-------
5.2
Results and Discussion
In figure 7 the removal results are compiled based on soil vapour
analyses.
Total amount
removed (kg)
4000
3000-•
2000 ••
1000-
Vaporlzed and blodegraded
(3)
Time (weeks)
Figure 7: Cumulative Amounts of Gasoline Removed During SVE
Based upon the speed of the oxygen consumption and the carbon
dioxide production the biodegradation rate of gasoline can be
estimated to be 7 mg/kg soil/day (3 sampling days, week 30-50).
The in situ biodegradation seems to take place by zero order
kinetics.
The hydrocarbon concentrations in the withdrawn soil vapour varied
enormously. Initial concentrations of up to 80 g/m3 were measured.
After 20 weeks the concentration dropped to 3" g/m3 and gradually
decreased until 240 mg/m3 (week 60). Due to the dimensions of the
soil vapour treatment system the withdrawn flow of the soil vapour
could gradually be increased from 25 Nm3/h to 50 Nm3/h in week 36
and until 63 Nra3/h in week 46. The hydraulic residence time of
(ambient) air in the soil is approximately 1 hour.
In tabel 2 the biological activity is given measured with total
plate counts (T.P.C.) plate counts with gasoline degradable micro-
organisms and the oxygen uptahe rate (O.U.R.) at different depths.
92
-------
label 2 Measured concentrations and biological activity
Depth
(m)
0.5
0.9
1.5
2.0
2.5
2,7
non vol.
oil
ing/kg dw
100
2.700
130
vol.
oil
mg/kg dw
1.200
1.300
<100
T.P.C.
CFU/g soil
2,7x103
0,7x106
4,6x106
0,3x10*
gasoline
degr.
CFU/g soil
3.4x104
1,6x106
0.9x106
5.6x106
0.4x106
O.U.R.
mg02/kg.d /
3,6
1,3
6,8
18
34
14
CFU * Colony Forming Unit
The highest activity (T.P.C., O.U.R. and gasoline degraders) and
the highest concentration left in the ground is at the groundwater
surface.
5.3 Progress
After being operational for more than ^year there are still soil
samples which show relatively high gasoline concentrations. This
might be as a result of soil heterogenities or stagnant areas
(aerodynamically). In order to speed up the remedial action the
groundwater level will be discontinuously lowered and the flow
rate of the separate SVE wells will be varied. Additionally, hot
air (30-40"C) will be infiltrated at the other side of the road.
93
-------
ON SITE APPLICATION OF SVE - SITE 3
6.1
Introduction
Should an in situ application of SVE not be suitable for whatever
reason, an on site application of SVE can be a good alternative to
other soil treatment methods. At this site the soil, contaminated
with perchloroethene, had to be excavated. As an alternative to
washing or disposing, TAUW Infra Consult B.V. designed an on site
SVE depot treatment system.
The main aim was to perform a soil treatment at relatively low
costs at an industrial site and to have the treated soil re-used
at the same site.
6.2
Site Characteristics
After excavation the soil was stored at a SVE depot. A cross-
section is given in figure 8.
£
> i
1 1 ' 1
1 1 ' 1
J
J L
1 - 1
1 — 1 — 1 — 1
i — i — i
4*
i^^^
i-
Figure 8: Crossection of SVE depot
The depot consists of several vapour extraction drains and a water
drainage system. Percolated water was treated in a water treatment
system.
The soil, deposited into two depots with different concentrations
of contaminanants. Depot A consisted of sand and humous sand and
was contaminated with perchloroethene 87 mg/kg (measured in the
depot). Two depots were constructed, each containing 2200 tons.
94
-------
The depots were treated in succession so only one air pump and
activated carbon unit was necessary. The experiences achieved with
the SVE at depot A could be used to optimally treat depot B at
minimum costs. An important factor was the estimated duration of
the SVE. A good estimate for the duration of treatment at depot B
could be made using the "spread sheet" model of TAUW Infra Consult
B.V. and during treatment the prediction was easy to verify.
6.3 Results and Discussion
The results of the SVE in depot A are given in figure 9 and 10.
Within 5 months 95 kg of perchloroethene was withdrawn. The under
pressure applied was 40 mbar and the air flow was approximately
60 tn3/hour.
After 5 months the SVE was concentrated on the humous sand because
concentrations did not decrease anymore. It appeared that it was
not possible to further diminish the concentrations by increasing
the air flow during 5 months.
With these data the treatment at depot B was performed with mini-
mum monitoring costs and under optimal conditions. Within 4 months
31 kg per chloroethene was withdrawn and 99.3X of the contam-
ination was removed.
1»OO
1MO
1700
1«OO
1500
two
1X0
1200
1100
1000
too
•00
700
•00
900
400
MO
900
100
0
PROJECT WENUM
CO**CCKHUT»CW PP»O«.O*OOHO* « «t
Figure 9: Concentration in the
withdrawn soil vapour
Jwi-ll .M-»l Jtet*M $*>#-t1
O UC/U3 WITHOtUWN «M
Figur 10: Total amount of
Perchloroethene withdrawn
95
PROJECT WENUM
imvum PWCH.OKOCTMCMC <«)
D KC PC*O. WTMCNUWM
-------
In Table 3 the measured soil contamination in depot A is given
prior to SVE and after 2, 5 and 10 months upon completion.
Table 3; Concentret!on of Perchoroethene in the Soil at Depot A (mq/kq d.m.)
Drain
1
2
3
«
5
Average
0 Months
256
35
12
16
114
87
2 Months
4
-
2
-
32
13
5 Months
1.3
1.7
0.16
1.0
1.9
1.2
10 Months
1.9
0.9
0.22
0.6
3.1
1.3
X Removed
99.3X
97.4X
98. 2X
96.3X
97.3X
98. 5X
As was previously mentioned the situation was ideal to verify the
model of the duration of the SVE. Figure 11. gives the measured
amount and calculated amount in the withdrawn soil vapour.
60 -r
SOIL VAPOUR EXTRACTION
PERCHLOROETHENE
*o
I
1C -
0 20 40 60 SO 100 120 140 160 180 200
TIME (DAVS)
Figure 11; Measured (4) and calculated (-) amount
96
-------
The measured concentrations match quite good with the calculated
concentrations in this situation and the model can be used to
estimate the duration of the SVE.
The costs for this treatment were $ 30-40/ton, including liners,
air treatment and monitoring.
The sandy soil has been re-used as a stabilizing layer under the
car park on site, with the approval of the Province of Gelderland.
97
-------
CONCLUSIONS AND RECOMMENDATIONS
In soil remediation the application of the in situ SVE technique
can often be a good alternative to conventional remedial tech-
niques. The removal of volatile contaminants can be done effect-
ively even in less permeable soils such as loamy sands. The SVE
technique becomes even more feasible when air based enhanced bio-
degradation is applied. Important advantages of SVE in situ biode-
gradation are the reduced costs for on-site treatment for with-
drawn soil vapour and the breakdown of less volatile (hydrocarbon)
contaminants in the subsoil.
For an accurate estimate of the costs involved for the SVE it is
important to calculate the duration of the SVE. The application of
computer simulations are therefore necessary. More attention
should be paid to the further development of these models. In
order to validate the simulation it is recommended to carry out
column tests with contaminated soil from the site itself.
The treatment costs of the withdrawn soil vapour are substantial,
therefore the search for alternative techniques must be continued.
TAUW Infra Consult B.V. have successfully applied a combined soil
vapour and groundwater treatment at a gasoline polluted site. The
treatment is based on biodegradation.
SVE is a simple treatment system at low costs, able to treat the
contaminated soil in-situ or allowing the soil to be re-used on
the same site after on site treatment. So even if excavation is
necessary SVE is still a good alternative to treating the soil.
98
-------
LITERATURE
Miller, R.N. , et al. "A Field Scale Investigation of Enhanced
Petrol Hydrocarbon Biodegradation in the Vadose Zone at Tyndall
AFB, Florida", Proceedings NATO/CCMS meeting France, December
1990.
Hutzel N.J., et al. "State of Technology Review, Soil Vapour Ex-
traction System", EPA/600/2-89/024, 1989.
Hinchee, R.E. Miller R.N. "Bioreclamation of Hydrocarbons in the
Unsaturated Zone, in Hazardous Waste Management Contaminated
Sites, and Industrial Risk Assessment" et ed. by W. Pillmann and
K. Zirm, Vienna 1990, 641-650.
Schutzle, R., R. Casanova, R. Rentemann and H. Finke (1989).
Praxisbezogene Grundlagen und Kriterien fur eine schadensfallge-
rechte Anwendung der Bodenluftabsaugung. Band 5 Landesanstal Fur
Umweltschutz Badenwurttemberg.
99
-------
PH0NIX MIU0
A/S Phanix Contractors
Fuglesangsalle 14
DK-6600 Vejen, Denmark
Tel.: +45 75 36 11 11
Fax: +45 75 36 46 09
PH0NIX MIU0
MOBILE EXTRACTION TECHNOLOGY
FOR ON-SITE SOIL DECONTAMINATION
CONTEX SYSTEM
Prepared by:
Steen Vedby
and Allan Skov
June 1, 1991
100
-------
1. ABSTRACT
Ph0nix Milj0 has developed new technology for
decontamination of soil from industrial sites and
material from drilling operations.
Three years of extensive development work has
resulted in a continuous extraction process, the
Contex process. The Contex System is arranged in a
semi-mobile containerized plant designed to
operate on the site of the contamination.
The Contex process makes it possible to purify
contaminated soils containing all organic con-
taminants, as well as some forms of heavy metals.
The extraction takes place in a closed circuit
including regeneration and complete recycling of
the extraction liquid. The extraction liquid is
completely removed from the. material in a steam-
distillation step followed by efficient recovery
of the extraction liquid.
The first generation of the process was based on
batch technology. During 1990 the continuous ver-
sion of the process, the Contex System, was
developed, offering high capacity and attractive
process economy.
A documentation program has demonstrated the
plant's decontamination efficiency for a number of
soil types from Denmark, West Germany, and Norway,
and for very different contamination sources.
The successful tests have included soil contamina-
tion ranging from solvents over oil products to
heavy tar fractions. These trials prove that the
plant can be used to combat very concentrated
pollution and that the process can meet the strict-
est decontamination requirements. -
Concurrently with technical tests, the authorities'
approval and comprehensive safety analyses, in-
cluding the fire and emission risks, have taken
place as documentation.
101
-------
The cost of operation is competitive compared to
other technologies but dependent on type of soil
and specific contamination.
2. PROCESS DESCRIPTION
The Contex System is a high-capacity technology
which cleans 10 tonnes of contaminated soil per
hour. The technology is particularly efficient
against heavy organic contamination. The cleaning
process is characterized by simplicity and con-
sists of a combined liquid extraction and steam-
stripping process in a closed loop, figure 1.
Fig 1
EXTRACTION + STEAMSTRiPPING = DECONTAMINATED SOIL
Conlominenl
for deitrudion/
recycling
102
-------
1. The soil is screened (mesh size 5-8 cm) and led
by belt conveyors to the inlet lock of the
plant situated on top of the extraction unit.
2. The contaiminated soil is now led via the lock
and into the extraction unit. Shaftless augers
transport the soil through a tube system (cf.
fig 1).
3. In the extraction unit, the extraction liquid
(dichloromethane) flows through the soil
dissolving and washing out the contaminants
from the soil in a counter current extraction
process.
The extraction process is adapted to each
specific case, in a way that .the contamination -
is totally removed, i.a. by adjusting the soil
feed rate as well as the flow rate of the
extraction liquid.
4. The soil passes through a drainage screw which
drains the extraction liquid out of the soil by
gravitation.
5. Following the extraction step, the soil is
passed through a steam-stripping unit in which
residual extraction liquid is removed from the
soil by heating up to approx. 212°F/100°C.
Complete removal of extraction liquid is
ensured by steam injection in a. polishing step.
The steam is generated in a separate steam
unit.
Hereupon, the decontaminated soil leaves the
plant through an outlet lock.
6*. During the decontamination in the extraction
unit, the extraction liquid containing the dis-
solved contaminants is led to a distilling
unit. The extraction liquid is regenerated by
distillation and recycled. The dichloromethane
vapor phase is condensed and led to the buffer
tank from which all of the extraction liquid
can be re-used.
J
7. The contaminants have now been isolated in the
distillling unit as an aqueous emulsion. This
emulsion (remanence) is pumped into tanks which
are transported to an authorized treatment
103
-------
plant for harzardous waste for destruction. In a
considerable number of cases, the remanence is
regenerated by further processing or refine-
ment . ,
8: The Contex System is equipped with an emission
control system which efficiently prevents
. emission to the environment. The general concept
- is that leaks will cause flow of seal water or
air into the process with subsequent detection
(alarm).
A central vent system connects to an emission
guard consisting of a freezing trap and an ab-
sorption .step. :
3. DEVELOPMENT STATUS
The Contex System represents the 4th process gene-
ration based on continuous liquid extraction with
dichloromethane followed by steam-stripping.
In the first three generations, the cleaning
process was based on batch cleaning using dif-
ferent washing tanks with various handling and
mixing equipment.
The first two generations were developed from 1987
to 1989, mainly as pilot plants for research and
development of the new extraction technology.
• The 3rd generation was operating until the end of
1990 and was equipped with a batch cleaning tank
with a capacity of 1-5 t/h soil.
The new Contex System (4th generation) was opera-
tive in February 1991 and has by now been installed
in a new receiving and development facility at
Ph0nix' premises in Denmark. The test run results
have proved the capability of this new Contex
process which by now is starting to operate
commercially.
The Contex System is covered by patents.
104
-------
MEDIA TREATED
The Contex extraction plant has basically been
developed to clean contaminated soil with textures
from sandy soils to heavy clay soils.
However, the use of screw conveyors in combination
with the actual extraction liquid makes it pos-
sible to treat even rather inhomogeneous material,
such as soils containing gravel and stones,
organic soils, drilling mud, etc.
Fragments above 10-20 cm are treated in a separate
container with a batch washing tank, using the
same extraction and steam-stripping technology.
This washing tank has another potential for re-
moval of oil and fat from metal fragments or
scrub.
Several test results have proved the efficiency of
the Contex System for different soil textures. In
figure 2 different soil texture classes are shown
for which the efficiency of the technology has
been demonstrated.
Fig 2
100
JO/SANDY ^CLAY ^^
LOAM/ \# \ / \
-7^—-^--fe-^M —
t|ANpY/LOAy^ \< / \
Percent sand
Soil texture classes (U.S.D.A.) for which the
Contex System has proved efficient.
105
-------
5. POLLUTANTS TREATED - INITIAL AND FINAL POLLUTANT
CONCENTRATION
The Contex System is based on dichloromethane
extraction which is efficient against almost all
kinds of organic contaminant, such as oil, gaso-
line, solvents, tar, phenols, a number of pesti-
cides, polychlorinated biphenyls (PCBs), etc.
During the past period of development and opera-
tion, the extraction technology has been tested on
diffent soil textures and types of organic con-
taminant.
The1 results of all these, experiments are presented
in Table 1 which shows the maximum content of
pollutant before and after treatment for selected
types of contaminants and solvents.
Table 1
Test Resuffs
Tor
Oil products:
•Jel fuel/gosoline
- Fuel oil/diesel oil
- Heovy fuel/oil sludge
• Drilling cuttings
Chlorinated solvenfs
Max. value
Before cleaning
(ppm)
270,000
1 5,000
22,000
30,000
400,000
3,600
Afler cleoning
(ppm)
<30
<10
<10
<30
<60
<1
Specific components:
BTX compounds
Naphthalene
Phenonthrene
5,000 | <1.5
5,300
23,000
<1.5
<1.0
106
-------
All the results indicated are based on samples
taken by an independent laboratory using gas-
chromatography after extraction with dichloro-
methane or freon.
The results obtained are in general sufficient
according to the standards of the Danish autho-
rities, as well as to the limit values in many
countries. Moreover, the process can be adjusted
to bring the residual concentration of any speci-
fic component or hazardous material below a
specific value (Table 1).
6. RATED THROUGHPUT
The treatment capacity of the Contex System varies
between 10-20 tonnes contaminated soil per hour,
dependent on soil texture, water content, type of
contaminant, initial and final pollutant concen-
tration, etc.
In general, the capacity is higher in a dry and
sandy- soil with volatile components compared with
a wet and heavy clay soil with non-volatile ;
components such a tar.
The capacity in the washing tank for decontaminated
large stones, concrete and metal fragments is
about 2 tonnes per hour.
All containers, including the soil treatment units
and supply units, may be disassembled and trans-
ported to the contaminated site and operating in
the course of a few days.
7. PROCESS LIMITATIONS
The new Contex System is primarily developed to
extract hazardous organic chemicals from different
soils and soil materials (such as i:a. drilling
mud) .
On the other hand, several other applications are
possible at this stage and in the future. Present
efforts are focused on adapting the Contex process
to treat sludge material.
107
-------
Furthermore, research and development concerning
the; treatment of inorganic components, including
heavy metals and cyanides, is taking place at pre-
sent and Ph0nix Milj0 also expect to be able to
treat these kinds of contamination in the future.
8. PROCESS WASTE STREAMS
The only waste material produced during the opera-
tion of the Contex System is a concentrated
residue of contaminants (emulsion) or a rema-
nence. This residue has to be transported to an
authorized receiving station or may be re-used
after a work-up process.
Water which enters the process with the polluted
material is typically used to emulsify the pollu-
tant concentrate. In cases of excessive water
content, the water phase can be taken out as a
separate stream for treatment in a standard waste
water unit.
In general, the treated'soil is unaffected by the
cleaning process. The soil can therefore be used
as filler or topsoil in the excavated areas.
TREATMENT COST
The cost of operation is competitive compared with
other technologies but varies dependent on type of
soil and specific contamination".1
In general, easily extractable components in sandy
soils have the lowest treatment'cost whereas the
decontamination of heavy clay soils and drilling
mud with high initial concentrations of tar
compone'nts, heavy complexes, etc. is somewhat more
expensive.
The: price of on-site treatment will normally vary
in the range from USD 100-1000 per tonne treated
soil.
108
-------
ZERO AIR EMISSIONS
GROUNDWATER AND SOIL REMEDIATION
USING THE
AWD INTEGRATED SYSTEM
A SITE Project in Operation for 3% Years
Selected for the Superfund Innovative
Technology Evaluation Program
Robert G. Hornsby, PE
AWD Technologies, Inc.
Ron Helgerson
Lockheed Aeronautical Systems Company
109
-------
ZERO AIR EMISSIONS
GROUNDWATER AND SOIL REMEDIATION
USING THE AWD INTEGRATED SYSTEM
A SITE Project in Operation for 3V£ Years
Robert G. Hornsby, PE
AWD Technologies, Inc.
Ron Helgerson
Lockheed Aeronautical Systems Company
INTRODUCTION
The Lockheed Aeronautical Systems Company (LASC) has over 200 /acres of aircraft
manufacturing facilities located in Burbank, California. Among the famous aircraft that have
been assembled at this facility are the P-38 Lightning, the F-104 Starflghter, the U-2, and the
L-1011.
In late 1987, solvent-contaminated soil and groundwater were identified near Building
175. As a result, the Los Angeles Regional Water Quality Control Board (RWQCB) issued a
Cleanup and Abatement order requiring soil and groundwater remediation to commence by
August 1, 1988, and September 15, 1988 respectively.
LASC selected AWD Technologies, Inc. (AWD) to design, install, and operate a
treatment facility to meet the requirements of the RWQCB. AWD, a wholly owned
subsidiary of The Dow Chemical Company provides a comprehensive range of services for
remediation of contaminated soil and groundwater and can draw upon the specialized
resources and expertise of its owner company.
110
-------
THE TECHNOLOGIES
Two technologies were integrated in an innovative way: AquaDetox, a low-pressure
steam stripping technology developed by Dow Chemical to extract volatile organic
compounds (VOCs) from the groundwater, and Soil Vapor Extraction (SVE) treatment of the
VOCs in the vadose zone. Each process discussed in the following paragraphs would be
effective if used independently. Integrating the two represented a significant operating cost
reduction. The integrated process is discussed in a separate section.
AquaDetox
Over the past several years, an effort has been under way to improve the efficiency of
air stripping in removing contaminants from groundwater. This work has led to the
development of the AquaDetox technology, which surpasses more conventional approaches to
air stripping in terms of reduction efficiency. In most cases, AquaDetox can reduce
contaminants in groundwater to below Maximum Contaminant Levels (MCLs) without liquid-
phase carbon bed treatment. Moderate vacuum and deep vacuum AquaDetox steam stripping
go even further, allowing the near total recovery of contaminants for possible recycling.
AquaDetox technology can be used to remove a wide variety of volatile compounds,
and many compounds that are normally considered "non-strippable" (i.e., those with boiling
points in excess of 200°C). The application of AquaDetox for the removal of compounds
with boiling points greater than 200°C and the use of vacuum are patented by The Dow
Chemical Company.
Stripping is commonly defined as a process to remove dissolved, volatile compounds
from water. A carrier gas, such as air or steam, is purged through the contaminated water,
111
-------
with the volatile components being transferred from the water into the gas phase. While the
physical principles involved are straightforward, the practice of stripping has undergone
considerable development since the early 70s.
Dow's effort has focused on:
1. Development vof the proper technical relationships that provide a clear understanding
of the stripping process.
2. Application of these relationships, along with the correct hardware, to attain higher
(
levels of contaminant removal than previously possible.
3. Development of the proper scale-up parameters to go from pilot units handling < 1
gpm to production units handling over 3000 gpm.
4. Development of the conditions under which compounds with very high boiling points
(e.g., 200°C) can be stripped from water.
5. Compilation of a vapor-liquid equilibrium data base with special emphasis on EPA
priority pollutants.
The effort necessary to address these criteria has been carried out by the Separations
Section of the Applied Science and Technology Department of Dow. The research and
development has been under the direction of Dr. Lanny Robbins.
By the early 1980s, the result of this effort was the AquaDetox process, an innovative
technology for the high efficiency stripping of organic contaminants from water.
AquaDetox is capable of effectively stripping over 90 of the 110 volatile compounds
listed in CFR 40, July 1, 1986, by the EPA (see Table 1). The ability of AquaDetox to
efficiently attain low levels of contamination in the effluent represents a major breakthrough.
112
-------
Conventional strippers will normally achieve only 95 to 98% removal of the contamination,
whereas AquaDetox can achieve up to 99.99%.
There are three versions of the basic AquaDetox technology:
1. Air Stripping AquaDetox.
2. Moderate Vacuum AquaDetox (requires source of steam).
3. Deep Vacuum AquaDetox (does not require source of steam).
Typical schematic flow diagrams for each of the types of AquaDetox technology are
included in the paper by Street, Robbins, and Clark (Reference Report).
The AquaDetox air stripper system can eliminate the need to carbon polish the
effluent water, which conventional air strippers sometimes require. Both need carbon beds
to clean the contaminated air before discharging it to the atmosphere. The AquaDetox steam
stripper (moderate or deep vacuum) condenses the contaminated steam to form a multi-phase
liquid from which the liquid phase contaminants can be decanted for possible recycling.
Only a small stream of non-condensable gases is emitted following carbon treatment.
Soil Vapor Extraction
Soil vapor extraction (SVE) is a technology commonly applied for the in-situ removal
of VOCs from soil. A vacuum is applied to vadose zone extraction wells to induce air flows
within the soil toward the wells. The air acts as a stripping medium which volatilizes the
VOCs in the soil. Soil-gas from the extraction wells is typically treated in carbon beds
before release to the atmosphere. Alternatively, the treated soil-gas is reinjected in the soil
to control the direction of air flow in the soil.
113
-------
THE PROJECT
On February 1, 1988, LASC awarded AWD a contract for pilot testing, design, and
installation of an integrated 1200 gpm groundwater treatment plant and 300 SCFM SVE
system. Fast-track project techniques were used, and seven and half months later all systems
of the $4 million project were operational.
AWD was responsible for the permitting, design, construction and operation of the
system. The effluent met and continues to meet drinking water standards for VOCs.
Integrated System
The integrated system consists of two basic processes: an AquaDetox vacuum
stripping tower using low pressure steam, and a soil-gas vapor extraction/reinjection process.
The system removes VOCs from the groundwater and soil with no gaseous emissions to the
atmosphere. Figure 1 shows a schematic flow diagram of the integrated system.
Integrating the two technologies creates a unique system. While the AquaDetox
system extracts and treats contaminated groundwater, an array of SVE wells removes
contaminated,soil-gas from the vadose zone. The soil-gas treated b;y the carbon beds is
readmitted into the ground to sweep through the soil and remove additional contamination.
The AquaDetox and SVE systems share a 3-bed granulated activated carbon (GAG)
unit. When one of the GAG beds is regenerated, the steam and organic vapors are
condensed in the secondary condenser of the AquaDetox system. Condensed organics are
pumped to a storage tank for recycle, water condensate is pumped to the recycle tank for
further treatment by the AquaDetox process, and non-condensables are transferred to the
active GAG bed.
114
-------
Groundwater Treatment Facility
The groundwater treatment technology at the Lockheed site is the Moderate Vacuum
Steam Stripper (MVSS) AquaDetox system. Process flow diagrams are shown on Figures 2
an 3.
Contaminated groundwater is fed from extraction wells to a cross exchanger, where it
is heated by the treated water. The heated water then enters the top of the stripping column
(9' diameter x 60' tall) and flows down the column, contacting the rising vapor flow
generated by the introduction of steam to the bottom of the column. Under a pressure of 100
mmHg abs., the contaminants are stripped from the liquid into the vapor stream, which exits
from the top of the column. The treated water leaves the bottom of the column. The treated
water passes through the heat exchanger, where it is cooled and the contaminated feedwater
is heated. The water exiting the treatment facility is thereby controlled to 9 to 10° F higher
than the incoming groundwater.
The overhead vapors flow to a water-cooled condenser, where the water vapor is
condensed and recycled back to the contaminated feedwater. The water for cooling the
condenser is provided by diverting a portion of the cool feed stream through the condenser
and back to the main feed stream.
f Total condensation of the overhead vapors is not possible due to non-condensable
gases from "vacuum leaks" and dissolved gas contained in the contaminated groundwater.
These non-condensable vapors, carrying some water, inert gases, and VOCs, enter a vacuum
pump where they are compressed to atmospheric pressure. Cooling of this compressed vapor
stream results in condensation of water and VOCs.
115
-------
The water phase is recycled to the contaminated feedwater and the organic solvent
phase is withdrawn for reclamation by a contract recycler. The coolant for this secondary
condenser is supplied from the feedwater as is done for the first condensing unit.
The vent stream from the secondary condenser contains the non-condensable and an
equilibrium quantity of VOCs. This stream is passed through vapor-phase GAC prior to
discharge into the reinjection wells of the SVE system.
Soil Vapor Extraction (SVE) System
Soil vapor extraction (SVE) is being used at the Lockheed site for remediation on
contaminated soil because of the relatively volatile character of the reported contaminants,
depth to groundwater in the range approximately 140 to 150 feet, and the predominantly
coarse-grained nature of subsurface soils.
/
i
Figure 4 shows the locations of the extraction and reinjection wells. The design of
the SVE system focused on the distribution of the wells to produce an effective and non- '
disruptive pneumatic flow regime. "Effectiveness" of the SVE was judged to depend on
establishing radially inward flow (toward an extraction well) throughout the areas of probable
soil contamination; "non-disruptive pneumatic flow regime" refers to injection well placement
such that (1) fugitive atmospheric emissions are not created, and (2) soil-gas within the areas
of probable soil contamination is not displaced from the zone of extraction well influence.
Extraction wells connected to a common header feed up to 300 CFM of contaminated
soil-gas to the system for processing and decontamination via carbon adsorption. Liquids
collected in the SVE scrubber sump are pumped to the water recycle tank for processing
116
-------
through the AquaDetox tower. Vapors are exhausted to the GAC beds for hydrocarbon
removal prior to reinjection.
Three GAC beds remove chlorinated hydrocarbons from SVE system extraction well
soil-gas, along with vent gases from the AquaDetox system. The GAC beds are operated
alternately, with two beds on-line in series while the remaining unit is being regenerated.
Once each 8 hours, the regenerated off-line bed is placed in service and spent carbon bed is
removed from service and regenerated. Steam is used to strip chlorinated hydrocarbons from
the GAC units. The vapors from this regeneration process are condensed and processed in
the AquaDetox separator. .
Treated soil-gas is reintroduced into the ground at depths ranging from 50 to 150 feet
through the vadose zone. The soil-gas then sweeps horizontally through the contaminated
soil, picking up additional hydrocarbons, and is once again collected in the soil-gas extraction
well system, where hydrocarbons are again removed.
SYSTEM OPERATION
The groundwater treatment plant operates at an average flow rate of 1000 gpm and
the SVE at 170 SCFM. The contaminants treated are listed in Table 2. Initially, total VOC
concentrations were 12,000 ppb in the groundwater and 6,000 ppm in the soil-gas. After the
integrated system had been operating several months, these concentrations dropped to 5,000
ppb and 450 ppm, respectively. Currently the concentrations are 3,500 ppb and 400 ppm.
To date, the AquaDetox/SVE facility has removed thousands of pounds of VOCs.
Table 2 lists the major contaminants in the groundwater feed to the treatment plant.
Effluent analyses show that all contaminants have been reduced to below the analytical
117
-------
detection level (1 ppb for most contaminants). This equates to a removal efficiency in excess
of 99.99%. The soil-gas treatment by two of three 3,500 pound carbon beds removes VOCs
to below 2 ppm before the air is reinjected in the ground. This equates to a removal
efficiency of better than 99%.
While the treatment plant has operated consistently (95% availability factor) at
average design flow rates and the outlet VOC concentrations from the AquaDetox tower
remain non-detectable, it has not been devoid of typical start-up problems and one
operational problem. The start-up problems were typically failures of instrumentation and
control software bugs, which have been since been resolved. A more persistent problem,
however, has been caused by the high alkalinity of the groundwater and resulting calcium
carbonate scaling in parts of the treatment plant.
Solubility of the calcium carbonate in the groundwater is reduced in two ways as the
water is processed through the AquaDetox system. First, the water is heated and, second,
carbon dioxide is removed during the stripping process in the column, thereby increasing the
^
pH. The principal disadvantage of scaling is the reduction in heat transfer efficiency of cross
exchanger, resulting in greater steam consumption.
Initially, an anti-sealant was injected in the feed water. This treatment slowed the
a
scaling due to the subsequent removal of carbon dioxide and concomitant pH increase, but
periodically the heat exchanger required acidizing to maintain its heat transfer properties.
A sulfuric acid injection system was later installed to control pH and eliminate
scaling. The costs associated with the addition of sulfuric acid were more than offset by: (a)
the savings in eliminating anti-sealant injection; (b) savings in eliminating phosphoric acid
used to clean the heat exchanger periodically; and (c) lowering average steam consumption
118
-------
by improving heat exchanger efficiency. Less than 20% of the steam consumption in the
AquaDetox facility is needed to strip contaminants; the other 80% is needed in raising the
incoming water to its boiling point of 120°F at 100 mmHg. The cross exchanger helps
reduce this steam requirement by using heat from the effluent water. This is a highly
energy-efficient and cost-effective approach, and future systems will have even larger cross
exchangers.
*--»*
OPERATING COSTS
Annual operating costs for the AquaDetox/SVE plant are:
Labor: One individual was initially assigned full-time for the maintenance and
operation of the facility, but after the first six months of operation his time was reduced to
three days per week. It is expected that after another year of operation, no more than one
day per week will be needed. Current labor costs are about $8,000 per month.
Steam: Steam, which is provided by an existing Lockheed boiler, is costed at $5.70
per 1,000 pounds. At a 1000 gpm flow rate, the steam consumption is 3,500 Ib/h before
calcium carbonate scaling shows its effect on the cross exchanger efficiency. An additional
340 Ib/h of steam (equivalent continuous average) is used to regenerate the carbon beds.
This results in a total monthly steam consumption of 3,840 Ib/h at a cost of $14,700.
Chemicals: Initially, significant amounts of anti-scale and scale-removing chemicals
were being consumed, with limited success. This was changed to a pH control system using
sulfuric acid. Current scale control costs are $1,800 per month.
119
-------
Power: The power requirement to operate the treatment plant is 88 kW. At a cost of
$0.70 per kWh, this represents a monthly cost of $4,200. This does not include the power
consumption for the groundwater extraction well.
Supplies: Miscellaneous supplies such as oil, replacement gauges, pump seals, spare
parts, etc. cost about $5,000 per month.
Based on the above breakdown, monthly operating costs average $33,700 or $.77 per
1000 gallons. Ffirther reductions (particularly labor) are anticipated with time, and savings
\
in steam costs can further be accomplished by installing larger cross exchangers.
EPILOGUE
AWD Technologies was chosen by Lockheed as the contractor to design and build a
facility to treat 12,000 gallons a minute of groundwater from the same aquifer. The $86
million cleanup effort, the largest pump and treat project in the nation under CERCLA, will
have the same project manager, Roger Heumann, and the same process design engineer,
Gene Cutler, that contributed to the AquaDetox/VES project's success. The scope of work
includes design, procurement, construction and operation.
120
-------
REFERENCES
V
Reports: Gary Street, Lanny Robbins, and James Clark, "AquaDetox Stripping System
for Groundwater Remediation," presented at HazMat Central 1989, Chicago,
Illinois.
Ron Derammelaere, AWD Technologies, Inc. and Ron Helgerson, Lockheed
Aeronautical Systems Company, "Integrated Zero-Emission Groundwater and
Soil Remediation Facility at Lockheed, Burbank", selected for the Superfund
Innovative Technology Evaluation Program (SITE).
Article: Greg Braxton, "Lockheed to Pay Most of Water Cleanup Cost."
121
-------
TRANSPORTABLE DEBRIS WASHING SYSTEM:
FIELD DEMONSTRATION RESULTS AND STATUS
OF FULL-SCALE DESIGN
Michael L. Taylor, Majid A. Dosani, John A. Wentz, and Avinash N. Patkar
IT Corporation
Cincinnati, Ohio
Naomi P. Barkley
U.S. Environmental Protection Agency, Risk Reduction Engineering Laboratory
Cincinnati, Ohio
Charles Eger
U.S. Environmental Protection Agency, Region IV, Office of Emergency Response
Atlanta, Georgia
Presented at the
Third Forum on Innovative Hazardous Waste Treatment Technologies:
Domestic and International.
Dallas, Texas,
June 11-13, 1991.
122
-------
INTRODUCTION
EPA recently published (Federal Register, May 30, 1991) an Advanced Notice
of Rule Making (ANPR) in which Potential Best Demonstrated Available Technology
for Contaminated Debris was addressed. In this publication, EPA sets forth suggested
regulatory definitions for debris and contaminated debris, indicates the applicability of
existing Land Disposal Restriction Treatment Standards as well as Superfund 6A and
6B Guidelines and describes in general the available technologies for treating
contaminated debris.
below.
The suggested definitions for debris and contaminated debris are quoted
Debris means solid material that: (1) has been originally manufactured or
processed, except for solids that are listed wastes or can be identified as being
residues from treatment of wastes and/or wastewaters, or air pollution control
devices; or (2) is plant and animal matter; or (3) is natural geologic material
exceeding a 9.5 mm sieve size including gravel, cobbles, and boulders (sizes
as classified by the U.S. Soil Conservation Service), or is a mixture of such
materials with soil or solid waste materials, such as liquids or sludges, and is
inseparable by simple mechanical removal processes.
Contaminated Debris means debris which contains RCRA hazardous waste(s)
listed in 40 CFR Part 261, Subpart D, or debris which otherwise exhibits one or
more characteristics of a hazardous waste (as a result of contamination) as
defined in 40 CFR Part 261, Subpart C.
In the ANPR it is stated that "promulgating land disposal restrictions (LDRs)
including treatment standards for solvents and dioxins, California list wastes and the
First Third, Second Third, and Third Third wastes, the Agency regulated debris
contaminated with these restricted wastes. The land disposal restrictions in 40 CFR
268 thus generally apply to contaminated debris, including such debris generated
from corrective actions and closures at RCRA-regulated land disposal sites, remedial
and removal actions at Comprehensive Environmental Response Compensation and
Liability Act of 1980 (CERCLA) (Superfund) sites; and private party cleanups."
In conjunction with the promulgation of LDR's, the EPA Risk Reduction
Engineering Laboratory funded a project under the SITE program to develop
technology which could be applied on-site for the decontamination of debris. The
results of initial field testing of the pilot scale Debris Washing System (DWS)
[Performed at the Ned Gray Site (PCB/transformers)] were presented at this
conference in May 1990.1
In this paper we describe results of a second field demonstration in which the
utility of the pilot scale DWS was demonstrated for decontaminating debris found at a
pesticide-contaminated site in Northern Georgia.
123
-------
DEMONSTRATION OF DWS AT SHAVER'S FARM DRUM DISPOSAL SITE
A demonstration of the DWS was conducted in August 1990 at the Shaver's
Farm drum-disposal site in Chickamauga, Georgia. Fifty-five gallon drums containing
varying amounts of a herbicide, Dicamba (2-methoxy-3,6-dichlorobenzoic acid), and
benzonitrile, a precursor in the manufacture of Dicamba, were buried on this 5-acre
site. An estimated 12,000 drums containing solid and liquid chemical residues from
the manufacture of Dicamba were buried there during August 1973 to January 1974. .
EPA Region IV had excavated more than 4000 drums from one location on the site
when this demonstration occurred in August 1990. Figure 1 presents an aerial
photograph of the site.
The pilot-scale DWS and the steel-framed temporary enclosure were
transported to this site on a 48-foot semi-trailer and assembled on a 24 ft x 24 ft
concrete pad. Both the temporary enclosure and the DWS had previously been
erected and used at a PCB-contaminated site in Kentucky. Figure 2 shows the
temporary enclosure and the assembled DWS at the Shaver's Farm site. Ambient
temperature at the site during the demonstration ranged from 75 to 105 degrees
Fahrenheit.
Prior to the initiation of the cleaning process, the EPA removed the 55-gallon,
pesticide-contaminated drums from the burial site. The contaminated drums were cut
into four sections and the contaminated surfaces were sampled using a surface wipe
technique.2 Pretreatment surface-wipe samples were obtained from each section.
/
The drum pieces were placed into the spray tank of the DWS, which was
equipped with multiple water jets that blast loosely adhered contaminants and dirt from
the debris. After the spray cycle, the drum pieces were removed and transferred to the
wash tank, where the debris was immersed in a high-turbulence washing solution.
Each batch of debris was cleaned for a period of 1 hour in the spray tank and 1 hour in
the wash tank. During both the spray and wash cycles, a portion of the cleaning
solution was cycled through a closed-loop system in which the contaminated cleaning
solution was passed through an oil/water separator, and the aqueous solution was
then recycled into the DWS. After the wash cycle, the debris was returned to the spray
tank, where it was rinsed with fresh water. Figure 3 presents a schematic of the pilot-
scale DWS.
Upon completion of the debris cleaning process, posttreatment wipe samples
were obtained from each of the drum pieces to assess the residual levels of
benzonitrile and Dicamba. In the case of the metallic debris sampled in this study, the
posttreatment wipe sample was obtained from a location adjacent to the location of the
pretreatment sample. This was necessary because wiping the surface removes the,
contamination, and if one were to wipe the same surface after cleaning, the results
obtained would be biased low.
All field demonstration activities performed under the SITE program were
governed by an EPA-approved, site-specific Health and Safety Plan.3 Hydrogen
cyanide was of particular concern at this site. In one instance during excavation,
inadvertent mixing of drum contents resulted in a release of hydrogen cyanide.
124
-------
r\> •
en
On-Site Water
Treatment System
PEI Pilot-Scale
Debris Washing System
Figure 1. Aerial photograph of Shaver's Farm site.
-------
Figure 2. The temporary enclosure and assembled pilot-scale
DWS at Shaver's Farm site.
126
-------
ro
Step 1 - Spray Cycle
Step 2 - Wash Cycle
• • Step 3 - Rinse Cycle
DE Filter
• Water Treatment Step
Pump
Surfactant
Tank
Activated Carbon
OCX)*
Cartridge
Filters
Oil/Water Separator
Treated Water
Storage Tank
Diesel-
Powered
Generator
240V.30
Cartridge
Filters
Figure 3. Schematic of pilot-scale Debris Washing System.
-------
However, no airborne hydrogen cyanide was detected during operation of the DWS
(Draeger Tubes were used to assess airborne cyanide concentrations). Personnel
donned Level C protective gear while working near the contaminated drums.
As stated above, surface wipes were obtained using the same method as
described for assessing PCB contamination. The Dicamba and benzonitrile in the
surface wipes were quantitated using SW 846 analytical methods. Dicamba was
extracted from the gauze wipes using Method 3540 and quantitated using Method
8150. Benzonitrile was extracted using Method 3540 and concentrations in the
extracts were measured using SW 846 Method 8270.
RESULTS
The results obtained during this demonstration are summarized in Tables 1
and 2. The data provide an indication of the effectiveness of the DWS technology for
removing a pesticides and a related contaminant (benzonitrile) from the internal
surfaces of excavated drums. Pretreatrnent concentrations of benzonitrile in surface-
wipe samples ranged from 8 to 47,000 |ig/100 cm2 and averaged 4556 jig/100 cm2.
Posttreatment levels of benzonitrile ranged from below detection limit to 117 ^g/100
cm2 and averaged 10 ng/100 cm2. Pretreatrnent Dicamba values ranged from below
detection limit to 180 ^ig/100 cm2 and averaged 23 |ig/100 cm2, whereas posttreat-
ment concentrations ranged from below detection limit to 5.2 ng/100 cm2 and
averaged 1 jj.g/100 cm2.
Upon completion of the treatment, the spent surfactant solution and rinse water
were treated in the water treatment system, where they were passed through a series
of paniculate filters, and then through activated-carbon drums. The treated water was
temporarily stored in a 1000-gallon polyethylene tank pending analysis. The before-
and after-treatment water samples were collected and analyzed for benzonitrile and
Dicamba. The concentration of benzonitrile in the pretreatment water samples was
250 and 400 ng/L (analyzed in duplicate), and the posttreatment concentration was
below the detection limit of 5 |ig/L The concentration of Dicamba in the pretreatment
samples was 6800 and 6500 ^ig/L (analyzed in duplicate), and the posttreatment
concentration was estimated to be 630 jig/L (value estimated due to matrix
interferences).
Because the concentration of Dicamba in the posttreated water sample was
630 n/L, the treated water stored in the polyethylene holding tank was pumped into an
onsite water-treatment system for further treatment before its discharge into a nearby
creek. Although the concentration of Dicamba in posttreatment water was an
estimated value, it was decided to send the water to the onsite water-treatment system
prior to discharge as a precautionary measure.
The test equipment was decontaminated with a high-pressure wash. The wash
water generated during this decontamination was collected and pumped into the
onsite water-treatment system. The system and the enclosure were disassembled and
transported back to Cincinnati in a semitrailer.
128
-------
Batch
Number
1
2
3
4
5
6
7
8
9
10
Number
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
1
2
1
2
Benzonitrile
Pretreatment
180a(50)b
130a(50)
125
90
43
28
4400
2700
47000
22000
10a(5)
8a(5)
200
320
1400
3000
3500
22a (5)
1400
Pos treatment
NDC
ND
117
7,8a (5)
ND
ND
ND
ND
10a(5)
7.9a (5)
ND
ND
ND
10a(5)
28
ND
7a(5)
ND
ND
a Estimated result less than 5 times detection limit.
b Numbers in parentheses indicate the minimum detectable concentration of the analyte.
c None detected in excess of the minimum detectable concentration of 5 u.g/100cnf
unless otherwise specified.
Table 1. Results obtained in analyzing surface wipe samples
for Benzonitrile (u,g/100 cm2).
129
-------
Batch
Number
4
5
6
7
8
9
10
Sample
Number
1
2
1
2
1
2
1
2
1
2
1
2
1
2
Dicamba
Pretreatment
1.9
3.4
NDP
ND
ND (2.7)
ND (2.7)
7.3a (2.7)
15
55
13
1.7
ND (2.7)
41
180
Posttreatment
0.63a
(0.27)b
ND
ND
2.6
ND
ND (2.7)
1.8
2.3
5.7a (2.7)
0.62a (0.27)
0.63a (0.27)
ND
0.30a (0.27)
0.34a (0.27)
a Estimated result less than 5 times detection limit.
b Numbers in parentheses indicate the minimum detectable concentration of the analyte.
c None detected in excess of the minimum detectable concentration of Dicamba at 0.27
unless otherwise specified.
Table 2. Results obtained in analyzing surface wipe samples
for Dicamba (u,g/100 cm2).
130
-------
FULL-SCALE DEBRIS WASHING SYSTEM: CONCEPTUAL DESIGN
This section describes the conceptual design of a full-scale version of the DWS,
which is based upon results obtained during bench- and pilot-scale work. The lessons
learned from these latter development stages are incorporated into the full-scale
design, and the elements that worked well have been retained. Figure 4 presents a
schematic block diagram of the full-scale DWS.
The debris will be loaded in a cylindrical basket, lifted by a crane, and lowered
into the wash/spray/rinse tank, in which the basket will rotate. The debris will then be
washed and sprayed with hot surfactant solution and finally rinsed with clean water. A
small bleed stream will be sent to the water treatment system to recondition the
surfactant solution while the process is in progress.
The full-scale system will be about 3 1/2 times (1000-gallon) the capacity of the
pilot-scale DWS and will be permanently mounted on two 48-ft flat-bed trailers. The
system will be semiautomatic and will be capabie of cleaning 3 to 5 tons of debris per
8-hour day.
CONCLUSIONS
Field-test results reported in this paper and previously obtained using the pilot-
scale transportable DWS showed the unit to be highly reliable and rugged. Extreme
high ambient temperatures had little effect on the operation of the equipment. The
system was successfully previously used to remove PCBs from transformer casing
surfaces and in this present demonstration was shown to be efficacious for removing
certain pesticide and herbicide residues from drum surfaces. Although the system has
not been proven effective for removal of all types of organic contaminants from the
surfaces of debris, results obtained to date are considered promising.
The cleaning solution was recovered, reconditioned, and reused during the
actual debris-cleaning process, which minimized the quantity of process water
required for the decontamination procedure. The water treatment system was effective
in reducing contaminant concentrations to below the detection limit.
The planned progression of this U.S. EPA-developed technology is continuing
with design, development, and demonstration of a full-scale, transportable version of
the DWS unit.
131
-------
Monorail/
Crane
Spray/Rinse/Wash
Tank
Wash
Pump
Oil/Water
Separator
J\sv> -Cy ^
tv.
Water Treatment
System
1
»*««**« Step 1 - Wash Cycle
step 2 - Spray Cycle
step 3 - Rinse Cycle
Water Treatment Step
Pump
Figure 4. Schematic diagram of full-scale Debris Washing System.
-------
REFERENCES
1) Taylor, M. L, Dosani; M.A., Wentz, J.A., et al. "Results of Field Demonstration of
Debris Washing System," Presented at the 2nd Forum on Innovative Hazardous
Waste Treatment Technologies: Domestic and International, Philadelphia, PA,
May 1990.
2) Field Manual for Grid Sampling of PCB Spill Sites to Verify Cleanup, U. S.
Environmental Protection Agency, EPA 560/5-86/017, May 1986.
3) Standard Operating Safety Guides, Office of Emergency and Remedial
Response, Hazardous Response Support Division, Edison, NJ, November
1984.
ACKNOWLEDGMENTS
This research was funded in its entirety by the United States Environmental
Protection Agency's Risk Reduction Engineering Laboratory under Contract No.
68-03-3413. Naomi Barkley is the Technical Project Monitor.
DISCLAIMER NOTICE
This paper was prepared as an account of work sponsored by an agency of the
United States Government. Neither the United States nor any of its employees, nor
any of the contractors, subcontractors, or their employees make any warranty,
expressed or implied, or assume any legal liability or responsibility for any third party's
use or the results of such work or of any information, apparatus, product, or process
disclosed in this paper or represent that its use by such third party would not infringe
on privately owned rights. The views and conclusions contained in this document are
those of the author and should not be interpreted as necessarily the official policies or
recommendations of the U. S. Environmental Protection Agency or of the U. S.
Government.
This paper has not yet undergone peer review by EPA Risk Reduction
Engineering Laboratory. Subsequent to peer review the contents may be revised.
133
-------
A- progress report on the developments in cleaning
soils with chlorinated hydrocarbons, and the develop-
ment of a wet cleaning method for contaminated sand
NBM Bodemsanering B.V.
P.O. Box 16032, 2500 BA The Hague
Netherlands
Phone 31 70 3814331 / Fax 31 70 3834013
Ir. H.J. van Hasselt / Ir. A.. Costerus
NBM Bodemsanering,
June 1991
134
-------
1. "Introduction
This paper discusses the developments in cleaning soils conta-
minated with chlorinated hydrocarbons in a thermal -system, and the
development of a low cost, high throughput system for the cleaning
of contaminated sand.
NBM Bodemsanering B.V. subsidiary of NBM Amstelland NV, the second
building and construction group in The Netherlands, developed an
indirect heated thermal treatment system and became one of the
leading firms in the Netherlands in the cleaning of soil and soil
pollution related projects.
After four years of operating experience this paper presents the
cleaning results of soils contaminated with several types of
contaminants.
Also, NBM felt that for specific types of sand and contaminants,
it is possible to design and build specialized cleaning equipment
that can clean this sand on a low cost basis.
This system is a "wet" system, and the results of the first series
of test are reported in this paper.
2. Thermal system
2.1 Process
Since 1987, NBM operates a commercial thermal soil cleaning plant.
The plant is. located in Schiedam, in the Rotterdam harbour area,
and has a yearly capacity of approx. 80.000 tons.
The system was developed in the years 1982-1987.
The requirements, set for development of a soil cleaning system
were:
- It must be possible to clean any type of soil (sand, clay, peat)
- It must be possible to clean all organic compounds and pollutants
that can be removed by volatilization and for pyrolysis at
temperatures up to 650 oC
NBM chose to develop an indirect heated thermal process.
This means, that offgases from the burners do not come into direct
contact with the contaminated soil.
135
-------
chimney
(Oil Input !
j
( — I
burner [
• — . Rotary drum
n
Schematic direct heating
chimney
J.
incinerator
1
j
Schematic indirect heating
Doing this, the contaminated gas stream is much smaller compared
to a direct fired system.
The principal arguments for the choice between direct and indirect
heated process are given in the following table.
Direct hea-
ted
Advantage
Unlimited
heat trans-
fer capacity
Relatively
short hea-
ting and
cooling time
of equipment
Disadvantage
High volume
of gas stream
in rotary
kiln
High volume
of gas stream
in afterbur-
ner ;
Dust filtra-
tion from gas
stream diffi-
cult
Indirect hea-
ted
Advantage
Low volume of
gas stream
Low volume of
gas stream in
afterburner
Filtration of
dust at high
temperature
possible
Di s advantage
Limited heat
transfer
Long heating
and cooling
time of
equipment
In the past years, cne system has proved to be able to clean all
types of soils, contaminated with chlorinated and not chlorinated
compounds.
By nature, thermal systems do not remove heavy metals for soil.
Their very high boiling points make evaporation in the NBM-process
practically impossible.
136
-------
2.2-Equipment
The commercial plant of NBM Bodemsanering has been in operation
since 1987.
The plant has a yearly throughput of 80,000 - 90,000 tons.
The plant is operating continuous, 7500 hours per year, at a
capacity which varies with type of soil from 8-15 tons per hour.
From 1987 till February 1991, 300,000 tons of contaminated soil
were cleaned.
The plant consists of the following items:
2.2.1 Soil input
The polluted soil, from which very coarse pieces, exceeding 100 mm,
have been removed, is loaded into a feed unit.
2.2.2 Feed unit
The soil coming from the bin drops into a picking belt, where
coarse pieces and non-ferrous metals may be removed. Then the soil
is moved upwards by a conveyor belt to the dryer. A magnet mounted
above the belt removes ferrous metals as much as possible from the
soil.
2.2.3 Dryer
The dryer comprises mainly a rotary drum having a diameter of 3 m
and a length of 21,5 m. The drum is heated externally, that is
indirectly, by means of hot gases. The heat sources for the dryer
are:
- burner gases exiting from the indirect heated rotary tube
furnace,
- the flue gases from the afterburner, and
- auxiliary burners.
2.2.4 Screen
When leaving the dryer, practically all of the free water has been
evaporated from the soil. At this stage the structure of the soil
is such that fragments of stone, rubble and pieces of wood may be
removed by screening. This is necessary because the transportation
and sealing means of the rotary tube furnace cannot cope with very
coarse pieces.
137
-------
2.2.~5 Transfer to the tube furnace
The fine fraction passing the screen, together with the dust
filtered from the exhaust gas from the dryer is passed through a
screw conveyor to an elevator. The elevator carries the soil into
an intermediate bin, where it is introduced into the tube furnace.
2.2.6 Rotary tube furnace
The dried and screened soil is heated in the tube furnace to a
temperature between 450 and 600° C. The achievable maximum tempe-
rature of the soil and the residence time of the soil in the drum
are, obviously, a function of the quantity of soil introduced per
hour. The tube furnace is provided with gas-sealing gaskets and
seals.
2.2.7 Soil cooler
The cooler consist of a rotating drum. A plurality of pipes are
located inside the drum, through which cooling air is blown. The
cleaned soil introduced into the cooler is cooled to approx. 150°
C. The air leaving the cooler has a temperature of approx. 250° C
and is used as preheated combustion air for the burners of the tube
furnace and the dryer.
2.2.8 Mixer humidifier
V
The soil is continuously humidified in a mixer. The soil leaving
the mixer has a temperature of approx. 50° C and a moisture content
that can be adjusted in advance. A conveyor belt carries the soil
to a bin.
2.2.9 Treatment of the gas leaving the dryer
The gases exhausted from the dryer include mainly steam, light
pollutants evaporated from the soil and air sucked in via the feed
mechanism and some of the- seals. This, gas contains a certain
quantity of dust which is removed from the gas stream by a filter.
The dust collected in the filter is fed back to the main stream of
soil.
Depending on the type of the soil and the type of the pollutants,
the gas stream from the dryer may be processed by either one of two
routes.
138
-------
Steam condensation
By lowering the temperature of the gases exiting in the dryer the
major part of the steam is condensed in a condenser. The condensed
water is passed to a water treatment unit.
Direct exhaust to the afterburner
If it is not desirable to condense the steam from the dryer, the
dryer effluent passes directly from the filters to the afterburner.
The gas stream to the afterburner is now much more substantial and
so is the energy consumption.
2.2.10 Treatment of the gas leaving the tube furnace
The gas stream from the tube furnace comprises the following groups
of substances:
- pollutants evaporated from the soil,
- moisture evaporated from the soil,
- products of pyrolysis of organic substances present in the soil,
- inert gas entering into the gas stream through the flushing
device of the ceramic filter located on the line between the tube
furnace and the afterburner.
The temperature of the gases leaving the tube furnace is a function
of the selected soil temperature. It has the same order of
magnitude as the temperature of the existing soil. The gaseous
mixture is then passed to the ceramic filter unit.
2.2.11 Ceramic filter unit
The ceramic filter unit comprises the following parts:
a) A settling chamber where the coarser dust particles contained
in the gas stream and entering the chamber are pre-separated by
settling.
b) Two parallel filter units, each provided with valves and an
exhaust fan. The gases pass from the settling chamber through
one or both of the filters and go on to the after-burner.
2.2.12 Afterburner
All the gases having been in contact with the polluted soil are
burnt in the afterburner. If they are burnt in the afterburner for
a sufficient period of time at an adequately high temperature and
with an adequate oxygen percentage, all pollutants will be conver-
ted into non-toxic compounds such as H2O, CO2 and N2.
The quantity and composition of the gas stream fed to the after-
burner depend greatly on the composition of the soil, its moisture
content and the type and quantity of the pollutants.
The afterburner has proven to give adequate destruction efficien-
cies, for all process conditions.
139
-------
2.2 .-13 Exhaust
When the installation is in normal operation, practically all of
the gases produced in the afterburner, together with the gases from
the tube furnace, will be exhausted through the heating jacket of
the dryer.
An exhaust fan removes the gases from the system and ensure a sub-
atmospheric pressure in the entire system.
If the gas stream from the afterburner is too great to, be handled
through the dryer, this stream (or a part of it) will be removed
from the installation by another exhaust fan.
2.3 Cleaning results
The NBM plant has been in operation since 1987.
Since then, over 300,000 tons of very various types of soil, with
also greatly varying contamination levels, were cleaned.
The cleaning result has always been below the dutch 'A' values,
independent of input concentrations.
This is understandable, because as soon as the temperature is
reached at which a contaminant will evaporate from the soil matrix,
it will totally boil off, if there is the necessary residence time.
Cleaning results
CN Total
Aliphatic hydrocarbons
Aromates
Naphtalene
PAH's (16 EPA)
Input (ppm)
0 - 10,000
0 - 20,000
0 - 5,000
0 - 10,000
0 - 20,000
Output ( ppm )
< 5
< 50
0.05 - 0.10
< 0.5
< 1
140
-------
During a series of test runs with soils, contaminated with
chlorinated hydrocarbons, the same type of results was observed.
The soil was cleaned to detection level, independent of input
concentration.
Cleaning results
HCH
PCB
Dioxine
Drins
Input ( ppm )
0 - 200
0 - 200
0 - 200
0 - 850
Output ( ppm )
< 0.002
< 0.002
< 0.002
< 0.002
The soils were cleaned to detection level, a factor 10.000 lower
than the input concentrations. A removal efficiency of 99,9999 %
was achieved
In the soil matrix, no undesired products of decomposition were
found.
'i
Also, the behaviour of the afterburner was tested.
In a first series of tests, "the destruction efficiency of Principal
Organic Hazardous Constituents, (POHC's) was measured.
Destruction efficiencies were measured to be bigger than
99,9999 % at 1200 C.
*'•--...-.
While cleaning batches of soil, the stack emissions of the
afterburner were measured.
No rests of HCH, PCB or drins were detected. Dioxin levels were
varying in 1990 between 0.074 - 2.1 ng/nm3 TEQ.
In the 1991 test runs, dioxin levels ranged from 0.2 to 3.1 ng/nm3
TEQ . _
Without a flue gas cleaning system, the stack emissions are below
the dutch standards till 1992. In 1992 - 1995, new limits will be
imposed on soil cleaning plants.
In 1992, a flue gas cleaning system will be installed, which will
keep a.o dioxin emission into the atmosphere at levels lower than
0.1 ng/nm3 TEQ.
Without a flue gas cleaning unit, the emissions of the plant are
as follows
141
-------
Fluoride
SO2
HCL
Cyanide
Aromates
PAH's
Cadmium
Mercury
Dust
CO
NOx
Dioxines (TEQ) .
Mean
ppm
0.12
200
9.15
0.05
0.28
0.0035
0.00029
0.001
21
Old limit
ppm
5
460
75
1
5
0.002
0.1
0.1
75
New limit
ppm
1
40
10
_
_
_
0.05
0.05 '
5
50
70
0.1 PPt
2.4 Operational factors, price
The operational costs are principally energy, labor, maintenance
and investment. The relative costs*of the selling price are shown
below.
Breakdown of cleaning price
Item
Energy
Labor
Maintenance
Capital cost
Analyses
%
20
20
20
35
5
The operational costs greatly depend on the production capacity.
Only the energy cost is related to the production. The production
capacity depends on type of soil, moisture content, energy content,
and type of pollutants
The total selling price under Dutch conditions and at an exchange
rate of 1 $ - fl. 2.00 is $ 100 to $ 125 per ton.
142
-------
CO
, lHtHMIi.Cllt
GRONDHtlNIUiNfi-
INSTALLAtlE Nt»M
VUUHUHUUIiIHUMMtl
14 hAVIHUHANUtt
UUl>LI UtVAlUH
HUlAMt MIMIINUIHtCI HI
CUOllHu UHUM (1NUIKICI I
I iHtHMISCHt
| BODfNHtlNlGU
ANLAQE NBM
-------
3. "Wet system
3.1 Introduction
t • -
A thermal system as described above can clean very many types of
soils with contaminants that are difficult to remove, it is also
a high priced system.
For specific types of soil, cleaning processes are possible at a
very much lower cost.
NBM, being one of the Netherlands biggest building contractors, is
also involved in recycling of building materials. Demolition waste
is crushed into aggregate that is used in the building and road
building industry. A byproduct of this activity is a sandy
material, with a granular dimension of 0 - 10 mm.
This sand is contaminated with PAH's and heavy metals, to levels
high above the dutch standards for building materials.
Because the tipping fee of this contaminated sand will go up to $
50-60 per ton, a low cost cleaning method was developed.
3.2 Process '
When analysing the problem of this specific type of contaminated
sand, NBM learned that the heavy metals in the sand are adhered to
the finest (< 60 mu) particles, and that the PAH's are present as
particles, mostly not adhered to a "sand particle.
Removing of the finest particles, approx. 10 % of the dry solids,
resulted in a drastic reduction of the heavy metals concentration
in the main stream.
This method does not remove the bigger tarry particles, so a second
technology was needed. *
A combination of particle size separation by means of hydrocyclones
and a separation based on the difference in density between sand
and tarry materials, flotation, proved to be successful.
3.3 Equipment
The plant consists of the following items:
3.3.1 Screen
The polluted sand is loaded into a feed unit, from which it is
conveyed to a screen.
On this screen, water is added, and the material is separated in
the fractions 0-2 mm, and 2-10 mm.
144
-------
3.3 .-2 Hydrocyclones
The fraction 0 - 2 mm is pumped to the hydrocyclones. - ,
The input is separated in two streams:
- 0 - 60 mu, approx. 10 % of the dry solids of the input of the
plant, with the bigger part of the heavy metals
- 6 mu - 2 mm, approx. 90 % of the dry solids.
This stream is still contaminated with tar particles, and is fed
into a flotation unit.
3.3.3 Flotation
In the flotation unit, the sand is flowing downward.
In counter current flow, a small stream of water is flowing
upwards, entraining the lighter particles. Since the density of the
tarry particles is 1,05 - 1,30 t/m3 compared to 2.65 t/m3 for sand,
the tarry particles will flowing. They are separated from the water
stream, and added to the waste stream.
The clean sand is leaving the flotation unit at the bottom, and
dewatered on a dewatering screen.
3.3.4 Dewatering
The 0 - 6 mu stream from the hydrocyclones, with the waste
stream from the flotation unit, flow through a thickener into a
screen belt press.
This stream is disposed of on a landfill.
3.4 Cleaning results
Up till now, the system has been in operation on a test
basis. The typical cleaning results that have been achieved are the
following.
Cleaning results
Pb
Zn
As
PAH ' s
Input (ppm)
150 - 600
200 - 800
3-7
150 - 450
Output ( ppm )
< 75 .*
< 100
< 1
< 20
The output values meet the dutch standards for recycling building
materials.
The material will be used as sand in the building- and roadbuilding.
industry.
145
-------
3.5 "Operational factors, price
The wet cleaning plant will be in operation in 1992. It will
have a throughput of 100 tons/hr.
The yearly throughput will be approx. 300.000 tons.
The total selling price is $ 15 - $ 25 per ton, excluded the tip-
ping fee for the separated contaminated fine fraction.
In Holland tipping fees for this type of materials are approx. $
60 per ton at this time.
The percentage of fines in normal sieve sand is 10 % per ton of
sieve sand. NBM calculates with $ 5 - 10 additional tipping fee for
dewatered sludge above the processing price of $ 15 - 25 per ton.
The relative costs are shown below:
Breakdown of cleaning price
Item
Energy
Labor
Maintenance
Capital cost
Utilities, etc.
Analyses"
%
5
20
10
30
25
10
Cleaning of contaminated sand
Input
1
Screening —
1
Cyclones —
Flotation _
1
Clean building material
To crushing plant
t
•*• Flotation > Waste
•+- Sludge dewatering.
Landfill
Waste
146
-------
E ACQUIRED WITH TEE OBCOTEC HIGH - PRESSURE SOIL
WASHING PLANT 2000 IN CLEANING CONTAMINATED SOIL
paper presented
Winfried Brull
and
Michael Mackeprang
Klockner Oecotec
Neudorfer Strasse 3-5
4100 DUISBURG
Germany
at the
EPA THIRD FORUM ON INNOVATIVE
HAZARDOUS WASTE
TREATMENT TECHNOLOGIES:
DOMESTIC AND INTERNATIONAL
June 11 - 13, 1991
The Fairmont Hotel
Dallas, Texas
147
-------
EXPERIENCE ACQUIRED WITH THE OECOTBC HIGH - PRESSURE SOI]
WASHING PLANT 2000 IN CLEANING CONTAMINATED SOIL
INDEX
1. Introduction
2. Description of the high-pressure soil
washing process
2.1 Wet liberation
2.2 Classification
2.3 Gravity separation
3. Process water treatment
4. Waste water treatment
5. Process air treatment
6. Case histories
7. Limitations of the process
8. Services offered by Klfickner Oecotec GmbH
1. INTRODUCTION
As you are aware, soil exchange has been the most commonly used
form of land reclamation up to now.
This soil exchange does not solve the problem but only
relocates it. Moreover, this is becoming increasingly
difficult, as landfill space is becoming more and more scarce,
the fees for depositing at the landfill sites are increasing
correspondingly, and new administrative regulations for
transporting and depositing contaminated substances have been
issued. There is, therefore, a demand for an innovative
148
-------
solution to soil exchange and for an improvement from the
ecological point of view.
\
KlSckner Oecotec has been working on the new technology of
high-pressure soil washing since 1986. To date, some 250,000
metric tonnes of different soils have been cleaned with three
industrial - scale plants. Other plants will be started up in
near future. The high-pressure soil washing process is used for
rehabilitating the former sites of steel and gas works, coking
plants, chemicals factories and shipyards. The capacity is up
to 50 metric t/h. This technology can be used both in the form
of a mobile plant on site and as a stationary plant in a waste
disposal centre. It is characterised by a high degree of
acceptance among the population which has become sensitized to
matters of pollution.
The high - pressure soil washing plant consists of sixty 40
feet containers or modules. The single containers can be
transported either by truck or train. Complete dismantling and
re-contruction of the plant including transportation from one
site to another takes about two weeks.
As the plant is supplied with power from its own diesel
generators, it is not reliant on the public mains.
Thanks to its modular design, the plant may be readily adapted
to changing soil conditions and alternating contamination
because the required treatment stages could, and still can, be
added to the system, just as others, which are not necessary,
can be left out and used for other purposes.
Flexible pipes connect the individual assemblies and modules.
Each module has its separate control cabinet which is either
mounted on the module itself or plugged into the central
control panel. By virtue of this feature, the individual ^.
assemblies, especially those provided for process water
processing and waste water treatment, can be used flexibly in
response to problems arising in connection with water
processing or cleaning.
•\
In winter times a light weight hall can be surrounded around
the plant to minimize down time and to avoid noise.
149
-------
2. DESCRIPTION OF THE HIGH - PRESSURE SOIL WASHING PROCESS
The high-pressure soil washing process is an extractive on-site
process. The plant consists of five interconnected main
components. The necessary process steps are the physical soil
treatment, process air treatment, gravity separation, process
water treatment and waste water treatment.
(Figure 1: Principle flow diagram of the high-pressure soil
washing plant)
The flow diagram clearly shows that the high-pressure soil
washing process requires only a small number of uncomplicated
components. However, these components must be interconnected in
the correct sequence and coordinated with great precision.
Before the-treatment of the contaminated soil can take place,
the material needs to be pre-treated. These necessary pre-
treatment steps include the
- transportation of the excavated material,
- crushing to a maximum size of 50 mm (< 2 inches),
- preliminary storage, and
- feeding of the material by front loaders
into the low capacity feed hopper.
The soil then passes on a troughed belt conveyor to the
homogenization stage. This is done in a log washer, in which a
horizontal shaft with log elements are-arranged. The effect of
this treatment is breaking up and dispersion of the soil
agglomerates by shearing forces. From there the soil is sucked
through the high-pressure jet pipe where the wet liberation
takes place. A phase separator subsequently separates the
mixture into an exhaust air stream and a fluid material stream
which also contains the soil particles that have already been
freed from the adhesive pollutants.
