Hi aleah
19
-------
above or below the water table. The final soil/HWT-20/water
slurry was targeted to contain approximately 1.6-1.7 Ib of
water/lb of HWT-20. Sodium silicate was added at the bottom 3-4
ft of each column to provide a quick-setting low-permeability
boundary-layer (this is not part of the IWT additive
technology). The manually controlled ratio of sodium silicate to
HWT-20 was about 0.05 on a dry weight basis. The sodium silicate
was provided as an approximately 40% solution.
The control system consisted of the following:
o Water flow meter (totalizer), for flow to the slurry mix ^-
tank.
o HWT-20 rotary feeder, to feed additive from the storage silo
to the mix tank.
o Magnetic flow meter, for measuring slurry flow to the drill
rig.
o Flow meters (totalizers), for water and sodium silicate
feeds.
o Control package that controls the ratio of slurry feed to the
drill-rig auger-penetration rate, and water-to-slurry feed
ratio.
Figures 3 to 7 are photographs of the process equipment.
20
-------
TABLE 1. EQUIPMENT SPECIFICATIONS
Item
Speci fi cations
1. Water supply pump
2. Water flow meter
3. Mixing tank
4. Grout pump
5. Slurry control valve
6. HWT-20 storage silo
7. Slurry flow meter
8. Sodium silicate flow meter
9. Sodium silicate feed pump
10. Sodium silicate tank
11. Reagent feeder
12. Control box
6-in. centrifugal Gorman Rupp,
Model 1682, 800 gpm
Liquid Controls Model M-7, 100
gpm
Tank - 1,000 gal
Mixer - Two-impeller, 4 blades
each, 230 rpm, 10-hp motor
Roper 7x428 6-in. rubber-lined
screw pump, 120 gpm
2-in. kni fe gate
Brooks Wafer-Magnetic, Model
7402BIWICGAA
0.1-10 gpm range
Moyno 2L4, 5 gpm
150 gal
Delta Extended-Life
Airlock-14-in. Rotary Feeder
I .S. E. Inc. - Monitors:
Flowrate, total volume, total
penetration, production rate,
pitch and roll, pressure
21
-------
Note: Shaded circles are- primary columns, and non-shaded areas
are secondary columns.
no
po
Figure 1. Overlapping column arrangement.
-------
ro
OJ
AIR
CONTROLLED
VALVES
FLOW METER-
SODIUM SILICATE BIN
SO
PUMP
MAGNETIC
FLOW METER
AIR COMPRESSOR
V |
FLOW CONTROL BOX
L_
L..
AIR
CONTROLLED j
VALVES
V
REAGENT
SILO
FLOW METER
LIGHTNING MIXER
PUMP
FLOW LINE -
CONTROL LINE-
COMMUNICATION LINE-
PUMP,
VALVE
WATER
Figure 2. Batch mixing plant
-------
BULXLIU 110$ •£
-------
Figure 4. Mixing auger on downstroke
25
-------
ro
en
-------
IQ
-S
CD
CD
oo
-s
-s
ro
O3
CL
oo
<<
to
r+
O5
-------
Figure 7. Overall view of auger assembly,
28
-------
SECTION 5
SITE CHARACTERISTICS
5.1 SITE DESCRIPTION
The GE electric service
is located
28th Street, just east of 10™ Avenue. The
on
in Hi aleah, Fl a
property is 210
x 132 ft in plan. It is approximately 7-8 ft above sea level.
The service shop consists of a one-story masonry-block building
60 x 120 ft in plan. A high bay of steel construction, 40 x 120
ft, exists on the western end of the building.
The following information was provided from reports prepared
for GE. The east-west geologic section through Dade County (see
Fig. 8) shows that the upper stratum is unconsolidated quartz
sand, typically about 5 ft thick. The Miami limestone, which
immediately .underlies the surface sand stratum, is typically 5
to 10 ft thick in this area. The Miami limestone in this
portion of Dade County has been found to be generally soft,
porous and slightly oolitic. It character- istically contains
solution channels, up to about 1 in. in diameter.
Uncemented fine quartz-sand immediately underlies the Miami
limestone. The uncemented sand zones have been encountered in
thicknesses of about 30 to 50 ft. Below this, the Fort Thompson
formation includes zones of sandstone, limestone, and
unconsol idated sands. The Fort Thompson formation is
wedge-shaped in cross section, having its maximum thickness
along the coast and pinching out along the western boundary of
Dade County. The base of the Biscayne Aquifer is a relatively
impermeable greenish-marl of the Tamiami formation. Fig. 8
shows that, in the east-west direction through the Hialeah
vicinity, the Biscayne Aquifer is wedge-shaped, having its
maximum thickness along the coast (maximum thickness of about
200 ft in the Fort Lauderdale area) and wedging out along the
western boundary of Dade County.
Permeability of the Biscayne Aquifer probably averages
between 40,000 and 70,000 gal/d/ft^. These values correspond
to 2.4-3.3 cm/s.
The general groundwater gradient in and around north Dade
County is relatively flat and slopes toward the east and
southeast. This natural slope is locally modified by canals
water-supply wells.
and
The average yearly 1owest-and-highest groundwater-1evels in
the service shop vicinity range from about 4 to 7 ft below
grade. Measurements performed for GE showed a relatively flat
29
-------
(LCVATIOM WEST
r«r. tm.
.10 T-
CO
o
IKVATIOM
rt(T. M*L
- • 0 MSL
- • -IS
_L «e
•GEHIALEAH
SERVICE SHOP
LOCATION OP CMOM-ICCTIOM A*A*
Figure 8. Miami area regional geological cross-section.
-------
water table, without a
that groundwater could
any direction.
discernible gradient. It is possible
flow, under certain conditions, in almost
Permeabilities on the order of 2.4 cm/s reflect the
influence of interconnected cavernous zones in the limestone.
Permeabilities in surface sand stratum and unconsolidated Fort
Thompson sands would most likely be in the range of 10"3 to
10"^ cm/s.
5.2 WASTE CHARACTERISTICS
Bborings and soil analyses 1n a five-phase study were
performed by GE and reported between July 6, 1984, and April 10,
1986. During this period, 536 soil samples were analyzed for
PCBs. In addition, they analyzed water samples from each of the
onsite and offsite monitoring wells for PCBs and volatile
organic compounds (VOCs). Results from onsite well-monitoring
data showed some low levels of VOCs in most of the wells.
Therefore, a few untreated-soil samples were analyzed for VOCs.
The principal site contamination is considered to be from
PCBs, although some VOCs were measured near a drainage drum next
to the east wall of the service shop. Appendix H provides PCB-
concentration profiles within the GE property line.
PCBs in the subsurface are primarily: on the east side of
the site in the area of the shop building; under the
southeastern portion of the low-bay shop building; and south of
the low bay shop building. Some shallow concentrations are
indicated west of the shop building. The maximum PCB
concentration is 5,639 mg/kg at a 1-ft depth in the northeast
corner of the site. Concentrations above 100 mg/kg occur
infrequently below the 8-ft depth level. The maximum
concentrations in each of the test sectors, as measured by LE,
were 2,150 and 435 mg/kg. EPA did not measure such high values
in the SITE Project pretreatment sampling of the test sectors.
PCBs found at the GE site are among the variety of highly
chlorinated biphenyls.
PCBs are pale-yellow viscous fluids with a mild hydrocarbon
odor. In September 1977 the National Institute for Occupational
Safety and Health (NIOSH) classified PCBs as potentially
hazardous substances . The Immediately Dangerous to Life or
Health (IDLH) level -- the maximum concentration at which 30-min
exposure will not result in any health impairment -- is 5
mg/m . PCBs are considered carcinogenic.
31
-------
SECTION 6
DEMONSTRATION PROCEDURES
6.1 SITE AND WASTE PREPARATION
EPA's principal roles for the onsite demonstration were to
conduct field sampling -- which included soil borings before and
after the test sector remediation -- and to make observations of
the operations. However, EPA did not perform the site
preparation for the SITE Project demonstration because GE had
prepared the entire site as part of their site remediation
obligation to Metropolitan Dade County Environmental Resources
(MCDERM). GE provided utilities and building
from the existing plant for the test. The sampling
was able to use the plant's utilities, except for
sanitary and communication facilities, which EPA provided during
the two sampling periods.
Management
facilities
contractor
6.2 OPERATIONAL PLAN
Geo-Con, Inc. personnel injected and mixed the HWT-20
additive with the soil as a slurry. The additive was delivered
in dry powder form to the site in trucks, and conveyed by air to
a storage bin. The slurry was prepared in a blending plant, as
described in Section 4.2, which was set up in the high-bay shop
building.
IWT, the HWT-20 supplier, provided a proposed mix design for
the treatment slurry prior to the start of work. The mix design
provided the amounts of additive, water, and other ingredients
required in each batch of treatment slurry.
During batching, the quality control inspector, provided by
GE, spot checked and documented the correct amount of
ingredients, HWT-20 additive and sodium silicate, in each
batch. The blending plant had the volume- and weight-measuring
capability necessary to permit this documentation. EPA
collected similar data concurrently.
Prior to the start of work, Geo^Con made calculations to
determine the volume of treatment slurry required to be injected
at each location in order to provide the planned ratio of
HWT-20-weight to dry-soil-weight of 0.15. During auger
withdrawal at each penetration location, the quality control
inspector observed and documented that the required quantity of
slurry has been injected. The quality control inspector also
took detailed measurements of the location of each column in
Sector B.
Two test sectors on the site were used for the in situ
treatment. The locations selected by GE and IWT were high in
32
-------
PCBs; values up to 2,000 mg/kg were expected. The planned
depths of in situ treatment in each area were 1.6 ft (0.5 m)
deeper than the maximum depth to which PCB concentrations
greater than 1.0 mg/kg had been measured. This provided for a
treatment depth in Sector B of 18.0 ft, and 14.0 ft in Sector
C. This was part of the criteria established by the local
regulatory authorities.
The drilling pattern consisted of alternate and overlapping
primary and secondary strokes. All the primary strokes were
performed first in each sector before the secondary strokes were
performed. In test Sector B, the secondary-stroke feedrates
were reduced in proportion to the untreated area remaining --
about 75% of that of the primary stroke. This variation was at
the request of GE, to determine if the use of less additive in
Sector B would have an impact on the results. In test Sector C,
both the primary and secondary strokes received the same
quantity of HWT-20. Sodium silicate was added to form a
fast-setting zone below both test sector areas to help contai
the soil-slurry additive mix, which was expected to be of low
permeability than the treated soil without sodium silicate.
n
ower
in Sector B
In Sector
. The actual location of each treated soil column
was measured and compared to the targeted locations
C column locations were not measured.
6.3 SAMPLING AND ANALYSIS ACTIVITIES
6.3.1 Samp!i nq Locati ons
EPA performed soil sampling two weeks before, and five weeks
after, test sector remediation. The sampling locations were
selected primarily to obtain information on the following
evaluation criteria:
1.
Mobility of the PCBs around areas of high concentration or
hot spots. Mobility of VOCs and heavy metals, if detected
from spot sampling.
2. Uniformity of the in situ solidified soil mass.
(a) Impact of sodium silicate on producing a more
impermeable boundary layer, compared to the body of the
solidi fi ed mass.
(b) Soil properties determined at depths above and below
the water table in order to ascertain the impact of
moisture content.
(c) Sampling in unconsolidated sand and limestone layers,
to indicate the process impact of the process on
different soils.
33
-------
Diagrams providing the planned sampling locations are shown
in Figs. 9, 10 and 11. Samples were taken at approximately the
same locations before and after the remediation.
One location in each sector was anticipated to be a hot
spot. At this location, five samples were taken at a selected
depth below the surface -- 1-2 ft in Sector B and 7-8 ft in
Sector C. A central sample was taken, along with four
additional samples about 18 in. away along a circumferential
arc.
To evaluate uniformity of the in situ solidification,
samples were taken at three or four levels: from the top layer
of unconsolidated sand, the limestone layer, and the lower
unconsolidated sand layer. For two of the locations in Sector
B, a fourth sample was taken in the boundary layer at the bottom
of the solidified mass. The sample depths with respect to the
surface were as follows:
Top unconsolidated sand
Limestone
Lower unconsolidated sand
Boundary layer - Sector B only
1-2 ft
7-8 ft
11-12 ft
16-17 ft
In Sector C, with a 14-ft treatment
ft depth) sample was in the boundary
taken during pretreatment and posttr
Appendix G.
depth, the third (11-12
„ layer. The sampling logs
posttreatment sampling are in
6.3
were
more
1.1 Determine PCB Mobility--
For the two PCB hot spots, the posttreatment samples
taken at the treatment-auger injection center, with four
samples taken up to 18 in. away within the diameter of a
treatment column. Not only did these four peripheral samples
provide additional samples for leaching tests of highly
contaminated soil, but they also indicated the treated soil
uniformity, by sampling at locations of column overlap. Section
5.5.4 describes the analyses in detail.
6.3.1.2 Determine In Situ Solidification Uniformity--
Solidified hazardous wastes are multi-phased materials whose
microstructure controls their leaching behavior and long-term
stability. Since cement setting reactions are complex, it was
important to characterize the microstructure to identify
potential durability problems.
Small-scale non-homogeneities or porosity can lead to
degradation of mechanical properties over time and possibly
allow the release of contaminants. Treated soils were
characterized by using scanning electron microscopy (SEM) and
X-ray diffraction. These techniques provided information about
porosity, uniformity, degree of mixing and mineral content of
the cured material .
Unconfined compressive strength, a test to be performed on
all treated soil samples, is a measure of the homogeneity and
34
-------
N
B
Scale
20'
Figure 9. Sampling sector locations.
35
-------
00
cri
18.0'
r-
T
r--
PRIMARY STROKE
LEGEND
O
SECONDARY STROKE
O SAMPLE LOCATION
Figure 10. Sampling locations - Sector B.
-------
CO
18.0'
PRIMARY STROKE
LEGEND
O
SECONDARY STROKE
O SAMPLE LOCATION
Figure 11. Sampling locations - Sector C.
