&EPA
United States
Environmental Protection
Agency
EPA/540/SR-93/517
September 1993
SUPERFUND INNOVATIVE
TECHNOLOGY EVALUATION
Technology Demonstration
Summary
Toronto Harbour Commissioners
(THC) Soil Recycle Treatment
Train
A demonstration of the Toronto
Harbour Commissioners' (THC) Soil Re-
cycle Treatment Train was performed
under the Superfund Innovative Tech-
nology Evaluation (SITE) Program at a
pilot plant facility in Toronto, Ontario,
Canada. The Soil Recycle Treatment
Train, which consists of soil washing,
biological treatment, and metals chela-
tion, is designed to treat inorganic and
organic contaminants in soil.
During the demonstration test, soil
from a site that had been used for met-
als finishing and refinery and petro-
leum storage was processed in the pilot
plant. The demonstration test results
were mixed. The primary developer's
claim to produce gravel and sand that
met the THC target criteria for medium
to fine soil suitable for industrial/com-
mercial sites was achieved for the sand
and gravel products. The fine soil from
the biological treatment process exhib-
ited anomalous oil and grease behav-
ior and, although exhibiting a significant
reduction in polynuclear aromatic hy-
drocarbon (PAH) compounds, did not
meet the target level of 2.4 ppm for
benzo(a)pyrene.
This Technology Demonstration Sum-
mary was developed by EPA's Risk Re-
duction Engineering Laboratory,
Cincinnati, OH, to announce key find-
ings of the SITE Program demonstra-
tion that is fully documented in two
separate reports (see ordering infor-
mation at back).
Introduction
In response to the Superfund Amend-
ments and Reauthorization Act (SARA) of
1986, the U.S. Environmental Protection
Agency (EPA) established a formal pro-
gram called the Superfund Innovative
Technology Evaluation (SITE) Program.
The SITE Program was established to ac-
celerate the development, demonstration,
and implementation of innovative technolo-
gies at hazardous waste sites across the
country. The program is a joint effort be-
tween EPA's Office of Research and De-
velopment (ORD) and Office of Solid
Waste and Emergency Response
(OSWER). The purpose of the program is
to assist the development of hazardous
waste treatment technologies necessary
to implement new cleanup standards that
require greater reliance on permanent rem-
edies. This is done through technology
Printed on Recycled Paper
-------
demonstrations designed to provide engi-
neering and cost data on selected tech-
nologies.
The Toronto Harbour Commissioners
conducted an extensive evaluation of this
treatment train at a 55 tons/day pilot plant
located on the Toronto Harbour Front,
Toronto, Ontario, Canada. An EPA SITE
demonstration was conducted in April
1992. The SITE project examined, in de-
tali, the processing of soil from one of the
sites being evaluated as part of the over-
all project. The treatment train consists of
_,
Sand
Coal/
Contaminated
Soil
three processes shown conceptually in
ure 1. The first process uses an attrition
soil wash process to separate relatively
uncontaminated soil from a more heavily
contaminated fine slurry. The contaminated
fine slurry is then further processed in a
metals removal process or a bioslurry re-
actor process or both to remove organic
and heavy metal contamination. THC has
estimated that as much as 2.2 million tons
of soil from locations within the Toronto
Port Industrial District (PID) may require
some form of treatment because of heavy
metal, organic contamination, or both.
Gravel
Feed
Hopper
^ '
Trommel
Washer
V
Contaminated slurry to
biological treatment
Metals
Figure 1. Simplified process flow diagram of the Toronto Harbour Commissioners' Soil Recycle
Treatment train.
The THC claims that the treatment train
technology will meet the following perfor-
mance criteria:
1. Produce gravel (sized between 0.24
and 1.97 in.) and sand (sized be-
tween 0.0025 and 0.24 in.) from the
soil washer that will meet the THC
target criteria for coarse textured soils
described in Table 1 for both organic
and inorganic compounds indepen-
dent of the initial contaminant levels.