150
-------
2.1 WET LIBERATION
The two major objectives of wet liberation are
- the complete dispersion of the particles
- to remove the contaminated substances from the surfaces of
the soil particles.
The particular advantage of the technology is to be found in
its basic principle: washing is untertaken with clean water
under high pressure so that no chemical additives are required.
The cleaning process takes place in what is known as
the high-pressure jet pipe, the core of the plant.
(Figure 2: High - pressure jet pipe; cross section)
The pollutants which firmly adhere to the surfaces of the
individual soil grain are blasted off by high pressure water
jets in a jet pipe specially designed for this pupose.
In the high-pressure soil washing process, jets of pure water,
arranged in a ring of nozzles, converge in a cone shape at a
focal point. The water jets at a pressure of up to 350 bar
(5150 psi) emerge from this set of jet nozzles. The wet
contaminated soil, after preliminary homogenization with water,
is sucked through this focal point. This means that the kinetic
energy of the high-pressure water jets is transformed into
blasting energy. During the acceleration phase to which each
single soil particle is subjected, the pollutants are separated
from the soil and transferred into the process water. The soil
compound is thus broken up, the surface of the soil particles
exposed and the pollutants blown off.
As a consequence, there is no need to employ any surfactants at
all. This is a requirement that other extractive washing
processes are unable to fullfil.
At the end of the washing process, there is a mixture of
process water, clean soil particles and pollutants in a
dissolved or undissolved form". The pollutants separated from
the soil are transferred to the carrier media "air" and
"water". Both streams are treated separately using following
steps.
151
-------
2.2 CIASSIFICA.TION
The wet liberation in the high-pressure jet pipe is followed by
classification operations. The main purposes for classifying in
the high-pressure plant are:
- to prepare a closely sized feed to the gravity concentration
process, and
- to transform the major toxic substances into the fines
fraction.
First of all, the clean soil particles with diameters over 8 mm
(5/16 inch) are screened off on a vibratory screen, rinsed with
fresh water, drained and discharged.
Then the entire remaining water / soil suspension is separated
step by step from the process water. The removal of particles
between 8 mm (5/16 inch) and 0.3 mm (1/64 inch) is accomplished
by an inclined classifier or screw classifier. The pulp feed is
introduced into an inclined trough. Particles of high falling
velocity quickly fall to the bottom of the trough and are
conveyed up the inclined trough by a helical screw. The fines
are removed over the overflow weir of the trough. These
particles, having a maximum particle size of 0.3 mm (1/64 inch)
are treated on two different banks of hydrocyclones.
A hydrocyclone is a continuously operating classifying device
that utilises centrifugal forces to accelerate the settling
rate of particles. Hydrocyclones have proved extremly efficient
at fine separation sizes. The first bank of parallel cyclones
has a cut point at 63, the second at 25 microns. Each of the
two cyclone underflows can be treated alternatively in the
gravity separation module, if necessary, or can directly be
dewatered and discharged as clean soil.
The cleaned soil particles from all classifying operations are
dewatered on vibratory screens by using fresh water and a
counter current wash circuit.
152
-------
2.3 GRAVITY SEPARATION
Extensive operational experience has meanwhile shown that a
cleaning of contaminated soil in combination with the high-
pressure jet pipe leads to a successful decontamination. But
what happens to those solid particles which are the carriers of
the pollutants?
Wood? coal, ash and slag are characterized by a high specific
surface in comparison to materials which belong to the soil
itself. The high inner specific surface bonds contaminants to a
great extent on .one hand and on the other hand the hydrophobic
surface character selectively adsorbs liquid and soluable
organic substances, such as phenols, benzene, aromatic
hydrocarbons etc.
Particles of this kind are very often much lighter in the
specific gravity, so that conventional gravity separation
equipment, as widely used in the mineral processing industry,
should remove such material by easy technology. Jigging
machines and spiral concentrators fullfil these demands.
Cleaning in jigs takes place by allowing a pulsing flow of
water to pass through a layer or bed of material. The pulsing
flow of water transforms the layer into a fluidized condition
or suspension. Thus the particles are arranged in layers
according to density, i.e. the light particles move to the
,/surface of the bed while particles of higher specific gravity
are arranged in the lower zone of the material bed. This
bedding is one basic process of jigging. The optimum water
movement required for a correct formation of layers is
primarily dependent upon the composition of the feed material
in relation to material and particle size. The other basic
process is the discharge i.e. separation of heavy product from
the layers of the bed. Here it is useful for separation of the
heavy product by discharging the heavy product from the largest
possible reserve layer. A star gate for bottom discharge is
being used for this purpose.
Spiral have been designed to treat slurries containing small
amounts of coal, ash or slag. A spiral is tolerant of changing
feed conditions so few operational adjustments are necessary
making it an ideal unit for most applications where the feed
material is constantly changing in both the long and short
term.
The spirals, having a polyurethane separating surface, is
supplied in duplex configuration. Each spiral trough has 2.25
turns beginning at the feed box and ending at the product box.
As the slurry flows from the feed box around the helix of the
spiral concentrator, stratification occurs in a vertical plane.
The heavies proceed to the lower velocity zone near the central
153
-------
shaft, while the lights tend to stratify in the higher velocity
zone or at the outer radius of the trough. Adjustable splitters
at the product box permit regulation of the amount of the
stream taken to a highly contaminated light concentrate.
Both light concentrates, from the jigging'machine and the
spiral concentrators, are dewatered and discharged into a
separate waste container for later landfill or incineration.
3, PROCESS WATER TREATMENT
The process water from the second cyclone overflow still
contains solids smaller than 25 microns. There are also
polluted particles in the slurry, e.g. ash, tar and slag, which
are removed by froth flotation.
(Figure 3: Process and waste water treatment)
Flotation is a process which utilizes the different surface
properties of the particles for separation. The pollutant is
usually transferred to the froth product, leaving the gangue or
clean soil particles in the pulp.
The froth product is a highly enriched concentrate of
pollutants, which is dewatered and for example incinerated.
The flotation pulp is subsequently treated with an anorganic
and organic flocculant, before the pulp is fed to the
sedimentation step, which consists out of two lamella
thickeners. It utilizes a pack of inclined parallel plates
which reduce settling distance and at the same-time increase
effective area.
This ensures that all extremely fine particles suspended in the
entire water cycle are removed. In addition, organic
substances, soluble in water, are adsorbed at this stage by the
previously mentioned anorganic adsorbant, which is a bentonite
type of product.
The cleaned water or thickener overflow is recirculated to the
physical soil treatment of the high-pressure soil washing
process for reuse. The volume of process water equals 1 m^ (264
gallons) water per ton of contaminated soil. The process water
is treated in such a way that it can be re-used for soil
washing several times. For this reason, the fresh water
requirement is limited to only 0.3 m^ (88 gallons) per metric
ton of soil. The same volume of bleed-off process water is
cleaned in a chemo-physically operating waste water treatment
plant.
154
-------
The pre-thickened sludge of the lamella thickener underflow Is
treated again with a flocculant, this time with a cationic
polyacrylamide, before it enters two belt filter presses. Here
the sludge is dewatered and transformed into a compact filter
cake with a solid content of approximately 65 %.
4. WASTE WATER TREATMENT
135 m3/h (35,640 gallons) of process water is recycled and re-
used. About 5 m3/h (1,320 gallons) is treated in the waste
water treatment facility, before it leaves the plant as clean
water. *
(Figure 3: Process and waste water treatment)
.The waste water treatment plant consists of an emulsion
breaking tank, an oxidation and reduction stage, a classical
neutralization, a gravel filter and an activated carbon filter
stage. The use of these treatment stages was coordinated with
the responsibility of the water authorities. The process stages
and their cleaning efficiencies meet state-of-the-art
requirements.
5. PROCESS AIR TREATMENT
Due to the partial vacuum arising in the closed high-pressure
jet pipe for physical reasons, a large volume of air is sucked
in together with the contaminated soil. As a result, an
excellent stripping effect is obtained in the high-pressure
soil washing process. Easily volatile substances as well as
those substances which become volatile only at reduced
pressure or in steam, are transferred into the process air
stream.
(Figure 4: Activated carbon adsorption for process air
treatment)
Volatile pollutants are captured by the connected activated
carbon filters. The exhaust air stream carrying the volatile
contamination is force-fed into two activated carbon adsorption
filters.
155
-------
Before the contaminated process air is treated in one of the
two activated carbon adsorbers a pre-filter adsorbs the long
chain solvents. Each adsorber contains 1.2 metric tonnes of
activated carbon and is able to adsorb about 10-15 weight-%
of solvents.
A continuously operating clean gas monitoring system equipped
with flame ionization detector (FID) checks that the
requirements of the relevant clean air regulation are met at
all times. This procedure ensures that the pollutants cannot
escape uncontollably at any time and do not endanger the
operators or the environment.
After the break-through of solvents in one of the two
adsorbers, the FID automatically switches over to the second
adsorber and initiates the steam regeneration of the first
adsorber. Steam is generated at a temperature of 120 °C (248
°F) causing solvent desbrbtion. The resulting solvent / steam
mixture is then copied to room temperature. A following phase
separator separates a swimming solvent phase and a water phase
in which solvents are still in solution. The cleaning of the
solvent loaded water is by means of a stripping process.
6. CASE HISTORIES
An example of the mass distribution of the products obtained
after washing of the contaminated soil is demonstrated in
figure 5.
Figure 5: Mass distribution of the soil after washing
The flotation concentrate, the lights and the filter cake are
the three concentrates which have to be incinerated or
deposited in a landfill. Theses concentrates, rich in
pollutants, represent only 15 weight - %, whereas 85 weight - %
of clean soil can be refilled. This clean soils meet the
requirements of the limit values given by the authorities.
There are no fixed regulations for limit values in Germany
today. Figure 6 gives you some required limit values from
different regions of Germany, which Oecotec had to fullfil in
the past.
Figure 6: Required limit values (ppm)
156
-------
From April 1989 to May 1990, the soil washing plant
rehabilitated a site formerly used by the steel industry. A
total of some 80/000 metric tonnes of soil contaminated with
hydrocarbons, mineral oils and polyaromatic hydrocarbons
were washed. The following rehabilitation objectives were set
but it was possible to improve upon them by up to 70 %:
Mineral oil hydrocarbons
Total hydrocarbons
Polyaromatic hydrocarbons
< 1,000 ppm
< 2,000 ppm
10 ppm
Figure 7: Cleaning efficiency at a former steel plant
After cleaning and sampling, more than 900 tonnes of every
1,000 metric tonnes of soil to be washed were replaced on
site, 8 % of the soil fed through the plant had to be disposed
of as polluted filter cake and only 1 % in the form of
severely polluted flotation sludge.
From May to November 1990, the plant was used for
rehabilitating a former ship yard site in the free port of
Hamburg. A total of some 35,000 metric tonnes of soil were
washed. The following rehabilitation objectives had to be met:
Mineral oil hydrocarbons
Polyaromatic hydrocarbons
500 ppm
10 ppm
Figure 8: Cleaning efficiency at a former shipyard
The soil was severely polluted by polyaromatic hydrocarbons. At
the same time, the soil was also interspersed with non-soil
substances, such as ash and slag, the porous surface structure
of which does not permit complete
separation of the pollutants in the high-pressure jet pipe.
The plant was therefore expanded at that time by the described
gravity separation facility.
7. LIMITATIONS OF THE PROCESS
The high-pressure soil washing process can be used to free soil
from both organic (hydrocarbons, mineral oil hydrocarbons and
polyaromatic hydrocarbons) and inorganic (heavy metals,
cyanides) pollutants. Appropriate technology is also available
157
-------
for treating soil contaminated by mercury. The degree of-
contamination and the soil structure are the decisive factors
for cleaning efficiency.
In cleaning soils with different kinds of pollution, the
process has proved its marketing ability both from the economic
point of view as well as with reference to the achieved
cleaning results.
Based on initial test results, pollutant loads of almost 5,000
ppm was anticipated in the soil to be treated. It was sandy
soil with a high silt and clay percentage of up to 20 - 25 %,
i.e. soil which did not contain more than 20 - 25 % fines with
particle diameters under 25 microns (< 0.025 mm).
8. SERVICES OFFERED BY
IKNER OECOTEC
KlSckner Oecotec sees itself as a service company. A custom -
made rehabilitation concept is developed for each damage or
pollution case. In addition to carrying out the rehabilitation
work with its own plant and team of workers, the company also
assumes responsibility for the disposal of residues and
auxiliary substances remaining and provides all the advice
required when applying for permits under public law.
KL&CKNER OECOTEC GmbH
Gesellschaft fur Rohstoffriickgewinnung
und Umweltsanierung
Postfach 10 08 51
D-4100 DUISBURG 1
Germany
Phone: (0203) - 181
Fax: (0203) - 33 19 17
158
-------
PRINCIPLE FLOW DIAGRAM OF THE
HIGH-PRESSURE SOIL WASHING PLANT
FRESH WATER
CONTAMINATED SOIL
171
ID
PHYSICAL
SOIL
TREATMENT
V
GRAVITY
SEPARATION
V
CLEAN SOIL
V
PROCESS AIR
PROCESS AIR
TREATMENT
PROCESS WATER
PROCESS
WATER
TREATMENT
LIGHT
CONCENTRATE
(WASTE)
V
WASTE WATER
TREATMENT
CLEAN PROCESS AIR
==> •
FILTER CAKE (WASTE)
WATER
o
a
w
H
-------
HIGH-PRESSURE JET PIPE
(cross section)
HIGH PRESSURE
RING
WATER JETS
CTl
O
INPUT:
SOIL-AIR-MIXTURE
JET NOZZLE
H
Q
d
50
W
M
OUTPUT:
SOIL-WATER-AIR-MIXTURE
WATER CONNECTION
TO HIGH PRESSURE PUMP
-------
PROCESS- AND WASTE WATER TREATMENT
Flotation Cone.
Sludge Treatment
Process Water
CT1
FROTH
FLOTATION
FLOCCULATION
SEDIMENTATION
Circuit Water (135 m /h)
WATER
RESERVOIR
V
Waste Water (5 m /h)
FILTRATION AND
ACTIVATED CARBON^1"
ADSORPTION
NEUTRALIZATION
OXIDATION /
REDUCTION
Cleaned Water
EMULSION
BREAKING
-------
ACTIVATED CARBON ADSORPTION
FOR PROCESS AIR TREATMENT
DRY AIR
STEAM
Oi
ISi
CONTAMINATED
PROCESS AIR
SOLVENTS
FID
HEATING
-AN-
N
PRE-
FILTER
O is^l
\ f
I
>
COOLING
PHASE \
SEPARATION
H
n
a
w
*.
STRIPPER
PROCESS WATER
-------
MASS DISTRIBUTION
of the soil after washing
WEIGHT - % (dry)
Shipyard / Hamburg ,
100 % • 28.600 t contaminated material
M
a
a
»
M
en
CO
CLEAN SOIL 8 5 %
FLOTATION CONG. 1 %
LIGHTS 4 %
FILTER CAKE 10 %
-------
REQUIRED LIMIT VALUES (ppm)
TYPE
OF POLLUTANT
MINERAL OIL
HYDROCARBONS
POLYAROMATC
HYDROCARBONS
BENZENE,
TOLUENE, XYLENE
CHLORINATED
HYDROCARBONS
PCB
PERIOD
TONS (t)
STEEL PLANT
DUSSELDORF
< 1,000
< 20 (EPA)
< 3.5
< 0.1
-
April 89-May 90
80,000
SHIPYARD
HAMBURG
< 500
< 10 (TVO)
-
-
-
May-Nov. 90
35,000
PCB-OIL
FILLING STATION
STUTTGART
< 100
< 10 (EPA)
-
-
< 1
since Jan.91
40,000
HJ
H
O
a
#
w
* Tonnage for Oecotec's Soil Washing Plant 2000 only
-------
CLEANING EFFICIENCY
(at a former steel plant in Dusseldorf)
CTl
O1
TYPE OF
POLLUTANT
MINERAL OIL
HYDROCARBO
POLYAROMATC
HYDROCARBO
BENZENE,
TOLUENE, XYL
CHLORINATED
HYDROCARBOI
WASHED SOIL
MINIMUM
VALUE
(ppm)
Ms110
1.7
MS
.ENE
< 0.006
MS
MEAN
VALUE
(ppm)
440
6.8
< 0.6
0.009
MAXIMUM
VALUE
(ppm)
980
8.2
-
0.087
UNWASHED SOIL
MINIMUM
VALUE
(ppm)
1,230
6.8
85
MEAN
VALUE
(ppm)
2,100
14.5
2.8
113
MAXIMUM
VALUE
(ppm)
9,040
56.3
-
130
w
-d
-------
01
cr>
CLEANING EFFICIENCY
(at a former ship yard in Hamburg)
TYPE OF
POLLUTANT
MINERAL OIL
HYDROCARBO
POLYAROMATIC
HYDROCARBO
WASHED SOIL
MINIMUM
VALUE
(ppm)
20
MS
,
' 0.5
NS
MEAN
VALUE
(ppm)
161
7.8
MAXIMUM
VALUE
(ppm)
410
21
UNWASHED SOIL
(MINIMUM
VALUE
(ppm)
110
4
MEAN
VALUE
(ppm)
2,170
30.2
MAXIMUM
VALUE
(ppm)
12,000
110
H
o
d
50
M
-------
EXTRACTION AND DRYING OF SUPERFUND WASTES
WITH THE CARVER-GREENFIELD PROCESS ®
By Theodore D. Trowbridge, Thomas C. Holcombe, and Ernest A. Kollitides
Dehydro-Tech Corporation
East Hanover, New Jersey
C-G PROCESS DESCRIPTION AND ADVANTAGES
The Carver-Greenfield (C-G) Process is an innovative, energy efficient, multi-effect
evaporation, extraction process which can be applied to the clean-up of Superfund sites
holding hazardous contaminated soils, petroleum sludges, and industrial/municipal wastes.
The C-G technology is licensed by Dehydro-Tech Corporation (DTC) and its parent,
Hanover Research Corporation.
This unique process efficiently separates solid/liquid feed materials into three product
streams convenient for disposal: (1) clean, dry solids, (2) water virtually free of solids and
oils, and (3) mixture of extracted oil-soluble compounds. In so doing, any hazardous/toxic
materials present in the feed are typically concentrated into one of the three product
streams each of which is more readily handled than the feed stream. This simplifies the
handling for recycling, destruction, and/or disposal of any deleterious compounds present in
the feed.
PRESENT APPLICATIONS
f
To date, DTC has licensed over 80 C-G Process facilities in a dozen countries to solve
waste disposal problems in a wide variety of fields. More than half of the of the plants are
designed to dry and deoil slaughterhouse wastes (rendering plants). The other units are
used to evaporate water and extract indigenous oil-soluble compounds from a broad
spectrum of materials, including municipal and industrial sewage sludges, wool scouring
wastes, petrochemical sludges, wood pulp wastes, pharmaceutical wastes, dairy and food
products, textile and dye wastes, animal manure, etc.
Presented at the US EPA's Third Forum on Innovative
Hazardous Waste Treatment Technologies: Domestic and
International, June 11-13, 1991, Dallas, Texas
167
-------
PROCESS DESCRIPTION
A simplified generalized C-G Process flow diagram is shown in Figure 1. The feed
solids/water/oil mixture is screened to a maximum particle size of about 1/4 inch and then
mixed with a water-immiscible solvent to form a slurry. Typically the solvent is a food grade
mineral oil with boiling point of 400F although other solvents may be used depending on
the characteristics of the feed, the objectives of the process application, and the ultimate use
of the end products. The feed/solvent slurry is circulated through an energy efficient
evaporator system to evaporate virtually all of the water from the solids. Multi-effect
evaporation or mechanical vapor recompression is used to achieve very low energy
requirements, typically 300-500 btu/lb of water evaporated, well below the 2000-1- btu/lb
required by alternative drying processes.
The solvent fluidizes the solids and ensures a low slurry viscosity, even as the solids are
drying. The solvent also prevents scaling and fouling of the heating surfaces, thereby
maximizing heat transfer. While evaporation of the water is proceeding, the solvent extracts
oil-soluble contaminants from the solids. Any light volatile compounds present are steam
stripped out of the mixture and are recovered. By operating at low average water
concentrations in the slurry at all times, no problems are encountered with emulsions, even
with very oily sludges.
After evaporating the water, the slurry is fed to a centrifuge to separate the bulk of the
solvent from the solids. The residual solvent is removed from the solids by heating and
purging the solids with steam or inert gas. The extracted hazardous compounds, such as
PCBs, PNAs, etc., are recovered from the solvent by distillation. Any heavy metals
complexed with the oil-soluble compounds will also be extracted by the solvent and
recovered with the oil-soluble compounds. As part of the distillation process, virtually all
of the solvent is recovered and recycled to the front end of the process.
Each of the three products generated with the C-G Process is in a form that is
convenient for final disposition. In most cases, the solids are decontaminated and can be
applied directly to the ground; in the case of municipal sewage sludge the solids can be used
as a fertilizer or soil conditioner. If the treated solids still contain toxic levels of inorganic
compounds such as heavy metals, they can be transported to a hazardous landfill or
chemically "fixed" to make the metals non-leachable. Removal of the water and
hydrocarbons from the initial feed substantially reduces the volume of solids to be landfilled.
If a non-leachable form of solids is required, removal of the petroleum contaminants can
greatly improve the effectiveness of chemical fixation techniques. The extracted oil-soluble
compounds can often be refined and reused, or, alternatively, combusted to destroy the
hazardous compounds and to produce steam for processing. The evaporated water
recovered from the feed can usually be returned directly to the ground or processed in a
wastewater treatment facility.
168
-------
The C-G Process is completely enclosed and operates under a slight negative pressure,
thereby minimizing concerns about air pollution and odors during operation; it is a compact
process that utilizes conventional process equipment.
DEVELOPMENT STATUS FOR NEW APPLICATIONS
Utilization of the C-G Process to soil remediation is a simple extension of its current
applications to municipal and industrial waste streams. Presently, the process is used to
dehydrate and separate waste materials into three components (oil, water and solids) and
it is relatively easy to add one or more extraction steps to better concentrate any hazardous
material present into one of the separated components. Soil remediation applicability is
illustrated with the results of two sets of laboratory tests described below."
PAB OIL & CHEMICAL SERVICE. INC. (PAB} Oil Site.
Vermilion Parish, Louisiana
The 17-acre PAB Oil Site between Lafayette and Abbeville, Louisiana was used from
1979 until 1982 as a disposal facility for oil field waste and contains three surface
impoundments that were used to separate oil and water from drilling muds. As a result of
this operation, site soils became contaminated with petroleum hydrocarbons. Because of the
contamination and potential threat to local ground water the site was placed on the National
Priority List in 1989. The DTC laboratory processed via triple extraction/dehydration a feed
sample containing 29.5% solids, 29.1% water and 41.4% oil. Analyses revealed that there
was a reduction of the toluene content from 18,600 ppb on feed solids to less than 350 ppb
on the final solids product. Acetone, ethylbenzene, and xylenes present at 2,000-10,000 ppb
on solids in the feed were undetected in the solid product. A larger scale demonstration of
the C-G Process on material from the PAB Oil Site in a C-G Process Demonstration Unit
at Edison, NJ is presently in progress.
AMTRAK Site. New York City
The 105 acre AMTRAK Railroad property at the Sunnyside Railroad Yard in Queens,
New York is a maintenance facility for electric locomotives and railroad cars. An
investigation in 1986 revealed petroleum hydrocarbon-contaminated soil and separate-phase
petroleum hydrocarbons near a former diesel fuel storage area which contained 9
underground storage tanks which had been abandoned and filled with sand. PCB's at levels
of 5 to 360 ppm were found in the separate-phase petroleum hydrocarbons. Three samples
containing 20-28% oil, 67-74% solids and 4-6% water underwent triple
extraction/dehydration in the DTC laboratory. In all cases, following treatment with the C-G
Process, the solids products contained less than 1% oil and non-detectable amounts of water.
The feed in one sample contained 7 ppm of PCB (Aroclor 1260) and was reduced to less
than 0.11 ppb on solids in the treated sample indicating a highly efficient PCB extraction into
the hydrocarbon solvent phase. Similar results would be expected in larger scale operations.
169
-------
Other Potential Uses for Remediation
Comparable results have been achieved on other samples of remediation candidates.
Since the C-G Process is flexible, solvent selection and operating conditions can be
determined quite easily to achieve optimum processing and contaminant removal targets and
end product utilization objectives.
PILOT PLANT FLOW DIAGRAM AND DESCRIPTION
The C-G Mobile Demonstration Unit as shown in Figure 2 has design capabilities for
both batch and continuous operation. Batch operation is being used for processing the PAB
Oil Site feed described earlier. Approximately 300 pounds of feed are fluidized in 200
gallons of solvent, a food grade petroleum hydrocarbon with a boiling point of about 400 F,
in a tank in the extraction/evaporation section. After thorough mixing, the mixture is
circulated through a steam heated heat exchanger until the desired extraction temperature,
somewhere between 150 and 200 F, is reached. After maintaining the temperature for the
desired time, 15 min. to 1 hour, the mixture is transferred to a centrifuge feed tank.
The mixture is then centrifuged and the solids are separated from the solvent which
contains the majority of the indigenous oil from the feed. The solvent/oil mixture is retained
for further processing while the intermediate solids product, containing some solvent and a
very small amount of indigenous oil, is returned to the extraction/fluidization section and the
process steps described above are repeated as many times as necessary. During one of the
extractions, dehydration also takes place and the water in the feed is also evaporated by
heating the mixture to a higher temperature, typically 225 to 275 F. Lower temperatures
can be used if necessary by operating under vacuum. The water is condensed and separated
from a small amount of vaporized/condensed solvent by decanting.
After completing the desired number of extractions/evaporations the centrifuge cake is
routed to a desolventizer where the solvent is vaporized away from the solids in a
recirculating stream of hot (350 F) inert gas such as nitrogen or steam. The final solids
product is cooled and is ready for analysis and disposal. The solvent vapor is condensed by
contacting with cold liquid solvent in a scrubber and removed from the system. The nitrogen
is reheated and recirculated to the desolventizer.
The solvent containing the indigenous oil from the centrifuge is returned to the
extraction/evaporation section and fractionated under vacuum at 150-250 F in a batch mode.
Different cuts corresponding to various boiling ranges of the indigenous oil and solvent are
recovered for analysis and disposal.
Although the mobile demonstration unit is presently being operated in a batch mode,
equipment has been provided for continuous operation. In this mode, feed/solvent mixtures
are made up, processed, and transferred continuously from the extraction/evaporation
section to the centrifuge feed tank and to the centrifuge where the solids and solvent/oil
170
-------
phases are separated. For multiple extractions, the solids may be transferred back to the
extraction/evaporation section for additional processing. Alternatively, the centrifuge solids
(cake) may be desolventized and solvent/water free solids recovered. Solvent/indigenous oil
recovery would be done separately in a batch or continuous mode depending on recovery
requirements.
From this quick description, it is apparent that the C-G Process Mobile Demonstration
Unit as well as the process itself is very flexible and can be designed and operated to
accommodate a variety of feeds and feed processing requirements and objectives. For
example, the following can be readily varied: choice of solvent, extraction temperature,
evaporation temperature, number of extractions, processing order of extraction and
evaporation, centrifuging and desolventizing temperature, solvent and indigenous oil recovery
operating temperatures and pressures. .
EXPECTED OPERATING RATES
The operating rates of the demonstration unit are very much a function of the feed
properties and the required processing to meet the desired objectives. For example, for the
PAB Oil Site demonstration the feed batch size is about 300 pounds of feed at about 50%
solids, 25% water and 25% indigenous oil. It is planned to do three extractions/evaporation,
solid desolventizing and demonstrate indigenous oil recovery in about 2.5 eight-hour working
days to recover about 150 pounds of solids. Capacities will change depending on feed
composition, extraction and evaporation requirements, and indigenous oil recoveries needed.
Nominally, the unit has a capacity of about 50 Ib/hour of water evaporation and 400 Ib/hour
of solvent evaporation. The centrifuge and desolventizer can process solids at the rate of
about 30 Ib/hour.
GENERALIZED COMMERCIAL UNIT FLOWSHEET
Because of the flexibility of the C-G Process there is no single flowsheet that defines the
process for soil remediation work. Rather, the flowsheet for a particular operation will be
a combination of "modular" type operations adapted for a particular application. Guidelines
for designing these modules which are shown in Figure 3 would include the criteria described
below.
The C-G Process can accommodate numerous varieties of feedstocks containing oil,
water, and solids. Since it is a dehydration process, water should be present in amounts of
4 to 95%. Solids' particle sizes must be smaller that 1/4".
Since energy-efficient, multi-effect evaporation is the heart of the process the extraction,
evaporation, and solids separation (centrifuging) steps should be closely integrated to provide
optimum operation. This integration is dependent on the desired extraction and evaporation
requirements, which are functions of both the feed composition and the product quality
criteria, which are set primarily by regulatory requirements and end-use specifications.
171
-------
The desolventizer would be a larger unit of the same design as the one used in the
demonstration unit. Based on DTC's commercial experience, it is probable that a number
of desolventizer units in series rather that a single large unit will be most economical for
remediation work. Similar to other process variables, operating conditions will be set by
feed properties and product quality requirements.
Solvent recovery typically requires a multi-stage distillation system to separate the solvent
from various fractions of indigenous oil. The recovery system design is determined by the
firactionation requirements of the solvent and indigenous oil.
C-G PROCESS ECONOMICS
C-G Process economics are a strong function of the feed properties and capacity
requirements. Typically, the extraction/evaporation section investments are determined by
the water evaporation capacity while the centrifuging and desolventizing section investments
are a function of the solids handling requirements. The oil/solvent recovery section
investments are set by the volume of solvent used, i.e. the extraction requirements, and the
recovery specifications for both the solvent and indigenous oil. Commercial experience to
date indicates that a complete C-G Unit can be built and operated for about $50 per dry
ton of solids processed.
SUMMARY
The C-G Process is an innovative, energy efficient, and economic process which can be
applied to the clean-up of Superfund and other sites. This unique process separates
solid/liquid feed materials into three product streams (solids, oil, and water) which are
convenient for disposal or reuse. In so doing, any hazardous/toxic materials present in the
feed are concentrated into one of the three product streams each of which are more readily
handled than the feed stream. This simplifies the handling for recycling, destruction, and/or
disposal of any deleterious materials present in the feed.
172
-------
FIGURE 1
SIMPLIFIED CARVER-GREENFIELD PROCESS FLOW DIAGRAM
co
WCUUM
VENT PUMP CONDENSER
FEED
SOLVENT WPOR AND STEAM
SOLIDS
PRODUCT
EXTRACTED
COMPOUNDS
MAKEUP SOLVENT
-------
FIGURE 2
MOBILE PILOT PLANT FLOW SHEET
FEED
SOLVENT
EXTRACTION
EVAPORATION
SOLVENT
SOLVENT
WATER
SOLIDS
AND
SOLVENT
CENTRIFUGE
SOLIDS
DESOLVENTIZER
SOLVENT AND
INDIGENOUS OIL
DRY SOLIDS
TO EXTRACTION
EVAPORATION
SECTION FOR
RECOVERY
174
-------
FIGURE 3
GENERALIZED COMMERCIAL UNIT FLOW SHEET
FEED
FEED
PREPARATION
EXTRACTION
EVAPORATION
CENTRIFUGING
SOLIDS
INDIGENOUS
OIL
AND
SOLVENT
DESOLVENTIZING
SOLVENT
DRY SOLIDS
MAKE-UP
SOLVENT
RECOVERED
SOLVENT
LIGHT
INDIGENOUS
OIL
SOLVENT
RECOVERY
HEAVY
INDIGENOUS
OIL
175
-------
REFERENCES
Bress, D. R, "Energy fr,om Sludge-derived Fuels: the Hyperion Energy Recovery System", April,
1986.
"Carver-Greenfield Process ® for a Cleaner Environment", Dehydro-Tech Corporation brochure,
1988.
Cmmm, Clifton James, and Kathryn A. Pluenneke, "Development of an Efficient Biomass Drying
Process and its Commercial Use for Energy Recovery", presented to the Institute of Gas
Technology Symposium, Orlando, Florida, February 1, 1984.
'EPA Examines Oil-based Extraction Process", Process Engineering. June, 1990.
Greenfield, Charles, 'Treatment of Industrial Wastes by Multi-Stage Evaporation Drying",
presented at the 67th Annual Conference of the New Jersey Water Pollution Control
Association, Atlantic City, New Jersey, May 12, 1982.
Holcombe, Thomas G, et al, "Use of the Carver-Greenfield Process for the Cleanup of
Petroleum-contaminated Soils", presented at the New York-New Jersey Environmental
Expo '90, Secaucus, New Jersey, October 17, 1990.
Johnsson, P., et al, "Possibilities and Benefits of Bitumen Recovery in Peat Drying", presented at
the International Symposium on Peat/Peatlands Characteristics and Uses, Bemidji,
Minnesota, May 16-19, 1989. -
"The L.A. Solution", Heat Engineering. January-April, 1986.
Makansi, Jason, "Power From Sludge", Power. February 1984, p. S-l.
U.S. Environmental Protection Agency, "Dehydro-Tech Corporation (Carver-Greenfield Process
for Extraction of Oily Waste)", The Superfund Innovative Technology Evaluation Program:
Technology Profiles. EPA/540/5-89/013, November 1989, p. 31.
Walters, Sam, "Benefits from Biowaste," Mechanical Engineering. April 1985, p. 70.
176
-------
(R
B.E.S.T.^ for Treatment
of
Toxic Sludge, Sediment and Soil
Lanny D. Weimer
Resources Conservation Company
3630 Cornus Lane
Ellicott City, MD21043
(301)596-6066
Presented at
Third Forum on Innovative Hazardous
Waste Treatment Technologies: Domestic and International
Dallas, TX
June, 1991
177
-------
INTRODUCTION
During the 1960's, the Boeing Aerospace Co. invented arid patented the B.E.S.T. process
for use aboard the manned space station. Subsequent to this early development
Resources Conservation Company's (RCC), focus was on the dewatering of municipal
water treatment sludges. The process was successfully demonstrated several times and its
ability to consistently produce dry, oil free solids was confirmed with a seven ton per day
truck mounted pilot unit.
Further development of the B.E.S.T. process for treating hazardous oily wastes began in in
1984 following the re-authorization of the Resources Conservation and Recovery Act
(RCRA). Laboratory work by RCC demonstrated that the process can separate RCRA
listed petroleum refining oily sludges into three fractions; oil free water, recyclable oil and
dry, oil free solids. A second generation pilot plant was designed, constructed and operated
to demonstrate the feasibility of the process for treating oily sludges.
Based on data from the operation of this second generation pilot unit, RCC constructed a
commercial scale B.E.S.T. unit designed to treat petroleum refining sludges (pumpable
sludges). In 1987, this transportable B.E.S.T. unit completed the clean up of 3700 cubic
yards of acidic, oily, PCB-contaminated sludges at the General Refining Superfund Site
near Savannah, GA.
Recently, RCC has constructed a third generation pilot unit to develop data for a B.E.S.T.
process to treat non-pumpable wastes, such as soils and sediments. This unit is currently
being operated at the U.S. Army Corps of Engineers Waterways Experiment Station. This
project is part of US EPA.s soil and debris BOAT development program.
During the last seven years, RCC's Laboratory has evaluated many different oily sludges
and contaminated soils. These bench scale treatability test results have demonstrated that
the B.E.S.T. process is applicable for treating a wide range of wastes contaminated with
volatile organics (VOC), polynuclear aromatic hydrocarbons (PAHs), polychlorinated
biphenyls (PCBs) and pesticides.
The main objective of the B.E.S.T. process is to separate oily sludges, soils and"sediments
into a water phase that can be treated by conventional treatment and discharged a dry
treated solid phase that can be used as back fill on site and an oil phase containing the
organic contaminants that can be destroyed or recycled cost effectively.
THE B.E.S.T. PROCESS
The B.E.S.T. process is a patented solvent extraction technology utilizing triethylamine as
the solvent. Triethylamine is an aliphatic amine that is produced by reacting ethyl alcohol
and ammonia.
178
-------
The key to the success of the B.E.S.T. process is triethylamine's property of inverse
miscibility. At temperatures below 65°F, triethylamine is completely soluble with water.
Above this temperature, triethylamine and water are only partially miscible. Figure 1 shows
a graphic representation of this phenomenon. The property of inverse miscibility can be
utilized since cold triethylamine can simultaneously solvate oil and water. B.E.S.T.
produces a single phase extraction solution which is a homogeneous mixture of
triethylamine and the water and oil (containing the organic contaminants, such as PCBs,
PAHs and VOCs) present in the feed material. In cases where extraction efficiencies of
other solvent extraction systems are hindered by emulsions, which have the effect of
partially occluding the solute (oil containing the organic contaminants), triethylamine can
achieve intimate contact at nearly ambient temperatures and pressures. This allows the
B.E.S.T. process to handle feed mixtures with.high water content without penalty in
extraction efficiency.
Triethylamine is an excellent solvent for treating hazardous wastes because it exhibits
several characteristics that enhance its use in a solvent extraction system. These
characteristics include:
• A high vapor pressure, therefore the solvent can be easily recovered
from the extract (oil, water and solvent) via simple steam stripping.
• Formation of a low-boiling azeotrope with water, therefore the
solvent can be recovered from the extract to very low residual levels
(typically less than 100 ppm).
• A low heat of vaporization (1/7 of water), therefore solvent can be
recovered from the treated solids by simple heat with a very low
energy input.
• Triethylamine is alkaline (pH=10), therefore some heavy metals are
converted to the hydroxide form, which precipitate and exit the
process with the treated solids.
• Triethylamine is only moderately toxic and readily biodegrades. Data
available in EPA document EPA-600/2-82-001a shows that a level of
200 ppm triethylamine in water was degraded completely in 11
hours by Aerobacter, a common soil bacteria.
A fact sheet for triethylamine is presented as Figure 2.
A block diagram for the B.E.S.T. process is presented as Figure 3. First extraction of the
contaminated material is conducted at low temperatures, below 40°F. At this temperature,
the triethylamine is soluble with water. Therefore, the extract solution from the first
extraction stage contains most of the water in the feed material. If the first stage extract
contains sufficient water to allow a phase separation of the triethylamine and water, the
extract is heated to a temperature above the miscibility limit (130°F). At this temperature,
the extract separates into two distinct phases, a triethylamine/oil phase and a water phase.
The two phases are separated by gravity and decanted.
179
-------
RCC
Resources
Conservation
Company
FIGURE 1
U.S. PATENT (AUG. 12,1975) 3,899,419
00
o
Invi
I
1
erse Miscibility
K
Valer and Trlethylamlna K
Immlsclbla L\
1
1
I
^
Temperature
Degrees C
50
40
30
20
\\ Water and Triethylamlne
\N Mlsclble \N
% Water 0 20 40 60 80 100
100 80 60 40 20 0 % Tricthylamine
-------
J?CC
Resources
ConservJt/on
Company
TRIETHYLAMINE
(C H5)3N
FIGURE 2
00
u Inversely Miscible in Water
p Heat of Vaporization - 130 BTU/lb (BP - 194°F)
a Forms 90/10 Azeotrope With Water (BP-170 °F)
u Olfactory Detection Level In Air - 0.1 ppm (OSHA Standard - 15ppm)
Q Regulatory Status
- CERCLA Spill Reporting Level - 5000 Ibs. (max.)
- No TSCA, RCRA, or EPCRA status
- Clean Water Act Discharge Limits - None
a Mutagenicity - None
a Biodegradation - 200 ppm in water, 100% in 11 hours
u Other Properties
- Flammable
- Alkaline. pH11
• "Fishy", Ammonia-Hke Odor
-------
RCC
Resources
Conservation
Company
FIGURE 3
B.E.S.T. PROCESS CONCEPT
00
ro
Extraction
Recycled
Solvent
Waste —>
Extraction
Solids
Solvent Recovery
Solvent (to recycle)
Solvent / Oil
•>i
Subsequent
Extractions
@130°F
First
Extraction
@40°F
Separation
1
Steam
Stripping
Oil
Solvent / Water
Solvent (to recycle)
I
Steam
Stripping
Water
Solvent (to recycle)
t
•>
Solids
Drying
->Solids
-------
Subsequent extraction stages are conducted at about 130°F. At this temperature, solubility
of oil (organic contaminants) in triethylamine increases which enhances removal of the oil
from the contaminated solids. Extract from these stages are combined with the decanted
triethylamine/oil phase from the first extraction stage. The solvent is recovered from the
from the two phases via steam stripping. Residual triethylamine in the water and oil
products is quite low, typically less than 100 ppm. The recovered solvent is recycled to the
extraction vessels, directly to the solvent recovery portion of the process.
Triethylamine is removed from the treated solids by indirect heating with steam. A small
amount of steam may be added directly to the dryer vessel to provide the water required to
form the low boiling azeotrope. Typically the residual triethylamine remaining with the
treated solids is less than 150 ppm. The residual triethylamine in the treated solids
biodegrades readily. Typically, the dry treated solids can be used as back fill at the site.
The B.E.S.T. process operates near ambient pressure and temperature and at a mildly
alkaline pH. Temperatures of liquid streams within the unit vary from about 40 - 170
degrees F and high pressures are not required. The system runs at virtually atmospheric
pressure. An advantage of this type system is that it uses standard off-the-shelf processing
equipment. A low pressure nitrogen blanket creates a small positive pressure of tanks and
vessels. Since the process operates in a closed loop, with one small vent for removal of
non-condensible gases, there are virtually no air emissions.
COMMERCIAL SCALE OPERATIONAL EXPERIENCE WITH THE B.E.S.T. PROCESS
The General Refining Superfund Site located near Savannah, GA, was operated as a waste
oil re-refining facility from the early 1950's until 1975. The operation collected used
crankcase oil from local gas stations and recovered the oil by adding acid and heating
followed by filtration to remove suspended material. The treated oil was then re-packaged
and sold. The sludge produced in the process was placed in un-lined pits. The acidic, oily
sludge contained high levels of lead because the sulfuric acid added to the used oil during
the recovery process was supplemented by spent battery acid.
Analysis of the impounded sludges detected heavy metals including lead (up to 10,000
ppm) and copper (83-190 ppm). PCBs were detected in all samples at low concentrations
(5-15 ppm). The acidic, oily sludges had a pH of less than 2. Composition of the sludges
and soils at the site varied widely from point to point laterally and vertically within the
lagoons. Nominal composition was about 10% oil, 70% water and 20% solids. During
actual operation at the site, oil concentrations ranged from 0-40%, solids from 2-30% and
water from 60-100%.
183
-------
US EPA investigated several treatment alternatives for clean up of the site including;
incineration (on-site and off-site), excavation and transportation to a commercial landfill, and
on-site processing with the B.E.S.T. unit. Decision criteria included; economics, schedule
and performance. The B.E.S.T. process was selected as the most cost effective and
suitable treatment alternative for this site.
B.E.S.T. process equipment was moved to.the site and set up during 1986. Processing of
sludges and soils at the site proceeded through early 1987 and was completed in March,
1987. About 3700 cubic yards of sludges were processed at the site.
Water recovered by the B.E.S.T. process was transported to a nearby industrial waste
treatment plant and discharged. Recovered oil was heavy (API gravity = 15), but was
suitable for recycle as fuel oil. The B.E.S.T. process isolated the PCBs in the oil fraction.
Residual PCB levels in the solids were quite low (<0.1 ppm). PCBs were not detected in the
water. (Detection limit = 10 ppb).
Near the end of the project at the General Refining site, RCC and EPA cooperated in a
comprehensive sampling program. Over 160 samples of input and output streams were
collected during a 24 hour period with the B.E.S.T. unit operating at capacity. These
samples were shipped to a CLP laboratory for analysis. Results of these analyses are
presented in Figure 4. US EPA has published a report detailing ail the results, entitled,
"Evaluation of the B.E.S.T.tm Solvent Extraction Process - Twenty-Four Test". The report is
available through NTIS. '
During the final months of the General Refining cleanup, the system operated twenty four
hours per day, achieving a 70 ton per day capacity. A photograph of this unit is presented
as Figure 5.
B.E.S.T. PROCESS TREATMENT OF SOILS AND SEDIMENTS
A B.E.S.T. system configured for soils and sediments uses extractor/dryer vessels for
extraction, solids separation and solids drying. The use of the extractor/dryer vessel allows
the contaminated soil to be handled only once. The soil is not moved from the vessel until
the treatment level is reached. Extraction and drying of soil is accomplished in the vessel a
"batch mode" which provides process flexibility and increases overall system reliability.
Following excavation, contaminated soil is screened to remove material larger than 1/2 inch
in diameter. The screened material is loaded into the specially, designed hoppers that hold
about three cubic yards of soil. As illustrated in Figure 6, each hopper is transported to the
B.E.S.T. unit and lifted by crane to the feeding port on the washer dryer vessel. After the
hopper is secured, a bottom gate is opened and the material flows into the vessel. The
hopper is then removed and the feeding port is sealed.
184
-------
RCC
Resources
Conservation
Company
FIGURE 4
TYPICAL GRI OPERATION
EPA DATA
oo
tn
23% OIL
72% WATER
5% SOLIDS
5-10 mg/l PCBs
PRODUCT SOLIDS
BS * W < 27.
SOLVENT < 100 mg/l
API GRAVITY 14
PCBs 40 mg/l
0 & G 60 mg/l
SOLVENT < 100 mg/l
PCBs < 10 ppb
0 it G 0.3%
WATER < 0.5%
SOLVENT 500 mg/l
PCBs < 0.1 ppm
SOLID TO
- WASTE OIL
BLENDER
J
TO INDUSTRIAL
WATER TREATMENT
- STORED ON SITE
TOTAL METALS
TCLP METALS
As
Ba
Cd
Cr
Pb
2n
Se
< 5.0 ppm
410
3.4
21
23.000
1000
< 5.0
As
Ba
Cd
Cr
Pb
Hg
Sa
Ag
< 0.03 ppm
<0.03
< 0.05
5.2
< 0.001
0.006
-------
(Top) The B.E.S.T.
unit with oily sludge
in the foreground.
(Right) The B.E.S.T.
unit being assembled
on the waste site near
Savannah, Georgia.
B.E.S.T. is a registered
trademark of Resources
Conservation Company.
FIGURES
General Refining
Superfund site
Garden City, Georgia
The B.E.S.T® unit
cleaned 3700 tons
of contaminated
sludge during EPA
emergency response
action.
Equipment:
Skid-mounted B.E.S.T. unit
Feature:
Uses triethylamine in a
patented process to
separate toxic sludge
into solids, water and oil
Capacity:
Up to 100 tons per day
Input:
Oily sludge of variable
consistency: 25% oil,
20% solids, 55% water
Output:
Water to local industrial plant
for treatment and discharge
Oil recovered as fuel
Inert solids retained on site
Client:
Environmental Protection
Agency
Operational:
August 1986 to February 1987
(Resources
Conservation
Company
3006 Northup Way
Bellevue, Washington
98004-1407
Phone: 206 828-2400
Telex: 350166 RCC SEA
Fax: 206 828-0526
186
-------
FIGURE 6
RCC
Resources
Conservation
Company
B.E.S.T.® PROCESS STEPS
BATTERY LIMITS OPERATION
GO
SOLIDS SETTLING
AND SOLVENT DRAINING
WASHER/DRYER CHARGING
-------
Figure 7 shows a schematic diagram of a three stage wash soil cleanup unit. After
extracting the feed material with triethylamine, the solids are allowed to settle and the
solvent mixture and particulates are separated. This is accomplished by simple decantation
of the liquid contents of the washer dryer vessel. Carryover of particulates from this process
can be controlled by backwash filters or disc centrifuges if necessary.
After each extraction step, the spent solvent is discharged from the extractor/dryer vessel
into the spent solvent storage tank. The decanted extraction solution of triethylamine,
water, and oil is then heated to about 140°F to separate water. Because much of the water
can be separated without distillation there is an energy savings. In some cases, however, it
may be necessary to distill the water if triethylamine/water separation is adversely affected
by matrix interferences such as detergents, surfactants, or some types of synthetic oils.
Empirical treatability testing is currently the only means of determining the efficiency of the
separation of triethylamine/oil from water for a given waste sample.
The solvent/oil phase is steam stripped to recover the solvent for recycle to the
extractor/dryer vessel to extract additional soils. The recovered oil is stored for subsequent-
treatment or recycle. Water from the steam stripper is treated and may be suitable for direct
discharge.
Upon final extraction of the soil, the extractor/dryer vessel is heated by direct and indirect
addition of steam. The solvent is vaporized and recovered for recycle. The energy required
for evaporation of the solvent is low, approximately 1/7th that of water. Upon completion of
the drying step the solids are unloaded from the extractor/dryer into a 3 cubic yard hopper
and sent to final disposition. Some water (about 5% by weight) will be left in the solids to
control dusting.
BENCH-SCALE TREATABILITY TESTING
In order to evaluate each potential application for the B.E.S.T. process, RCC has developed
a low cost glassware test protocol that requires only one kilogram of sample material. A
schematic for the bench scale treatability test protocol is provided in Figure 8.
Initially in the bench-scale treatability test, the waste material is analyzed to characterize its
phase composition; oil, water and solids. A 400 gram sample is adjusted to a pH = 10 and
mixe'd with triethylamine to three extraction stages. Following extraction, the treated solids
are separated from the extract (oil, water and solvent) in a floor mounted centrifuge.
188
-------
RCC
Resources
Conservation
Company
FIGURE 7
B.E.S.T.®SOIL CLEANUP UNIT SCHEMATIC
00
ID
SCREENED
CONTAMINATED:
SOIL
CENTRIFUGE
CENTRATE
•—M-
CONDENSER
V
CONDENSER
r
T T- T T
WASHER/DRYER
IPENTJ | 1st
VENr,-
WATER
STRIPPER
DRY, WASHED SOLIDS
SOLVENF
EVAPORATOR
PRODUCT
WATER
CHILLER
-------
RCC
Resources
Conservation
Company
FIGURE 8
SOLVENT EXTRACTION TECHNOLOGY
GLASSWARE TESTING APPROACH
ID
O
SLUDGE
SAMPLE
STEAM STRIP FOR
SOL VENT RECOVERY
RESIDUAL SOLVENT & H2O
OIL TESTS (API GRAVITY. BS & W)
ICP METALS ANALYSIS
SLUDGE
CHARACTERISTICS
ANALYSIS
COMPOSITION
- %OIL (SOXHLET EXTRACTION)
- % SOLIDS (DRY@ I05*C)
- % WATER (BY DIFFERENCE)
TOTAL METALS
- DRY @ I05*C. ASH @ 550'C
- HEAVY METALS BY HNO3 DIGESTION
HANDLING CHARACTERISTICS
-pH
- VISCOSITY (VISCOMETER)
• STEAM STRIP FOR
SOLVENT RECOVERY
• TOTAL ORGANIC CARBON
• OIL & GREASE
TOTAL SOLIDS
TOTAL DISSOLVED SOLIDS
RESIDUAL SOL VENT
TOTAL METALS
(HNO3 DIGESTION)
DRY @ I05*C
OIL & GREASE RESIDUAL
TOTAL METALS (AQUA REGIA)
EP TOXIOTY PROTOCOL
-------
The extract solution is then heated and allowed to decant in a separatory funnel into distinct
oii/TEA and water phases. Stripping of the triethylamine from the oil and water phases is
conducted in a rotovap apparatus operating at atmospheric pressure. The recovered
fractions are analyzed to determine the partitioning of the contaminants and to determine
the quality of the treated solids.
B.E.S.T. SOLVENT EXTRACTION PROCESS
The B.E.S.T. solvent extraction technology has been demonstrated as an effective process
for treating PCB contaminated wastes at bench, pilot and full scale. Recently, RCC has
successfully completed bench scale treatability testing with PCB contaminated soils
samples from natural gas pipeline compressor stations. .
The bench scale treatability testing was conducted with three samples from two compressor
stations. One "Hot Spot" soil sample was collected from one compressor station. Two
additional "Composite" samples were collected from another compressor station. Results of
the bench scale treatability testing are summarized below:
PCB Removal Summary
Sample
"Hot Spot"
"Composite" I
"Composite" II
PCBs in
Feed, mq/kq
2,000
550
510
PCBs In Product
Solids, mq/kq
1.5
0.77
1.2
Removal
Efficiency. %
99.9
99.9
99.8
PILOT-SCALE TREATABILITY TESTING
RCC has available a pilot-scale B.E.S.T. unit configured to demonstrate the full scale
performance of the process. The unit has a capacity of 100 pounds per day of treated dry
solids. Recently, RCC received a nationwide TSCA permit to use this unit for on-site
demonstration testing with PCB contaminated wastes. A photograph of the pilot unit is
presented in Figure 9.
The B.E.S.T. pilot plant is a solvent extraction system capable of processing liquid sludges
or contaminated soils. Its main function is to demonstrate the process fundamentals
through the use of scale-down versions of actual equipment components, namely, the
extractor/dryer, the solid-bowl centrifuge, decanters, stripping units, etc.
The pilot plant consists of two portable skids. The main skid (20 ft X 8 ft) contains all of the
B.E.S.T. process equipment including the extractor/dryer, the centrifuge, tanks, pumps, and
heat exchangers required to process contaminated sludges or soils. All electrical
equipment is Class 1, Group D, Division 1-rated.
191
-------
B.E.S.T. Pilot Plant.
Reverse side of pilot plant.
FIGURE
B.E.S/T® Pilot Plant
Mobile field unit
demonstrates the
effectiveness of the
B.E.S.T. process
on location with
on-site feeds and
actual operating
conditions.
Equipment:
Skid-mounted module contains
all B.E.S.T. unit operations to
mix, extract, centrifuge,
evaporate, decant, strip and
dry. Module is easily
transported on highways.
Capacity:
6-10 gallons per batch
Input:
Pumpable and non-pumpable
sludges, sediments or soils.
Output:
Dewatered oil
Stripped water
Dry solids
Site requirements:
20' x 50' process area.
Industrial air and electricity.
(If needed, all utilities can be
supplied by RCC utility skid.)
Operational:
March 1989
RCC
Resources
Conservation
Company
192
3006 Northup Way
Bellevue, Washington
98004-1407
Phone: 206 828-2400
Telex: 350166 RCC SEA
Fax: 206 828-0526
-------
The second, smaller skid (10 ft X 8 ft) contains several utility systems to support the
operation of the main skid. The utility skid includes a refrigeration unit and a tempered
cooling water system. Steam, nitrogen, and instrument air are provided separately.
When processing soils the pilot plant is configured with a extractor/dryer vessel for
extracting and drying the soils. The extractor/dryer is a horizontal cylindrical vessel which
has a rotating shaft with mixing paddles attached. Performance of the extractor/dryer is key
for successfully treating soils with the B.E.S.T. process. A Littleford Model FM-30
extractor/dryer vessel is used on the pilot plant. This model extractor/dryer vessel is
identical to the units Littleford uses in their pilot tests for scale up to commercial scale units.
Therefore the extraction and drying performance of unit is comparable in the larger scale
units.
Resources Conservation Company (RCC) has successfully completed a pilot-scale
demonstration of the B.E.S.T. process configuration for treating PCB contaminated soils.
The demonstration was conducted in November, 1989, at a site in the Mid-western United
States.
Objectives of the on-site pilot-scale test were:
• Demonstrate the B.E.S.T. process batch extraction configuration for treating
soils.
• Verify that the B.E.S.T. process can achieve the PCB clean up level (<10 ppm)
for the site.
• Demonstrate complete solvent recovery and recycle.
• Provide an overall mass balance.
Approximately 1000 pounds of soil with a PCB contamination level of 130 ppm was
processed in eighteen distinct batches. On-site testing included process optimization of
PCB extraction and soils drying. The PCB clean up standard for the site was established at
<10 ppm. The process was optimized throughout the test such that the levels of residual
PCBs consistently decreased. The final five batches contained residual PCBs in the treated
soils near 2 ppm. No PCBs were detected in the recovered water at a detection limit of 20
ppb.
The residual solvent levels in the B.E.S.T. treated solids was 130 ppm. The recovered
water had a residual solvent concentration of less than 10 ppm. With these very low
residual solvent concentrations in the treated products, the solvent recovery and recycle
rate is greater than 99.99%.
Mass balances calculated with the test results showed excellent accounting for PCBs and
solids. The test results verified that the nearly complete solvent recovery and recycle can
be achieved with soils treatment system configuration. The solids recovery was 98% and
the PCB recovery was 95%.
193
-------
The process was rigorously field tested for sensitivity to a wide range of possible parametric
variations. For example, extraction times were varied by as much as 50%. Results
indicated that B.E.S.T. treatment offers substantial process flexibility while meeting cleanup
standards. Process flexibility is important when treating sites containing heterogeneous
contaminated matrices, variable contaminant concentrations and multi-component
contamination which is so often found on Superfund sites while meeting the mandates of
site clean-up standards or permit restrictions. The favorable results and demonstrated
process flexibility support the assumptions which serve as the basis for full scale design,
operation and scheduling.
CONCLUSIONS
• The B.E.S.T. process is a proven technology for treatment of
contaminated sludges, soils and sediments.
• The B.E.S.T. process is an effective method of concentrating PCB's in the
oil fraction, thereby significantly reducing the cost of PCB destruction.
• The B.E.S.T. process offers a cost effective, viable alternative to land
disposal of hazardous wastes.
• The B.E.S.T. process can be used as an on-site materials handling step
which can significantly reduce the amount of hazardous material that is
transported off-site for disposal. This reduces the overall cost of site clean
up and limits the risks associated with transporting hazardous wastes over
public highways.
• Solvent recovery from the B.E.S.T. treated products is excellent
(>99.998%).
• Removal efficiency for PCBs, polynuclear aromatic hydrocarbons (PAHs)
and pesticides with the B.E.S.T. process is excellent (>99%).
194
-------
PURIFICATIOH OF LANDFILL LEACHATE
BASED ON REVERSE OSMOSIS AND
ROCHEM DISC TUBE MODULE DT
THOMAS A. PETERS
Dr.' Peters Consulting for Membrane Technology
and Environmental Engineering
Broichstr. 91, 4O4O Neuss, Germany
Tel. 49-2101-548127, Fax 49-2101-545040
EPA Third Forum on Innovative
Hazardous Waste Treatment Technologies:
Domestic and International, June 11 - 13, 1991
Dallas, Texas
195
-------
PURIFICATION OF LANDFILL LEACHATE
BASED ON REVERSE OSMOSIS AND
ROCHEM DISC TUBE MODULE DT
THOMAS A. PETERS
Dr. Peters Consulting for Membrane Technology
and Environmental Engineering
Broichstr. 9.1, 4040 Neuss, Germany
Tel. 49-2101-548127, Fax 49-2101-545040
SUMMARY
The ROCHEM Disc Tube Module Type DT is a innovative module
system for reverse osmosis applications in desalination and
purification of liquids that can operate effectively and
economically at increased turbidity and Silt Density Index
levels.
The disc-membrane stack of this module is assembled using
hydraulic discs alternating with membrane cushions on a central
tension rod with metal end flanges and is covered by a standard
8 inch pipe. This design allows for an easy alteration of
module length and tube materials and facilitates inspection or
changing of membranes in a quick and easy manner. The open feed
channel between the hydraulic disc and the membrane surfaces
ensures an unrestricted feedwater flow and consequently an
optimal membrane cleaning from fouling deposits.
The open channel ROCHEM RO DT module can operate without
problems on sea water which has been pretreated only with
normal mechanical filters such as a sand filter and a cartridge
filter, usually making acid and antiscalant pretreatment
unnecessary. Due to the low pretreatment requirements of the
module, it can be used for industrial waste water applications
like landfill leachate purification.
Design details of the DT module and results from practical
field work in landfill leachate applications are discussed.
1. INTRODUCTION
Similar to the days of the development of reverse osmosis
technology as a classical application in the desalination of
seawater in order to produce potable water, RO has lately
turned into1 an important instrument in environmental
engineering.
Due to the ability of modern high rejection reverse
osmosis membranes to retain both organic and inorganic
contaminants dissolved in water with high efficiency, a wide
variety of industrial waste waters can be treated. Thus reverse
osmosis is useful as a main step in the purification of liquid
waste like landfill leachate and contributes to solving growing
water pollution problems.
196
-------
Besides requiring highly resistant membranes
leachate demands the use of open
W*»~W^>~(W^> I WWf I WU krlll W I I kf
of landfill leachate demands the use of open channel module
systems. In comparison to sea water desalination there are more
problems with scaling, fouling, and especially biofouling.
Fu rt h*=. rmnre» -f oa-t-ii rp>e= 1 i k*=> h-inh fa-f -f n r-i *=nnv oaerw h^nrfl •? no "I nuj
Furthermore features like high efficiency., caoy i
energy consumption and optimized operation costs
by the potential customers, who at the same time
membrane technology under rough working condition
by the potential customers, who at the same
membrane technology under rough working condi
as possible in an unattended operation mode
usina select membranes and a special olant
easy handling, low
costs are expected
want to use
using select mcmur
these requirements.
far
DT module,
can meet
2. DISC TUBE MODULE SYSTEM
2.1 History
Based on the plate .and frame module technology which was
originally developed by the West German Research Center in
Geesthacht (GKSS), over ten years ago the ROCHEM company
started to produce reverse osmosis systems for water
desalination. In the initial phase, the company's activities
were focussed on systems for the production of potable water
from sea water to be installed on ships and for the offshore
industry. The standardized PLC controlled modular systems were
designed for a fail-safe unattended operation and in order to7
ensure for the end user 24 hours per day reliability and high
product water quality.
Utilizing the long term experience gained in the
desalination of sea water and later in brackish water and tap
water treatment a new module system was developed - the Disc-
Tube Module. The successful design of this module was made
possible among others by its development as an integrated part
of a total RO system concept, as ROCHEM possesses the rare
capability of being both the module manufacturer and the
systems manufacturer.
2.2 Fundamentals
The patented DT module represents a innovative concept for
membrane applications. It can operate effectively and
economically at increased turbidity and Silt Density Index
levels for reverse osmosis applications.
The DT-module is a fully modular design comprising the
following components /!/:
- membrane cushion
- hydraulic carrier disc with internal 0 ring
gasket
- end flanges with lip seals
- tubular pressure vessel
- central rod with tailpiece and nut
2.3 Membrane Cushion
The membrane cushions for the DT-Module are produced from
flat membranes with internal fabric- fleece (for drainage bf the
permeate). By contrast with other design concepts, the membrane
cushion is octagonal and has a round opening in the center
which is used to drain the permeate during operation (Fig. 1).
The outer edges of the membrane cushions are welded by an
ultrasonic method. This • welding method is based upon a
197
-------
development by the GKSS Research Center, Geesthacht, and has
already been used succesfully in another module system since
1985.
2.4 Hydraulic Disc
The hydraulic discs of the DT-Module (Fig. 2) have at their
outer edge a washer which extends equally on both sides. In the
center are a ring gap formed by ribbing through which the raw
water runs during operation, a gasket groove which takes the 0
ring seal between membrane cushion and hydraulic disc, and
the opening for the tension rod with the drainage gaps for the
permeate.
Fig. 1: Membrane cushion for the DT module
Fig. 2: Hydraulic disc for the DT module
2.5 Me«brane Element Stack
These hydraulic discs and membrane cushions are stacked
alternately on the tension rod to form a membrane stack (Fig.
3). Open flow channels are formed in this way on the raw water
side between the hydraulic discs and the membrane cushions
fitted inbetween. -
-------
The completed • membrane stack is bordered at both ends by
end flanges with lip seals. The stack is compressed . by
thightening the nut on the tension rod. The reaction is
provided by a joining flange, through which the permeate is
drained during operation. -
Fig.
Membrane
element
stack of
the DT-
Module
2.6 Module Assembling
The preassembled stack of membrane elements is inserted
into a pressure pipe vessel (Fig. 4). This pressure vessel not
only holds the stack of membrane elements but also acts as a
hydraulic component, since a gap is formed between the inner
wall of the pipe vessel and the gaskets at the edge of the
hydraulic discs, which guides "the incoming feed to the ring gap
in the first hydraulic disc of the membrane element stack.
Fig. 4: Insertion
of the membrane
elemtent stack
into the
pressure vessel
2.7 Principle of Operation
The principle of operation of the DT-Modul-e assembled from
the components described above is shown in Fig. 5.
199
-------
DT DISC-TUBE
»"•» MEMBRANE
W.^ M O D U L E
Hydraulic Flow Schematic
Fig.5: Principle of
operation of
the DT-Module
As a result of the special design features of the DT
module, there are a number of aspects which have a favourable
effect on the process and the operation of the module:
* The extremely short flow paths (6 cm) along the membrane,
before renewed mixing of the feed by the deflection at the
outside edge of the hydraulic disc or the ring gap, limit the
concentration polarisation.
* As a result of this limiting, the system is unlikely to
block, keeping the permeate flow at a high level.
* The delayed formation of a secondary membrane allows
longer cleaning intervals and thereby extends the service life.
* The clearly defined sealing functions, i.e. the sealing
off of the feed-side high pressure area from the permeate-side
low pressure area with O rings, and the sealing of the feed-
side high pressure area against the ambient pressure by lip
seals, avoid sealing problems in the high pressure area.
* The transfer of the operation pressure stress from the
membrane element stack to a commercially available 8 inch
pressure pipe vessel increases the operational reliability,
simplifies the system and the assembly procedure and also helps
to reduce the cost of the module.
* Since the sealing pressure- of, the edges of the hydraulic
disc in the compressed state is limited deliberately, the
optimised emergency flow characteristics enhance the
operational reliability of the module. Even extreme pressure
differences do not result in destruction of the module, but
only in a limited local breakdown within the system.
* Due to the open ring gaps or flow paths between the
membrane cushion and the hydraulic discs, even fluids with high
colloid or solids content can be processed without problems.
* The open channels allow efficient cleaning of the module,
since the interfering substances which have been removed from
the membrane by the cleansing agent can be carried away freely.
* The relatively high packing density for an "open" module
system of 200 m2/m3 allows the construction of compact modules
and installations.
-------
As further advantages one can mention:
* Distortion of the membrane element stack internally or
during insertion into the pipe vessel is avoided by the
symmetrical fully modular design.
* Flexibility of the application of the^Tnodule, because the
hydraulic discs as well as the membranes used with them can be
chosen to suit the particular problem (membrane index).
* Reduction of operating costs through quick and easy
assembly and dismantling of the module for membrane
replacement.
* Cost effective operation, because the narrow gap
technology only requires small fluid volumes and therefore low
energy input to achieve sufficient flow across the membranes.
3. Treatment of Landfill Leachate
3.1 General Aspects
Due to its comparatively low requirements i^n the
pretreatment, the DT module system can be used in a wide range
of industrial waste water treatment.
Among others which have gained importance in recent years
are the treatment of landfill leachate, as toxic and hazardous
compounds can originate from landfill leachates as a result of
the soluble components of solid and liquid wastes being leached
into surface and groundwater.
The landfill leaches are comparable to complex industrial
waste streams which contain both toxic organic and inorganic
contaminants. In cases where drinking water supplies may be
impacted or where there is surface breakout, remedial actions
must be undertaken. At .some landfill sites leachate which is
collected is discharged to municipal sewers to treatment _in
conventional sewage treatment processes. At other landfill
sites where sewer lines are not available the leachate water
may be hauled off site for treatment, may be recycled back to
'the landfill site or may be discharged to a surface water
source. Hauling leachate for treatment by conventional
processes is expensive and new regulations may limit the
discharge of such complex wastes to municipal sewers since the
contaminants are often not appropriate for treatment by
conventional municipal biological processes. Recycling leachate
within the landfill site itself can reduce the amount of liquid
leaving the site but can only provide a short term solution.
Effective and inexpensive methods of treatment of the material
need to be,developed /2/.
One possible solution is the use of reverse osmosis as a
main step in a 'landfill leachate treatment chain composed of
different processes. The residual organics and dissolved solids
are concentrated and reduced in volume by .further treatment,
and a treated water stream is generated with only very _low
levels of inorganic and organic contaminants meeting mainly
potable wafer standards /3/.