-------
potential long-term durability of the treatment. A higher
strength indicates a more uniform soil/additive mix, which
provides an increased treatment benefit. Inability to
incorporate the organics into the physical cement microstructure
could retard the development of the unconfined compressive
strength.
In addition to sample analysis, all treated soil borings
were inspected and documented for fine cracks, void areas,
material consistency, apparent sample integrity, and
insufficiently treated areas. Any cracks, voids, or poorly
mixed areas are potential problem areas for degradation of the
solidified mass and the leaching of contaminants. However, this
1s a qualitative judgement, and 1t may take decades before
problems exist.
any
It was also of value
mixing of soil above and
existing between treated
and below the limestone)
to analyze the results of the in situ
below the water table. Any differences
limestone and the sand layers (above
were checked in the data analysis.
6.3.2 Sample Recovery Procedures
Unconsolidated soil (pretreatment), and consolidated soil
(posttreatment), samplings were performed with a rotary drilling
rig. Untreated soil samples were obtained with a split spoon
sampler. Samples were collected for - - -
tests, moisture level measurements,
measurements. Separate VOC samples
spoon, refrigerated in 40-mL vials,
on the day collected. Shelby tubes
undisturbed samples for bulk density
determinations.
chemical analyses, leaching
and grain size
were taken from the split
and sent to the laboratory
were used to collect
and permeabi1i ty
Solidified soil samples were extracted with a core barrel
assembly. Core samples of 2.875-in. diameter were used for bulk
density, compressive strength tests, permeability, and chemical
analyses. Core samples of 2.125-in. diameter were used for
wet/dry and freeze/thaw weathering tests. These multiple
samples of different diameters required the boring of at least
two holes at each sample location. Due to the loss of the
2.875-in. corer for a few days, some of the samples were taken
with the smaller corer. See Tables 7 and 8 for definitions of
samples using only 2.125-in. corer. Air cooling of the corer
was used for the first 40% of the samples collected. The
cooling procedure was then changed from air to water to
eliminate the soil dusting that existed during coring. See
Table 7 for samples taken with air cooling. All of the Sector C
samples were water-cooled.
Standard diamond coring-bits were used. All sampling
equipment was cleaned after each sample was taken to avoid cross
contamination. All sediment and rock samples were wrapped in
38
-------
aluminum foil, placed in glass jars, closed with
placed in zip-loc bags, then in cans packed with
and stored with ice packs in coolers closed with
a custody seal
vermi culi te,
a custody seal
Two duplicate sample sets were taken for each test sector
(one for each ten samples as required by the Quality Assurance
Project Plan). Duplicate samples required additional borings,
which were located as close as possible to the original
samples. They were analyzed for PCBs, all physical properties,
and leach test data. For each soil boring, logsheets describing
the core samples were prepared. For each boring, photographs of
the cores were taken to complement the logsheet descriptions.
6.3.3 Analytical Procedures
Soil samples were taken before and after the site
remediation. The purpose of the pretreatment soil analyses was
to characterize the soil and determine the contaminant levels at
specific locations; these locations also were tested after the
remediation, so that a direct comparison of the physical and
chemical properties before and after remediation could be made.
However, since the treatment process involves a high degree of
soil mixing, it is difficult to obtain comparable samples from
the same location before and after treatment. Table 2 presents
the analyses performed on the pretreatment soil samples, along
with procedures used.
The posttreatment core samples were taken at approximately
the same locations as the pretreatment samples. Their purpose
was to evaluate the changes in the soil properties and the
ability of the contaminants to migrate from the treated soil.
Table 3 gives the analyses
taken five weeks after the two
along with procedures used.
performed on
test sectors
the core samples
were remediated,
In addition, formulation tests were performed as a baseline
for the demonstration in the laboratory of the analysis
contractor. PCB-contaminated soil and cement (without HWT-20)
were blended at the same dosage rate as the IWT additive. In
addition, clean soil from the site and cement also were
formulated. Analyses performed were for moisture, bulk density,
compressive strength, permeability, and the Toxic Characteristic
Leaching Procedure (TCLP) test for PCBs.
Pretreatment analyses provided a range of important
information. The grain size, pH, moisture, and bulk density
define basic soil characteristics. Oi1-and-grease and total-
organic-carbon are both measures of organics in the soil, which
may interfere with the stabilization/solidification process.
Organics, usually above 10 wt%, interfere with many cement-based
fixation processes.
39
-------
TABLE 2.
PRETREATMENT ANALYSES
Test type No.
Grain size
PH
Moisture
Bulk density
of samp!
34
34
34
34
es
ASTM D 42
SW 9045
ASTM D 22
American
Procedure^3'13)
2-63 (reapproved in
16-80
Society of Agronomy
1972)
-
Oil and grease
Total organic carbon 34
Total metals (Sb, As, 12
Be, Cd, Cr, Cu, Pb,
Hg, Ni, Se, Ag, Th, Zn)
Total PCBs in soil 34
%
Permeability 34
Leaching
Volatile organics in
soil (32 on NPDES 1ist)
Hicrostructural
analyses^3'
Formulation test (e)
Methods of Soil Analysis - p.375
34 Standard Method 503D - American
Public Health Assn., 16th Edition
Walkley-Black
Digestion and Atomic Absorption -
See Table 5.1 and Appendix A of
QAPP, Section 6.0
SW 846 Method 8080
American Soc. of Agronomy - Methods
of Soil Analysis (in conjunction
with bulk density)
34 EP TCLP - Federal Register 11/7/86,
Vol. 51, No. 216, Appendix 1, Part
268; SW 846 Method 8080
SW 8240
Scanning Electron Microscope
and X-ray Diffraction
(f)
(a)
(b)
Where analyses showed significant metals or VOCs, leachate
analyses for those components were added.
All procedures are defined in more detail in Section 6.0 of the
QAPP.
(c) Sample locations B-l, 6, 7, 8; C-l, 9, 10, 11.
(d) Soil was taken from locations B-l and C-l for laboratory formu-
lation tests. Clean soil was taken from the site.
(e) Soil was taken from locations B-l, 6, 7, 8, 21, 22 and C-l, 9,
10, 11, 12, 13.
(f) See Section 5.4 for analyses to be performed.
40
-------
TABLE 3. POSTTREATMENT ANALYSES
Test type
No. of samples
Procedure
(a)
Moi sture
Bulk density
Unconfined compres-
sive strength
Wet/Dry weathering
test
Freeze/Thaw weather-
ing test
Unconfined compres-
sive strength after
weathering tests (test
specimen and control)
Permeability after
weathering tests
Total PCBs in soil
Permeabi1i ty
Leachi ng
43
43
45
41
38
2()(b)
42
41
42
Microstructural
analyses
40
Test Methods for Solidified Waste
Characterization (TMSWC) -Secti on 4
TMSWC - Section 2
ASTM D 2166-66 (1981)
TMSWC - Section 12
TMSWC - Section 11
ASTM D 2166-66 (1981)
Falling Head TMSWC - Section 13
SW 846 Method 8080
Falling Head - TMSWC- Section 13
EP TCLP - Federal Register 11/7/86,
Vol. 51, No. 216, Appendix 1, Part
268; Method 8080
MCC-lP-Static Leach Test (Materials
Characterization Center); Method
8080
ANS 16. 1-Multiple Extraction
(American Nuclear Society);
Method 8080
Scanning Electron Microscope
and X-ray Diffraction
(a) Samples from B-6 and B-7 were analyzed for both tests with site
water and deionized water. In Sector C, ANS 16.1 was performed
on samples from C-2,4, and MCC-1P was performed on samples from
C-1,3 using site water.
(b) For approximately half the weathering samples, UCS and
permeabilities were performed.
41
-------
Analyses for PCBs, volatile organics, and priority pollutant
metals defined the contaminants in the soils and indicated those
contaminants to include in leaching tests. The metals content
only three sampling points in each sector, since
expected to be present in significant
one of these three points, samples were taken at
depths; at another two depths; and at the third,
just one depth. VOC content was measured
in each sector — one at each sampling
was measured at
metals were not
quantities. At
three different
a sample was taken at
at only two locations
depth, plus one at the 1-2 ft level in Sector B and at the 7-8
ft level in Sector C--since VOCs also were not expected to be
present in the soil. The permeability and Teachability tests
provide data on two properties related to the mechanism of
contaminant mobility that should change dramatically with the
soil treatment.
The posttreatment analyses characterized the treated soil.
The moisture and bulk density tests provided information on soil
properties and changes in soil volume as a result of the
treatment. The unconfined compressive strength indicated a
measure of the product mix uniformity and potential soil
durability. The wet/dry and freeze/thaw weathering tests
provided indications of life expectancy of the solidified
material through moisture and temperature cycles. Unconfined
compressive strength tests were performed on the weathered test
specimens to determine if there was any loss of strength. Some
permeabilities of the weathered samples also were performed.
Permeability and Teachability are measures of the likely
mobility of the contaminants into groundwater. A comparison of
treated and untreated soil was made. Two special leach tests
were used to attempt to simulate the leaching of the material as
it would occur in the ground. They were MCC-1P, Static Leach
Test, and ANS 16.1, Multiple Extraction Leach Test, both
developed for monitoring low-level radioactive wastes for the
nuclear industry, but modified for use with hazardous wastes.
These tests were performed on samples from each of the two PCB
and VOC hot spots, using site water and deionized water in
Sector B and the PCB hot spot in Sector C. The standard TCLP
test requires grinding of the solidified mass.
6.3.4 Range of Testing
The range of variables tested to evaluate the technology was
restricted, since the Demonstration Test was part of the site
remediation process. Therefore, existing operating conditions
had to be used. However, some variability did exist at the
Hialeah site that provided valuable information. These
variables were:
1. PCB contamination level ranged from 1 to 950 mg/kg.
2. The moisture in the soil above the water table (5-7
the surface) averaged about 5.5 wt% and in the soil
the water table averaged about 18 wt%.
ft below
bel ow
42
-------
3. Soils included unconsolidated sand as well as soft, porous
1imestone.
4. Small variations in total organic carbon.
5. Ability to mix the additive with soil at various depths --
to a depth of almost 18 ft.
6. Impact on permeability, solidification integrity
soil properties of adding sodium silicate to the
for the bottom 3 ft in each sector.
and other
additive
6.4 SCREENING ANALYSIS
Prior to the demonstration, EPA performed tests on samples
previously prepared for 6E as a preliminary evaluation of the
technology before proceeding with the demonstration (see Table
4). Treated soil from the GE site was obtained from GE along
with the following information:
PCB
Desig- Concentration
nation (pom)(c)
SL-21 1,130
SL-24 5,628
TG-17 5,628
HWT-20
Additive rate
(Ibs/lb drv soil)
0.15
0.20
0.12
Moi sture
condition(a)
Formulati on
date(b)
Above
Bel ow
Bel ow
12/86
12/86
8/86
(a) Simulating above, or below, water table.
(b) The 12/86 samples were prepared by mixing soil with a slurry
additive, while for the 8/86 sample, the additive was mixed
with the dry soil, -then water was added.
(c) Values reported by GE.
TABLE 4. SCREENING SAMPLES RESULTS
Parameter
SL-21
SL-24 TG-17
Moisture, wt%
Bulk density, g/mL
Permeability, 10 cm/s
Unconfined compressive strength, psi ..
9.84
1 .97
3.6
876
16.09
1 .67
4.2
418
8.59
1 .86
0.44
1185
Leachabilitv - PCB concentration
PCB in solid being leached, mg/kg
TCLP, Mg/L
MCC-1P, jig/L
4,100
ND
ND
4,900
ND
ND
5,700
ND
ND
ND
Not detected. Detection limit, 1 M9/L
43
-------
EPA results summarized in Table 4 show low permeability,
approximately 10 a cm/s; high unconfined compressive strengths,
418 to 1,185 psi; and immobilization of the PCBs.
In addition, a microstructural analysis was performed. The
samples examined were porous and incompletely hydrated. These
results indicated a potential for durability problems. These
results will be compared to those of the demonstration in Section
/ •
6.5 PHYSICAL TESTS
The physical tests described below were used to analyze the
soil and leachate samples during this SITE project.
ASTM D 422-63: Grain Size Analysis
This method covers the quantitative determination of the
distribution of particle sizes in soils. The distribution of
particle sizes larger than 75 microns (retained on the No. 200
sieve) was determined by sieving, while the distribution of
particle sizes smaller than 75 microns was determined by a
sedimentation process using a hydrometer to secure the necessary
data.
EPA-600/4-79-020: Methods for Chemical Analysis of Water and
Wastes
EPA Method 600 was used to determine the water content of
untreated soil samples. Moisture was determined by measuring the
mass of water removed by drying the sample to a constant mass at
103°-105°C.
TMSWC-4: Water Content (Moisture) - Solid Cores
A 50-g sample was ground to pass an ASTM No.
mass of the sample was measured before and after
maintained at 60° +3°C. The dry weight must be
weight (mass change of less than 0.03 g in 4 h)
10 sieve. The
dryi ng in an oven
a constant
The wet sample
mass was divided into the difference of the wet sample mass minus
the dry sample mass.
Bulk Density
Bulk density was determined in the demonstration using the
Core Method described in Methods of Soil Analysis, American
Society of Agronomy, 1965. The mass of the samples was calculated
by difference, using a top-loading balance. The dimensions of the
specimen (cube or cylinder) were measured using a 30-cm ruler
having a precision of ±.1 mm. The bulk density was calculated by
dividing the volume into the mass.
44
-------
ASTH D 2434: Permeability Coefficient-Constant Head
Permeability coefficient was determined by a constant head
method for determining the laminar flow rate of water through
granular soils. This procedure was limited to disturbed granular
soils containing not more than 10% soil passing the 75 /zm (No. 200
sieve). For the demonstration measurements were made on minimally
disturbed soil samples collected in Shelby Tubes (when the bulk
density samples were taken.)
THSWC-13; Permeability Coefficient-Falling Head-Solid Cores
This test was carried out in the demonstration on the solidified
7- and 28-day core samples. A cylindrical sample 7.62x7.62 cm was
used. Permeability was determined using a triaxial cell measuring
changes of water volume over time under controlled conditions of
temperature and pressure.
