United 'States
Environmental Protection
Agency
Risk Reduction
Engineering Laboratory
Cincinnati OH 45268
Research and Development
EPA/600/S2-90/026 July 1990
&EPA Prpject Summary
Treating Chlorinated Wastes
with the KPEG Process
i
M. L. Taylor, J. A. Wentz, M. A. Dosani, W. Gallagher, and J. S. Greber
The two reports summarized here
describe development of the alkali
metal (polyethylene gylycolate (APEG)
chemical technology to dechlorinate
hazardous hydrocarbons in soils and
its application at four demonstration
sites: field-scale application to
contaminated soils on the island of
Guam; pilot-scale demonstration in
Moreau, New York; and drum-stored
contaminated materials at the
Bengart & Memel site in New York
and ir|i Omaha, Nebraska. The Omaha
site involved tetrachlorodibenzo-p-
dioxiri (TCDD) in waste herbicides.
Equipment for the field-scale
application on Guam included a
stearji-jacketed mixer, a steam-
generating plant, and a condensate
collection system. Approximately 15
yd3 of contaminated soil (with
concentrations averaging 3535 ppm
Aroclbr 1260) were batch treated with
potassium polyethylene glycoiate
(KPEG). The polychlorinated biphenyl
(PCS) concentrations were reduced
by more than 99%, with no individual
PCB congener exceeding 2 ppm.
The reports provide detailed
information about preparation of
contaminated soils before treatment,
equipment and reagents used during
treatment, and analyses done before,
during, and after treatment. Potential
users are given sufficient information
of th:e KPEG process to ascertain
which version of the process is best
suited for a particular site. '•.
This Project Summary was develo-
ped by EPA's Risk Reduction Engi-
neering Laboratory, Cincinnati, OH, to
announce key findings of \the
research projects that are fully docu-
mented in two separate reports (see
Project Report ordering information at
back).
Introduction
The two reports concern a chemical
dechlorination process that employs a
reagent, typically prepared by reacting a
base (e.g., potassium hydroxide) with one
of several polyethylene glycols (e.g.,
PEG 400). Early laboratory- and drum-
scale studies are described that proved
the feasibility of the KPEG technology to
treat PCB-contaminated soils. In-depth
descriptions are given of the KPEG-
process demonstrations that have been
done to date.
A typical laboratory-scale procedure for
dechlorinating polychlorinated dibenzo
dioxins (PCDD)- and PCB-contaminated
soil involves mixing potassium hydroxide
(KOH) and polyethylene glycol-400
(PEG-400, average molecular weight of
400 Daltons) to produce the reagent
KPEG. This KPEG reagent is mixed with
the contaminated soil, heated to approx-
imately 150°C (302°F), and maintained
at that temperature while being con-
tinuously stirred for 1 to 4 hr. Excess
reagent is then decanted, the soil is
neutralized with acid and rinsed two or
three times with water, and the decon-
taminated soil is discharged.
The reaction of the KOH and PEG-400
produces an alkoxide (ROK) (Equation 1),
which in turn reacts with one of the
chlorine atoms on the aryl ring to
produce an ether (ArCI,^ OR) and potas-
sium chloride (KCI) salt (Equation 2).
Replacement of the chlorine atom on the
aryl ring with an ether-linked PEG
-------
detoxifies the compound. The dechlo-
rination process is described in Equations
1,2, and 3:
ROH + KOH
ROK + ArCln
ROK + H20 (1)
! OR + KCI (2)
AH
Ar Cln., OR — •*" Ar Cl n.t OH + CH2 =
(3)
H
I -OR
C
Pilot-Scale and Field-Scale
KPEG Treatment
Demonstrations
Four site demonstrations of the KPEG
process are discussed. A pilot-scale
demonstration was conducted at the
Moreau, New York, site, and a field-scale
demonstration was conducted at the
Guam site. At the Bengart & Memel site
in New York and a site in Omaha.