201
-------
3.2 Examples for Landfill Leachate Treatment Plants
Using ROCHEM disc-tube module and the adequate membrane
optimized results are obtained in different landfill leachate
treatment plants installed in Germany, f.e.:
* Schwabach, start up June 6, 1988, two RO-stages with DT-
Modul.
* Ihlenberg (former Schonberg), start up January 8, 1990,
36 m3/h feed capacity (reverse osmosis ,as main step of a
landfill leachate purification process avoiding any waste
water, the most modern and largest multi-stage plant that has
been build up to now for this application).
* Kolenfeld, start up in June 1990, contanerised plant with
two stages'with DT-Modul, ca. 30 m3/d feed capacity.
The data related in table 1 show some examples for the
plant with two stages at Schwabach (see figure 6).
Table 1: Landfill leachate treated by the reverse osmosis plant
with DT-module in two stages in Schwabach.
Source: ZVSMM
Parameter Feed water Permeate
mg/1 second stage
mg/1
Retention
sulf ate
chloride
ammonium
nitrate
COD
TOC
hydrocarbons
nickel
chromium
copper
vanadium
22093.00
6364.00
1955. OO
455.00
912.00
v 289.00
13.40
2.78
2.18
0.97
290.00
4.80
14.00
42.00
18.00
15.00
4.00
0.30
0.10
0.10
0.10
2 -20
99.90
99.80
97.90
96. OO
98.40
98.60
97.80
96.40
95.40
89.60
99.20
The plant itself was installed turnkey from ROCHEM RO-
Wasserbehandlung GmbH, Hamburg, at the waste treatment center
of the Zweckverband Sondermullplatze Mittelfranken (ZVSMM) in
Schwabach (near Nurnberg). Start up was June 1st, 1988.
202
-------
Fig. 6: Reverse osmosis plant with two stages for
treatment of landfill leachate with DT module
at Schwabach
Source: ROCHEM
Another example for landfill leachate treatment with DT-
module is a plant at the Deppnie of the Ihlenberger
Abfallentsorgungsgesellschaft (former Deponie Schonberg), where
the production of a minimum of 100,000 m3 per year of permeate
was guaranted by UNION RHEINBRAUN Umwelttechnik GmbH, who
designed and installed the whole leachate purification plant
turnkey, where reverse osmosis is the main treatment step.
As far as we are aware this purification plant at the
Schonberg disposal site is the most modern and largest multi-
stage plant that has been built till now on an industrial
scale for this application /4/.
Figure 7 shows the first stage of the reverse osmosis
treatment, where 3 identical systems equipped with the DT
module are operating in parallel. Table 2 and table 3 relate
some results of this plant, that in the first year of operation
(start up was January 8th, 1990) has produced 126.OOO m3 of
permeate.
203
-------
Fig. 7: Supply units and DT module banks for
landfill leachate treatment at Schonberg
Source: IAG/UNION RHEINBRAUN Umwelttechnik
Table 2: Results obtained from landfill leachate purification
by the two-stage reverse osmosis plant at Schonberg, analysis
from March 1990.
Source: IAG/URT
leachate
permeate
stage 2
rejection
pH-value
el. conductivity mS/cm
COD
BOD
sodium
chloride
calcium
magnesium
ammonium
arsenic
cyanide
mg02/l
mg02/l
mg/1
mg/1 •
mg/1
mg/1
mg/1
mg/1
mg/1
7.7 6.6
12.250 382
2,619
184
3,255
3,091
192
97
380
0.25
2.35
1.2
2.5
2.4
2.7
0.9
0.3
O.4
n.m
n.m
99.9
99.9
98.6
99.9
99.9
99.5
99.7
99.9
>99.99
>99.99
204
-------
Table 3: Results from same plant as in table 2, analysis from
October 5, 1990.
Source: IAG/URT
leachate
pH-value
elec. conduct
(MS/cm)
COD
mg/1
Ammonium
(mg/1)
Chloride
(mg/1)
Sodium
(mg/1)
heavy metals
7.7
17,250
1,797
366
1 2,830
4,180
0.25
permeate I
6.8
382
<15
9.8
48.4
55.9
<0.005
permeate II
6.6
2
< 1 5
0 . 66
1.9
2.5
99.
99.
99.
99.
> 98
s .
9
2
9
9
9
4. FUTURE
The combination of an open channel design with the narrow gap
technology and a new membrane cushion concept in the advanced
ROCHEM module system type DT' (Disc-Tube) makes possible the use
of reverse osmosis in applications like desalination,
industrial waste water treatment or aqueous fluids processing,
even if the water to be treated has a high Silt Density Index
respectively high turbidity. Furthermore the high flexibility
regarding the choice of material for the membrane and the other
module components allows the adaptation of reverse . osmosis
systems to particular problems within a wide range of chemical,
physical and technical requirements, like it is necessary f.e.
for a successful treatment of ,landf ill leachate /5/.
The cost effective operation because of comparatively low
energy and chemicals demand, a favourable price/performance
ratio and the convincent results obtained during operation of
this type of module for cleaning landfill leachate demonstrate
the innovation potential and the growing importance of reverse
osmosis technology in environmental engineering and pollution
control.
205
-------
5. REFERENCES
1. ROCHEM RO-Wasserbehandlung GmbH., Company
brochure, 1989, Hamburg
2. Krug Th. A. and S. Me Dougall: Preliminary
Assessment of a Microfiltration/Reverse
Osmosis Process for the Treatment of Landfill
Leachate, Proceeedings 43rd Purdue Industrial
Waste Conference, Chelsea, Michigan, 1989
3. Peters, Th.: Deponie Sickerwasser durch
Umkehrosmose aufbereiten, UMWELT, 20 (1990)3
4. Peters, Th., G. Braun and H.-H. Gierlich:5
Treatment of Landfill Leachate by a Multi-
stage Process Avoiding any Waste Water,
Proceedings ENVIROTECH, Vienna, 1990
t
5. Peters, Th. : Desalination and Industrial Waste Water
Treatment with the ROCHEM Disc Tube Module DT, Proceedings of
the Twelfth International Symposium on Desalination and Water
Re-Use, Malta, 1991
206
-------
Innovative Concept
for Evaluation of In-Situ
Treatment of Contaminated
Soil and Groundwater
Presented by
Leo B. Langgaard
Presented to
United States Environmental Protection Agency
Third Forum on Innovative Hazardous Waste
Treatment Technologies: Domestic and International
June 11 to 13,1991
Dallas, Texas, U.S.A.
B. H0jlund Rasmussen
Consulting Engineers and Planners A/S
N0rregade 7 A
DK - 1165 Copenhagen K
207
-------
Abstract
A peninsular shaped site in a fishing harbour situated close to
the greater Copenhagen area was found to be contaminated with
volatile aromatic compounds, water-soluble organic solvents, chlo-
rine-containing organic solvents and hydrocarbons. Since 1914 the
site has been used as a storage and handling area. At first it as
used for fuel oil products and later for organic solvents. Leaking
of materials from the storage area uncovered the pollution and a
number of separate investigations were carried out. The storage
area has been characterized horizontally and vertically with the
following activities: drilling, chemical analyses, ecotoxicologi-
cal tests, soil characterization, groundwater modelling, groundwa-
ter treatment, and full-scale in-situ treatment of soil and
groundwater combined with a laboratory study on the degradation
potential on soil from the actual site.
Two separate test areas (each of 25 m2) on the site were erected
with strippers down to a depth of 4 m2 in combination with air ex-
traction units. The test area was characterized with chemical ana-
lyses on soil and groundwater before and after a test period of
about 2 months. Soil samples were also tested for the content of
micro-organisms.
The test results showed that values of the residual concentration
in soil and groundwater could be accepted by the Copenhagen Coun-
ty, especially for the soil but also for the groundwater after a
treatment on site and recirculation to the site.
As a result of the promising full-scale studies, a total full-
scale in-situ and on site treatment plant is now in the phase of a
detailed planning. The treatment plant will be erected in the be-
ginning of 1992, when the frost has left the soil and the weather
is getting warmer.
208
-------
1.
2.
2.1
2.2
Introduction
Pollution from a storage area was discovered in the very
cold winter in 1986 when leaking materials coloured the
ice black in the surrounding fishing harbour. Since 1914
the site has been used for storing and handling of fuel
oil products and a range of different organic solvents.
Several separate investigations have been carried out
both to characterize the contamination horizontally and
vertically and to identify the technically and economic-
ally optimal remediation technology.
Some of the investigations are summarized in the follow-
ing.
Status
Figure 1 shows the site with the monitoring wells and
test areas.
Soil characteristics
More than 20 monitoring wells were established to give
information on the soil and groundwater quality both ho-
rizontally and vertically. As a result of this, the geo-
logical profiles show that the upper 4 m consist of a
mixture of sand, gravel, peat and fill. Underneath, the
formation consists of 7 m of undisturbed moraine sedi-
ments (clay), followed by Danian limestone. The main part
of the upper 4 m is sand probably marine sand pumped'from
the coastal areas nearby.
Hydrogeology
The groundwater level has been registered throughout the
investigation period. The hydraulic properties have been
tested. They showed a relatively sm^ll transmissivity in
the upper 4 m of the soil (2-4-10 m2/sec.). The tests
indicated that the water should be abstracted from cen-
tral wells combined with horizontal drains.
2.3
Chemical analyses
Chemical analyses of soil and water samples showed that
the contamination consisted of fuel oil products and
organic solvents from the following groups of products:
- Volatile aromatic compounds and fuel oil products
(e.g. benzene, toluene, xylenes, MIBK, trimethylbenze-
ne, total hydrocarbons)
- Water-soluble organic solvents (e.g. methanol, ethan—
ol, acetone, isopropanol, N-butanol).
'209
-------
2.4
- Chlorine-containing organic solvents (e.g. chloroform,
1.1.1-trichlorethan, trichlorethylene, tetrachlorethy-
lene).
A summary of constituents in groundwater and soil is
shown in Table 1. As indicated in the chemical analyses
an unsystematic variation must be expected for the total
handling and storage area.
Ecotoxicological tests
The toxicity of selected groundwater samples were deter-
mined by microtox tests. Effect concentrations which
cause 20% and 50% inhibition (EC 20/EC 50) were measured.
The values had the following variation
2.5
EC 20 ml/1
EC 50 ml/1
Variation
18 - 350
76 - >500
One of the ecotoxicological tests showed that the ground-
water had to be diluted more than 1.250 times to avoid
acute toxicity to bacterial systems.
Pollution distribution
The distribution of contaminants on the site is evaluated
with the indicator variables VOC (volatile organic car-
bon) and VOX (volatile organic halogen) combined with the
toxicity tests EC 20 and EC 50. The measured values are
shown in Figure 2.
2.6 Aeration and aerobic stabilization
To evaluate the potential for remediation of polluted
groundwater, aeration combined with aerobic stabilization
were tested in laboratory scale. The tests were followed
by VOC, NVOC (non-volatile organic carbon), VOX and the
microtox tests.
After an aeration of 48 hours the concentration of VOC
was reduced to half of the measured value and VOX was re-
duced to below the detection limit. The toxicity decreas-
ed from 40 ml/1 to 130 ml/1 for the EC 20 value.
To evaluate the possibility of microbial degradation
after aeration, tests for aerobic stabilization with
microorganisms from a municipal waste water treatment
plant were carried out. As a result of this it was con-
cluded that the residual products from aeration were
degradable in an aerobic biological treatment plant.
210
-------
Strategy for remediation technologies on the actual site
The total battery of national and international technolo-
gies crystallized into ideas with in-situ treatment of
soil and groundwater and on site treatment of contaminat-
ed groundwater combined with a recirculation of water
with additives (nutrients, bacteria, oxygen etc.).
To follow up and document the ideas it was necessary to
carry out further investigations. The normal procedure
for such studies can be outlined as:
Phase I:
4.
4.1
Phase II:
Phase.Ill
Studies on a laboratory scale (batch) for a
relatively simple characterization of stripp-
ing (degradation) on soil and groundwater
separately or in combination. Soil and
groundwater are characterized with chemical
variables before and after testing together
with the content of micro-organisms.
Laboratory studies with soil columns are
normally the next step in order to evaluate
the rate of stripping in soil and the micro-
bial degradation.
The next step is to transform laboratory
studies to full-scale in-situ pilot tests on
representative areas of the total storage
area.
Even with the greatest amount of carefulness in sampling,
transportation and execution of studies in the labora-
tory, the transformation from the in-situ environment to
the laboratory involves great , risk that the results be
affected. This is common knowledge and not further ela-
borated in this context.
As a result of these considerations it was concluded to
study decontamination directly in full-scale combined
with studies on the degradation potential in laboratory
scale. The activities are further described in the fol-
lowing text.
Studies on remediation technologies
Introduction
The following technologies were selected to remediate the
site, an area of 8,000 m2 down to a depth of 4 m:
- In-situ stripping with air and extraction of air from
the aerobic zone.
211
-------
- Abstraction '- of contaminated water, oil separation,
stripping and microbial degradation of water and re-
circulation to the site.
4.2 In—situ stripping
In-situ stripping in soil and groundwater partly affects
the volatile contaminants, which are stripped off and
partly the microbial activity, which is expected to be
optimized as a result of the aerobic conditions. The
purpose of the activities was to measure the combined
effect in soil and groundwater. To evaluate the degra-
dation potential, laboratory studies were carried out
after the full-scale testing period.
On two selected localities with areas of approx. 25 m2,
studies were carried out in-situ, stripping in the
groundwater zone and extraction of air in the aerobic
zone.
The areas were selected after an evaluation of several
monitoring wells had shown the most promising geological
profile for the experiments. An outline of in-situ
stripping and extraction of air in soil and groundwater
air is shown in Figure 3. It is illustrated that air is
stripped in each corner of a square and in the middle.
Air is extracted from four wells. Before and after the 2
month test period, samples of soil and groundwater were
analysed for the variables mentioned earlier. The results
of the chemical characterization is shown in Tables 2 to
7.
4.3
Microbiology
Soil and water samples have been characterized, with re-
spect to the following parameters:
- Total bacterial counts (Acridin Orange Total (AOTC)).
- Plate counts .with a variable content of nutrients.
- Enzyme activity.
- Growth tests at 10°c and 20°c with different types of
substrates including contaminated groundwater from the
site.
The results were:
- Thentotal .ppntent of micro-organisms were measured to
1CTU to 10Xi per g.
- The existing population/microorganisms would be able
to degrade the components aerobically.
212
-------
Micro-organisms in soil combined with groundwater from
the site accelerate the degradation of the contami-
nants.
The studies on the decomposition potential of the or-
ganic components indicate a microbial degradation
which can be optimized in an environment with oxygen
as arranged with stripping. Additives with substrate
is not necessary to start up the microbial degrada-
tion.
Air emission
As an indicator on the emission of components to the air,
selected constituents were measured once a day during the
first period and later, depending on actual concentra-
tions with longer time intervals. The following test
tubes were examined: 1.1.1-trichloroethan, toluene, al-
cohol, perchlorethylene. Only toluene and perchlorethy-
lene gave an indication on the tubes.
Concept for remediation
As a result of the different investigations, the follow-
ing remediation activities will be carried out on the
site:
Stripping in-rsitu (depth 4m).
Extraction of contaminated air from the vadose zone
combined with carbon treatment.
- Abstraction of groundwater.
- Oil separation,
of groundwater.
stripping and biological degradation
Recirculation of treated water combined with additives
as nutrients, oxygen, aid chemicals etc.
- Optimization of the biological degradation of contami-
nants in soil and groundwater in-situ.
The concept is illustrated in Figure 4.
It should be mentioned that the costs of the full-scale
studies have been reduced significantly be.cause ,the
industry has a stationary installation with compressors,
extraction of air and activated carbon filters for normal
daily production purposes, if this equipment should have
been supplied to carry out the experiments it is a ques-
tion whether the studies economically could have been
carried out to the extent mentioned earlier.
213
-------
Copenhagen County has worked out the following acceptable
values for residual concentrations of organic solvents in
soil and groundwater:
Soil
Chlorinated solvents:
per constituent
totally
Benzene
Toluene
Xylenes
Acetone
Methylisobuthylketon (MIBK)
Methanol
Isopropanol
Total for solvents
Groundwater
Toluene
Xylenes
Benzene
Chlorinated solvents
Methanol
Acetone
MIBK
Isopropanol
5 ppm
10 -
1 -
5 -
5 -
10 -
10 -
10 -
10 -
50 -
0.3 ppm
0.3 -
0.1 -
0.3 -
50 ppm COD
There is a worldwide interest for the in-situ decontami-
nation solution. Yet, there is a lot of difficulties in
the optimization of the process because of lack of infor-
mation on what is going on 'down there'. The lack of
documentation on residual concentrations are one of the
main problems in accepting the process on a routine
basis. (It should be mentioned that air extraction is
documented to a great extent when dealing with organic
solvents). *
With respect to this concept, the documentation will be
developed in connection with contaminants in soil, water
and air.
The documentation will be developed as part of the fol-
lowing activities:
- Construction and further development of air strippers
applicable in soil
Statistical sampling of soil,
document the decontamination
water and air to
Development of special soil sampling equipment to
document the decontamination in-situ in the laboratory
To develop microbiological indicators/monitors for
optimization of the biological degradation in-situ
214
-------
To develop a 'model for acceptance' of the measured
values of chemical variables in combination with the
outlined accept criteria for soil and groundwater
To establish a connection between air emission compo-
nents and the corresponding constituents in soil and
water.
6.
Costs
The following costs are estimated for the different acti-
vities:
Stripping of soil and
extraction of air
Striping of groundwater
and activated carbon
treatment
Microbiological treat-
ment of water
Oil separation
(existing equipment)
Infiltration plant
Total
Investment
costs-
US$ 1,000
690
75
30
30
75
900
Operating
costs per year
US$ 1,000
150
15
15
15
195
These values give the following costs of the total acti-
vities:
Investment
Operation
US$ 25/m3
US$ 6/m3
215
-------
Figure 1:
Actual site with monitoring wells
^ B 5 - Monitoring well
B 32, B 34 - In-situ study are;
Drains
Figure 2:
Concentrations of VOC and VOX together with
toxicity measured as microtox EC20 and EC50
for water samples from 11 monitoring wells.
VOC:i10
BK\VOX:0.(KB
\EC20:120
ECSO:350
„ . voc: 4.8
• B I voy;
-------
Figure 3
Stripping
S: Stripping
E: Extraction
S: Control
Cloy
S
«-
5m
5m
Outline of in-situ stripping and extraction of air in soil and groundwater.
Figure 4
Seperation of oil, water treatment (stripping,
biological degradation)
Reeirculation
Compressor
Harbour
Steolpile wall
Stripping > aay Stripping
Illustration of the concept for remedial activities:
Stripping irv-trtu (depth 4 m)
Extraction of contaminated air and carbon treatment.
Attraction af groundwater and treatment.
Reeirculation of treated water.
Biological degradation of contaminants in toil
and groundwater.
217
-------
Table 1: Variation in concentrations horizontally and vertically.
Maximum and minimum concentrations are not corresponding to
the same monitoring well or horisontal distribution.
Sample/Concen-
tration
varia-
tion
Ground water
Soil
1.1.1-
Trichlor-
ethan
ppm
<0.0001-
3.6
0.001-
9.8
Trichlor-
ethylena
pp»
0.0001-
6.6
0.001-
68
•Tetrac-
chlor-
ethylene
ppn
<0.0001-
7.6
0.001-
330
Hotha-
nol
ppm
<0.2-
29000
<0.2-
1.0
Acetone
ppm
<0.2-
21000
<0.2-
34
Isopro-
panol
ppm
<0.2-
1600
<0.2-
110
Benzen
ppm
<0.001-
360
<0.01-
90
Toluon
ppm
0.002-
2400
<0.001-
3500
Xylener
ppm
0.02-
1300
<0.02-
2800
Total
Hydro-
carbons
ppm
0.34-
5800
<1-
34000
ro
i-»
oo
Table 2; Ghlorine-containing organic solvents in water
Hater sample
B 32 short
B 32 deep
B 34 short
B 34 deep
Chloroform
Before
stripping
PPB
0.0002
i.p.
i.p.
i.p.
After
stripping
ppo
i.p.
i.p.
i.p.
i.p.
1.1. 1-Trichlore than
Before
stripping
ppm
0.003
0.0003
0.021
0.035
After
stripping
ppn
i.p.
i.p.
i.p.
i.p.
Trichlorethylene
Before
stripping
ppm
0.013
0.0005
0.082
0.023
After
stripping
ppm
0.0001
0.0001
0.0064
0.0013
Tetrachlorethylene
Before
stripping
ppm
0.015
0.0003
0.016
0.019
After
, stripping
ppm
0.0002
0.0002
0.0029
0.0006
i.p. = not detected, detection limit 0.0001 mg/1 (ppm)
-------
Table 3; Chlorine-containing organic solvents in soil
Soil sample
''
Depth
m
B 32 0.5
B 32 1.5
B 32 2.5
B 32 4.0
B 34 0.5
B 34 1.5
B 34 2.5
B 34 4.0
Chloroform
Before
stripping
ppm
0.005
i.p.
i-J?.
i.p.
i.p.
i.p.
i.p.
i.p.
After
stripping
ppm
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p._
1.1.1-Trichlorethan
Before
stripping
ppm
i.p.
i.p.
0.036
i.p.
i.p.
0.087
0.034
0.004
After
stripping
ppm
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
Trichlorethylene
Before
stripping
ppm
0.002
i.p.
0.001
0.001
0.012
56
9.7
0.1
After
stripping
ppm
i.p.
i.p.
i.p.
i.p.
0.015
i.p.
i.p.
i.p.
Tetrachlorethylene
Before
stripping
ppn
0.016
0.004
0.011
0.002
0.030
230
28
0.29
After
stripping
ppm
0.003
i.p.
i.p.
i.p.
0.14
0.003
0.001
i.p.
ro
!-•
vo
i.p. = not detected, detection limit 0.001 mg/kg (ppm)
Table 4: Water-soluble organic solvents in water
Water sample
B 32 short
B 32 deep
B 34 short
B 34 deep
Methanol
Before
strip.
ppm
i.p.
i.p.
0.42
3.4
After
strip.
ppm
i.p.
i.p.
i.p.
i.p.
Ethanol
Before
strip.
ppm
i.p.
i.p.
0.59
8.8
After
strip.
ppm
0.7
1.2
i.p.
0.5
Acetone
Before
strip.
ppm
0.43
0.49
7.1
240
After
strip.
ppm
0.32
i.p.
90
703
Isopropanol
Before
strip.
ppm
0.39
i.p.
120
1300
After
strip.
ppm
i.p.
i.p.
i.p.
- 1630
N-Butanol
Before
strip.
ppm
i.p.
i.p.
1.1
18
After
strip.
ppm
i.p.
i.p.
i.p.
i.p.
i.p. = not detected, detection limit 0.1 ppm
-------
Table 5; Water-soluble organic solvents in soil
Soil SBBple
Depth
•
B 32 0.5
B 32 1.5
B 32 2.5
B 32 4.0
B 34 0.5
B 34 1.5
B 34 2.5
B 34 4.0
Hetl
Before
strip.
ppm
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
0.65
icnol
After
strip.
ppm
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
Etna
Before
strip.
ppn
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
1.4
nol
After
strip.
ppm
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
Acel
Before
strip.
ppm
i.p.
i.p.
i.p.
i.p.
i.p.
22
34
13
one
After
strip.
ppa
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
83
111
Isoprc
Before
strip.
ppn
i.p.
i.p.
i.p.
i.p.
i.p.
21
77
110
panol
After
strip.
ppa
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
14
78
H-Bul
Before
strip.
ppn
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
2.0
anol
After
strip.
ppa
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
i.p.
ro
ro
o
i.p. = not detected, detection limit 0.2 ppm
Table 6; Volatile aromatic compounds and oil products in water
Water .
sample
B 32 short
B 32 deep
B 34 short
B 34 deep
Ben;
Before
strip.
ppm
0.04
0.02
i.p.
0.06
sen
After
strip.
ppm
0.1
0.04
i.p.
0.02
Toll
Before
strip.
ppn
61
0.08
7
10
ien
After
strip.
ppm
16
0.52
0.06
5.7
Xyl<
Before
strip.
ppm
93
0.08
2.1
3.1
mer
After
strip.
ppm
10
0.9
1.7
0.03
Ethyll
Before
strip.
ppm
10
i.p.
0.89
1.2
jenzen
After
strip.
ppm
1.6
0.09
i.p.
0.7
Trimat
benzol
Before
strip.
ppm
4.3
0.01
i.p.
0.45
.hyl-
ler
After
strip.
ppm
0.13
i.p.
0.05
0.08
Tot
Hydroci
Before
strip.
ppm
360
0.34
76
51
:al
irbons* •
After
strip.
ppm
34
2.8
0.8
15
* Determined from a mixed hydrocarbon standard
i.p.: not detected, detection limit 0.001 mg/1 (ppm)
-------
Table 7: Volatile aromatic compounds and oil products in soil
Soil
sample
Depth
m
B 32 0.5
B 32 1.5
B 32 2.5
B 32 4.0
B 34 0.5
B 34 1.5
B 34 2.5
B 34 4.0
Benz
Before
strip.
ppm
0.05
0.01
0.19
0.02
<0.01»
0.92
i.p.
0.18
en
After
strip.
ppm
<0.1
<0.1
0.1
0.03
0.03
0.01
0.04
' <0.1
Tolu
Before
strip.
ppm
0.11
0.11
46
2.4
<0.01tt
360
72
6.2
en
After
strip.
ppm
0.01
1.4
11
0.01
i.p.
i.p.
0.3
0.01
Xyle
Before
strip.
ppm
0.21
34
75
4.9
<0.01»
65
12
11
ner
After
strip.
ppm
i.p.
2.5
20
0.02
i.p.
i.p.
0.03
0.05
Ethylb
Before
strip.
ppm
0.06
4.5
10
0.46
<0.01|
19
3.3
1
enzen
After
strip.
ppm
i.p.
0.2
2.6
i.p.
i.p.
i.p.
i.p.
0.01
Trimet
benzen
Before
strip.
ppm
1.3
19
0.55
i.p.
<0.01|
480
83
3.8
hyl-
er
After
strip.
ppm
i.p.
i.p.
0.7
i.p.
i.p.
i.p.
i.p.
i.p.
Tot
Hydroc
Before
strip.
ppm
3.8
780
350
19
-------
WEHRWISSENSCHAFTLICHE DIENSTSTELLE
DER BUNDESWEHR FOR ABC-SCHUTZ
INCINERATION PLANT FOR TOXIC WASTE
OF THE FEDERAL ARMED"FORCEs"DJFENCE SCIENCE AGENCY FOR NBC PROTECTION*
IN MUNSTER
HISTORY
In World War I a part of the present Training Area Munster-North was used
for production and field testing of chemical agents. The place at that
time was called "Gasplatz Breloh" with buildings used for the production
of chloropicrine and for,filling of chemical ammunition. After the war,
on October 24, 1919 by a powerful explosion all buildings were destroyed
and 1000 t of warfare 'agents, 1 million chemical shells and 40 tank
waggons with chemical warfare (CW) agents distributed on this area.
In World War II the expanded area Munster-North was used again for fill-
ing and field testing of CW munition. Additionally there was some produc-
tion activity, too, in a pilot plant for nerve agents (6A and GB).
After surrender and dismantling by the British occupation army almost all
infrastructural facilities were blown up again with dispersion of chemi-
cal agents in the surroundings. The red regions to be noticed on the map
of Munster-North can be derived from all the demolition measures.
Besides this, during World War II failure batches of the war gas produc-
tion, particularly mustard gas (HD), Which was adjusted intentionally to
a high viscosity and called "Zahlost", were buried in the ground. Unfor-
tunately no records about such actions were left after 1945.
After 1948 the area was started to be made free of danger by the Explo-
sive Ordnance Disposal (EOD) Command Lower-Saxony. This was carried out
for the most part only on the surface of the ground until in 1956 the
Training Area was taken over by the Federal Armed Forces. Now systemati-
cal scouring of the area began.
REMOVAL CONCEPT FOR FOUND AMMUNITION
The technical steps for an ultimate elimination of the dangers by chem-
ical duds and contaminations in the ground which are still being faced in
the Munster area are the following
1. Searching and discovering,,
2. Removal and transport,
3. Storage,
4. Demilitarization of the CW ammunition and
5. Destruction of the chemical agent contents.
* Wehrwissenschaftliche Dienststelle der Bundeswehr fQr ABC-Schutz
(WWDBw ABC-Schutz)
222
-------
The steps 1. to 4. are in the responsibility of the Munster Training Area
Command, which has available a demilitarization plant with a team of
EOD people. It is their task to separate fuzes and explosives from the
chemical ammunition. The chemical agents, waste produced during demilita-
rization and empty shells are filled into polyethylene barrels and taken
into storage up to incineration. The Federal Armed Forces Defence Science
Agency for NBC Protection is responsible for step 5: the environmentally
acceptable incineration of the chemical agents and final disposal of the
burned residues.
3. FUNCTION OF THE INCINERATION PLANT
(see functional diagram)
3.1 Incineration of chemical agents
The incineration plant of WWDBw ABC-Schutz, planned since 1975 and taken
in full operation since 1980 is being operated as a batch-type double
chamber furnace. This installation is unique in its function and was
primarily designed to manage the viscous mustard gas problem. It is the
only incinerator for thermal destruction of chemical agents in Western
Europe.
Before its incineration the initial material will be analyzed to deter-
mine problematic components, mainly arising from the presence of arseni-
cals. Based on the analytical results suitable batches of toxic waste
material are put together and provisions for a most effective waste water
and effluent air treatment are possible.
The opened polyethylene barrels are placed on one of the charging waggons
which are covered with high-temperature proof stones. At first the waggon
is pulled through a gas lock into the evaporation chamber. Here, at a
temperature of 300°C and in an inert gas atmosphere (N^+COj+HpO) the
chemical agent vapours are released during 10 to 12 hours and introduced
through an insulated duct into the main combustion chamber.
In this chamber equipped with highly fire-proof and chemical resistant
brick walls mustand gas (HD), for example, is oxidized at a temperature
of 1000° to 1200°C within a reaction time of 2 seconds to the components
sulfur dioxide (S02), hydrogen chloride (HC1), which still imply an
environmental burden, plus carbon dioxide and water.
The daily destruction rate during two overlapping working shifts, result-
ing in a total processing time of 12 hrs, amounts to 350 kg mustard gas
or about 70 tons a year.
Then the charging waggon with inevaporable organic chemical remnants and
metal parts is displaced into the burn-out chamber. Particularly metal
parts, e.g. shells are annealed in this chamber in air at 1000°C over 12
to 18 hrs. The effluent air from this chamber is conducted through the
main combustion chamber for final combustion of toxic components.
Charging area and displacement room are adjusted to an underpressure
(0.5 to 1.0 mbars) against atmosphere to prevent any egress of toxic
compound to the exterior.
223
-------
3.2 FLUE GAS SCRUBBING
Flue gas from chemical agent incineration.at first is cooled down to 80°C
by injection of water into the cooling (quench) tower. By passing two
washing towers (scrubbers) arranged one after the other the noxious gases
S02 and HC1 are eliminated from the flue gas with additional water. By
simultaneous injection of sodium hydroxide (NaOH) a fixed pH profile can
be achieved. After passage of a waste water neutralizer the highly toxic
blister agent HD has been transformed into such harmless salts like
sodium sulfate and sodium chloride which can be released unobjectionably
into the municipal waste water sewer.
The scrubbed flue gas leaves the incineration plant through a 30 m high
stack, not without further aerosol elimination in an aerosol separator.
Sampling probes and analyzers for continuous monitoring the emissions with
reference to the noxious components S02 and HC1 as well as hydrocarbons
(as a total) and dust (including arsenic trioxide) ,are attached to the
stack. The indicated values are transmitted to recorders installed in the
switchboard panel of the process operation control center.
The recorded values are far below the legally allowed limits of exhaust
air emission.
3.3 PRECIPITATION OF ARSENIC COMPOUNDS
Arsenicals found in the war gas mixtures are requiring additional opera-
tional steps:
All washing waters and collected aerosols with arsenical compounds in it
have to be jointly subjected to a subsequent oxidation and arsenic pre-
cipitation.
The mineralized arsenic (Ill)-compounds are oxidized to sodium arsenate
by reaction with potassium permanganate. Addition of ferric chloride
leads to an effective flocculation and dragging effect of the originating
• ferric hydroxide, accompanied with precipitation of ferric arsenate. This
compound is being dehydrated by a filtration step and can be deposited in
an old salt mine.
The filtration waters finally can be disposed of into the sewer.
OUTLOOK
More war gas remnants are still lying in the ground of the Training Area
Munster-North and on other sites of the Federal States of Western Germany.
Mostly these chemical agents and their decomposition products are mixed
with soil, sand or other solid materials. Thermal destruction of the
toxic ingredients of solid masses cannot be achieved economically in the
existing muffle-type furnace chambers because of the lack of poking
devices. Only very thin layers of earth can be handled to obtain a
complete burn-out.
Therefore it is planned to build an additional destruction facility as a
second incineration line on the basis of a rotary kiln.
224 '
-------
A rotary kiln has the capability to burn up almost all types of waste at
temperatures up to 1200°C. Thus an environmentally acceptable disposal of
a wide category of waste products endangering ground water resources and
the ambient air will be available.
225
-------
CHRRGING
MHGGON
CHRRGING HRGGON
WITH CH RGENT
ro
cr>
•"^A^"^
r=^.
fe=^r
<= =>
CHRRGING
RRER
F52^
fc^f
EVRPORRTION
CHflflBER
12 HRS RT 300»C
flIRLOCK
1
MRIN COMBUSTION
CHRMBER
2 SEC HT 1200*C
h
BURN-OUT
CHRMBER
DISPLRCEMENT 1B HRS flT '°°°"C
ROOM
COOLING
WRTER
SODIUM HYDROXIDE
EMISSION
CONTROL
aUENCH
TOMER
tSCRUBBERS! I
FERRIC=0>
CHLORIDE
URSTE MRTER
TRNKS
RRSENIC
PRECIPITRTION/
FILTRHTION
FERRIC RRSENRTE
(SPECIRL URSTE)
SEMRGE
(SRLT LORD)
oincRnn OF iNciNERnnoN PLRNT
-------
IN SITU GROUNDWATER REMEDIATION OF STRIPPABLE
CONTAMINANTS BY VACUUM VAPORIZER WELLS (UVB):
OPERATION OF THE WELL AND REPORT ABOUT CLEANED
INDUSTRIAL SITES
by
Dr.-Ing. B. Herrling*, Dipl.-Ing. J. Stamm*, Dr. EJ. Alesi", Dr. P. Brinnel*",
Dr. F. Hirschberger***, Dr. M.R. Sick"
Institute of Hydromechanics, University of Karlsruhe
Kaiserstrasse 12, D-7500 Karlsruhe 1, Germany
Tel.: (721) 608-3896 ; Fax: (721) 608-4290
GfS mbH
Dettinger Strasse 146, D-7312 Kirchheim/Teck, Germany
Tel.i (7021) 83335 ; Fax: (7021) 81794
HYDRODATA GmbH
Gattenhoferweg 29, D-6370 Oberursel, Germany
Tel.: (6171) 57068 ; Fax: (6171) 51406
f presented at the
Third Forum on Inovative Hazardous Waste Treatment Technologies :
Domestic and International
June 11-13, 1991, Dallas/Texas, U.S.A.
sponsored by U.S. EPA
227
-------
IN SITU GROUNDWATER REMEDIATION OF STRIPPABLE CONTAMINANTS BY
VACUUM VAPORIZER WELLS (UVB): OPERATION OF THE WELL AND REPORT
ABOUT CLEANED INDUSTRIAL SITES
B. Herrling*, J. Stamm*, EJ. Alesi", P. Brinnel*", F. Hirschberger***, M.R. Sick"
Institute of Hydromechanics, University of Karlsruhe, Kaiserstrasse 12, D-7500
Karlsruhe, Germany
GfS mbH, Dettinger Str. 146, D-7312 Kirchheim/Teck, Germany
"* Hydrodata GmbH, Gattenhofer Weg 29, D-6370 Oberursel, Germany
\
INTRODUCTION
The contamination of groundwater by strippable substances is a significant
problem in all industrial countries. For remediating aquifers in situ technologies are
favored to reduce the. investment and operating costs. The paper presents an in situ
method that can remove strippable substances, e.g. volatile chlorinated hydrocarbons, and
BTEX, from the subsurface (groundwater zone, capillary fringe, and unsaturated zone);
it is currently being used at numerous locations in Germany. This technology is an
alternative to conventional hydraulic remediation measures (pumping, off-site cleaning,
and reinfiltration of groundwater). The contaminated groundwater is stripped in situ by
air in a below atmospheric pressure field within a so-called "vacuum vaporizer well"
(German: Unterdruck-Verdampfer-Brunnen, UVB). The used air, charged with volatile
contaminants, is cleaned using activated carbon.
The UVB technique produces a vertical circulation flow in the area surrounding
the well, which catches the total aquifer. The vertical velocity component yields a desired
flow through the horizontal structure of a native aquifer. Numerical results demonstrate
the size of the sphere of influence and the capture zone of a well or well field; extended
field measurements have been and continue to be taken (Herrling et al. 199la).
The advantages of the UVB technique concerning the vertical circulation system
around the wells instigated thought about other applications, even without stripping the
groundwater. The realization of in situ biodegradation is such an example and seems to
be an appropriate alternative to other existing hydraulic systems. The different nutrients
and/or electron acceptors needed for biological activity can be added when the
groundwater passes the well casing (Herrling et al. 1991b).
This paper presents the UVB technique for in situ removal of strippable
contaminants. The circulation system, sphere of influence, and capture zone of a UVB
or UVB field as essential components of the hydraulic flow system are discussed in detail.
Further diagrams for dimensioning a UVB or UVB field are presented.
Two extended examples demonstrate the groundwater and soil remediation at
different sites located in the Rhine-Ruhr area and in Berlin using different installations
of the UVB system. The short remediation period and the low cost niveau for the UVB
investment and the well operation are only two among other advantages of the UVB
228
-------
technology which will be listed.
IN SITU REMEDIATION OF VOLATILE CONTAMINANTS BY THE UVB METHOD
The UVB helps to remove volatile substances from the groundwater, the
unsaturated zone, and the capillary fringe. When using the UVB method, a special well
with two screen sections is employed, one at the aquifer bottom and one at the
groundwater surface (Fig. 1) or below an aquitard in a confined aquifer. The borehole
reach between the two screen sections should be made impermeable. One well should
be used to remediate only one aquifer (phreatic or confined) and should not connect
different aquifers.
fresh air activated carbon
ventilator filter
stripping zone
working water level
additional pump to
support the air bunble
effect
separation plare
extensive groundwater
circulation
borehole filling :
*••/. filter gravel sealing material
'acuifer bottom
Figure 1: Vacuum vaporizer well (UVB) with additional pump and separating plate.
The upper, closed part of the well is maintained at below atmospheric pressure
by a ventilator. This lifts the water level within the well casing. The fresh air for the
upper part of the well casing is introduced through a fresh air pipe: the upper end is
open to the atmosphere, and the lower end terminates in a pinhole plate. The height of
the pinhole plate is adjusted such that the water pressure is lower there than the
atmospheric pressure. Therefore, the fresh air is drawn into the system. The reach
between the pinhole plate and the water surface in the well casing is the stripping zone,
in which an air bubble flow develops. The rising air bubbles produce a pump effect,
which moves the water up and causes a suction effect at the well bottom. In recent wells,
a separating plate and an additional pump (Fig. 1) are used to reinforce the pumping
effect of the air bubbles. Additionally, soil air is drawn from the surrounding
229
-------
contaminated unsaturated zone at many sites. Stripped air and possibly soil air are
transported through the ventilator and across activated carbon, onto which the
contamination is adsorbed. Thus, only clean air escapes into the atmosphere.
The cleaning effect of the well is based on reduced pressure, which reinforces the
escape of volatile contamination out of the water, and as a result of the air intermixing,
onto the considerable surface area of the air bubbles and onto the concentration
gradient. In this sense, the permanent vibration caused by the air bubbles is beneficial
to the escape process of the contamination. This vibration is transmitted as compression
and shear waves into sediment and fluid, and presumably influences the mobility of the
contaminants, even outside the well.
The upward-streaming, stripped groundwater leaves the well casing through the
upper screen section in the reach of the groundwater surface, which is lifted in a phreatic
aquifer by the previously explained pump processes and the below-atmospheric pressure.
It then returns in an extensive circulation to the well bottom. In this way, the
groundwater surrounding the well is also remediated. The expansion of groundwater
circulation is positively influenced by the anisotropy existing in each natural aquifer
possessing greater horizontal than vertical hydraulic conductivities. The artificial
groundwater circulation determines the sphere of influence of a well and is overlapped
with the natural groundwater flow (as described below).
The pinhole plate and all the installations within the well casing are designed as
a float so they can adjust automaticly to changing groundwater levels.
For special contaminants of lower density than water, a special installation within
the well is available: the contaminated water enters the well through the upper screen,
is stripped there, and with help of the additional pump, leaves the well through the lower
screen. Both installations can be used within the same well casing.
At many remediation sites, the UVB is used without an additional pump
and separating plate (see Fig. 2). In this case, a circulation flow occurs within the well
casing, which is produced by the strong pumping effect of the rising air bubbles. For the
most part, the stripped water follows the path of least resistance and flows down to the
end of the suction pipe. Thus, a water of uniform temperature and oxygen content
appears in the entire well casing . The water temperature is influenced by the withdrawn
evaporation heat in the stripping zone and by the temperature of the fresh air.
Depending on the groundwater temperature around the well, the water leaves the well
casing through the upper screen section and contaminated water enters the UVB at the
lower screen section. This occurs when the groundwater is colder than the circulation
water in the well casing. On the other hand, when the water in the well is colder than the
surrounding groundwater, an outer circulation occurs which is opposite to that shown in
Figure 2. The water leaves the well at the lower screen section and enters it at the upper.
Both cases, influenced by density differences of the involved water bodies, have been
observed at different sites.
SPHERE OF INFLUENCE AND CAPTURE ZONE OF A UVB OR UVB FIELD
The extended circulation field outside the well is of special interest. In this paper
numerical results of only UVB installations with additional pump and separation plate
will be discussed (Fig. 1). The effect of the above-mentioned permanent vibrations,
caused by the air bubbles, will not be considered. In principle, two different cases have
230
-------
fresh air activated carbon
ventilator
soil air removed
via suction
stripping zone
working water level
extensive groundwgter
/circulation
t*;'.'k filtergfavel sealing material
aquifer bottom
Figure 2. UVB method, driven by the air bubble effect.
been considered:
• When there is no (or negligible) natural groundwater flow, the sphere of influence
(or the range, R) of a UVB is of interest.
• When natural groundwater flow is significant, the extent of the capture zone has to
be determined for locating the well installations at a remediation site.
The resulting flow field of one or several UVB installations differs from the
natural groundwater flow field only in a limited area around the UVB. This is because
sinks and sources are located at the bottom and top of the same aquifer, each at places
with the same horizontal coordinates. The effected area can, therefore, be limited to the
sum of the areas of influence of all the UVBs. When only confined aquifer conditions are
considered to reduce the computational effort, the flow field of each UVB can be
superimposed onto those of other UVBs and of the natural groundwater flow field.
To estimate the sphere of influence and the capture zone of a UVB, numerical
investigations have been performed. To calculate the complex three-dimensional flow
field of a single UVB or a UVB field with minimal effort, the following simplifications
and assumptions have been used:
• The aquifer thickness is constant.
• Only confined aquifer conditions are considered in the calculation, even if the
natural aquifer is phreatic.
• The aquifer structure is assumed radially homogeneous to hydraulic conductivities.
Horizontal layers, each with different conductivities, can be used. The hydraulic
231
-------
conductivities may be anisotropic, but each horizontal layer may have only one
vertical and one horizontal conductivity.
• The local below-atmospheric pressure field near the wells is neglected.
• Density effects are neglected.
• The computations assume steady-state conditions.
• For estimating the capture zone, only convective transport is considered.
The three-dimensional flow field in the above-defined, limited aquifer region is
obtained by superimposition of a horizontal uniform flow field, computed in a vertical
cross section and representing the natural groundwater flow, and of radially symmetric,
vertical flow fields for each UVB. The superimposition of the different flow fields with
their own discretization is achieved by interpolating and adding the different flow vectors
at the various nodes of a simple rectangular grid with variable grid distances that are
•independently chosen for each Cartesian coordinate. The rectangular grid can be quickly
and simply set up and allows for some refinements near the wells and their screen
sections. More details of the numerical computations are given in Herrling and
Buermann (1990).
Resulting Flow System
Before going into more detail, the complex flow field near an individual UVB is
clarified for a vertical longitudinal section in the direction of the natural groundwater
flow (symmetry plane of the flow problem). In Figure 3, the streamlines of three case
studies are illustrated with Darcy velocities (v) of natural groundwater flow of 0.0 m/day,
0.3 m/day, and 1.0 m/day. All other parameters remain constant: the discharge (Q)
through the well casing is 20.16 m3/hr, the thickness (H) of the aquifer is 10 m, the
anisotropic hydraulic conductivities are KH = 0.001 m/sec (horizontal) and Kv = 0.0001
m/sec (vertical), and the lengths of the screen sections are aB = 1.2 m at the bottom and
aT = 2.1 m at the top. • • . ,
Figures 3b and 3c show that the groundwater, flowing from the left, dives
downward to the lower screen section and is transported upward within the well casing,
and that the cleaned water flows out to all sides at the upper screen section. The flow
situation can only be calculated and plotted in such a simple way in this longitudinal
section, otherwise the complex three-dimensional flow field has to be considered.
For a deep aquifer contaminated only in the upper groundwater zone, a UVB
installation can be used at a hydraulically imperfect well. The resulting flow system is
demonstrated in Figure 4, clarified for a vertical longitudinal section in the symmetry
plane (Fig. 4b). The used parameters are the same as for Figure 3b. The only difference
is that the aquifer thickness (H) is 30 m (well length = 10 m, as before).
At most of the UVB installation sites, a natural, nonnegligible groundwater flow
will exist. For a normal withdrawal well, a separating streamline can be determined: all
the water within this line is captured by the well, and all water outside of it passes the
well. In principle, the situation is the same when using a UVB. In contrast to a normal
withdrawal well, where the flow can be considered horizontal, the flow around a UVB
must be regarded as three-dimensional. Thus, the water body, flowing toward the UVB
from upstream and being captured by the lower screen section, cannot be delimited by
a simple separating streamline, but by a curved separating stream surface. This can be
calculated as described in Herrling and Buermann (1990): on the basis of the three-
dimensional flow field, a three-dimensional, particle-tracking method is used. The
232
-------
Figure 3. Streamlines clarified for a vertical longitudinal section with natural velocities:
(a) 0.0 m/day; (b) 0.3 m/day; (c) 1.0 m/day.
Figure 4. Streamlines at a hydraulically imperfect well clarified for a vertical longitudinal
section with natural velocities: (a) O.Om/day: (b) 0.3 m/day.
233
-------
water body within the separating stream surface is captured by the UVB, and that outside
of it, which flows from upstream, passes the well.
In Figure 5 the outer surface of the capture zone, calculated numerically, and the
surrounding horizontal aquifer bottom and aquifer top are plotted for two situations (the
natural groundwater flows from the background at the right side to the UVB, as shown
by the vectors). Figures 5a and 5b were calculated for the situation described for Figure
3b; the only difference is that for Figure 5a the vertical hydraulic conductivity is Kv =
0.001 m/sec, which means the calculation is performed for isotropic conditions. The
figures have a visible basis area of 50 m • 50 "m (Fig. 5a) and 100 m • 50 m (Fig. 5b).
(a)
UVB
K= K= 0.001 m/s
(b)
KH= 0.001 m/s
Kv= 0.0001 m/s
Figure 5. Separating stream surface of the capture zone for the situation of Figure 3b:
(a) KH = 0.001 m/sec (isotropic); (b) anisotropic KH/KV = 10.
The captured water is cleaned within the well and leaves it through the upper
screen section in all directions (not shown in Fig. 5). Parts of it are again captured by the
lower screen section, and the rest flows directly downstream.
If a wide plume of contaminated groundwater is to be cleaned, one UVB might
not be enough to capture the whole plume. Different UVB installations can be arranged,
for example, in one line normal to the natural flow. An important question concerns the
maximum distance that allows no contaminated water to flow between two neighbouring
wells without being cleaned. Figure 6 demonstrates such an example for the situation of
234
-------
Figure 5b where the maximum well distance is 46 m. The visible basis area of Figure 6
is to 150 m • 150 m.
UVB
KH= 0.001 m/s
Kv= 0.0001 m/s
Figure 6. Separating stream surface of the capture zone for the situation of Figure 5b,
but for two UVB installations at a maximum distance.
(a)
(b)
Figure 7. Separating stream surface of the different water bodies in the outside flow of
a UVB: captured, circulating and flowing downstream water in (a) a real
situation, and (b) water bodies separated for clarification.
235
-------
Figure 7 presents a view of the separating stream surfaces of all three water
bodies in connection with the flow around a UVB, The natural groundwater flow comes
from the left side. (In Figure 7b the three water bodies were artificially separated for
clarification.)
At the left side of Figure 7, the separating stream surface "of the contaminated
groundwater captured by the UVB can be seen. In the center a water body is shown
which consists of cleaned groundwater and shows the circulation flow around the UVB.
At the right side of Figure 7, the separating stream surface of the cleaned groundwater
flowing downstream is displayed. The calculation has accounted for the following
dimensionless parameters: Q/(HV) = 30, a/H = 0.25, and KH/KV = 5. The screen lengths
at the bottom and top are the same: aT = aB = a.
Diagrams for the Dimensioning of UVB Installations
Absence of Natural Groundwater Flow. At sites without natural groundwater flow, the
sphere of influence (R) of a UVB is of special interest. R is dependent on the anisotropy
(horizontal over vertical hydraulic conductivity: KH/KV), on the thickness (H) of the
aquifer, and on the length of the screen sections aT and aB at the top and bottom of the
aquifer (see Fig. 8) or the ratio a/H (when the same length of the screen section is used
for both, then only a is referred to). Although R is mathematically infinite, it is, in
practice, defined as the horizontal distance from the well axis to the farthest point at
which circulation flow is still significant. In a dimensionless description, R has been made
dependent on the ratio QR/Q, where QR is that water quantity, which circulates within the
distance R from the well. The ratio QR/Q, which is prescribed for practical reasons,
describes the strength of a circulation flow at the distance R from the well.
In Figure 9a, results are presented for ratios QR/Q = 0.98 and 0.8 and for a = aT
= aB in a dimensionless diagram. The sphere of influence (R) is independent of the dis-
charge through the well, but strongly dependent on the anisotropy KH/KV. Within usual
proportions, the length of the screen sections has only a small influence. For a UVB with
separating plate and additional pump, a totally screened well casing should be avoided
because of hydraulic shortcircuiting.
Figure 9b presents a dimensionless diagram that describes the differences (Ah) of
the hydraulic heads between the top and bottom of a double-screened well through which
Ah=hT-hB
Qr =Q -2TrrJTvrdz
:) z = H/2
K
H
K
H
z-h
°T
! -*v
HO
! *+^
i *
-R-
UVB
Figure 8. Notation in a vertical cross section.
236
-------
(a)
(b)
01 234 56789 10
Figure 9. (a) Sphere of influence (R) for a site without natural groundwater flow, (b)
differences (Ah) of the hydraulic heads between the top and bottom of a
well.
a discharge (Q) is pumped. Ah is dependent on the parameter Q/(H2KH) and the ratios
KH/KV and a/H. Abiding by the above-described assumptions, the rise of the hydraulic
head at the top of the well amounts to Ah/2, and the decrease is -Ah/2 at the bottom
(both referring to the position of rest). When using the UVB for stripping, the falling,
stripped water in the reactor causes a dynamic effect that will influence the upper
hydraulic head within the well.
, For the dimensioning or examination of a site, Figure 9b is a valuable expedient.
When KH is known (e.g., by pump test) - along with H, Q, and a - Figure 9b and the
measured Ah allow an estimate of the anisotropy at a site.
Presence of Natural Groundwater Flow. At most remediation sites a natural groundwater
flow exists. Figure 11 shows numerical results represented in dimensionless form for the
dimensioning of UVB installations under these conditions. Figure 10 introduces the
notations for an upstream cross section through the capture zone normal to the natural
groundwater flow direction (comparable with the open influx region to the left of the
capture zone in Figure 7) for one and two UVB installations. It is often the case when
remediating a wide contamination plume, that several wells are used in a line normal to
the direction of the natural groundwater flow. The length (D) denotes the maximum well
distance at which the contaminated groundwater cannot pass between the wells without
being cleaned. The results of Figure 11 have been calculated for an upstream distance
237
-------
of 5H from the well and for a constant ratio of a/H = 0.25 (screen length over aquifer
thickness). The results are discussed for wells which pump upward.
-Be
UVB
UVB
UVB
Figure 10. Notations in an upstream cross section through the capture zone for one
and two UVB installations (for wells pumping upward).
The widths By and BB of the upstream capture zone, measured at the aquifer top
and bottom, are shown in Figure lla. The ratios Bj/H and Bg/H are dependent on the
ratios Q^EPv), KH/KV, and a/H. v denotes the Darcy velocity of the natural groundwater
flow; all other variables are explained above. For small values of Q^H^v), the upper part
of the capture zone does not reach the top of the aquifer. This implies that for
remediating a plume, a minimum well discharge (Q) is required. Again, the results are
quite sensitive to the degree of the anisotropy (see Fig. 5, as well).
Figure lib shows the results for the influx area (A) of the upstream capture zone,
and Figure lie the maximum well distance (D) of two wells between which contaminated
groundwater cannot pass without being treated. The ratios A/H2 as well as D/H are
dependent on the same parameters as the widths Bj- and BB. When a plume of width W
is to be cleaned, the number (n) of UVB installations can be estimated by n =
(W-Br)/D+l.
When a plume is remediated, the contaminated water of quantity Q0, flowing into
the capture zone of a UVB from upstream, is diluted with water that has already flowed
through the well and circulates around the UVB. Thus, the contaminant concentration
of the water within the well casing will be lower than in the upstream plume; near a
contamination source the situation is contrary. Figure lid illustrates the portion Q0 of the
total well discharge Q. The ratio QJQ is again dependent on the same parameters as the
widths of the upstream capture zone. Figure lid can be used to estimate the expected
concentration value of the water within the well casing for the dimensioning of a UVB
installation. It may help to evaluate the progress of remediation at a site when
concentration data of the upstream plume and the water within the well are determined.
In Figure 12 the upstream distance (S) of the stagnation point at the aquifer top
from the well axis is described (see Fig. 3b and 3c, as well). The ratio S/H is also
dependent on the parameters Q^H^), KH/KV, and a/H. The location of the stagnation
238
-------
(a)
(b)
5 10 15 20 25 30 35 40 45 50 Q
1C 15 20 25 30 35 40 IS 50 Q
(c)
(d)
0 5 10 15 20 25 30 35 40 45
Q 5 10 15 20 25 30 35 40 45 50 Q
H2v
Figure 11. (a) Widths B-j- and BB of the upstream capture zone at the aquifer top and
bottom; (b) Influx area A of the upstream capture zone; (c) Maximum well
distance (D) at which the contaminated groundwater cannot pass between the
wells without being treated; (d) Upstream discharge (Q0) in the capture zone,
which is diluted with the circulating water to the total well discharge (Q).
239
-------
point is highly sensitive to the anisotropy of the aquifer. The length of the screen section
is of small importance within usual proportions (as described above). The knowledge of
the distance (S) from the stagnation point can be used to determine the positions of
measuring equipment. The operation of a UVB can also be supervised using depth-
dependent measurements between the stagnation point and the well.
The sphere of influence of the circulation around a UVB at sites with natural
groundwater flow is of special interest. This sphere of circulation is limited in a quite
different way than at a site with absence of natural flow (Fig. 9a) as can be seen in
Figures 3b, 3c, 4b, and 7. In the direction of natural groundwater flow, this sphere has
a maximum expansion of (S) (see Fig. 12) to the upstream and downstream sides.
Normal to this direction, the maximum radius of the sphere of circulation is
approximated by (Be-HB^M (Fig. lla), and, in the case of several wells in one line, by
D/2 (Fig.llc).
Figures 9, 11, and 12 can be used for the dimensioning of a UVB or UVB field
when the parameters KH/KV and Q/(HV) can be estimated, where Q depends on the well
size and on the additional pump. For an irregular well field, a layered aquifer, or special
critical cases, numerical calculations can be performed.
_§_
H
Q3 \ Kv
ai i KH_C
Q3 J Kv
Q1
0.3
0 5 10 15 20 25 30 35 iO 45 50 Q
Figure 12. Upstream distance (S) of the stagnation point from the well axis.
240
-------
\JVB-TECHNIQUE FOR THE CREATION OF A CIRCULATION FLOW IN AN
AQUIFER FOR THE REMOVAL OF CHLORINATED HYDROCARBONS AT A SITE
IN THE RHINE-RUHR AREA
(Contribution by E.J. Alesi and M.R. Sick)
Original Situation
A contamination of the groundwater with volatile chlorinated hydrocarbons (CHC)
was detected on the premises of a former steel processing plant. The primary component
of the contamination was identified as trichloroethene (TCE).
This underground contamination was centered at the location of a former
processing area where trichloroethene was utilized in a degreasing procedure. The first
investigative measures were conducted in the middle of the 1980's and revealed a
maximum trichloroethene contamination of 5 mg/1 in the groundwater. Moving outward
from the center of the contamination, a pollutant-front possessing a CHC concentration
of 2-3 mg/1 extended over a distance of 250 m up to the former border of the plant site.
The pollutants could be detected up to a distance of 2 km away in the direction of the
groundwater flow.
An investigative survey of the entire plant site was not carried out as the exact
location of entry of the pollutants into the ground was known. In 1987, after a new
structure was built over the contaminated area, a groundwater remediation featuring a
removal rate of 80 m3/h was planned. Due to the high costs and the problems which
arose from the attempt to pump the groundwater up to the surface, this remediation
method was later abandoned.
The firm GfS, a licensed supplier of the UVB remediation process, received a
contract for the implementation of a UVB test run using an existing well on the site.
Due to the high rate of contaminant removal achieved, the initial test set-up
remained in operation for a period of two years. During this time period, the ongoing site
investigation revealed that the CHC contamination existed also in the deeper zones of
the aquifer. Due to this new information, the client decided to drill a new remediation
well which would encompass the entire aquifer.
By positioning the second remediation well in the proximate vicinity of the original
well, a larger predictable radius of effectivity could be obtained as well as a shortening
of the time required for the remediation. Installed in this new well was a UVB unit
equipped with a separating plate and additional pump, which amplifies the circulatory
flow.
An automatically regenerating activated carbon filter unit was installed for the
treatment of the exhaust air which was heavily loaded with CHC.
Geological and Hydrogeological Data
The analysis of numerous boreholes gave rise to a uniform geological portrait.
Situated under an up to 4 m thick artificial filling are naturally lying silts possessing
different proportions of fine- to medium-grained sands.
Further below follows a 30-35 m thick irregularly alternating bedding of Rhine
sediments which are also composed of fine- to medium-grained sands, and in addition
contain medium and coarse gravel. Occasionally encountered at various depths are single,
241
-------
destination
100m
GfS
Gesallschaft fur
Boden- und Grund-
wassersanierung mbH
Dettinger Straflo 146
7312Klrchheim/Teck
Telefon07021/83335
Telefax 070 21/8 17 94
Figure 13. Former production area of a steel processing plant.
thin silt lentils. The base of the aquifer consists of very tightly layered, silty fine sands of
Tertiary age which possess a low water permeability. The aquifer possesses a thickness
of approximately 30-35 m and lies 7 m below ground level.
From the geological profile surveys and pump tests carried out on the site,
hydraulic conductivity coefficents from 1.0 • 10'3 to 5.0 • 10"4 m/sec were obtained. From
these a groundwater transport velocity range of 0.05-0.2 m/day was calculated.
242
-------
Bore Profile and Construction Plan of the UVB I
The well used for the UVB test run possessed a fine- to medium- grained sand
profile down to a depth of 12 m and was filtered from a depth of 4 to 12 m with a 400
mm slit-screen filter (Figure 14). A UVB unit designed to function on the airlift principle
was built into the well. At a groundwater level of approximately 6 m, soil air was ad-
ditionally vented through the free filter area above the groundwater table.
Bore Profile and Construction Plan of the UVB II
Due to the fact that the center of the CHC contamination was located under the
former degreasing area, the new remediation well was positioned downstream from this
area. The bore hole for the remediation well was sunk down to the bottom of the aquifer
at a depth of 39,70 m.
The bore profile and the corresponding well construction plan are displayed in
Figure 15. Three filter segments, positioned in the upper, middle, and lower portion of
the aquifer, were* built into the remediation well. Each segment had a length of 5 m. The
positioning of these filter segments allows the circulation flow to be positioned for
functioning either solely within the upper portion of the aquifer or over the entire aquifer
through the corresponding positioning of the packers.
The upper 2 m of the upper filter segment was fitted with special IEG double-
cased screen filters.
Well Design
Remediation Well According to the Airlift Pump Principle (UVB I). The discharge of the
groundwater into the well shaft could not be directly determined due to the set-up of the
test unit. The exact distribution of the contaminant load (ratio of the contaminant
removal from the saturated vs. unsaturated zones) could not be calculated because the
exhaust air flow also contained contaminated soil air from the capillary zone, which was
additionally pulled into the system. For an average CHC total removal of 1.6 kg/day, a
rough calculation of the portion originating exclusively from the groundwater was
estimated at 0.6 kg/day. Using this value and the average groundwater CHC
concentration of 5 mg/1, the circulated amount of water can be roughly estimated at 5
m3/h.
With the UVB, an additional contaminant removal is achieved by soil air venting
through the highly contaminated capillary area.
Remediation Well with Separating Plate and Support Pump (UVB II). In contrast to the
UVB I (airlift pump principle; see Figure 2) of the test run phase, the new UVB II was
equipped with an underwater pump for the amplification of the circulatory flow (see
Figure 1).
When the UVB II was put into operation, the route of the circular flow was
established so that through the closing of packer 2 (see construction drawing in figure 15)
the incoming flow would be directed over the middle filter segment. The reason for
establishing this flow pattern was that the CHC concentration in the groundwater of the
243
-------
5!
Bore Profile UVB I
Well Construction UVB I
ro
w
o
1
CO
3
8
3
on
a
o
3
3
O
0,00
3,50.
ground level
artificial fill
5,10
9,80
10,40 _
12,50
c\ uO • Ci &
? 0 ? C
0 „ O „
?°o ?°c
O . 0 „
°o? °oc.
V'o°?°c
0 „ 0 .
^"o'^'c
o „ o .
o 0 o s
'? " o" 'c
O a O .
"o . 'o1;
"0 , "0 ,
fine - medium sand
,
sandy gravel
fine - medium sand
sandy gravel
L J -
4,00
groundwater level
6,00
12,50
600
Filling
Filter Gravel
Slot Screen Filter, 400 mm
-------
r\i
4=>
on
t
•-t
0)
W
o
o
S
3
a
8
3
en
o'
3
3
O
3"
CD
Bore Profile UVB II
0,0
3,2
4,3
5,6
9,3
17,4
19,8
23,3
28,5
31,2
35.1
39,5
40.0
^
^/
'////
o.00-o.oo
0.0.
o.oo-o.00
O . 0 .
?-°oT°o
0.0.
'o . 'o .
'o . 'o .
' • ' «
. • '• ' . • '•
'o . 'o .
o-o- o-o,
'o . 'o .
••'..'•".
•o" 'oV
'o . 'o .
artmclal fill
fine - medium sand
sandy gravel
One • medium sand
sandy gravel
fine - medium sand
sandy gravel
fine - medium sand
sandy gravel
fine - medium sand
sandy gravel
>-i^ sllty sand
Well Construction UVB II
Ground Level
ca. 7m
V
. 6,0 m
Sealing
IEG Double-Cased
Screen Fitter
Filter Gravel
Solid Pipe (Steel) 400mm
Packer I
Slot Screen Filter, 400mm
,.-JU Packer II
Optional Positioning
Aquifer Bottom 39,70 m
W//M
-------
upper section in the aquifer was higher than that of the lower section and via this route
of circular flow, a cleansing within this area should occur.
After 4 months of operation the circular route was changed so that the middle
filter segment would be sealed off by the closing of packers 1 and 2.
Effectiviiy Radius of the UVB I Test Well.
A numerical determination of the effectivity radius of the circulation flow was not
attempted for this situation as theoretical design diagrams for imperfect wells do not yet
exist and because the problems associated with the density flowes arising from UVB's
operating without separating plate and additional pump have not yet been scientifically
clarified.
From the distinct concentration changes detected in the surrounding groundwater
monitoring wells, namely a decrease in the contaminants and an increase in the oxygen
content (oxygen enrichment), it is clearly evident that with this test design a far-reaching
range of effectivity was achieved.
Computation of the Circulation Flow for the Remediation Well with Separating Plate
and Additional Pump (UVB II)
On the basis of the following cited geological and hydrogeological parameters,
calculations were performed according to Herrling et al. 1991b for the determination of
the upstream capture zone of the UVB, of the sphere of circulation, and, corresponding
to the upstream capture zone, of the zone of cleaned water flowing downstream (see
Figure 7). Hydraulic parameters of the contamination area and well specifications are:
I
n
Kv
H
KH/KV
Q
aT
0.00035
0.15
0.198
0.03
1 x lO'3
IxlO"4
33.0
10
20
5.0
5.0
[m/day]
[m/day]
[m/sec]
[m/sec]
[m]
K/hr]
[m]
[m]
(hydraulic gradient)
(pore volume)
(transport velocity)
(Darcy velocity)
(horizontal hydraulic conductivity)
(vertical hydraulic conductivity)
(thickness of the saturated aquifer)
(discharge through the UVB)
(length of the upper screen section)
(length of the lower screen section)
Using the corresponding dimensioning diagrams (Figures 11 and 12), the following
values are obtained for the upstream capture zone:
&r
B
56
214
5445
7
69
[m]
[m]
K]
[mVhr]
[m]
(width at the top of the aquifer)
(width at the bottom of the aquifer)
(cross section area)
(upstream discharge in the capture zone)
(distance of the stagnation point from the well axis)
A graphic presentation of the above values is given hi Figures 8 and 10.
246
-------
Direction of Groundwater Flow
LB32
KLP10M
100m
Calculated according to
HERRL1NG, STAMM & BURMANN (1991):
I"[ 1 | Capture Zone of UVB
Release Zone of UVB
Area of Circular Flow
GfS
Gesellschaft fur
Boden- und Grund-
wassersanlerung mbH
D«ttlngerStrafie146
7312Wrchheim/Teck
TaIefon07021/33335
Telefax 070 21/817 94
Figure 16. Flow conditions for the UVB II.
247
-------
The flow conditions determined for this remediation case from the above values are
graphically displayed in Figure 16. The corresponding flow zones are represented as
surface areas. However, their actual (spatial) proportions are significantly more complex.
From this graphic representation it is evident which groundwater monitoring wells lie in
the circulation flow, downstream and upstream of the UVB well.
Due to different anisotropic conditions in the underground (i.e. local silt lentils or
areas possessing higher hydraulic conductivity), the actual area of influence deviates from
the calculated effectivity radius of the UVB generated flow.
Previous results have shown that the actual area of influence is normally greater
than the calculated effectivity radius. This is shown here, for example, by the monitoring
well TB191 which according to the calculations lies outside of the area of intensive
circulation flow, but nevertheless experiences a pronounced decline in contaminant
concentration. This clearly shows that this point was also influenced by the UVB.
Groundwater Contours and the Movement of the Groundwater Surface
Based on the changes of the groundwater levels caused by the hydraulic effect of
the UVB II as well as from the displayed groundwater contour diagrams, the area of
influence of the UVB circulation flow can be estimated and compared with that obtained
from the theoretical calculations.