ASTM D-1633: Unconfined Compressive Strength Test
This test method covers the determination of the unconfined
compressive strength of molded soil-cement cylinders using
strain-controlled application of the axial load.
TMSWC-12: Wet/Dry Weathering Test
This test, which provides indications of short-term durability
of the solidified mass, was performed in the demonstration using two
4.5x7.4-cm cylindrical core specimens of solidified wastes. It was
carried out in conjunction with TMSWC Method 4.0, Water Content.
One of the specimens was used as the test specimen, the other as the
control.
Two solidified test samples were compared by weight difference.
One sample, the control, was placed in a humidity chamber, and the
other was dried in an oven at 60°-65°C for 24 ±1 h. The sample
specimen then was cooled in a desiccator, and 230 ml of water was
added to each sample. The sample and control then were placed in
the humidity chamber for 24 h. This was repeated 11 times. The
weight loss of test specimen and control and the relative weight
loss were then calculated for each cycle.
TMSWC-11
Freeze/Thaw Test
This test, which provides information on short-term durability
of the solidified mass, is a more severe cycling than would occur in
nature and was performed using two 4.5x7.4-cm cylindrical core
specimens of solidified waste. The test was carried out in
conjunction with the water content determination. One of the two
specimens was used as a control. The test specimen was placed in a
freezer at -20 i3°C for 24 ±1 h. Water was then added to the
frozen specimen and control and maintained at 22 i3°C for 24 +.1 h.
45
-------
The process was repeated 11 additional times, with relative weight
loss calculated after each cycle.
Modified Bulk Density
Eight untreated soil samples, which could not be collected in
Shelby Tubes, were analyzed as described in the following
paragraphs.
Non-pourable samples were placed into a tube of known
dimensions and vibrated until a light film of water surfaced, at
which time weight and height measurements were taken. The density
was calculated using the core method.
Pourable samples were poured into a 100-ml graduated cylinder
of known weight. The graduate containing the sample was weighed.
The density was calculated using the formula,
D - W/V
where D is the density, W is the weight of the sample, and V is
the volume.
6.6 CHEMICAL TESTS
%
The chemical tests and definitions of the
described below were used to analyze the soil
during this SITE project.
Polvchlorinated Biohenvls (PCBs)
contami nants
and leachate
as
samples
Polychlorinated biphenyls are a group of related isomers of
chlorinated organic compounds characterized by having 1 to 10
chlorine atoms substituted on the biphenyl group.
Priority Pollutant Metals
Thirteen priority pollutant metals have been specified to be
of particular environmental concern by the U.S. Environmental
Protection Agency. The metals are: antimony, arsenic, beryllium,
cadmium, chromium, copper, lead, mercury, nickel, selenium,
silver, thallium, and zinc. Four of these metals were found in
measurable concentrations in contaminated soils from the Hialeah
site -- chromium, copper, lead, and zinc. To conserve resources,
only the four metals found at the site were analyzed for in many
of the samples tested.
Volatile Organic Compounds fVOCs)
Volatile organic compounds were determined by purging
volatiles from the samples tested. Compounds measured at Hialeah
were xylenes, chlorobenzene, and ethyl benzene.
46
-------
Toxicitv Characteristic Leaching Procedure (TCLP1
The TCLP was designed to determine the mobility of both
organic and inorganic contaminants present in liquid, solid, and
multiphase wastes. For wastes comprised of solids, the particle
size of the waste was reduced and analytes are extracted for 18
hours with an acetic acid solution. Two liters of extractant
(fluid no. 2) were used for 100 g of solid. The extract was then
separated from the solid phase and analyzed for PCBs, Priority
Pollutant Metals, and VOCs. This procedure was developed to
measure a wider variety of contaminants, including volatile
organics, than is measured by EP Toxicity.
MCC-1P: Modified Static Leach Test
The static leach test establishes the maximum concentrations
of elements in a quasi-static groundwater regime that has been in
contact with a stabilized waste. The samples are kept as solid
cores to simulate an in situ condition. For the SITE
demonstration, cylinders cured for at least 28 days were used and
were taken from the core barrel drilling. Four test specimens for
each test were leached with organic-free ultra pure water, at
40°C, for four varying time periods up to 28 days. Leachates
then were analyzed for all contaminants.
ANS 16.1: Leach Test
The intact samples for the demonstration, cut from the solid
cores, were leached, using ultra pure water. The sample specimen
was placed in fresh leachates at five different time intervals,
with the total leaching time being 28 days. (This differs from
the TCLP and the MCC-1P, where each of four specimens for MCC-1P
is placed in water once and held there for varying time frames up
to 28 days.) Therefore, five leachates were analyzed for the
organic and inorganic contaminants.
SW846 Method 9045; Soil oH
The pH of a sample was determined in the demonstration
electrometrically using either a glass electrode in combination
with a reference potential or a combination electrode. In soil
samples, pH was determined by preparing a slurry, using equal
volumes of soil and deionized water and measuring the pH of the
decanted liquid.
APHA 503D: Oil and Grease
Method 503D is a modification. of the Soxhlet extraction
method, which is suitable for sludges. Magnesium sulfate
monohydrate was combined with the sludge to remove water (as
MgS04 x 7H20). After drying, the oil and grease was extracted
in a Soxhlet apparatus with trichlorofluoromethane and after
solvent evaporation was measured gravimetrically.
47
-------
Total Organic Carbon - Waiklev-BIack Method
Oxidizable matter in a soil sample was oxidized by the
chromate ion (Cr207~^), and the reaction was facilitated by
the heat generated when two volumes of sulfuric acid were mixed
with one volume of potassium dichromate (K2Cr207). The
excess chromate was determined by titration with ferrous sulfate,
and the quantity of substances oxidized was calculated from the
amount of Cr207"^ reduced.
SW846 Method 3510; Liquid-Liquid Extraction
Method 3510 is a procedure for isolating organic compounds
from aqueous samples. A measured volume of sample was serially,
extracted with methylene chloride using a separatory funnel. The
extract was dried, concentrated, and, as necessary, exchanged into
another solvent compatible with the cleanup or determinative step
to be used.
SH846 Method 5030: Purae-and-Trao
Method 5030 describes sample preparation and extraction for
the analysis of volatile organics by a purge-and-trap procedure.
An inert gas was bubbled through the aqueous sample at ambient
temperature, and the volatile compounds were transferred from the
aqueous to the vapor phase. The vapor was swept through a sorbent
column where the volatile components were adsorbed. After purging
was completed, the sorbent column was heated and back-flushed with
inert gas to desorb the components onto a gas chromatographic
column.
EPA Method 8240:
Volatile Proanics
Gas Chromatographv/Mass Spectrometrv (GC/MS for
Method 8240 i
concentration of
waste matrices.
organic compounds
are insoluble or
low-molecular-wei
nitriles, ketones
volatile compound
to the purge-and-
spectrometer.
s a GC/MS procedure used to determine the
volatile organic compounds in a variety of solid
Method 8240 can be used to quantify most volatile
that have boiling points below 200°C and that
slightly soluble in water. These include
ght halogenated hydrocarbons, aromatics,
, acetates, acrylates, ethers, and sulfides. The
s were introduced into the GC by a method similar
trap method; detection was by a mass
SW846 Method 8080: GC/ECD for PCBs
Method 8080 was for the gas chromatographic analysis of PCBs.
Prior to analysis, samples were subject to appropriate extraction
procedures. Samples were injected into the GC using the solvent
flush technique. Compounds in the GC effluent were detected by an
electron capture detector (ECD).
48
-------
SW846 Method 680: GC/MS for PCBs
Method 680 is used to determine pesticides and polychlorinated
biphenyls (PCBs) in waters, soils, and sediments by gas
chromatography/mass spectrometry (GC/MS). It is applicable to
samples containing single congeners or to samples containing
complex mixtures, such as Aroclors. Polychlorinated biphenyls
(PCBs) are identified and measured as isomer groups by levels of
chlorination.
SW846 Method 3050: Acid Digestion for Metals
Method 3050 is an acid digestion procedure used to prepare
sediments, sludges, and soil samples for analysis by flame or
furnace Atomic Absorption (AA) spectroscopy. A representative
sample is digested with HNOo and HoOo. The digestate is
then refluxed with either HN03 or HCT.
SW846 Method 3010: Acid Digestion for Metals
Method 3010 is a digestion procedure used to prepare samples
for analysis by flame Atomic Absorption. The sample was mixed
with HN03 and allowed to reflux in a covered Griffin beaker,
followed by refluxing with HC1.
SW846 Methods 7060/7740: Furnace AA
Methods 7060 and 7740 are graphite furnace atomic absorption
techniques approved for determination of arsenic and selenium.
Following sample digestion, an aliquot of sample was placed in a
graphite tube in the furnace, evaporated to dryness, charred, and
atomized. The metal atoms to be measured were placed in the light
path of an atomic spectrophotometer.
SW846 Methods 7470/7471: Mercury by Cold Vapor Atomic Absorption
(CVAA)
Method 7470 is a cold-vapor atomic-absorption procedure for
determining the concentration of mercury in mobility-procedure
extractions. Method 7471 is prescribed for solid and sludge-type
wastes. Sample preparation is specified in each method. Following
dissolution, mercury in the sample was reduced to the elemental
state and aerated from solution in a closed system. The mercury
vapor passed through a cell positioned in the light path of an
atomic absorption spectrophotometer.
SW846 Methods 7040/7090/7130/7190/7210/7420/7520/7760/7840/7950
These methods are used
cadmium, chromium, copper,
zinc. The method of analysis is
absorption spectroscopy, where a
in a flame. A light beam from a
to analyze antimony, beryllium,
lead, nickel, silver, thallium, and
direct aspiration atomic
sample is aspirated and atomized
cathode lamp whose cathode was
49
-------
made of the element to be determined was directed through the
flame into a monochrometer and onto a detector that measured the
amount of light absorbed. Since the wavelength of the light beam
is characteristic of only the metal being determined, the light
energy absorbed by the flame is a measure of the concentration in
the sample.
50
-------
SECTION 7
FIELD ACTIVITIES
7.1 OPERATIONAL HISTORY
The sampling contractor arrived at the electric service shop on
March 17, 1988, to start preparations for the pretreatment
sampling. This included the following:
o Staking out the two test sectors
o Locating and flagging each sample location
o Constructing an equipment decontamination area
o Setting up the sample preparation area
o Providing for personnel facilities - clean office, office
furniture, sanitary facilities
o Setting up health and safety facilities
o Purchasing necessary materials for decontamination, health
and safety, and sample shipping.
The drilling crew arrived on site on Monday, March 21, and the
first samples were collected the next day. From March 22-26, 17
sets of samples were taken from each test sector and sent to the
laboratory for analyses.
Geo-Con, Inc. arrived on site on March 28, 1988, and spent two
weeks setting up their equipment. This included the mixing plant,
crane with drilling auger, a template the size of each test sector
for locating each column of treated soil, control instrumentation,
and decontamination facilities. The equipment then was tested by
treating a preliminary column outside the test sectors. A calibra-
tion curve of slurry concentration versus solids concentration was
prepared for the HWT-20 additive.
The remediation of the two sectors, which required the
production of 36 columns of treated soil in each, started on April
11. Each sector took three days to treat, and the remediation was
completed on April 16. Sector B was treated first. For Sector B
only, GE's quality assurance officer took measurements to determine
the actual location of each column of treated soil (see Fig. 12).
Geo-Con, GE, and EPA recorded operational data and observations
separately.
The operations started with the preparation of the additive
slurry batch. A known volume of water was added to the slurry
preparation tank, and then a predetermined quantity of HWT-20 was
added by a calibrated rotary valve. When a uniform slurry was
attained, usually in less than five minutes, a sample was taken and
the specific gravity was measured on a mud balance. The target
value was 1.51 g/mL for a 4/3 weight ratio of additive to water.
If the value deviated, either water or additive was injected to
51
-------
10.47-
837-
6.48-
o
I
3.99-
cn
ro
1.50-
0-
\ \ I I I I I I I I I I ! I I . I III
0 1.50 2.62 3.75 4.88 6.00 7.12 8.25 9.38 10.50 11.62 12.75 13.88 15.00 16.12 17.25 18.88 19.50 20.62 22.12
DISTANCE (FT.)
LEGEND
+ PLANNED COLUMN CENTERS
• FIELD MEASURED COLUMN CENTERS
WITH CIRCUMSCRIBED 3-FOOT DIAMETER
COLUMN AREA BY LE.
x 27/8" CORES
© 2V8" CORES
Figure 12. Location of soil columns in Sector B.
-------
bring the slurry density to the targeted value. The slurry tank
was sufficiently large to prepare enough slurry to treat three or
four soil columns with each batch.
The operator then pumped
preset and controlled ratio
also was fed to the auger on
rate. Two water ratios were
penetration was above or bel
on the downstroke, and addit
auger withdrawal. Typically
locating of the crane and al
mi nutes.
the slurry to the mixing auger at a
to the soil penetration rate. Water
a ratio controlled by the slurry
used, depending upon whether auger
ow the water table. Slurry was added
ional mixing with the soil occurred on
, each column of treated soil -- with
igning the auger -- took thirty
In Sector B all primary strokes were performed before any
secondary strokes. Thus, Geo-Con was able to auger the secondary
columns 24-48 hours after the primary columns started to set. In
bector C, primary and secondary columns were performed alternately
due to logistical difficulties in moving the crane.
The soil was treated to a depth of 18 ft in Sector B and
in Sector C. For the bottom 3 ft, sodium silicate was added
provide a quick setting, more impervious treated-soi1-mass.
14
to
ft
The sampling contractor returned to the site on May 16, 1988
-- approximately thirty days after the remediation. Three days
were required for site preparation, similar to that performed for
the pretreatment sampling. Sampling started on May 19 and was com
pleted on June 3. Samples were collected from nineteen locations
in Sector B and eighteen locations in Sector C. This is three
more locations than originally intended, and these samples were
added because they are close to high PCB concentration samples
previously collected. All SITE project work at the electric
service shop was completed by June 6, 1988.