Nebraska, drum-treatment demonstra-
tions of the KPEG process were con-
ducted. Discussions of these four site
demonstrations should assist potential
users of the KPEG process with sufficient
information to assess which of the four
versions of the KPEG process is best
suited for implementation at a particular
site.
Pilot-Scale Demonstration of
KPEG Process in Moreau,
New York
The objectives of the demonstration at
the Moreau site were to document that
the KPEG reagent could effectively de-
chlorinate PCBs to acceptable levels at
the 40-gal scale (from hundreds and
thousands of parts per million to less than
3 to 4 ppm) and to gather process data
that would subsequently be used to
design a larger (2-yd3 or 400-gal) reactor.
The test at Moreau was not intended to
be a site cleanup; only a small portion of
the contaminated soil was treated as part
of the test.
The Moreau site was formerly a drag
strip. The drag strip was oiled period-
ically, and some of the oil used for this
purpose was transformer oil containing
PCBs. The PCS concentrations in the soil
at the site ranged from nondetectable up
to tens of thousands parts per million.
The pipe reactor (Figure 1) used during
the pilot-scale tests was constructed from
a 40-in. length of 16-in. steel pipe. It was
mounted on a stand that allowed it to
rotate about the indicated pivot point, and
because it rotated through an arc of 180°
in about 10 sec, mixing inside the pipe
was gentle. A baffle along the side of the
pipe (midway down its length) redirected
the flow of slurry along the pipe wall and
enhanced mixing.
Four electric heating straps heated the
reactor. One 1000-watt heating strap was
wrapped around the top of the reactor,
and another was wrapped around the
bottom. Both straps were about 10 in.
wide. Two 3-in.-wide 1500-watt straps
were wrapped around the center portion
of the reactor
Hose Connection for
Draining and
Adding Reagent
1000-W
Strap Heater
Thermometer
0°-/50°C
Sample Valve
1500-W
Strap Heater
Internal Baffle
1500-W
Strap Heater
Arc of Rotation
1000-W
Strap Heater
16-Bolt
Flange
Screen Stack for
'Draining Reagent in
Top Flange
1/2 Area Internal Baffle to
Protect Screen Stack from Rocks
-60 psi Gage
Vent Line
Welded
End Plate
Removable Insulation Is Used During Operation
Figure 1. Reactor detail for the Moreau, New York pilot-scale demonstration.
-------
During the reactor operation, tempera-
ture and pressure were recorded every
15 min. The reactor motion and its
direction were controlled by a hand-held
switch; therefore, during the entire time
the reactor was operating, someone held
the switch and manually reversed the
reactor's direction at the end of each
180° turn.
Vapor from the reactor was vented
through a smalt fan-cooled condenser
(Figure 2). The condenser condensate
was collected in Condensate Drum No. 1,
which was vented to an ice condenser
consisting of a coil submerged in an ice
bath that drained to Condensate Drum
No. 2. Condensate Drum No. 2 was
vented through an activated carbon
canister to remove any trace volatiles
before atmospheric discharge.
The treatment procedure involved pre-
paring the solvent PEG mixture —
dimethylsulfoxide (DMSO), triethylene
glycol methyl ether (TMH), and PEG-400
[ratio 2:1:1]—in a 55-gal drum. The soil,
solvent mixture, and 45% KOH solution
were then loaded into the reactor and
mixed. The reactor was heated to 150°C,
and the samples were withdrawn by
removing the thermometer and pouring
the slurry (-50 g) out of the thermometer
insertion port.