The situation under naturally occurring conditions, without the effects of the
remediation equipment, is shown in Figure 17. The groundwater contours depict a small
groundwater gradient running from northeast to southwest.
During UVB remediation, an elevation of the groundwater surface in the area
surrounding the UVB and a sinking of the groundwater surface in the area surrounding
the monitoring wells RSB211 and T-P31 occurs (Figure 18).
The filtering segment length of the individual groundwater monitoring wells had a
significant influence on the deviations of the water levels within the well. According to
the positioning of the filtering segment within the circulation flow, distinct patterns of
change emerged which ultimately also provide evidence of the effectivity radius.
Contamination Concentration and Removal in the Exhaust Air of UVB I
s
During the 2 year operational period of the UVB I, a clear trend with regard to the
contaminant concentration in the exhaust was not recognizable.
At volumetric flow rates between 300 and 500 m3/hr the daily contamination
removal revolved around an average value of 2.0 kg/day. The contaminant concentrations
in the exhaust air, in which trichloroethene was always the predominant single
component, fluctuated between 120,000 and 300,000 Mg/m3. During the course of the
remediation, the content of cis-1.2-dichloroethene rose from an initial value of 150 /zg/m3
to 9,000 /ig/m3.
Over an operational period of approximately 15,000 hours, a combined total of
1,500 kg of CHC was removed from the saturated and unsaturated soil zones.
248
-------
350 392 433 475 517 558 600 642 683 725 767 898 850
400
350 392 433 475 517 558 600 642 683 725 767 808 850
400
Figure 17. Groundwater contours without the effects of the UVB
Contaminant Concentration and Removal in the Exhaust Air of the UVB II with
Separating Plate and Additional Pump
After the UVB II was put into operation, the CHC concentrations in the exhaust
air totaled 30,000 /ig/m3. Using the concentration and volumetric flow values of the
exhaust air, CHC removal rates from initially 375 g/day to finally 20 g/day couid be
calculated. This strong decline in the removal rate correlates with the decline of the total
CHC concentration, which at this time point was approximately 100 ;ug/l in the incoming
flow of the lower filter segment.
This reduction in the CHC removal from the groundwater, in contrast to the first
remediation phase (UVB I), is explained by the intensive flushing of the aquifer with
about 13 m3/hr of "clean" circulation water and the simultaneous inflow of 7 m'/hr of
contaminated groundwater from upstream. Of note here is that the entrance of
contaminated soil air into the system is impossible due to the position of the groundwater
level and the filter segment with respect to one another.
Following an operational period of 4,000 hours for the UVB II unit, an additional
50 kg of CHC was removed from the aquifer.
249
-------
392 433 r 475 517
600 642 683 725 767 808
400
350 392 433 475 517 558 600 642 683 725 767 808
400
850
Figure 18. Groundwater contours during UVB remediation.
Concentration Changes in the UVB Monitoring Tubes
The time-zero sampling (the sampling done before the UVB was put into
operation) in the deep-lying monitoring tubes of the UVB II yielded CHC concentrations
of 1,260 jig/1 (11 m depth), 1,218 nf\ (24 m depth), and 1,635 ng/l (38 m depth).
Within one month after the UVB was put into operation, the CHC concentration
in the incoming flow of the middle filter segment declined to a value of 100 /ig/1.
After the CHC concentrations in the remediation well had significantly declined, the
circular flow rate of the circulation influx was shifted onto the lower and upper filter
sections of the UVB well. In the neighboring monitoring wells and in the monitoring
tubes of the UVB II no further significant differences with regard to the vertical
contaminant distribution in the aquifer could be detected.
Through these measures, the CHC concentration in the incoming flow of the lowest
segment of the UVB II also declined during this time interval to a value of 100 p.g/1.
250
-------
Concentration Changes in the Individual Groundwater Monitoring Wells
Groundwater Monitoring Wells Upstream of the UVB. Monitoring well KB221, which lies
in the upstream area of the UVB II, showed a strong deviation in CHC concentration
(460 - 5,075 /ig/1) during the remediation period. In striking comparison to the other
monitoring wells, KB221 possessed relatively high concentrations of tetrachloroethene
(630 jig/1) and 1,1,1-trichloroethene (119 /xg/1).
Since the beginning of the remediation, the CHC concentrations in the upstream
flow displayed a clear increase, which was also evident from the rise of the regression
curves (Figure 19). If one considers, however, the comparatively short time period since
the new remediation well was put into operation, a reduction in the concentrations also
in these upstream wells can be ascertained.
Total CHC Content
6000
5000
4000
3000
2000
1000
0 100 200 300 400 500 600 700 800 900 1000
Remediation Time (Days) - Remediation Starting Date: 24.8.88
8 CHC Concentrations (Sampling Depth = 10m)
—e— Regression Curve
Figure 19. Concentration changes in groundwater monitoring well KB221.
1100
In any case, due to the composition of the contaminant spectrum and the
significantly higher CHC concentrations, as in monitoring well TB191 (Figure 20), the
presence of an additional, upstream situated source of contamination must be assumed.
A consequence of this is that the current CHC concentrations in the area of the
circulation can not be further significantly reduced due to the high concentrations in the
upstream flow.
Already before the beginning of the remediation, the monitoring well TB191
displayed a strong deviation of up to 4,000 jug/1 in the analysis values. After the UVB I
was put into operation (airlift pump principle), a distinct reduction of the CHC
concentrations was observed, which continued down to a concentration of 300 /ug/1 after
the UVB II was put into operation.
251 /• .
-------
1500
Total CHC Content
1000
•a
500
s
Starting Date
UVB 2 (14.11.80)
100 200 300 400 500 600 700 800 900 1000 1100
Remediation Time (Days) - Remediation Starting Date: 24.8.88
s CHC Concentrations (Sampling Depth = 10m)
—o— Regression Curve -
Figure 20. Concentration changes in groundwater monitoring well TB119.
Groundwater Monitoring Wells downstream of the UVB. The downstream monitoring
wells T-P11 and T-P21 were regularly sampled at three different depths. One striking
result is the low CHC concentration in the 5-10 m depth of the aquifer. The initially high
CHC concentrations at a depth of 25 m (2,000 /xg/l in T-P11 and 5,000 /ig/1 in T-P21) fell
(also clearly recognizable in the regression curves) to values between 800 and 400 /ig/1
(Figures 21 and 22). Monitoring well T-P21, which lies on the border of the UVB
downstream zone, displays a rather low reduction in contaminant concentration compared
to monitoring well T-P11, which lies in the downstream center.
Groundwater Monitoring Wells in the Circulation Flow. Monitoring well LDB201, which
lies 10 m from UVB I and 40 m from UVB II, possessed since the beginning of the
measurements in 1986 a CHC concentration of 2,000 /ig/1. At the beginning of the test
run in the fall of 1988 this concentration reached an equilibrium value of 1,500 /ng/1. After
the UVB II was put into operation in November 1990, the concentration declined
continuously down to a value of 170 jug/1 (Figure 23).
The UVB test well, which since the end of the first remediation phase was further
utilized as a monitoring well, possessed CHC concentrations up to 5,000 fj.g/1 during 1987.
Within this time period, the trichloroethene concentration declined to 20 jig/1 whereby
the concentration of the breakdown product cis-l,2-dichloroethene increased up to 570
Mg/1.
252
-------
CD
2,
2500
2000 -
1500 -
1000 -
500 -
Total CMC Concentration
"300 400 500 600 700 800 900 1000 1100
Remediation Time (Days) - Remediation Starting Date: 24.8.88
CHC Concentrations at 25m Depth
CHC Concentrations at 10m Depth
Regression Curve for 25m Depth
Regression Curve for 10m Depth
Figure 21. Concentration changes in groundwater monitoring well T-P11
Total CHC Content
6000
300 400 500 600 700 800 900 1000
Remediation Time (Days) - Remediation Starting Date: 24.8.88
H CHC Concentrations at 24m Depth -&- Regression Curve for 24m Depth
A CHC Concentrations at 10m Depth -x- Regression Curve for 10m Depth
Figure 22. Concentration changes in groundwater monitoring well T-P21.
253
1100
-------
o>
5000
4000
3000
2000
1000
Total CHC Content
Value not Included in Regression Analysis
Starting Oats
UVB 2 (14.1140)
300 400 500 600 700 800 900 1000 1100
Remediation Time (Days) - Remediation Staring Date: 24.8.88
a CHC Concentations (Sampling Depth = 10m)
-e- Regression Curves (Dual)
Figure 23. Concentration changes in groundwater monitoring well LDB201
Monitoring well RSB211, which lies 40 m from UVB I and 20 m from UVB II and
in the direct downstream flow from of the contaminant center, displayed a very different
change in CHC concentration. The course of change for this monitoring well can be
divided into three stages (Figure 24). Before the beginning of the remediation, the
concentration fluctuated around a value of 1,000 jug/l. After UVB I was put into
operation the concentration rose suddenly to values of over 7,000 ng/l.
This rise can be explained by the mobilization of CHC out of the capilllary zone
caused by the circulation flux and by the compression waves which propagate within the
aquifer. Pressure impulses, produced by the bursting of air bubbles on the body of the
nozzle, lead to a better solvation of the CHC present in the pore spaces and in the
adhesive (retained) water. This effect results in a faster cleaning of the contaminated
area.
Since the UVB II was put into operation, the preliminary trend of a distinct
reduction of the CHC concentration in monitoring well RSB211, which began in the
middle of 1990, has continued (Figure 25). Presently, the concentration is revolving
around a value of 350 /xg/1. An intensive flow is indicated by the cis-l,2-dichloroethene
concentrations which rise during the remediation period up to a value of 440 /Lig/1 after
initially lying below the detection limit.
Monitoring well T-P31, which lies 10 m from UVB II and reaches the bottom of the
aquifer, displayed a reduction in CHC concentration down to 150 /ig/1 after an initial
concentration of 2,000 /ig/1 (Figure 26).
254
-------
Total CMC Content 1986-91
8000
7000 -
6000 -
5000
O) 4000
3000
2000
1000
0
14.11.84
29.03.86
11.08.87
23.12.88
07.05.90
19.09.91
31.01.93
Figure 24. Concentration changes in groundwater monitoring well RSB211 since 1986
Total CMC Content
100 200 300 400 500 600 700 800 900 1000 1100
Remediation Time (Days) - Remediation Starting Date: 24^8.88
B CHC Concentrations (Sampling Depth = 10m)
—€>— Regression Curves (Dual)
Figure 25. Concentration changes in groundwater monitoring well RSB211
255
-------
Total CMC Content
3500
3000
2500
2000
1500
1000
500
650 700 750 800 850 900 950
Remediation Time (Days) - Remediation Starting Date: 24.8.88
CMC Concentration at 30m Depth
CMC Concentration at 20m Depth
• CHC Concentration at 10m Depth
1000
Figure 26. Concentration changes in grounwater monitoring well T-P31
Oxygen Content in the Groundwater
The oxygen content in the groundwater can be utilized for the evaluation of the
effectivity radius of the UVB as long as no strong biotic consumption of oxygen is
present. During the course of the remediation, a clear correlation between rising oxygen
concentration and falling CHC concentration was evident.
The monitoring wells' RSB211 and T-P31 were used as a representation of the
enrichment of oxygen within the aquifer. During the test run remediation phase of
UVB I, the oxygen concentration in RSB211 remained at a very low level (0.1 to 0.5
mg/1). After UVB II was put into operation, the oxygen concentrations in the upper 3
meters of the aquifer rose up to values of 2.0 mg/1 (Figure 27). UVB II produced a
similar trend of oxygen enrichment in T-P31. Here, the different sampling depths all
displayed the same change in oxygen concentration. The strong rise in values up to 9 mg/1
can clearly be attributed to the operation of ,UVB II.
256
-------
Oxygen Content
3.5
3.0
2.5
=• 2'°
E
'-' 1.5
1.0
.0.5
StartogDato
UVB Z (14.11.90)
Starting Date
UVB 1 (24.8.88)
0
06.06.88
23.12.88 11.07.89 27.01.90
Measuring Depth 10m
15.08.90 03.03.91
Figure 27. Oxygen content in the groundwater monitoring well RSB211
Oxygen Content
O)
5 -
o
26.06.90
15.08.90
04.10.90 23.11.90 12.01.91 03.03.91
Depth 8m -e- Depth 14m
22.04.91
Figure 28. Oxygen content in the groundwater monitoring well T-P31
257
-------
Remediation Costs
The following itemized costs are given in US-Dollars, however they correspond to
the German market prices (exchange factor: 1 US$ = 1.70 Marks). Due to the different
price structure found in the USA as compared to Germany (i.e. patent costs, insurance
regulations, etc.) they can not be directly transferred, and thus do not reflect the actual
cost of installing and operating a UVB remediation well in the USA.
The remediation costs incurred up to this point amount to $ 352,000, see Table 1.
The expenditures for the contaminant site investigation and the drilling of the
groundwater monitoring wells are not included in this amount.
This total cost figure includes the remediation planning, organization, and project
management, all measurements performed, sample taking, fieldwork, and the supervision
of the remediation operation. Additionally included are the 'expenditures for the boring
of the two remediation wells (respective depths of 15 and 40 m), the supervision of the
boring, the connecting equipment between the UVB and the activated carbon
regeneration unit and its installation, the laboratory analysis work, provision of two UVB
remediation units, the self-regenerating activated carbon unit, and the disposal of the
recovered chlorinated solvents. The cost of the electrical energy consumed by the entire
remediation system was also included, based on a price for electricity similar to the
current typical price in Germany of $ 0.18 per kWh.
The electrical consumption of the UVB I unit is 3.1 kW and for the UVB II unit
4.5 kW. During the elapsed operational time of approximately 19,000 hours, the total
electrical power usage for both 'UVB units amounted to 64,500 kWh.
The self-regenerating activated carbon unit has during its regeneration phase an
electrical consumption of 8 kW. The regeneration cycles were adjusted to coincide with
the incurred contaminant concentration in the exhaust air of the UVB.
Of note with respect to the total remediation cost is that a very comprehensive
monitoring program was required by the legal authorities due to their lack of experience
regarding the mode of operation of the UVB remediation procedure.
The average monthly cost of conducting the remediation (not including initial
equipment costs) was approximately $4,000.
Advantages of the UVB remediation technology which reduce the required
expenditures as compared to a standard hydraulic remediation are:
• No groundwater is brought to the surface, and no waste water is produced.
Thus, the need for a reinfiltration of water into the ground and for a
canalization of waste water are eliminated.
• The need for a conditioning of the exhaust air due to high air humidity is not
present.
• The amount of air required for the stripping process is much less (air/waterratio
of 10:1 for the UVB, but 50:1 for hydraulic stripping units). This becomes
especially signifigant when the quantity of groundwater to be remediated is very
large.
258
-------
compilation of costs
planing, organization, project management,
provision of the UVB units
remediation monitoring, field work
laboratory analytical work
boring of two remediation wells, equipment
installation
activated carbon unit and disposal of the
recovered contaminants
total electrical energy costl*
total
%
21.8
21.5
8.2
15.3
24.1
9.1
100.0
us$
77,000
76,000
29,000
53,000
85,000
32,000
352,000
Table 1. Total remediation costs (see first paragraph of this chapter).
GROUNDWATER AND SOIL REMEDIATION USING THE UVB TECHNIQUE
WITHOUT SEPARATING PLATE AND ADDITIONAL PUMP AT A SITE IN BERLIN
(Contribution by P. Brinnel and F. Hirschberger)
Brief Description of the Original Situation
An underground contamination of chlorinated hydrocarbons (CHC) was found on
the plant site of a metal treatment firm in West Berlin. 1,1,1-Trichloroethane and
dichloromethane were identified as the primary contaminants. The exact location of the
original source of the contamination was found to be in an area where degreasing
operations were performed. The existence of the contamination had been discovered
earlier, but the initially collected data was not sufficent for the development of a
remediation concept and plan as the degree and distribution of the contamination in the
groundwater was not determined. Due to the fact that a network of groundwater
monitoring wells, intended for a previous remediation plan, had already been installed
on the site and that the location of the source of the contamination was known, an initial
groundwater sampling of all operational monitoring wells was done before the UVB
remediation was begun (April 3, 1989). The results of the analysis of these 'time-zero'
samples are presented as the original values in Table 1 and are graphicly displayed in
Figure 29.
During the course of the remediation, four additional groundwater monitoring wells
were installed (July 1989). The local hydraulic data of the aquifer were, unfortunately,
not determined. Also, we were unable to carry out a short pump test on the site.
259
-------
Underground Characteristics, Hydrogeological Data
Deposited under an approximately 0.4 m thick cement layer or a sporadiciy occuring
1.5 m thick fill layer is a 3 to 4 m thick layer of fine sands which possess a slightly silty
composition. The groundwater table exists within this layer, approximately 3.8 m under
ground level. Under these pure fine sands lies a medium-grained and sporadiciy fine
gravel structured layer of fine sand which is 2 to 3 m thick. Below this layer, coarse
gravel containing, medium-grained sand is locally found.
The bases of the uppermost aquifer form fine and coarse-grained clays or dense marl.
The upper edge of the clay and marl layers rises in the direction of the groundwater flow,
southward, from 10.5 m (UVB, well SB1) to 8.3 m (P3) below ground level.
Correspondingly, the aquifer thickness increases in the upstream direction and reaches
a minimum in the downstream direction (see profile diagram in Figure 30).
According to the geological surveys, the hydraulioiionductivity coefficent for this area
can be very roughly estimated in the range from 10^ to 10"6 m/sec. The largest yearly
range of groundwater deviation is reported to be approximately 30 cm (see Figure 31).
During the course of the remediation, the groundwater gradient remained constant at
approximately 2 x 10"3. From this data a very rough estimate of 1 m/day for the transport
velocity can be derived.
P7
P6
BL2
prottiithvntcr flow
9
P3
€
BL1
€
0 10 20 30m
i 1 1 1
remediation well
monitoring well:
groimdwnfer
soil nir
predominance of
dichlorntnctlinne nnd
1,1,1 -trichloroethnne
predominance of
lefracliloroefhcne
Figure 29. Site plan, goundwater flow direction, and delineation of primary
contaminant distribution.
260
-------
Flange
Values in
neters below
Groundlevel
1
Concrete
Seal ing
ti
a
a.
01
c
VI
1}
o
a
U>
-1 2-3 mr
i.
13
to
C
_l
a
c.
uj
Cenent + Beto
Gravel Lined
Filter,,
SBrl.7
Grauel 2-3
So 1 id Cos i na
Pipe
ravoi 2-3mm
c
0
a t.
C 0> '
0 h>
a i nod.
• ' • * *"
t ' ^ • . p
" . _5^Q f ine— gr a i ned nediun sand
• 0 » •
*. '
" • '" .5 5Q slightly fine granul-r s-nd
. " '" ^yiQ slioh'tiu aranular- sand
"* "^™ * • — *
• 1 ;
_ • slightly fine— grained
- " • -8.30 nediun =ar.d
• • *
' ; -Q/n f i "e— ur a i ned nediun sand
• • ^ - *
. •-gyg "ediun-yt-ained coarse sand
"°
-------
Remediation Measures, Remediation Goal
The remediation well (bore diameter 500 mm, construction diameter DN 400) was
constructed according to the local geologic conditions (see Figure 30). Integrated into the
well design was an IEG Absorption Filter, especially designed for UVB operation. Due
to the confined physical conditions at the well location (inside a factory building), a well
sounding tube could not be installed. The type of well construction employed allows for
the simultaneous remediation of contaminants from the groundwater, the capillary zone,
and the soil air.
The goal for this remediation case is the attainment of a residual concentration of
the primary contaminant, 1,1,1-trichloroethane, of 50 /xg/1.
After the initial sampling was performed, a Model 400 UVB (without separating plate
and without supporting pump) was installed in the remediation well. The UVB was put
into operation on April 4, 1989.
Monitoring Program, Course of Remediation, Follow-Up Verification
The monitoring program consisted of the measurement and recording of contaminant
concentrations in the groundwater, groundwater temperature, groundwater oxygen
content, operational data of the UVB, contaminant removal data, and groundwater table
UVB-Remediation of Groundwater
Groundwater Table Levels PI -P3
32. 60
0 6 12 18 26 32 38 44 50 56 64
3 9 15 23 29 35 41 47 53 59
Elapsed Remediation Tune (Week)
Figure 31. Groundwater levels during the remediation period.
262
-------
levels (due to the construction of the monitoring wells, only the upper level of the aquifer
could be measured). The monitoring cycles were continuously adjusted according to the
data obtained. This resulted in a gradual increase in the sampling interval over the course
of the remediation.
Figure 31 demonstrates the groundwater levels during the remediation. The
groundwater level data showed no striking influence on the groundwater flow direction
within the effectivity radius of the remediation well.
As explained above there is a strong circulation flow within the well when no
separating plate is used (see Figure 2). Figure 32 demonstrates that the oxygen saturation
is high and Figure 33 shows that the temperature is nearly constant over the total depth
within the well.
The contaminant concentration curves of the individual monitoring wells indicate -
as would be expected with the implementation of a circulation process in the central
position of a contaminant plume - a discontinual course, as shown by the peaks, of
contaminant removal from the groundwater within the remediation area. The
concentration curves in the Figures 34 to 36 show that this removal is relatively quick
and thorough.
Oxygen-Concentration Change - DVB
65 ' . i i i i i i i i i i i . . i i i i i i . . .
1 3 6 10 14 18 22 26 32 36 >53 64
2-4 8 12 16 20 24 30 34 42 59
Elapsed Remediation Time (Week)
Figure 32. Oxygen saturation(%) in the UVB(well SB1) during the remediation period.
263
-------
i y
8 •
7 •
ID"
<5
g 1 C
! 15 -
03
& H
£ U"
w
o>
3 1 -T
r
~^J * ^ A-f
\f \ I
' IK V j
I
0 "^ 1 1 1 1 1 1 1 1 T" 1 1 1 1 1 1 1 1 1 ' ' ' i
1 3 6 10 14 18 22 26 32 36 53 64
2 4 8 12 16 20 24 30 34 42 59
Elapsed Remediation Time (Week)
-*#-
Temp./above
Temp./below
Figure 33. Temperature in the UVB(well SB1) during the remediation period.
Three different contaminant distribution patterns could be detected within the
monitored area of contamination. In the area surrounding the remediation well and
monitoring well PI, the contaminants dichloromethane (DCM) and 1,1,1-trichloroethane
(111TCA) predominate (see Figure 35). Around monitoring well P2, tetrachloroethene
(PCE) dominates the contaminant spectrum (see Figure 36). By monitoring well P3, the
contaminants remain uniformly distributed. An examination of the time course of the
remediation shows that over the first 12 weeks the central 'dichloromethane province' has
been remediated and that the 1,1,1-trichloroethane concentration, after an initial increase
caused by the UVB process, has been significantly reduced. Further on during the course
of the remediation, a shift towards higher proportions of tetrachloroethene in the total
contamination spectrum appears in the individual monitoring wells at different times.
From the 32nd to the 36th week of the remediation, according to monitoring point,
relatively low., compared to the original values, total CHC concentrations are observed
within the entire area of remediation.
After consultation with the responsible civil authorities, the operation of the UVB
was terminated in the 66th week, on July 5, 1990. The first follow-up verification was
conducted in the 69th week and detected no appreciable renewed increases in the
contaminant concentrations (see Figures 34(a) and 35(a)).
After the disconnection of the UVB equipment on July 5, 1990, the factory building
in which the well was located was torn down. The construction of a new storage building
was then begun on the site. This construction work and the accompanying soil removal
have not yet been completed.
264
-------
(a)
c nnn -•
bUUU
5000 -
5*000 •
e
e
3 3000 -
2
*J
c
e 2000 •
o
oou
o i
ITVB-Remediation of Ground water ||
CHC-Concentration SBl /Upper Level 1
| moron o»w «••* |
l\.
1 \
jj -
/\\A
.c. 3 5 7 9 12 16 20 24 28 32 36 42 52 64 83
2 4 6 8 10 14 18 22 26 30 34 38 46 58 69
Elapsed Remediation Time (Week)
(Lc. = initial coocentntioa)
1 1UTCA
""•
IPCE
(b)
6000 -
5000 -
54000 -
e
3 3000 -
1
v
e 2000 -
1 000.-
o -
UVB- Remediation of Ground water |
CHC-Concentration SBl /Lower Level |
JHffiBDOiTl QmUS. Obvnral j
.
-V
X \r^ •
V '"-V"\. /J*^L_
;>4---^T^r--»^l. 7^ T „ T^ ,
.c. 3 5 7 9 12 16 20 24 28 32 36 42 83
2 4 6 8 10 14 18 22 26 30 34 38 46
" Elapsed Remediation Time (Week)
(Id = Initial cmcnlraUonl
-O-
E-CHC
111TCA
->w-
PCE
Figure 34. CHC concentration (Mg/1) in the UVB (well SBl) during the remediation
period; (a) upper level, (b) lower level.
265
-------
(a)
5000
UVB-Remediation of Groundwater ||
CHC-Concentration Change PI/Upper Level I
i.e. 3 5 7 9 12 16 20 24 28 32 36 42 52 64 83
2 4 6 8 10 14 18 22 26 30 34 38 46 58 69
. Elapsed Remediation Time (Week)
{Lc. = imtiil conceolnUoa)
(b)
ODUO
3000 •
^2500 •
a
2000 •
a
15
•3 1500 •
i
1 000 •
500 •
UVB-Remediation of Groundwater |
CHC-Concentration Change PI /Lower Level|
1 nwomo»M obmmi 1
•
1
1
1
Vl
\\ « *
\\ AAV^ r\
\ j/ \ \ lr~~^\
^^^ -^Ayk,. T^rX
.c. 3 5 7 9 12 16 20 24 28 32 36 42
2 4 6 8 10 14 18 22 26 30 34 38 46
Elapsed Remediation Tune (Week)
(Lc. — inilbJ cooccutnUoo)
T-THf
1UTCA
-4K-
PCE
Figure 35. CHC concentration Og/1) in monitoring well PI (upstsream of the UVB)
during the remediation period; (a) upper level, (b) lower level.
266
-------
(a)
UVB-Remediation of Groundwater
CMC-Concentration Change P2/Upper Level
2500
i.e. 3 5 7 9 12 16 20 24 28 32 36 42 52 64 83
2 4 6 8 10 14 18 22 26 30 34 38 46 58 69
Elapsed Remediation Time (Week)
(ic. = imUil co
(b)
UVB-Remediation of Groundwater j|
CHC-Concentration Change P2/Lower Level I
i-c. 3 5 7 9 12 16 20 24 28 32 36 42
2 4 6 8 10 14 18 22 26 30 34 38 46
Elapsed Remediation Tune (Week)
(La = initial cooccntntim)
Figure 36. CHC concentration (Atg/1) in monitoring well P2 (downstream of the UVB)
during the remediation period; (a) upper level, (b) lower level.
267
-------
We subsequently learned that an additional follow-up verification was performed
at a later date by the responsible civil authorities. However, we have been unable to
obtain the data from this verification. During the 83rd remediation week a follow-up
verification was once again performed. The results showed the greatest total CHC
residual concentration to be 30jug/l (well P2). Thus, up to 4 months after the end of the
UVB remediation, the contamination levels remain well under the original remediation
goal, and they only slightly exceed the values of the 'Drinking Water Allowable Limits'
(German: Trinkwasserverordnung, TVO from 1986) 25 Mg/1 total CHC (see Tables 2 and
3).
number of
remediation
week
0
18
48
64
69
84
concentration [fig/1] in the monitoring well
UVB (SB1)
top
3000
34
28
11
9
11,
bottom
1067
68
29
n.s.
n.s.
8
PI
top
3032
136
19
3
4
9
bottom
3092
16
12
n.s.
n.s.
n.s.
P2
, top
749
1503
53
3
3
30
bottom
420
122
22
n.s.
n.s.
n.s.
P7
top
-
372
5
0.1
n.s.
n.s.
bottom
-
16
0.5
n.s.
n.s.
n.s.
operation
before start
stop
control
control
n.s. = not sampled
Table 2. Total of CHC concentration (^g/1) in the groundwater at selected monitoring
wells.
number of
the
remediation
week
0
18
48
64
69
84
concentration [/jg/1] in the monitoring well
P3
top
52
24
19
0.5
n.s.
3
bottom
55
28
17
n.s.
n.s.
n.s.
P4
top
-
35
2
0.1
n.s.
0.6
bottom
-
50
0.4
n.s.
n.s.
n.s.
P5
top
-
16
3
4
5
n.s.
bottom
-
33
38
n.s.
n.s.
n.s.
P6
top
-
14
27
2
n.s.
n.s.
bottom
-
17
2
n.s.
n.s.
n.s.
operation
before start
stop
control
control
n.s. »• not sampled
Table 3. Total of CHC concentration (Mg/1) in the groundwater at the selected
monitoring wells.
Before the beginning of the remediation, four soil air monitoring wells were installed
for the determination of the extent of remediation in the unsaturated soil zone, site plan
see Figure 29. A signifigant reduction in the total CHC concentration appeared in three
of the four wells (Table 4). The values obtained from well BL4, located 30 meters away
268
-------
from the UVB well, show in contrast no appreciable influence Irorn tne remediation
process. Upon completion of the construction work at the. site, an additional soil air
remediation project is planned. This remediation will utilize a soil air circular flow
process which will also remediate the capillary zone. Such a technique is practical in this
situation as the ground surface is covered with two thick layers of concrete.
number of
the
remediation
week
0
18
48
64
concentration [mg/m3] in the monitoring well
BL1
256
92
23
9
BL2
99
39
28
20
BL3
2818
1331
656
352
BL4
257
230
247
186
operation
before start
stop
Table 4. Total of CHC concentration (mg/m3) in the soil air.
Operational Data and Contaminant Removal
The operational data obtained during the running of the UVB equipment reveal a
relatively continuous amount of air withdrawal from the underground. The proportion
of soil air within the total exhaust air averaged 30%. This was sufficent for the
maintenance of a roughly continual remediation of the unsaturated and capillary zones.
The humidity of the exhaust air ranged between 31 and 40 % and its temperature
between 31 and 33 °C.
The contaminant removal at the beginning of the remediation was very high due tc
the central positioning of the well. As a result, the first charge of activated caru .n
reached its absorbtion capacity after only 148 hours of operation. The filter.unit,
absorbed a total of 473 kg of matter within this time. This yields an average loading rate
of 40 %. The second charge of activated carbon reached its absorbtion capacity nfier
1510 hours of operation and absorbed a total of 425 kg, yielding an average loading rate
of again 40 %. The third charge of activated carbon reached its absorbtion capacity alter
8136 hours of operation and absorbed a total of 405 kg, yielding an average loading • KI--
of 33 %. In all, a total of approximately 1300 kg of contaminants were removed rrora
the underground.
Remediation Expenses
The following itemized remediation costs are in accordance with the West Gerrnar
market prices existing at the time of the remediation. For this reason they are given ir
269
-------
German Marks (DM). What should be especially, noted « that in l.«KO. when the
remediation was carried out, Berlin existed as an economic 'island'. Thus, thp pn«'.e
structure with respect to conducting business in Berlin was distinctly different as
compared to the rest of Germany. It should also be noted that the implementation of
this, at the time relatively unknown, technology in Berlin required a relatively large
monitoring expenditure in comparisioh to what would currently be necessary. Due to the
different legal and economic structures found in Berlin, the former West Germany,
Canada, and the USA, the remediation costs mentioned here should neither be directly
converted nor simply transfered over to the currency of the other country. Instead, tb^y
can best be used within a cost comparison where the incurred expenses for a UVB
remediation versus those for a conventional hydraulic remediation (pump and treat
technique) of an individual contaminated site in Berlin or Germany are directly compared
with one another. Likewise for a contaminated site located in the USA or Canada, only
the incurred costs for each remediation technique should be directly compared and not
with the costs from a different country.
For the remediation case presented here, the total costs amount to approximately
430,000 DM. The costs relating to the original contamination detection and survey are
not included. This amount includes project planning, organization, and management, the
UVB remediation equipment and its installation, travel and shipping expenses, the
required field work, drilling costs, and the entire monitoring program, including
laboratory analytical work. Also included are the expenditures for the activated carbon
and its regeneration. The accompanying energy (electrical) costs are not included as they
were paid for directly by the firm on whose property the remediation was performed.
The power consumption of the UVB equipment is 3.1 kW. Over the course of the
remediation, the total running time was approximately 11,000 hours. Thus, the amount
of electrical energy consumed was approximately 30,000 kWh (approximately 9,000 DM;
equivalent 5,300 $). During the remediation process no waste water was produced, thus
avoiding further handling costs.
compilation of costs
planning, organisation, project management,
remediation equipment (including shipping)
field work
laboratory analytical work
drilling costs
activated carbon and regeneration
total
total costs
%
25.5
17.3
29.1
11.6
16.5
100.0
DM
110,000
74,000
125,000
50,000 ,
71,000
430,000
equivalent $*
64,000
44,000
74,000
29,000
42,000
253,000
* 1 US$ = 1,70 DM
Table 5, Total remediation costs for the special situation in Berlin.
270
-------
Keeping in mind the previously described special situation of Berlin, the total
remediation cost is distributed as compiled in Table 5.
The average monthly cost with respect to the total cost was approximately 28,500
DM. However, in such a project the expenditures are not linearly distributed. Thus, the
actual incurred average monthly cost (excluding the installation and regeneration costs)
was approximately 14,000 DM. When excluding the extra cost of the special monitoring
program employed in this case, the average monthly operational cost for such a UVB
remediation system (including electricity, activated carbon and its regeneration, and a
standard minimal monitoring program) amount to approximately 7,100 DM (equivalent
4,200 $).
CONCLUSION
The UVB technique can be used for in situ stripping of volatile contaminants from
the groundwater by itself or in combination with added nutrients and/or electron
acceptors for in situ biodegradation. Further, the circulation flow around a UVB can be
utilized exclusively for in situ biorestoration. For all of these cases, the hydraulic system
offers many advantages, particularly when compared with a typical hydraulic remediation
system of pumping, off-site treatment, and reinfiltration of the groundwaier. Such
advantages include:
• No lowering of the groundwater level
• No groundwater extraction
• No waste water
• Less permeable, horizontal layers are penetrated vertically
• Remediation of the groundwater takes place down to the bottom of the aquifer
• Even at low well capacity, remediation operation is continuous
• Soil air extraction is possible at the same time
• Low space requirement
• Investment and operating costs will be considerably lower.
When the water discharge through the well casing is directed downward, the
hydraulic head is lowered at the well top (-Ah/2, Fig. 9b), but this amount is mucti smaller
than that caused by a normal withdrawal well.
The total aquifer is caught by the circulation flow of a UVB. When using different
wells for extraction and infiltration, only those areas of an aquifer which are more
permeable are penetrated. The other areas are reached mainly by diffusion. The
groundwater flow system will only locally be influenced, there is no need for large
extended groundwater flow investigations. Further, a plume can directly be treated
without pumping lots of clean water as in case of using pump and treat methods.
A layered aquifer enlarges the sphere of influence or the distance between the well
and the stagnation point. This has been found by numerical simulations and by.
comparison with field measurements of a tracer test. On the other hand, the positive
effect of a layered aquifer is limited when an aquitard is present. Here, several
remediation systems must be installed, one for each aquifer.
271
-------
ACKNOWLEDGMENTS
The first two authors thank IEG mbH, D-7410 Reutlingen, for financial support or
the numerical investigations. In particular, B. Bernhardt, IEG mbH, D-7410 Reutlingen,
inventor and patent holder of the UVB method; W. Buermann, Institute or
Hydromechanics, University of Karlsruhe; W. Kaess, D-7801 Umkirch; and HJ. Lochte,
UTB mbH, D-4020 Mettmann, are gratefully acknowledged for many helpful discussions
and contributions to the operation and development of the vacuum vaporizer well.
REFERENCES
Herrling, B.; Buermann, W. "A New Method for In-Situ Remediation of Volatile
Contaminants in Groundwater - Numerical Simulation of the Flow Regime." In
Computational Methods in Subsurface Hvdrologv: Gambolati, G.; Rinaldo, A.; Brebbia,
C. A.; Gray, W. G.; and Finder, G. R; Eds.; Springer: Berlin, 1990; pp 299-304.
Herrling, B.; Buermann, W.; Stamm, J. "In-Situ Remediation of Volatile Contaminants
in Groundwater by a New System of 'Vacuum-Vaporizer-Wells'." In Subsurface
Contamination bv Immiscible Fluids: Weyer, K.U., Ed.; A.A. Balkema: Rotterdam, 1991a;
[in press].
Herrling, B.; Stamm, J.; Buermann, W. " Hydraulic Circulation System for In Situ
Bioreclamation and/or In Situ Remediation of Strippable Contamination". Proc. Int.
Symp., 19-21 March, 1991, San Diego, California; In Situ and On-Site Bioreclamation:
Hinchee, R.; Ed.; 1991b; [in press].
272
-------
REMEDIATION OF GROUNDWATER
AND PROCESS WASTEWATER
CONTAMINATED WITH NDMA
AND OTHER TOXICS USING RAYOX®
This paper was presented by Keith G. Bircher.
Until recently few people had ever heard of NDMA.
It is not in the EPA Priority Pollutant list and is not
generally looked for when contaminated waters are
analyzed for pollutants. However, the extensive
occurrence of NDMA precursors in industry and in the
environment, its high mobility in groundwater and its
toxicity have assured that it will receive increased
attention in the future.
WHAT IS NDMA?
CH
CH
N—N=O
N-Nitrosodimethylamine is formed by the nitrosation
of dimethylamine (DMA) with the nitrite ion (N02~)
and is reportedly found in trace amounts in tobacco
smoke condensates and in cured meat products such as
bacon. NDMA is a chemical of lexicological interest
because it has been shown to be carcinogenic and
mutagenic in animals, and is probably so in humans.
Following US EPA estimates that NDMA in water at
concentrations of 14 ppt may increase cancer risk by
one case in 100,000, the Ontario Ministry of the
Environment recommends an interim drinking water
guideline of 9 ppt.
Other sources of NDMA would be as byproducts in
production or use of unsymmetrical
dimethylhydrazine (UDMH) which is a rocket
propellant and dimethylamine (DMA) which is used
as an accelerator in vulcanizing rubber and the
manufacture of detergents.
THE BAD NEWS....
NDMA is polar, and highly soluble in water, and as
such, has a low partition coefficient in water. It
therefore moves through soil as readily as chloride,
indicating it is not appreciably retarded as it moves
through groundwater. It does not absorb onto carbon
and is not amenable to air or steam stripping from
water. It is also resistant to biodegradation and is
unlikely to degrade in the subsurface.
THE GOOD NEWS ....
NDMA absorbs UV light and readily photo
dissociates. Figure 1 shows the absorption spectrum
of NDMA between 200 and 400 nm at a
concentration of 10 ppm. The extinction coefficient
at the peak wavelength of 228 nm is 10,000.
Table 1 shows the wavelength of peak absorbance
lmax, extinction coefficient e and Quantum yield
for some typical compounds found in contaminated
water.
SOLARCHEM ENVIRONMENTAL SYSTEMS
40 West Wilmot Street, Unit #6
Richmond Hill, Ontario.
Canada L4B 1H8
Telephone: (416) 764-9666
Fax: (416) 764-9669
273
\
7320 Smoke Ranch Road, Suite H
Las Vegas, Nevada, U.S.A.
89128
Telephone: (702) 255-7055
Fax- (702) 255-7280
-------
1.6
g 1.2
o
0)
o
c
ffl
O
(0
0
FIGURE!
UV ABSORPTION SPECTRUM OF NDMA
7C-7I*'
n-71*
200 240 280 320 360 400
Wavelength nm
Table 1
UV Absorption of various Compounds
Compound
NDMA
Benzene
Chlorobenzene
Pentachlorophenol
Trichloroelhylene
1,1,1 Trichloroethane
2,4,6 Trinitrotoluene
^•max
228
215
215
220
215
200
230
e
10,000
533
8,000
25,000
7,000
155
18,000
*
~1
.1
.4
.01
.3
.1
.001
e.cfr
10,000
53
3,200
250
2,100
15
18
As can be seen, although Pentachlorophenol and TNT
absorb more strongly, when you take the product of
extinction coefficient and the quantum yield as a
rough measure of relative photodegradation efficiency,
NDMA comes out on top. This is just as well since
NDMA does not react very fast with the hydroxyl
radical which, as is explained later, is the other
pathway to destruction employed by the Rayox®
enhanced oxidation technology.
HOW DOES IT WORK ....
The major UV absorption is by the nitroso group
with a degradation mechanism postulated by Polo
etal.1 presented in figure 2.
274
-------
FIGURE 2
MECHANISM FOR PHOTODESTRUCTION OF NDMA
(CH3)2N-N=O + H+
(CH3)2N~
(b)
(a)
H
NO
ho>
J
(CH3)2 NH2 + HNO2
H+
CH3 ^
>-i-i
-3/ 6
(CH3>2
'NO
i
As can be seen from Figure 1, the major absorption
of NDMA is n to K* at a lmax of 228 nm. It is,
therefore, important to have a lamp that is efficient in
the production of light in this UV band (200 to 240
nm). As can be seen Jrom Figure 2, the major
products from this photolysis are dimethylamine and
nitrite (NO2~). The nitrite can be oxidized by the
addition of hydrogen peroxide to produce nitrate and
hence preclude the possibility of reformation of
NDMA, although this reformation has not been
observed to occur to any significant degree in the
laboratory.
GROUNDWATER
Solarchem currently has several Rayox® systems
installed or under final system design for treating
NDMA in water on a continuous basis.
One of the Rayox® systems is for treating
groundwater from a well that has been sunk to protect
adjacent drinking water wells from being
contaminated. This system has eight 30 kw
Rayox® reactors and treats up to 600 gpm water,
removing over 99.9% of the NDMA from 20 to 40
ppb to the drinking water standard of 9 ppt as shown
in Figure 3.
FIGURE 3
DESTRUCTION OF NDMA IN GROUNDWATER
lOO-i
1 OR .
N 1r
D 4\
M 1.M
A
n 0.1.
P V.M..
P
001.
0 001 .
\
\
.11.
A Groundwaur
•° Spiked to
20ppb
"Detection
Limit
0 51015
Rayox Dose
A Solarchem lamp that has been developed to
maximize the efficiency of UV production in the 200
to 240 nm range is employed and achieves the
discharge target The power required is only 2 kWh
per 1,000 gallons per decade or order of magnitude of
removal. Thus for 99.9% removal, only 6 kWh per
1,000 gallons is required to clean the water.
275
-------
PROCESS WASTEWATER
Another of the Rayox® systems is for cleaning
process wastewater from a chemical manufacturer
before discharge to the local POTW (remember,
NDMA does not readily biodegrade). In this case,
nine 30 kW Rayox® reactors are used to clean up to
100 gpm of water from 30 ppb to 0.25 ppb.
A flow schematic of this system is shown in Figure
4. As can be seen, the nine 30 kw Rayox® reactors
are arranged in three groups of three, each mounted on
a separate skid. The piping is arranged such that one
skid can be bypassed for maintenance while the others
remain on-line.
FIGURE 4
LAYOUT OF 270 KW RAYOX ® SYSTEM
Influent
Effluent
Skid with 3 Reactors
30 kW Rayox®Reactor
Reactor
Power Supply
This arrangement gives considerable flexibility where
a additional skid could be added, should that prove
necessary in the future, or one of the skids could be
moved to a different location.
THE ENHANCED OXIDATION PROCESS
Another means for the removal of organic
contaminants from water is through oxidation by the
hydroxyl radical (HO"). The hydroxyl radical is a very
powerful oxidizing agent which reacts very rapidly
with virtually all organic compounds. The rate
constant for HO" attack on organic compounds is
typically between 106-109 times that of ozone attack.
There are several methods of generating hydroxyl
radicals, which have been described by several
authors.2 Three common methods are: the photolysis
of hydrogen peroxide with ultraviolet light, the
reaction between ozone and hydrogen peroxide and the
use of Fenton's reagent. These approaches can be
described by the following simplified chemical
equations:
H2O2 + hu
20s + H2O2
-» 20H"
-» -» 20H* + 302
OH" + Fe+3 + OH'
276
-------
GASOLINE COMPONENTS...
Petroleum Hydrocarbons such as Benzene, Toluene
and Xylenes (BTX) are found in ballast water, in
groundwater contaminated by leaking underground
storage tanks, in waste water from tank cleaning
operations and in runoff water, etc.
Whilst BTX's can be removed by carbon, Solarchem
has come up with a process, involving UV and a
proprietary catalyst (Enox 510) that in parallel trials
against activated carbon yields operating cost savings
up to 80%. In addition other components present in
gasoline contaminated water such as MTBE are not
readily removed by carbon. Figure 5 shows the
performance of a Rayox* system on ballast water at
a major oil terminal.
FIGURES
DESTRUCTION OF GASOLINE COMPONENTS
BY RAYOX IN LARGE BALLAST WATER
FIGURE 6
DESTRUCTION OF BTX, MEK and ACETONE
in GROUNDWATER
\
CATALYTIC ENHANCEMENT
Figure 6 shows the extent of enhancement on the
performance or destruction of MEK of a Rayox®
system that can be achieved by Solarchem's
proprietary catalyst Enox 510. This particular
application had gasoline components and Acetone in
addition to the MEK in groundwater at a North
Western paint manufacturer.
Other applications that are amenable to treatment by
Rayox® include Volatile Organic Compounds such
as chlorinated solvents, Pentachlorophenol in the
wood preserving industry and dissolved explosives
produced in their manufacture such as TNT.
REFERENCES
1. POLO, J. & CHOW, Y.L. Efficient degradation
of Nitrosamines by Photolysis, IARC Scientific
Publications, p 473.
2. Stevens, R.D.S. Rayox® : A Second Generation
Enhanced Oxidation Process For Groundwater
Remediation, Symposium on Advanced Oxidation
Process for the Treatment of Contaminated Water and
Aii, Toronto, June 1990.
3. Hoine, J. and Bader, H. The Role of Hydroxyl
Radical Reactions in Ozonation Process in Aqueous
Solutions. Water Res. 10, 377-386(1976).
277
-------
W B G N
Edisonweg 52
Postbus 30
2950 AA Alblasserdam
The Netherlands
Telephone 01859 - 30200
Telefax 01859-30196
Banknummer 42.44.83.122
t.n.v. St. WBGN Alblasserdam
Handelsregister K.v.K.
's-Hertogenbosch 83767
REMEDIATION OF CONTAMINATED SEDIMENTS IN
THE NETHERLANDS
Paper to be presented at the
third forum on innovative hazardous waste treatment
technologies domestic and international
to be held in dallas, texas, U.S.A., June 1991
H.J. van Veen, TNO; Apeldoorn, The Netherlands
H.J. van Hasselt, NBM Bodemsanering; The Hague, The Netherlands
G. Roza, Herik Baggerwerken; Sliedrecht, The Netherlands
278
-------
1. INTRODUCTION
In the 1960s, the Dutch Water Boards were confronted with a deteriorating
water quality. Consequently, several actions were taken to reduce the dis-
charge of contaminants, such as heavy metals and organic micropollutants. At
the time, one did not realize that not only the surface water, but also the
suspended matter and sediment were polluted. Today, however, we are well
aware of the sediment being contaminated in many watercourses.
In the Netherlands, contaminated sediments are manifest as a dual environr
mental problem:
As contaminated aquatic soil with the corresponding environmental im-
pact; and
As a dredged-sludge problem: many Dutch watercourses must be dredged
for nautical reasons and for water management.
The dredged-sludge problem is currently dominant. This means that remedia-
tion of contaminated sediments in the Netherlands refers specifically to
dredged-sludge remediation. Until a few years ago, all dredged material was
disposed of in confined disposal facilities. Recently, treatment technology
has been applied to improve the quality of the dredged sludge. Thus, several
harbors that were seriously contaminated with specifically PAHs, oil, and
metals, have been remediated. Their remediation included the dredging and
processing of the sludge by classifying and dewatering into a fraction for
beneficial use, and into a concentrate for disposal.
The Dutch Ministry of Traffic and Public Works tried to accelerate the deve-
lopment of technology by some sort of competition. Dredging companies and
contractors were asked to make plans for the remediation of ten sites with
contaminated sediments. The winner of this competition is a group of coopera-
ting companies called WBGN.
The WaterBodem Groep Nederland (WBGN - Aquatic Soil Group Netherlands) spe-
cializes in dredging and treating contaminated sediments.
This paper gives a survey of technologies already applied and technologies
now being developed by WBGN.
2. APPLIED TECHNOLOGIES
Since 1985, technology has been applied to reduce the quantitative volume of
contaminated dredged sludge to be disposed of. The applied process consists
of a combination of two techniques: hydrocyclones and dewatering. In this
way, a relatively clean fraction is separated from the dredged sludge, while
the residual fraction is reduced in volume as much as possible.
279
-------
Hydrocyclones
Particle classification is performed by hydrocyclones (Fig. 1). A hydro-
cyclone has one inlet, two outlets, a vortex finder, and an apex nozzle.
The two outlet flows are called overflow and underflow. The fluid feed en-
ters the cyclone through the tangential inlet, bringing about a downward flow
that first swirls round the outside wall down to the apex, and then is forced
upwards in the center of the cyclone, passing up and out through the vortex
finder.
Heavy particles entering the feed will be thrown to the wall of the cyclone
due to the centrifugal force of the downward flow, leaving the cyclone
through the apex. Less heavy particles do not have enough time to reach the
wall of the cyclone and will, therefore, leave the cyclone together with the
larger part of the water in the overflow. Thus, a hydrocyclone separates
dredged sludge into heavy sand particles, on the one hand, and fines and or-
ganic material, on the other.
On account of the differences in sorption properties , fines and organic
material have a high contaminant content, compared to sand. This means that
hydrocyclones separate a relatively clean sand fraction from the slime frac-
tion in which a concentration of contaminants is found.
The effect of hydrocyclones is characterized by: 1) the distribution of the
dry matter [E^ ]l; and 2) the distribution of the contaminant [Ex]2.
As early as 1983, the applicability of hydrocyclones for treating contamina-
ted dredged sludge was recognized. The technique has been applied in several
dredging operations, but does not always offer a solution, particularly not
for dredged sediments with a high content of very small particles and a high
organic matter content (peat). Fig. 2 shows some results obtained in hydro -
cyclone experiments with dredged sludge from various sites, and with various
contaminants . Hydrocyclones have a more favourable effect when the data
point is closer to the origin of the diagram. Consequently, it can be con-
cluded that hydrocyclones often give good results , but not always .
Dcvatering
Various dewatering equipment is available. Three apparatus qualify for the
dewatering of dredged sludges and of the slime fraction of dredged sludges:
the belt press, the filter press, and the decanter. In general, it can be
stated that filter presses give the highest dry-matter content results,
whereas decanters give the lowest dry-matter content results.
Ban - separation efficiency for the dry matter; this is the percentage
of the dry matter that leaves the hydrocyclone as underflow (sand
fraction)
EX — separation efficiency for the contaminants; this is the percenta-
ge of the contaminants that leaves the hydrocyclone with the underf-
low.
280
-------
In most cases, it is necessary to use flocculants for dewatering. When ap-
plying belt presses and filter presses, flocculants bring about a good filte-
rability; when decanters are used, flocculants help to reach a clear decan-
tate. All three apparatus mentioned are used in practical dredged-sludge
treatment. ; •
Dewatering aims at reaching a volume reduction of the sludge or the slime
fraction produced by hydrocyclones. Fig. 3 presents the effect of dewatering
on volume, starting from a slime fraction with a dry-matter content of 5%
after using hydrocyclones. It shows that at increasing dry-matter content
levels, a considerable volume reduction is reached in the first instance.
However, higher dry-matter content levels (approximately 40%) lessen this
reduction.
If dewatering aims at volume reduction, Fig. 3 shows that further dewatering
becomes less cost-effective. Dewatering costs increase sharply as higher
dry-matter content levels are reached.
Cases
As described, the hydrocyclone/dewatering combination results in two frac-
tions; first the relatively clean sand fraction, second the dewatered slime
fraction.
A number of practical cases prove that the hydrocyclone/dewatering treatment
has been very successful (see Table 1). Volume reductions of up to 75% of
the material to be disposed of have been achieved.
*
In some cases, the hydrocyclone/dewater-ing operations were not successful; an
example of this was the remediation of nine harbors in the Oosterschelde
(Eastern Scheldt).
The remedial action plan was based on some preliminary investigations.
The plan involved:
Dredging, transportation, and disposal of the bulk of the material;
and *
Dredging, transportation, classification by a hydrocyclone, dewatering
and incineration of the slime fraction. This action was planned only
for the dredged sludge from hot spots.
Table 2 gives some information on this case. It appeared that there was a
large difference between the expectations in the planning phase and the real-
ization on full scale. The main reasons for this difference were:
It turned out that the information obtained in the preliminary invest-
igation strongly differed from the actual situation.
For instance, the sediment composition strongly deviated from the
composition expected on the basis of the "preliminary investigation.
This made it difficult for the contractor in charge of the remediation
to comply with the results described in his quotation.
281
-------
After using hydrocyclones, the sand fraction still showed a high PAHs
concentration, because the PAHs were not adsorbed to the slime frac-
tion, but were present as some kind of tar particle that could hardly
be separated from the sand by hydrocyclones.
This case clearly shows that saving money on research in the planning phase
produces much greater costs in later phases.
3. TECHNIQUES IN DEVELOPMENT
3.1 Biological remediation
General
Dutch research into biological remediation has specifically focused on the
biodegradation of oils and PAHs, because these organic micropollutants occur
most frequently. TNO has carried out laboratory exploratory research into
the biological remediation of dredged sludges contaminated with mineral oils
and PAHs. This research has shown that effective biological cleaning is pos-
sible for several dredged sludges. Spontaneous degradations have been found
in these dredged sludges, where conditions are for these sludges biologically
favorable (as in the case of a bioreactor). From a biological point of view,
such a degradation often passes quickly.
Present research is conducted along two lines:
1. Development of biological remediation techniques up to a practical
scale. This concerns the development of designs for the biodegrada-
tion process that link up with the dredging process.
2. Broadening the fundamental knowledge of the degradation of PAHs and
other substances, such as chlorinated hydrocarbons.
At present, Dutch research emphasizes £he former line.
There are two potential treatment methods for the practical application of
biological remediation:
1. Large-scale extensive treatment, e.g. in aeration basins, ?and landfar-
ming.
2. Intensive treatment in bioreactprs,
Intensive implementation methods
Intensive implementation methods aim at:
operating a process as intensively as possible (with much exertion)
thus realizing conditions that are as optimum and verifiable as pos-
sible
resulting in a treatment period that is as short as possible.
These implementation methods refer to process-type treatment methods (e.g.
bioreactors).
282
-------
Extensive implementation methods
Extensive implementation methods are meant:
to operate remediation methods with relatively slight exertion .(exten-,
sive)
to imply usually less optimum and verifiable conditions. '
These implementation methods refer to large-scale, more or less batchwise
treatments, like biodegradation by landfarming, or treatment as a slurry in
an aeration basin.
A plan has been developed for the extensive treatment of dredged sludges
•This plan comprises separation of the dredged sludge by hydrocyclones,
followed by treatment of the sand in a landfarm and treatment of the slime in
an aeration basin.
Whether an intensive or an extensive manner of implementation is chosen for
remedial operations, is determined by several choice criteria, such as treat-
ment time, costs, and needed space.
TNO has carried out an investigation into the intensive treatment of dredged
sludge in a bioreactor, and the extensive treatment of the slime fraction in
an aeration basin. Landfarming of the coarse fraction (sand) has not been
investigated.
Bioreactor for treating the total dredged sludge
The samples were treated in a laboratory bioreactor. The reactor is a 10
liter rotating drum with baffles. The total dredged .sludge was treated in
the reactor.
Aeration basin for treating the slime fraction
An aeration basin is a large-scale facility (up to 10,000 m3) for treating
the slime fraction of the dredged sludge, after hydrocyclonage treatment.
When aerating basins up to 10,000 m3, the distribution of air within the ba-
sin is an important aspect. In this respect> a. comparison was made with an
aeration basin of a sewage treatment plant, which strives to make the treat-
ment time as short as possible. This means:
1. Installation of aeration elements across the entire surface of the
aeration space, and
2. Sufficient mixing of the waste water (turbulence).
These two conditions are not considered feasible for aeration basins that
have to treat the slime fraction, in view of the size of such basins. The
plan considers the installation of a large-scale treatment depot for the
slime fraction, with intermittent aeration being realized by moving a pontoon
with aerators and mixers slowly to and fro across the length of the basin.
The feasibility of an aeration basin for treating the slime fraction has been
investigated in a 10 m3 pilot basin with intermittent aeration.
283
-------
Landfarming of the sand fraction
Landfarming is a proven technique that is frequently applied "in practical
(tetrestrial) soil remediation. Much experience has been gained with respect
to the degradation of mineral oils and PAHs in particular. Briefly, the con-
taminated soil is spread in 20-50 cm-thick layers in a field that is
especially equipped for this purpose.
Table 3 presents some results of the investigations into the bioreactor and
the aeration basin. It shows that:
Treatment time in a bioreactor is less than in an aeration basin.
Contaminant reduction in a bioreactor is less than in an aeration
basin. This is probably caused by the fact that the biodegradation
rate depends on the contaminant concentration. When starting with a
lower concentration, the biodegradation rate is lower.
Contaminant concentrations after treatment in both the reactor and the
basin are comparable.
Treatment costs involving an aeration basin are lower than those in-
volving a bioreactor, even when the treatment costs of both the slime
fraction and the sand fraction in a landfarm are included.
Expectations are that the biodegradation of PAHs in dredged sludge will be
demonstrated on a full scale by 1992.
3.2 Other remediation techniques
Other remediation techniques to be developed or improved are:
Classification techniques.
Apart from hydrocyclones, there are various other techniques to clas-
sify contaminated particles from dredged sludge, such as froth flota-
tion, magnetic separation, and other mining classification techniques.
Solvent extraction.
Solvent extraction is a process that has been developed for treating
various hazardous wastes. To make solvent extraction feasible for
treating dredged sludge, it is important to develop a high-capacity,
low-cost process.
4. FINAL REMARKS • '
In the Netherlands, the introduction of treatment technology for contaminated
sediments in dredging operations has only recently begun. The introduction
of a relatively simple technique like hydrocyclones already appears to cause
many problems. These problems are the result of, for instance:
An inadequate preliminary survey of the site to be dredged; in this
way, remediation plans are based on incomplete information which later
turns out to be incorrect. '
Underestimation of the degree of complexity of the remediation tech-
nology by the companies involved.
284
-------
Overly fast upscaling of the research results to a practical scale, as
researchers underestimate the implementation problems.
Two conditions must be met to accelerate the remediation of contaminated se-
diments :
A research program and a budget for developing and improving reme-
diation techniques. The Dutch government has allocated about 20 mil-
lion dollars for further research.
A certain guarantee that the techniques developed will indeed be
applied and that not all dredged sludge will be dumped - this in order
to increase the interest of companies in developing remedial techno-
logies. This political choice has not yet been made in the Nether-
lands .
285
-------
Table 1. Results of Practical Hydrocyclone Applications
Barendrecht
1985
Roozendaal
1986
Nijkerk
1986
Dordrecht
1988
Process (*)
Capacity (m3/h)
Outsize fim
Contaminants
Concentration
in sand (mg/kg)
1
20
20
metals, oils
50%
metals: + 15%
Zn: 169
Cu: 28
Cd: 1.8
Volume reduction by
hydrocyclones/
dewatering
2
18
50-60
metals, oils
20%
metals: 1-5%
oils: +0.5%
Zn: 63
Cu: 18
Cd: 0.2
oils: 93
75
3
300
50-60
PAH
70%
PAH: 5-10%
PAH: 1-2.9
300
50-60
PAH,,metals
60%
PAH: ± 5%
metals:
± 10%
PAH: 0.38
Zn: 150
Cu: 38
Cd: 0.9
50
* 1. Test installation consisting of a storage basin, a preseparator
(CBC - Circulation Bed Classifier), a buffer basin and hydrocyclones.
2. Installation consisting of a hydrocyclone and a sieve belt press.
3. Installation consisting of a sieve, three hydrocyclones, a sediment
tank and a sludge depot. Flocculants have been dosed in the delivery
pipe to the depot for a quick first sedimentation, thus making a quick
water drainage possible.
4. See 3. The sludge depot has been replaced by a flat-bottom craft in
which the fine fraction has settled.
286
-------
Table 2. 'Oosterschelde' Case
Expected1'
Realized
Total volume of dredged sludge
(m3)
PAH concentration in sludge
(mg/kg d.m.)
Volume of dredged sludge from
hot spots to be classified
and dewatered (m3)
Volume of clean sand (m3)
Quality of sand
PAH concentration
(mg/kg d.m.)
Volume of dewatered slime
fraction (tons)
Dry-matter content level of
dewatered slime fraction (%)
Costs of dredging and disposal ($)
Costs of classification and
dewatering ($)
110,000
100-200
11,000
5,000
10
1,075
70
8
32
190,000
300-400
7,000
1,000
50
3,500
60
14
100
i)
Expectation based on preliminary investigations,
287
-------
Table 3. Intensive Versus Extensive Treatment Methods
Material to be treated
Contaminant concentration
(mg/kg)
Treatment time (days)
Contaminant reduction (2)
Contaminant concentration
after treatment (mg/kg)
Total costs3'
($/m3 dredged sludge)
Total costs3' aeration basin2)+
landfarming
$/m3 .dredged sludge)
Bioreactor
dredged
sludge
PAH
100
30
60
40
200
Aeration
basin
slime fraction1'
PAH
200
180
85
30
24
50
1)
2)
3)
Slime fraction (after hydrocyclonage) of the same dredged sludge that
has been treated in a bioreactor.
Aeration basin treatment for slime fraction; landfarming for sand frac-
tion of same dredged sludge.
Total costs are capital costs and operational costs per m3 of dredged
sludge, with a dry-matter content level of 40%.
288
-------
feed
overflow
(slime fraction)
underflow
(sand fraction)
Figure 1. Hydrocyclone.
Part of contaminant
in underflow (Ex) X
Waalhaven
Seulhaxen
Donmel
» Naarden
5 Amsrel-Drecht-
kanaai
6 Kampen
7 Haas (Roerfflond)
8 Eemskanaal
9 Noordzeekanaal
10 Singelgracht
11 Zaan
12 Arnhen
13 Scheveningen
14 Dordrecht
15 Dodewaard
Part of dry matter
in underflow (Ed.m.)
1000
volume
(U
500 -
Based on:
- 1 m3 slime fraction
- d.m. content 5X
dry matter content (%)
Figure 2. Hydrocyclone results.
Figure 3. Volumereduction by dewafering.
289
-------
TOXIC CHEMICAL
SPECIALISTS
ECO LOGIC
DEMONSTRATION TESTING OF A THERMAL
GAS PHASE REDUCTION PROCESS
BY
D.J. HALLETT & K.R. CAMPBELL
ELI ECO LOGIC INTERNATIONAL INC.
143 Dennis Street
Rockwood, Ontario
Canada
NOB 2KO
2395 Huron Parkway
Ann Arbor, MI
U.S.A.
48104
143 Dennis Si, Rockwood, Ontario, Canada, NOB2KO
Rockwood (519) 856-9591
Toronto (416)450-7691
Fax (519) 856-9235
290
-------
DEMONSTRATION TESTING OF A THERMAL GAS-PHASE REDUCTION PROCESS
D.J. Hallett, K.R. Campbell,
ELI Eco Logic International Inc
143 Dennis Street
Rockwood, Ontario, Canada
NOB 2KO •
ABSTRACT
Thermal gas-phase reduction of organic hazardous waste is an alternative
to incineration suitable for processing aqueous waste such as harbour
sediment, lagoon sludges, and landfill leachate. The reaction is
conducted in a hydrogen-rich reducing atmosphere at approximately 900°C
and atmospheric pressure. The products of the reaction depend on the
waste constituents but usually include HC1 from the reduction of
chlorinated organics. such as polychlorinated biphenyls (PCBs) and methane
and ethylene from reduction of straight-chain and aromatic hydrocarbons.
The absence of free oxygen in the reactor prevents the formation of
dioxin compounds.
ECO LOGIC has set up a demonstration facility for processing polyaromatic
hydrocarbons (PAH) and PCB-contaminated harbour sediments in Hamilton,
Ontario and has been conducting destruction tests during the spring of
1991.
The demonstration-scale reactor is 2 m in diameter and 3 rn tall and is
mounted on a 15 m drop-deck trailer. A scrubber system and recirculation
gas heating system are also mounted on the trailer, as well as the
electrical control centre. A second trailer holds a propane boiler and
waste pre-heating vessel. The boiler also accepts a small portion of the
scrubbed dechlorinated recirculation gas as fuel. The processing rate
for the demonstration unit is 4-5 kg/min.
Results from the demonstration testing including destruction efficiencies
obtained and processing costs estimates will be discussed in the paper.
The complete demonstration program will consist of 15 characterization
tests of short duration and longer duration performance tests.
291
-------
1.0
INTRODUCTION
ECO LOGIC has been conducting research on a method of decontaminating
hazardous wastes using a patented thermo-chemical reduction process. This
process is particularly suitable for wastes that are primarily aqueous, such as
harbour sediments, landfill leachates and lagoon sludges. The research to date
has been funded by the National Research Council Industrial Research Assistance
Program, the Defence Industrial Research Program administered by the Department
of National Defence, the Great Lakes Clean Up Fund, the Ontario Ministry of the
Environment, the Environmental Technologies Program and ECO LOGIC.
Research and development has focused on bench-scale testing of surrogate
compounds, development of a larger lab-scale destructor for testing actual waste
samples, and construction of a mobile full-scale field unit for materials and
component testing.
292
-------
2.0
BACKGROUND
There is a growing sense of awareness and concern about the state of our
environment, and the lack of appropriate ways of dealing with some of the
problems we have created. Eco Logic was formed in 1986 specifically to address
the need for a clean-up tool for one of the most difficult problems, that of
severely contaminated aqueous wastes such as harbour sediments, landfill
leachate, and lagoon sludges. The criteria that Eco Logic used in developing
the process included:
* destruction efficiency
* possibility of dioxin or furan formation
* continuous monitoring arid process control suitability
* suitability for aqueous wastes
* mobility
* cost
The patented ECO LOGIC process addresses all of these criteria. It is
based on the gas-phase thermo-chemical reaction of hydrogen with organic and
chlorinated organic compounds at elevated temperatures. At 850°C or higher,
/
hydrogen reacts with organic compounds in a process known as reduction to produce
smaller, lighter hydrocarbons. In the case of chlorinated organic compounds,
such as polychlorinated biphenyls (PCBs), the products of the reaction include
hydrogen chloride, methane and ethylene. This reaction is enhanced by the
presence of water, which can also act as a reducing agent. Bench-scale testing
with trichlorobenzene (half of a PCB molecule) has shown that the reduction
reaction will achieve 99.9999% destruction efficiency or better. The first
measure of a good destruction technology is high destruction efficiency, and the
bench-scale testing demonstrated that the ECO LOGIC process was capable of this.
Figure 1 shows the four principal reduction reactions that occur in the patented
ECO LOGIC process. The first is the dechlorination and dismantling of a PCB
molecule to produce hydrogen chloride and benzene. The second is the reduction
of benzene to produce ethylene. The third is the reduction of straight-chain
hydrocarbons to produce methane, and the fourth is the reduction of a
polyaromatic hydrocarbon (PAH) compound, phenanthrene, to produce ethylene.
293
-------
FIGURE 1
Patented Thermo-Chemical Reduction Reactions
Cl Cl
Cl Cl
5 H2
4 HCl
"N + 3 H
3 CH
E "4
n C H4
9 H
-------
The second criterion for process design was elimination of the possibility
of dioxin and furan emissions. The ECO LOGIC process is not an incineration
technology. Incineration processes destroy chlorinated organic wastes by
breaking contaminant molecules apart with high temperatures and then combining
them with oxygen, usually from air. A PCB waste would first fragment to form
chlorobenzenes, which when combined with oxygen can form dioxins and furans,
which are more toxic than the original PCBs. The ECO LOGIC process uses hydrogen
to produce a reducing atmosphere devoid of free oxygen, and thus eliminates the
possibility of dioxin or furan formation.