For the 6-day demonstration plus mobilization/demobilization,
Geo-Con utilized eleven people, as follows:
1
1
1
1
2
1
1
1
1
1
overall coordinator
construction manager
control-panel and operations-control supervisor
outside operation supervisor
crane operator and helper
operator at auger
operator at slurry feed system
hydraulic power-pack mechanic
electrician
health and safety officer
A detailed log of the six days of operation is described in
Volume II, Appendices A and B. The log provides operating data
53
-------
collected from the control panel by EPA, and GE and records many
additional observations. The summary report prepared by Geo-Con
for GE is provided in Volume II, Appendix C.
7.2 UNIT PROBLEMS AND DEVIATIONS FROM DEMONSTRATION PLAN
7.2.1 Operations
Equipment operations during the demonstration were quite
satisfactory. However, some minor problems were encountered,
along with deviations from the plan. These operational deviations
are as follows:
o Sector B was to contain 20 primary columns and 16 secondary
columns, with the end columns of each of 4 rows of 9 being
primary (see Fig. 10). The deviation was that 2 rows started
with secondary columns (Rows 1 and 3 from the north end), and
thus there were 18 columns of each type. This deviation
should not have a significant impact on the results.
o The preferred procedure described by Geo-Con for treating soil
columns was to drill all the primary columns (strokes) in each
sector first before augering the secondary columns. In Sector
C this did not occur, as most secondary columns were done
after their corresponding primary column. Practical
difficulties of moving the crane and aligning the auger
necessitated a reduction in the number of times the crane
would be moved. A primary and a secondary column could be
mixed without relocating the crane. The impact of this
deviation was not detected and was not expected to be of
significance. The added curing time of the primary columns in
Sector B (compared to those in Sector C) did not impact
Geo-Con's ability to auger the secondary columns.
o Automatic control of the slurry and supplemental water feed
rates could not be maintained, and" some of these operations
had to be performed manually. Manual operation was used for
most of Sector C (the second one done), which resulted in an
overall reduction in the ability to control flow and in uneven
feed additions per foot of depth. This problem resulted in
part from the oversized design of the feed system, which was
sized for the larger, 4-auger commercial unit. In some
instances, slurry and water were not added for a depth of 6 to
9 in., then would be compensated for in the next 6 to 12 in.
Lesser variations were not uncommon. This would leave
significant variations in quantity of additive per foot of
penetration. However, the total slurry added was close to the
targeted value. Although auger mixing was good and tended to
blend the soil along the vertical column, this deviation may
have an impact on the solidified mass and samples collected.
This impact may be greater than the small deviations in slurry
density (solids concentration in additive slurry). A
54
-------
detrimental impact, due to the uneven HWT-20 feed rate was
not noted in the analytical results.
The sodium silicate was not added uniformly over the final 3
teet of depth. To compensate for this and for the incomplete
mixing in the bottom region, the auger was cycled one
additiona time over a 3 1/2- to 4-ft range before
withdrawal. This mixing penetrated the soil an additional 6
in., thereby spreading the two additives through more soil.
u ?v?r«ie !od,ium silicate feed to the bottom 3 1/2 ft was
about 4.2% of dry HWT-20 additive. This addition is low
compared to the target value of 5% for the bottom 3 ft.
Some difficulties were encountered in starting the auger pene-
tration in Sector C, particularly with the secondary columns
J™SRW!f d*e+J? large stones in the
top 5 ft of this sector. This added
columns' augering times, with Column
start. Some very minor difficulties
to the harder limestone layer, were
comprising the
5-10 min to many of the
66 taking about 40 min to
in auger penetration, due
encountered in Sector B.
A major water leak developed at the auger head where water
enters. Therefore, except when sodium silicate was added,
supplemental water was turned off for the last 21 soil
columns, on the instructions of GE, to minimize time loss.
However, based on laboratory results, the moisture content
these solidified soil columns appears to be approximately
equal to that of the earlier columns in this sector when
additional water was added.
the
for
Location of the auger head deviated from the target point by
many inches in many cases. The exact locations of the auger
or column centers in Sector B were measured, and a number of
untreated soil areas are apparent. Geo-Con's drilling plan
precluded untreated areas due to the overlapping of properly
situated columns. Geo-Con claims the actual mixing zone is
slightly larger than the 36-in.-diameter columns shown in Fig.
i<>, and thus the size of the poor overlap areas would be
reduced. This difficulty should be overcome with the use of
t,eo-Con s multi-augered machine, which should provide improved
penetration control. Samples at locations B-10, 11, 12, and
13 bordered an untreated area (unsatisfactory overlap of soil
columns); as the soil was loose or weakly bonded, treated soil
cores could not be collected. Therefore, additional samples
at locations B-21, 22, 23, and 24 were collected at a soil
column center about 1.5 ft to the northwest. The auger head
also deviated from the target locations in Sector C, but no
measurements of the column locations were made, and void areas
were not noted during the core sampling.
The nature of the auger operation, injecting additive on the
55
-------
downstroke with additional mixing on the upstroke, causes
vertical soil blending. In addition, grout from one column
overflowed into others. Therefore, both vertical and
horizontal soil blending occurred, making it impossible (as
predicted) to maintain the integrity of local contaminant hot
spots for physical and chemical analyses.
o The summary sheets prepared by Geo-Con used slurry densities
about 5% higher than actually measured by GE. Thus, using the
Geo-Con numbers, the HWT-20 usage would be even greater than
calculated in Section 7.4. The correct numbers, based upon GE
provided density values, were used for the material balance in
Table 14.
o A few other minor operating difficulties were encountered, but
these caused only momentary delays. Many of these upsets are
recorded in Appendix A.
'.*•.-
7.2.2 Sampling and Analysis
Some deviations also occurred during the sampling and analysis
work. They are as follows:
Based upon
operations
collected,
additional
preliminary untreated soil analyses and the
log, some additional posttreatment samples were
two in Sector B and one in Sector C. The two
samples in Sector B were taken in areas of high PCB
concentration near poor column overlap areas to obtain
additional data based upon both characteristics. The
additional sample in Sector C was taken to obtain an
additional soi1-column-interface sample.
In the lower unconsolidated soil layer, Shelby tube samples of
untreated soils (for bulk density and permeability tests)
could not be collected, due to the fluidity of the soil.
Split spoon samples were collected, and a modified bulk
density test, as described in Section 5.7, was performed.
After approximately 40% of the posttreatment samples were
collected, coring operations were changed from air cooling to
water cooling. Air cooling had been attempted instead of
water cooling (the more commonly used method), to avoid
possible leaching of some contaminants by the cooling water.
However, after switching to cooling water, analysis of the
water, which is recycled, indicated only very minor losses in
PCBs, based on the core mass and PCB concentration in the
water. The losses would provide a small error, well below the
limits of accuracy of Method 8080 for soil analysis. This
change eliminated PCB contamination of the immediate area by
the cooling air and increased the coring rate, which had been
extremely slow (1/4 to 1/2 in./min). The change did not
appear to impact the analytical results. However, the benefit
was that the integrity of the sample cores improved, with less
apparent loss in core sample material.
56
-------
nc« 5a11;n?^He?J Permeabil^y test, the test samples were
presoaked outside the apparatus to start the saturation
process. Then pressure higher than specified in the test
procedures was used in the triaxial-cell test unit. Both
presoaking and high pressures were used to reduce the time
required for saturation, which had been taking many weeks.
The impact on the results should be negligible.
The diameter of the cores for the variously treated soil
samples deviated from those defined in the analytical
procedures. The sizes selected, 2.875 instead of 3.0 in., and
*.1Z5 instead of 1.77 in., were used because they were the
closest sizes available to the driller. The use of 2.875-in.
cores for the UCS tests did not cause any errors since the
length-to-diameter ratio of 2.0 was maintained. The 2.125-in
cores for the weathering tests may have had a more significant
impact, although the general trend of the results would not
change.
For the wet/dry weathering tests, a convective drying oven was
used instead of a vacuum oven. This increased the drying
temperature (60°C) by about 2°C, which should not affect
the results obtained.
57
-------
SECTION 8
PERFORMANCE DATA AND EVALUATION
8.1 PHYSICAL TESTS
The results of the physical tests, the details of which are in
Appendices E and F, are summarized in Tables 5 through 8 (at the
end of this section). The highlights of the results, with
discussion, are as follows:
8.1.1
Moi sture
The moisture content of the untreated soil varied with depth,
depending on whether the sample was -taken above or below the water
table. At a 1-2-ft depth, the moisture content averaged about 5.5
wt%. Below the water table, it averaged about 18 wt%. There was
no definitive trend between moisture content and depth below the
water table, or between moisture content in the two sectors
(moisture content does affect the strength of cement). The
average free-moisture content of the treated soil was 18.1 wt% and
was the same in both sectors at all depths. The range of
individual values was 12 wt% to 26 wt%. There was also no
apparent relationship between the soil moisture (treated and
untreated) and any of the other physical properties, such as
unconfined compressive strength (UCS) or bulk density.
8.1.2
The pH in water of the soil was
soil sample. The values were quite
value in S.ector B being 8.0, and in
with depth were seen.
measured for each untreated
consistent, with the average
Sector C, 8.5. No variations
8.1.3 Particle Size Distribution
The particle size distribution
showed for Sector B that 30 wt%-40
mesh (250 urn). About 1 wt%-3 wt%
is less than 200 mesh (74 fim), as
limestone layer. Data on the soil
(depth greater than 10 ft) is very
substantial fraction less than 60
the soil is in a proper size range
and should not have any impact on
for the untreated soil samples
wt% of the soil is less than 60
of the soil in the sand layers
is about 5 wt%-10 wt% in the
below the limestone layer
limited, but appears to have a
mesh. This data indicates that
for the preparation of concrete
the results.
8.1.4 Total Organic Carbon (TOC1 and Oil and Grease (O&G)
The TOC and O&G concentrations in the soil samples were very
low. Host of the O&G values were below the detection limit of
58
-------
and
in Sector C, and the TOC values were
s of 1,000-3,000 mg/kg. At sample locations
oily drainage drum in Sector B, values of TOC
, UP to 16,000 mg/kg for TOC and 1.6 wt% for
tn addition, it appears that at the higher TOC
uov^ =io« 4-u Tfns>4.hl9ner PCB concentrations were measured.
were also the locations where samples high in volatile
organic compounds (VOCs) were detected. For nlar y all the
samples TOC values were larger than 0&6 values, as would be
expected since O&G is the solvent extractable portion of TOC. It
is usually expected that organic contents above 10% bv wt mav
interfere with the hydration reactions of cement and other *
pozzolans Therefore, at the demonstration site, organics would
not have interfered with the cement hydration reactions.
8.1.5 Bulk Density
B
lndiia
landfill
?wer ^
i HUtShd
had the
to 1.67
--a
age-1 densi'ty of the untreated soil in Sector
g/mL 1n the tQP sand
ayer- In Sector c' th* top
1° rep ace contaminated soil taken to a secure
largest average bulk density of 1.65 g/mL
l'** 9/tnL for the Tinstone and 1 57
Which is a 11ttle IGSS than the Sector B
arage overal! bulk density of all the samples was
dent
depth
with a range of individual values from 1.21 to 1
85
fom
from
Thi
lone
zone
fn8?*8?"80!1 ^]k Density became greater with increased
In Sector B, the top layer averaged 1.76 g/mL, and the
J/T" W3,S !;97 9/mL' In Sect°r C' the values ranged
g/mL in the top layer to 2.00 g/mL in the lower layer,
PPr°Jimai"ly the same as s°lidified Portland cement.
mnpH ^H$ +Um Sllicate *° the additive mix in the bottom
sampled did not appear to have an impact on the bulk
th™ Jh8 ^iability of the results for the treated soil was
fv of ?h he- ""treated soil. On average, the overall bulk
ty of the soil increased by 21% with treatment. The bulk
ties obtained during the screening tests, described in
on 6.4, were about the same as for the field samples.
dPni
densi
densi
Secti
f *-e °V(rrall material balance in Section 8.4 showed
thP TWT H- Weighl increase of the soil after the addition of
the IWT additive, water, and sodium silicate averaged 31.7%.
Iherefore, the volume increase was about 8.5%, which for Sector B
was equivalent to 18 in. and for Sector C, 14 in of ground rise
!EvplSofmiLTS02able beca"se the field observations showed the '
level of Sector B rose 1 1/2 to 2 ft, and Sector C rose about 1 to
8.1.6 Permeabil i ty
The average permeability of the untreated soil in both
59
-------
sectors, at all depths, was approximately 1.8 x 10"^ cm/s. The
values for the limestone layer (5-10 ft depth) appeared slightly
greater than for the two adjacent sandy layers. The permeability
values were quite scattered, covering a range 0.1 x 10"^ to 12 x
10"z cm/s. For most of the samples in the lower sandy soil
layer, a modified bulk density (see Section 6.5 for procedure) was
obtained. Samples from the lower layer showed the lowest
permeabilities, but this may have been due to the procedural
change.
The treated soil permeabilities ranged from 10"6 to 10"7
cm/s. No discernible differences were noted between samples above
and below the water table. A guideline for satisfactory
permeability used by EPA is 10"' cm/sec. This is the maximum
allowable value for hazardous-waste landfill liners, as suggested
by EPA. However, the achieved four-orders-of-magnitude decrease
in permeability by the treatment will cause the groundwater to
flow around, not through, the treated soil.
The permeabilities performed on samples in the sodium silicate
layer, at sample locations B-24, C-8, and C-14, were similar to
values obtained without the sodium silicate. Permeabilities for
the weathering tests are discussed in Section 8.1.8.
8.1.7 Unconfined Compressive Strength (UCS)
The unconfined compressive strength (UCS) of all the samples
was satisfactory, easily meeting the EPA's minimum guideline of 50
psi, which provides for a high load-bearing strength. Those
collected in Sector B averaged 290 psi, ranging from 75 psi at
B-19 to 579 psi at sample location B-23. The average of the
unconfined compressive strengths in Sector C was 536 psi, with the
individual values ranging from 247 psi at sample location C-15 to
866 psi at sample location C-l. The average HWT-20 additive
dosage rate in Sector B was 0.171 Ib HWT-20/lb dry soil, and for
Sector C, 0.193 Ib HWT-20/lb dry soil.