2. Produce a fine soil fraction (sized less
than 0.0025 in.) after metals removal
or biological treatment or both that
will meet the THC target criteria for
fine textured soils described in Table
1 for both organic and inorganic com-
pounds independent of the initial con-
taminant levels.
The THC criteria have been developed
by THC by combining existing criteria for
conventional pollutants and metals with a
site-specific criterion developed for a con-
taminated soil associated with a refinery
site.
The goals of this demonstration were to
evaluate the technical effectiveness and
economics of the treatment process se-
quence and to assess the potential appli-
cability of the process to other waste and/
or other Superfund and hazardous waste
sites. These and other specific critical and
noncritical objectives may be found in the
Demonstration Plan [1]. This Project Sum-
mary summarizes the treatment train's
ability to meet the THC target criteria.
Procedure
The demonstration took place while soil
from a site that had been used for metals
finishing and refinery and petroleum stor-
age was processed in the pilot plant. This
soil was expected to exhibit relatively high
organic (oil and grease, PAH compounds)
and inorganic (heavy metals) contami-
nants. EPA's sampling was of relatively
short duration, but it was expected that
when combined with results of THC evalu-
ation, a sound basis for analysis of the
technology would be obtained.
Soil was fed to the treatment train, and
all solid products were sampled. Samples
were also taken to allow separate assess-
ment of the performance of each process
technology, as required. In addition,
samples of recycled water streams and
air emission streams were obtained to de-
termine the fate of contaminants. Process
flow data were accumulated to allow the
development of an estimated mass bal-
ance for the soil wash process.
The data collected during the demon-
stration were used to determine the fol-
lowing:
-------
Table 1. THC Target Criteria for Selected Parameters for Soils to be Used for Commercial/Industrial
Lands
THC Target THC Target Medium
Parameter
Course Textured
So/7*
& Fine Textured
Soil
Feed'
Soil
Gravel* Sand"
Fine*
Soil
Conventional
Oil and grease (%)
TRPHS (mg/kg)
Total Metals (mg/kg)
Copper
Lead
Zinc
225
750
600
Organic Compounds (mg/kg)
Naphthalene 8.0"
Benzo(a)pyrene 2.4 **
300
1000
800
8.0"
2.4 ป
0.82
2500
18.3
115.0
82.5
11.2
1.9
0.33
800
0.22
620
2.5
5400
6.4 13.8 84
45.3 46.0 548
46.0 34.0 343
2.5
0.6
2.1
0.5
1.3"
(2.6) ฎ
Defined as greater than 70% sand and less than 17% organic matter.
Average of six composite samples.
Average of six samples from bioslurry reactor batch 2.
Total recoverable petroleum hydrocarbons.
If these trigger levels are exceeded, the Ministry of the Environment will make a determination on
a case-by-case basis regarding the need for remediation.
Values reported are estimated detection limits for this parameter.
Cleanup levels are shown for organic compounds. If soils exceed these levels, then the soil is
considered hazardous and remediation is required.
Values shown are below quantitation limits for procedures. Values shown are estimated.
the quality of the gravel, sand, and
fine soil relative to the THC target
criteria,
the percent removal for organic con-
taminants (oil and grease, total re-
coverable petroleum hydrocarbons
(TRPH), naphthalene, and benzo-
(a)pyrene) from gravel, sand, and con-
taminated slurry for the attrition soil
wash process,
the percent removal of organic con-
taminants from a soil slurry for the
bioslurry reactor process, and
the percent removal of heavy metals
(copper, nickel, lead, and zinc) for a
soil slurry being processed in the met-
als removal process.
Soil Wash Process
Because the majority of contaminated
soils encountered at the PID are sandy,
silty soils, soil washing (Figure 1) is an
economical and effective process to sepa-
rate contaminants from the bulk of the
soil. A scrubbing action and selected
chemicals are used to separate contami-
nants from the larger soil particles. The
rotary trommel washer removes particles
larger than 0.24 in. as a gravel fraction.