Other non-chlorinated hazardous organic contaminants, such as PAHs, are
also reduced to smaller, lighter hydrocarbons, primarily methane and ethylene.
Because of the tendency of the reaction to produce lighter, more volatile gases,
the process lends itself to continuous monitoring of the destruction efficiency.
This satisfied the third criterion of process design. ECO LOGIC has purchased
a very sophisticated on-line mass spectrometer system which is capable of
measuring many organic chemicals on a continuous basis. It was used for bench-
scale and lab-scale testing, and is now part of the process control system in
the full-scale demonstration unit. PCB and PAH destruction efficiencies can be
measured very quickly by continuously monitoring chlorobenzene and benzene
concentrations. The information from the mass spectrometer is sent to the
process controller so that an increase in chlorobenzene or benzene concentration
(signalling a decrease in PCB or PAH destruction efficiency) halts the input of
waste and alerts the operator.
The fourth measure of a destruction technology is its ability to process
aqueous wastes. The ECO LOGIC process is suitable for many types of waste,
including those with a high water content which are very difficult to incinerate.
For example, water contaminated with 0.1% (1000 ppm) PCBs can be processed easily
with this reaction. The presence of water in the waste actually aids in the
destruction process since water itself can act as a reducing agent to help
dismantle the contaminant molecules.
295
-------
The final criteria ECO LOGIC used in designing the process were mobility
and cost. Because the process is not an incinerator, the reactor does not
require a large volume for the addition of combustion air. The small reactor
size and the capability to recirculate product gases from the reaction make the
process equipment snail enough to be mobile. As well, the smaller size reduces
the capital cost of the process equipment. The main processing costs are for
hydrogen, electricity, and personnel. A commercial-scale system processing 100
tonnes per day should be capable of operating for a- price of approximately $500
per tonne of waste processed depending on the organic strength of the waste.
This is roughly equivalent to the cost of long-term entombment, with the
advantage of actually eliminating the problem.
Figure 2 shows a schematic of the reactor designed to accommodate the
thermo-chemical reduction reaction. A mixture of preheated waste and hydrogen
is injected through nozzles mounted tangentially near the top of the reactor.
The mixture swirls around a central ceramic tube past glo-bar heaters and is
heated to 850°C by the time it passes through the ports at the bottom of the
ceramic tube. Particulate matter up to 5 mm diameter not entrained in the gas
stream will impact the hot refractory walls of the reactor, thereby volatilizing
any organic matter associated with the particulate. That particulate will exit
out of the reactor bottom to a quench tank, while finer particulate entrained
in the gas stream will flow up the ceramic tube into the exit elbow and through
the retention zone. The reduction reaction takes place from the bottom of the
ceramic tube onwards, and takes less than one second to come to completion.
Figure 3 shows a complete process schematic of the field demonstration
unit, which is now operating at Hamilton harbour. Waste liquid and suspended
solids are pumped fron a small storage tank (W) to a heat exchanger vessel (PHV)
for preheating to 150°C by a small boiler (B). Hot liquid and steam from the
watery waste are metered continuously using special metering valves (MV) and
injected into the reactor (R) using atomizing nozzles. A mixture of hydrogen
(H^) and recirculation gas (RG) also enters the reactor near the top after
passing through a gas-fired heat exchanger (RH).
296
-------
FIGURE 2
THERMD-CHEMICAL REDUCTION REACTDR
REFRACTORY
GLDBAR HEATER
CERAMIC TUBE
297
-------
FIGURE 3
PROCESS SCHEMATIC
RH
RG
H2
ro
VO
00
V
-©-
\
R
SL
DW
MV
MV
GB
HX
RG
EC
HC
n SG
PHV
B
-------
Heavy particulate exits as grit (G) out the bottom to a quench tank and
fine particulate and gases pass up the ceramic tube where the gas-phase reduction
reaction takes place. Additional residence time is provided by the retention
zone elbow and extension pipe prior to the scrubber. Once the gases enter the
scrubber (S), they are quenched by direct injection of scrubber water spray.
Hydrogen chloride and fine particulate are removed by contact with the scrubber
water as the gases pass through the scrubber media, which is carbon steel on the
down leg and polypropylene on the up leg of the scrubber. Scrubber water is
collected in a tank below via a large water-sealed vent that also acts as an
emergency pressure relief duct. The scrubber water is cooled to 35°C using a
heat exchanger fed by cooling water from an evaporative cooler (EC). Sludge (SL)
and decant water *(DW) are the two effluent streams from the scrubber and both
are held in tanks for batch analysis prior to disposal.
The gases that exit the scrubber consist only of excess hydrogen, reduction
products such as methane and ethylene, and a small amount of water vapour.
Approximately 95% of this gas is recirculated back into the reactor after
reheating to 500°C, and about 5% of the hydrocarbon-rich gas (HC) is used as
supplementary fuel in the boiler. The boiler uses propane (P) as its main fuel
to produce steam used in the heat exchanger which preheats the waste to 150°C.
The only air emissions are from the boiler in the form of stack gas (SG). Since
the fuel going into the boiler is very clean, and contains no chlorine, emissions
from the boiler should be insignificant.
In the case of a process upset where total destruction of hazardous organic
compounds was not occurring, the on-line mass spectrometer automatically diverts
all gases into a recirculation mode. No sidestream gas (HC) is sent to the
boiler and the waste feed is stopped. Recirculation continues until the
continuous analysis indicates the reaction is again occurring optimally. During
this time, the scrubber water may become contaminated and require treatment
itself, but no escape of or incineration of chlorinated organic compounds occurs.
Since 95% of the gas stream is recirculated under normal conditions, procedure
is not a drastic action.
299
-------
FIGURE 4
HAMILTON HARBOUR LAYOUT
GENERATOR
WATER
PROPANE
H2
TUBE TRAILER
12n X lOn SPILL PROTECTION
PROCESS CONTROL TRAILER
LAB
TRAILER
PERIMETER FENCE
60 n-
CO
o
o
THE EQUIPMENT DESCRIBED IN THIS TEXT IS LOCATED AT HAMILTON HARBOUR, AND IS OPERATING AND
SCHEDULED TO COMPLETE THE DEMONSTRATION TESTING ON ACTUAL HARBOUR SEDIMENT BY AUGUST 1, 1991.
THE PRCOESS CONTROL TRAILER CONTAINING THE ON-LINE MASS SPECTROMETER, PROCESS CONTROL SYSTEM,
AND OTHER ANALYSIS EQUIPMENT IS LOCATED ADJACENT TO THE TWO PROCESS TRAILERS, AND IS SITUATED
NEXT TO THE LAB TRAILER.
-------
Wastes Tested To Date
1) Pure Compounds (Bench Scale)
PCB askarel (Aroclor 1254)
Hexachlorobenzene
Trichlorobenze
24-D
Methoxychlor
2) Environmental Wastes (Laboratory Scale)
(1 Kg/hr)
Hamilton Harbour Sediment
(coal tar, PAH, PCBs, TCB, Fe, Zn,
PCDD/PCDF)
Thunder Bay Harbour Sediment (chlorophenols, PCDD/PCDF, wood treatment
waste)
Sheboygan Harbour Sediment (0.3% PCBs)
(Audited Program by Environment Canada)
3) Full Scale Testing
(2-5 kg/min)
Hamilton Harbour Sediment
(3-7 T/day coal tar, PAH, PCBs, PCDD/PCDF,
Fe, Zn)
(Audited Program by Environment Canada and the Ontario
Ministry of the Environment)
301
-------
3.0 GENERAL STRATEGY
ECO LOGIC will enter the market to supply hazardous waste destruction
services itself with its own machines, and to licence and sell, equipment to
conpanies that already supply services, or to licence and sell to large chemical
producers or users where ownership is economically advantageous. The service
sarket is the initial focus in order to demonstrate the machine and obtain
approvals in jurisdictions where units will be licensed and sold when the
technology has buyer (versus user) acceptance.
We would encourage interested parties with organic hazardous waste problems
to contact our offices in Rockwood and Ann Arbor.
This offers potential clients the opportunity to obtain direct information
on the application of this technology to the resolution of their hazardous waste
problems.
i
For more information, please do not hesitate to contact:
Wayland R. Swain, Ph.D.
Vice President
U.S. Operations
ELI Eco Logic International Inc
2395 Huron Parkway
Ann Arbor, Michigan U.S.A. 48104
313-973-2780
Jim Nash, B.A.
Manager
Sales & Business Development
ELI Eco Logic International Inc
143 Dennis Street
Rockwood, Ontario Canada NOB 2KO
519-856-9591
302
-------
4.0
WASTES SUITABLE FOR DESTRUCTION
CHEMICALS:
Non-halogenated / halogenated biphenyls
Non-halogenated / halogenated benzenes
Non-halogenated / halogenated phenols
Non-halogenated / halogenated cycloalkanes
Non-halogenated / halogenated alkanes
Non-halogenated / halogenated dioxins
Non-halogenated / halogenated dibenzofurans
Polyaromatic hydrocarbons
* Note: Halogenated means:
Chlorinated
Brominated
Fluorinated
TYPICAL WASTES;
PCBs
Pulp mill wastes
Chlorinated solvent waste
Contaminated coal tars
Solvent still bottoms
Chlorophenols / Wood treatment waste
Pesticide wastes
Landfill leachates
Lagoon bottoms
303
-------
RESEARCH
DETERMINING THE APPLICABILITY
OF X*TRAX™ FOR ON-SITE REMEDIATION
OF SOIL CONTAMINATED WITH
ORGANIC COMPOUNDS
Presented at
EPA - Third Forum on Innovative
Hazardous Waste Treatment Technologies:
Domestic and International
The Fairmont Hotel
Dallas, Texas
June 11-13, 1991
Presented by
Carl Swanstrom
Chemical Waste Management, Inc.
Geneva Research Center
1950 S. Batavia Avenue
Geneva, IL 60134
304
-------
Determining The Applicability Of X*TRAX™
For On-Site Remediation of Soil
Contaminated With Organic Compounds
by
Carl Swanstrom
Senior Project Manager
Chemical Waste Management
1950 S. Batavia Avenue
Geneva, IL 60134-3310
Phone: 708/513-4500
Fax: 708/513-6401
1.0 INTRODUCTION
The market for soil treatment technologies is expanding rapidly. One
source estimates the contaminated soils market to be $200-300 billion in
the next 30 to 40 years.1 This market expansion is driven by at least
three factors. First is the Superfund program, of which contaminated soils
comprise the majority of wastes requiring remediation of hundreds of sites
across the country. A second*driving force is the "landbans" (40 CFR Part
268) which will prohibit the landfilling of many contaminated soils
because of their organic content. The third driving force is the recent
legislation by many states requiring that real estate must be certified as
non-contaminated before the transfer takes place.
Many technologies can potentially be used to treat organically-contaminated
soils. These include solidification/stabilization, bioremediation, soil
washing, in situ vacuum extraction, solvent extraction, thermal desorption
(stripping), in situ and ex situ vitrification, incineration and others.
This paper will focus on the types of contaminated soils that are likely
candidates for low temperature thermal desorption. The data presented and
conclusions drawn apply only to the patented X*TRAX™ process developed by
Chemical Waste Management, Inc. (CWM).
2.0 PROCESS DESCRIPTION
Chemical Waste Management initiated development of its low temperature
thermal treatment process, X*TRAX, in 1987 after laboratory testing by CWM
had shown that at the relatively low temperatures of 250-450°C (475-850°F)
many organic compounds, including high boiling compounds (PCBs), could be
successfully separated from solids such as soils and sand. Since then, a
process has been defined and scaled up to a 115 metric tons per day
commercial unit. U.S. Patent No. 4,864,942 has been granted for the
process.
305
-------
The process can handle soils and dewatered solids such as pond sludge and
filter cakes. Organic contaminants can range from high boiling, semi-
volatile compounds such as PCBs, to low boiling, volatile compounds such as
RCRA regulated solvents.
In the course of the development effort, treatment systems have been built
at three distinct sizes, termed laboratory scale, pilot scale, and
commercial. The two laboratory units are used for performing bench-scale
treatability studies. One unit is operated by a CWM company called Chem-
Nuclear Systems, Inc. (CNSI) for performing treatability studies on mixed
(RCRA and radioactive) wastes. The other unit is operated by CWM's
Research and Development Division located in Geneva, IL. This unit is used
for RCRA and TSCA materials and is fully permitted. The pilot scale system
was used to confirm the design parameters and is now used as a
demonstration unit. The first commercial unit has been completed and will
be moved to a Superfund site during the second quarter of 1991.
The X*TRAX process uses an indirectly heated rotary dryer to volatilize the
water and organic contaminants in a sealed system. The hot treated solids
are cooled and dedusted using the condensed water removed from the feed.
An inert carrier gas (nitrogen) transports the volatilized components to a
gas treatment train. The gas treatment train removes the entrained
particulate solids with a scrubber and then cools the entire gas stream to
less than 5°C (40°F) to condense the volatilized organics. Ninety to
ninety-five percent of the carrier gas is reheated to 315°C (600°F) and
recycled to the dryer". The remainder of the carrier gas passes through a 2
micron filter and a carbon adsorption system before being discharged to the
atmosphere. The condensed liquid organic removed from the soil is disposed
off-site. Detailed descriptions of the X*TRAX systems have been presented
in previous papers.2-3
3.0 SOIL CHARACTERISTICS
Contaminated soil at a remediation site is rarely homogeneous and
frequently contains solid debris. Naturally occurring sandy locations may"
have areas of high clay content and vice-versa. On-site remediations may
require as little as the top six inches tq be treated, or as much as ten to
twenty feet may have to be excavated. It is possible to encounter rich
topsoil, clay, and sand when excavating only a few feet. In addition, many
remediation sites have areas of engineered fill that may contain compacted
clay, sand, gravel, and even building rubble. Suitable ex-situ on-site
treatment technologies will ideally handle all of the variabilities in soil
condition and makeup that may be encountered after the oversize material is
removed or crushed.
Virtually all ex situ technologies will have a maximum particle size that
can be fed to the process. The maximum size generally ranges from 1/2" to
3". Before pretreatment can begin, any underground tanks, vaults, or
containers must be removed. In locations where the soil is primarily sand,
a simple screening operation will usually suffice. The X*TRAX process
requires screening to less than two inches. When clays are encountered, a
306
-------
soil shredder may be required. This will depend on the type of clay and
its moisture content. In either case, the screened material should be
stored under a canopy to prevent rainfall from increasing the moisture
content and to reduce the amount of contaminated water that must be
handled.
When very large rocks or building rubble is encountered, three options may
be available. The most desirable solution is to return the oversize
material to the excavation. In many cases, the large rocks or cement
pieces do not contain a significant amount of organic contamination/ though
testing on a site-specific basis is usually necessary in order to confirm
this. If this approach is not acceptable to the regulatory agencies, then
the oversize material can be transported off-site to a suitable landfill or
crushed on-site and treated. The last two options are highly site-
specific.. '
The soil type has a significant effect on process rates and attainable
treatment levels. The presence of clay or naturally-occurring humic
material will increase the adsorptive capacity of soil.4 Experience has
shown that soils with very high clay content are more difficult to treat
than soils comprised mainly of sand or silt. An example is shown in Table
1. Both of the soils from Illinois and South Carolina had the appearance
and texture of a high clay content soil. The laboratory soil
characterization results have not been received yet.
Table 1. PCS Removal for Different Soil Types
Sample
SCH32
SAW
SAH
Soil Location
Illinois
South Carolina
South Carolina
%
Moist
26
24
14
Initial
PCS
Concentration
2,900 (1)
3,760 (2)
2,960 (3)
Treated
Soil
Concentration
(mg/kg)
7.0
164
170
%
Removal
99.8
95.6
94.3
Notes: (1) Primarily Aroclors 1248 & 1254
(2) Primarily Aroclors 1248 & 1254
(3) Primarily Aroclor 1248
All three soil samples referenced in Table 1 were processed through the
laboratory X*TRAX unit under similar conditions. The variability in
residual PCB levels in the treated soils dramatically confirms the need for
performing treatability studies. Both soils had similar moisture contents,
PCB concentrations and PCB Aroclors, yet the Illinois soil could be treated
to less than 10 mg/kg, while the two samples from South Carolina could not
be treated to less than 100 mg/kg.
307
-------
Soils rich in humic material, such as topsoil, should be avoided where
possible. In addition to the previously mentioned increase in soil
adsorptivity, many of the organic acids driven off during the thermal
desorption are water soluble. The condensed water from the process
requires substantial treatment before reuse in the system or discharge. If
a relatively thin layer of ;topsoil is present, it can be blended with the.
underlying soil to reduce the humic content of the feed. Any vegetative
cover should be removed before excavation.
4.0 CONTAMINANTS
Contaminated soils seldom contain only a single contaminant. It is not
uncommon to have more than four principle organic hazardous constituents
(POHCs) along with numerous other organic compounds at measurable levels.
The following sections discuss the treatability of five broad categories of,
contaminants, including metals.
4.1 PCBs .
The X*TRAX thermal desorption process was designed to be capable of
removing PCBs from soil to levels that would allow the soil to be left on-
site. Since there are not many organic compounds with vapor pressures
lower than the common PCB Aroclors, a process that effectively treats PCBs
should be suitable for most any organic contaminant.
The ultimate goal was to be able to reduce the PCB level in any soil to
less than 2 mg/kg, which is a treatment level often imposed by EPA. For
soils with more than 100 mg/kg PCBs, this goal.has not yet been attained.
Except for a few isolated cases, the X*TRAX process has been able to reduce
the PCB level to less than 25 mg/kg and frequently to below 10 mg/kg when
the PCB concentration in the feed materials tested ranged from 10Q to 7,800
mg/kg. Treatment levels of 10 mg/kg or 25 mg/kg are consistent with the
latest EPA guidance document for remedial actions at Superfund sites with
PCB contamination.5 .
Treatability studies on over sixteen samples from ten PCB-contaminated
sites have been performed using the laboratory X*TRAX unit. The data from
several of these treatability studies are presented in Table 2. The data
show that at least a 95% reduction can be obtained when the contaminated
soil contains greater than 100 mg/kg of PCBs. These treatment levels are
consistent with the Alternate Treatability Variance Levels for PCBs
presented in the EPA's Superfund LDR Guide #6A.6 For PCB concentrations
greater than 100 ppm, a 90-99.9% reduction is recommended. The X*TRAX
process can- certainly meet this requirement. It is interesting to note
that this document does not list low temperature stripping (desorption) as
a technology capable of achieving the guidelines.
The pilot X*TRAX system has processed ten different PCB contaminated soils
under a one-time U.S. EPA TSCA R&D permit. The results from the pilot
testing are summarized in Table 3.
308
-------
Table 2. Laboratory X*TRAX Results - PCB Contaminated Soils
Sample ID
SCH17
PB
SAD
GMB
JDR
PRY
GER
SCH32
SAH
SAW
Feed PCB
Concentration
(mg/kg)
4.6
5.2
12
150
330
770
805
2,900
2,960
3,760
Product PCB
Concentration
(mg/kg)
0.94
2.0
BDL (1.0)
3.2
BDL (5.0)
12
17 ,
7.0
170
K 164
%
Reduction
79.6
61.5
> 91.7
97.9
> 98.5
98.4
97.9
99.8
94.3
95.6
Table 3. Pilot X*TRAX Results - PCBs
Run ID
0921B
0124
0914
0929
0926
0727
1003
0810
0919
0122
Amount
Processed
(Ibs)
2,145
4,205
2,820
2,240
1,830
4,940
2,790
4,570
4,530
4,750
Feed PCB
Cone .
(mg/kg)
68
120
190
630
640
1,480
1,600
2,800
5,000
7,800
Product
PCB Cone .
(mg/kg)
16
3.4
9.6
17
18
8.7
4.8
19
50
24
%
Reduction
76.5
97.2
95.0
97.3
97.2
99.1
99.7
99.3
99.0
99.7
309
-------
Again, we see that at least 95% removal can be obtained when the PCS level
is greater than 100 rag/kg. We also see that even with 7,800 rag/kg PCBs in
the feed the treated soil had less than 25 mg/kg PCBs.
4.2 Pesticides
Only one pesticide contaminated soil has been tested to date. We
anticipate performing two more studies in the; near future. One will be at
the laboratory scale and the other at both the laboratory and pilot scale.
The results from the completed laboratory treatability study are presented
in Table 4. For all identified pesticides at least a 97% reduction was
achieved. The TCLP was also performed on the treated soil to assess
whether it exhibited a toxic organic characteristic (new D codes). These
results are presented in Table 5.
Table 4. Laboratory X*TRAX Results
Pesticide Contaminated Soil
Contaminant
4,4' -DDE
4,4' -ODD
Alpha-Chlordane
Gamma-Chlordane
Feed
Concentration
(ppm)
32
320
100
110
Product
Concentration
(ppm)
0.57
1.3
2.1
3.0
%
Removal
98.2
99.6
97.9'
97.3
Table 5. Comparison of Treated Soil to
Toxicity Characteristic Regulatory Level
EPA HW
NO.
D020
D031
D013
Contaminant
Chlordane
Heptachlor
Lindane
Treated
Soil
(ppm)
5.1
< 0.03
< 0.03
Treated
Soil TCLP
(mg/L)
< .0025
< .00025
< .00025
Regulatory
Level
(mg/L)
0.03
0.008
0.4
The TCLP concentrations for the three listed pesticides are well below the
regulatory level. Although only one test has been completed, the data
indicate that the X*TRAX process is suitable for certain pesticide-
contaminated soils.
310
-------
4.3 Semi-Volatile Oraanics
The samples tested at both the pilot and laboratory scale have all had
relatively low levels of semi-volatile organic compounds. The results for
three different compounds are presented in Table 6. Sample IDs that are
letters are laboratory scale tests, and the number IDs are from pilot
tests. When the concentration in the treated soil was measurable, the
percentage reduction was greater than 95%. The data indicate that most
semi-volatile organic compounds can be reduced to less than 10 ppm and
frequently below l ppm.
Table 6. Laboratory & Pilot X*TRAX Results
Semi-Volatile Organics
Compound
Bis(2-
Ethylhexyl)
Phthalate
Phenanthrene
Naphthalene
ID
SAW
SAWR
1003
0727
PB
MC
MBL
PB
MBL
MC
Feed
(ppm)
3.2
3.9
3.4
9.1
14
19
30
34
110
450
Product
(ppm)
BDL (.33)
BDL (.33)
0.30
0.18
BDL (.33)
0.29
BDL (13)
0.74
BDL (13)
7.9
%
Reduction
> 89.7
> 91.5
91.2
98,. 0
> 97.6
98.5
> 56.7
97.8
> 88.2
98.2
4.3 Volatile Oraanics
As was the case for semi-volatile organics, very few soil samples have been
received that contain significant quantities of volatile organics. It is
not surprising that the most often detected volatile organics were the
BETXs (benzene, ethylbenzene, toluene and xylene). Several examples of
removal rates are presented in Table 7. In most cases the semi-volatile
organics were reduced to well below 1 ppm.
It is interesting to note that the highest volatile organic contamination
levels were in the feed sample labeled SSM-1/5. The material SSM
(Synthetic Soil Matrix) is a surrogate Superfund soil created for the EPA
for treatability studies.7 CWM did not participate in the original study
since it was initiated before the laboratory X*TRAX system became
operational. A sample of the Type I soil was obtained from the contractors
archive to be used for CWM's own testing.
311
-------
Table 7. Laboratory and Pilot X*TRAX Results
Volatile Organics
Compound
Benzene
Ethylbenzene
Tetrachloroethylene
Xylene
ID
PB
MC
MBL
TFC
BL9
PB
MC
MBL
SSM-1/5
SSM-1/5
PC
SAHK
PB
BP
MC
MBL
SSM-1/5
Feed
(ppm)
0.35
3.0
6.9
30
980
0.40
13
50
1,600
150
0.22
1.6
3.8
7.2
77
130
2,400
Product
(ppm)
BDL (.05)
BDL (0.10)
1.6
£DL (0.13)
BDL (0.21)
BDL (.05)
BDL (0.10)
0.33
5.2
.094
.030
.043
BDL (.050)
BDL (.095)
BDL (0.10)
0.84
9.5
%
Reduction
> 85.7
> 96.7
76.8
> 99.6
> 99.98
> 87.5
> 99.2
99.3
99.8
99.9
86.4
97.3
> 98.7
> 98.7
> 99.9
99.4
99.6
Based on the contaminated soil samples received by CWM for testing, it
appears that the actual or average organic contamination levels at many
sites may be much lower than many have predicted. Consider a site with a
maximum compound X concentration of 1,000 ppm that must be excavated to 10
ppm. Assuming a linear concentration, the average contaminant
concentration will be 505 ppm. Since the contaminant concentration level
from the source is rarely linear, the actual average concentration could be
as low as 50 to 100 ppm. This should be considered when collecting samples
for treatability studies. Samples from known "hot spots" should be
avoided.
The sample ID BL9 in Table 7 was high in benzene. This sample was actually
a pond sludge. In an actual remediation, one would not process the sludge
alone, but it would be excavated and processed with the surrounding soil.
This would allow for a fairly consistent feed material throughout the
remediation. It does not make sense to optimize a process for 1,000 ppm of
a POHC and then readjust for a feed with 10 to 50 ppm.
4.5 Metals
With the exception of mercury, the X*TRAX process will do little to remove
non-volatile metals (W10) or volatile metals (Wll) from contaminated soils.
This should not preclude the evaluation of X*TRAX when metals contamination
is significant. One of the most common treatments for metals is
312
-------
stabilization. Most stabilization treatments involve a pozzolanic reaction
using cement, kiln dust or other materials. Organics interfere with the
pozzolanic reaction. Also, stabilized samples frequently fail because of
organic leaching, not metals leaching.
The treated solids exiting the X*TRAX dryer are conveyed to a mixer, where
condensed water from the process is used to cool and dedust the treated
soil. Normally the solids will exit the mixer with 5 to 15% moisture.
Since the mixing hardware already exists, stabilization can also be
performed by adding a reagent silo and a conveyor prior to the mixer.
Since the majority of the organic compounds that interfere with the
stabilization chemistry have been removed the reagent usage will be very
low.
When metals levels are high, incineration may not be an acceptable
treatment technology. Hazardous waste incinerators normally limit the
metals in the feed in order to stay within the constraints of their air
permit. At the low temperature that X*TRAX operates, the metals are not
volatilized (except for mercury). Also, the process vent gas is filtered
through a 2 micron filter before venting to atmosphere which will remove
any particulate solids that may contain metals.
4.6 Other
In addition to contaminated soils, the X*TRAX process has been evaluated
for one RCRA waste. Filter cakes from refinery operations carrying waste
codes K048 to K050 have been tested. Two laboratory studies have been
completed and a pilot test is scheduled for the second quarter of 1991.
The treated solids.easily met all of the applicable BDAT requirements. The
oil recovered from the phase separator had an energy content of 18,000 Btu
per pound, and the condensed oil had an energy content of 15,000 Btu per
pound. The filter cake from the phase separator blowdown amounted to
approximately 5%'of the feed and had an energy content of 8,700 Btu per
pound.
5.0 SUMMARY
The CWM-developed X*TRAX process has been shown to be a suitable
alternative treatment technology for a wide variety of organically
contaminated soils. Successful treatability studies have been performed on
soils and sludges contaminated with the following categories of
contaminants:
W01: - Halogenated Non-Polar Aromatic Compounds
W02 - Dioxins/Furans, PCBs, and their Precursors
W03 - Halogenated Phenols, Cresols and Other Polar Aromatics
WO4 - Halogenated Aliphatic Compounds
W05 - Halogenated Cyclic Aliphatics, Ethers, Esters, Ketones
W07 - Simple Non-Polar Aromatics and Heterocyclics
W08 - Polynuclear Aromatics
W09 - Other Non-Halogenated Polar Organic Compounds
313
-------
The W06 (Nitrated Aromatic and Aliphatic Compounds) were not included in
the list only because no soil samples containing any of these compounds
have been tested. PCBs were included, since most soils can be treated to
less than 25 ppm and in many cases to less than 10 ppm.
Organically-contaminated soils that also contain metals may be suitable for
treatment by X*TRAX with stabilization of the treated soil. The additional
cost for stabilization will be small, since only minimal additional
hardware is required and reagent usage will be low.
Due to the complexity and variability of contaminated soils, treatability
studies are recommended. The laboratory X*TRAX system can accurately
determine treatment levels. Samples for treatability studies should not
usually be collected from the "hot spots." The author can be contacted for
information on arranging a laboratory treatability study.
*
References
1. "Industry Tests New Technologies for Soil Cleaning," Environmental
Business Journal Vol. Ill, No. 2, February 1990.
2. Swanstrom, C., Palmer, C., "X*TRAX™ Transportable Thermal Separator
for Solids contaminated with Organics,!1 presented at the Air and
Waste Management Association International Symposium, Cincinnati,
Ohio, February 5-8, 1990.
3. Swanstrom, C., Palmer, C., "X*TRAX™ Transportable Thermal Separator
for Organic Contaminated Solids," presented at the Second Forum on
Innovative Hazardous Waste Treatment Technologies: Domestic and
International, Philadelphia, Pennsylvania, May 15-17, 1990.
4. Dragun, James, PhD., "The Soil Chemistry of Hazardous Materials,"
Hazardous Materials Control Research Institute, Silver Springs,
Maryland 1988.
5. U.S. EPA "Guidance on Remedial Actions for Superfund Sites with PCB
Contamination" EPA/540/6-90/007. Office of Emergency And Remedial
Response, Washington, DC, 20460, August 1990.
6. U.S. EPA "Superfund LDR Guide #6A - Obtaining a Soil and Debris
Treatability Variance for Remedial Actions." Office of Solid Waste
and Emergency Response Directive 9347.3-O6FS, July 1989.
7. Esposito, Pat et al. "Results of Treatment Evaluations of a
Contaminated Synthetic Soil," JAPCA Vol. 39, No. 3, March 1989.
314
-------
Paper presented at the Third Forum on Innovative Hazardous Waste Treatment
Technologies: Domestic and International, June 11-13, 1991. Dallas, Texas, USA.
Removal of Arsenic and/or Other Amphoteric Elements
from Sludge and Solid Waste Materials
A.N. van Breemen
Faculty of Civil Engineering, University of Technology
P.O. Box 5048, 2600 GA Delft, The Netherlands
Introduction
One of the major challenges in environmental engineering nowadays is a well-
considered reuse of hazardous waste. Reuse of waste is most often blocked by the
presence of certain constituents endangering the new environment involved or
inhibiting a proper application. Development of waste treatment methods aiming at
reuse often results in a method which can have a much wider application than the
original objective.
The method presented here was originally developed for the removal of arsenic out
of water treatment sludge often dumped in the environment. Due to new legislation
in the Netherlands reuse of arsenic bearing waterworks sludges exceeding the 50 mg
per kg dry solids value has been forbidden. Controlled dumping of these wastes is
getting much more expensive. Good reasons to invest more time in the removal of
arsenic out of arsenic-bearing sludges.
For the removal of arsenic compounds from arsenic-bearing waterworks sludges a
method has been developed based upon the amphoteric properties of the element
arsenic. Due to the presence of more amphoteric elements in solid wastes the method
is multi-purpose. Elements like phosphorus, aluminium, chromium, vanadium and
antimony behave in the same way. As example of application the research reported
here has been focussed on the removal of arsenic from iron hydroxide sludges. The
arsenic present in these sludges originates both from natural and industrial sources.
Concentration values exceeding the 1000 mg arsenic per kg dry solids have been
registered. 315
-------
In table 1 the reactions between arsenate ions and ferric hydroxide are presented,
showing a strong bond between the iron hydroxide matrix and the arsenate ions.
Table 1. Reactions between arsenate ions and ferric hydroxide.
4 < pH < 7
2FeOH2+
H2AsO4~
FeO O
V
As
/ \
FeO OH
2 H,0
pH « 8
2FeOH
HAsO4
2~
FeO O
Y'
As
/ \
FeO O-
HO
8 < pH < 10
FeO O
FeOYFeOH + H AsO4:
2—
FeO
As
\
2OH-
Essential in waste reduction or recycling/reuse studies is the avoidance or limitation
of additional waste production either in a dissolved or gaseous form. The impurities
themselves must be concentrated as good as possible. The method for the removal of
arsenic out of sludges holds this principle.
Occurrence of arsenic in sludges and sediments
In literature only scattered information about the occurrence of arsenic in sludges and
sediments is available. Up to now the presence of minor amounts of arsenic has not
been considered as a problem.
The mobility of arsenic being a toxic element in the environment deserves particular
attention. A lot of studies on arsenic nowadays has been focussed on that aspect next
to the removal of arsenic from water.
316
-------
Table 2. Concentrations of arsenic in natural waters.
Type of water
Rain
Lakes
Rivers
Groundwater
Sea
Concentration (/xg/1)
1
1-60
1 - 20
1-30
1- 7
In the Netherlands the introduction of the Chemical Waste Act resulted in a survey of
limiting values of concentrations of toxic metals and metalloids. A selection is
presented in table 3.
Table 3. Some limiting values of concentrations of toxic metals and metalloids (in
mg per kg dry solids) as formulated in the Dutch Chemical Waste Act
Concentration value
(mg/kg dry solids)
Element
50
5,000
20,000
Arsenic (As), cadmium (Cd),
mercury (Hg), antimony (Sb),
selenium (Se), tellurium (Te)
Cobalt (Co), chromium (Cr),
copper (Cu), lead (Pb),
vanadium (V), wolfram (W)
Zinc (Zn), barium (Ba)
So if the arsenic content of the sludge exceeds the 50 mg per kg dry solids value the
sludge involved is considered as hazardous waste according the Dutch Chemical
Waste Act. Reuse of that sludge is not possible whereas dumping is getting much
more expensive.
317
-------
In figure 1 the cumulative frequency distribution of the arsenic content of ground and
surface water in the Netherlands is presented. It clearly indicates that more than half
of the waterworks has to deal with hazardous sludge. Given the relatively low
concentrations of arsenic in ground and surface water (see table 1) the figures clearly
demonstrate the high affinity of arsenate ions for ferric hydroxide. Next to the
concentration of arsenic in water the Fe/As-ratio is decisive for whether the resulting
sludge falls under the Dutch Chemical Waste Act or not.
Figure 1. Cumulative frequency distribution of the arsenic content of ground and
surface water in the Netherlands.
Water Works %
100-
75-
50-
25-
0
Surface Water
Groundwater
Limiting Value
0 50 100 200 300 400 500 600
Concentration of Arsenic (rng/kg da)
Methods for the removal of arsenic out of sludges
Some methods have been reported in litterature concerning the possible removal of
arsenic out of solid waste especially when the analogy as amphoteric element with
phosphorus is taken into account (1-10). Most of them are quite complicated,
moreover resulting in additional voluminous liquid waste. The method of removal
applied here is based upon the amphoteric properties of arsenic. In presence of an
alkali arsenic oxide behaves as acidic compound resulting in the formation of soluble
arsenate compounds.
318
-------
Basic reactions for the dissolution of arsenic at high temperature from the solid
matrix in presence of soda are:
Na,CO,
co
2 Fe(OH)3 . Fe AsO4
2 Fe2O3 + As2O5 + 3H2O
As2O5 + 3 Na2O
or as overall reaction:
2 Fe(OH)3.FeAsO4 + 3 Na2CO3
2 Fe2O3 + 2 Na3AsO4 + 3 CO2 + 3H2O
Due to the thermal treatment at 800 - 900°C in an oxic environment organic matter
present in the sludge is completely oxidized. The formation of sodium arsenate is a
matter of discussion due to the instability of arsenic pentoxide at high temperatures.
Conversion of arsenic pentoxide into arsenic trioxide is most likely, but has not been
proven yet here.
Other amphoterics like aluminium, chromium, phosphorus, antimony and vanadium
behave in the same way and will coprecipitate.
Sodium arsenate (arsenite) can be easily extracted from the solid phase after heat
treatment resulting in a dilute solution. Precipitation of the dissolved arsenate with
lime in a so-called pellet reactor results in a very condensed form of calcium arsenate
(pellets).
The advantage of the method applied is that the chemicals involved are quite
common and easy to apply. Moreover the liquid phases are fit for reuse as alkaline
reagents in water treatment.
A more direct approach for the removal of arsenate out of sludge is leaching with
caustic soda. The arsenate present will be exchanged for hydroxyl ions leading to a
substantial lowering up to complete removal of arsenate out of the iron hydroxide
matrix. A disadvantage here is that usually organic matter is leached out too resulting
in a rather polluted alkaline solution of arsenates and other amphoterics which is
more difficult to treat.
319
-------
Removal of arsenate ions from waterworks sludge
The goal of the treatment of waterworks sludge originating from surface water
treatment was reuse of the purified sludge (mainly ferric hydroxide) as coagulant.
Next to the removal of arsenic the removal of the organic matter from the sludge was
essential. Thermochemical treatment of the arsenic-bearing iron hydroxide at 850°C in
presence of soda was preferred. Moreover while taking treatment at 850°C water
phases being set free during the treatment are in principle recyclable.
As for composition and dewatering behaviour sludge originating from groundwater
treatment is principally different from surface water treatment sludge.
For the evaluation of the method a typical iron-rich groundwater and surface water
treatment sludge were taken. A dosage dry solids/soda ratio (on weight basis) of 5 : 1
results in a reduction in the arsenic content of the solids of more than 90%. (see
table 4).
Table 4. Removal of arsenit from different type of solid waste resulting from
groundwater treatment.
Arsenic (mg As/kg d.s.)
Fresh precipitate
Dump (age > 1 year)
Filtering material
(covered by iron oxides)
original
302
296
580
after thermochemical
treatment
11
20
28
The method of removal of arsenic from surface water treatment sludge is identical
with that for groundwater sludge: same process conditions and process efficiency.
(more than 90%). Decomposition of soda into sodium oxide proved to be essential: at
temperatures substantially lower than 800 -.900°C (temperature of decomposition for
soda) the removal efficiency for arsenic is much lower.
In order to gain more insight in the processes going on during the thermal treatment
of the sludges thermogravimetric analysis was carried out for the groundwater
treatment sludge and the surface water treatment (coagulation sludge).
In figure 2 thermogravimetric analysis (TG/DTG) of dried groundwater sludge with
and without soda addition is presented. In figure 3 the same information on surface
water treatment sludge.
320
-------
Figure 2. Thennogravimetric analysis (TG/DTG) of dried (105°C) groundwater
sludge with and without soda addition.
roundwater sludge + soda
(5:1)
groundwater sludge + soda
(1:1)
40.
200
400
600
800
1000
1200
TGA
soda
groundwater sludge
groundwater sludge + soda^
(5:1)
groundwater sludge + soda
^' :' ' y temperature (~C)
H , 1 ^—« . , f H-
0,0
-0,1
-0,2
0,0
-0,2
200
400
600
800
1000'
1200
DTGA
321
-------
Figure 3. Thennogravimetric analysis (TG/DTG) of dried (105°C) coagulation
sludge with and without soda addition.
coagulation sludge -f soda (5:1)
40
200
400 600
800
1000 1200
TGA
0,0
-0,1
-0,2
weight
loss
rate,
0,0
-0,1 •
-0,2-
• V/
soda
coagulation sludge
coagulation sludge + soda (5:1)
tempera
ture (°cy
0,0
-0,1
-0,2
200 400 600 800 1000 1200
DTGA
322
-------
Thermogravimetric analysis shows a fixed residue for the groundwater and
coagulation sludge of 31.5% and 29.5% respectively. The thermogravimetric curve for
soda exhibits a loss of weight of 3.1% between 40°C and 150°C, mainly due to the loss
of adsorbed water and a substantial loss of weight after 830°C due to the decomposi-
tion of soda and volatilization of the sodium oxide formed.
In presence of soda a remarkable loss of weight appears for both sludges at
temperatures between 750 and 950°C. Comparison of the soda curve and the sludge
+ soda curve shows an influence of the matrix on the decomposition of the soda.
There is clear evidence that the sodium oxide formed also reacts with ferric oxide. At
dosage ratio (one weight part soda - five weight parts dried sludge) sublimation of
sodium oxide does not occur. At higher ratios sublimation of sodium oxide appears.
At lower ratios all the sodium oxide interacts with the arsenic oxide and other
amphoterics and with iron oxide so volatilization is less likely.
Conclusions
A method has been developed on lab.scale to remove amphoteric elements like
arsenic, phosphorus, chromium and aluminium out of sludges by thermochemical
treatment with soda.
The amphoterics present after thermochemical treatment as sodium salts can be
extracted and separated as their calcium salts in the form, of pellets.
Apart frpm the presence of the amphoterics as pellets no additional solid or liquid
wastes are produced which cannot be reused.
Pilot plant experiments are necessary for the set-up of the process and the costs
involved.
References
Ripl, W. et al.
Recovery of phosphorus and precipitants (iron) from post-precipitation sludge.
Vom Wasser 70 (1988), pp. 179-185 (Ger).
Nikandrov, I.S. et al.
Recovery of phosphorus from sludge in the presence of surfactants.
Ih. Prikl. Khim. Leningrad 61 (1988), no. 3, pp. 621-623 (Russ.).
323
-------
3. Kugler, S. and G. Szalai
Nonselective processing of galvanic and phosphatic sludges.
Hu. 85-3371 t Sep., 1985.
4. Grynkewick, G.W.
Recovering phosphate values from precipitates formed during the
neutralization of impure wet process phosphoric acid.
US 4428915 A 31 Jan., 1984.
5. Gninwald, A. and J. Koller
Recovery of phosphates from activated sludge.
Sb. Vys. Sk. Chem.-Technol. Praze [Oddil] F.
Technol. Vody Prostredi, F24, (1982), pp. 97-110 (Czech.).
6. Scott, D.S.
Removal and recovery of metals and phosphates from municipal sewage
sludge.
Report, EPA-600/2-80-037 (1976)
Order No PB 80-221245, 45 pp.
Avail. NITS.
7. Plummer, D. .
Sludge incineration and precipitant recovery. Volume I.
A selective coded bibliography.
Res. Rep. - Res. Program Abatement
Munic. Pollut. Provis. Can.-Ont.
Agreement Great Lakes Water Qual. 31, 31 pp. (1976).
8. Powell, H.E. et al.
Recovery of phosphates and metals from phosphate sludge by solvent extrac-
tion.
VS Nat. Tech. Inform. Serv., PB Rep., No. 211933, 17 pp.
Avail. NTIS.
9. Sawhill, D.L.
Sodium phosphate recovery process.
US 70-98134 14 Dec. 1970.
10. Waters, R.F. et al.
Recovery of metals and phosphate from waste phosphate sludge.
US 71-120235, 2 Mar. 1971
324
-------
IN SITU VITRIFICATION APPLICATIONS
James E. Hansen and Vincent F. FitzPatrick
Geosafe Corporation
303 Parkplace, Suite 126
Kirkland, WA 98033
(206) 822-4000
Presented at the 3rd Forum on Innovative ,
Hazardous Waste Treatment Technologies:
Domestic and International
June 11-13
ABSTRACT
In Situ Vitrification (ISV) is an innovative solids remediation technology that involves
the electric melting of contaminated solids for purposes of destruction, removal, and/or immobili-
zation of hazardous contaminants. The ISV process is applied onsite and in situ. Contaminated
materials may be treated where they presently exist, or they may be staged in a more economic
configuration for treatment. Host natural soils may be treated without modification; rare cases
may require addition of fluxants or conductivity enhancing materials.
Typical processing temperature for contaminated soils is 1,600-2,000°C. ISV treatment
results in the removal or pyrolytic destruction (i.e., thermal decomposition) of hazardous organic
compounds. Heavy metals are either removed or permanently immobilized in a high quality, high
strength, glassy and/or crystalline residual product capable of surpassing TCLP leach testing
criteria. Typical volume reduction is 25-45%.
The ISV process can simultaneously process hazardous organics, inorganics (heavy metals),
radioactive contaminants, and mixtures of these. It can also tolerate the presence of varying
stratigraphy and debris and rubble, all within .reasonable limits. Application limitations exist
relative to: 1) water recharge rate if treatment is occurring in a recharge zone, 2) depth of
processing, 3) organic loading, and 4) presence of inclusions. The process does not cause adverse
electrical or thermal affects in the surrounding soil or nearby environment.
Relative to Superfund RI/FS technology evaluation criteria, ISV's capability to permanent-
ly destroy, remove, and/or immobilize hazardous contaminants results in excellent short-term and
long-term effectiveness and permanence, and reduction of toxicity and mobility. Its capability to
produce significant volume reduction in soils, sludges, sediments, and tailings is unique to ISV
and other vitrification technologies. ISV's current implementability is limited in that Geosafe
Corporation is a sole source for commercial applications (exclusive government licensee), and by
the existence of a single commercial large-scale equipment system which is currently undergoing
some redesign effort.
ISV costs categories include: 1) treatability testing at engineering-scale ($50-80,000
depending on contaminants and analytical requirements), 2) mobilization and demobilization ($125-
200,000 depending on transport distance), and 3) vitrification operations ($300-400/ton depending
on price of electricity, amount of water present, and depth of processing). Costs are not
included for site characterization, permitting/ARARs compliance, remedial design, site prepara-
tion, and other non-direct ISV activities.
The ISV technology is capable of meeting national and state ARARs. It is considered
highly protective of human health and the environment, and enjoys excellent regulatory and public
acceptance.
ISV technology has been selected as a preferred technology at ten private, EPA-Superfund,
and DOD sites.
INTRODUCTION
In Situ Vitrification (ISV) is an innovative, mobile, onsite
remediation technology for contaminated solids. ISV has been under
development for the U.S. Department of Energy (DOE) since 1980 by
Battelle Memorial Institute's Pacific Northwest Laboratories. DOE
has licensed the technology to Battelle, who has in turn exclusively
sublicensed it to Geosafe Corporation for commercial application
purposes.
325
-------
The ISV technology has been widely published and exhibited. The
interested reader is referred to Geosafe for ISV bibliographies and
detailed reports on various aspects of the technology. This paper
presents a brief description of ISV followed by a summary of applica-
tion and evaluation considerations of interest to regulatory and
engineering organizations involved with the comparative evaluation of
alternative technologies. The status of ISV technology development
and commercialization is also reviewed.
GENERAL PROCESS DESCRIPTION
The ISV process involves in situ electric melting of contaminat-
ed solids at very high temperatures, typically in the range of 1,600-
2,000'C for most soils. Figure 1 illustrates progressive stages of
ISV treatment, and Figure 2 presents typical process conditions. An
array of four electrodes is either placed to the desired treatment
depth in the volume to be treated prior to treatment (fixed elec- ,
trodes), or the electrodes are lowered into the treatment volume as
the melt progresses (moveable electrodes). A conductive mixture of
graphite and glass frit is placed on the surface between the elec-
trodes to serve as an initial conductive (starter) path. As electric
potential is applied between the electrodes, current flows through
the starter path, heating it and the adjacent solids to the solids
melting point. Upon melting, typical soils become electrically
conductive; thus the molten mass becomes the primary electrical
conductor and heat transfer medium allowing the process to continue
beyond startup. The molten mass grows downward and outward as long
as electric power is applied.
An off-gas collection hood gathers gases that evolve from the
treatment zone during processing. Water vapor is usually the predom-
inant evolved gas present in the hood, since most soils contain 15-
30% moisture above the saturated zone. Secondarily, organic contami-
nant pyrolysis products and soil decomposition products will evolve
to the surface under the collection hood. A large amount of ambient
air is allowed to enter the hood, where it supplies oxygen for the
combustion of flammable pyrolysis products and for purposes of
cooling the hood. The air and other gases are then drawn through an
off-gas treatment system to ensure their acceptability for release.
Significant volume reduction (25-45 vol% for most soils) occurs
as solids particles melt and interstitial void volume is removed.
Volume reduction results in a subsidence of the melt surface below
the starting grade (see Figure 3). When power is terminated to the
melt, it cools to a monolithic, vitrified (glassy with microcrystal-
linity) residual product which resembles natural obsidian (natural
volcanic glass) for most soil applications. Single melts as large as
1,000 tons can be produced by existing large-scale equipment capable
of processing 120 tons/day. Adjacent melts fuse together to produce
a single impermeable monolithic structure. Completion of each melt
setting involves placement of clean backfill to the desired depth in
the subsidence volume.
326
-------
Graphite and
Glass Frit
Starter Path
Electrodes
to Desired
Depth
Subsidence
Backfill Over
Completed
Monolith
\
<
I
i
Contaminated
Soil Region
Vitrified Monolith
(1)
(2)
(3)
FIGURE l. Stages of ISV Processing
Off-Gas Collection Hood
(-0.5 to 1.0 in HZO)
Controlled Air \
Input v \
X /
-/
Angle of Repose ^
Unaffected Soil
(minimum permeability)
?
S
f
\
Conductive Heating
(melt advance rate
of 1 to 2 in/hr)
r
Off-C
Trea
Subs
\
M
aases to
tment
dence
olten Soil
. int tin ho
/- 3,750 kva power level
- 0.3 to 0.4 kwh/lb treated
Soil Surface
flOO°C Isotherm
. Melt Surface
*• Dry Zone
- thermal gradient of
— 1 150 to250°C/ln
/ • maximum permeability
Region
• 1,600 to 2,000°C
- melt rate 4-6 tons/hr
- molten oxides and contaminants
- chemically reducing environment
- convection currents
FIGURE 2. Typical ISV Process Conditions
327
-------
-T^KirsTsrTSi'* • - - •'•"
,v^H5:v^ :••.;;
'A, • A-. .' -.-.ME^
@W«L -»"''15 ' » ** •"ir *«•• ••** i"C • • •• 'ts*'-:."'-"*-*•;.•*««*
•iff * i^< ' ' •-'- ' , '*••**.-• r'- '•', V1 •• -'; .-V™1 1J "i1? jij .y,f-
is'IF^" " -..*-,-;•>- -.' - •'"^.•i^™tfi?**f9ft
?T^'• i ":,: i,->i;vV/T,' .tV; i'vv'^rv^^JTta
FIGURE 3. Surface of Treatment Zone Showing
Subsidence Volume Over 750 Ton Melt
PROCESS EQUIPMENT
The ISV equipment is mounted on three over-the-road trailers so
that it is truly mobile in nature. The equipment is designed for
quick interconnection at the site. The ISV equipment system is
illustrated in simplified schematic form in Figure 4. Figure 5
presents an aerial view of Geosafe's large-scale commercial system.
The major portion of the equipment system is the off-gas collec-
tion and treatment system. A 55-ft diameter off-gas collection hood
directs ambient air and evolved gases/vapors from the treatment zone
to the off-gas treatment system. This system utilizes quenching,
venturi scrubbing, mist elimination, humidity control, filtration,
and carbon adsorption unit processes to ensure clean air emissions.
The quenching and scrubbing solution is cooled by a self-contained
glycol cooling system so that a continuous supply of onsite water is
not required. Periodically, contaminants collected in the scrubber
solution, filters, and/or carbon beds may be recycled back to a
subsequent ISV setting. In this way, only the secondary waste
present at the end of the last setting requires further treatment or
disposal.
328
-------
Off-Gas Hood
V
Controlled
Air Input
Power to Electrodes
f
Electrode <^
Location (typ)
Power / ,
Conditioning
ft
Utility
Diesel
Gener
Power
t
X Dewater ^
~VVVW\ ^VVVVV
;> Heat £
ated
Scrub j
^X^JBvi
Pillar •'
•ty
i
MM^M
r,tf$y$'
\
MMMMl
\ Quench
•>>>>:%";<
orb
.;
-
Clean t
Emissions
•S-S\. Backup :'.-
— — — » Generator,
jj;?/ Cooler, ••'.'••
'~y Filter,;ahd
y Adsorber
. ^Rlyrol -"
> , Cooling"' ',
'J'"'. '.'r-''5'-/-
FIGURE 4. Simplified ISV Equipment Schematic
FIGURE 5. Aerial View of Geosafe Large-scale ISV System
329
-------
APPLICATION CONSIDERATIONS
A
1. Application Types
The process is designed to treat contaminated soil in the
ground; however, it may also be applied in a large container. ISV
processing is termed "in situ" when the soils are processed where
they presently exist, as in a landfill or impoundment. When they are
placed in a trench or container for treatment, it is termed "staged"
processing. Some applications may involve consolidating contami-
natedsoil by removing and staging some soil on top of existing (in
situ) contaminated soil.
Since ISV is a batch or setting type process, its time-operated
efficiency increases with depth of processing. The process is most
economical when dealing with large quantities (e.g., 300-1,000 tons
treated/setting of electrodes). Processing depths greater than 10-ft
are ideal, but not necessary.
ISV applications may also be categorized relative to the primary
location and/or condition of the waste. Such categorization in-
cludes: 1) contaminated soil, 2) buried waste, and 3) underground
structures. Most ISV development work has focused on contaminated
soil applications wherein the contaminated media is primarily soil.
The soil has typically become contaminated in such cases through
exposure to contaminated liquids. In many cases the contaminated
liquid is water that has percolated through impounded or buried waste
that may or may not have been removed prior to addressing remediation
of the contaminated soil. Contaminated soil applications are rela-
tively straightforward compared to other types of applications; and
the ISV technology is considered to be developed and demonstrated for
many such applications.
Buried waste applications address wastes that have been covered
by soil such as backfilled impoundments and landfills. Substantial
amounts of test work has been performed on a variety of process
sludges, ash, and containerized waste. The ISV technology is not
considered generically ready for such applications; at this time a
specific test and demonstration plan is necessary for each one.
Buried waste applications involving wastes which were highly
heterogeneous at time of burial, typically pose a problem of site
characterization. It is necessary to know worst case conditions
within the treatment zone to allow appropriate remedial design for
the site. Homogeneous wastes, such as some settled lagoon and
impoundment sludges and sediments, pose less of a characterization
problem. However, the chemical composition of such wastes must be
analyzed relative to the soil in the treatment zone to allow predic-
tion and evaluation of melt behavior when the sludge/sediment zones
are encountered. In some cases it may be necessary to intermix the
soil and waste layers to allow proper treatment. The effect of the
wastes on overall residual product chemistry and properties must also
be evaluated.
Containerized wastes such as buried drums, crates, and cartons
pose additional problems. Whereas the ISV process conditions may be
adequate for treating such materials,' the site characterization
challenge becomes even more severe. The ISV technology is not
330
-------
considered ready for application to such sites at this time except on
a test and demonstration basis. ISV is being developed for such
applications within the DOE community because of the high cost of
alternative technologies. Treatment of such sites may require use of
equipment with larger than normal off-gas treatment capacity and/or
the use of secondary off-gas containment to protect against unforseen
high gas generation events.
2.
Solid Media
The primary qualification regarding type of soil that may be
treated by ISV is whether or not the soil will form and support a
melt. ISV test results have indicated that most natural soils may be
processed by ISV without modification. Various sludges, sediments,
and process tailings have also been successfully tested. For proper
application, it is necessary that the soil and/or other solids
contain sufficient inorganic material that will remain in the molten
state during treatment. It is the molten mass that serves as the
electrical conductor during ISV; and the flow of electricity through
the melt results in the generation of heat which is then passed into
adjacent soil by thermal conduction.
Molten soil must possess sufficient electrical conductivity to
allow the process to be performed economically. Electrical conduc-
tivity within a soil melt is typically provided by the monovalent
alkali earth cations (e.g., sodium, potassium). It is desirable that
such cations be present in the 2-5 wt% range, which is common for
most soils. In the event a soil possesses insufficient molten
conductivity, it is possible to obtain the needed conductivity
through addition of other materials (e.g., materials that provide
Na2O and/or CaO, such as suitable soil, soda ash, and lime).
The chemical (oxide) composition of the soil is important in
determining the quality of residual product produced. Soil is the -
result of weathering of rocks, and rocks are made up many minerals
(complex metal oxides). Upon melting, minerals decompose to a melt
mixture of major oxides, in which silica is predominant for most
soils. Silicate melts typically produce a residual product of
excellent properties relative to environmental exposure. Other low-
silica soils (e.g., limestone/dolomite) have also been treated by ISV
to produce a high quality residual product. It is possible to
determine the applicability of ISV to various soils by performing and
evaluating oxide composition analyses and small-scale melt tests.
•V
3. Contaminant Disposition ,
As the high temperature ISV melt moves slowly downward and
outward through the contaminated solids, a very steep thermal gradi-
ent (150 to 250°C/inch) precedes the melt. At appropriate tempera-
ture regimes within this gradient, or within the melt itself, the
solids and,contaminants undergo change of physical state and decompo-
sition reactions. The possible dispositions of particular contami-
nants include: 1) chemical and/or thermal destruction, 2) removal
from the treatment volume to the off-gas treatment system, and 3)
chemical and/or physical incorporation within the residual product.
Many site- and application-specific variables affect the disposition
of specific contaminants. The primary variables include: 1) con-
taminant physical and chemical properties, 2) melt chemistry, 3) melt
331
-------
temperature, 4) contaminant dwell time in the treatment zone (in turn
dependent on melt viscosity, depth, and other variables), 5) adjacent
soil properties, 6) soil moisture content, and 7) extent of over-
melting (i.e., amount of soil melted beyond the limit of contamina-
tion) . Because of the many site-specific variables involved, it is
necessary to consider each remediation project individually.
The ISV testing program has indicated that certain classes of
contaminants may be expected to undergo basic types of response and
ultimate disposition during treatment. Hazardous compounds undergo
the phenomenon of pyrolysis (i.e., thermally induced decomposition of
compounds into their elements, usually in the absence of oxygen;
applicable to organics) and thermal decomposition (applicable to
inorganics). For example, chlorinated organics decompose to carbon,
hydrogen, and chlorine; and nitrates break down into nitrogen and
oxygen. In addition to the pyrolysis products, it is also possible
that limited quantities of highly volatile materials may evolve from
the treatment volume during processing. All materials evolved are
captured in a collection hood and are subjected to off-gas treatment
processes to ensure all emissions are within regulatory limits.
The solid media itself may also decompose during processing.
For example, the inorganic portion of soils, which consist of complex
mineral compounds, typically break down into major oxide groups such
as silica and alumina. Upon cooling of the ISV melt, which is
relatively rapid in terms of the time required for minerals to form,
a residual product is formed which is glassy (a supercooled liquid of
the oxide mixture) and may have varying amounts of crystallinity
(from precipitated minerals) present. Such residual product typical-
ly has outstanding environmental exposure properties.
Regulatory criteria of interest regarding the residual monolith
produced typically relate to: 1) structural, 2) weathering, 3)
chemical leaching and 4) biotoxicity properties. EPA has performed
tests on typical ISV product in these areas (Ref. 1). Structural
strength tests indicated approximately 10 times the strength of
unreinforced concrete, both in tension (ISV values of 4-8,000 psi)
and compression (ISV values of 30-45,000 psi). Freeze/thaw and
wet/dry weathering tests indicated the ISV residual was unaffected by
repeated exposure. Chemical leaching tests consistently indicated
the ISV residual is capable of surpassing the EP Toxicity (EP-Tox)
and Toxic Characteristic Leaching Procedure (TCLP) leach tests. EPA
also found the ISV residual to be non-toxic to near surface life
forms (Ref. 2).
The above properties of the ISV residual product make it truly
unique among remediation alternatives. Because of its unequalled
ability to immobilize arsenic, as indicated by TCLP testing, vitrifi-
cation has been identified as the best demonstrated available tech-
nology (BDAT) for arsenic-bearing wastes as defined in the current
Resource Conservation and Recovery Act (RCRA) landban regulations.
The ISV residual product is considered to be permanent; that is,
capable of withstanding environmental exposure for geologic time
periods (e.g., thousands to millions of years).
In typical soil applications inorganic elements which do not
evolve from the melt during processing become part of this residual
332
-------
product through physical and/or chemical incorporation. The reader
interested in heavy metals applications considerations is referred to
Reference 1.
Tables 1 and 2 present typical results from the ISV development
and testing program, indicating performance on various types of
contaminants. Table 1 presents organic destruction and removal
results, and Table 2 presents heavy metal retention, removal and
leach testing results.
4. Presence of Water
The presence and movement of water during ISV is a major consid-
eration in evaluating potential applications and in project remedial^
design. During ISV, the thermal gradient which moves in front of the
melt evaporates water within the 100°C isoband that starts less than
1-ft away from the melt. Water vapor moves to the surface through
and adjacent to the melt, accomplishing some vapor stripping of other
volatiles (e.g., organics) as it moves. The water pathways to the
surface are illustrated in Figure 6.
Research studies have indicated that, while a slight steam
pressure (1-2 psi) may exist within the dry zone and 100°C isoband,
the water flow path is to the surface as opposed to into the adjacent
soi'l. This occurs because vapor phase permeability within the 100 °C
isoband is at least several orders of magnitude lower than within the
dry zone due to the presence of liquid water within the isoband.
Water mass balance experiments have verified that substantially all
water present in the treatment zone is removed to the surface' during
processing (Ref. 3). ,
Since significant energy is required to vaporize water, its
presence represents an economic penalty. Therefore, it is economi-
cally advantageous for the treatment volume to be as dry as practi-
cable immediately prior to ISV treatment. In similar manner, consid-
eration should be given to employing means to minimize/eliminate
water recharge if ISV is applied in an active water zone. Typical
means for limiting recharge include use of barrier walls, well
points, or french drains.
Water may also influence processing cost in that its removal
from the treatment zone relates directly to processing rate. Since
the water vapor flows to the surface through the dry zone and the
melt itself, it contributes to the agitation (flow activity) level of
the melt. Depending on melt and adjacent soil conditions, processing
rate must be controlled to maintain acceptable water vapor generation
and removal rates.
5. Affects on Nearby Surroundings
The volume reduction characteristic of ISV processing results in
a subsidence volume over the residual product mass (refer back to
Figure 3). Such subsidence and adjacent soil sloughing may result in
undercutting surface or near-surface structures, unless engineering •
provision is made to prevent sloughing. It should also be noted that
the off-gas collection hood is 55 ft in diameter compared to a
maximum large-scale melt width of 30-35 ft. Thus, a minimum side
clearance of 15-20 ft is desirable.
333
-------
TABLE 1. Typical Organic Destruction/Removal Efficiencies
Contaminant
Concentration
fppb)
Percent
Destruction
Percent
Removal fl)
Total
ORE f%)
PESTICIDES
4,4 DDD/DDE/DDT
Aldrin
Chlordane
Dieldrin
Heptachlor
VOLATILES
Fuel Oil
MEK
Toluene
Trichloroethane
Xylenes
SEMI -VOLATILES
PCP
NON-VOLATILES
Glycol
PCBs
Dioxins
Furans
**************
21-240,000
113
535,000
24,000
61
**************
230-110,000
6,000 (2)
203,000
106,000
3,533,000
**************
>4, 000, 000
**************
8,000 (3)
19,400,000
>47,000
>9,400
**************
99.9-99.99
>97
99.95
98-99.9
98.7
**************
>99
>99
99.996
99.995
99.998
**************
99.995
**************
>98
99.9-99.99
99.9-99.99
99.9-99.99
**************
>99.9
>99.9
>99.9
>99.9
>99.9
**************
>99.9
>99.9
>99.9
>99.9
>99.9
**************
>99.9
**************
>99.9
>99.9
>99.9
>99.9
*************
99.9999
99.99
99.9999
99.99
99.99
*************
99.999
99.999
99.99999
99.99999
99.99999
*************
99.99999
*************
99.99
99.9999
99.9999
99.9999
oo
CO
(1) Percent removed from off-gas after destruction; percentages are additive for the
total ORE.
(2) 98% MEK in container, yielding 6,000 ppm in layer of container thickness
(3) 50% ethylene glycol in container, yielding 8,000 ppm in layer of container thickness
-------
TABLE 2. Typical Inorganic Removal/Retention/Leach Results
Percent Percent Total Initial TCLP (3)
Contaminant Retention Removal (1) Ret/Rem (%) Concen. (2) Result Allowable
VOLATILE
Hg
SEMI-VOL.
As
Cd
Co
Cs
Pb
NON-VOL.
Ba
Cr
Cu
Ni
Ra-226
Pu/Th/U
Zn
*********
0
**********
70-85
67-75
99-99.9
99-99.9
90-99
**********
99.9
99.9
90-99
99.9
99.9
99.99
90-99
*********
97->99
**********
>99.9
>99.9
>99
>99
>99.9
**********
>99.98
>99.9
>99.9
>99.9
>99.9
>99.9
>99.9
*********
>97
**********
99.98
99.96
99.99
99 . 99
99.99
**********
99.9999
99.9999
99.99
99.9999
99.9999
99.99999
99.99
**********
5,360
*********
43,900
37
17
—
1,550
**********
185
290
65,000
47
6,000 pCi/g
—
14,200
**********
ND
**********
0.9
0.001
<0.01
—
<0.063
**********
0.140
0.020
3.3
ND
6.4 pCi/g
—
<0.05
*********
0.2
**********
5.0
1.0
—
—
5.0
*********
100
5.0
—
—
100 (4)
—
—
oo
oo
tn
(1) Percent removed from off-gas not retained; thus, percentages are additive for
retention and removal
(2) Concentration in ppm unless otherwise noted
(3) TCLP values in mg/L unless otherwise noted
(4) DOE limit for concentration in drinking water
Either not available or not applicable
ND Non-detectable
-------
Fusion Zone
Ambient Soil (little or
no temperature affect;
no vapor movement)
100 C isoband
(little or no vapor
movement due to
low permeability)
Melt(1,600-2,0008C)
Net flow of water vapor
is from the dry zone side
of the 100°C isoband
through the dry zone to
the surface or through
the melt and thence to
the surface
FIGURE 6. Water Vapor Pathways to Surface
Cooling of the residual monolith may result in some heating of
the adjacent soil. During processing, the 100°C isoband extends from
less than 1 ft away from the melt to 3-5 ft away (typical case).
During cooling, the 100°C isotherm has been observed to extend as far
as 10 ft from the melt. This occurrence has the effect of drying out
the adjacent soil.
The ISV process has never been observed to induce significant
magnetic or electric fields beyond the treatment zone.
6.
Production Rates
The large-scale ISV equipment is capable of a 3.5 MW power
level. This corresponds to a maximum soil melting rate in the range
of 4-6 ton/hr. This in turn results in a melt advance rate of 1-2
in/hr, depending on soil density. The process is operated 24 hr/day
until a melt setting is completed. Downtime for movement of equip-
ment between settings is less than one day.
Other factors, such as allowable water vapor generation and
removal rates may govern processing rates to lower than those possi-
ble based on available power level. Allowable production rates can
be predicted during remedial design.
336
-------
7. Secondary Waste
The ISV off-gas treatment system may collect small quantities of
hazardous materials in the quencher/scrubber solution, the HEPA
filter, and the activated carbon. These media are either subject to
disposal as secondary waste, or they may be placed within a subse-
quent ISV setting for disposal (of the media) and recycling of the
contaminants. In this way the site-wide contaminant destruction/-
removal/immobilization efficiencies may be maximized and secondary
waste minimized. In a similar manner, limited quantities of protec-
tive clothing and other site secondary waste may be treated. Demobi-
lization decontamination solution and the quencher/scrubber solution,
filters and activated carbon from the last setting at the site are
subject to disposal as secondary waste.
8.
Cost
Typical ISV remediation projects involve the following work
elements: 1) site characterization , 2) treatability/pilot testing,
3) remedial design, 4) permitting/compliance analysis and documen-
tation, 5) site preparation, 6) equipment mobilization, 7) onsite
vitrification operations, 8) equipment demobilization, 9) site
restoration, and 10) delisting and/or long-term monitoring. The cost
of these activities is dependent upon specific conditions at the
site, and the overall project criteria and objectives. Typical cost
ranges are discussed below for those activities that are peculiar to
•ISV processing.