Altogether, four runs were made with
this npactor. Table 1 lists the reactor
contents for each of these runs. After the
samples were taken out of the reactor,
both soil and reagent were analyzed for
RGBs] The reductions in PCB concentra-
tion in| the soil achieved during Run No. 1
ranged from 93.9% to 99.8% ^and
averaged 98.3%. The PCB reduction in
the rehnaining three runs (Runs 2 through
4) was greater than 99%. PCBs were jalso
founcj in the recovered reagents,
generally at higher concentrations than
thosel in the soil; this indicated that
although the PCBs were being extracted
from the soil into the liquid phase,
chem cal dechlorination was not yet
comp ete. The PCB concentrations in the
reagent were 0.16 ppm in Run 1, 74.8
ppm In Run 2, 72.4 ppm in Run 3, and
4.25 ppm in Run 4.
F/e/o-Sca/e Demonstration of
KPEp Process in Guam
The; U.S. Navy Public Works Center on
Guarr was selected for the field-scale
rrj
o
demostration of the KPEG chemical
dechlprination system. The PCB concen-
trations in the soil at the site averaged
3535 jppm, with "hot-spots" as high as
45,860 ppm (4.59%). Soil contamination
(found primarily in a nearby storm
drainage ditch) was the result of leaks
from a transformer rework building and
waste PCB stored outside the building.
Excavated, contaminated soil was stored
in a 12.2- by 18.3-m (40- by 60-ft) metal
building with a heavy rubber liner erected
on a concrete pad. Also another 30.5- by
30.5-m (100- by 100-ft) concrete pad was
poured for the installation of the field-
scale KPEG treatment system.
Soil and particles smaller than 1/2 in.
were passed through the screens and
collected separately so that particles
would not be jammed in the mixer.
Laboratory studies of contaminated soil
samples from Guam determined the
soil's potential for KPEG treatment and
established the reagent formulation and
other operating parameters. Figure 3 is a
simplified mechanical flow diagram of the
KPEG system at Guam. As illustrated, the
mixer was the primary component of the
system in which the chemical dechlo-
rination process occurred. An extensive
pipe network connected the ancillary
equipment with each other and with the
mixer.
The principal components of the Guam
KPEG treatment system (mixer, platform,
liquid reagent loading system, heating
system, nitrogen system, condensate col-
lection system, process cooling system,
reagent collection system, and ventilation
Condenser
Center
Pivot Point
Pipe
Reactor
Thermometer - "" ,./
Vent
Arc of
Rotation
Condensate
Drum #7
Condensate
Drum #2
Figure 2. Equipment arrangement for the Moreau Test.
Run
1
2
3
4
Soil
33.5
32.5
34.0
39.0
Sulfolane
0.0
0.0
0.0
15.0
DMSO
11.0
10.0
12.0
0.0
TMH
PEG-400
5.5 5.5
5.0
6.0
5.0
6.0
7.5 7.5
45% KOH
11.0
10.0
12.0
13.0
Total ,
Liquid
32.9
30.0
35.9
43.0,
Nitrox
Drum
-------
Contaminated
Soil
KOH Pellets
Liquid Reagent
(PEG-400)
Process
Cooling Water
Steam
Generator
Diesel Fuel
Air
Compressor
~ System Air Requirements
Condensate
Collection
Neutralization
Acid
Nitrogen
Reagent
Collection
Required Process
Optional Process
Figure 3. Mechanical How diagram of KPEG field-scale treatment system.
system) are described in detail in the full
report.
Six 55-gai drums were placed on a
metal pallet, the pallet was positioned in
the metal building next to the stockpiled
soil, and the soil was shoveled into the
drums. The six drums of soil were
relocated to the scale, where a small
forklift with a drum-lifting attachment
placed them on the scale. A grab sample
was collected from each of the six drums
of soil. These samples were placed in a
container and mixed thoroughly to
provide a homogeneous sample for PCS
analysis.
A crane hoisted the drums of soil, and
two persons on the platform used the
manual chainfall on the drum lifter to
dump the drum contents into the mixer.
After every second drum of soil was
loaded, the mixer was turned on for a
short time to disperse the soil evenly
throughout the mixing cylinder.