For both sectors, the UCS appeared to increase with depth.
For Sector B, it ranged at increasing depth from 249 to 410 psi,
and for Sector C, from 420 to 688 psi. The highest averaged value
in both sectors was in the region where sodium silicate was added.
In Sector B, it appeared that the samples taken from the
center of the primary columns produced the highest average UCS --
420 psi. For the center of the secondary columns,
slurry flow was reduced by almost 30%, the average
185 psi. The average UCS of the interface samples
degree of column overlap varied, was 210 psi.
where the
value was about
where the
In Sector C, where the HWT-20 addition for the primary and
secondary columns was equal (at approximately 0.193 Ib.HWT-20/lb
dry soil), the UCS values in the primary column centers and at the
interfaces were about equal. This was not unexpected, since the
60
-------
the
faceDon e
iir* JI2 ?i,P nts ^here treated material overlapped. However t
UCS for the secondary column centers appeared tb be highlr than
of
s1tjf or "sture content. The
wlth the
8.1.8 Weathering
F the wet/dry weathering tests, which showed very
it losses (0.25% to 0.50% for both the test
^controls), were quite good. The weight losses
respective controls, usulny"by 1 esf thanTl'I^ "" f°r th61>
incc f of the freeze/thaw tests showed very dramatic
losses for the test specimens, while the weiaht Toll nf th*
^^olLr!mai^ irlUlVlocJtlJnV?% ™* "?1^* ]°SSe°
i^JM^a^r8" a^?age valSe being about"1?^^ in
,.8/» m sortnv. r Nevertheless, this degradation may
ncal bonding of HWT-20 to tf
u" T"T T i, IWT hi
3
of the specimens that passed through the
1 rllu^ ei*her,UCS or'permeabi 1 i tf tests were
UCS results showed no strength loss for the
:r{ \* aPPe?red that a "S" «t'«Sth loss
test specimens where weight losses above
* i;""P
were performed
arf^ds^e;ih^edt%L°^s,f^„]i^u^0r[^§^p0"^^i^s!--«
8.1.9 Laboratory Formulations
Soil samples from Sector B, Sector C, and an uncontaminated
61
-------
area, were solidified, using Type 1 Portland cement at ,15 wt% and
20 wt% addition rates to provide a comparison to treatment witn
HWT-20. The results, shown in Table 9 (at the end of this
section), are as follows:
o For Sector B, untreated soils from locations 4, 5, 14, 16, and
17 were blended and, based on calculations, had a PCB content
of 177 ppm by wt. For Sector C, untreated soils from
locations 1, 2, 5, 7, 9, and 15 were blended and had a
calculated PCB content of 62 ppm by wt.
o The bulk densities were approximately the same as the field
samples.
o The moisture content ranged from 3.6 wt% to 8.9 wt%, which was
less than the treated soil samples.
than
The
for
The UCS for the samples containing 15 wt% cement was less
for the 20 wt% samples by approximately a factor of two.
values using Sector B soil were the lowest, and were
approximately equal to the field core samples. The values
Sector C and the uncontaminated soil (both with a lower
moisture content than Sector B) were higher than the field
samples at Sector C. The low moisture content for the
uncontaminated soil and Sector C soil formulations may account
for the higher UCS values observed. The quantity of water
added was based upon obtaining 100% for the ASTM slump test,
which was less than the amount of water used in the field
operation.
1eachates
1.0
The TCLP
limit of
results.
8.2 CHEMICAL TESTS
for all samples were
which was equivalent
below the detection
to the field sample
Tables
results of
Appendices
discussion,
9 to 11 at the end of this section summarize the
the chemical tests, the details of which are in
E and F. The highlights of the results, with
are as follows:
8.2.1 Soils
The untreated soil samples contained PCBs (Aroclor 1260), up
to a maximum concentration of 950 mg/kg at location B-ll. This is
on the eastern end of the sector, at a depth of 7-8 ft, away from
the anticipated hot spot represented by sample locations B-l, <;,
3, 4, and 5. Samples from locations B-6 and B:7> approximately 7
ft west of B-2, contained PCB concentrations of 650 and 460 mg/kg,
respectively. Other relatively high concentrations of PCBs
occurred at sample locations B-13, 16, and 17, all higher than any
value from samples taken in Sector C. The largest PCB
concentration measured in Sector C was 150 mg/kg at 1ocation C-7,
which is near the southeast corner at a nominal depth of 7-8 rt.
62
-------
Aroclor 1260 was the only PCB Aroclor detected in any sample taken
from either sector.
The treated soil in Sector B contained PCBs up to a
concentration of 170 mg/kg at 1ocation B-ll, the same location
where the maximum concentration was measured for untreated soil.
The maximum concentration of PCBs in Sector C was 110 mg/kg at
location C-3 (7-8 ft depth), which is in the central portion of
the sector, and not close to location C-7 (where the highest
untreated concentration was found). A comparison of treated soil
versus untreated soil concentrations showed no consistent
relationship, only an approximate general trend. High PCB
concentrations in untreated soil locations produced relatively
high PCB concentrations in the treated soil at these same
locations.
A comparison of treated soil to untreated soil PCB
concentrations by sector and by depth was performed. In Sector C,
the quantity of PCBs was approximately equivalent before and after
soil treatment. However, in Sector B, the quantity of PCBs
measured in the treated soil appeared to be only about one-third
of that measured prior to treatment, with slightly better
accountability obtained from the near-surface samples. The most
likely explanation for the PCB reductions was the vertical and
horizontal dispersion of the PCBs due to the mixing involved in
the remediation operation, along with a 30% dilution due to the
addition of HWT-20 and water. This explanation is supported by
the large change observed after treatment in VOC and heavy metals
concentrations at sample locations B-6, B-7, and B-8. Another
possibility that might explain this is a chemical interaction, as
claimed by IWT, between the PCB molecules and the additive
HWT-20. However, the PCB concentrations observed during the
screening tests performed in July 1987 on samples provided by GE
showed that the intended concentration level of PCBs -- about
5,000 mg/kg -- was measured during the laboratory analyses, which
if chemical bonding occurred would have been much lower.
Therefore, the most likely explanation is blending with
surrounding soils of lower concentration.
Eight untreated samples were analyzed for volatile organic
compounds (VOCs), four samples from each sector. VOCs were
detected only at sample locations B-6, B-7, and B-8, with the
maximum concentration of 1,485 mg/kg measured at B-6. The VOCs
detected were total xylenes, ethyl benzene, and chlorobenzene, with
xylenes existing in the greatest concentrations (see Table 12 at
the end of this section) for individual component
concentrations). Therefore, the leaching tests on these samples
included measurements of VOCs.
For the treated soil samples, analyses for VOCs were made at
sample locations B-6, B-6 duplicate, B-7, and B-8. The maximum
VOC concentrations -- primarily xylenes -- were observed at
location B-6, with a concentration of 41 mg/kg. This compares to
the untreated soil, where the concentration of total VOCs at
location B-6 was 1,485 mg/kg. The dilution effect of the
63
-------
additions would reduce the concentration of VOCs by 30%, which
does not come close to accounting for the concentration changes
seen. The large observed change was likely due to a combination
of factors. The largest one was probably the vertical and
horizontal mixing, which would blend high and low concentration
soils. Other factors contributing to the VOC reduction may include
volatilization in the field due to soil disturbance, laboratory
preparation procedures before analysis (e.g., sample crushing),
and the extended hold time before analysis. Although the samples
were kept refrigerated, the maximum allowable hold time of 10 days
was exceeded by 3-4 weeks. However, any analyses for VOC
immobilization compares the ratios of a contaminant in the soil to
\ts leachate for the treated and untreated soils, thus, the
absolute values are of less importance. Another possibility is
that the loss was due to chemical interaction with HWT-20,
although no direct evidence from the Demonstration Test supports
this.
Analyses for the thirteen priority pollutant metals in
untreated soil were performed on six samples from each sector.
The only significant concentrations were found at sample locations
B-6, B-7, and B-8, and are summarized in Table 13 (at the end of
this section). The primary metals detected were chromium, copper,
lead, and zinc, with some samples containing small amounts of
nickel, cadmium, antimony, and arsenic. No analyses were made for
these latter metals in the treated soil or leachates. The maximum
untreated concentration of metals was at location B-6, where the
total quantity was about 5,000 mg/kg. The metals concentration in
the treated soil ranged from 80 to 279 mg/kg. This is a major
change from the untreated soil, a reduction of about 90%, which is
probably due to the soil- mixing/treatment operations. This metal
data provides corroboration for the above explanation on the
reduction of VOC and PCB concentrations in the treated soil.
8.2.2 Leachates
For each of the untreated soil samples, PCBs were analyzed for
in the TCLP leachates. In Sector C, where the maximum PCB
concentration in the untreated soil was 150 mg/kg, all leachates
(except in sample C-ll) contained PCBs below the normal detection
limit of 1.0 ng/L of Aroclor 1260. In Sector B, PCBs were
detected in more than half of the untreated soil leachates.
Except for sample B-7, (a wild point) where the concentration was
400 ng/L , the values ranged from 1 to 13 fj.g/1 .
Of the eleven Sector B untreated samples with detectable PCBs
in the leachates, five samples had soil concentrations of 63 to
140 mg/kg. Except for B-15 (63 mg/kg), the concentrations in the
leachate were 1.1 to 1.6 ng/L . The other six samples, with soi"\
concentrations of 300 mg/kg and above, had leachate concentrations
ranging from 1.8 to 12 ng/l , except for B-7 (400 jitg/L ) . Two soil
samples, B-8 and B-13, with PCB concentrations of 200 and 250
mg/kg, respectively, had leachate concentrations below
64
-------
detection limits.
mg/kg in the soil
leachates.
Therefore, with PCB concentrations below 300
PCBs cannot always be detected in the
For the treated soils in both sectors, Aroclor 1260 was not
detected in any leachate based on the original analyses with a
detection limit of 1.0 M9/L.. Only four treated-soil samples
reported PCB concentrations of 100 mg/kg or more, with a maximum
of 170 mg/kg. Based on
the PCB detection limit
analyses of some of the
4°C, were performed. A
obtain a detection limit
expressed concerns of both IWT and GE that
of 1.0 (J.g/1 was too high, additional
TCLP leachates, which had been stored at
modification in Method 8080 was used to
of 0.1 /jg/L . Four of the seven treated
soil leachates were below the new detection limit. Of the six
untreated soil leachates, five were less than previously measured
and in general were in only fair agreement with the earlier
results; see Tables 10 and 11 at the end of this section.
Based primarily upon these additional leachate analyses, it
appears that the process immobilized PCBs. However, since almost
all of the PCB concentrations in the TCLP leachates were very
close to the detection limits, some uncertainty remains whether
the PCB immobilization took place or not. The screening sample
results performed under the direction of EPA, using laboratory-
prepared samples supplied by GE, reported TCLP leachate analyses
for three samples of less than 1.0 /ig/L . The solidified soil
samples contained 4,100 to 5,700 ppm PCBs, significantly higher
than the demonstration samples (170 ppm maximum). The formulation
tests, using cement as a'substitute for HWT-20, also showed that
the PCBs in the leachates were below the detection limit of 1.0
TCLP leachate results for VOCs were obtained on both the
untreated and treated soils for samples from locations B-6, B-7,
and B-8. The results showed leachate concentration of VOCs from
the untreated soils of 2,490 to 7,890 /ig/L . For the treated soil,
the leachate concentrations ranged from 320 to 605 /jg/L . For each
of the three VOCs -- total xylenes, chl orobenzene, and
ethylbenzene -- concentrations were reduced by an equal factor
from the untreated soil leachate to the treated soil leachate;
leachate reductions were less than the corresponding soil
concentration changes. However, the treated-soil leachate
concentrations were quite low and may not decrease very readily
below the levels measured. IWT claims the composition of their
additive was tailored only to PCB immobilization, and could be
changed should the immobilization of VOCs be required.
TCLP leachate results for the four heavy metals detected in
the soil were obtained on both the untreated and treated soils for
sample locations B-6, B-7, and B-8. The results showed untreated
soil leachate concentrations ranging from 320 ng/l for sample
location B-7 to 2650 /tg/L for sample location B-6. For the
treated soil, the leachate concentrations ranged from 120 M9/L for
sample location B-7 to 210 /zg/L for sample location B-6. The
65
-------
results showed lower treated-soil TCLP leachate concentrations
compared to untreated-soil leachates for lead and zinc; values for
chromium and copper increased. However, the leachate values for
the soil samples after treatment were very low, and were obtained
only for three samples, so immobilization of heavy metals could not
be determined in this project.
Soil samples from sample locations B-6 and B-7 were leached
using the ANS 16.1 and MCC-1P procedures both with site water (PCBs
were not detected in the site water) and deionized water. In
addition, the ANS 16.1 leach test was used for sample locations C-2
and C-4 and leach test MCC-1P was used for sample locations C-l and
C-3; all with deionized water. For MCC-1P, samples of leachate
were collected after 3, 7, 14, and 28 days. For ANS 16.1, samples
of leachates were collected after 1, 3, 7, 14, and 28 days. The
latter test uses fresh leach water at each time interval for the
same solid sample, while MCC-1P involves four parallel leach tank
operating at 40°C, each running for a different time period.
VOCs and PCBs were not detected in any of the leachates, for all
six sets of samples, for both leaching tests. This differs from
the TCLP results, where VOCs were measured in some of the
leachates. Thus, as expected, leaching from a solid sample is less
than from a crushed sample, which contains more surface area.
8.3 MICROSTRUCTURAL STUDIES
Microstructural studies were performed on untreated and treated
soil samples. All analyses were performed three to four months
after soil processing. All samples were studied by scanning
electron microscopy (SEM), optical microscopy, and X-ray
diffraction (XRD). Energy dispersive X-ray spectrometry was also
performed on some samples. The type of information to be obtained
from each test is:
o X-ray diffractometry:
hydration products.