The contaminated soil particles less than
0.24 in. and the washwater pass through
the screen in the trommel washer into a
holding tank where belt-type oil skimmers
remove free oil from the water. The re-
maining soil and washwater are pumped
through a separation hydrocyclone where
the contaminated fines (less than 0.0025
in.) are separated from the coarser soil
particles. Larger sand particles are easily
separated from the fines, where the con-
taminants are concentrated. The fines are
pumped to a lamellar separator and then
to a gravity thickener, while the coarse
sand is pumped to the attrition scrubbers.
Three attrition scrubbing cells agitate
the soil particles, causing them to rub
against each other and scrub the fine par-
ticles and contaminants from the surfaces
of the soil particles. Detergents or surfac-
tants and acids or bases, if required, may
be added to the third cell to aid in dislodg-
ing contaminants from the soil particles or
in dissolving certain contaminants. The
treatment processes subsequently used
to treat the contaminated slurry restrict
the types of chemicals that can be used in
this treatment train.
Scrubbed particles and washwater from
the attrition scrubbing units are pumped
to a second hydrocyclone at the top of the
plant, where sand particles (>0.0025 in.)
are separated from the process water and
the remaining fines. The sand stream from
this separator is then put through a den-
sity separator to remove the light materi-
als, such as coal, wood, and peat particles
from the heavier soil particles. The coal
and peat are collected separately, as a
potentially contaminated waste stream.
The sand is discharged by conveyor to a
collection bin and is combined with the
gravel from the trommel washer for return
to the original site. This washed material
is expected to include approximately 70%
to 80% of the soil feed to the wash plant.
Contaminated fines with a grain size
smaller than 0.0025 in. pass through the
lamellar separator and sludge thickener to
remove water. The contaminated slurry
from the sludge thickener is fed into two
large holding tanks at the front end of the
metals removal system or directly to the
bioslurry reactor process. The contami-
nated slurry is expected to represent ap-
proximately 15% to 30% of the soil feed
to the wash plant.
The contaminated process water re-
moved by the lamellar separator and
sludge thickener is discharged to an out-
door storage pond for recycle. Any sludge
recovered from the ponds is then added
to the deep cone sludge thickener where
it joins the slurry for further treatment.
Bioslurry Reactor Process
The bioslurry reactor process (Figure 1)
involves a series of reactors (tanks) where
organic contaminants are treated. Before
introduction into the reactor, the slurry is
pretreated with a proprietary inorganic oxi-
dant.
The slurry to be treated is gently mixed
in two surge tanks to prevent particles
from settling and then pumped to one of
three 20,000-gal upflow bioreactor tanks
where submerged pumps and the upflow
of air from the medium to fine bubble
aerators provide constant mixing condi-
tions and the suspension of fines.
The biological system is prepared for
each soil to be treated by inoculating the
system with bacteria that have developed
in the soil; that is, a limited amount of fine
slurry obtained from the soil wash process
is pumped directly to the bioreactors with-
out passing through the metals extraction
process where the highly acidic condi-
tions would destroy the desired bacteria.
This allows a bacterial population in the
bioreactor to developa population based
on strains in the soil to be treated. Fine
slurry is accumulated until a single reactor
is fully charged.
Nutrients in the form of urea and phos-
phoric acid solutions are added periodi-
cally, and oxidants may be added.
Once the organics content of the slurry
is reduced to a level below the THC guide-
lines for industrial soils, the slurry is re-
turned to the excavation site since the
dewatering process originally selected did
not produce a solid product.
The developer of this treatment train
had planned to use a continuous process
for the biological treatment system, but
earlier experience with the bioslurry reac-
-------
tor process disclosed variable analytical
results. Because the developer had not
moved beyond batch evaluations at the
time of the SITE demonstration sampling,
the discharge from two bioslurry reactor
batches was extensively sampled.