After all the soil was loaded into the
mixer, the components of this KPEG
(PEG 400 and KOH) formulation were
added. The mixer was run at low speed
(30 rpm) while the PEG-400 was pumped
into it; this allowed thorough dispersion of
the PEG-400 and KOH to initiate the first
reaction shown in Equation 1. When the
reagents and soil were completely mixed,
the mixer was switched to high speed (60
rpm). The vent line from the mixer to the
condensate collection system was
opened, and the fan condenser was
turned on.
The steam-generating plant was ignited
and 80-psi steam was circulated through
the mixer jacket. The mixer contents
reached 150°C in approximately 2 hr,
and the temperature and mixing were
maintained for 4 hr. At the end of 4 hr,
the steam generator and mixer were shut
down, the fan condenser was turned off,
and the contents were allowed to cool
overnight.
After cooling overnight, all the treated
batches of soil were neutralized with
concentrated sulfuric acid and dis-
charged. Samples were collected from
the sample-collection port, and the slurry
pH was measured. If the collected
sample had a pH in the 6 to 9 range, it
was also used as the posttreatment soil
sample. The treated soil and reagents
used in the treatment process were dis-
charged into the soil-collection hoppers
pending analytical results.
A high-pressure (2000-psi) water
washer was used to decontaminate the
process equipment, and the resulting
water was collected in 55-gal drums. The
collected condensate and water were
passed through a series of cartridge
filters and activated-carbon canisters. The
treated water was then collected,
sampled, and held until results of
-------
analyses for PCB contamination were
obtained. Treated waters with nonde-
tectable PCB concentrations were either
collected by a liquid transport truck and
discharged into the sanitary sewer
treatment plant or pumped back onto the
hot zone.
PCB analyses employed an EPA-
developed field-screening method, which
was followed with a more vigorous,
corroborative analysis. A review of
treated soil analytical data from both
analyses indicated that the destruction of
the total PCB concentration exceeded
99%. In both sets of data, however,
analysis of each of the congener peaks
showed that a tetrachlorobiphenyl
congener concentration in a portion of the
treated batches was slightly above the
R&D permit requirement of 2 ppm or
lower per PCB peak. The batches con-
taining greater than 2 ppm of the tetra
congener were loaded back into the
mixer, mixed with the new KPEG formu-
lation of 1.3 moles KOH and 1.0 mole
PEG-400, and retreated. Analytical re-
sults of the retreated batches indicated
the tetra congener in each batch was well
below 1 ppm.
Drum Treatment Demonstration
of KPEG Process at Bengart &
Memel Site, New York
The preceding two KPEG demonstra-
tions were performed to prove the effec-
tiveness of the KPEG process at larger
scales and to promote the development
of a full-scale system. The demonstra-
tions described in this section and the
one that follows concern the in situ
treatment of smaller quantities of con-
taminated materials that may have
already been placed into drums for
storage. This process offers an alterna-
tive to the costly and hazardous process
of emptying drums of hazardous mate-
rials.
When Bengart & Memel, Inc., a whole-
saler of nonferrous scrap metals, dis-
mantled PCB transformers and capaci-
tors, PCBs were released into the soil. In
the mid-1970"s, soil samples from this
site contained PCB concentrations ex-
ceeding 50 ppm. Soil sampling and
analysis indicated that PCB concen-
trations were excessive in seven loca-
tions on the property. As part of the
remedial program, the contaminated soil
was excavated and placed in 166 55-gal
steel drums. Because of a consent order
deadline, only 51 drums of soil were
treated in this demonstration. All the
remaining drums were shipped, un-
processed, directly to a landfill.