Crystalline structure of the soil and
o Microscopy: Characterizes crystal appearance, porosity,
fractures, and the presence of unaltered soil waste-material
o Energy dispersive x-ray spectrometry:
e.g., calcium, aluminum.
Elemental analysis,
Microstructural and microchemical analyses are proven methods
for understanding the mechanism of structural degradation in
materials similar to those in this demonstration. The literature
is replete with examples of SEM and XRD analyses of soils, cement,
soi7-cement mixtures, and each of those mixed with various
inorganic and organic compounds. However, there have been
relatively few studies of the microstructure of complex waste/soil
mixtures like those resulting from stabilization/solidification
procedures. Consequently, in some cases interpretation of
microstructural observations may be difficult. The microstructural
66
-------
report is intended to be complete in its reporting, yet
conservative in its conclusions.
The detailed report with photographs and X-ray diffraction
patterns is included in Appendix D.
The results can be summarized as follows:
The two most important mineral
were quartz and calcite. This
geological
phases in the samples studied
would be expected, based on the
structure of the Hialeah area.
The morphology of the samples showed subangular clastic grains
of quartz or calcite, cemented by much finer-grained binder
material. The bulk of the binder has a crude layered appear-
ance, usually at most a few micrometers thick. Needles of
ettringite are extremely common, much more common than is
observed in a Portland cement sample of a typical water/cement
ratio of 0.4. The presence of large amounts of ettringite in
a Portland cement sample is a symptom of sulfate attack, which
can in some cases lead to structural failure due to
expansion. The extra sulfate required for extensive
ettringite formation could come from the gypsum that was found
in the untreated soil. However, IWT indicated that the HWT-20
formulation contains a higher content of gypsum than a normal
known whether ettringite observed in
necessarily lead to expansive failures,
a part of the IWT chemical fixation
cement mix. It is not
the treated soils will
as it is claimed to be
technology. The presence of the ettringite, according to
Professor Perry in London (see Appendix D), may aid the
immobilization of metals.
Treatment of the contaminated soil by the IWT process produced
a dense, homogeneous mass with low porosity. Low porosity
reduces the susceptibility of damage from wet/dry, and
particularly freeze/thaw weathering cycles, by reducing the
quantity of water in the pores of the solid that could freeze
and fracture the solid.
All the ettringite analyses showed the presence of high
amounts of silicon, which is not expected in the normal
formula. This is also claimed to be a part of the IWT
techno!ogy.
The quantity of portlandite, a common crystalline phase in
cement, was less than usual. This is probably not a
significantfactor.
Variation of properties in the vertical and horizontal
directions appeared to be absent. No significant difference
in hydration products between quartz-rich and calcite-rich
samples was observed. The mixing operation probably led to
thorough vertical and some horizontal mixing, thus explaining
apparent consistency.
67
-------
8.4 MATERIAL BALANCE
A material balance, summarizing daily operations, is provided
in Table 14 (at the end of this section). It shows the actual
additive usage (not including spills and line flushing) was
approximately 61.2 tons; sodium silicate usage (41° Baume) was
2,440 Ibs.; and overall dosage rate of the HWT-20 additive was
0.171 Ib dry additive/lb dry soil for Sector B, and 0.193 Ib/lb
dry soil for Sector C. Part of the reason the additive rates of
fWT-20 were higher than the targeted value of 0.15 Ib/lb dry soil
was that the average bulk density of the untreated soil was 1.55
g/mL (96 Ib/ft1*) compared to the previously measured (by GE)
value of 1.68 g/mL (105 lb/ft3).
8.5 DATA QUALITY ASSURANCE
8.5.1 Sampling and Analysis
In Section 7 of the approved Quality Assurance Project Plan
(QAPP - Level II), it was indicated that various Quality Control
(QC) samples would be taken to control and/or assess data
quality. These are:
o Laboratory blanks - deionized water taken through sample
preparation steps.
o Field blanks - clean water samples brought from the fie.ld and
then analyzed in the laboratory to check for field
contaminations.
o Surrogate standards - deuterated or halogenated compounds that
respond similarly to the compounds of interest were added to
all samples and blanks for PCB and VOC analyses during sample
preparation. The recoveries of the surrogate compounds are
used to isolate problems that will occur throughout the entire
analytical procedure.
o Spiked samples - samples were spiked with known contaminants
to confirm analytical recoveries and thus accuracy of the
analyses. Duplicates on the spiked samples were also
performed.
o Duplicate samples - duplicate samples from the field were
collected and analyzed to confirm soil sample data.
To verify that correct sampling procedures were used, EPA sent
a Quality Assurance (QA) person to the field to observe sampling
procedures. In addition, QA personnel went to the analytical
laboratory to observe and correct, if necessary, procedures being
used.
68
-------
The purpose of the QA/QC program was to fulfill two related
purposes:
o To provide an organized frame work for sampling and analytical
efforts.
o To control data quality within preestablished limits to ensure
that it was adequate to achieve the objectives of the program.
The laboratory provided the following information on the
quality control data:
8.5.1.1 Calibration Data--
o PCB Analyses - All of the initial three-point calibrations and
calibration verifications met acceptance criteria outlined in
Section 7 of the Quality Assurance Project Plan (QAPP) during
the pretreatment phase of the project. During the
posttreatment phase, calibration verifications at the end of
runs on 7/11/88 and 7/12/88 did not recover within the
required + 15 wt%. New initial three-point calibrations were
generated the following day as required. Samples analyzed
prior to the out-of-control calibration verifications were
quantitated using the response factors from the last
in-control standard.
0 Volatile Analyses - Tune initial and continuing calibration
acceptance criteria outlined in Section 7 of the QAPP were met
for each twelve-hour analysis period during the pretreatment
and posttreatment phases of the project.
o Metals Analyses - Instrument calibration and calibration
verification acceptance criteria outlined in the QAPP were met
for all metals analyses during both the pretreatment and
posttreatment phases of the project.
8.5.1.2 Method Blank Analyses--
o PCB Analyses - For the pretreatment phase, four method blanks
were extracted with soil samples to be analyzed for PCBs.
Seven water method blanks were extracted with leachate samples
to be analyzed for PCBs. During the posttreatment phase,
three soil and twelve water blanks were extracted, along with
soil and leachate samples for PCB analysis. Acceptance
criteria were met for all blanks.
0 Volatile Analyses - Three low-level blanks and one medium-
level blank were analyzed, with soil samples analyzed for
volatiles during the pretreatment phase. Six low-level and
three medium-level soil and eighteen low-level water blanks
were analyzed with soil and leachate samples during the
posttreatment phase. Three pretreatment and seventeen
posttreatment blanks contained detectable concentrations of
69
-------
laboratory solvents and/or target compounds at less than the
maximum concentration allowed in the QAPP.
When samples associated with these blanks were found to
contain compounds also present in the blank, the reported
result was flagged with a "B" qualifier to indicate possible
blank contamination. These results should be considered
accordingly. The laboratory contaminants were usually acetone
and methylene chloride, which are not the VOCs found at the
site.
o Surrogate and MS/MSP Analyses - The mean and standard
deviation of percent recoveries of surrogate and matrix spike
standards added to samples and blanks is listed below for PCB
analyses. The number of values used to calculate the ranges
is given in parentheses ().
8.5.1.3 PCB Analyses--
o Pretreatment Phase
PCB/Soil
PCB/Leachate
Nonochlorobiphenyl
fCI-91
96+13 (38)
92+20 (59)
MS/MSP
111+16 (10)
96±20 (10)
Posttreatment Phase
PCB/Soil
PCB/Leachate
106+17 (29)
99+11 (63)
100+13 (12)
86+_12 (12)
The surrogate standard nonochlorobiphenyl (Cl-9) was added to
each solid sample, leachate, and blank prior to extraction for
PCBs. The recovery of the surrogate was calculated as the percent
ratio of the concentration found divided by the concentration
added.
The matrix spike/matrix duplicate (MS/MSD) analyses for PCBs
were one in ten samples by the addition of a known amount of an
Aroclor to the sample selected prior to extraction. Aroclors
1016, 1232, 1242, 1248, 1254, and 1260 were alternately added.'
The recoveries were outside acceptance limits for three samples.
A method blank for PCBs was prepared for each day's samples
were extracted. There were a few days (see Appendix F) where the
surrogate recoveries were high.
Surrogate recoveries were above acceptance limits for the ten
70
-------
leachates analyzed during the pretreatment phase for PCBs.
Corrective action was not taken because of the limited amount of
sample. Results reported for these samples may be ten to thirty
percent higher than what is actually present. The samples
involved are defined in Appendix F.
The relative percent difference (RPD) was within acceptance
limits.
8.5.1.4 Volatile Analyses--
The mean and standard deviations of percent recoveries of
matrix spikes were as follows:
o Pretreatment Phase
To1-d8 BFB DCE-d4 MS/MSP
VOC/low-level soil 107+10 101+10 92+8 (17) 102+8 (10)
o Posttreatment Phase
VOC/low-level soil 99+6
VOC/medi um-1 eve! soil 96+_4
VOC/leachate 97+4
103+7 95+9 (10) 117+23 (10)
99+4 89+5 (10) 72+14 (10)
99+4 96+6 (109) 87+8 (50)
For VOCs, 1,2-dichloroethane-D4, 4-bromof1uorobenzene, and
08-toluene were used as surrogate standards. All recoveries for
water samples for both pretreatment and posttreatment samples were
within acceptance limits, but a few of the VOC analyses for soil
samples were high and outside the acceptance limits. Agreement
between duplicates was good.
A matrix spike/matrix spike duplicate analysis was performed
for one in ten samples on both the soil and leachates by the
addition of the following: 1,1-dichloroethylene,
trichloroethylene, chlorobenzene, toluene, and benzene. Three
MS/MSD spikes in the pretreatment phase were performed. Two of
the three were improperly performed by the laboratory technician.
The other sample and the surrogate spikes met acceptance
criteria. Therefore, it is expected that the accuracy of the
results should be acceptable.
For the posttreatment samples, hold times to perform the
volatile analyses were missed for all the samples. Additionally,
some of the leaching analyses performed for the project did not
use the zero headspace extractor. Thus, some volatile material in
these tests were lost during the leaching tests. Values reported
in Table 12 used the zero headspace apparatus. Although the
71
-------
samples were held beyond the 14-day hold time allowed (at 4°C),
the evaluation of VOC immobilization is based on ratio of soil
concentration to leachate concentration and would probably not be
affected by unnecessary VOC losses.
The relative percent difference (RPD) was within acceptance
limits for volatile organics MS/MSD analyses.
8.5.1.5 Metals Analyses--
and
For the heavy metals, soil samples for the pretreatment
the posttreatment were analyzed for method blanks, spike
recoveries, and duplicates. The spike sample recoveries were aAT
within acceptance limits of 70% to 130%. The relative percent
difference between duplicates with only two exceptions was within
20%. For the four metals found at the Hialeah site, all values
were within 12% except for one value of 22% for copper. All
atomic absorption instruments were zeroed with a blank solution
prior to analysis. Therefore, the accuracy of the results should
be satisfactory.
o For the physical tests -- moisture, bulk density, unconfined
compressive strength, weathering (wet/dry and freeze/thaw),
and permeability -- a quality assurance program was not
performed, other than some equipment calibrations. However,
sufficient samples were taken in the program to provide
confidence in the results obtained.
o For a few of the bulk densities of the untreated soils, the
moisture content of the soil was so great, causing the sample
to flow, that a Shelby tube sample could not be collected.
Therefore, the sample was collected in a split spoon and a
modified bulk density was performed, as described in Section
6.5.
8.5.2 Operati ons
An operations quality assurance plan was prepared for GE as
part of the Remedial Action Work Plan for the site in December
1987. The following description is taken from the quality
assurance section of this work plan.
The quality assurance officer was provided by GE. His
functions for the site remediation in which the remediation
Sector B and Sector C is the first part are as follows:
of
o Oversee quality assurance aspects of operations.
o Review and approve project planning documents.
o Monitor remedial operations for adherence to QA procedures.
n
-------
Review data for adherence to data quality objectives
Injection point locations were to be
drawing, and these points were to be pos
untreated areas were allowed between inj
remediation of the two test sectors, the
auger positioning, but did not control i
areas, which did occur. Only Sector B,
was monitored. In addition, the quality
the depth of most of the soil columns in
B).
plotted on a site survey
itioned so that no
ections. In fact for the
QA officer monitored the
t to prevent untreated
the first sector treated,
control officer monitored
Sector B (see Appendix
The quantity of HWT-20 was also monitored and controlled by
the QA officer. A curve of slurry density versus percent solids,
based on actual field measurements of various HWT-20 slurry
concentrations, was prepared. A one-liter graduated cylinder was
filled and weighed on a triple beam balance. The measurements for
specific gravity of each slurry batch, using a mud balance, were
taken by Geo-Con, and the feed to the auger would only occur at
the targeted density of 1.51 g/mL. A check of slurry density by
the former method against the mud balance showed good agreement.
The targeted flow rate from the batch plant to the drill auger
was intended to be 0.15 +. 0.005 Ib of HWT-20 per pound of dry
soil, based on a soil density of 105 Ib/ft . In actuality, the
soil densities were lower (96 Ib/ft^), causing the additive
dosage rate to be high; this accounts for a major part of the high
average dosage rates described in Section 8.4.
The batch plant flow meters, mud balances, and rotary valve on
the HWT-20 silo were all calibrated by Geo-Con before coming to
the field. The documentation for these calibration tests is in
the Geo-Con report to GE, and is included as Appendix C.
73
-------
TABLE 5. PHYSICAL PROPERTIES OF UNTREATED SOILS - SECTOR B
Samp
desl
/?at?