Metals Removal Process
The contaminated slurry from the attri-
tion wash plant or the bioslurry reactor
process is fed into two large holding tanks
in the central area of the facility, at the
front end of the metals removal process
(Figure 1). The slurry consists of approxi-
mately 24% solids by weight and 76%
process water. Mild acid is added to the
slurry from the acid storage tanks to des-
orb and solubilize any metal contaminants
from the soil particles.
The contaminated slurry is then pumped
Into the first tubular reactor. This screw-
type rotary reactor brings the slurry into
countercurrent contact with solid metals
chelating agents that have an affinity for
specific metal contaminants. From here,
the slurry, which now contains only soil
particles, organic contaminants, residual
metals not removed by the process, and
process water, is pumped to a holding
tank, where it is neutralized.
The solid chelating agent, which moves
countercurrent to the slurry, now contains
the extracted metals. The solid chelating
material is selected from a family of metal-
specific, ion-exchange chelating resins to
preferentially remove heavy metals. It is
washed to remove solid soil particles and
is fed through a second tubular reactor
where a mild acid breaks the bond be-
tween the chelating agent and the con-
taminant metals. The chelating agent is
then recycled to the first reactor for reuse
in the metals extraction process. Mean-
while, the metals/acid mixture is recycled
in the second reactor until it becomes
sufficiently rich in metal to be pumped to
an electrowinnlng unit, where the metals
are removed by electrolysis. The result of
the buildup of metal concentration in the
regenerating acid will somewhat reduce
the absorption capacity of the resin beads
being returned to the slurry contractor.
The system has, however, been designed
to provide excess absorption capacity of
the chelating resin in relationship to the
metals being absorbed. Little change in
performance is expected. Another, more
long-term deterioration of the resins ab-
sorption capacity is associated with the
oxidation of the active sites on the resin
bead. This reaction is expected to be mea-
sured In months to years and should not
affect this demonstration. Nevertheless,
resin replacement costs can be a signifi-,
cant cost factor in such a system. The
metals may be removed singly or as one
composite mass. During electrowinning,
the metal-depleted acid is pumped back
to the holding tanks for reuse as regener-
ating acid, or it may be neutralized and
become a part of contaminated slurry.
Results and Discussion
Soil Product Criterion
The gravel (sized between 0.24 and
1.97 in.) and sand (sized between
0.0025 and 0.24 in.) products met the
THC target criteria.
Fine soil did not .meet the THC target
criteria because oil and grease and
benzp(a)pyrene levels exceeded the
criteria.
Soil Wash Process
For gravel, removal rates for organic
contaminants (oil and grease, TRPH,
naphthalene, and benzo(a)pyrene)
were 67% or greater. This gravel ac-
counted for 11.5% of total process
mass output and 4% or less of the
organic contaminants in the product
streams.
For sand, removal rates for organic
contaminants (oil and grease, TRPH,
naphthalene, and benzo(a)pyrene)
were 78% or greater. This sand ac-
counted for about 68% of the process
output and 15% or less of the organic
contaminants in the product stream.
The process concentrated the organic
contaminants into a contaminated fine
slurry (<0.0025 in.) which accounted
for about 19% of the process output
mass and 74% or more of the or-
ganic contaminants.
The process also produced a con-
taminated coal/peat product (<0.24 in;
>0.00025 in) that represented about
1.6% of the process output and 6%
or more of the organic contaminants.
This waste stream will require dis-
posal (most likely by incineration).
The feed soil exhibited low heavy-
metals contaminant levels (copper, 18
ppm; lead, 115 ppm; and zinc, 83
ppm). The wash process concentrated
these contaminants in the fine slurry
(19% of the process mass output and
59% or more in the process output
streams).'
Data from the soil wash process that
were developed during the SITE demon-
stration are presented in Table 2.