During the KPEG treatment process, a
weighed quantity (typically 150 to 170 Ib)
of a single-phase reagent mixture
(2:2:4:9:5-PEG:TMH:DMSO:45% KOH:
water!) was added to each of the' soil
drums. A bimetal thermometer was in-
serted into the 3/4-in. bung on each drum
coveij, and the drums were heated :from
the bottom to ensure that all of the soil
contaminants would react. The soil |tem-
perature was monitored 9 to 12 in. below
the top of the drum, and the soil temper-
atures were typically maintained between
105° and 110°C for several hours. ;
The drums were vented to a central
condenser system with a capacity pf 16
drums (see Figure 4). During heating,
vapors exited from the drum through an
insulated flexible line connecting the 2-in.
drum bung with the main header.
Because part of the header was: ice-
jacketed, much of the water vapor was
condensed and drained directly into a 55-
gal holding drum. The remaining vapor
was drawn through an air-cooled radiator
(used as a condenser), which also
drain'ed liquid to the 55-gal holding drum.
An ice-jacketed scrubber containing a
dilute sodium hypochlorite solution (for
odor
control) trapped most of the rernain-
ing condensables. From the scrubber, the
vapo' passed through a 55-gal drunrvfilled
with a mixture of activated carbon and
containing a molecular sieve. A vacuum
pump provided negative pressure for the
vapoj' control system.
A 30-in. auger attached to a hand drill
was {used to collect treated soil samples
from the drums. Four samples (one; from
the center, one close to the drum; wall,
and two spaced between the center and
wall); were collected from the drum by
penetrating the soil as deeply as possible
with the auger.
After the treated soil had been
decontaminated, the soil had to be
brought from a highly alkaline to a neutral
pH bjy the addition of dilute acid.
Soon after the KPEG treatment
(October 1986), EPA determined the con-
centrations of PCBs in selected drums. A
second group of drums was sampled and
analyzed in February 1987 after the
reagent had been in contact with the soil
for 5 mo. Group 2 included fresh samples
froml the drums that EPA found contained
PCB concentrations exceeding 50; ppm,
samples from treated drums not analyzed
.by the EPA, and one soil sample that
contained less than 50 ppm when EPA
analyzed it. The low-concentration j sam-
ple Was analyzed to check the potential
reacfion progress in drums with PCB
concentrations below 50 ppm. A :com-
parison of the reduced concentrations
reported in February with the October
results indicated that the reaction con-
tinued during the holding period.
A review of treated-soil analytical data
indicated that PCB concentrations were
reduced from 108 to 27 ppm (a 75%
reduction). The PCB levels in the soil in
one drum with an initial concentration of
1300 ppm were reduced to 78 ppm (a
93% reduction).
Drum Treatment Demonstration
of KPEG Process at an Omaha,
Nebraska, Site
This 1987 EPA study was designed to
demonstrate the effectiveness of the
KPEG process for treating drummed,
contaminated herbicide wastes containing
dioxins. The contaminated waste con-
tained 17,800 ppm 2,4-D, 2800 ppm
2,4,5-T, and 1.3 ppm 2,3,7,8-TCDD.
The KPEG reagent used in this demon-
stration was prepared with PEG and
KOH. The field-scale testing of the KPEG
technology was performed on one 20-gal
batch of the contaminated waste.
The treatment equipment consisted of
a 55-gal drum, a clamp-on heating band,
and a stirring motor. The drum containing
the KPEG reagent and the contaminated
waste was heated to a temperature of
70° to 85°C for 2 days while being con-
tinually stirred. After 2 days, the heater
was turned off and the reaction mixture
was allowed to cool to ambient tem-
perature. The treated waste samples
were then analyzed to determine the
posttreatment concentration of the con-
taminated waste.
Table 2 presents the analytical results
of samples collected before treatment
and after the first and second treatments.
The KPEG treatment reduced the con-
centration of 2,3,7,8-TCDD in the waste
to less than the detectable range. The
concentrations of both 2,4-D and 2,4,5-T
were reduced by 98%.