B-l
B-2
B-3
B-4
B-5
B-6
B-7
B-8
B-9
B-l
B-l
B-l
B-l
B-l
B-l
B-l
B-l
(a)
(b)
le Moisture
g- content
on(b) %
2.8
3.0
6.4
4.7
4.1
3.6
13.3
13.3
16.8
24.8
0 6.3
1 34.9
2 15.6
3 22.5
4 3.2
5 9.7
6 7.5
7 12.4
12.3
Modified bul
Sample locat
B-l, 2, 3, 4
B-7, 11, 15,
Bulk
density
g/mL
1.50
1.56
1.21
1.41
1.55
1.28
1.46
1.74
1.85
1.59(a)
1.25
1.58
1,52,
1.63(a)
1.46,
1.73(a)
1.52
1.83,
1.46
1.30
1.85
1.32
k density us
ions
, 5, 6, 10,
17 at a dep
PH
7.7
8.4
7.6
7.5
7.3
11.2
8.3
8.1
7.8
8.5
7.8
8.1
7.8
7.7
8.1
7.9
8.2
8.3
i ng spl i t
14, 16 at
th of 7-8
Oil &
grease
%
<0.1
0.1
<0.1
0.1
<0.1
0.1
0.8
1.6
0.4
<0.1
<0.1
0.2
<0.1
0.3
0. 1
0.8
0.1
0.8
0.7
spoon .
a depth
ft
P
TOC
mg/kg
2,100
1,300
2,900
2,050
1,600
2,600
16,000
12,000
3,100
< 100
< 100
8,100
920
1,500
320
960
11,000
9,650
9,400
of 1-2 ft
ermeabi 1 i ty
x 10^
cm/s (c)
1.6
1.0
1.0
0.76
0.50
1.2
1.4
6.0
0.98
0. 15
2.6
0.05
0.91
0.05
0.98
2. 1
3.7
0.13
0.55
B-8, 12 at a depth of 11-12 ft
B-9, 13 at a depth of 16-17 ft
(c)
All the values shown are the
For example, B-l permeability
multiplied by 10 , is reported
permeability multiplied by H
is 1.6x10"^ cm/s and, when
as 1.6
74
-------
TABLE 6.
PHYSICAL PROPERTIES OF UNTREATED SOILS - SECTOR C
Sample Moisture Bulk
desig- content density
nation(b) % g/ml
C-l
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C-9
C-10
C-ll
C-12
C-13
C-14
C-15
C-16
C-17
16
14
17
14
9
9
5
20
14
5
8
19
5
12
23
5
15
23
.7
.6
.1
.8
.1
.1
.7
.2
.7
.9
.2
.5
.7
.5
.5
.3
.1
.2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
.37
.29
.67
.49
.66
.62
.61
.69
.60(a)
.60,
.74
.63(a)
.39(a)
.63
.82
.63
.59(a)
.66
.20
.65,
.60(a)
pH
8.6
8.5
8.2
8.2
8.7
8.3
8.6
8.5
8.4
8.6
8.5
8.5
8.7
8.4
8.3
8.7
8.6
7.9
Oil &
grease
%
0
<0
0
0
0
0
<0
<0
<0
<0
<0
<0
<0
0
<0
<0
<0
<0
.4
.1
.4
.2
.2
.2
.1
.1
.1
.1
.1
.1
.1
.2
.1
.1
.1
.1
Permeabil i ty
TOC x 10^
mg/kg cm/s
5,
1,
8,
6,
1,
1,
2,
3,
2,
1,
1,
1,
2,
1,
2,
1,
1,
200
800
200
600
700
100
000
200
800
500
800
100
600
400
500
400
300
300
0
24
3
0
2
7
1
0
0
3
1
1
1
2
0
0
0
12
0
.58
.0
.6
.27
.1
.0
.0
.84
.83
.5,
.7
.3
.1
.0
.18
.35
.81
.0
.27
(a) Modified bulk density using split spoon.
(b) Sample depth
C-6, 9, 12, 15 at 1-2 ft
C-l, 2, 3, 4, 5, 7, 10, 13, 16 at 7-8 ft
C-8, 11, 14, 17 at 11-12 ft.
75
-------
TABLE 7. PHYSICAL PROPERTIES OF TREATED SOILS - SECTOR B
Sample Moisture
desig- content
nat
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(J)
fon %
l(i) 15
2(i) 9
3(i) 20
4(i) 17
5(i) 31
6 23
12
7 24
8(i) 19
9(i) 15
1 0 ( i )
ll(i) 12
12(1) -
13
14(i) 20
15(h) 21
16(i) 26
17(i) 13
18
19(i) 19
20(h) 17
21(j) 18
22(j) 20
22
23(j) 17
24(j) 20
Reported
weight 1
Permeabi
Permeabi
Permeabi
Permeabi
Permeabi
Permeabi
Sampl es
Sampl es
2-1/8 in
cool ed .
.7
.9
.3
.6
.1
.1
.9
.7
.0
.5
-
.9
-
-
.2
.2
.5
.3
-
.1
.6
.1
.9
.9
.2
.1
as
Bulk
density
g/mL
1.78
1.72
1.74
1.81
1.66
1.77
1.75
1.81
1.88
1.96
2.24
2.15
1.78
1.83
1.81
1.82
1.58
1.79
1.92
1.99
1.76
1.98
1.90
percent
osses of the
lity
lity
lity
lity
lity
lity
coll
coll
after 1
after 1
after 1
after 1
after 1
after 1
ected wi
ected wi
Compres- Permea-
sive bility
strength x 10'
psi
492
330
508
172
206
86
114
115
173
303
470
321
204
221
256
413
507
75
199
479
428
177
579
351
loss of starti
cm/s
1.
3.
0.
2.
4.
21.
5.
11.
2.
8.
4.
3.
3...
ng wei
W/D and F/T control
2 wet/dry cycl
2 wet/dry cycl
2 freeze/thaw
2 freeze/thaw
2 freeze/thaw
2 freeze/thaw
th only 2-1/8
th air cooling
es = 2
es = 4
cycles
cycl es
cycl es
cycl es
Weathering tests
specific wt 7oss, %(a)
W/D
4
-
3
79
3
2
0
9
0
6
3
1
5
5
ght on
s were
.7x10-;
.9x10-'
= 8.9x
= 1.2x
= 3.9x
= 5.9x
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
a
0
1
1
1
1
.38
.32
.37
.42
.37
.43
.49
.34
.53
.44
.40
.26
.46
.39(b)
.39(c)
.27
.41
dry ba
.3-0.4%
cm/s .
cm/s .
0-« cm/
0-7 cm/
0-° cm/
0"' cm/
F/T
0
1
2
3
1
3
27
29
1
6
4
1
0
1
6
23
10
sis.
.
s .
s .
s .
s .
.65
.48
.07
.34
.84
.04
.92
.53
.66(
.06(
.37
.10
.87
.34(
.05(
.28
.73
The
f)
g)
d)
e)
in. corer.
,
. cores taken air cooled and 2-7/8 in.
cores ta
ken water
76
-------
TABLE 8. PHYSICAL PROPERTIES OF TREATED SOILS - SECTOR C
Samp] e
desig-
nati on
Moisture Bulk
content density
% g/mL
Compres-
si ve
strength
psi
Permea-
bility
x 10'
cm/s
Weathering tests
speci f i c wt
W/D
loss, % ( a )
F/T
C-l
C-2
C-3
C-4
C-5
C-6
C-7
C-8
C-9(d)
C-10(d)
C-ll(d)
C-12(d)
C-13(d)
C-14(d)
C-15(d)
C-16(d)
C-17(d)
C-18(d)
18.
14,
20,
17,
20,
14,
12,
11,
20,
15,
20,
18,
19
23
15
13
18
15
16
1.97
1.93
1.95
2.01
1.96
1.95
1.91
1.82
1.91
2.00
1.95
1.93
1.97
1.84
1.99
1.99
1.80
02
02
16.1
1.91
866
528
482
611
656
294
567
343
524
813
460
466
783
409
553
636
247
435
521
530
0.24
1.0
6.4
4.1
1.6
1.9
2.2
4.6
2.5
0.35
0.41
0.27
0.31
0.38
0.40
0.34
0.31
0.38
0.39
0.32
1 .68
0.27
0.40
0.25
0.31
0.33
0.30
0.29
0.32
2.06
8.11
3.94
30.75
2.53
3.12
1.65
1.97(c)
0.72
1.70
0.88
0.99
4.20(b)
8.04
20.98
2.14
2.57
2.95
14.45
(a) Reported as percent loss of starting weight on a dry basis. The
weight loss of W/D and F/T controls were 0.3-0.4%.
(b) Permeability after 12 freeze/thaw cycles = 3.0xlO~° cm/s.
(c) Permeability after 12 freeze/thaw cycles = 2.3x10"' cm/s.
(d) Sample collected with only 2-1/8 in. corer.
(e) All samples collected used water cooling.
(f) All values shown are the permeabilitv multiplied by 107. For
example, C-8 permeability is 4.1x10"' cm/s and, when multiplied
by 10', is reported as 4.1.
77
-------
TABLE 9. KESUIIIS OF POIMILATION STUDIES
property
Slump flow, %
jyioisture content, %
Bulk density, g/mL
UCS, psi
TCLP Aroclor 1260, /jg/L
Clean
15% cement
139.8
3.6
2.01
740
<1.0
Soil
20% cement
102.8
4.0
2.02
1770
<1.0
Sector B
15% cement 20% cement
58.1 79.7
5.1 5.0
2.01 2.03
1332 —
<1.0 <1.0
Sector C
15% cement
129.7
8.9
1.88
170
<1.0
20% cement
116.5
8.9
1.81
318
<1.0
-------
TABLE 12
TOTAL VOLATILE ORGANICS IN SOILS AND LEACHATES
Sample
designation(b)
Total xylenes
Chlorobenzene
Ethyl benzene
Total
B-6d(a)
Total xylenes
Chi orobenzene
Ethyl benzene
Total
B-7
Total xylenes
Chi orobenzene
Ethylbenzene
Total
B-8
Total xylenes
Chi orobenzene
Ethylbenzene
Total
Untreated
soil
mg/kg
1,300
65
120
1,485
_ _
560, 1,000
20, 150
74, 28
916
avg.
-
140, 190
5, 7
13, 23
189
avg .
Untreated
soil
leachate
3,700
280
440
4,420
_ _
—
6,600
290
1,000
7,890
2,100
100
290
2,490
Treated
soil
mg/kg
35.0
1.9
4.4
41.3
32.0
2.2
4.6
38.8
34.0
2.5
4.5
41.0
1.7
< 1 . 2
0.66
2.4
Treated
soi 1
1 eachate
W/l
30
<5
<5
30
<13
< 13
< 13
<13
430
54
120
604
270
19
36
325
(a) Duplicate.
(b) Depth of samples are:
11-12 ft.
B-6 at 1-2 ft; B-7 at 7-8 ft; and B-8 at
81
-------
TABLE 13
TOTAL OF
IN SOILS
FOUR PRIORITY
AND LEACHATES
POLLUTANT METALS
Sample
designation*
Untreated
soil
mg/kg
Untreated
soil
leachate
Treated
soi 1
mg/kg
Treated
soil
1eachate
Chromium
Copper
Lead
Zinc
Total
400
910
2,500
1.000
4,810
10
240
200
2.200
2,650
50
39
140
50
279
40
60
70
40
210
Chromium
Copper
Lead
Zinc
Total
43
70
310
240
663
10
20
<50
290
320
47
12
55
80
194
40
50
<50
30
120
Chromium
Copper
Lead
Zinc
Total
84
59
280
190
613
10
20
100
300
430
46
6
11
17
80
30
40
<50
100
170
* Depth of samples are B-6 at 1-2 ft; B-7 at 7-8 ft; B-8 at 11-12 ft
82
-------
TABLE 14. MATERIAL BALANCE
Date,
1988
4/11
4/12
4/13
4/14
4/15
4/16
Number
of
columns
7
17
12
9
18
9
Totals
Sector
B
B
B
C
C
C
Additive
slurry
Ib
29,720
55,407
27,622
24,138
48,603
25,823
211,313
Dry
additive
Ib
18,575
31,660
15,783
13,792
27,772
14,755
122,337
Supple-
mental
water
Ib
4,621
10,408
6,978
5,315
4,077
708
32,107
Sodium
Sili-
cate
Ib
257
577
321
326
643
319
Total
addi
Ib
34
66
34
29
53
26
2,443 245
tion
,598
,392
,921
,779
,323
,850
,863
Total
column
soil (a)
Ib
85,027
206,499
145,764
84,908
169,816
84,906
776,920
Material Balance (continued)
Date
1988
4/11
4/12
4/13
4/14
4/15
4/16
Number
of
columns
7
17
12
9
18
9
Total
Sector
B
B
B
C
C
C
s
Dry
soil(b)
Ib
75,008
182,158
128,582
72,849
145,198
72,849
677,144
Total soil
addi
1
1,
tions(b)
b
119,625
272,891
180,685
114,687
223,139
111,756
022,783
Soil
weight
increase
%
40.7
32.2
26.1
35.1
31.4
31.6
31.65
Additive
addition,
HWT-20 dry/
soil dry
0.248
0.174
0.123
0.189
0.191
0.203
0.181
(a) Average bulk density of untreated soil, based on laboratory results,
is 96 lb/ftj. In Sector B, column depth was 17.9 ft, and in Sector
C, 13.9 ft.
(b) Correction from wet soil to dry soil is based on the results; average
moisture in Sector B is 11.8% by wt. and in Sector C is 14.2% by wt.
83
-------
SECTION 9
ECONOMICS
9.1
INTRODUCTION
A cost
costs; and
analysis addresses two main categories: capital
operating and maintenance costs.
Capital costs include both depreciable and nondepreciable
cost elements. Depreciable costs include direct costs for site
development, capital equipment, and equipment installation; as
well as indirect costs for engineering services prior to unit
construction (such as feasibility studies and consultant
costs), administrative tasks (such as permitting, construction
overhead and fee), and contingencies. Nondepreciable costs
include startup costs (including operator training, trial or
test run expenses) and working capital. Operating and
maintenance costs include variable, semivariable, and fixed
cost elements. Variable operating cost elements include raw
materials, utilities, and residual water disposal costs.
Semivariable costs include unit labor and maintenance costs,
and laboratory analyses. Fixed costs include depreciation,
insurance, and taxes.
The above cost element breakdown, however, is based on a
permanently sited hazardous-waste cleanup device. The
IWT/Geo-Con system is a transportable unit that will not be
installed at a fixed site. Thus, it involves some cost
elements that are different in their impact on a cost analysis
compared to cost elements occurring with the more typical
permanent installations.