Treatability/pilot testing is utilized to: 1) demonstrate that
the technology is applicable to the specific soil/waste combination
at the site, 2) produce contaminant-related performance data neces-
sary to support permitting activities, 3) produce operation-related
performance data necessary to support cost estimates and quotations,
and 4) produce samples of residual product for use in community
relations efforts. Treatability testing involves performance of
various physical and chemical tests on actual contaminated materials
from the site, followed by engineering-scale ISV melt testing on the
materials. The cost of treatability testing for non-PCB, non-diox-
in/furan wastes is $25,000 plus the cost of necessary analytical
work, which usually falls in the $15,000 to $25,000 range. For PCB
and/or dioxin/furan wastes, the base treatability testing cost is
$30,000 plus analytical costs in the range of $25,000 to $50,000.
These cost ranges do not include data validation costs or other
unusual analytical requirements. Treatability testing can usually be
completed within 10 to 12 weeks after initiation of the project.
The cost of equipment mobilization and demobilization depends on
the transport distance to and from the site. Typical total mobiliza-
tion/demobilization costs fall in the range of $125,000 to $200,000.
The on-site service cost of ISV processing typically falls in
the range of $300 to 400 per ton of material processed. This cost
includes all elements of direct and indirect cost, such as labor,
materials, energy, equipment amortization, and contractor overhead
and profit. The most significant variables affecting this cost
include: 1) the cost of electrical power, 2) the amount of water to
337
-------
be removed during processing, 3) depth of processing, and 4) analyti-
cal chemistry requirements associated with process control and permit
compliance. Given information on these and other pertinent vari-
ables, application-specific cost estimates may be developed.
ADVANTAGES AND LIMITATIONS
Advantages
ISy advantages relative to alternative technologies include its
capability to: 1) simultaneously process mixed waste types (organic,
heavy metal, radioactive), 2) achieve destruction, removal, and
immobilization performance beyond regulatory criteria, 3) be per-
formed onsite and in situ, 4) accept significant quantities of rubble
and debris in the treatment zone, 5) achieve a significant volume
reduction (25-45% for most soils), and 6) produce an unequalled
residual product with a geologic time life expectancy (thousands to
millions of years). ISV also possesses significant differences from
other vitrification technologies (e.g., plasma, joule-melter, slagg-
ing kiln), many of which are considered advantages (Ref. 4).
2. Limitations
The primary limitations on ISV applications relate to: 1) total
organic concentration, 2) water recharge rate, 3) depth of process-
ing, 4) and presence of inclusions. Since organics become gaseous
pyrolysis products during ISV, the concentration of organics must be
limited in relation to the off-gas collection and treatment equipment
capacity. The average allowable concentration for most organics
falls in the range of 5-10 wt%.
Fully saturated soils may be processed; however, it is economi-
cally advantageous to minimize soil moisture content and water
recharge rate. These factors influence cost through consumption of
energy and impacting of processing rate. Processing rate may be
limited by the amount of energy going into water removal, or by
operating at less than full power to maintain acceptable water vapor
generation and removal rates peculiar to a specific application.
The maximum depth processed by ISV to date is 19-20 ft. Greater
depths will be attempted in the continuing ISV development program.
The ISV process is capable of accommodating significant inclu-
sions within the treatment zone (e.g., rocks, roots, drum remnants
and other metal scrap, concrete, asphalt, construction debris, etc.);
however, the concentration of these must be limited so as to not
interfere with proper formation and advancement of the melt. All of
the above limitations are subject to consideration during applicabil-
ity analyses, treatability testing, and project remedial design.
338
-------
CHARACTERISTICS PERTINENT TO RI/FS EVALUATION CRITERIA
ISV typically fares well in evaluation of short- and long-term
effectiveness, permanence, and reduction of toxicity and mobility
because of its excellent capability to destroy, remove, and/or
immobilize contaminants. It is outstanding in the area of volume
reduction in that vitrification is the only known means to achieve
significant volume reduction (25-45%) in silica-based soils.
ISV is limited in regards to the implementability criterion in
that it is commercially available from a single source. The technol-
ogy is covered by a basic patent which has been licensed by DOE to
Battelle, and has been exclusively sublicensed to Geosafe for commer-
cial application. At this time Geosafe owns a single commercial
large-scale ISV machine capable of processing 15-25,000 tons of soil
per year. The off-gas collection hood portion of that machine is
presently undergoing redesign to increase its design capacity.
As noted in the cost discussion above, total ISV project costs
are highly site specific. When total project costs are evaluated,
use of ISV may be the most cost effective alternative for specific
sites, particularly when considered in relation to the effectiveness,
permanence,
and volume reduction criteria.
ISV is considered capable of meeting state and federal ARARs
where it is being considered for use. ISV has enjoyed good support
by the regulatory community. The onsite and in situ nature of ISV,
and the quality of its residual product have resulted in generally
excellent acceptance by the public.
DEVELOPMENT AND COMMERCIALIZATION STATUS
1. U.S. DOE Development Program
The current DOE ISV development program is addressing potential
applications at DOE's Hanford, Idaho Falls, Oak Ridge, Rocky Flats,
and Savannah River plant sites. The program is addressing contami-
nated soil, buried,waste, and underground tank type applications.
The largest program (estimated $10-15 M/yr) is addressing a buried
waste site at the Idaho National Engineering Laboratory (INEL).
Battelle's Pacific Northwest Laboratories provides national coordina-
tion of DOE's ISV program. Experimental and demonstration work is
being performed by Geosafe, Battelle, and EG&G Idaho.
2. Vapor Retreat Issue <*"
The ISV development and commercialization program has long been
plagued by a competitor who claims that vapors generated during ISV
treatment do not rise toward the surface as claimed by the ISV
technical community, but rather "retreat" into the soil adjacent to
and underneath the melt. Geosafe and Battelle have thoroughly
investigated these allegations and have found them to be without
technical merit and in exact opposition to observed ISV processing
performance (Ref. 5).
339
-------
EPA has participated in the resolution of this issue by requir-
ing treatability test work plans to include the attainment of inde-
pendently qualified performance data relative to whether or not
contaminants move into the adjacent soil during processing. One such
recent test on PCB-contaminated soil included a large number of
samples (for statistical significance) and an independent data
validation effort. The test results confirmed the complete absence
of contaminant migration into the adjacent soil (Ref. 6).
3.
Moveable Electrode Development
The DOE ISV program has developed a moveable electrode concept
for application to very high metals content sites. The need for such
electrodes was identified relative to an INEL site containing very
high levels of metals in buried waste (20-40 wt%). The original ISV
fixed electrode concept, wherein electrodes are placed to full depth
in the treatment zone before initiation of processing, was found to
be limited in such high metals applications due to shorting between
the electrodes caused by pooling of molten metal at the bottom of the
melt. Moveable electrodes are lowered into the melt as melt depth
increases. In the event of metal pooling in the melt, the moveable
electrodes can be maintained at least a minimum distance above the
pool, thereby avoiding shorting.
Moveable electrodes avoid the necessity of preplacing electrodes
in the treatment zone. They also have the advantage of being made
only from graphite, whereas the fixed electrodes are of a combination
graphite/molybdenum core construction. At this time all ISV test
equipment has been converted to the moveable electrode capability.
Battelle will be testing moveable electrodes at large-scale during an
underground tank treatment test which is presently scheduled to occur
in June, 1991.
4. Geosafe Operational Acceptance Testing Event
Geosafe recently experienced an event during large-scale opera-
tional acceptance testing of a new fabric hood design that resulted
in significant damage to the hood. The test results indicated that
the hood containment fabric was not acceptable for worst case operat-
ing conditions. Geosafe is currently investigating the cause(s) of
the event, and is redesigning the large-scale ISV off-gas collection
hood. Geosafe is returning to an all metal hood design, such as has
been the standard during the ISV development program. Commercial ISV
field operations are currently on hold pending conclusion of the
investigation and completion and testing of the new hood.
5.
Cobble Walls for Melt Control
Recent tests employing cobble (2-4 in diameter rock) barrier
walls have shown the concept to be effective in retarding, but not
eliminating, melt growth rate in the direction of the cobble wall.
The cobble material, which has a high solid density and large void
volume between stones, melts slower than soil particles with small
void volumes. The objective of these tests has been to explore the
340
-------
use of cobble walls for applications wherein melt shape control is
desired for various purposes.
6. Preferred Remedy Selections
ISV has been selected as a preferred remedy at 10 private, EPA-
Superfund, and DOD sites within the U.S. These selections include:
1) Parsons Chemical/ETM Enterprises (EPA-V), 2) Northwest Transformer
(EPA-X), 3) Arnold AFB Site 10 (DOD), 4) Rocky Mountain Arsenal M-l
Holding Ponds (DOD), 5) Ionia City Landfill (EPA-V), 6) Crab Orchard
National Wildlife Refuge (EPA-V), 7) Anderson Development Company
(EPA-V), 8) Crystal Chemical (EPA-VI), 9) Wasatch Chemical (EPA-
VIII), and 10) Transformer Service Facility (EPA-X). Remediation
contracts currently exist for two of these sites. The others are at
various stages of treatability testing, remedial design, or are under
'negotiation between regulators and responsible parties.
REFERENCES
1. Paxton, James. 1985. "Environmental Protection Agency (EPA),
Test Program on Raw, Stabilized and Vitrified Soil, Western
Processing Inc.". NPDEN-GS-1, December 12, 1985, Department of
the Army/ North Pacific Division Materials Laboratory, Corps of
Engineers, Troutdale, OR.
2. Green, Joseph C., et al. 1988. "Comparison of Toxicity Results
Obtained from Eluates Prepared from Non-Stabilized and Stabi-
lized Waste Site Soils". Proceedings of the 5th National
Conference on Hazardous Wastes and Hazardous Materials. April
19-21. 1988. Las Vegas, NV.
3. Bonner, W.F. and J.L. Buelt. 1989. "In Situ Vitrification:
Test Results for a Contaminated Soil Melting Process". PNL-SA-
16584. Paper presented at the 1989 Incineration Conference, May
1-5, 1989, Knoxville, TN.
4. Hansen, J.E. 1989. "Vitrification Technologies". Immobiliza-
tion Technology Seminar. U.S. EPA, Center for Environmental
Research Information, Cincinnati, OH; CERI-89-222.
5. Geosafe Corporation. 1990. "Geosafe Corporation Comments on
Claims by Larry Penberthy, President of PEI, Inc., Against In
Situ Vitrification Technology". Geosafe Corporation. November
22, 1990.
6. Geosafe Corporation. 1991. Engineering-Scale Test Report for
Application of In Situ Vitrification Technology to Soils Contam-
inated with Polychlorinated Biphenyls at the Northwest Trans-
former Superfund Site. GSC 1006, Rev. 1, February 6, 1991.
Geosafe Corporation, Kirkland, WA.
341
-------
Rapid rehabilitation
of a former coking plant site
Ulrich Jacobs
RWE Entsorgung Aktiengesellschaft
BamlerstraBe 61
D-4300 Essen 1
342
-------
•\.PT\or use o1 the s'rte
'The Friedrich-Ernestine coking plant operated in the Essen district of
Stoppenberg between 1929 and 1959. Besides coke furnaces, on the site
were also several additional tar. production facilities. Old plans of the site
show that it also contained ammonia and benzene scrubbers, a tar
pressure separator and a naphtaline plant.
RWE Entsorgung
Rehabilitation Friedrich-Ernestine
the coking plant (1950)
After the plant has been shut down it was demolished and the site was later
used by a construction company. Two administration buildings were
erected. The site of the addtional facilities was filled In, paved over and
used as a car park.
343
-------
machine
room
ammonia
and
benzene
scrubber
DDDD
tar condensation
pressure
separator
Looo
^00
RWE Entsorgung
Rehabilitation Friedrich-Ernestine
prior use of the site
2. Geological situation
The site's subsoil initially comprises an approx. 1 m thick anthropogenic
backtili consisting largely of building and mining debris.
Below the backfill material lies quaternary, silty/sandy loess clay, slope
wash and possibly boulder clay of varying thickness (1-5 m). The site in
question slopes northwards down towards the Emscher river lowland. At
the base of the Quaternary lie layers of upper cretaceous matter which
begin with thin, slightly weathered and shifted greensands of fine to
medium coarseness. The clay content of the greensand varies according
to the degree of weathering. The greensand understratum consists of
faulted marl which is impounded at the upper level.
There are two separate groundwater strata beneath the site in question.
The upper groundwater stratum is in the vicinity of Quaternary silt, silty
344
-------
sand as well as greensand, where impounded water. collects at Jeve/s
between approx. 6 and 10m beneath the surface of the site. The lower
stratum, which forms the actual aquifer, is located in the faulted marl at a
depth of 10 - 13m below the surface. The clay surface of the marl functions
as a barrier dividing the two groundwater strata.
upper
groundwater
stratum —,
iquaternarylil
!•!• •-. A... -»••.••_•.••»».•.••.•••*.•••..•.^"•»*« • •**.*•*.•*.•
lower —J
groundwater
stratum
AWE Entsorgung
Rehabilitation Friedrich-Ernestine
geological situation
3. Expert Investigations
Soil contamination at the site was first discovered during a routine
investigation conducted by the Essen municipal authority, in which all
industrial sites were subjected to random testing. A subsequently ordered
hazard estimation survey localised substantial contamination caused by tar
oil residue in the vicinity of where the additional plants had once stood. The
analysis revealed that the largely organic contamination was due to
polycyclical aromatic hydrocarbons present in concentrations of between
1,000 and 3,000 ppm. In addition, the survey established substantial
contamination caused by volatile aromatic compounds (BTX) in the air at
ground level. A prior investigation had already served to establish the main
areas of contamination. An examination of old plans enabled the surveyor
to pinpoint the tar pressure separator and the ammonia and benzene
scrubbers as the source of the lion's share of the contamination.
345
-------
machine
room
ammonia
and
benzene
scrubber
tar
pressure
separator
condensation
RWE Entsorgung
Rehabilitation Friedrich-Ernestine
main areas of contamination
In the opinion of the surveyor, the contamination necessitated urgent
rehabilitation measures since an emergency well belonging to Essen City
Council was located downstream from the site and its water already
contained hazardous substances.
In mid-June 1989 the surveyor was commissioned with the compilation of a
rehabilitation report on the basis of the results of the previous surveys. The
main areas of concentrated contamination were sampled by compiling a
detailed matrix of the site, enabling their extent to be determined with great
precision. Attempts were also made to
establish the depth of contamination. The survey pinpointed two clearly
defined areas of severe contamination and ascertained that the level of
concentration dropped by a factor of 10 - 20 at a depth of between 4 and 5
metres below the surface.
4. The rehabilitation scheme
The rehabilitation report together with the rehabilitation scheme, compiled
by the companies involved served as the document submitted to the
supervisory authorities for approval.
346
-------
The rehabilitation scheme recommended excavating the areas of major
contamination down as far as the drop in contamination but under no
conditions into the aquifer, in order to avoid an hydraulic soil collapse. This
involved making two excavations, precisely located by the results of the
rehabilitation investigation and lined for safety. The surface area of the
smaller excavation was 510 sq.m., that of the larger one amounted to 1,750
sq.m. The soil was to be separated and classified by the surveyor during
excavation work:
machine
room
ammonia
and
benzene
scrubber
tar condensation
pressure
separator
RWE Entsorgung
Rehabilitation Friedrich-Ernestine
location of the excavations
Uncontaminated matter would remain on site whilst contaminated matter
would be subject to a decontamination process.
Contaminated soil was to be cleaned in a thermic plant. As no such plant
existed in West Germany, it was planned to transport the material by truck
to the intended plant, located in the Netherlands, where it would be cleaned
and returned for refilling. The return of the decontaminated soil was
required by the Essen City Council ordinance authority, as it is wished to
constantly monitor the success of the cleaning process on the one hand
and prevent dumping in Holland on the other.
347
-------
The substantial contamination of the ground level air by BTX gases,
especially benzene, meant that wide-ranging safety measures were
necessary to prevent surface water from seeping into the excavations, as
this would have inevitably transmitted contamination to the groundwater.
For these reasons it was planned to cover both excavations with airtight
tents fitted with a ventilation system. Air suctioned out of the tent would be
passed through an activated charcoal filter before expulsion. In addition, all
items of machinery used on the site were to be fitted with airtight and air-
conditioned' cabins, also' drawing their fresh air supply via activated
charcoal filters.
RWE Entsorgung
Rehabilitation Friedrich-Ernestine
black and white area
In order to clearly distinguish between contaminated and uncontaminated
areas, the scheme involved marking the entire site ^ either black
(contaminated) or white (uncontaminated). The black area would comprise
the excavations themselves plus all sections in which contaminated
material was moved or stored, with the white area consisting of parking and
loading zones for the trucks as well as storage areas for cleaned soil. Since
348
-------
the-excavation and decontamination processes could not be conducted
simultaneously due to the material having to be transported to the
Netherlands and back, a hall located on the site was transported into a
buffer storage facility for contaminated soil and fitted with a channel for
loading onto the trucks. Changing and shower rooms in the hall were
converted into a personnel filter between the black and white areas,
ensuring that neither workmen nor trucks could carry contamination with
them out of the site.
•'-."^i •" 9^Vej-t-&''f.rsAf±\''-'f- ' ' " """fat*
fi ^ *v*-",vi.ft t*y&f fi^xfe .» *
Rehabilitation Friedrich-Ernestine
depot channel
RWE Entsorgung
5. Authorisation permits
The authorisation applications for the proposed rehabilitation measures
were submitted to the relevant authorities'in mid-December. The individual
applications in question were for contstruction authorisation, water law
authorisation and an exemption permit in accordance with Section 4,
Paragraph 2 of the West German Refuse Law. In addition, a transport permit
from the President of the state government of North Rhine-Westphalia was
also required for the transportation of contaminated soil to the Netherlands.
Following intensive negotiations, the authorities issued all necessary
349
-------
permits in complete accordance with the rehabilitation scheme, thus giving
the go-ahead for work to begin.
With regard to the objective of the rehabilitation scheme, it was agreed that
the necessary excavation depth would be reached at the point where there
was' a sharp drop in contamination. This objective would require
confirmation from the surveyor as well as the authorities.
6. Preparatory measures
Work on the excavation lining began in mid-January, prior to the actual
issue of authorisation but in agreement with the authorities. ,
A Essen-type lining was used. For this we inserted stell girders of approx.
10.5 m length down pre-bored holes at intervals of 2 m before commencing
excavation. Wooden planks would then be gradually wedged horizontally
between the girders as the excavation proceeded. In order to ensure the
safety of the lining, two anchor tiers had to be installed, the first
immediately beneath the surface and the second, as work progressed,
around 5 m below.
This was followed by the erection of the tents, sealing of the canvas and
installation of the-ventilation system.
Rehabilitation Friedrich-Ernestme
Entsorgung
entrance of the tent .
350
-------
Two deep wells were sunk directly downstream from the two excavations to
act as additional hydraulic barriers. Permanent monitoring for contaminants
in the wells ensured that the release of any pollutants mobilised by soil
movement could be ascertained.
The foundations of the buffer storage facility in the hall were sealed off by
means of seepage shafts and, since it was possible that slightly volatile
gases might also escape, this was also fitted with a ventilation system.
7. Industrial safety
The entire site workforce was medically examined before excavation got
underway, the purpose of the examination being to serve as a control base
in case of possible illnesses during work. Protective equipment and masks
within constant reach were compulsory when working in the buffer storage
facility and the excavations themselves.
The gas levels in these areas were constantly monitored using photo-
ionisation detectors (PIDs). Trade supervision regulations stipluated that
the wearing of a protective mask was compulsory when gas concentrations
exceeded 5 ppm.
Gas concentration levels were commonly over 5 pp, during the entire
excavation period, both in the storage facility as well as in the pits
themselves, with the effect that soil movements were mainly carried out
wearing full protective gear.
Air passing into the outside atmosphere after active charcoal filtering was
also subject to constant monitoring. At no point during the excavation work
was the maximum limit exceeded.
8. Soil cleaning
The soil was cleaned on the premises of Afvalstoffen Terminal Moerdijk
(ATM), near Rotterdam.
The ATM plant operates in-three stages comprising two directly fired rotary
tubular kilns and one atterburn chamber. In the first rotary kiln the soil ist
preheated and dried, reducing its moisture content by around 70 %. In the
second kiln the organic impurities are steamed out at temperatures of up to
800 ° C before being thermically destroyed in the atterburn chamber.
ATM's plant has been operating since 1984 and achieves a throughput of
approx. 30 t/h, depending on the moisture content of the soil.
9. Implementation of the rehabilitation work
In February, when authorisation had been issued, we began excavating the
uncontaminated areas, storing the material in the white marked section of
the site.
351
-------
The first contaminated soil from the .smaller pit was excavated on 6th
March, at which point work on the larger pit began. From April onwards the
excavations proceeded in parallel. ,•
RWE Entsorgung
Rehabilitation Friedrich-Ernestine
excavation in the small pit
Two skip trucks were used to move the contaminated material within the
black area from the pits to the buffer depot, where it was loaded through
the depot channel by means of a wheel loader onto trucks for
transportation to the cleaning plant. At peak times the pits were filling 27
trucks per day, each carrying 25 tonnes of soil.
352
-------
Rehabilitation Friedrich-Ernestme
truck loading
RWE Entsorgung
As it had been expected, given the history of the site, the excavation
yielded large amounts of structural 'debris, including steel tanks and pipes
filled with tar oil residue. Concrete and masonry was broken up coarsely
during excavation, removed and comminuted prior to thermic treatment in a
crushing plant. Stell sections, some of which were seriously contaminated,
had to be pre-cleaned manually and then sandblasted. The sand used in
the process was naturally added to the contaminated soil and taken to the
decontamination plant.
In accordance with the results of the initial surveys, we found that the
degree of contamination in the eastern pit decreased the deeper excavation
proceeded, the contamination drop being at a depth of around 6 m, well
above the aquifer. We had therefore, in accordance with the agreement with
the surveyor and authorities, attained the rehabilitation objective at this
depth and could now begin refilling with cleaned and uncontaminated
material.
353
-------
RWE Entsorgung Rehabilitation Friedrich-Ernestine
areas of contamination
Whereas foundation residue - and thus the main source of contamination in
the smaller pit - was located solely in the Quaternary layer, the foundations
of.the previous construction in the larger excavation stretched down only
as far as the marl layer of upper cretaceous matter, the depth of which was
considerably less.
Crevices in the marl had allowed pollutants to spread, which is why all
exposed or excavated crevice material was contaminated. The crevice
contamination did not subside as the excavation depth increased.
Excavation was therefore continued until directly above the first
groundwater level. In order to prevent additional contamination which had
not been removed during excavation, but was still mobile, from being
carried further via the groundwater, we decided, in agreement with the
authorities, to lay a drainage system and pump shaft into the aquifer along
the entire length of the excavation.
This was first and foremost .for safety purposes, but it also intended as an
additional rehabilitation measure. However, the water and contamination
levels have so far been extremely low.
354
-------
Both the surveyor and the authorities, confirmed that the rehabilitation
objective had been reached in the larger pit after installation of the drainage
system. The pit was then refilled with decontaminated soil.
By mid-August 1990 the entire project, which had lasted six months, was
completed with the resurfacing procedure.
in all, 18,000 m3 of soil was moved in both pits. Of this, 27,000 tonnes, or
approx. 15,000 m ^ was thermically cleaned by ATM, who had no trouble
adhering to the contractually agreed cleaning output (PAH < 1 ppm).
10. Costs
Since the site is situated right in the centre of residential and office
buildings, special measures were required to protect the personal, the
environement and the buildings. As a result, the conventional structural
engineering work was substantially more expensive than for comparable
construction work.
Total costs amounted to DM 17 million with soil cleaning and excavation
accomiting for the lion' s share of DM 10 million and about DM 5 million
respectivaly.
Soil cleaning
and transport
Construction works
Excavation lining
Ventilation tents
and storage
Experts
Analysis
Total costs
9.950.000 DM
4.850.000 DM
1.290.000 DM
440.000 Diyi
340.000 DM
145.000 DM
17.01 5.000 DM
BWE Entsorgung Rehabilitation i Friedrich-Ernestine
cost distribution
355
-------
DM 1,3 million had to be .spent on the excavation lining required for static
reasons and DM 440.000,™ were necessary for the tents and the ventilation
systems.
The specific costs related to the corresponding construction quantities
were as follows:
Soil cleaning
Transport
Excavation
Experts + analysis
Total costs
270.- DM/t
100.- DM/t
155.- DM/t
20.- DM/t
545.- DM/t
RWE Entsorgung
^Rehabilitation Friedrich-Ernestine
specific costs
356
-------
SANITATION OF THE
CRESOL ACCIDENT
IN SYLSBEK
Presentation at the Third Forum on
Innovative Hazardous Waste Treatment Technologies
Dallas, June 1991
357
-------
Sanitation of the cresol accident in the area
S y 1 s b e k
Highway Al, West-Germany
Dipl.Ing. Hannes Parti
TBU GmbH Innsbruck
6020 Innsbruck, Defreggerstr. 18
Tel. 0512/493733, Fax 493022
Dipl.Ing. Christoph Tiebel
ITU GmbH Hamburg
2000 Hamburg 11, Deichstr. 9
040/362677, Fax 378015
Descriptors: Cresol; neuro-toxine; contamination of soil, water and air; hazardous
waste; protection of labour.
Summary
After the spill of a considerable amount of toxic substance there was
not only the need of evaluating the degree of contamination and to treat
and dispose the contaminated soil in a way which would be accepted by
the population of the district, but also to protect the working staff and
the surrounding environment.
Introduction
In January 1989 the driver of a hazardous material transporter spilled about 5,000
liters of ortho-cresol (ortho-methylphenol) onto a parking area of the highway. The
toxic substance caused the death of fish in the nearby creeks, the neighbouring
houses were evacuated, the road closed.
Cresol is used for the production of resins, dis-
infectants, pesticides and tanning agents
(Gefahrstoffverordnung, 1988). Cresol is a strong
neuro-toxine; at temperatures around 30 Celsius
it is a yellowish glutinous liquid, at lower tempe-
ratures it becomes more and more solid, turning
into white to yellow crystals, but still poisoning
the environment through athmospheric
emissions.
358
Fig. 1: Ortho-cresol
-------
Cresol is dangerous to humans not only when inhaled but also by diffusing through
the skin. First signs of intoxication are headache, chest pain and drowsiness. The
LD50 - concentration in water is between 1 and 20 mg/1 for fish. The MAK - value
(direct translation: maximum work place concentration) for Germany is 5 ppm (M
118, 1987), that is around 22 mg/m3. The German MAK - values are comparable
with the US PEL -values (permissible exposure limit over an 8 hour time-weighted
average to which any employee may be exposed; for o-cresol also 5 ppm). MIKD-
value (long-time immission) is 0,2 mg/m3 or 0,05 ppm, MIKKvalue (short-time
immission) 0,6 mg/m3 or 0,15 ppm (Leithe, 1968).
After unsuccessful attempts to get the mostly solidified substance off the bitumen
and out of the drainage system, the bitumen was cut off and piled with the soil
from underneath - unfortunately on a cresol-permeable foliage.
The only analysis of the contaminated material (because of the dangerous emis-
sion - situation) at that time showed a concentration of 22.000 ppm for an estima-
ted 60 m3 and 63 ppm for about 1.500 m3.
*.
The next weeks were filled with the search for disposal possibilities. During this
time immediate measurements to reduce toxic gaseous emissions were made: Two
tents were built over the area. This step was successful: The concentration of
immission was reduced from 5 -10 mg/m3 to less than 0,5 mg/m3.
Calling in of TBU/ITU
Because of the fact that the contaminated soil would not be accepted anywhere the
idea of biological on-site-technique was brought up. TBU/ITU got the job of pro-
viding concepts, managing and supervising the sanitation (the accident happened in
the 5th week, our job started in the 21st week). TBU/ITU also worked out a
feasibility study in which all sanitation- and disposal possibilities were discussed.
Results of the study: 500 tonnes of higher contaminated material should be brought
to a hazardous-waste dump and disposed there in the ash of waste incineration
plants; the lower contaminated material should be mixed with fresh compost,
stored in ventilated containers on site and allow the cresol to be biologically
degraded (comp. Beyer et al., 1988; Deeley et al., 1985; Molin et al., 1985; Rubin et
al. 1984; Slump et al., 1984). TBU/ITU did laboratory research and successfully
repeated the experiment in technical dimensions.
. 359
-------
Circumstances did not allow the start of this solution: Inhabitants and politicians of
the area were afraid of this 'combination of poison and bacteria' and wanted to
have a fast solution to getting rid of the problem. After quite a few discussions the
majority believed that a washing of higher contaminated soil in special machines,
and delivery of the washed material plus the low contaminated soil in gasproof 'big
bags' to the hazardous-waste dump would be the best way. The estimated concen-
tration of the one and only analysis (from the first days after the accident) was set
as the contamination limit for the delivered material.
Sanitation, part I:
TBU/ITU installed a laboratory on the site and monitored the air inside the tent
and in the area regularly. TBU/ITU designed biological filters which were instal-
led, connected to ventilators producing a partial vacuum inside the tent. The filters
for the ventilation of the tent worked more than satisfactorily for the whole time.
Except the fact that a comparison between fig. 2 and fig. 3/4 shows the efficiency
of the 5 installed filters (filter material: waste compost from a nearby waste-treat-
ment plant) there is the need of discussing exposure limits: Cresol is a pretty
unknown substance and not many people have to deal with it. Fig. 2 shows that the
average weekly cresol-concentration inside the tent exceeded the MAK - value
only in two weeks during the whole sanitation.
30
26
20
15
10
mg/m3
a,c7*i=if~af^fza / /
27 29 31 33 35 37 39 41 43 45 47 49 51 1 3 6 7
Calendar - Weeks
1989 1990
Fig. 2: Cresol-concentration inside the tent
360
-------
ug/m3
12UU -
400-
| \
p
HR
1
-
;
||
' PI
x /
_, /
7\
/
p.
00I0R
27 29 31 33 35 37 39 41 43 45 47 49 51 1 3 5 7
Calendar - Week
1989
1990
Fig. 3: Cresol-concentration of the biofilter-outputs
But the experience showed that already the immissions outside (concentrations 100
times lower (fig. 4) than inside!) caused slight signs of intoxication when working
there the whole day. To work in an environment with athmospheric cresol-concen-
trations of just a fifth to a tenth of the MAK (PEL-) value would be unacceptable
and cause serious damage to health.
160
140
ug/m3
21 23 25 27 29 31 33 36 37 39 41 43 46 47 49 51 1 3 5 7
Calendar - Weeks
1989 1990
Fig. 4: Immissions in the area outside the tent
361
-------
In addition to that we examined the contaminated soil including the washed bat-
ches, which had to be released for package and delivery to the dump. Because of
the order not to deliver any material over a value of 63 ppm to the hazardous waste
dump (which is very low; the setting of this value was a decision made more out of
political than of security reasons) controls and examinations had to be made of
defined batches (which were sieved < 100 mm - especially batches from excava-
tion - because of the danger of leakage in big bags) inside the tent and finally
examined again in the bags on the other side of the bag-filling station. The labora-
tory work had to be done quick because on average there were around 30 to 40
samples per day to analyse (just soil) and results to be brought out on the following
day. The method used was the determination of the phenol index (DIN 38409,
1984).
For the purpose of organising and supervising the work and all safety precautions
TBU/ITU installed a sealed container inside the tent which was accessible from
the outside and stayed under slight overcompression.
ppm Cresol
I4UU •
n -
•f-i — a
$
$
$
^
H
/?'
// .
* «r
'•• 42 % add 78 %
1 V
I A
^
^
f'^'l-
^
s. 50 %
vv 9 % * *
^\ 8 % o %
X-H:T
O v r''x *^ vl-' '*• 'vl v [''>( ^> v"^"
£. A f^ O /j/^, M-. X '/^ O /^
Charge 71/72 12.-14.
B B'
Name of the batch
B"
B'"
efficiency
Cone, of batch
Trend
1 x, 2 x washed one time, two times,...
Fig. 5: 'Non-efficiency of soil-washing'
After 3 weeks there was evidence of the fact that the contamination of the stock-
pile was a lot higher than estimated by other people at an earlier stage, and the
washing-unit did not work well, mainly because it was designed for gravel and
debris and not for soil with a considerable content of clay minerals (see fig. 5).
The soil monitoring showed that there was also a lot of contaminated material
362
-------
under the stockpile. The immediate 'big stop' for this sort of sanitation came when
the company which was in charge of the soil washing and -movement packed not
released soil and the contamination limit for the delivered material was over-
stepped.
* •'
At this point there were the following possibilities:
* continue very slowly and with enormous costs (e.g. treating of washing-water)
* increase the contamination limit for delivery in 'big bags' to the dump (from the
environmental view o.k., but politically not acceptable)
* a different disposal method
Because of the fact that the whole case was very unusual, all involved administra-
tion offices were absolutely inflexible: they were bound to wait for political deci-
sions, and each step of decision making took a considerable amount of time.
So this was the point were TBU/ITU started a complete ground-water monitoring
program in the area. Earlier examinations in the nearby creek and fountains didn't
show any contamination, but apparently the substance migrated further into the
underground even if the content of clay in those layers of soil was very high.
Sanitation, part II:
After one month of consideration the responsible people chose the possibility to
ship the higher contaminated material to the Netherlands to treat it thermally, al-
though nobody really liked the idea of exporting hazardous waste.
Finally 1.850 tonnes of soil went to an incineration plant in the Netherlands and
3.650 tonnes of (lower contaminated) soil in big-bags to the toxic waste dump in
Rondeshagen.
363
-------
Financial aspects
There is no doubt that some of the above costs are markedly higher than expected.
Nevertheless it is impossible to objectively compare them with costs of other sani-
tation projects because - fortunately - there is hardly *a situation comparable to the
one shown here. Especially the fact that always again there was a lot of stoppage
due to long decision and/or permission periods (so that it took about a year to
finish the job) rose some of the costs evidently.
The costs of the sanitation - about 5 million US-Dollars - split up as following:
US-$
First measurements (traffic, evacuation, etc.)
Gasproof suits (disposable), etc.
Equipment (tent, machinery, etc.)
Excavations, drillings
Gravel washing (incl. waste water treatment)
Laboratories, engineering
Material to toxic waste dump (incl. packing)
Material to incineration (incl. oackine)
750.000,
250.000,
500.000,
500.000,
500.000,
500.000,
800.000,
1.000.000,
Sum
4.800.000.-
16 %
Measurements
21 %
Work on sitej
Transport/Disposal
38 %
16 %
Equipment
10 %
Lab. &
Engineering
Fig. 6: Cost distribution
364
-------
Protection of labour
TBU/ITU was in charge of the protection of labour on the sanitation site, but
many different institutions were 'somehow' responsible for the according regula-
tions, too. This also caused a few delays with the sanitation job because every
change of conditions or equipment needed to be approved by several administra-
tive institutions (TBG, GAA, BAMD, AM, etc.).
Some main precautions for the protection of labour:
* Everybody working on the site had to have the according medical examination.
* All machines working on the site had to have an installed pressure-ventilation
with activated carbon filters.
* Everybody on the site had to get sufficient instruction about the special dangers
and necessities while working on the site.
* Respiratory equipment and gas-proof suits had to be worn inside the tent. In the
summer time very high temperatures built up inside the tent so that the maxi-
mum working time had to be restricted to 30 min. with at least one and "a half
hours of rest. This fact slowed down the sanitation progress a lot. In those weeks
the company supporting the gas-proof suits developed a new generation of suits;
the difference was that in addition to the support-ventilation for the gas-masks
there was a smaller second 'blower' mounted to the back which pumped air
straight into the suit. This brought improved comfort so the maximum working
time could be prolonged to two hours with only 30 min. of rest, for at the most
three times a day.
The experience showed that the main problem with the protection of labour is that
all those precautions have to be stipulated before work starts because the com-
missioner must be able to estimate the costs. Still there must always be enough
leeway to add or change precautions quickly during the project responding to
changed conditions. <•••
Therefore ITU GmbH Berlin-Hamburg-Munchen-Saarbriicken and TBU GmbH
Innsbruck suggest that the responsible government institutions establish a commit-
tee which can help to improve efficiency of environmental protection measures
and sanitation jobs for possible similar accidents in a manner of decision making
outside the usual bureaucratic paths.
365 . f
-------
Fig. 7: Taking soil samples inside the tent
Literature
,4
Beyer.M., Schacht S., Klein J. (1988): Bergbau-Forschung GmbH Abteilung Ent-
sorgungstechnik, Essen. Sonderdruck aus Bio-Engineering 2/88, S. 120 -125.
Deeley G., Skierkowsky P., Robertson J. (1984): Biodegradation of (14C)Phenol in
Secondary Sewage and Landfill Leachate Measured by Double Vial Radiorespi-
rometry. Applied and Environmental Microbiology, Apr. 1985.
366
-------
DIN 38409 (1984): Teil 16, Bestimmung des Phenol-Index.
Gefahrstoffverordnung (1988): Datenblatter fur gefahrliche Arbeitsstoffe nach der
Gefahrstoffverordnung. Hamburg, 1988.
Kuhn-Birett (1980): Merkblatter fiir gefahrliche Gefahrenstoffe, Erg.Lfg. 6/80.
Leithe (1968): Die Analyse der Luft und ihrer Verunreinigung in der freien Ath-
mosphare und am Arbeitsplatz. Wissenschaftliche Verlagsgesellschaft Stuttgart.
Lewis D., Kollig H., Hodson R. (1985): Nutrient Limitation of Microbiological
Populations to Chemical Transformations. Applied and Environmental Micro-'
biology, Mar. 1986.
. M 118 (1987): Merkblatt 118 des Handbuchs der gefahrlichen Giiter. Springer-
Verlag Berlin Heidelberg.
Molin., Nilsson I. (1985): Degradation of Phenol by Pseudomonas putida ATGC
11172 in Continuous Culture at Different Ratios of Biofilm Surface to Culture
Volume. Applied and Environmental Microbiology, Oct. 1985.
.•
NFPA (1982): Exposure Limits and Hazard Ratings. National Fire Protection
Association, USA.
TBG (Tiefbauberufsgenossenschaft), GAA (Gewerbeaufsichtsamt), BAMD
(Berufsgenossenschaftlicher Arbeitsmedizinischer Dienst), AM
(Autobahnmeisterei), and several 'Amter und Behorden'; e.g. (1987): Arbeiten
im Bereich kontaminierter Standorte - Mafinahmen zum Schutz der Beschaftig-
ten. TiefbauBerufsgenossenschaft, Miinchen.
Shimp R., Pfaender F. (1984): Influence of Easily Degradable Naturally Occurring
Carbon Substrates on Biodegradation of Monosubstituted Phenols by Aquatic
Bacteria. Applied and Environmental Microbiology, Feb. 1985.
Shimp R., Pfaender F. (1984): Influence of Naturally Occuring Humic Acids on
Biodegradation of Monosubstituted Phenols by Aquatic Bacteria. Applied and
Environmental Microbiology, Feb. 1985.
367
-------
SOIL SEPARATION, WASHING AND DISPOSAL;
A Low COST REMEDIATION PROCESS TRAIN
PRESENTED TO: THIRD FORUM ON INNOVATIVE
HAZARDOUS WASTE TREATMENT TECHNOLOGIES
JUNE 11 -13, 1991 DALLAS, TEXAS
BY
PETER G. HANNAK, ucc & p
SUNIL l. SHAH, ucc & p
JEAN P. LACOURSIERE, UCC & P
PHILIPPE GUERIN, LAVALIN ENV.
368
-------
Soil Separation. Washing and Disposal:
A Low Cost Remediation Process Train®
Authors:
Peter G. Hannak1, Sunii I. Shah2
Union Carbide Chemicals and Plastics Company Inc.
Technical Center
P.O. Box 8361, So. Charleston, WV. 25303
Jean Paul Lacoursiere3
Union Carbide Chemicals and Plastics Company Inc.
Montreal East, Quebec H IB 1A1
Philippe Guerin4
Lavalin Envirotech Inc.
Montreal, Quebec H3B 4P3
Abstract:
Construction activities at industrial sites often lead to cleanup needs. For an expansion of Union Carbide's
(UCC & P) Montreal East Plant an estimated 6750 m3 (12,000 Tons) of polycyclic aromatic hydrocarbon
(PAH) contaminated soil were excavated and stored under secure conditions. The storage facility
included a concrete pad and plastic cover to prevent release of, the contaminants and infiltration of the
precipitation, respectively.
This site is under the jurisdiction of The Quebec Ministry of the Environment (MENVIQ). The Plant was
ordered by MENVIQ to propose a cleanup technology and to initiate the cleanup by March, 1991.
MENVTQ required compliance with a list of criteria "B" . These criteria (in mg/Kg) limit Mineral Oil and
Grease levels to 1000, Naphthalene and Phenanthrene to 5, total PAHs to 10. Some specific PAHs,
assumed or proven to be carcinogenic, are limited to a maximum concentration of 1 mg/Kg.
Due to the heterogeneity of the soil, the screening analysis indicated that these levels for some samples
were exceeded.
Several remediation technology options were reviewed before the technology selection was made.
Treatment processes were studied and sites visited by the project team.
. For studying the feasible remediation options, a pilot scale particle separation study was carried out at
Union Carbide's Technical Center. It was hypothesized, by the members of the project team that
separation of the soil into size fractions would make the treatment simple, and could possibly reduce the
volume remaining for washing and/or off-site disposal. Thus, a sequential combination of size separation,
washing, and treatment/disposal technologies offered a viable, low-cost management option.
Based on the results of the laboratory, pilot, and field experiments, a full scale soil treatment process was
designed and implemented. The full scale implementation of the project was contracted to Lavalin
Envirotech Inc. of Montreal, Quebec.
© 1991 Union Carbide Chemicals & Plastics Technology Corporation. All rights reserved.
1 Project Scientist UCC & P, Technical Center
2Group Manager UCC & P, Technical Center
3EP Department Head at UCC & P, Montreal Plant
4Vice President, Lavaiin Envirotech Inc.
369
-------
The COSL of the process is anticipated at US S 110 for each ton of contaminated soil. The comparison with
alternative off site treatment and disposal options, our pre-treatment and soil washing strategy realizes a
saving of US S 140/Ton.
The screened and washed residue was shipped to a secure disposal site in Montreal's vicinity. The clean
residue of the process will be used for fill at the plant. The process includes plans for wash water
recycling collection and local disposal.
Introduction;
The construction of a new polyethylene production facility at Union Carbide's Montreal East Plant lead to
the excavation, removal and safe storage of an estimated 6750 m3 contaminated soil. The site is under the
jurisdiction of The Quebec Ministry of the Environment (MENVIQ).
MENVIQ established three levels of groundwater and soil contamination requiring specific actions.5
A: Background level or detection limit
B: Level above which a serious study must be considered and,
C: Level above which decontamination of the site must be considered.
UCC carried out a comprehensive environmental assessment of die contaminated site. The resulting
report recommended that the disposal of the excavated soil in accordance with MENVIQ Level B criteria
would be necessary. The Plant was ordered by MENVIQ to propose a clean up technology and to initiate
the clean up by March, 1991.
Soil Chnracterization;
The soil characterization was intended to determine the weight and volume of the contaminated soil to be
disposed. Furthermore a detailed sampling plan was prepared for the collection of representative samples
to determine the types and concentrations of the contaminants. To assure safe storage the soil had been
stored on a concrete pad and covered with reinforced plastic tarp.
Subsequently detailed sampling and characterization plans were developed in preparation of a pilot scale
treatability study at the site. Figure 1 shows the layout of the soil piles. Figure 2 shows the photographs of
the sampling trenches, whereas Figure 3 and 4 show the heterogeneity of the a pile.
Representative sampling was a major concern due to the heterogeneous nature of the contamination and
due to the fact that the soil contained large amounts of rocks, and pebbles. Furthermore, several types of
debris were encountered including wood panels and beams, paper boards and bags, plastic films, concrete
and brick pieces, metal rods, wires, drums and pipes and polyethylene lumps and pellets.
*
For representative sampling six locations were chosen, where approximately 0.9 m wide, 1.5 m long and
2 m deep sampling pits were excavated. Figure 1 also shows the location of the sampling pits.
5Ministere de I'Environment du Quebec: " Guide standard de caracterisation de terrains contamines",
1985.
370
-------
For the protection of the sampling personnel, for assuring compliance with the applicable air emission
regulation's, and for establishing potential material handling requirements the combustible vapor
concentration of the soil samples was also measured. A Gastechtor Model 1238 portable hydrocarbon
vapor analyzer was used for this survey. Two types of measurements were performed (i) Soil samples
were placed in a sealed mason jar and allowed to stand four hours to equilibrate before the measurement.
and (ii) the monitoring probe of the analyzer was inserted to the pile through the plastic cover and the
measured concentrations were recorded. The results indicated safe concentrations of combustible vapors.
The excavation was done by a backhoe and a 10 ton capacity truck. The dump truck was weighed before
and after loading. The sample volumes were derived from the size of the pit while the weight was
calculated as a difference of two weighings and also on the basis of the backhoe's bucket count. An
average bulk density of 1800 Kg/m3 was calculated.
Table 1 shows the result of an earlier soil chemical characterization. It also shows the respective
MENVIQ B and C criteria. The results of these previous studies showed, that the concentration of several
contaminants was in excess of criteria B some of which also exceeded criteria C.
TABLE 1
Characterization of the soil piles
Contaminant Sample Sample MENVIQ Criteria
mg/Kg B-9 S-4 B C
Mineral oil and grease
Naphthalene
Acenaphtalyne
Acenaphtene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo anthracene
Benzo fluoranthene
Benzopyrene
Total PAH
2400
652
57
82
64
124
ND
53
51
242
ND
ND
>1325
38100
4600
226
887
659
705
418
226
361
173
86
66
>8520
1000
5
10
10
10
5
10
10
10
1
1
1
20
5000
50
100
100
100
50
100
100
100
10
10
10
200
Project objectives and annrnarh;
The project objectives were adopted as follows:
To establish the volume and weight of the soil
To determine the grain size distribution
To evaluate the feasibility of size separation
371
-------
To provide technical information for the scale up of the pre-treatment
technology
To select a suitable and cost effective treatment technology
Our approach could be outlined as follows:
Assumption: Concentration of the PAHs is inversely proportional with the
grain-size
Statistical assessment of replicate samples
Scale up: Pilot trial - Field trial and pre-treatment
Pilot Screening/Bench Scale Experiments
A size separation pilot trial was carried out at UCC & P's Technical Center. The purpose of the screening
was to (i) separate boulders, improving treatability of the material, (ii) separate fractions with larger
panicle size that are assumed non hazardous due to the large proportion of inert, non-contaminated
material and (iii) separate material fractions for the further characterization work.
The equipment used was a Soitest brand, pilot scale, five fraction mechanical vibrator sieve with
exchangeable trays. The vibration rate was adjusted to medium position. The screen opening of the four
trays chosen were: 1/2,3/4,1, and 2 inches. All trays and the pan were tared and weighed before and
after the experiment. The weight of the contaminated soil used for the experiment was 56 pounds.
The results of the test are shown in Figure 4.
The separation of the soil was relatively easy. During the run, no plugging of the screens was
encountered. The larger fractions (>!") did not contain any soil agglomerates even after 2 minutes of
vibration time. Only boulders and rocks were retained on the 1" and 2" screens. However, some rocks
were coated with the sticky soil residue, in spots up to 1/8 " in thickness.
These results demonstrated that,
* the separation of this waste is feasible
* 80% separation efficiency is attainable within 1 minute , however additional
processing could improve the resolution of the separation.
The experiments have also provided size distribution (Figure 5) .information important for process
economic estimates.
Since these preliminary results indicated the feasibility of a field scale particle sizing/separation
operation a site test was prepared and carried out.
Site pilot test
The major phases of the site test included (Figure 6):
* ; representative sampling
* dry screening
* wet screening
* analytical characterization
372
-------
Details of the representative sampling has been given earlier.
To accomplish dry screening the bulk samples were passed through two large screens made of welded
stainless steel bars. The first screen had openings of six inches, the second screen had openings of four
inches. Material retained on the screens was weighed using a 5000 Ibs capacity "hook" scale. Material
retained on the 6" screen and coarser than 12 inches was manually segregated and weighed.
The soil passing the 4" screen was sub-sampled and submitted to a manual screening using commercial
field screens of nominal 3", 2" and 1" openings. All material was screened at natural moisture content
which was changing throughout the test due to hot and dry weather.
Washing of coarse particles was carried out to evaluate the proportion of fine soil particles adhering to
the various coarse fractions. Sub-samples of the coarse material were washed. The sub-samples were
weighed, washed on screens placed into open top barrels, allowed to dry and weighed again.
Washing was done using ditch water supplied by a pump delivering 0.5 L/sec at approximately 20 KPa
pressure. The volume of the wash water collected varied from 125 to 180 L. Washing was performed in a
manner to approximate performance of high pressure jets on moving screens. Although not all the soil
particles could be removed, the washed rock pieces were largely clean with some occasional signs of
adhering soil and hydrocarbon stains.
Drying was done by spreading and rotating the soil on a clean surface. The drying took place in 2- 6
hours.
The analytical program encompassed the analysis of 35 soil and 12 water samples. The types of soil
analyses were as follows: oil and grease, PAHs, metals, and sulfur. Water samples were analyzed for
PAHs, MAHs, phenols, COD, BOD5, pH, alkalinity and conductivity.
Figures 7 and 8 present the critical findings of the experimental work. It was demonstrated that dry
screening results in the retention of marginally contaminated particles on the 4" screen. Supplementary
soil washing will result in compliance of ail contaminated soil fractions retained on the 2" screen with the
exception of benzo-anthracene due to its low 1 mg/Kg regulatory limit.
It was concluded that a combination of screening, soil washing using water and secure disposal is a
feasible option for the treatment and disposal of the contaminated soil.
Soil treatment and disposal
Alternative treatment and disposal options were reviewed including: incineration, low temperature
thermal treatment, biological treatment, solidification/stabilization.and secure disposal without pre-
treatment. The project staff visited operating and pilot facilities. Several process options were excluded
due to their experimental states. Some others were excluded because of their high cost. The detailed
selection process will not be discussed here.
For the secure disposal of the pre-treated but yet contaminated soil a MENVIQ permitted disposal site
near Montreal, Quebec was selected. This multi cell site has a well designed triple liner system, landfill
cover and it is equipped with water treatment and recycling facilities.
A conceptual process design (Figure 9) was developed to perform the full scale treatment of the
contaminated soil. Lavalin Envirotech Inc. was selected as the primary contractor for the treatment and
disposal of the soil.
373
-------
Concurrent soil washing experiments were performed by using cationic. anionic and non-ionic surfactants
at UCC's Technical Center for investigating the potential application of surfactants. Recently developed
and yet evolving in-situ soil flashing models6'7'8 were taken from the literature and modified to a
sequential batch washing scenario.
The results showed no significant advantage of the application of surfactants, and in addition, there
appeared to be difficulties in recycling the surfactants. Thus the surfactant aided washing attempts were
terminated.
The treatment started in October 1990 and was temporary halted due to the winter conditions in
Montreal. However, cold weather improved the feasibility of the dry screening. The weather conditions
between 5 and 15 °C were found the most suitable for the separation of this sticky soil.
Preliminary data suggest that after the physical separation and washing at least 25 percent of the soil will
be clean enough for reuse or replacement at the site of UCC & P's Montreal Plant.
Experiments are on-going to characterize the contamination level of the wash water. Several waste water
treatment options are being considered including on and off site treatment
The cost of the process averaged at USS 110/Ton of contaminated soil.The savings resulting from the
selected process train may be estimated at USS 140/Ton using a USS 250. esu'mated cost for an
alternative solidification/stabilization solution.
The treatment and disposal of the soil is expected to be completed by the time of this presentation.
6WUson, J. David: "Soil Clean Up by In-situ Surfactant Flushing. I. Mathematical Modeling". Separation
Science and Technology, 24 (11) pp.863-892, 1989.
7Wayt, J. Howard and Wilson, J. David: "Soil Clean Up by In-situ Surfactant Flushing. II. Theory of
Micellar Solubilization". Separation Science and Technology, 24 (12&13), pp. 905-937, 1989.
8GannoncO. Keith, atal.: "Soil Clean Up by In-situ Surfactant Flushing. III. Laboratory Results".
Separation Science and Technology, 24 (14), PP. 1073-1094, 1989.
374
-------
"WORK AREA ON APRON
N
f!T."
56 m
•:-x
•:-x
NO COVER
ON -SOIL
3.35 • • • S.37
;X CENTER PILE .v.v.v.v.v.v.-.v.-.v.-.v.-.v.v.v.vi- .-\
A
2.00
S.2Z
• 3.78 •s.ae
•s.sa
i i
Jt-o;
-------
\
oo
^J
CTl
FIGURE 2
Overview of sampling trench No6
Silty soil with pieces of wood, steel and traces of hydrocarbon
Vertical scale: 1.2 meter
Horizontal scale: 0.75 motor
-------
FIGURE 3
Sampling Site No4
East side. Scale 0.6 X 0.6 meters
'"*• w
FIGURE 4
Sample No4
Silty Soil with pieces of steel, wood, polyethylene
Scale: 1.2 meter
377
-------
CHANGE IN PARTICLE SIZE DISTRIBUTION BY SIEVING
CO
^•J
00
Weight % retained
Retaining Screen
D
Screening time (min.)
FIGURE 5
-------
PROCESS CHART
SOIL CHARACTERIZATION
PILES
BULK
SAMPLES (6)
WEIGHING
COARSE SEPARATION
SUBSAMPLE
"DRY" SIEVING
DUPLICATE *
/\ A
\ /6-12"\
4-6'
WASHING (WEIGHED SUBSAMPLE)
FIGURE 6
379
*: LABORATORY SAMPLE
COLLECTED
-------
to
00
o
Concentration in Washed Soil
ENVIQcriteria: Total PAH 20, MOGT60W(mg/kg)
3>JRetaining screen
Mineral oil and grease
Total PAH
FIGURE?
-------
I8£
e
vs
P3
00
(Q
5?
(Q
m
5
CD
O
f
53'
Naphtalene
Acenaphtalyne
Acenaphtene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo anthracene
Benzo fluoranthene
V V V A
co ro ->• _,.
-------
CO
00
ro
TO STORAGE
PAD Jill
FEEDIHO It SCREENING CAPACITY
PUUPIHC CAPACITY
TANK CAPACmf
25m3/bour
SSmS/hour
23m3
FIGURE 9
( °"cl- UNION CARBIDE OF CANADA LTD
Title:
Conceived Dv Iwm-r !•>
V A. CHEVALIER 0. G
PRE-TREATMENT OF THE CONTAMINATED SOIL
- PROCESS DIAGRAM -
/enlifd D>
RANI A. CHEVALIER,
bale-
27/08/90
file nc :>coi?
36-247-005 N.
..Ol-'t '- =
t.S. X J
-------
INNOVATIVE GROUND WATER AND SOIL REMEDIATION
AT THE
USCG AIR STATION
TRAVERSE CITY, MICHIGAN
INTRODUCTION
Soils and .ground water beneath the Traverse City Air Station have been contaminated by
liquid hydrocarbon fuel from two sources. To remediate this contamination, innovative
processes are being employed. The first of these projects uses the injection of hydrogen
peroxide to enhance aerobic biodegradation. The second project is a study in anaerobic
biodegradation and uses nitrate as the electron acceptor in the metabolism of the
contaminants. Upcoming projects include the implementation of vacuum extraction and in
situ ^bio-treatment of vapors during vacuum extraction. Clean up of the site using these
technologies is a cooperative effort between the United States Coast Guard as proprietor,
the US EPA R.S. Kerr Laboratory providing research support and The Traverse Group
providing engineering and design support.
SITE HISTORY
The first confirmed use of the present day air station site was by the United States Navy
from 1942 to 1945 for secret research and development of pilotless aircraft. The Coast
Guard began using the site in 1944 and has since equipped it with a full range of aircraft and
maintenance facilities. Contamination was first identified in 1980 when the water wells of
residents hydraulically downgradient from the air station became contaminated with
hydrocarbon compounds.
Suspected sources of contamination were the air station, the municipal airport located
southwest of the air station and industrial facilities between the air station and the residential
area. Subsequent site assessments indicated that contamination from the air station was
contributing to the water well fouling problems downgradient. After a detailed historical
review of the air station uses, all possible sources ie; underground tanks, drywells, pits, spills,
etc. were investigated. Ultimately, research conducted during litigation with the state
regarding clean-up identified the source of contamination as the instantaneous release of
gasoline in 1969, when a pressurized line from an underground storage tank ruptured and
released at least 2500 gallons of aviation gas. Identification of this source, located near the
NW corner of the present hangar, matched models hypothesizing the contaminant source
location.
383
-------
USCG — Groundwater and Soil Remediation
A second source of contamination at the air station was identified in 1985. Underground
storage tanks containing JP-4 jet fuel and located south of the hangar were found to be
leaking.
The contaminated ground water and product plumes resulting from both releases are
currently contained by a purge and treat system using a row of interdiction wells transecting
each plume. Remediation of the dissolved contaminant plumes by purge and treat has
halted the plume migration, but contamination of the ground water continues because of
water table fluctuations and vertical recharge through the contaminated interval. In-situ
bioremediation is among the technologies being employed in the clean-up of contaminated
soils above the water table in the source areas.
HYDROGEN PEROXIDE ENHANCED rAEROBIQ BIG-RESTORATION PROJECT
In 1988, the United States Environmental Protection Agency (USEPA) and the United
States Coast Guard (USCG) installed an aerobic in-situ biorestoration project (referred to
as Bio I) near the source of the aviation gas release. This project was designed by the
USEPA and the Traverse Group, Inc. (TGI) to demonstrate the use of hydrogen peroxide
as the oxygen source (combined w/ phosphorous, nitrogen and trace elements) in enhancing
aerobic biodegradation of organic contaminants for large scale clean up of hydrocarbon fuel
spills. Alkylbenzenes are the object of the regulatory concern at the site, so the
bioremediation is considered complete when their concentration is brought to a level
specified by the Michigan Departmenfof Natural Resources (MDNR).
Initiation
Soil cores were acquired from the source area of the aviation gas spill to determine the
vertical and horizontal extent of contamination, the concentration of total hydrocarbons in
the contaminated interval and concentrations of individual alkylbenzenes. This information
was used to identify the most contaminated flow path through the spill. A series of
miniature monitoring wells was then installed along and below the most contaminated flow
path. These wells were constructed of 3/8-inch stainless steel with a stainless steel screen
that was 6 inches long.
The design of this system required installation of a series of fresh water injection wells and
a series of injection wells for injecting water amended with hydrogen peroxide and nutrients
(Figure 1). The wells are installed in parallel rows approximately one foot apart, transecting
the treatment area and the hydraulic gradient. Fresh water is injected in the upgradient row
of wells below the water table. Water amended with hydrogen peroxide and nutrients is
injected in the downgradient row of wells. This water contains 380 mg/liter of ammonium
chloride, 190 mg/liter of disodium phosphate and 190 mg/liter of potassium phosphate. The
temperature of the water is 11-12 degrees Centigrade, and the pH near neutrality.
384
-------
CO
00
o m n i
o o -*i •-• 1
I'll Ui
• -« - - CO,— S> BW
31 ,. n - "Ti
O • Y* O 35 ~ *
m l-oi I
Ji n n •* ~ 1
« r i * « I ' I
c m c. y H
ntn^rm •*--•* u
O H " O ~~ M fl
z KI n -» Tl - OI n
- jo a - m - H
-< r o I" - »
on^r "1-^lr
- r m t ~* in /V
z r m m I/
" ° r X
o . , yi
ro K/^ n)
o , a S co
' S "X^ "• OB
^ CD ""
PI m 1
r r - I
r r CONCRETE FIRE LflHE • . .
OJ
5 . . • • D
0 3 ' ' J,
n 3 Q)
T> 5 >1
2r S /
So r >^
-1 • 0 S
en 3) _ *^
H W 0) .g ,
dr e «
o""" Tl §
" 2"1 n C
jj°m ^ at n
<"r " Q i ' S
(rt O o • **
(n 3j n ji
«On
HZo O
- -out T
-*? *" *n
«• C. ^ '*
nm fr 1
•-« *H 1 ^"^
o i -^ -n
3 * 1
Z M (0
n 1*
***
i
0}
C
(D
• 1
a »to
o m to
' a
•* i
to en
3> ru
CD •* tO • «CO
a o cf a
1 . 1 CD o '
«•»?.{,. w
3) •aj « J:: °
1 1 a 5 «5 !. ' • _. ~
n!
a
i
-3
3
o
T|
rt-C
^-1 *-* **'"**—• — IVI
V • 1 1 to . 0 ' TJ
•ni " S" tnn i
01 M -° J: -C
/ft
a
o
i
•CO
CO
a
1 «tD
M a
CO 1
V i" °
1-4
OI
-C
i +*i
0
HflNGER RDMINISTRRTIOH BUILDING
-------
USCG — Groundwater and Soil Remediation
Injection of the fresh water is necessary to mound the water table through the contaminated
zone (Figure 2). The mound serves two purposes. The first is to saturate the contaminated
zone as the water is the media for metabolite and nutrient transport. Secondly, mounding
keeps the peroxide/nutrient mix on top of the water table concentrating it in the
contaminated zone. Water for mounding is injected at approximately 3.1 gpm. This rate is
sufficient to build a mound of approximately 1 foot and saturate the one foot contaminated
zone. Monitoring of hydraulic and biological performance is accomplished by a combination
of monitoring and cluster wells located at various distances downgradient from the lines of
injection weUs.
The monitoring wells are equipped with dedicated pumps for sampling groundwater. The
cluster wells are groups of five or six 3/8" stainless steel tubings, each screened at one end
and each a different length. The cluster is installed in a borehole and backfilled. A
bentonite plug is placed between each cluster point to prevent communication between
points and to allow sampling of discrete vertical points throughout the aquifer.
Operation
Injection of fresh water for mound building began in February 1988. After approximately
one week of injection, in the first week of March, the injection stream was split and reduced
for chemical addition. An eleven gallon per minute stream was used to carry the chemical
mix and a 20 gallon per minute stream was used to maintain the mound. Nutrients
(ammonium chloride, disodium phosphate and potassium phosphate) and liquid oxygen were
added.
Use of Oxygen
Liquid oxygen was used initially to acclimate the biological activity of the aquifer to a higher
than normal oxygen level before adding hydrogen peroxide which could be toxic to the
microorganisms. Oxygen was injected for three months before hydrogen peroxide was used
as the oxygen source.
The aviation gas was confined to a narrow interval between 15 and 17 feet below the land
surface. This interval corresponds closely with the seasonal high and low water table at the
site. The concentration of fuel hydrocarbons in the most contaminated interval averaged
7,500 rug/kg of aquifer material. The porosity of the contaminated sand is 0.4, and its bulk
density is 0.2 g/cubic cm. Therefore, the water content of the aquifer is 0.2 liter/kg, and each
liter of pore water was in contact with 37,500 mg of fuel hydrocarbons.
386
-------
CROSS SECTION OF PILOT SCflLE UELLS
GO
00
•-J
ELE
IN
615-
610-
605-
600_
595-
590-
SRS-
.VHTION inutu i IUH
FEET UELLS v
UG-4.1 V,
•
•
1
• .
_
-
-
fl
-
i r
m
h~
1
n
m
V
— i
(
i (
! i
i
G-8 D
-
__ — —
" ~~~~
G-25 DG-
DG-37
m
.-4
ro
I
o
CO
•"••••••••.
i
(
(
>
>
'
_
49B DG_61 DG-109
ii — -
;; "1
i
> i
OJ
to
•
D
CO
' J
)
> (
m co
ro o
00 -<
i i
• i
Q iD
OQ CD
> P
• —4
i t
) (
> i
i j ^
H
1 UpTER TRBLES
n
1 — 4/20/88 HI
T - — 9/28/88 LOU
0' 10' 20' INITIRL
•
HO
III II 1 M 1 1 Mil 1 1 1 III!
RIZONTRL SCfiLE
v URTER TRBLE
FIGURE 2
-------
USCG — Groundwater and Soil Remediation
The oxygen demand for microbial respiration of total fuel hydrocarbons was estimated
assuming the following stoichiometry:
CH2 + 1.502 CO2 + H2O
The oxygen demand of the alkylbenzene fraction alone was estimated from:
CH + 1.2502 CO2 + 0.5 H2O
Of the 31 feet between the injection wells and the monitoring well at BD 31-2, 15 feet was
considered to be contaminated. Of the 50 feet to the next monitoring well, BD 50B-2, 35
feet was considered to be contaminated. The concentration of hydrocarbons, the length of
the contaminated portion of the flow path and the assumed stoichiometry for microbial
respiration were used to estimate the total oxygen required to remediate the flow paths to
the two monitoring wells.
Tracer Surveys
System hydraulic performance was monitored by periodic tracer surveys conducted after
oxygen and nutrient injection was started and by baseline and monthly sampling of the
monitoring and cluster wells. The tracer surveys used dissolved oxygen, chloride, ammonia
and phosphate to determine oxygen and nutrient transport rates and profiles. The seepage
velocity of the injected water in the aquifer averaged 10 feet per day. The cluster wells were
sampled during tracer surveys to measure breakthrough and vertical distribution through the
aquifer.
Sampling
Baseline sampling consisted of collecting samples from all monitoring wells, cluster wells, the
injection wells and the supply well. These samples were analyzed for BTEX, pH,
conductivity, dissolved oxygen (DO), chloride, phosphate, ammonia and methane.
Temperature was also taken, from buried soil probe located nine feet down gradient from
the injection well row and set two feet below the water table.
Monthly monitoring is conducted by collecting samples from the cluster and monitoring wells
and analyzing them for BTEX, DO, and hydrogen peroxide.
f
System Performance
Monthly monitoring of the cluster and monitoring wells is conducted to identify and track
the treatment front and vertical distribution of nutrients and peroxide. With this
information, the performance of the system can be monitored. Peroxide and DO levels
388
-------
USCG — Groundwater and Soil Remediation
identify the zones of varying levels of contamination and clean-up. Tracking these levels
through time indicates the rate and extent of clean-up achieved.
The interval between the injection wells and the monitoring wells was considered remediated
when detectable oxygen broke through and alkylbenzenes were below the detection limit (1
part per billion). The interval to the monitoring well at 31 feet was remediated after 220
days, and the interval to the monitoring well at 50 feet was remediated after 270 days.
The aquifer was purged of alkylbenzenes very quickly. Aviation gasoline is composed
primarily of branched-chain alkanes. The material spilled at Traverse City was 38%, 2,2,4-
trimethylpentane, 15% 2,2,5-trimethylhexane, 14% 2,3-dimethylpentane, 13% 2,4-
dimethylhexane, 7% 2,3-dimethylhexane and 5% 2,4-dimethylpentane. Only 10% of the
original spill was alkylbenzenes. The quantity of oxygen and hydrogen peroxide required to
remove alkylbenzenes from the wells agreed closely with the projected oxygen demand of
the alkylbenzenes alone.
To some degree, this indicated positive results. Some of the alkylbenzenes may have been
washed from the source area by simple physical weathering. Some of the alkylbenzenes may
have been removed by anaerobic biological processes before the front of oxygen swept
through. Water from anaerobic regions of the demonstration contained significant
concentrations of volatile fatty acids (by-product of degradation) and was visibly turbid with
microorganisms. The flow paths to the monitoring wells at 31 and 50 feet from the injection
wells were remediated when a small fraction of the oxygen demand of the spill had been
supplied.
To date, monitoring has shown clean-up of BTEX contamination out to the 81 foot gallery.
Cluster well profile data also indicates clean-up of contaminated zones by comparing
contaminated zone DO, which increases upon clean-up, with deeper clean zone DO.
Additionally, peroxide is monitored to determine how far oxygen can be carried as hydrogen
peroxide. The ability to transport oxygen as hydrogen peroxide to the contamination front
before the peroxide degrades is essential to the potential full scale application of this system.
Peroxide has been detected out to the 31 foot gallery.