Conclusions
The pilot-scale KPEG demonstration in
Moreau, New York, represented the first
attempt to dechlorinate PCBs in a
reactor/mixer at a scale larger than that
used in the laboratory. Results of the
demonstration indicated that PCBs could
be reduced by an average of 98.3%
(reductions ranged from 93.9% to
99.8%).
The field-scale demonstration at Guam
proved the effectiveness of the KPEG
technology for treating soils with initial
PCB concentrations averaging 3535 ppm
-------
Drum
Heaters
To
Vacuum
Pump
Reaction
Drums
Water
Sorbent
Figure 4. Central condenser system, KPEG drum treatment demonstration, Bengart &
Memel, Buffalo, New York.
Table 2. Drum Scale Treatment of TCDD, 2,4-D and 2.4,5-7
Concentration in Treated Waste
Contaminant
TCDD (2,3,7,8)
2.4-D
2.4.5-T
Cone, in Untreated
Waste, ppm
1.3
17,800
2,800
First Treatment,
ppm
0.086 :
9,000
7,700
Second
Treatment, ppm
ND"
334
55
reroef/i
Reduction of
Contaminants
99.99
98.12
98.04
'ND = None Detected
(3260 to 3828 ppm). PCB concentrations
were reduced to an average concentra-
tion of 6.74 ppm {1.01 to 13.9 ppm),
which represents an average reduction of
99.84% (99.58% to 99.98%) with no
resolvable PCB congener exceeding 2
ppm (after retreatment). The process
implemented in Guam was successful in
its treatment of PCB-contaminated soil
without the use of DMSO or TMH.
The KPEG process used at the Bengart
& Memel site successfully reduced PCB
levels in soil contained in 55-gal drums to
below the 50-ppm control limit set for the
soil at the site.
The drum-scale KPEG process at the
Omaha site proved to be capable of
degrading 2,3.7,8-TCDD, 2,4-D, and
2,4,5-T. Initial concentrations of 1.3 ppm
2,3,7,8-TCDD, 17,800 ppm 2,4-D, and
2,800 ppm 2,4,5-T were reduced to con-
centrations of none detectable, 334 ppm,
and 55 ppm, respectively. The demon-
stration also proved the effectiveness of
KPEG without the use of DMSO or TMH.
Recommendations
The following recommendations were
made concerning the field-scale KPEG
system:
1. The system should be used at addi-
tional sites to provide more back-
ground data on the technology.
2. Additional field-scale demonstrations
should be done in an attempt to
optimize the reagent formulation and
operating parameters.
3. The labor intensity of the system
should be reduced by modifying it
with materials handling equipment
and automation.
4, Attempts should be made to identify
alternative equipment manufacturers
that may be capable of further
increasing the efficiency of the
system.
5. Design of a full-scale, fully portable,
self-supportive KPEG treatment
system should be initiated for soils
contaminated with halogenated aro-
matic compounds.
6. Efforts should be initiated to delist the
KPEG-treated soil judged to be
"clean" per the requirements of the
R&D permit.
The full report was submitted in ful-
fillment of Contract No. 68-03-3413, Work
Assignment No. 1-2, and IAG
RW17933910 by PEI Associates, Inc.,
under the sponsorship of the U.S. Envi-
ronmental Protection Agency.
-------
-------
M. L Taylor, J. A. Wentz, M. A. Dosani, W. Gallagher, and J, S. Greber are with PEI
Associates, Inc., Cincinnati, OH 45246.
T. D. Ferguson and C. J. Rogers are the EPA Project Officers (see below).
The complete reports entitled "Field Applications of the KPEG Process for Treating
Chlorinated Wastes" (Order No. PB 89-212 7241 AS: Cost: $17.00, subject to
change), and "Comprehensive Report on the KPEG Process for Treating
Chlorinated Wastes" (Order No. PB 90 163 643; Cost $23.00, subject to change)
will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, V'A 221'61
Telephone: 703-487-4650
The EPA Project Officers can be contacted at:
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268 ,
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/600/S2-90/026
------- |