In general, the cost for a transportable hazardous-waste
remediation-facility falls into three categories: capital
costs, including all costs that can be amortized over the
service life of the unit; mobilization/demobilization costs
associated with startup and shutdown at a given site, and that
can be amortized over the duration at the site; and operating.
costs to operate and maintain the system. Capital costs can be
subdivided into direct, indirect, and nondepreciable cost
elements. Mobilization/demobilization costs can be accrued as
semivariable operating and maintenance costs. Operating costs
include variable utility and raw material costs, semivariable
labor and maintenance costs, and fixed costs such as
depreciation, insurance, and taxes.
Several capital cost elements defined for the permanently
sited unit need to be redefined into cost categories pertinent
to a mobile unit. These include the direct costs for site
development and the direct costs for engineering studies, which
84
-------
on a site-specific basis, become mobilization/demobilization
costs. These factors are not included here because of the
complexity of the analysis and planning in this area. Total
site cleanup is the responsibility of other contractors, with
the in situ stabilization/solidification technology used for
only a section of the total site remediation.
Based on the above,
illustrated in Table 15,
9.2 COST ELEMENTS
an overall cost element breakdown, as
can be developed.
A detailed
Table 14 is
discussion of each of the cost elements defined
provided in the following:
9.2.1 Capital Costs
9.2.1.1 Direct Costs--
Equipment fabrication/construction and/or purchase--The current
costs for the design, engineering, materials and equipment
procurement, fabrication and installation of the in situ
stabilization/solidification process, are included as direct
costs. The costs include all the subsystems and components
installed. Waste preparation equipment is not included as it
can be rented or provided by the site-responsible party.
Pretreatment or posttreatment of the soil is not required.
9.2.1.2 Indirect Costs--
Admini strative/permittinq--Administrative costs associated with
regulatory compliance issues could be numerous and varied, and
these costs are not included in this analysis. The costs that
are being accrued under this cost element are directed to the
overall non-site-related regulatory activities in establishing
federal and state permit requirements, preparing initial permit
applications, and supporting permit application information
throughout the permit issuance process. Once the final permits
are issued, recordkeeping, inspection, survey response to
permitting agencies, and additional reporting activities may be
required. These costs include the preparation of technical
support data, sampling/analysis project plans, and quality
assurance project plans by in-house engineering personnel;
preparation of RCRA/TSCA permit forms (if applicable); time,
travel, and per diem for consultant and in-house staff
interfacing with federal EPA officials; and in-house
administrative and clerical staff.
For this cost analysis, administration costs, taken as
percent of direct costs, include office expenses such as
supplies and furniture, but not salaries (included elsewhere).
85
-------
TABLE 15. COST ELEMENT BREAKDOWN
Direct
Indirect
Nondepreciable
Variable
Semivariable
Fixed
Capital cost
Equipment fabrication/construction or
purchase
Administrative/permitting
contingency
Operations procedures/training
Initial startup/shakedown
Working capital
Operating and maintenance costs
Raw materials: HWT-20, sodium silicate
Power
Water
Fuel
Waste disposal
Labor
Maintenance
Analyses
Mobi1izati on/demobi1ization
Site preparation/logistics
Transportation/setup
Onsite checkout
Working capital
Decentamination/demobi1ization
Depreci ation
Insurance
Taxes
Contingency — A contingency cost,
capital cost, is usually allowed
cost definitions.
9.2.1.3 Nondepreciable Costs--
approximately 10%
for unforeseen or
of direct
poorly defined
Operations procedures/training--In order to ensure the safe,
economical, and efficient operation of the unit, the creation of
operating procedures and a program to train operators is
necessary. Costs that may accrue include: preparation of a unit
86
-------
health/safety and operating manual, development and
implementation of an operator training program, equipment
decontamination procedures, and reporting procedures. All
documentation must be site-specific, though they can be derived
nwnl th3"?^ documents. The preparation costs can be amortized
over the life of the equipment.
Initial startup/shakedown--After the unit is brought to a site
it must initially be started and operated to check out the
mechanical and technical integrity of the equipment and its
controls. This cost is assumed to be one week of labor expenses.
Workinq capital --Although the unit is a transportable system, it
nnILreqU-rKia suPPjy of maintenance materials attributable to a
nondepreciable capital cost. Maintenance materials typically
account for approximately one-half of the total maintenance cost,
and three-month inventories are usually maintained. This cost is
included in the Geo-Con equipment costs and facility
modifications.
9.2.2 Operating and Maintenance Costs
9.2.2.1 Variable Costs--
Variable operating cost elements for this unit include fuel,
power, water, chemicals, and process waste disposal. They are
defined as variable operating cost elements because they can
usually be expressed in terms of dol1ars-per-unit flow of soil
treated and, as such, these costs are more or less proportional
to overall facility utilization during specific site operations.
It is also assumed for the tabulation of costs that there are no
process waste by-products.
Fuel--The
power the
fuel
requirement for the unit includes diesel fuel to
crane and hydraulic power pack. In addition, fuel is
used for supporting vehicles--backhoe, front-end loader--and for
generators for lights and possibly heat
diesel
Power--The power requirement for the unit includes the electrical
requirements for the trailers, sampling equipment, auxiliary
lighting, etc. This is assumed to come from plant facilities.
Water — Water use is based upon the water content of the
feedstock, to bring the cement-like final slurry to about 18% by
wt. water. In addition, 500 gal of water is used for
decontamination.
Chemicals--The IWT proprietary additive is HWT-20
a rate of 15 wt% to dry contaminated soil.
It is used at
Decontamination wat.er--Tf the unit is not operated 24 h/d it
needs to be cleaned with high-pressure water or steam to prevent
plugging. Costs will accrue for the containment and disposal
87
-------
of this waste
water is used
stream.
on site
It is assumed that all decontamination
9.2.2.2 Semivariable Costs
Labor—This analysis determines total operating personnel based
upon 1 shift per day, 5 shifts per week for a total of 8 people.
This includes 5 process operators, 1 supervisor, a site safety
and health officer, and 1 overall coordinator.
In addition, there are 3 support personnel for office
operations, including a combined office manager-purchasing agent,
a secretary, and a part-time sample technician. This totals 11
people.
Maintenance--Maintenance materials and labor costs are extremely
difficult to estimate and cannot be predicted as functions of a
few simple waste and facility design characteristics, because a
myriad of site-specific factors can dramatically affect
maintenance requirements. Annual maintenance cost will be
assumed as 10% of capital cost.
Analvses--In order to ensure that the unit is operating
efficiently and meeting environmental standards, a program for
continuously analyzing waste feed and treated solids is
required. Initially sample sets will be taken every day, and
less often as operation efficiency improves. A sample set is
assumed to cost $1,200.
Mobilization/demobilization--As discussed in Section 8.1, the
following costs will accrue to the Geo-Con unit at each specific
site. The costs are site-specific and may vary widely, depending
on the nature and location of the site. They include site
preparation/logistics, transportation/setup, construction
supervision, onsite checkout, site-specific permitting/
engineering services, working capital, and decontamination/
demobilization. Notes to Table 16 indicate the items included in
the cost analysis.
Site preparation/1oqisties--The costs associated with site
preparation/logistics include advanced planning and management,
detailed site design and development, auxiliary and temporary
equipment and facilities, water conditioning, emergency and
safety equipment, and site staff support. Soil excavation,
feedstock preparation, and feed handling costs are also
included. This may be performed by companies other than Geo-Con
or IWT but still comprises part of the site remediation costs.
Due to the temporary and transient nature of the setup at an
assumed Florida site, the costs incurred for the demonstration
were based in part on Geo-Con estimates. Costs for advanced
planning, detailed site design and development, and water
conditioning, if needed, are assumed to be part of the site prime
contractor's expenses, and are not included.
88
-------
Transportation/setup--The cost of transportation and setup
includes transport to a new location. The costs for the Geo-Con
deep-soil-mixing unit are included in mobilization/demobilization
and subcontract costs.
Onsite checkout--0nce the unit has been set up, it is necessary
to shake down the system to ensure that no damage occurred as a
result of disassembly, transport, and reassembly.
Working capital--Fuel inventory, sodium silicate, and HWT-20-
additive storage facilities will exist at each site and, as such,
are semi-variable operating costs specific to the site-specific
mobilization/demobilization cost-element breakdown. These
storage facilities are included as part of the capital costs.
Decontamination/demobilization--With the completion of activities
at a specific site, the unit must be decontaminated and
demobilized before being transported to its next location. Costs
that will accrue to this cost element include field labor and
supervision, decontamination equipment and materials, utilities,
security, health/safety activities, and site staff support.
9.3 OVERALL COST EVALUATION
A primary purpose of the economic analysis is to estimate
costs for a commercial-size remediation. However, it was assumed
for this analysis that part of a large Florida site would be
remediated. The costs used were provided by Geo-Con and were
based on the unit used during the demonstration. For this case,
the analysis assumes that the treatment unit processed sixteen
soil columns per day; and that the additive consumption rate was
0.15 Ib additive/lb dry soil, based on soil conditions found at
the GE site. The results of the analysis are presented in Table
1 6 .
These results show a cost per ton (without
soil of $194 for the 1-auger unit used for the
can be seen from the results, 85% of the costs
materials, equipment rental, and labor.
fee) for untreated
demonstration. As
consist of raw
89
-------
TABLE 16. ESTIMATED COST
1-auger
assembly
Capital cost
Direct
Equipment costs, $ 77,000
Indirect, $/ton
Administration (10% direct costs) 0.45
Contingency (10% direct cost) 0.45
Nondepreciable, $/ton
Initial startup/shakedown (5 days of labor expenses)
and operator training 0.31
Working Capital
Operating and maintenance costs
Variable, $/ton
HWT-20 52.45
Sodium silicate 0.23
Fuel ($1.00/gal-diesel) 2.16
Electricity ($0.04/kWh) 0.21
Water ($0.80/1,000 gal) 0.02
Semivariable $/ton
Salaries and living expenses 45.73
Equipment rental and subcontracts 67.90
Consummable 19.29
Analytical services 3.28
Maintenance (10% direct costs) 0.45
Mobilization/demobilization 0.62
Site preparations
Misc. (Insurance, taxes, etc.) 0.45
Depreciation 0.45
Totals, $/Ton 194.45
90
-------
Table 16. Notes
1. The demonstrated Geo-Con unit could process one 3-ft diameter
soil column to a depth of 16 ft in 30 min. Thus, 16 columns/d
on the a.verage were treated.
2. Operations were based upon 5 d/wk, 8 h/d.
3. Equipment life estimated at 10 yr.
4. It is assumed that 50,000 ft2 of soil were processed to a
depth of 16 ft; this is equivalent to 38,400 ton.
5. Labor and living expenses for Geo-Con operating supervisory
personnel were provided by Geo-Con at $20 yd3. Support
personnel costs were estimated by EPA.
6. Utilities consumption estimates:
Electricity - provided by others at battery limits at
$0.04/kWh
Water - the daily average rate for the process,
decontaminating, and miscellaneous was
4.3 gpm.
Diesel Fuel - 140 gal/d
7. Average bulk density of soil is 96 lb/ft3, or 2,592
lb/yd3.
8. Chemical consumption:
HWT-20 - 0.15 Ib/lb of dry soil; soil moisture content 8%
by wt.; delivered cost $380/ton.
Sodium silicate - (41° Baume); 5% by wt. (100% basis) of
HWT-20 additive; delivered cost $177/ton of
soluti on.
9. Labor estimate: 1 shift, 5 days per week; includes overhead
(no profit). Data provided by Geo-Con - $50/yd for 1-shaft
augers. Includes all living and travel expenses for 8
people. Since the onsite time is about 2 1/2 yr, assume an
office manager and a secretary. In addition, a sampling
technician is required. Salaries plus overhead for the
non-operating personnel are $30/h for office manager, $16/h
for a secretary, and $25/h for a sampling technician.
10. Mobilization/demobilization: Labor, subcontracts, etc. - One
week of labor charges for each is assumed.
11. Capital cost of equipment as provided by Geo-Con:
Mixing plant - $50,000
Flow control system - $20,000 (1-auger)
Augers & auxiliaries - $7,000 (l-auger--some added equipment
assumed to be rented)
91
-------
samples daily;
remedi ati on.
Table 16. Notes (continued)
Totals for 1-auger system, $77,000.
12. Laboratory analysis: First 2 weeks, 1 set of
then 1 sample set weekly for remainder of the
Sample set cost is $1,200.
13. Rental equipment: Crane, backhoe, hydraulic power pack,
pickup,truck, other vehicles, miscellaneous. One-shaft auger,
$70/yd3. ^Basic data were provided by Geo-Con with some
adjustments assumed.
14. Subcontract expenses: Trucking, electric wiring, piping is
$18/yd3. Data provided by Geo-Con.
15. Miscellaneous: Purchases, insurance, health and safety
supplies. Geo-Con estimated cost of $25/yd .
16. Cost for permitting, recordkeeping, inspection, and other
related activities were not included in this analysis. Site
preparation, since it would be so interrelated with the
overall planning and costs of the prime contractor for the
entire site, was not included.
17. The maintenance and working capital costs were prorated to the
actual time on site, 28 mo for a 1-auger system.
18. Administration and contingency are assumed at 10% of capital
cost (per annum) and prorated to the actual time on site.
19. Operator training assumes 5 days of training for Geo-Con oper-
ators, the Health and Safety Officer, and sample technician.
20. Many of the costs shown were provided by Geo-Con, and included
a fee. Some adjustments were estimated, so that the bottom
line in Table 16 more closely estimated an actual cost without
fee.
*U.S. GOVERNMENT PRINTING OFFICE: 1989.6^8.163/00309
92
-------
-------
-------
-------
Environmental Protection
Agency
Information
Cincinnati OH 45268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use, $300
Please make all necessary changes on the above label,
detach or copy, and return to the address in the upper
left-hand corner.
If you do not wish to receive these reports CHECK HERE o;
detach, or copy this cover, and return to the address in the
upper left-hand corner.
EPA/540/5-89/004a
------- |