In addition to the routine performance monitoring, the USEPA has conducted additional
tracer studies subsequent to the start-up tracer study and collected soil cores from the study
area. The tracer studies are used comparatively to identify changes in aquifer flow
characteristics (i.e., hydraulic conductivity, vertical flow profile, etc.) which may occur as a
result of bio-fouling, bio-enhancement, mineral precipitation, or chemical plugging.
389
-------
USCG — Groundwater and SoD Remediation
Results
JSJ
The spiU was cored in August of 1987 to provide information to design the demonstration,
then again in March of 1988, before the demonstration began, to define the initial
conditions. The proportion of alkylbenzenes in the spill declined modestly over that time
interval. This was probably due to anaerobic microbial degradation.
Shortly after the breakthrough of oxygen and disappearance of oxygen in monitoring well
BD 31-2, the area near the monitoring well was cored and analyzed for alkylbenzenes and
total fuel hydrocarbons. Results were compared to earlier cores to determine whether any
contaminants remained in the aquifer. The aliphatic hydrocarbons remained at their initial
concentration, but the alkylbenzenes were below the analytical detection limit. Subsequent
soil borings have shown elimination of alkylbenzenes out to approximately 80'.
390
-------
oo
IO
In the most contaminated interval:
Fuel contamination averages 7,500 mg / kg aquifer material.
f
Because 7.53 mg of hydrogen peroxide or 3.5 mg of oxygen
is required to completely oxidize 1.0 mg of fuel, this
corresponds to 56,000 mg of hydrogen peroxide or 26,000 mg
of oxygen per kg aquifer material.
Each kilogram of aquifer contains about 0.1 liter of water.
560,000 mg of hydrogen peroxide or 260,000 mg of oxygen
must be consumed to oxidize the fuel in contact with each liter
of aquifer water.
-------
Water Quality In Hydrogen
Peroxide Demonstration
08/16/89
co
ID
PO
Well Location
(feet from injection)
7
31
50
62
, 83
108
280
Benzene Toluene
nn / Ijtpr I
H-y / Hit'i \
<0.1 <0.1
<0.1 <0.1
<0.1 <0.1
<0.1 0.3
<0.1 0.3
<0.1 2.4
45.8 6,987
Ethyl-
benzene
lA/ofpr-N .......
<0.1
<0.1
1.0
1.7
12
6.4
97
Total
Xylenes
<0.1
0.8
10.4
37.2
367
393
1,452
Total
Avgas
mn / Ifn
1 1 ly / r\y
770
2,500
1 1 ,000
-------
Course of Bioremediation
Near Monitoring Well BD 31-2
CO
ID
GO
Core Date Conditions Total Total O2 Delivered
Hydrocarbons BTEX (mg / liter of
(mg / kg) (mg / kg) pore space)
50114 8-87 Preliminary 6,500 ±2,000 6041200 0
site charac-
terization
50AE4 10-88 Eight months 8,40012,000 <0.3 3,000
operation
50BD2 8-89 Eighteen 2,5001500 <0.3 28,000
months
operation
-------
Rates of Oxygen Consumption Along
the Most Contaminated Flow Path
Esti
mg 02 / liter day
co
10
Depletion of oxygen during
transit from:
Injection to 7B-2 (4/88)
7B-2 to 318-2(8/89)
Theoretical oxygen demand of
depleted gasoline near 31B-2
(10/88 to 8/89)
Column study (Huling and Bledsoe)
83
80
350
114
-------
/ _ / _ /
t
t7 |9A9"|
8 |9A9"]
2 |9A9"|
I
t _ /_
ID
CT>
CO
n
II
-------
96£
\
M
.i X
X
\
-------
NITRATE ENHANCED BIODEGRADATION PROJECT
In February 1985, four underground storage tanks comprising a fuel farm at the U.S. Coast
Guard facility in Traverse City, Michigan were found to be leaking JP-4 jet fuel, through
faulty piping, to a shallow underlying aquifer. By the time the leaks were discovered and the
tanks and contaminated soil were excavated, several thousand gallons of JP-4 jet fuel had
been lost to the subsurface. In one monitoring well alone, over 50 inches of free product
were recovered from the water table in the immediate site vicinity. A series of interdiction
wells was installed to contain the contaminated ground water. Although this successfully
retained the plume within the facility's boundaries, a large area predominantly underlying
the concrete apron and runway had been contaminated. Moreover, the seasonal rise and
decline of the water table has effectively smeared the contamination over a five to eight foot
depth interval and has resulted in residual saturation of JP-4 in the unsaturated zone as well
as free product at the water table.
The use of nitrate to promote the removal of aromatic hydrocarbons is being investigated
in the JP-4 fuel spill area (Bio II). A series of laboratory tests was conducted to evaluate
whether denitrification would be suitable for biorestoration of the aquifer and to determine
the conditions necessary for the process. Based on the results of the laboratory evaluation,
a pilot study is underway to determine the applicability to full scale site clean-up.
A 30 ft. x 30 ft. area within the contaminated area was selected for this study. The study
area is underlain by a contaminated zone approximately 5 feet thick immediately above the
water table. An infiltration gallery was placed at the study area to build a mound sufficient
to saturate the contaminated zone with water and evenly distribute nitrates and required
nutrients over the study area.
397
-------
USCG — Groundwater and Soil Remediation
Hydraulic computer modelling was conducted to design the infiltration system. Results of
the model indicated an infiltration rate of 300 gallons per minute would be required to
saturate the contaminated zone. To achieve this rate, a row of five withdrawal wells was
drilled along the line of existing interdiction wells. These wells are pumped at high rates (50
to 60 gpm) from the bottom of the aquifer to supply relatively clean infiltration water and
maintain a sufficient draw down so that contamination is captured at the interdiction wells.
Figure 3 and 4 are a map and cross-section showing well placement and configuration.
Monitoring of the system is accomplished with cluster and piezometric wells located in the
study area and down gradient, between the study area and withdrawal wells.
Infiltration of fresh water into the study area began on April 20,1989. The initial infiltration
rate was 50 gallons per minute. The system was brought to hydraulic equilibrium by
gradually increasing the fresh water infiltration rate until the desired mound was obtained.
Fresh water infiltration was then continued at a constant rate until water levels in the study
area and surrounding monitoring wells stabilized.
During the start-up period, baseline sampling was conducted. The first sampling occurred
prior to infiltration. Samples were collected from the cluster wells, interdiction wells and
infiltration supply wells and were analyzed for BTEX, DO, metals, conductivity, pH, nitrate,
nitrite, ammonia) alkalinity, sodium, and potassium.
After infiltration of fresh water began, samples were collected weekly to monitor BTEX
levels, DO and conductivity.
Once hydraulic equilibrium was achieved and before chemical infiltration began, a second
round of sampling was conducted, and the samples were analyzed for the same parameters
as the baseline sampling. This identified changes in the chemical characteristics of the
system resulting from the high infiltration rate flushing the contaminated zone.
In addition, soil cores were taken periodically during the study to evaluate the relative
contribution of flushing and biodegradation to the clean up process.
Tracer studies before and after chemical addition were used to determine the vertical flow
profile and transport velocities.
Nitrate was delivered to the contaminated zone as sodium nitrate at a nitrate concentration
of 45 ppm. The same nutrients as used in the aerobic study were used for Bio II but at
lower concentrations of 20 ppm for ammonium chloride, 10 ppm for disodium phosphate
398
-------
HANGAR/ADMINISTRATION
CO
VD
in
MW-21'
APRON
0 30
iillllllllllllll
SCALE IN FEET
MW-34-* BASE MONITORING WELL
PP-6* PUMPING WELL
N-2 • PIEZOMETRIC WELL
C3 • CLUSTER MONITORING WELL
24" RECOVERY WELL
N-3.N-3A
N-6 CLUSTER
N-4.N-4A
MW-20 Ct\
PP-5
••PP-5A
PP-6
"PP-6A
•PP-7A
PP-8
..PP-BA
SAMPLE
BUILDING
DIRECTION OF
GROUND WATER FLOW
PP-9A
PT-1
MW-34
N-2B
CD
N-5A.N-5B.N-5C
Figure 3
JP-4 BIODEGRADATION SITE
TRAVERSE CITY COAST GUARD STATION
LOCATION OF WELLS AND AUGER BORINGS
-------
CROSS SECTION OF BIO II STUDY AREA
2" PIEZOMETRIC WELL
ns
no
ui
UJ
u.
O
103 .
800 .
MS.
BQfl •
C WELL --v N.
CE
603.37 < '
\7 601.36 «'
^^ 599.36 <•
597.37 ..
595.37 «
593.36 1
*^\
'2M-2B N-5C
CD
i
rORAVtL IED
~»
i 603.37
601.40
" 599.42
N_5B N-5A N-1
^"
C
fj=-
' •
i
> 597.40
" 595.41
^ 593.83
C| CA
WntRATlOH
IUWW— v .
\ J •
M— ^ ^
/Tt?. SUIFtf fOWT
603.77 G04-03
601.77 601.77
599.80 600'12
597.81 598'12 '
595.81 596'13 '
593.82 594-11
^^
\. ^^^
^-3" WELL SCREEN
589.48'
•
-
;
i
; '
—
>-
t nc.AUMc.mio wcu.
N-UX- 2" PIEZOMETRIC
J-r^^ SURFACE ELEVATION 813.4'
CB CF
ssa==«.
603.67
601.65 601.59 '
599.69 599.63 (
597.66 597.56
595.63 595.57
593.67 593.60 '
i 591.61 .
cc
/raw
SU171C
601.78 (
599.84 '
597.80 '
595.75 •
593.78 '
591.78 '
__
s~
B 589.75'
•CLUSTER WELL (TYP)
Figure 4
REVISION DATE: 6/1/89
-------
USCG — Groundwater and Soil Remediation
and potassium phosphate each.
Performance
During the project, BTEX in the study area and infiltration and supply water, total
infiltration rate, individual pump pressures, total system pressure, incoming and infiltration
nitrate and nitrite levels were monitored, and adjustments made daily to assure stable
operation. In addition, water levels in the study area piezometers, pumping wells and
surrounding monitoring wells were measured daily. Once each week, water samples were
collected from cluster wells and pumping wells for analysis of BTEX, dissolved oxygen,
conductivity, pH, nitrate, nitrite, ammonia, chloride, ortho-phosphate and sulfate.
The infiltration rate was maintained at or near 200 pgm for the first three months when,
because of ponding in the study area, the flow rate was reduced to 150 gpm. Upon making
this adjustment, individual well flow rates were reapportioned. After three weeks the
infiltration rate was again reduced, because of continued ponding, to 130 gpm. After three
days at 130 gpm, daily water level monitoring in the study area indicated loss of saturation
in the contaminated interval so the rate was returned to 150 gpm while ponding recurred.
Because of continually dropping water levels in the piezometers, pore space plugging
beneath the gallery appeared to be the cause of the ponding. In June, two split spoon
samples were taken in the study area form immediately below the gallery to investigate the
plugging. One sample was 'opened on site and inspected visually. The pore spaces in the
top 6" -10" appeared plugged with a secondary mineral deposition. Calcium carbonate was
the suspected mineral and this was confirmed by its rapid reaction when contacted with
hydrochloric acid. The second core sample was taken to Kerr Lab for biological analysis to
investigate the possibility of bio-fouling. The results indicated there was no evidence of
excess bio-mass in the plugging material.
To monitor nitrate level and .biological usage, nitrate and nitrite concentrations were
measured in the combined pumping well stream before nitrate addition and downstream of
nitrate, addition. Sodium nitrate feed was continually adjusted to remain at or below the
regulatory nitrate limit of 10 mg/1 as nitrate nitrogen. Incoming water was monitored to
determine the extent of nitrate carry through and to adjust feed accordingly. Nitrites were
measured as an indication of denitrification (biological reduction of the nitrate).
Potassium chloride tracer studies conducted by EPA researchers indicated no change in
aquifer characteristics over time. Results indicated that the vertical flow component through
the study area extend to the bottom of the aquifer. This large vertical contact allowed most
of the aquifer between the study area and pumping wells to act as a bioreactor providing the
necessary retention time for biotreatment.
401
-------
USCG — Groundwater and Soil Remediation
After three months of hydraulic loading and two months of nitrate and nutrient addition,
virtually all of the BTEX was removed from the study area soils. Three processes appear
to have been the major removal mechanisms: aerobic oxidation, denitrification, and water
flushing with an eventual loss from the hydraulic system. Removal of BTEX from the
contaminated soil may have been due primarily to water washing, but there is strong
evidence that the majority of the dissolved contaminants were then biodegraded within the
recirculating hydraulic loop. With the exception of O-xylene, it appears that BTEX was
actively degraded either aerobically or by denitrification as water was recirculated between
the infiltration area and the purge wells.
Nitrate levels in the injection water varied from 9-14 mg/L as nitrogen during chemical
injection. By the time infiltrated water reached the water table, concentrations were reduced
to 2-4 mg/L of nitrate as nitrogen. Very little nitrate was returned from the purge wells with
source water concentrations typically being less than 2 mg/L nitrate as N. Nitrate
concentrations rapidly increased when injection first began, reaching a peak concentration
after 20 days, then rapidly declined. This data may represent the lag phase that was
observed in the microcosm study conducted by RSKERL or an inhibition of denitrification
due to dissolved oxygen. Nitrite concentrations also rose sharply and peaked after twenty
days. Within ten days after the peak concentrations, nitrite had declined. The return water
from the purge wells also showed a peak of nitrite at 20 days after injection began and then
rapidly declined over the next twenty days. Nitrite levels remained below 1.0 mg/L as
nitrogen for all points within the study area for the remainder of the project.
After a 20 to 30 day lag phase denitrification appeared to be very rapid within the study
area. Samples of the escaping gas bubbles in the infiltration area consisted almost entirely
of nitrogen with no evidence of ammonia production from nitrate reduction.
Results
Clean-up of the study area was successful although the actual contaminant removal resulted
from a variety of processes. BTEX removal occurred sequentially with benzene disappearing
first, followed by toluene, ethylbenzene, and the xylenes. The removal sequence followed
results observed in laboratory denitrification biodegradation experiments indicating that
denitrification was one of the processes occurring. The disappearance of benzene occurred
before the addition of nitrate to the infiltration stream indicating that benzene removal
resulted from aerobic processes. Other aromatic compound removal however coincided with
the reduction in dissolved oxygen and the addition of nitrate.
In addition to BTEX removal, total hydrocarbon reduction occurred as a result of flushing
and miscible and immiscible displacement of the residual fuel. Following the start of fresh
water infiltration, free product recovery increased as a result of the additional drawdown
near the pumping wells and flushing through the contaminated interval.
402
-------
C • Stainless sleel cluster wells
-P>
o
co
Elev
(meters)
188 -T-
187
186 -J-
185 4-
184
183
182 --
181 —
180 --
CE CD
CC
CA CB
Core
Interval
TTt
4
5
e
Cluster
well lines
Infiltration
feed line
Contaminated
Figure 13. Relationship between the infiltration gallery, tht monitoring vtli
clusters) and th« fuel-contaminated interval in the aquifer.
-------
Dissolved Oxygen
—•— Injection Water
— Level 2
Level 4
Level 6
Begin Nitrate
Injection
0.0 H
Time (days)
Figure 14. Concentration of dissolved oxygen In the reclrculatlon vater, and
•t various levels In the veil clusters. See Figure « for the depth of the
wells. ^ The veils are numbered froa shallov to deep. Level 2 is in the Jp!*
contasunated interval, level 4 is Just under the conta-lnated interval and
level 6 is deeper under the contaminated Interval. LSD Is the least
significant difference betveen.aeans at the 90* confidence level.
404
-------
Nitrate-Nitrogen
Begin Nitrate
Injection
14
12-
Injection Water
Level 2
Level 4
40
r
60 80
Time (days)
100
120
Figure 15. Concentration of nitrate-nitrogen In the recirculation vater, and
at various levels in the veil clusters. See Figure 4 for the depth of the
veils. The veils are numbered froa shallov tc deep. Level 2 is in the JP-4
contaminated interval, level 4 is just under the contanlnated interval and
level 6 is deeper under the contaminated interval. LSD is the least
significant difference between. Beans at the 95Z confidence level.
405
-------
3.0
Nitrite-Nitrogen
Injection Water
Level 2
Level 4
Level 6
r
60 80
Time (days)
Figure 16. Concentration of nitrite-nitrogen in-the reclrculatlon vater, and
at various levels in the well clusters. See Figure A for the depth of the
veils. The veils are nuabered from sballov to deep. Level 2 is in the JP-4
contaminated interval, level 4 is Just under the contaminated interval and
level 6 is deeper under the contaminated interval. LSD is the least
significant difference between, means at the 95Z confidence level.
406
-------
Benzene
Concenlralion (mg/L)
b _. «
-4 -» -* O
* — Injection
Level 2
i — • air— Level 4
\ "« Level 6
i A /" '* LSD
• \f \
A V *
f \A \ Begin Nitrate
• A \ \ y Injection
\ \X
^^M> *
.IXJl -J «-Jfc — £1 — ju — at 1 J*. »1 a* l m .m. ***»*» «K ^ w
O 20 40 60 60 1OO
Water
120
Time (days)
Figure 18. Concentration of Benzene in the reclrculatlon vater, and at
various levels in the veil clusters. See Figure 4 for the depth of the veils.
The veils are numbered fron shallov to deep. Level 2 is in the JP-4
contaainated interval, level 4 is just under the contaminated interval and
level 6 is deeper tinder the contaminated interval. LSD is the least
significant difference betveen »eans at the 95X confidence level, appropriate
only for the first 40 days.
407
-------
Toluene
Begin Nitrate
Injection
i
Injection Water
Level 2
Level 4
Levels
.0014
LSD
-a—a*-a-*-a-
60 80
Time (days)
100
120
figure 19. Concentration of Toluene in the recirculation vater, and at
various levels In the veil clusters. See Figure 4 for the depth of the wells.
The veils are numbered fron shallow to deep. Level 2 is in the JP-4
contaminated interval, level 4 is just under the contaminated interval and
level 6 Is deeper under the contaminated interval. LSD is the least
significant difference between iseans at the 95Z confidence level, appropriate
only to the first 55 days.
408
-------
m,p-Xylene
10
Begin Nitrate
Injection
-«— Injection Water
Level 2
Level 4
I Level 6
.001
100
120
Time (days)
Figure 20. Concentration of ••4-p-Zyleoe in the reeirculation vater, and at
various levels in the veil clusters. See Figure'4 for the depth of the veils.
The veils are numbered from shallov to deep. Level 2 is in the JP-4
contaminated interval, level 4 is just under the contaminated interval and
level € is deeper under the contaminated interval. LSD is the least
significant difference between, means at the 95Z confidence level.
409
-------
o-Xylene
Injection Water
Level 2
teveU
Level 6
Begin Nitrate
Injection
A I
.001
80
100
120
Time (days)
Figure 21. Concentration of o-Xylene in the recirculation vater, and «t
various levels in the veil clusters. See Figure 4 for the depth of the veils.
The wells are nuabered from shallov to deep. Level 2 is in the JP-4
contaminated interval, level 4 is just under the contaminated interval and
level 6 is deeper under the contaminated interval. LSD is the least
significant difference betveen neans at the 95Z confidence level.
410
-------
1
o
o
O
®
c
e
IM
C
O
CD
O)
O
Daily Benzene Levels
80 100
Days Since Startup
200
Figure 23. Concentration of Benzene in recirculstion veil PP-7A compared to
the predicted concentration based on dilution and wasting of the recirculation
vater.
411
-------
Table 4. Comparison of -actual BTX aqueous Concentrations with
predicted values at equilibrium, based on partition
•theory using residual BTX and fuel mass.
Compound
TREATMENT
Partition
Coefficient*
Prior to After Flood- After Two
Hydraulic ing, Before Months of
Loading Nitrate Nitrate
_ {concentrations in jig/L)
Benzene
Toluene
Ethyl-
benzene
Xylene
o-Xylene
3.39 Actual t
Predict
3.44 Actual
Predict
3.68 Actual
Predict
3.75 Actual
Predict
3 . 85 Actual
Predict
760
58
4540
440
840
450
2550
310
1380
480
5
17
12
44
9
490
200
260
74
2
15
6
6
23
27
37
18
* Partition coefficients from Smith et alf 1981. Value for xarp-
xylene based on average of values for m-xylene and p-xylene.
*t Actual concentrations are averages of samples from each of the
five basin cluster monitoring wells at Level 2 (or Level 3
prior to hydraulic loading) and Level 4.
412
-------
598
Background
After Nitrate
0 1 2 3 4 5 6 7 8 9 10
Log Cell Number
Denitrifiers
Total Counts
607 = Unsaturated, Clean
604 = Unsaturated, Contaminated
600 = Water Table, Contaminated
599 = Saturated, Contaminated
598 a Saturated, Clean
Figure 5
-------
USCG - Groundwater and Soil Remediation
SOIL VENTING
Two additional remediation projects using variations of vacuum extraction technology were
started up during the summer of 1990. The first of these is a soil venting (vacuum
extraction) system. This venting project is focused on venting a thin layer of aviation gas
contamination that has migrated down gradient from the source area and is concentrated
at and throughout the capillary fringe.
The challenge of this treatment is to effectively vent an area with air flow properties
governed by a three phase system of varying saturations. Air flow through the soil is
governed by the amount of pore space within the soil. Since Avgas contamination is present
throughout the capillary fringe, air flow characteristics will change if the pore spaces are
filled with air, liquid phase product, or water. The venting system was designed to
accommodate these varying saturations. Three withdrawal and six injection wells screened
across the contaminated layer at the capillary fringe and into the water table were installed.
The injection wells are located and screened to concentrate air flow through the
contaminated interval to the withdrawal wells, thereby increasing the rate of mass transfer
and soil remediation (Figure 5). Extracted vapors are treated with activated carbon.
The soil venting system began operation in August 1990, and has removed over 60% of the
estimated 602 gallons of Avgas present above the water table in the soil venting area in five
months. The system operates with a total flow of 150 cfm from the three withdrawal wells
and 250 cfm into the six injection wells. Vacuum withdrawal and air injection is
accomplished using two positive displacement blowers; one for withdrawal and one for
injection. Two fiberglass vessels connected in series hold 800 pounds of granular activated
carbon each for vapor treatment. Clean air discharged from the carbon tanks is vented to
the atmosphere. An air use permit has been obtained from the Michigan Department of
Natural Resources for this process.
Vapor concentrations from the venting system peaked at 31 mg/I total Avgas when the
system was started, and rapidly decreased after startup. Vapor concentration's had decreased
to below the detection limit for Avgas within two months of operation. At this point, cycling
was begun to maximize removal. The system now runs at an average of two to three days
every two to three weeks. The system will be expanded in the summer of 1991 to remediate
a larger area. An air use permit application for the expansion is currently being reviewed
by the Michigan Department of Natural Resources.
414
-------
PLAN VIEW
100*
LJ —
•D •
•D
r
o • •
•D • «D
75'
LEGEND
AIR INJECTION WELLS
O AIR WITHDRAWAL WELLS
• PRESSURE/CONCENTRATION
MONITORING CLUSTERS
n WATER LEVEL MONITORING WELLS
SECTION VIEW
2' AIR INFLUENT
TYPICAL
CLUSTER
4* AIR EFFLUENT
1.
2" AIR INFLUENT
•GROUND LEVEL
NOTE: NO SCALE ON SECTION VIEW
AVGAS CONTAMINATED
JAYER
WATER TABLE
ELEVATION
The Traverse Group, Inc.
2525 Aero Pork Drive
Travarae City. Michigan 49684
US078022
FIGURE 5
SOIL VENTING SUBSURFACE LAYOUT
US COAST GUARD
US COAST GUARD AIR STATION
TRAVERSE CITY. MICHIGAN
3-14-90
415
-------
USCG — Groundwater and Soil Remediation
BIOVENTING
A 90 by 75 foot study area has been divided into two equal areas of 45 by 75 feet to
evaluate the effects of different flows and extraction patterns. The northern area has an
injection system while the southern area has an injection and extraction/reinjection system
(Figure 6). The pneumatic properties of the unsaturated zone where evaluated by the
performance of a pneumatic pump test, resulting in a design radius of influence of 10 feet.
The work plan calls for ambient air injected into both areas at an initial rate which replace
the volume of calculated air filled pore space in 24 hours. The flow rate will be increased
to a vapor recharge rage of 8 hours or higher as the system becomes acclimate.
The blower package had to be capable therefore of extracting vapors in the south study
area, at depths of 15 to 18 feet (depth of the water table), at flow rates ranging from 5 to
65 cubic feet per minute (cfrn) and then reinjecting the vapors at the same rate, at a depth
of 10 feet. Additionally, the system has to be able to inject ambient air at the same flow
rate within both the extraction/reinjection plot (south area) and the air injection plot (north
area). Accordingly, because the ambient air injected will be placed in twice the are (two test
plots), the blower has to be able to inject air at flow rates ranging from 10 cfm to 128 cfm.
Prior to start up, the Michigan Department of Natural Resources Air Quality Division
determined that because the air contaminants would not be discharged to the atmosphere,
an air use permit would not be required.
Construction
The construction of the Bioventing project consisted of installing in the north area fifteen
aeration injection points, placed on 10 foot centers, in a three by five grid and screened just
above the water table. In the south area, eight sets of injection points coupled with seven
extraction points, ten feet on center, were installed with screens placed just above the water
table. Eight reinjection wells were installed with the screens placed at a depth of 10 feet.
Monitoring
The monitoring requirements of the EPA Work Plan called for the installation of several
different types and depths of monitoring equipment and/or sample points. To monitor vapor
hydrocarbon and oxygen concentrations, six 5-point cluster wells were installed with three
cluster wells per plot. The cluster wells consisted of 1/4-inch diameter copper tubing with
a wire mesh screen covering the tip. The 5 points of each cluster well were installed at 3.28
foot (1-meter) depth increments throughout the unsaturated zone. Additionally, we installed
three 14-point cluster monitoring well, (well screens at 1.5 foot intervals from ground surface
to 21 feet-one per plot and one at an upgradient location ) and one set of moisture
temperature probes per plot. The moisture/temperature probes are Soil Test Series 300
416
-------
USCG - Groundwater and Soil Remediation
moisture-temperature cells consisting of thermistor soil cells buried at depths of 5, 10 and
15 feet below grade.
The development of a sufficient microbial population to degrade the hydrocarbon vapors
requires adequate quantities of nitrogen, phosphorous and potassium. The EPA Bioventing
Work Plan called for an initial application of these nutrients consisting of 64 pounds of
nitrogen, 13 pounds of phosphorous and 5 pounds of potassium be applied to each area
prior to start up. Additionally during the growing season, 10 pounds of nitrogen, 2 pounds
of phosphorous and 1 pound of potassium is to be applied to each area monthly. These
nutrients were added in an aqueous solution by sprinklers until they were detected in the
ground water indicating they had moved completely through the treatment zone.
Operation
The system was started in October 1990 and ran until January 1991. After the application
of nutrients the system was restarted on April 29, 1991.
_^
Results
i
The results of vapor monitoring indicate that all of the degradation occurs in the upper 3
feet and the theoretical kinetics predict all of the treatment should occur in the 12 to 18
inch increment. The surface emission samples indicate that the surface emission are well
below regulatory limits for traditional soil venting systems. Soil samples obtained in
February indicate approximately 40% removal of contaminants contained above the water
table.
417
-------
00
FRESH AIR INJECTION WEIL MTH IF OT
WELL SCHEKH LOCATED AT A DEPTH
OF APPRO*. 16ft BELOW OKADE
US07S044
wry ACTION tvra NTH w or
Mai SIXtTII LOCATED AT A DEPTV
Of AWKOX. Utt BttOW ORAOE
REHWECTION Wat WTH HP Of
WU. SIWffH IOCATCD AT A DEFIH
OF AHhKOX. 10ft btl.OW 6KAUE
AW nueutiM v-nj. V.ITH TIH
Wt'LI. SfthEfll LOBATEU AT A lieVTil
OF APPKOX. lifl Btl.OW tthAUE
~~!\n The Traverse Group, Inc.
UH2S A».ro Pfjrk Drlvu
I TroveriiB City, Michigan 406B4
FIGURE 6
AREA II VENTING SYSTEM
U.S. COAST GUARD
U.S. COAST GUARD AIR STATION
TRAVERSE CITY, MICHIGAN
7-9-90
-------
CORE PROFILE OF FUEL CARBON, mgtkg
CONTROL
DEPTH
INCHES
160
175
182
192
198
206
212
DEPTH
INCHES
160
175
182
192
198
206
212
V
1
SEPT. 1990-FEB. 1991
234 97
461 1080
1030 1000
701 731
6500 8240
5620 3020
N.D. <8
f\ct£r\ OQ1 A
2862 2ol4
CONTROL2
SEPT. 1990-FEB. 1991
<8 <8
193 84
238 85
212 140
549 639
34 10
<8 <8
O/K ..^^IQ')
NORTH PLOT
SEPT.1990-FEB.1991
52
923
1253
926
6740
5780
39
SOUTH
SEPT.1990-
11
144
253
1970
1880
2830
<8
<8
18
153
972
1420
2200
53
1
100
PLOT
FEB.1991
<8
16
70
946
1860
17
<8
100
419
-------
WATER QUALITY - FEBRUARY 1991
Control
Recirculated
o
SAMPLE
jMW-1
IMW-I
{MW-2S
(MW-2S
-------
Comparison to Three Performance Evaluatations at USCG Air Station
in Traverse City, Michigan.
Hydrogen Peroxide
Nitrate
Bioventina
for Aviation gasoline for JP->4 for Aviation gas
m earth
in demo
200
235
354
Liters fuel 3,200
in demo
2,500
839
Remediation
accomplished
Benzene to
0.1 ppb
TPH still
>700 mg/kg
BTEX to
Standard,
Benzene to
0.1 ppb
30 % TPH
removed
Benzene
removed above
water table
50% TPH removed
Time expended
18 months
6 months
4 months
421
-------
Cost Cmparison of Three Bioremedial Technologies of Fuel Spills
in a Shallow, Sandy, Water-Table Aquifer under the USCG Air
Station at Traverse City, Michigan.
Hydrogen Peroxide
Category $/liter $/m
Construction* 2.8 45
Labor/ Monitoring 4.5 72
(per month) (0.25) (4.0)
Nitrate
. Bioventing
$/liter $/m $/liter $/m3
11
118
11
26
9 96 17 40
(0.9) (9.6) (4.2) (10)
Chemicals
(per
month)
Electricity
(per
Total
month)
32
(1
1
(0
41
-8)
.5
.08)
500
(28)
24
(1.3)
641
2
(0
1
(0
24
.8
.3)
.1
•1)
30
(3.0)
12
(1.2)
256
. 0
(0
2
(0
31
.20
.05)
.9
.72)
0
(0
6
(1
73
.44
.11)
.8
.7)
* Prorated to a five-year service life on buildings, pumps, and
blowers.
Figures for bioventing only reflect the first four months of an
ongoing demonstration.
422
-------
THE NEW LYME LANDFILL SUPERFUND SITE
GROUNDWATER TREATMENT FACILITY
BY
Ted Streckfuss, Environmental Engineer
U.S. Army Corps of Engineers, Omaha District
Engineering Division
Robert Curnyn, Program Manager
U.S. Army Corps of Engineers, Missouri River Division
Program and Project Management Directorate
I. INTRODUCTION. The New Lyme Landfill Superfund Site is located in a
remote area of Ashtabula County, approximately 70 miles east of Cleveland, Ohio.
The final remediation at the site consisted of the construction of a 43 acre RCRA
landfill cap and associated fencing, a 100 gpm groundwater and leachate treatment
facility, and 13 groundwater/leachate extraction wells. The extraction wells 'are
located around the periphery of the landfill cap. This paper will delve into the
operational aspects of the New Lyme Treatment Facility, construction
difficulties, and modifications necessary to maintain an operational system.
II. PREVIOUS STUDIES. The selected remediation of the New Lyme Superfund Site
was determined based upon information gathered during the Remedial Investigation
(RI) and Feasibility Study (FS) periods, conducted by CH2M HILL Ecology &
Environment for the U.S. Environmental Protection Agency (EPA), Region V. The
Final RI/FS was dated February 6, 1985. The Record of Decision (ROD), which
provides "the specific design 'direction required by the EPA for full site
remediation (based upon the RI/FS reports), was signed on September 27, 1985.
EPA Region V provided the U.S. Army Corps of Engineers, Omaha District with
a design directive in 1986. The Omaha District contracted with Donohue &
Associates, Inc. of Sheboygan, Wisconsin to perform a Field Investigation, a
Treatability Study, and the full site Design for the New Lyme Superfund site in
llarch 1986. The Treatability Study was conducted during 1986 and 1987 and the
Design was prepared during 1987 and 1988. The contract for construction services
was awarded in the fall of 1988 to Sevenson Environmental Services, Inc. of
Niagara Falls, New York. A portion of the design completed by the Omaha District
Corps of Engineers involved the borrow area for the landfill cap. .Innovative
design methods associated with cooperation between all agencies involved allowed
the construction of a contoured wetland area on state owned property, adjacent
to the site. The use of this borrow area provided for significant cost savings.
The Treatability Study was conducted by Donohue & Associates to verify the
suitability of the treatment process specified within the Record of Decision, as
well as to identify the specific loading parameters, the potential influent
characteristics, and the chemical feed rates required for the intended unit
treatment processes. The analytical information presented in the Initial and
Final RI reports and the FS report needed to be expanded upon to define the
parameters required for the design of the treatment processes. Difficulty in
obtaining representative samples of the pump leachate were identified in the
RI/FS reports, Treatability Study, and Design. It was known that the
groundwater/leachate extractions well pumping system would not be pumping pure
leachate, rather a mixture of groundwater and leachate. This led to the
difficulty in developing what could be considered a "representative sample" upon
423
-------
which design could be based. This dilemma has been verified through the early
operational results obtained regarding the characteristics of the waters to be
treated. The groundwater at the site is under an artesian head with monitoring
well water levels higher than the surrounding topography. It was anticipated
that there would be a significant difference in the analysis between the samples
that were collected for the Treatability Study and actual pumping conditions
which would be encountered under full site operation. The intent of the pumping
system was to dewater the landfill and to prevent future flow of groundwater
through the contaminated landfill by artesian action.
The Treatability Study performed by Donohue consisted of laboratory bench
scale studies, field pilot scale studies, and off-site activated carbon analysis.
The laboratory bench scale studies tested for metals precipitation and biological
kinetic rates. The field studies were conducted on site and included a multiple
train (four cell) pilot Rotating Biological Contactor (RBC), metals precipita-
tion, neutralization, and sand filtration. The pilot RBC system was operated in
parallel mode. Operation of the RBC pilot facility was handled in a batch mode.
This allowed coordination and correlation with ongoing EPA research laboratory
studies using pilot RBC systems.
The design for the New Lyme Treatment Facility provided a great deal of
flexibility between individual unit processes because of the potential for
changing waste characteristics from the groundwater/leachate extraction wells.
The Treatment Facility design allows for partial or complete by-passing of the
various treatment processes. Also incorporated within the design is the
capability to route return flows from a specific process unit to the head of
other another process unit.
The analytical data provided within Tables 1 & 2 was presented in the Final
RI report, Volume 1, Table 1.
TABLE 1
SUMMARY - ANALYTICAL DATA
VOLATILE ORGANIC
COMPOUNDS fVOC'SV
Aero le in
1 , 2-Dichloroethane
" 1,1-Dichloroethane
Trans 1,3 Dichloropropene
Ethylbenzene
Methylene Chloride
Chloromethane
Toluene
Trichloroethene
Vinyl Chloride
2-Butanone
2-Hexanone
4 -Methyl 2-Pentatione "
Xylene
Acetone
Leachate
(VG/L)
234
37.9 -
30.8 -
71.4
21.3 -
2870
10.9 -
92.5 -
15.2 -
20
82.6 -
6.3 -
2230 " -
41.4 -
' 0
180
23
13700
44000
12.2 ,
12600
162,4
101
49400
2780
5610
415
328000
Groundwater
(Waste 'Cell)
' OJG/L) . * /
430 -
., 570 -
' 430 -
570' -™
130
12
6000 -
1100 -
1500 -
1000 r
9700
10000 2
"
9700
"10000 '' '" '" "
"' ' ,/ ' '
^
76000
2300 ' , ~C1
15000 fr^ ' "_^
'"'/
46000
424
-------
Additional analytical data was collected and presented within the Initial RI
report, and is provided in Tables 3 & 4. This data was the basis for the
selection of the treatment train constructed for the site. The "Leachate" column
in Tables 1 & 2 consisted of data on samples gathered from seeps emanating from
the landfill. The "Groundwater - Waste Cell" column depicts the results from a
sample taken from a depression located near the center of the landfill, 'and also
represented the static water at the site. The high levels of contamination
documented through these sampling efforts was the driving force in the selection
of the ultimate treatment process.
TABLE 2
SUMMARY - ANALYTICAL DATA
SEMIVOLATILE
ORGANIC COMPOUNDS
P-Chloro-M-Cresol
Pentachlorophenol.
Phenol!
Benzoic Acid
.•2-Methylphenol'
1.4-Dichlorobenzene
N-Nitrosodiphenylamine
Benzyl Alcohol
PAH's
alpha-BHC
Leachate
(UG/L)
10.8
14
38400
11.4
14.6
6
6.8
16
13
11
99.2
- 21,7
4 - 15
0.006
III. SYSTEM DESIGN. The preceding design information contained in Tables 1 thru
4 was utilized in the preparation of the design documents for the New Lyme
Treatment Facility. The completed groundwater/leachate treatment facility
consists of the following unit processes:
Equalization Tank
pH Adjustment
Chemical Clarifier
Neutralization Tank
Rotating Biological Contactor
Biological Clarifier
Anthracite Filtration
Granular Activated Carbon Units
Effluent Storage Tank
Gravity Thickener (for chemical and biological sludges)
Sludge Filter Press
The treatment facility is also equipped with an analytical laboratory for
process monitoring and an automated computer maintenance system. The design flow
for the treatment facility was established as 100 gallons per minute (gpm), with
a peak flow of 150 gpm.
425
-------
TABLE 3
SUMMARY - ANALYTICAL DATA
Average Leachate
- Concentration
INORGANIC COMPOUNDS UG/L
; Aluminum
Arsenic
Asbestos (as mass concentration)
Barium ,
Boron
, Cadmium
• .. / Chromium
Cobalt •!. • -
. . ":" : iaCopper, '•''••.•'
: . '•;' 'Iron ;•>/;.. ,
.-.'•'• .- .^ Lead!'.-. ':=. •.':••';
••;'•..' ••? Manganese
' ,i',:Nick'el ./.. ;: :
• ':-•'-. '':'Tin':': .''"'• :•' :.:.'''
,-;:./ ;;l,,:.;;,:,vZinc:'';'\:;''1:; .•••
12800
9
2.46
2200
1200 *
2
28 ,„
160
67
900000 ,
160'
4900
,63
64
' 430
TABLE 4
SUMMARY - ANALYTICAL DATA
Average Leachate
" Concentration ^
VOLATILE ORGANIC COMPOUNDS" ' "" " UG/L
Acrolein
1»2-Dichloroethane
1,1"Dichloroethane
1,1-Dichloroethene
Trans 1.3 Dichloropropene
Ethylbenzene
Methylene Chloride
Chloromethane
Toluene
Trichloroethene
Vinyl,Chloride
2-Butanone
2-Hexanone ,, ,
4-Methyl 2-I'entanone
Xylene
150
70
11
3
45
2500
11900
9
" 2200
47
34
2500
119
As previously alluded to, this project was initiated through placement of
the site on the National-Priorities List (NPL) for remediation through CERCLA
legislation. The design for this project was based upon data collected during the
investigative and treatability phases. Analytical data collected during these
426
-------
phases provided the basis for system design. The design soluble biochemical
oxygen demand (SBOD) was found to be 424 mg/1, and the soluble chemical oxygen
demand (SCOD) was established as 1,439 mg/1. Other parameters encountered during
site investigations which were not incorporated within the previous tables are
as .follows:
TABLE 5
SITE SPECIFIC WASTE CONSTITUENTS
Parameter
Maximum Value Identified
NH3
TKN
N03
TSS
P
pH
118 mg/1
121 mg/1
13.8 mg/1
1510 mg/1
17.8 mg/1
between 5 and 8 units
Various metals were also encountered in the landfill leachate. Specific
metals of concern included arsenic, chromium, copper, cyanide, lead, mercury, .and
zinc. The concentration of these metals were generally minor, typically lower
than 50 parts per billion. Volatile and semivolatile constituents monitored
during the remedial investigation/feasibility study were instrumental in the
selection of the treatment process. Table 1, as indicated in the former
paragraphs, summarized data obtained during the field investigation period. The
extremely high levels associated with specific constituents led to the inclusion
of a biological process within the treatment plant flow scheme.
TABLE 6
CONTRACT DISCHARGE REQUIREMENTS
Pollutant of Concern
BODS
SS
Ammonia
PH . .
Acetone
Arsenic
Bis (2-ethylhexyl)pthalate
Cadmium (total)
Chromium (+6)
Chromium (total)
Copper
Cyanide , free
Ethylbenzene
Lead (total)
Mercury (total)
Methylene Chloride
Phenol
Toluene
Zinc
30 dav average
10 ing/1
15 ntg/1
2 flrg/1
6.5 to
85.8 mg/1
209 ug/1
9.2 ug/1
5 ug/1
11 Ug/1
54 ug/1
-
8.9 ug/1
68 ug/1
33 ug/1
0.2 ug/1
473 ug/1
407 ug/1
1870 ug/1
122 ug/1
Daily maximum
15 mg/1
22.5 mg/1
3 mg/1
9.0
577.5 mg/1
378 ug/1
1155 ug/1
58 ug/1
20 ug/1
2587 ug/1
27 ug/1
40 ug/1
1470, ug/1
174 ug/1
2.2 ug/1
10200 ug/1
5600 ug/1
2520 ug/1
. 464 ug/1
427
-------
The treatment plant was provided with enforceable discharge limits in the
operating contract. The limits required of the effluent stream are as shown in
Table 6. These discharge limitations were developed in conjunction with the Ohio
Environmental Protection Agency.
IV. SYSTEM OPERATION. Operation of the treatment plant consists of chemical
precipitation for metals contained within the influent stream, followed by
biological treatment for the organic, biodegradable portion, and activated carbon
for the refractory organics. Chemicals used in the metals precipitation portion
of the process consist of sodium hydroxide for pH elevation, and sulfuric acid
for neutralization after metals removal. Provisions for the addition of
phosphoric acid were incorporated in the event influent nutrient levels were
growth limiting. Specifically, the typical operation of the groundwater/leachate
treatment facility consists of the following scenario. Note that specific units
may be bypassed in the event the influent concentration does not warrant their
use.
Groundwater and leachate are pumped from thirteen extraction wells through
the forcemain located adjacent to the foot of the landfill. The flow currently
being treated at the facility is approximately 67 gpm, due to plugj^ng problems
within the piping preceding the pH adjustment tank. This problem will be
discussed in depth later in this paper. The maximum flow that has been treated
to date is 100 gpm. The contaminated influent is pumped into the equalization
tank located outside the treatment plant, adjacent to the structure. The water
contained within the equalization tank is aerated through the use of an air
header located approximately one foot off of the bottom of the influent tank
structure. Aeration within the equalization tank oxidizes the iron constituents
present within the raw water. Theoretically, the iron is subsequently removed
within the chemical clarifier. After travelling through the equalization tank
(with a detention time equal to approximately 150 minutes at the 100 gpm design
flow), the contaminated water enters the treatment facility through gravity flow,
and travels to the pH adjustment tank. Within this tank, the pH is elevated to
approximately 9.5 through the addition of a fifty percent sodium hydroxide
solution. Detention time within the 1000 gallon pH adjustment tank is
approximately 10 minutes at the maximum 'flow rate of 100 gpm. A flocculating
mixer is mounted on the pH adjustment tank to aid in the formation of flocculent
particles. Water exiting the pH adjustment tank is routed through four inch
piping to the chemical clarifier. The chemical clarifier is a thirty foot
diameter, 10 foot sidewall height, conventional circular clarifier. Hydroxide
precipitates formed through the addition of the sodium hydroxide in the pH
adjustment tank are precipitated in this unit. Water exiting the chemical
precipitation unit is considered relatively devoid of metals, and is capable of
meeting the compliance limitations targeted within the contract specifications
for those parameters. The constituents remaining within the" process stream
consist of the organic and volatile fractions. These contaminants are removed
through the subsequent treatment processes to be discussed.
Process water leaving the chemical precipitation clarifier enters a
neutralization tank where 93 percent sulfuric acid is added to neutralize the
waste stream prior to it entering the biological portion of the treatment plant.
The pH is typically lowered into the neutral range. Provisions to add phosphoric
acid to the waste stream were incorporated in the event that the flow did'not
contain sufficient phosphorous to maintain the microbial population. A
428
-------
modification to the contract during the plant start-up period incorporated a mix
tank and chemical addition to supplement the nitrogen and carbon concentrations
in the process water, again to enhance bacterial growth. Sodium citrate and
ammonia sulfate were the chemicals selected to supplement the carbon and nitrogen
constituents, respectively. The mix tank added to the contract is used to
prepare a batch for nutrient addition, which is subsequently pumped into the
neutralization tank for dissolution prior to entering the Rotating Biological
Contactor (RBC) train. After being neutralized, the water exiting the
neutralization tank travels to a splitter box where the flow is directed to one
of the three RBCs. The mechanically operated RBCs are 20 feet in length, have
100,000 square feet of surface area per contactor, and are constructed in
individual basins. Each contactor is fifty percent submerged. The individual
RBC units are covered with fiberglass reinforced plastic (FRP) covers which are
vented to,the roof of the facility. The process water flows by gravity through
the RBC units, which requires a stepped configuration in the plant layout.
Typical operation warrants flow through all three RBCs in series in order that
removal of the organic constituents may be optimized. Specific units may be
bypassed at the operator's discretion. Air headers are provided on the RBC units
to aid in media removal and water oxygenation. The principal use for the process
air is the removal of excess biomass which will tend to occur after extended
operational periods, provided that sufficient food is available within the plant
flow to support biomass growth.
Water exiting the RBC basins travels to an above grade, rectangular tank,
where the water is stored prior to pumping to the biological clarifier. A
modification to the contract incorporated an air header into this lift station
to prevent deposition of solids from occurring within the tank. Flow from the
rectangular tank is controlled through the use of a level controller and variable
speed pumps. Operation of the plant entails operator selection of a specific
level, typically fifty percent depth, within the lift station. The variable
speed pumps-then operate at the rate required to maintain the specific elevation
programmed by the operator.
Water is pumped from the RBC lift station to the biological clarifier. The
biological clarifier is configured similarly to the chemical clarifier. Water
entering the circular clarifier deposits the floe material sluffed from the RBCs
during system operation. The biological clarifier is 22 feet in diameter, with
ten foot'sidewalls. Flow exits the biological clarifier through the peripheral
V-notch weirs, and travels on to the gravity anthracite filtration system.
The anthracite filtration system is designed to remove particle sizes larger
than approximately twenty microns. Four sixteen inch gravel beds are provided
for support of the anthracite media. The filter bed is a uniform grade
anthracite. Two units are provided within the treatment plant. Of concern in
terms of operability of the system is the fact that the operation of the filters
is in parallel. Upon activation of the backwash system, the unit is taken out
of service in order to rejuvenate the media. This can be detrimental to
operation of the facility, should both filters need backwash at one specific
time. Backwash of the sand filter system is either automatic or through operator
intervention. Problems associated with plugging of the anthracite filters arose
during the plant start-up period. This problem will be discussed in depth in
subsequent paragraphs. As a result of the plugging of the filter by 'algal
populations, the backwash cycle was initiated every three to four hours. Typical
operation of ther unit process requires backwash on a daily basis. As the
headless through the filter system increases, water level within the filtration
429
-------
chamber rises, activating the suspended floats. These floats within the
anthracite filter chamber are activated at preset levels. The activation of the
float system in turn initiates the backwash operation. Water for use in the
backwash of the sand filtration system is taken from the effluent storage tank.
Waters generated during backwash of the sand filters are directed to the recycle
tank located adjacent to the sand filter system. A submersible pump located in
the recycle tank returns the water to the headworks of the plant for subsequent
re-treatment. Other inputs to the recycle tank include supernatant from the
gravity thickener, and filtrate from the filter press. After the process water
is routed through the anthracite filtration system for suspended solids removal,
the water is pumped to the second "lift station" contained within the treatment
plant.
The lift station located after the anthracite filtration system is designed
to transfer the water from the filtration system to the Granular Activated Carbon
(GAG) system. The GAG was incorporated within the process design to remove those
refractory organic contaminants not removed in the RBCs. Two identical GAG
contactor vessels are provided in the treatment plant. The pressure vessels are
charged with a total of 40,000 pounds of activated carbon. Process water is
pumped from the previously discussed lift station, through the use of variable
speed pumps, to the GAG contactors. Typical operation is in series mode, for
optimum removal efficiency. The GAG contactors are operated in a downflow mode,
with sampling ports located at intermediate points to detect exhaustion of the
carbon. Carbon is slurried to the contactors from the front bay of the facility.
No carbon washing or storage facilities are provided within the process area.
Pressure differential indicators are provided to indicate the headloss through
the GAG contactors. No provisions were incorporated for backwash of the
activated carbon.
Water treated by the GAG contactors is routed to the effluent storage tank
for storage prior to disposal to Lebanon Creek. The contract documents
established the discharge limitations which were required to be met by the
treatment facility prior to the discharge of any treated water. The treated
process water was also designed to be used for backwashing the anthracite
filtration system. The other primary use for the treated water included process
water for the lime slurry equipment. Quicklime is slaked in the lime slurry tank
before being fed to the sludge for conditioning prior to processing. Lime is
stored on- site in a storage bin located adjacent to the equalization tank. A
mechanical shaker system meters the appropriate quantity of quicklime into the
lime slurry tank for processing prior to use in sludge treatment.
Sludge within the treatment plant is generated from one of two sources, the
chemical clarifier and the biological clarifier. Sludge from these two units is
pumped through the use of diaphragm pumps to the gravity thickener located
adjacent to the effluent storage tank. As previously alluded to, the thickened
sludge is treated with lime prior to the use of the sludge filter press.
Supernatant generated during operation of the gravity thickener is decanted from
the sludge thickener and returned to the recycle tank, where it is subsequently
routed to the headworks of the treatment plant. Thickened, conditioned sludge
is conveyed to the sludge filter press through the use of a diaphragm pump.
After completion of the filter press operation, the cake sludge (typically
between thirty and fifty percent solids) is discharged to a sludge hopper for
ultimate disposal. Since completion of the treatment plant, it has been
unnecessary to process sludge due to the paucity of sludge generated in the
treatment process.
430
-------
Operation of the treatment facility is highly dependant upon data obtained
from in-plant sampling. The treatment facility has been provided with a
laboratory capable of monitoring COD, BOD, solids, and pH in order that process
control may be maintained. All metals analysis, nitrogen analysis, and
phosphorous analysis are performed by a contracted laboratory. Operation of the
treatment facility is limited to a qualified individual with an Class III Ohio
Wastewater Treatment Plant operators license.
The construction cost associated with the specific components of the project
was as follows:
Treatment Plant = $ 3.5 million
Landfill Cap = $ 9.3 million
Extraction Wells =- $ 1.0-million
Monitoring Wells = $ 0.5 million
Operation and Maintenance costs for the facility have been projected at
$0.7 million dollars. This cost includes power, chemicals, analytical., and
personnel costs. Costs associated with the treatability phase of the project
were approximately $0.3 million dollars.
In terms of annual operation of the treatment facility, an estimated 1.1
million pounds of 50% sodium hydroxide; 140,000 pounds of lime (for sludge
stabilization); 5,782 pounds of 75% phosphoric acid; 1.34 million pounds of 93%
sulfuric acid; and 60,000 pounds of activated carbon are expected to be used for
plant operation. ^H
V. OPERATIONAL DIFFICULTIES. As is normal with any new treatment facility, a
number of problems were encountered during-. the plant start-up period.
Immediately after initiation of plant operations, a problem with algae growth was
manifested. This was perhaps the least expected problem anticipated to occur
during the start-up of a groundwater treatment facility. The appearance of algae
within the influent could only be explained by a cross connection between surface
water and the groundwater/leachate pumping wells. Within one week of initial
plant start-up, the entire facility was inundated with a thick green, "pea-soup"
mixture which effectively halted operation of the facility. The initial solution
recommended to remedy the problem involved chlorine addition to the chemical
clarifier in an attempt to provide a 30 mg/1 residual, which would effectively
kill the algae present within the system. Extraction well operation was halted,
and the RBCs were bypassed to prevent the destruction of the seeded biomass. The
system was run in this recycle mode for a period of four days, after which it was
evident that the algae had been effectively eliminated from the system. After
initiating operation of the extraction wells a second time, the problem quickly
resurfaced, leading to the second solution, which has been effective to date.
The second solution involved the placement of black plastic over all exposed
process units within the plant, as well as the equalization tank located outside
the facility. In conjunction with the placement of the plastic, the extraction
well operation was again halted, and the treatment facility was operated in a
recycle mode, with chlorine added again (in the form of calcium hypochlorite) to
kill the algae growth as had been done in the previous remediation attempt.
Without a light source, the algae quickly went into remission, and the facility
operated without problem for nearly a week before the second start-up dilemma
occurred.
Problem two, arose as a result of premature precipitation of material
between the pH adjustment tank and the chemical clarifier. The problem
431
-------
originally manifested itself through excessive buildup on the butterfly valves
between the two process units. The quick fix for the problem involved
replacement of the butterfly valves with full port ball valves. This temporarily
solved the problem, but deposition continued until the process was forced to be
taken out of service in order that the line could be cleaned to remove the
accumulated precipitate. In order to permanently solve the problem, it was
recognized that excessive detention time was being provided within the pH
adjustment tank, leading to;early deposition within the piping network. The
final solution involved bypass of the pH adjustment tank through the installation
of a static mixer, caustic solution feed port, and an in-line pH monitoring unit.
At the time of this report preparation, this modification has not yet been
incorporated within the process. Until such time as the change is initiated,
caustic feed to the pH adjustment tank has been limited. This action is possible
due to the limited metals concentration currently experienced within the influent
waste flow.
The third major problem associated with operation and start up of the
hazardous waste treatment facility also involved premature precipitation within
the process piping. The four inch line from the equalization tank to the pH
adjustment tank was experiencing' a severe buildup of precipitate within the line
and on the valves preceding the tank, similarly to the second problem statement.
It was felt that the aeration within the equalization tank was oxidizing the
reduced metal lens within the waste flow and precipitating the resultant on the
piping interior and butterfly valves. The precipitation was occurring prior to
the chemical clarifier in the process piping, where the reaction was not desired.
The solution for this problem again involved the replacement of the butterfly
valves with full port ball valves. A second part of the solution involved the
cessation of air flow to the equalization tank except for one eight hour cycle
per week, with this cycle necessary to agitate any settled material for transport
into the chemical clarifier for subsequent deposition at that location. This
solution appears to be working well, in that no additional scaling appears to be
occurring in the piping and on the valves.
Other start-up problems have typically been minor in nature, and easily
solved. During actual construction of the facility, there were a number of
significant deviations which forced modifications to the construction contract.
The following paragraphs attempt to define those areas changed during the
construction cycle:
The most costly modification involved the placement of underdrain trenches
around the toe of the landfill. This change was necessary due to the excessive
amount of leachate which was "weeping" out the side of the compacted fill. The
modification involved the placement of six inch perforated piping, connected
every 400 feet with a manhole collection system. Periodically, the treatment
plant operator evacuates the contaminated waters which collect in these sump
areas for transport and subsequent treatment within the plant.
The blower system for the equalization tank was determined to be undersized
based upon a typographical error contained within the contract specifications.
The error was directly responsible for the purchase of a blower system which was
incapable of operating under the head conditions required under the actual
operation. The problem was remedied through the replacement of the existing
blowers with larger motors.
Other modifications to the construction contract were necessary, as can be
expected from a contract of this magnitude. Overall, the modifications were
432
-------
limited in scope, and resulted in the more efficient operation of the finished
product.
VI. CONCLUSIONS. This paper has provided an in depth review of this
multifaceted treatment facility. Information gleaned from treatability studies,
design, construction, start-up and operation was presented. This paper also
discussed specific problems encountered during the design and construction
phases, and presented the selected alternatives which may be of use in the
prevention of future problems at other site remediations using this type of
treatment scheme. Discussion regarding plant start-up difficulties and current
operating conditions was pursued and developed. The uniqueness of the New Lyme
Landfill Superfund Site lies not in the complexity of the individual unit
processes, but in the number of unit processes required to remediate all of the
waste constituents encountered at the site. Under this circumstance, the
complexity of the individual components is significantly less than the complexity
of the system as an entity. The coordinated effort between all of the individual
process units provides a unique system capable of treating the myriad of
contaminated compounds present at this site. This treatment plant has thus far
proven capable of meeting the designers requirement for the remediation of the
contamination caused by the landfill located in this remote area of Ashtabula
County, Ohio.
VII. REFERENCES.
1. CH2M Hill, Final Remedial Investigation, New Lyme Landfill Site, Ashtabula
County, Ohio. U.S. EPA Work Assignment No. 52.5MA8.0, Feb. 6, 1985.
2. U.S. EPA, Superfund Record of Decision. New Lyme, Ohio. Sept. 27, 1985,
Report No. EPA/ROD/R05-85/023.
3. CHZM Hill, Feasibility Study, New Lyme Landfill Site, Ashtabula County, Ohio.
U.S. EPA Work Assignment No. 52.5MA8.0 Aug. 15, 1985. •
4. CH2M Hill, Predesign Report, New Lyme Landfill, Ashtabula County, Ohio. U.S.
EPA Work Assignment No. 52.5MA8.0, June 13, 1986.
5. Donohue & Associated, Inc. Treatability Testing and Field Investigation
Report, New Lyme Landfill. U.S. ACE Contract No. DACW45-86-D-0007, Jan. 1987.
6. Donohue & Associates, Inc. Site - Specific Quality Management Plan, Site -
Specific Safety Plan, New Lyme Landfill Superfund Site. U.S. ACE Contract No.
DACW45-86-D-0007.
7. Donohue & Associates, Inc. Final Design Analysis - Hazardous Waste Cleanup,
New Lyme Landfill Superfund Site, Ashtabula County, Ohio. U.S. ACE Contract No.
DACW45-86-D-0007.
433
-------
Field Demonstration of Environmental Restoration Using Horizontal Wells (U)
by
B. B. Looney, D. S. Kaback, and J. C. Corey £,.15.
Westinghouse Savannah River Company
Savannah River Laboratory
AikcnSC 29808
A paper proposed for presentation at the Third Forum on Innovative Hazardous Waste Treatment Technologies:
Domestic and International. June 11, 1991 to June 13, 1991, Dallas TX.
The information in this article was developed during the course of work under contract No. DE-AC09-89SR18035
with the U. S. Department of Energy. By acceptance of this paper, the publisher and/or recipient acknowledges
the U. S. Governments right to reproduce, and to authorize others to reproduce all or part of the copyrighted
paper.
434
-------
Field Demonstration of Environmental Restoration Using Horizontal Wells (U)
by
B. B. Looney, D. S. Kaback, and J. C. Corey
Westinghouse Savannah River Company
Savannah River Laboratory .
AikenSC 29808
Abstract
Under sponsorship from the U.S. Department of Energy, technical personnel from the Savannah River
Laboratory and other DOE laboratories, universities and private industry have completed a full scale
demonstration of environmental remediation using horizontal wells. The 139 day long test was designed to
remove volatile chlorinated solvents from the subsurface using two horizontal wells. One well, approximately
90m long and 45m deep drilled below a contaminant plume in the groundwater, was used to inject air and
strip the contaminants from the groundwater. A second horizontal well, approximately 50m long and 20m
deep in the vadose zone, was used to extract residual contamination in the vadose zone along with the
material purged from the groundwater. The test successfully removed approximately 7250 kg of
contaminants. A large amount of characterization and monitoring data was collected to aid in interpretation
of the test and to provide the information needed for future environmental restorations that employ
directionally drilled wells as extraction or delivery systems.
Introduction and Summary
Remediation of soils and groundwater contaminated with organic and inorganic contaminants is an important
objective of the Savannah River Site (SRS - Figure 1) environmental restoration efforts. A wide range of
technologies are available and emerging technologies are being developed and demonstrated to remediate
contaminated subsurface materials. The goal is to provide a wide range of "tools" for this work.
Development of a diverse and robust "toolkit" of technologies is the best path toward efficient environmental
restoration that provides maximum protection of health and ecology at a minimum cost. Various extraction
and in-situ remediations are currently being developed. Recently, applications of directional drilling
technologies developed in the oil industry have been identified as a potentially important "tool" to allow
improved access to the subsurface for all types of remediations.
Normally, extraction of contaminants or addition of reactants for remediation must be performed through
vertical wells or boreholes. Directional drilling, including horizontal wells, offers a new and promising means
to optimize these operations. Just as horizontal wells have improved the performance of oil recovery
systems, they may similarly improve the performance of in-situ remediation, contaminant extraction, or
monitoring technologies for environmental restoration. The geometry of horizontal wells conforms to typical
subsurface systems, relatively thin but laterally extensive zones. Horizontal wells can be installed to
remediate beneath buildings and waste sites, to remediate linear sources of contamination such as pipelines
or streams, to prevent the spread of the edge of a plume, or to introduce reactants (e.g., for bioremediation).
A variety of competing directional drilling methods have been developed. Each of these represents a
possible new approach to installing delivery/removal systems to improve environmental restoration.
The information in this article was developed during the course of work under contract No. DE-AC09- "
89SR18035 with the U. S. Department of Energy. By acceptance of this paper, the publisher and/or
recipient acknowledges the U. S. Governments right to reproduce, and to authorize others to reproduce all
or part of the copyrighted paper.
435
-------
The first application of horizontal wells for environmental restoration is the In-Situ Air Stripping test at SRS.
For this test, two wells were installed (Figure 2). One well, below an area of contaminated groundwater, was
used to purge volatile chlorinated solvents from the groundwater by injection of air. The second well, was
installed in the vadose zone beneath the location of an abandoned process sewer line that was known to have
leaked chlorinated solvents; this well was used to remove solvent that has not reached the groundwater and
to collect solvent stripped by the purge well. In-Situ Air Stripping has the potential to significantly reduce
clean up time when compared to standard groundwater pumping followed by air stripping in an above
ground tower. The demonstration site has been carefully characterized and monitored using surface and
borehole geophysics, cross hole geophysical tomography, chemical analysis of soil, soil gas and groundwater,
microbial characterization of soil and water, and pressure monitoring hi all affected areas. Additionally, each
of the horizontal wells has been fitted with a bundle of tubes that allow pressure/concentration
measurements at various points along the well bore. The data that result from all of these tests will improve
the technical basis for design of future directional drilling applications. Kaback et al. (1989) provide a
detailed description of the horizontal well installation methods and experiences.
The 139 days of field operation at the demonstration site were successful; the operational and data collection
goals identified at the beginning of the demonstration were met. We operated at approximately 90% utility
and extracted 45 to 60 kg of VOC from the subsurface each day. A total of almost 7250 kg of chlorinated
solvents were removed from the subsurface during the test. Pressure and concentration data from the
bundle tubes and vadose zone piezometers provide valuable data for future projects of this nature. These
data, combined with the helium tracer test data, provide information on the types and importance of various
heterogeneities in this typical natural system. Both low permeability and high permeability zones influenced
the performance of the system. The importance of these zones to mass trarisfer in subsurface remediations
(both in situ and extractive methods) will be evaluated using the collected data. Initial data from Sandia
National Laboratory (SNL) and Lawrence Livermore National Laboratory (LLNL) suggest that the seismic
and electrical tomography methods and the single point flow sensors will be able to distinguish changes in
moisture and fluid flow in the subsurface caused by the extraction/injection.
Synopsis of Test
The following list is a synopsis of the activities at the site (unless otherwise noted, vadose zone pressure
readings, extraction/injection well pressure and temperature readings, and extraction well concentrations
were determined at least three times per day; groundwater, vadose zone piezometer and bundle tube
concentrations were measured approximately weekly). Tunes are provided both in terms of date and elapsed
time from startup to facilitate interpretation of the data provided later in the report. Collection of water
samples continued following completion of the active demonstration phase in December.
Synopsis/Highlights of In Situ Air Stripping Test:
July 27 - Day 1 - Begin Test - Vacuum Extraction through horizontal extraction well
(AMH2) at 16.4 standard cubic meters per minute (scmm).
August 11 - Day 16 - Air Injection through horizontal purge well (AMH1) initiated at
low rate (1.8 scmm)
August 23 - Day 28 - Air injection rate increased to medium rate (4.8 scmm)
September 13 - Day 49 - Heating"cf injected air to approximately 64 degrees Centigrade
initiated.
October 3 - Day 69 - Air injection rate increased to high rate (7.5 scmm).
November 16 - Day 113 - Ah- Injection portion of the test completed - Injection stopped
and compressor demobilized
December 13 - Day 140 - Vacuum Extraction portion of the test completed - extraction
stopped and vacuum blower demobilized
436
-------
System Performance
Horizontal Vacuum Extraction Well Concentration and Pressure Data
The flow and vacuum conditions at the vacuum extraction well remained relatively constant throughout the
test. The flow (measured by a calibrated pilot tube) generally ranged from 15.6 to 17.0 scmm. The vacuum
at the wellhead stabilized at approximately 25 to 27 cm Hg. The temperature of the extracted gas was
relatively constant at approximately 15 degrees Centigrade. The heating of the injection air had no
measurable effect on the temperature of the gas extracted from AMH2.
The concentration and temperature of the contaminated vapors extracted from the horizontal vacuum
extraction well (AMH-2) were measured approximately three tunes per day (more frequently at each change
in operating parameters). As shown in Figure 3, the total concentration of chlorinated solvents decreased
rapidly during the first 2 days of operation and stabilized after approximately 3 days. Initial concentrations
were as high as 5000 ppm (volume/volume) in the gas and the total concentration stabilized at approximately
300 to 400 ppm. This concentration represents an extraction rate of approximately 100 to 140 pounds of
solvent per day. The average extraction rate for each of the injection/extraction conditions (e.g., vacuum
only, vacuum plus low injection rate, etc.) was calculated and resulting data are plotted vs injection rate in
Figure 4. The vacuum extraction process removed contaminants at a rate of about 50 kg/day; injection of an-
al the medium and high injection rates appear to result in the stripping/removal of an additional 9 kg of
solvents per day from the groundwater and the vadose zone below the extraction well. The cumulative VOC
removal is summarized in Figure 5. Almost 7250 kg of contaminant were removed from the vadose zone and
groundwater at the demonstration site during the short testing period.
The concentration and vacuum data from the bundle tubes in the horizontal vacuum extraction well (AMH-
2) indicate that, while operating, AMH-2 appears to draw water from the vadose zone in a manner similar to
a suction lysimeter. Since the well dips to its lowest elevation at the terminal end, the final sampling position
(at 67 m from top of casing) appears to be covered by water and the sampling position at 56 m may be
covered by water periodically. These two sample tubes were more variable than the first four and the
vacuum level becomes lower over time at these positions. Nonetheless, the vacuum data suggest that the
entry of air into the screen occurs over much of the horizontal section and that heterogeneity in the system
results in more air entering in some segments of the well (e.g., 19 to 31 m) than in others (e.g., 31 to 44 m).
Note that the vadose zone piezometers near the end of AMH-2 have significant drawdowns - suggesting that
the water covering the bundle tubes does not completely fill the wellbore (i.e., air is entering all the way to
the end of the well). The bundle tube concentration data support the conclusions from the pressure data.
Water covering the final two sampling tubes precludes contaminated air from entering these tubes for
sampling. Variations in concentration along the well are a measure of the concentrations of VOCs entering
the well and of the amount of air entering the well. The concentrations are relatively stable in the bundle
tubes closest to the wellhead; these sampling points represent a composite of the entire well screen.
Transient behavior of the vacuum levels in the AMH2 bundle tubes (e.g., after the system was down or after
a large ram event) also supported the above interpretation. The vacuum levels in the last two tubes were
highest when the system was restarted after being off (i.e., no water in the wellbore) and the vacuum levels
declined with time as water was drawn into the well following startup.