Bioslurry Reactor Process
When inlet samples were compared
with outlet samples, the oil and grease
reduction from the bioslurry process
was limited.
' A similar comparison for other pa-
rameters yielded the following reduc-
tion: TRPH, 52%; naphthalene, at
least 97%; and benzo(a)pyrene, ap-
proximately 70%.
Results for the bioslurry reactor pro-
cess are shown in Table 3.
Metals Removal Process
The levels of metal contamination ac-
tually encountered eliminated the need
to use the metals removal process for
this soil. Limited data were developed
for the efficiency of the metals re-
moval process by sampling a process
run of a metal-rich slurry from an-
other soil. The reactor achieved the
following removal efficiencies based
on metals concentrations in the inlet
versus the outlet samples: copper,
96%; lead, 71%; nickel, 71%; and
zinc, 63%.
Because the metals removal process
became fouled with oil and grease,
the operation shut down prematurely.
This may be a limitation on the pro-
cess in that slurries with free oil and
grease cannot be processed.
Results for the metals removal process
are shown in Table 4.
Fine Product Dewatering
Because the hydrocyclone device
used for final dewatering of the fine
soil was not successful, the final prod-
uct from the process was a slurry.
Dewatering will require further evalu-
ation by the developer or the applica-
tion of other technology.
Emissions Assessment
Emission sampling of the ventilation
system serving the biological treat-
ment system did not detect PAH com-
pounds, but detection limits were very
high due to a high concentration of
light hydrocarbons in the exhaust
stream. These light hydrocarbons
. were tentatively identified as a pe-
troleum distillate in the range be-
tween diesel oil to Stoddard solvent
(C9 - C16 paraffins). Total gaseous,
nonmethane organic compounds
were detected at levels that indicated
220 Ib/day of emissions. (The data
illustrate that significant air stripping
is occurring in the bioreactor, and this
must be accounted for in the design.)
The facility has a biological filter sys-
tem and carbon adsorption bed in
place to control these emissions.
-------
Table 2. Selected Feed and Product Characteristics of the Attrition Soil Wash Process
Characteristic
Feed
Soil'
THC
Criteria
Gravel, *
<1.97in.
X).24 in.
Coal/Peat
Fraction, *
<0.24 in.
Sand, *
<0.24 in.
>0.0025in.
Contam.
Fines, *
<0.0025 in.
Percent of output based
on site demo data
Percent of output based
on THC overall analysis
11.5
10.5
1.6
2.5
68.1
70.2
18.8
16.7
Oil, grease (mg/kg)
TRPH (mg/kg)
Copper (mg/kg)
Lead (mg/kg)
Zinc (mg/kg)
Naphthalene (mg/kg)
Benzo(a)pyrene
(mgfcg)
8,200
(6,708-9,700)*
2,500
(2,270-3,430)
18.3
(9.2-42.2)
115 '
(63.3-127)
82.5
(40.4-181) '
11.2
(5.3-18)
1.9
(0.9-2.9)
10,000
225
750
600
8
2.4
3,300
(1,200-10,400)
800
(270-1,370)
6.4
(0.7-12.1)
45.3
(3.2-117)
46
(2.3-98.6)
2.5
(0.9-2.9)
0.6
(0.2-1.0)
38,000
(17,600-51,600)
11,900
(4,760-16,280)
32.9
(22.8-41.7)
406
(12.9-749
210
(46.8-406)
64
(34-110)
14.5
(9.6-23)
2,200
(1,400-3,900)
620
(380-960)
13.8
(32-32.4)
46
(23.6-82.9)
34.1
(15.9-71.4)
2.1
(1.5-3.1)
0.5
(03-1.2)
40.000
(26,900-50,500)
14.000
(8,500-19,800)
83.1
(48.2-135)
522
(421-680)
344
(192-593)
51.7
(17-82)
10.0
(9.0-12.0)
Average of six composite samples.
* Average of three composite samples.
* Range of results.
Conclusions
The results of the SITE Demonstration
Test showed that
Soil washing effectively produced
clean, coarse soil fractions and con-
centrated the contaminants in the fine
slurry.
The chemical treatment process and
biological slurry reactors achieved at
least a 90% reduction in simple PAH
compounds such as naphthalene but
Table 3. Selected Feed and Product Characteristics of the Bioslurry Reactor Process
Characteristic
Oil, grease
TRPH
Naphthalene
Benzo(a)pyrene
MOE
Criteria
1.0%
8 mg/kg
2.4 mg/kg
Contaminated
Fine
Slurry '
4.00%
(2.7-5.4) *
1.4%
(.85-1.98)
5 1.7 mg/kg
(17-82)
10 mg/kg
(8.4-12)
Bioslurry
Reactor
Batch 1 *
4.98%
(3.96-6.08)
.78%
(68-.9S)
<14 "mg/kg
3. 1 mg/kg
(2.0-5.1)
Bioslurry
Reactor
Batch 2*
2.53%
(3.98-2.17)
.54%
(.39-.76)
<1 3 "mg/kg
<16~i11
2.6 mg/kg
(2.3-3.4)
Removal
Efficiency
%*
6
52
97"
71
Average of six composite samples.
Average of 6 samples taken at20-min intervals during discharge of batch.
Removal efficiency based on average value for both batches.
Range of results.
Value reported is average ofquantitation limit reported. Detection limit is at least a factor 10 less
than the quantitation limit.
Removal efficiency calculated from detection limit; estimated by dividing quantitation limit by 10.
fell just short of the approximately
75% reduction in benzo(a)pyrene re-
quired to achieve the THC criteria.
The biological process discharge did
not meet the THC criteria for oil and
grease, and the process exhibited vir-
tually no removal of this parameter.
The developer believes that the high
outlet oil and grease values are the
result of the analytical extraction of
the biomass developed during the pro-
cess.
The hydrocyclone dewatering device
did not achieve significant dewater-
ing. Final process slurries were re-
turned to the excavation site in liquid
form. The development of an accept-
able dewatering process will require
further evaluation of alternative tech-
nologies.
The metals removal process equip-
ment and chelating agent were fouled
by free oil and grease contamination,
forcing the premature curtailment of
sampling. This establishes a limita-
tion for this technology since biologi-
cal treatment or physical separation
of oil and grease will be required to
avoid such fouling.
-------
Tablo 4. Selected Heavy Metals Date for
Removal of Metals from the Liquid Stream by
the THC Metals Removal Process
Influent
Metal mgfag
Copper 51.1
(492-532)'
Lead 100.5
(942-112)
Nickel 11.7
(10.7-12.7)
Zinc 277
(264-294)
Effluent
mg/kg Removal %
1.8 96
(0.9-3.0)
29.0 71
(13.5-46)
3.3 71
(0.9-7.3)
101 63
(53-183)
References
1. Science Applications International
Corporation. March 16, 1992. "Dem-
onstration Plan for the Toronto
Harbour Commissioners (THC) Soil
Recycle Treatment Train." .
Range of values.
.S. GOVERNMENT PRINTING OFFICE: 1993 - 750-071/80051
-------
-------
The EPA Project Manager, Teri Richardson, is with the Risk Reduction
Engineering Laboratory, Cincinnati, OH 45268 (see below)
The complete report, entitled "Technology Evaluation Report; Toronto Harbour
Commissioners (THC) Soil Recycle Treatment Train," (Order No. PB93-
216087; Cost: $27.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
A related report, entitled "Toronto Harbour Commissioners (THC) Soil Recycle
Treatment Train; Applications Analysis Report," EPA/540/AR-93/517, dis-
cusses application and costs.
The EPA Project Manager 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
Official Business
Penalty for Private Use
$300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/540/SR-93/517
------- |