Vadose Zone Piezometer Data
The pressure in the piezometers respond rapidly to the vacuum system and were relatively stable throughout
the test. Vacuum levels of several cm of water or greater are measured in all of the vadose zone piezometer
clusters indicating that the zone of capture in the vadose zone extends across the entire demonstration site
and that contaminants are not being spread in this zone. These data will be interpreted in detail in future
publications, however, a few observations are listed below. Concentration vs time plots in the vadose zone \
gases varied greatly over the demonstration site as contaminated vapors were swept from the system and
437 ' -
-------
purge gases entered from below. For example, MHV1A (the lowest tube in one of the vadose zone clusters)
exhibited an increase in concentration following the initiation of the medium injection rate. The
concentrations in the zone monitored by this tube stabilized after a short tune and appeared to decrease
following completion of the injection activities. The zone monitored by this tube is between the water table
and a clayey zone and is separated from the extraction well by clays. This zone appears to respond to the
injection process as the gas migrates up to the extraction well through discontinuities in the clays of the
vadose zone. The piezometer tubes that monitor the sandy zone in which the vacuum extraction well is
completed yielded concentrations that were more constant. These data are consistent with the observed
heterogeneity and dispersion of injected gases observed in the helium tracer test discussed later.
Groundwater Data
Groundwater samples for VOCs have been collected approximately weekly during the demonstration.
Samples for microorganisms and biochemical markers have been collected every two weeks. A base map of
the wells used in the plots is shown in Figure 6. Sampling the wells takes approximately 2 to three days;
each of the maps is labeled using the elapsed tune from the start of the test to the first day of sampling.
The time periods and operational conditions at each time are summarized in the figure captions. One map
was selected to represent each operational condition for this report.
Figures 7 through 10 show the groundwater concentrations for trichloroethylene at each of the selected
times. These time series data, combined with other data, provide valuable information about the
performance of the demonstration. The plume maps clearly indicate that the ak Injection is impacting the
groundwater. In particular, comparison of the initial conditions to the "final" conditions shows a significant
overall improvement in water quality for both trichloroethylene and tetrachloroethylene. Trichloroethylene
concentrations ranged from 500 jig/L to 1800 /tg/L at the beginning of the test and from 10 to 1031 ng/L
at the end of the test. Similarly, tetrachloroethylene ranged from 85 jtg/L to 184 /ig/L at the beginning of
the test and from 3 to 124 Mg/L at the end of the test. Concentrations in almost all wells declined between
day 28 and 39 (following initiation of injection at the medium rate). This reduction is assumed to represent
the purging of the in-situ air stripping process.
Examination of the later time periods reveals the appearance of a high concentration area in the southern
and distal portions of the site (near the end of the injection well). While the area of increased concentration
is near the edge of the demonstration site, it does not appear to result from lateral spreading of the
contaminants (lateral spreading of the plume would have yielded a transitional map for day 39 rather than
the relatively low concentrations at all monitoring points). Two hypotheses are currently being examined to
determine the cause of the concentration increases: 1) upward migration of contaminants caused by the
injection of air below the monitoring well screen, or 2) slight pressurization of the vadose zone between the
water table and a zone of clays resulting in downward migration from the water table to the depth of the
screen being measured. The microbiological measurements indicated that microrganism counts increased by
two to three orders of magnitude in the groundwater sampled during the injection test.
Tracer Tests
A test using an inert tracer (helium) to determine the behavior of the injected gas was performed. The test
consisted of adding 3 standard cylinders of helium to the injected air over a 24 hour period using a regulator
and flow meter. All of the identifiable potential exit points for gas to leave the system were sampled to
determine if: 1) the injected air is reaching the extraction well and 2) significant quantities of helium are
being forced out of the system through monitoring wells. Since flow of injected air had been qualitatively
noted by field sampling personnel, all 20 monitoring wells and the geophysical access holes that extended
below the water table were monitored.- Samples were collected using a 50 mL disposable syringe and the
samples were placed in 30 mL preevacuated serum vials. These vials were analyzed using a helium mass
spectrometer that has been modified to sample the serum vials at a constant rate. The mass spectrometer
438
-------
was calibrated in two steps. First, the mass spectrometer is tuned and the sensitivity adjusted to an internal
calibrated leak (diffusion) standard in units of standard mL of He per second; after this step, gas standards
prepared in the serum vials are used to convert the instrument reading to ppm (volume) and check the
stability of the tuning. The results of the test are listed below:
>
Elevated helium concentrations were measured in the horizontal vacuum extraction well
(AMH2), confirming "communication" between the wells. The maximum concentration in
AMH2 was approximately 30 ppm, significantly lower than the maximum injected
concentration of approximately 1800 ppm. Additionally, the helium peak in the gas from
AMH2 was much broader than the 24 hour injection (helium was still exiting AMH2 at the
end of the test period). These observations suggest that the gas traveling between the two
wells is dispersed as it flows through the several clayey/fine grained zones between the
wells. At the end of the reporting test, approximately 45 percent of the injected helium had
been extracted from AMH2.
Elevated helium concentrations have been measured in all 10 water table wells indicating
that the dry screen above the water table represents, a pathway for the injected gas to flow.
In two cases, MHT1D and MHT6D, the arrival times were very rapid (hours) and the
concentrations were almost the same as the injected air indicating that the gas exiting these
two wells was traveling through relatively high permeability flow paths. The volumetric flow
rate was measured for the gas exiting these well heads to allow helium mass balance
calculations. Despite the high concentrations and measurable flow, these wells account for
loss of under 0.5 percent of the injected gas. No flow was measurable at any of the other
water table wells.
Elevated helium concentrations were measured in 6 of the 10 wells completed in the air
injection zone (~46-48m deep). This suggests that some injected gas is entering the screens
and exiting at the wellheads. No volumetric flows could be measured at any of the injection
zone wells. Thus, these wells do not represent a significant component in the
helium/injected gas mass balance.
In summary, the data confirm the fact that the subsurface at the test site contains high permeability and low
permeability heterogeneities (typical of natural systems). Nonetheless, there is communication between the
injection well (AMH1) and the extraction well (AMH2). AMH2 represents the largest component of the
mass balance.
Related Tests/Reports:
No geophysical reports have been written to date, however, contact with LLNL indicates that the electrical
cross hole tomography has successfully mapped the moisture changes in the subsurface as the test conditions
have been modified. Also, the Sandia National Laboratory single location 3D flow sensor responded rapidly
to changes in injection conditions. Reports on these tests will be prepared in 1991.
Conclusions and Recommendations:
The SRS in situ air stripping test successfully removed approximately 7250 kg of contaminants. A large
amount of characterization and monitoring data was collected to aid in interpretation of the test and to
provide the information needed for future environmental restorations that employ directionally drilled wells
as extraction or delivery systems. Results from the in-situ air stripping test suggest that access to subsurface
systems can be improved through theNuse of directional drilling. Comparison of the horizontal vacuum
extraction well to a vertical vacuum extraction well operated at the same test site indicates that the horizontal
439
-------
geometry increased the VOC extraction efficiency by a factor of approximately five. Additionally, the test
indicates that in situ sparging operations or addition of reactants to aquifer systems may be a reasonable
technology for selected remediation conditions. For example, installation of a remediation system consisting
of horizontal extraction/delivery wells that are perpendicular to groundwater flow at the leading edge of the
plume may be potentially cost effective. In this scenario, with sufficient design data, cycling periods of
operation could be employed to reduce energy usage and remediation costs. Further development of
directional drilling methods (to reduce costs and minimize drilling fluids) and related characterization and
monitoring systems will assist in developing timely technical design data for remediation systems that employ
dircctionally drilled wells. Continued development of this and other emerging environmental technologies
will assist in developing a "toolkit" of options for remediation that can be combined in creative ways, resulting
in cost effective systems that provide a high level of protection to human and environmental health.
Reference:
Kaback, D. S., B. B. Looney, J. C. Corey, L. M. Wright and J. L. Steele (1989). Horizontal Wells for In-Situ
Remediation of Groundwater and Soils. Proceedings of the Third National Outdoor Action Conference on
Aquifer Restoration, Groundwater Monitoring and Geophysical Methods. Association of Ground Water
Scientists and Engineers, pp 121-135.
440
-------
TEST SITE
Figure 1. Location of the Savannah River Site in South Carolina
Injection point for flushing gas
Extraction of contaminated gas
03
I
0)
N
I
Figure 2. Schematic Diagram of In Situ Air Stripping
Using Horizontal Wells, a Flushing Gas, and Vacuum Recovery
441
-------
1000
900
800
700
VOC 600
Concentrations 500
(ppm) 400
300
200
100 -t
0
TCE
— PCE — TOTAL
5021 ppm total VOC (1052 TCE / 3969 PCE)
20 40 60 80 100
ELAPSED TIME (DAYS)
120 140
Figure 3. Concentration of Chlorinated Solvents in the Air Extracted from Horizontal Well AMH-2
150
140
VOC Extraction
Rate (Ibs/day)
50 100 150 200 250
Air Injection Rate (scfm)
300
Figure 4. Average VOC Extraction Rate as a Function of Air Injection Rate (metric conversions: 1 kg = 22
pounds, 1 cubic meter = 353 cubic feet)
442
-------
Cumulative
VOC Removal
(pounds)
20 40 60 80 100
ELAPSED TIME (DAYS)
120 140
Figure 5. Cumulative VOC Removal During The In Situ Air Stripping Demonstration
(metric conversion: 1 kg = 22 pounds)
103000
102900
£
o
1
102800
o 102700
gj 102600
102500
MHT-10C
MHT-BC •
-9C
'AMH-1
J_
I
48500 48600
48700 48800 48900 49000
SRS East Coordinate (feet)
49100
49200
Figure 6. Base Map of Groundwater Wells Contoured During In Situ Air Stripping Test. Well AMH-1 is
the horizontal injection well and well AMH-2 is the horizontal vacuum extraction well.
443
-------
I UOUUU
102900
^^
t)
^
~ 102800
j>
75
c
'•g
0
o
0 102700
t!
o
z
U)
% 102600
1 02500
— • !
/ /fc '•
•' • / i r
; ' / i
1600 v£/ / / /
^•^^ / $ •'
i2oo^?.;X>
-------
1 02900
o
o
102800
o
'I
o
"102700
ti
o
(/J
85 102600
102500 L
48500 48600
48700 48800 48900 49000
SRS East Coordinate (feet)
49100
49200
Figure 9. Trichloroethylene Concentrations (Mg/L) at the In Situ Air Stripping Demonstration Site
(Day 39 - Medium Injection Rate for 11 Days).
103000
1 02900
1 02800
o
o
" 102700
o
£ 102600
102500
'AMH-1
1
1
_L
J_
_L
_L
JL
48500
48600
49200
48700 48800 48900 49000 49100
SRS East Coordinate (feet)
Figure 10. Trichloroethylene Concentrations 6*g/L) at the In Situ Air Stripping Demonstration Site
(Day 144 - Final Conditions).
445
-------
Third Forum on Innovative Hazardous Waste
Treatment Technologies : Domestic and International
June 11 - 13, 1991, Dallas, TEXAS, U.S.A.--
Environmental Problems of the Czechoslovak
Chemical Industry : Cleanup Actions in SPOLANA
Ivan Zika
Spolana Neratovice
Czechoslovakia
Keratovice - Praha, May, 1991
446
-------
SPOLANA HERATOVICE, C z echo siovaki a
Production Programm-' , • •
••' ,'sT -•
-;-:;Vl:-- .'. •
iM+l^^--:
^"ligai^
Pounded in 1941,' after the large facility Spolek -Company.
for the Chemical and foundry Processing."
in tfsti n.L-. (previSudetenland) had been
occupied by the German.-Army
Area 3,5 sq km• . . .
Value app. 1 bil USD
Employees 6 .000 people
Main .operating plants : .
Viscose Sta'ple Rayon Fibre incl. Natrium Sulphate
Nylon 6 Monomere incl. Ammonium Sulphate
Vinylchloride and Polyvinylchloride
Chlorine and Natrium Hydroxide
Sulphuric Acid
Hydrochloric Acid
Sweeteners
'Pesticides
i. .
•Man made Hormones
Constr. Plastics «
In'April 1992 should staM; the new Alphaolephin Plant licensed.
by.'U.S. Chevron being now assembled. Spolana is going to; be; ••....;,
privatized ; auditory and corisulting; company-': Wright Killeri,~':*
Heuston, Tx, .U.S.A. '-";; ;. ' ' -:'"" '
447
-------
NERATOVICE
,^-C.Hl
J «»«-ft<
^T!
T CHEMICK? PROMYSL
.... vleintl 10000 1_
=®&i£§S° • »««<~™-
. O
W y» *• <*r ^?-v >K '' '*•• ^ •T^^°al*.
•f^/ H^hTa^y^^PVr°^> '.UkJi. ^ .r,
->' JjA/0 5^ !- ^^---'^F^f^^""1---^--, -—v^X'-s-f
VrY* ^-- ^&f^^^^^^^
\S\_ ! t '^ — V < r-«raN»«».M,.«,,)»-Sl^,/V>*^>-^x—^^1 li* •>""-• VrTN
^
fe
-------
SPOLAKA'.'' NERATOVICE, Czeclioslovalcia
Environmental Problems
They have resulted from the improper, • ideology preferring,
management; from the low grade energy base.and material
sources, from many backward production technologies and
outwarn -equipment; The failing centralized planning state
policy and the resulting lack of.finances for the proper-
extended reproduction as well as for the plain renewal •
were the main reasons' for the described state :
Energy Management :
Energy - Base - Brown coal '(600 let p.y. consumpted)
Capacities installed 550 t of!steam p.h. (420 MW Heat)
65 MW of electricity '
Condensate recovery 60 % plus turbine condensate
Water Management : • • . •
Up take of the Elbe - river water (40 mil m3 p.y.)
Water use app. 150 mil.m3 p.y.
'Water treatment with "ferric sulphate ; demineralizing
recovery condensate polishing
Waste v/ater treatment in both primary (chemical end ,
mechanical) and.secondary (biological - activated sludge)
steps (15 mil.nO p.y.);' primary and secondary sludges •
being stored in the lagoon, superna tant liquor-returned
back into the biological system ; mechanical dev/atering
(the belt filter pressing)'is being assembled, sludge cake
intended to be incinerated.
The plants having been, as environmentaly not sustainable,
'stoped : (producing) -
lindane (HCE) and pant-achlorophnnol (PGP) in 1965 ;
sulphuric acid from pyrite,. glue and .gelatiife in 1974,
chlorine and natrium hydroxide in graphite mercury
electrolysis (1-975), cyclohexanone from phenol (1983)
ferric chloride in 1 991 , s.acharine - intended to be
stoped in 1991.
Solving her environmental problems, SPOLANA has surveyed
hydrology and hydrogeology conditions, changes of the
meteorology situation, as well of .
449
-------
the emissions of wastes into the air, the surface and ground
water, and of solid wastes. Especially during the last ten
years the material ballance, the monitoring of ballance losses
as well as of emergency (accident) leaks have been improved ;
the satisfactors'- state has not yet been reached, lack of
technologies, money and e:cperience being the main .pause.
The greening of the SPOLAKA facility, according to the-pro-
ject dated on Oct. 1990, considering all production plants
running, should demand about 200 mil. USD (incl. investment
costs of the new membraneelectrolysis) for the next twenty
years.
Of course, the 'part of the Czechoslovak government, as of
the legal successor of the previous communist regime, in
sharing clean - up projects of newly privatized facilities •
is suggested as being controversial, though it is evident
that most of industry facilities were systematically deprived
of their property in last fourty years.
The environmental problems, generally, may be characterized
by the step of knowledge :•-
1 - unknown, hidden, not yet detected as such or from
the point of view of their consequencies
2 - known feasible, but still without any real reme-
dial technologies
3 --exactly detected, having more variant technologies,
real, defferring each from others by the economy,
by the economy, by the efficiency or by the time
for the implementation
4 - those, that could be prevented by the- approaches
accepted beforehand
Case histores I and III are illustrating the item I, cs.se
history II refers to the item 4, prevention being its goal.
The modesty prevents us from confessing there are any unknown
problems in SPOLAKA ...
Having prepared and on presenting my papers I wish to thank
my coleagues from SPOLAKA as well as from other cooperating
institutions helping SPOLAKA to solve her environmental problems
450
-------
especially projecting institutes Chemoprojekt Praha
and Eydroprojekt Praha, hydrogeology survey compa-
nies Aquatest Praha, Vodnl zdrojje Praha as well as'
VUGI Bmo, my coleagues from the Ministry of Industry
and from the Ministry of the Environmental Protection
of the Czech Republic.
Enclosed : Diagrams of Material Ballance respecting
hazardous substances
Enclosed : Case History 1 - the Contamination of the
Abandoned Mercury Slectrolysis Cell Eoom
and Hydrogen Gas - Holder with Metal Mercury
Case History II - the Prevention of the
Potential Light and Chlorinated Hydrocarbons
Mixture Ground Water Contamination
Case History-III - the Advantages of the Car-bo-
nation of an Alkalined Activated Sludge
451
-------
SPOLAKA - Production Plants Material Eallance
- carbondisulphide
- celulose
- metal zinc
- natrium hydroxide
- sulphuric acid
CS2» H2S
-emissions into air
-gcJods : staple rayon
fibre, natr.sulphate
-solid a.liquid wastes :
(hazardous :free acids,
zinc,carbon disulphide,
hydrogen sulphide)
Staple Viscose Sayori Fibre Production -
(goods: raw materials)..
0,47 incl. recovery of
natrium sulphate , carbon-
disulphide and sulphur
- minesalt
- water
-misc. metal mercury,
calcium chloride,sulpK
acid,graphite,act.
carbon
\ NoCl
/ H20
V
No OH
0,93
C12
.
> "
> )
•"2.
-^
-hydrogene
-goods:natrium hydroxide,
chlorine,hypochlorite
—solid a. liquid wastes :
(hazardous : chlorine,metal
a. inorg.mercury,free alkali)
Chlorine a.Natrium Hydroxide Production - *(
0,93
- molten sulphur
-air oxygen
-water
-misc.filter aids,
catalyst
jirq
emissions into air
-goods : sulphuric acid,
oleum
-solid a.liquid wastes :
(Hazardous :diatom.earth
filter cake,free acid)
Sulphuric Acid Production
452
0,985
-------
SPOLANA - Gross Material Ballance
•p
s
o
o3
•rl
CD
-p.
3
o
• w.
rH CO CO
•P -d CD
K »H T3
03 X -rl
CO «rl O
•d Pi
fj 0
•d A 3 bo
d *P XI O
3 In ft k
05 rH P
faO CD 3 -H
S 1fl
is co bo •>
co -p cd co
•H CO -H
o cd
o
CD K ra -H
*d CD o3 *d
ca.
•d
o
o
bO
•d
•H
H
o
CO
3
a
a1
•H
rH
O
CO
to
03
bO
CO
CD
•P
CO
cd
CO
h a
" CO O
CO ft,0
•H S3 O 03
A 03 o O
rH bO-d Si
tQ O 3 >s
•H a A
•d -H O
- a -d
' CD -rl CO
N -P
CD
0) >>
P, ~ 03
'•H 11 O W)
T3 CD fl
...H Tl
Cd rH
H O -d
c; ra CD co
o • -H -p
a bO -rj
03 ft
O i-H
.. 3
CO
3d «
5 §)s
cd
•> C
>»-H
~ fl
CO -H
•H -d -H
U CD C3
O >
pj «.
CO bO ti
a o
H -P
3 O
O rH
O
O
CO CO
to
3 O E
•• cd
co ft -d
O rH CO
to o3
"•H (S CD
3 T3 O O bO
o a a
03 -rl
cd -P H
N -d -P rH
td >> co oj
-C A 8
CO CD -P CO
' -p co co
O -H rH £1
«. O -H -
CQ -rl 03
co >d
fl
•H
rH
03
•H
to h
CO CD
S3 -P
•H oJ
^J E
o
05 bD
S d
-•H
C E
O (D
ra m
oi ra
•p
H
nJ
rH
Cd
O
O
I
•>.
£
.£
&
rH
3
CO
1
a
CD
-p
rH
O
S
CD
•d
•H
ft
:3
CO
• H
TJ
rj
O
_r^
05
O
cd~
•H
a
O
e
e
1
•P
rH
td
to
,
O
CO
0
rH
H
CD
O
CO
£5
CO
rH
£>j
-P
<1>
cJ
CD
^j
K
O
JH
•ri
cd
••
CO
a
o
CD
a
05
rH
rH
CO
0
m
•rl
E
1
^
3
0
bp
••H
^
CO
-p
rH
•rl
*l [
•t
CO
•p
CD
^>
rH
0
CO
i
•H
_V{
§
ft
h
CD.
•p
oJ
^
'*«
CO
T3
•rl
0)
hO
CO
CO
CD
•d rH
bO-H
S3 h
^H
co
a)
•H
(-.
CO
•p
03
a
I
453
-------
— cyclohexanone
- ammonia
- oleum
- molten sulphur
for S02
- carbondioxide
- air*kygen
- trichlorethylene
- natrium hydroxide
\cyklohexona
\ NH3
)
l^sot,
1 S
I ":o?
fichloretvlen
I
nes
.
koprolaktcrrt
0.95
..0.18
Ht^ri, '. '
S07/N0x,
trichloretylen
emissions into air
goods : caprolactam
ammonium sulphate!
- solid a. liquid vrastes
( hazardous :trichlorethy-|
' lene,hydroxiiaminesulphate]
Caprolactam (Kylon 6 nonom.) Production -- \ ... 0,95
(incl. recovery of amm. sulphate)
- ethylens
- gasaous chlorine
misc :natrium hydro-
xide j stabili-
zers,
Soar--"
- emissions into air
— goods — polyvinylchloride,
dichloroethane
- solid a. liquid wastes :
(hazardous : vinylchloride-
monomere, dichloroethane,
ethylchloride)
Polyvinylchloride Productin (from VCM -ethylene chlorination)
- ... 0,71.
454
-------
£ase History I - The Contamination of the abandoned Mercury
Electrolysis Cell Room and Hydrogen Gas -
Holder with Metal Mercury
(Mercury electrolysis - contamination - detection - remedy)
A; "S""":^ !'~-^ -£i iEb ^iz- izi fi
r
/x x / ,x -• ,-'
>' ///, X4
.. JT
4f/
The environmental problem of the mercury chlor - "alkali tech-
nology plants is vrell Known: abont 10 grams of metal mercury
consumption per ton of chlorine produced could be explained
by the unavitable losses-into the vent air, into the brine.-
treatment sludges (if not the pure salt is being used),'into
The hydrogen and natrium hydroxide0 In spite of it The speci-
fic consuption of metal mercury from 70 up to 300 and.even
more are still regarded as not surprising.
455
-------
The clean - up of the old abandoned chlor - alkali cell
room seems to be important, as well as difficulto It con-
cerns the change of the old -mercury technology to the
new membrans technology in the same electrolysis cell -
room, as well as another intended use of the cell - room
builolting.- The separation of the contaminated room from
the environment by the slurry walls seems to be insuffi-
cient*
In the Spolana cell - room building {reinforced concrete,
brick filling, wooden roof) 50 m wide, 200 m long amd 15^
heigh) chlorine and natrium hydroxide were produced since
1947 till 1975 • The by produced hydrogen was being stored
in two gasholders with the concrete underground vesselso
t
-Of 400 tons of metal mercury consumed since 1947 abou
50 tons were lost into the wastes ( sludge, vents incl. hy-
drogen, v/aste waters, natrium hydroxide )* Abont 200 tons
escaped into the Elbe river owing to the plant's once -
•through cooling water system - This was deduced from the
.amounts of metal mercury having been recovered from the
coaling watter circuit of the new chlor - alkali plant.
In the 'last 5 years 30 tons have been recovered, from the
sev/erage system and from the dismantled equipment o
Searching for The ballance difference of about 100 - 120 tons
both structures of The cell-room building 'and of gas - holders
•were surveyed by the drilled sampling - bores and by the
analysis of the samples*
The ana-lysis of 90 samples having been taken from the
bore holes 1,5 m and 3 m deep. located .regularly at the
10 000 sqm ground ''floor area indicated 1
the original cell --room (started in 1947, ended -in 1975) *
depth 1,5 m o»o average Hg content. 0,45 g kg"
(from 0,01 up%to 2,2) __
depth 3 m o»o average Hg"; content 0,97 g kg **
(from 0,04 up to 7,14)
The extended cell -. room (started, in 1956, ended in 1975
depth- 1,5 m •••o average Hg content 0,14 g kg"*
(from 0,02 up to 0,58)
. 456
-------
«."!
depth 3 m oo* average Hg content 0,10 g kg
(from 0,02 up to 0,10) .
The reinforced concrete slab having carried the cells . . -
is contaminated by the metal mercury droplets hidden under
the covering insulation resin layerj the intensity of the
contamination differs from 50 up to 400 g of mercury per aqnu
The soil in the 10 m distance of' two hydroge.n gasholders
was sampled and detected as beinq contaminated by the meial
mercury at the intensity ranging from 0,3 up to 3 mg'kg"" •
Aiming at preventing the cell - room building from
spontaneous colapse, that would result in the transport
and storage of 10 000 cuin of heterogeneous building,
material contaminated with metal mercury, the project
of demercurization plant has been accepted* Till the end
of 1993 the contaminated parts of the building and of the
underqround material will be desintegrated, the metal mercury
will be separated, the resting mercury fixed in the resulting
material, which will be mostly turned back into the under -
ground of the cell room building. The reconstructed cell -
room building should be prepared for the storage or for the
production needs. The decontaminated hydrogen gas - holders
are intended to reuse as pretreatment vessels for pesticide.-
waste waters* '
The analysis method: the atomic absorption spec trorne try
(spectrometer TMA 254) for the solid samples containing
metal mercury*
The contamination of the air in the cell room building:
before the metal mercury was excavated .from the sev/erage
svstem and from the surface garbage, the concentrations
ranged from 0,015 up* to 0,57 mg of mercury per m .of air;
they .'decreased at 0,009 up to 0,07 mg per m -- the .range
being influenced by the temperature.
457
-------
The economyi
though the price of the 99,9§ percent mercury decrea-
sed from 300 USD per flask in 1988 up to 130 in March
1991, most of the resting metal mercury should be
recovered till ihe end of 1993, as the other alterna-
tive , the demolishing of the structure and 'the storage
of the material in the land -fill seems to be impossibleo
The planned decontamination and renewal of the cell-room
building saves more than 70 per cent of the investment
costs needed for the similar new structure1.: the invest-
ment costs of the new building would cost at leas^
100 mil CzK; the demercurization of the building and its
renewal should cost not more than 50 mil CzK; the trans-
port and storage of the building material into the
hazardous waste landfill would cost another 50 mil CzK»
458
-------
i**-4* ••
Q^f^^^r^ '... •'.
ti!H«¥
-------
-40
\
-------
Case History II - the Prevention of the- Potential
light and chlorinated Hydrocarbons
Mixture Ground Water Contamination
(Contamination of the ground water - existing and
expected by potential leaks from new plant - survey -
model - project - hanging slurry walls - discharge wells)
The fifty years ' operation of different technologies
in SPOLAKA, the unefficient maintenance of the sewerage
system and still low discipline of the operational crews
as well as low effective organization and managing resulted,
among others, in the considerable contamination of the
ground water; though the inorganic pollution (chlorides,
sulphates) prevails, high leaks of chlorinated hydrocarbons
(AOX) were detected in last 10 years, resulting in increa-
sed concentrations of ethylene-dichloride, trichloroethy-
lene and others dangerous substances*
The extreme concentrations were detected at?
l,2dichlorethane 9 500
trichloroethylene 70
chlorobenzene 1 500
dichlorocenzenes
trichlorobenzenes
* 21 500 mig 1
•* 105 mig 1
•» 2 000 mig 1
-1
-1
-1
-1
up tolOO mig 1
5 100 •» 16 200 mig l"^
rems.rki higher concentrations were sampled and detected
from the sampling 'wells at the north of the facility
site as a consequuence of .previous wastes' dump
The new Alphaolephine (NAO) plant, thfough all potentially
contaminated areas are being protected with horizontal
insulating plastic coverings, could contribute to the
existing pollution by the leaks1 of light hydrocarbons,
that might interact with existing AOX in the groundwatero
The interaction would result in higher solubility of hydro-
carbons in the groundwater containing AOX, and their compli-
cated treatment.
Aiming at avoiding that , both production and storage areas
of the NAG plant were provided with vertical insulation,
separating the upper shifts of ground water under the NAO
plant from the surrounding ground waterl both sites were
surrounded by 1500 m long hanging slurry cut off: wallsJ the
control of the level and quality of the inside ground water
will be enabled by the systems of detecting wells as well
461
-------
as well as of controling withdrawal wellso The ground
water, if it would be contaminated with hydrocarbons,
v/ill be disch&rget into the oily waste water system and
treated in the new part of the waste water treatment
plant.
The project, asked by Spolana - investor of NAO plant
was completed by hydrogeology companies .(VUGI Brno and
Yodni zdroje Praha) in October 1987; the construction of
the walls as well as of the sampling and control discharge
wells was finished in March 1991*
The project aimed at four items I
1 - analog study of different possibilities of the ground
water quality protecton, •
2 — The design of the dimensions as well as of the depth
of the cut - off walls,
3 - The design of the discharging regime of the ground watee
at both NAO plant's sites, and the design of the main
features of the operating manual
4 - The disign of the monitoring and of the sampling and
analysing equipment
The results
The hydrocarbons, if they would leak from NAO plant
into the soil and would they contaminate the ground water
v/ill be prevented to escape by the cut-off slurry walls*
The analog model resulted in the possibility to use either
complete or hanging wall without any difference in influen-
cing of ground water regime* The economy and the technical
equiqment of the construction company decided to build
the hanging sturry wall.
The bottom edge of the wall at the level bf 155 00 aoSo,
is abont 3 m under the normal water table. Three controlinq
\vells equiped with pumps could decreaxe, by the ontput of
8,5 1 s each, the ground water level not more than at the
level of 158,30 a s,l* The output of the ground water discha
ge pumps could be controls, d'by analyzing the data collec-
ted from other eight, detecting bore - holes on each,
production and storage site*
462
-------
Geology and hydrogeology
The permeability of the grevel - sand material on both NAO
plant sites isucharacterized by the-jCoefficient of filtra-
tion 2,0 v 10 up to 8 . 10 m s~ o The bedrock, low
permeable calcareous clay marl - is about 10 m under the
ground level.
The walls' construction
The wals 0,60 m thick of the mixture I 702'1 water, 562 kg
loess, 140 kg blast furnare slag cement, 14 kg activated
bentonit, having compression strength 0,24 MEa after 28
days, filtration coefficient of the order of 10 m s ,
was built with customary technology covering the cuting
by the clayiskr-suspension. The crest of the wall is being
protected by the reinforced concrete panelso
The controling wells equipment
The wells of the diameter 630 mm with inner polypropylene
'screening and with the gravel sand outer strew ave equiped
with two submersible pumps, one for the control of ground
water table level, second for the colection of hydrocarbon
film.
Organization
The ground water quality system will be.operated and
maintained by the operational crews of NAO production
plant, being under supervision of the people responsible
for the enviroment protection.
Economy}
The investment costs of the NAO plant exceeds 3 bil. CzK.
The costs of the secondary safety system preventing the
contamination of .the environmental ground water in the
neighbourhood of Spplana amount to 25 mil CzK, less than
1 percent of NAO pla,nt investment costs* At avy case, the
safety, of people as well as of the nature, is a good
bussines where the searching for the right difference
between expensive and those being worth more than their cost
could mislead to father sophisticated conclusions.
463
-------
m^WiK^siivF' • !r*^m
I MJJ. ., .r'/Gij €fi'-::fi to1'-It?' •"•^vv-.. -413--A '••"*^^>H\
•*•.' X/l^: f.'-7 .* -n . n; ;'
•^WnV^, J-fa '-"• .v^.).
Jiii'^/lOl'F:;^
o
'i
• ? -
• •• i v • ,
-
-------
465
-------
9917
a
~o
-i
6
o
—J-
-------
^^f^^
/
-------
'Case History III - The Advantages of the Carbonation
of an Alkalined Activated Studge
(Industrial waste water - activated sludge processbulking
sludge - efficiency improvement - carbonation - better
settling, dewatering - decreased septic decay and mobi-
lity of heavy metals)
Waste waters from carpolactam (nylon 6) , viskose-rayon
and polyvinylchloride plants and from pesticides produc-
tion as well at the municipal sewage wastewater have been
treated for more than 20 years in the chemical - biological
wastewater treatment plant of the capacity 35 000 m d •
The problems of-decreased efficiency connected with fila-
mentous activated sludge in the system, due to poor set-
tling and thickening, v/ith low dry matter cencentration in
dewatered biological sludge have been effectively solved
by the modification of activated sludge process - carbonata-
tion of activated, sludge.
Carbon dioxide generated in aeration tanks and in final
clarifiers is capable to neutralize the alkalinity dosed
into the mixed-liguor. Alkalization of the inflow into the
activated - sludge tank up to pH 10,5 with 5 percent
lime .- milk, continuously or interrnitently, is succeeded
by the.rapid C0p neutralization* The reaction proceeds in
well operated activated sludge systems during not more than
30 minuteg. The velocity of neutralization reaction depends
on the rate of carbon dioxide generation during the oxidati-
on process; on the contrary the higher bacteria concentrati-
on due to the carbonation reaction increases the rate of CO,.
c
as metabolite generation, shen the alkalinity in the form
of lime-milk or calcium oxide powder is dozed, forming
microcrystaline calcium carbonate. The presence of this
result of carbonatation reaction in the mixed. - liguor elumj
nates the disadvantages of poor settling sludge, radically
improves the thickening and dexvatering processes. The final
sludge cake with high portion of calcium carbonate particle,'
is fit for land use, heavy metals and other impurities
mobility being decreased by high calcium and alkalicity
content.
468
-------
The process modification, compensating the usual chlori-
nation of mixed - liguor inflow to secondary clarifier,
being under suspicion of forming risky chlorinated hydro-
carb ns, has been successfully tested and continuously
improved in the chemical cornbinate-.waste water treatment
plant Spolana Keratovice (35 000 m d~ ) for 10 years,
and in the small foodstuffs factory w.\v. treatment plant
Vitana Bysice (600 m d ) for 8 years with desired re-
sults ! the efficiency of the biological step Treatment
process has been increased from previous 75 % up.to.93 %
•in the.ca.se of chemical industry Treatment plant, up to
97 % in the case of foodstuff factory treatment plant,
respectively. The demand for the de-watered sludge cake by
farmers raised rapidly.
As the concentrations of zinc in the resulting primary and
recondary waste water treatment plant sludges are high
(60 g of Zn per kg od a.m.), the future thickening and de«ra
terintrr of the sludge is the only alternative to the present
lagooning. Dev/atered sludge will be incinerated* The alka-
li zati on and carbone,tion of the activated surplus sludge
generally improve the settling and dewatering characteristi
values of the sludge
The economy:
The dose 0,2 upto-0,5 kg of CaO per 1 kg BOD 5 loaded
(at avg. 90 % efficiency) increased the efficiency of
SPOLANA v/.v/. treatment plant by 8 per cent (83 - 91) o
The cost of lime for 3 600 tons of BOD loaded varies
from 0,576 up to 1,44 mil CzK (cost of lime being 800 CzK
per ton) • The decrease of 200 t BOD 5 at the discharge cau
sed the decrease -of the discharge tax (10 CzK per kg BOD
5 discharged) e.to saving of 2 mil CzK.
469
-------
\jjAtHR
Jo-
Go
bo
10
o) •
-\oo
~3ao
470
-------
^
- ( Iff lU
, .
V
_2
i
*-
1012
10
1t
JJJS
to
^F
J
51
58
\ H
: NrrT2
10* '*'
Ot=
471
-------
CATALYTIC OXIDATION EMISSIONS CONTROL
FOE REMEDIATION EFFORTS
Captain Ed Karchand, HQ AFESC/RDVW, Tyndall AFB FL 32403-6001, USA
INTRODUCTION
The soils and groundwater tmder airfield facilities are often
contaminated with jet fuel components, chlorinated solvents, and degreasers.
This contamination has resulted from past disposal practices, leaking storage
tanks, and accidental spills. As a primary solution to this problem, the Air
Force established the Installation Restoration Program (IRP), to identify
contaminated areas, determine the type and extent of contamination, and
initiate appropriate cleanup actions. There are now over 3,500 IRP sites at
243 installations with an estimated 60% of the sites requiring cleanup action
(Reference 1). The Engineering and Services Laboratory (ESL), part of the Air
Force Engineering and Services Center, is responsible for environmental
quality research and development of more effective, cost efficient remedial
actions. This research targets the development of chemical, biological, and
physical treatment systems to meet this challenge. This paper reports the
findings from several field tests of remediation technology where catalytic
oxidation was used to control or treat the off gasses from the effort.
CONTAMINATED GROUNDWATER REMEDIATION
WURTSMITH AFB STUDIES
In the late 1970's, trichloroethlyene (TCE), a degreasing *gent, vas
discovered in the drinking water at Wurtsmith AFB, Michigan. Chemical
analyses of the groundwater shoved levels of TCE exceeding 6,000 micrograms
per liter (ug/L). The U.S. Environmental Protection Agency maximum
contaminant level for TCE is 5 ug/L. The source of the TCE was traced to a
leaking 500-gallon underground storage tank. Since the leaking tank went
undetected for years, the quantity of TCE leaked could only be estimated. The
subsequent plume of TCE was determined to encompass approximately 9 million
cubic meters, with a maximum concentration approaching 10,000 ug/L.
t
A review of the literature identified countercurrent packed-bed air
stripping as a possible treatment alternative. Countercurrent 'packed-bed air
stripping involves flowing contaminated water down a packed column, while
forcing air upward through the column. The packing breaks tip the flow of
water and air, increasing the air/water contact and enhancing transfer of the
contaminant from the water into the air. In many states air emission controls
are required to prevent release of these volatiles to the environment.
The Environics Division of the ESL performed laboratory and pilot-scale
tests at Wurtsmith AFB to verify the operating performance of packed-bed air
stripping. As a result of the study Wurtsmith AFB currently has two air
•tripping operations underway removing TCE from the groundwater from two
separate plumes. 1 third unit,, under construction, vill remove benzene from
472
-------
another plume of contaminated groundwater. The initial air stripper does not
have any emissions control device while the other two have (or will have for
the benzene unit) catalytic oxidation for emissions control. Catalytic
oxidation is a combustion process where the contaminant-laden air stream is
preheated and passed through a catalyst bed. Final products of the oxidation
are typically carbon dioxide, water, and inorganics.
Evaluations are underway at Wurtsmith AFB on the catalytic oxidation unit
installed to control the air stream coming from the 200 gallon per minute air
stripper used to remove TCE from the groundwater. The preliminary findings
are shown in Table 1. The catalytic unit is a fluidized bed reactor. The
catalyst particles are spherical shaped and the contaminated air stream is
passed through the reactor at sufficient velocity to churn or fluidize the
catalyst bed. This motion causes the particles to collide into one another
which breaks off small pieces of the surface. Since catalyst fouling occurs
on the surface, this type of reactor is contiuously self-cleaning. This
catalyst attrition is slow and at Wurtsmith AFB they are still running on the
same catalyst charge from 1988.
There is some concern though because the Wurtsmith AFB catalyst appears
to be forming a small amount of benzene when operating. Simultaneous sampling
of the preheater effluent and the stack emissions show an 40 - 60 percent
increase in the benzene concentrations. This is based on one sampling effort
and is a preliminary, and puzzling, finding. The Engineering and Services.
Laboratory is looking further into the situation to understand the reaction
mechanisms. The vendor indicates that the benzene formation is due to a low
catalyst bed volume (not enough residence time for the air to contact the
catalyst). This will be verified in the near future.
TABLE 1. PRELIMINARY DATA FROM THE EVALUATION OF A
CATALYTIC OXIDATION CONTROL UNIT AT WURTSMITH AFB MI
AIR STREAM CONCENTRATIONS:
1 part per million Trichloroethylene (TCE); 10
parts per billion of 1,2 Dichloroethylene
CATALYTIC OXIDATION UNIT SPECIFICATIONS:
CAPACITY: 1200 cubic feet per minute
OPERATING TEMPERATURE: 700 °F
NATURAL GAS CONSUMPTION (ave): 800 cubic feet per hour
TIME ON STREAM: SINCE JUNE 1988
DESTRUCTION EFFICIENCY OF TCE (as of Feb 1990): >97%
.(as of Oct 1990): >92%
PURCHASE PRICE: $113,000
EGLIN AFB STUDIES
In 1988-1989, at a large jet fuel spill site on Eglin AFB, Florida, we
evaluated (Reference 2) different packing materials for conventional
counter-current air stripping operations and compared their performance to a
new rotary air stripper. In addition several emissions control options were
also evaluated. The groundwater at the site contained a large variety of
soluble jet fuel components as well as inorganic materials that greatly
affected the research effort. Table 2 lists some selected parameters from the
Eglin site.
473
-------
TABLE 2 SELECT CONTAMINANTS AT THE EGLIN AFB FUEL SPILL SITE
CONTAMINANT
BENZENE
HAPHTHALEKE
TOLUENE
0-XYLENE
IRON 500
HENRY'S LAW
CONSTANT
(atm-mVmole)
.0047
.00041
.0059
.0040
Hot Applicable
Fot Applicable
The rotary air stripper is a new approach to
-------
Catalytic oxidation, carbon adsorption mud molecular sieves were
evaluated for the control of the emissions from the air stripping omits at the
Eglin site. The carbon units had a very low capacity for the lover molecular
veight compounds (C$ and belov). In addition the excess humidity from the
air stripping effluents further reduced the carbon adsorption capacity. Thus
a carbon bed large enough to adsorb the emissions from the air stripping
operations,would have a large capital and operating cost making carbon a very
expensive alternative at this site. Two molecular sieve materials, Union
Carbide's type 9102 and 1387-53, were tested because they are not Impacted by
humidity effects and they could be regenerated on-site vith ozone. Our data
showed that both molecular sieve materials were unsuccessful for adsorbing the
contaminants in the air stripping emissions. The unfavorable performance of
the molecular sieves may have been because their pore size were too small to
allow the contaminant molecules access to the active adsorption sites.
Another emission control technique evaluated was catalytic oxidation. An
Engelhard pilot-scale catalytic oxidation unit was tested at the Eglin site.
The unit uses an electric preheater to raise the inlet gas temperature to
I'OOO °F before passing it through a precious metal fixed bed catalyst
reaction chamber. The result is on-site destruction of the organic
contaminants. Enough of the hydrogen sulfide (see Table 2) was stripped out
of the water to cause a chemical reaction in the catalytic oxidation unit'.-"
which effectively and rapidly deactivated the catalyst. Cost estimates for a
100 gallon per minute air stripping unit, based on 99% removal of benzene from
contaminated groundwater, are: $3.19/1000 gallons just for the air stripping
unit, 11.70/1000 gallons for catalytic oxidation of the emissions (based on
other fluidized bed data) or $6.<47/ioOO gallons for activated carbon emissions
control.
CONTAMINATED SOILS REMEDIATION
There are several methods to remediate a site contaminated with volatile
organics such as jet fuel. The ESL tested the efficacy of using in situ soil
venting to remove JP-4 from a contaminated sandy soil site at Hill AFB UT.
During the ten months of operation 115,000 pounds of hydrocarbons were removed
from the site. The emissions from this effort were sent through one of two
catalytic oxidation units.
The first unit was a 500 cubic foot per minute fluidized bed unit that
operated for eight months. The second was a 1000 cubic foot per minute fixed
bed unit that used a precious metal catalyst and was operated for six months.
Thus there was a period of four months where the two units operated together
to treat the venting off gases. The fixed bed was operated between 470 and
625 °F while the fluidized bed unit was operated between 625 and 700 °F.
The results (reference 3) show that the fluidized bed unit had an average 89X
destruction efficiency and the fixed bed unit had a 97X destruction
efficiency. This gives a cost-per-volume-treated rate of $23.80/million ft3
air and $29.80/million ft3 air for the fixed and fluidized bed units,
respectively.
475
-------
While the fixed bed trait appears economically feasible it has it's
limitations. The unit would not be able to handle a large flow rate of the
initial highly concentrated air stream. This is because the process is one of
oxidation or burning of the contaminants. That means releasing heat in the
process. Fixed beds could get so hot that they actually melt the end of the
bed. Temperature safety controls prevent this from happening, however it does
limit the amount of contaminant you can treat. The fluidized bed unit,
because of the better heat transfer, can handle the higher concentration flow
rates, up to a point. The draw back is the need to add catalyst.
Approximately 150 pounds of catalyst were added to the reactor over the eight
aonth operation at Hill AFB UT.
CATALYST DEVELOPMENT AM) TESTING
Tvo laboratory studies are now being conducted to investigate catalysts
resistant to deactivation. The University of Akron is developing a catalyst
that resists deactivation when challenged vith a chlorinated air stream.
Akron researchers have found that chromium oxide and vanadium oxide materials
can reach greater than 95 percent conversion of chlorinated organics to "water,
carbon dioxide, and dilute hydrogen chloride (Reference 4). They are
continuing their research to find a superior catalyst that is resistant to .
chlorinated organics and sulphonated compounds present in air-stripping
emissions.
The second study is being done by the Research Triangle Institute (ETI),
H.C. They are evaluating off-the-shelf catalyst formulations from five
manufacturers. The initial step was to create a standard catalyst testing
protocol from vhich future catalyst formulations can be compared to this
study. The goal is to find out which catalyst is the best for a given
contaminated air stream.
After the catalyst has deactivated from constant exposure to a
synthesized air-stripper emissions stream, ETI will determine what caused the
catalyst to deactivate, which operating procedures will minimize deactivation,
and whether the catalyst can be effectively regenerated. This information
will be used in in an economic comparison of the different catalysts.
Catalyst formulations being tested are the AEI Econocat, a copper chromite
formulation from Harshaw, Carulite from Carus Chemical, three supported noble
metal catalyst formulations from DCI, and a Ealdor-Topsoe catalyst.
HEAPED - CROSSFLOW AIR STRIPPING VITH CATALYTIC
5SIONS CONTROL
Crossflow air stripping is a packed-column aeration process which
Involves changing the air flow path of a conventional countercurrent tower.
The main change is the placement of baffles Inside the tower which causes the
air to flow in a crisscross pattern up through the packing (Figure 2). This
forces the air to flow at 90 degrees to the flow of contaminated water rather
than in completely opposing directions, as in a countercurrent tower. Proper
selection of baffle spacing can produce a marked reduction in gas velocity,
lowering gas-phase pressure drop, and reducing blower energy costs compared to
conventional countercurrent mode of operation.
476
-------
Results show the crossflow tower can greatly reduce the blower energy
costs (Reference 5). However, for the highly volatile compounds, the blower
energy cost is not a significant factor in the total cost; therefore, a
countercurrent tower would be just as cost-effective as • crossflow tower.
Blower energy costs do have a significant impact on the total cost of air
stripping for the low and moderately volatile contaminants such as 1,2
Dichloroethane and Methyl Ethyl Ketone. Therefore, the crossflow tower could
be more cost-effective for removing these compounds from groundwater.
Gas Out
Liquid In
Gas Out
Liquid In
Gas In
Liquid Out
Gas In '
Liquid Out
COUNTERCURRENT
CROSSFLOW
Figure 2. Comparisons of Crossflow and Countercurrent Air Strippers
477
-------
A field study demonstrating the removal efficiency of crossflow air
stripping for low and semi-volatile organics vill be conducted during 1991 and
1992. During this test field validation of the RTI catalyst selection
procedure and the University of Akron formulations will be carried out as
emissions control from the crossflow air stripping operations.
REFERENCES
1. Statement of Mr. Gary D. Vest, Deputy Assistant Secretary of the Air Force
(Environment, Safety and Occupational Health) to the Readiness, Sustainability
«nd Support Subcommittee of the Senate Armed Services Committee, 4 April 1990.
2. AFESC, Air Stripping and Emissions Control Technologies; Field Testing of
Countercurrent Packings. Rotary Air Stripping, Catalytic Oxidation, and
Adsorption Materials, under publication.
3. AFESC, Field Demonstration of In Situ Soil Venting of JP-4 Jet Fuel Spill
Site at Hill Air Force BaseT under publication.
4. AFESC, Vapor-Phase Catalytic Oxidation of Mixed Volatile Organic
Compoundsf Greene, H. L., ESL TR 89-12, Sep 89.
5. AFESC, Laboratory Investigations of Cascade Crossflow Packed Towers for
AirStripping of Volatile Organics from Groundwater. under publication.
478
-------
Composting to Bioremediate Explosives Contaminated Soils
Kevin R. Keehan*, Wayne Sisk, and Craig A. Myler
U.S. Army Toxic Hazardous Materials Agency
. Aberdeen Proving Ground, Maryland, 21010-5401
Abstract
Composting is currently being-considered as a viable,
cost-effective alternative treatment method for the remediation of
explosives-contaminated soils. Results from bench studies, pilot
scale tests and field demonstrations have proven that composting
is capable of biologically degrading explosives to acceptable
levels. The advancement of composting explosives contaminated
soils to a full scale remedial technology has been hampered mostly
by economics. An optimization field demonstration was conducted
at a current National Priority List (NPL) site to access the
operational aspects, economic aspects and analytical aspects
required for full scale remedial operations. Maximum soil
throughput was evaluated using both aerated static piles and
mechanically agitated, in-vessel composting systems. Results from
this optimization study are currently being used to design a full
scale bioremediation process at the same NPL site. Environmental
regulators are supportive of full scale remediation using this
technology and the decision to use composting is expected in 1991.
Introduction
•
The Army is focusing it's technology development efforts on
the remediation of Army-owned sites which have been contaminated
from past operations. The production and handling of conventional
munitions has resulted in the generation of approximately one
million tons of explosives-contaminated soils at various military
installations. The principle explosive contaminants are
trinitrotoluene (TNT), cyclotrimethylenetrinitramine (RDX), and
cyclotetramethylenetetranitramine (HMX). There are three types of
operations which, in the past, resulted in explosives contaminated
soils: the manufacture of TNT, the load assemble and pack of
explosives into munitions and the demilitarization of obsolete
munitions. Each of these operations generated large amounts of
water contaminated with explosives. In the past it was considered
acceptable practice to discharge this contaminated water into
unlined washout lagoons.
There are two risks associated with explosives-contaminated
soils. The first is a reactivity hazard in which soils with
concentrations of explosives greater than 12 percent can propagate
a detonation. This does not preclude any explosion if explosives
content is below 12 percent but describes a limit below which
propagation will not occur. sThe second risk is the toxicity
hazard in which low concentrations of explosives in drinking water
479
-------
poses a health hazard. Although explosives are relatively
insoluble, these contaminants have at some installations migrated
into the groundwater. The Army is now actively pursuing means to
prevent this groundwater intrusion by remediating these sites.
Because of the high risk of groundwater intrusion, the Army has
incinerated explosives-contaminated soils at Cornhusker Army
Ammunition Plant and at Louisiana Army Ammunition Plant.
Until now, incineration has been the only available
technology for the remediation of explosives-contaminated soils.
This technology is publicly undesirable and essentially
economically infeasible for the remediation of small sites due to
the required disproportionate capital expenditures for the
incineration systems. The Army has identified 26 active or
inactive Army ammunition production and processing properties on
the National Priorities List (NPL) for explosives-contaminated
soils. Due to the large volume of explosive-contaminated soils,
estimated incineration costs for all of these sites could exceed
$1.5 billion.
Composting Process
Composting is a controlled process by which readily available
organic substrates are biologically degraded by microorganisms to
inorganic and stabilized organic by-products (compost). In most
cases this i? achieved by the use of indigenous microbial
consortia. The composting process is initiated by mesophilic
bacteria which are biologically active at temperatures between 30
and 45 C. As the organic substrates are degraded, the heat from
the biochemical degradation increases the temperature to
thermophilic temperatures, between 50 and"60°C. The mesophilic
bacteria are replaced by thermophilic bacteria. If not
controlled, the temperature could easily exceed 70 C; a
temperature at which biological degradation ceases. In order to
achieve maximum efficiency, the optimum biological activity is
controlled by maintaining moisture content, pH, oxygenation,
temperature, and the carbon-nitrogen ratio. Various options are
available for conducting composting which range from simple,
uncontrolled static piles to complex in-vessel composters which
mix and aerate. The selection of which type of composting
operation depends on the amount of waste to be treated, the amount
of area available for facilities, the degradation characteristics
and the regulatory requirements applicable to a given site.
Hazardous Waste Composting
While a significant amount of information is available on the.
design and operation of municipal waste compost facilities, there
is very little information available on hazardous waste
contaminated soils composting. The difference is in the process.
In municipal waste composting, volume reduction, pathogen
destruction and dewatering are the primary objectives. For
contaminated soils composting, the principle objective is
480
-------
reduction of the contaminant(s) of concern to innocuous end
products. Typical municipal waste treatment starts with a
carbonaceous waste stream to which minimal additional materials
are added. Normally these materials are bulking agents to improve
porosity. For contaminated soils, addition of a carbonaceous
amendment is typically necessary to provide sufficient substrate
for microbial action. This amendment mixture increases the volume
of the waste, requires additional materials handling and involves
an operating cost to purchase amendments. The contaminated soil
itself places a thermal burden on the compost system which must be
accounted for in the operation. Finally, reaction times for
hazardous waste treatment using composting can be significantly
longer than those for municipal wastes and cost of operations as
well as design strategies can be very different.
Previous Studies on Composting of Explosives
The development of composting as a technology for explosives
contaminated soil treatment has followed the classical research
and development cycle, starting with laboratory studies, moving
through bench scale studies to field scale demonstration.
Investigation into biological treatment of explosives, paticularly
TNT, began in the early 1970's. ' ' In 1982, serious laboratory
and pilot scale efforts were initiated to evaluate composting as a
remediation technology. Results indicated that TNT, RDX and HMX
could be biologically degraded in hay-horse feed and manure
composts. ' Explosives degradation was relatively fast and no
toxic or objectionable degradates accumulated-in the compost
materials.
In 1988, a field-scale composting of explosives-contaminated
soils demonstration was conducted at Louisiana Army Ammunition
Plant (LAAP). Static piles were constructed by mixing lagoon
sediments that were contaminated with TNT, RDX, and HMX with
horse manure, straw, alfalfa, horse feed and fertilizer and
composted for 153 days. A schematic of the static compost piles
is provide in figure 1. Both thermophilic (55 C) and mesophilic
(35 C) temperatures were compared. In the thermophilic static
pile the mean percent degradation of TNT, RDX and HMX was 99.9,
99.1 and 96.5 weight percent while the mesophilic pile exhibited
only 99.6, 94.8, and 86.9 weight percent degradation. Results in
one experimental compost pile, shown in figure 2, shows
significant explosives degradation as well as extinction of toxic
intermediates. Explosives degradation was also shown to undergo
first-order decay kinetics. The LAAP composting field scale
demonstration provided proof-of-principle that composting could
reduce the concentration of explosives in soils. To determine if
there was a reduction in toxicity, Oak Ridge National Laboratory,
Oak Ridge, Tennessee conducted toxicity testing of the LAAP
compost leachates. The results indicated a significant reduction
in toxicity in the composted product.
481
-------
COUWMO
\
Hay
bail
j ~j — uoropcair™ ^77;
fj ••• I
n /-- —L*.— •> -
/ .j -
/in ,
y tro^ryrT^ & \w//\
Iff
Woodctto
covwano
baaa
ConcnnaoaourxarXB-thick) ' /J
Figure 1: Schematic of Compost Static Pile used at
Louisiana Army Ammunition Plant.
Cfl
***
•\
Cfl
C
o
c
(U
a
c
o
zs r.b as
Time (weeks)
S7 Jul BB
Figure 2: Explosives Degradation in a Compost Static Pile at
Louisiana Army Ammunition Plant (from Williams, et.al., 1988)
482
-------
Full Scale Design Cost Estimate
A full scale design estimate was prepared which analyzed the
cost of several different scale aerated static pile and mechanical
compost facilities. Aerated static pile method of composting
was considered to be the most viable approach to explosives
composting because mechanical in-vessel composting had not been
demonstrated for explosives composting. Kinetics data from the
LAAP composting field demonstration was used in the cost analysis.
Assumptions used in the the cost analysis included a 50 dollar per
ton compost amendment, 90 day composting period and a requirement
for hardened structures. Direct implementation of composting was
determined to be more expensive than incineration for explosives
contaminated soils. Figure 3 presents costs estimated for both
capital and operations of a static compost facility. The
sensitivity to volume of soil treated is primarily due to capital
costs. Incineration.costs are typically estimated to be between
300 to 600 dollars/ton for installations with less than 20,000
tons of explosives-contaminated soils and 200 to 300 dollars/ ton
for installations with larger volumes. The unit cost for
incineration varies with the amount of soil to be treated be'cause
of the requirements for excavation, mobilization and
demobilization, incineration and ash disposal.
The cost analysis indicated that composting using static
piles could compete with incineration costs if the amendment cost
was could be kept below 50 dollars per~ton and a soil fraction of
greater than 20 percent was utilized. In order to keep the
costs of composting competitive with the costs of incineration,
savings in both capital and operation and maintenance costs were
required. Capital costs are influenced by the type of compost
design (static pile, windrow or mechanical in-vessel composting)
and amount of soil to be treated. Figure 4 presents the effect of
soil fraction on the cost per ton of composting explosives
contaminated soils. Operational costs are influenced by soil
throughput and degradation kinetics. The .effect of degradation
kinetic rates on the cost per ton of composting explosives
contaminated soils is presented in figure 5. Since economics is
the driving force in the implementation of new technologies, a
composting optimization field demonstration was initiated.
Composting Optimization Field Demonstration
A composting of explosives-contaminated soils optimization
field study has recently been completed using explosives
contaminated soil from a National Priority List (NPL).,site at the
Dmatilla Depot Activity (UMDA), Hermiston, Oregon. ' UMDA was
selected not only because of it's harsh environment but, also
because the two explosive washout lagoons contained sufficient
concentrations of TNT, RDX and HMX required to adequately access
the composting degradative process. The. objective of this study
was to maximize soil throughput and obtain the necessary data for
full scale implementation. Since soil throughput is dependent on
the rate of reaction and the percent soil loading, the key
variables investigated in this study were amendment mixture
483
-------
Untt
COM
(Won)
400 -T-
300 --
200
100 " •
Total protect cot
Capital coat
yaar opanrttona and
malntananca coata
•i f-
40,000 80.000 120.000 160.000
Systam capacity (tona traatad In 5 yeara)
200.000
Figure 3: Effects of Facility Size on Cost per Ton of
Sediment using Aerated Static Pile Composting (Lowe, et.al.)
484
-------
a
o
*J
u
Q)
a
w
O
u
700 -
600 •
500 -
400 •
300 •
200 -
100
— • 90 Day Compost Potted
Date from Low. at. aL
10 20 30
Soil Fraction (Volume %)
50
Figure 4: The Effect of Soil Fraction on Cost for Explosives
Contaminated Composting using Static Piles (Lowe, et.al.)
o
o
.4-) '
u
-------
composition, percent soil loading, and mixing. In order to
evaluate the effect of mixing, two compost design technologies
were compared; aerated static pile systems and mechanically
agitated in-vessel system.
Amendment selection was based on adiabatic testing using^a
combination of fifteen readily available agricultural wastes.
Using CO^ respiration as a measure of microbial activity and
compostaBility, two amendment compositions were selected for study
in the optimization test. The first amendment mixture, containing
20% poultry manure, 35% potato processing waste, 15% apple waste
and 30% sawdust, exhibited the highest respiration rate and was
used in the static pile tests. The second amendment mixture,
which was similar to that used in the LAAP composting field
demonstration, contained 50% horse manure, 10% buffalo manure, 32%
alfalfa and 8% horse feed. A comparison of these two amendments
was made in the mechanically agitated in-vessel composting tests.
A survey of the local agricultural businesses indicated that these
amendments could be purchased for approximately 15 dollars per
ton.
Percent soil loading was investigated using seven 3 cubic
yard aerated static pile systems which were constructed out of
fiberglass to model actual static pile conditions. A schematic
of a static pile system is shown in figure 6. Six of these were
used to investigate soil/amendment ratios using 0, 7, 10, 20, 30
and 40 percent by volume of explosives contaminated soil.
Contaminated soil was mixed with the first amendment mixture
containing chicken manure and composted for 90 days.
Soil amendment ratios and different amendment mixture
compositions were investigated using a special seven-cubic yard
pilot scale mechanical agitated in-vessel (MAIV) system.which was
constructed by Fairfield Engineering Co., Marion, Ohio. The
pilot scale system was designed and constructed according to
rigorous explosives safety standards. The mechanical pilot system
uses rotating augurs attached to the rotating cover to mix the
compost. In both systems temperatures were controlled using a
forced aeration system. Since there was only one mechanical
agitated in-vessel system, these tests were performed in
sequential order. The first two tests investigated two different
amendment compositions using 10 percent soil by volume. The final
two tests were used to investigate soil/amendment ratios using 25
and 40 percent by volume of explosives contaminated soil with the
optimum amendment composition selected from the first two tests.
Contaminated soil was mixed with the amendment mixtures and
composted for 44 days.
Both the static pile systems and the mechanically agitated
in-vessel system were housed in greenhouses to protect them from
the environment and prevent the spreading of contamination from
explosives dust. A computer-based data acquisition and control
system was used to regulate the environment" in each of the compost
systems. Temperature, oxygen level and moisture were
automatically sampled and recorded electronically.
486
-------
Deflector
\
To Blower
insulation
To Blower
•Wood Chips
Figure 6: Schematic of an Aerated Static Pile Reactor
used in Umatilla Depot Activity Composting Optimization
Study
Pile/%soil
TNT
RDX
HMX
Static Pile
0% (Control)
7%
10%
20%
30%
40%
Mechanical
10% (Amend #1)
10% (Amend #2)
n/a
90.7
95.9
94.2
97.8
78.8
97.4
99.8
n/a
72.6
46.2
16.2
21.5
-6.5
89.7
99.3
n/a
38.6
21.3
94.7
11.1
1.6
28.8
94.9
n/a - Uncontaminated soil, no explosives present.
Table 1: Percent Reduction of Explosives in the Umatilla
Depot Activity Compost Optimization Study
487
-------
Temperatures were maintained at 55 C using the forced aeration
system and the moisture content maintained between 45 to 55
percent. Compost samples from both systems were taken at various
time intervals, homogenized and split into two fractions. This
sample preparation strategy was developed by the United States
Geological Survey Office in Denver, Colorado. Chemical analysis
for the presence of TNT, RDX and HMX was performed by HPLC methods
while the second fraction is being subjected to toxicity testing.
Oak Ridge National Laboratory is conducting an investigation to
quantify the toxicity of the resultant compost and chemically
characterize the source of any residual toxicity.
Optimization Study Results
Although the majority of the data from the composting
optimization study is unavailable at this time, the data from
selected tests is presented. All three explosives, TNT, RDX and
HMX demonstrated significant reductions in all the tests performed
(Table 1). In the static pile tests, the best degradation of RDX
and HMX was observed in the the 7% contaminated soil test whereas
TNT was relatively unaffected by increasing soil amounts. The
second mechanical agitated test exhibited superior degradation
capabilities, and removed 99.8% of TNT, 99.3% of RDX and 94.9% of
HMX. TNT reduction was observed with all the static pile tests
regardless of the initial concentrations (Figure 7). The majority
of degradation in the static piles occurred in the first 44 days
with insignificant degradation observed after 44 days.
A direct comparison of aerated static piles versus
mechanically agitated in-vessel composting using a 10% soil
loading indicates that mixing significantly increases the
degradation rate (Figure 8). The majority of the degradation in
both mechanical tests occurred within the first 10 days of the
study. A final report on the Compost Optimization Study will be
available in September 1991 and it's associated report on
Toxicology and Chemical Characterization of Compost will be
available in December 1991.
Full Scale Implementation
Results from the composting optimization study are currently
being used to develop a full scale remediation design for Umatilla
Depot Activity, Hermiston, Oregon. The amount of soil to be
remediated is dependent on the depth of explosives contamination
in the lagoons and the regulatory clean-up criteria to be met.
Although these are currently being determined,, a rough estimate of
8,000 cubic yards has been used in preliminary planning. The
optimization study results indicate that with a composting design
system that provides mixing, composting of explosives-contaminated
soils can be economical. A controlled windrow system is being
considered for this site based on the amount of soil for treatment
and the disproportionate capital costs associated with large
mechanical in-vessel composters. Using a 15 dollar amendment
mixture composition, a 60 day retention time, 20 percent
488 /
-------
00
ID
CONCENTRATION (ppm)
-------
contaminated soil loading rate and the kinetic rates from the
optimization study, the initial estimates indicate that composting
can be implemented at approximately 200 to 250 dollars a ton.
Rough estimates for incineration for the same volume of soil are
500 to 600 dollars a ton. .These are only preliminary estimates
and are subject to change based on the final results of the
optimization study, amount of explosives contamination in the
lagoons, the specified clean-up levels and the toxicity results.
Conclusion
Composting offers a cost-effective innovative technology to
remediate explosives contaminated soils. The Army is proposing
the use of composting to remediate the explosive washout lagoons
at Umatilla Depot Activity, Hermiston, Oregon. Environmental
regulators are supportive of a full scale remediation using this
technology and the decision to use composting is expected in 1991.
If implemented, it will be the first application of composting to
remediate hazardous wastes at an NPL site.
This publication reflects the personal views of the author(s) and
does not suggest or reflect the policy, practices, programs, or
doctrine of the U.S. Army or Government of the United States.
References
1. Balasco, A.A. , 1987, Testing to Determine the Relationship
Between Explosive Contaminated Sludge Components and Reactivity,
Final Report, U.S. Army Toxic and Hazardous Materials Agency,
Aberdeen Proving Ground, MD, Report No. AMXTH-TE-CR-86096.
2. Turkletaub, R.B. and Wiehl, C.D., U.S. Army Installation
Restoration Incineration Program at Cornhusker and Louisiana Army
Ammunition Plants, Proc. SUPERFUND '88, pp. 569-574, HMCRI, Silver
Spring, MD, 1988.
3. White, M.K., and Fow, C.L., 1990, USATHAMA Installation
Restoration Program Research and Development Strategies, Final
Report, U.S. Army and Toxic Hazardous Materials Agency, Aberdeen
Proving Ground, MD, Report No. CETHA-TE-CR-90002.
4. Anderson, J.G., and Smith, J.E., Composting, in Biotechnology
of Waste Treatment and Exploitation, ed. J.M. Sidwick and R.S.
Holdom, pp.301-321, Ellis Harwood Limited, Chichester, -England,
1987.
5. Osmon, T.L. and Andrews, C.C., 1978, The Biodegradation of TNT
in Enhanced Soil and Compost Systems, Final Report, U.S. Army
Armament Research and Development Command, Dover, NJ, Report No.
ARLCD-TR-77032.
490
-------
6. Hoffsomraer, J.C., Kaplan, L.A., Glover, D.J., Kubose, D.A.,
Dickinson, C., Goya, H., Kayser, E.G., Groves, C.L., and Sitzman,
M.E., 1978, Biodegradability of TNT: A Three-Year Pilot Study,
Final Report, Naval Surface Weapons Center, Crane, IN, Report No.
NSWC/WOL TR 77-136.
7. Klausmeir, R.E. and Jamison, E.I., 1982, Composting of
TNT:Airborne Products and Toxicity, Final Report, U.S. Army
Armament Research and Development Command, Dover, NJ, Report No.
ARLCD-CR-81039.
8. Isbister, J.D., Anspach, G.L., Kitchens, J.F., and Doyle,
R.C., 1984, Composting for Decontamination of Soils Containing
Explosives, Microbioloqica. 7:47-73.
9. Doyle, R.C., Isbister, J.D., Anspach, G.L., and Kitchens,
J.F., 1986, Composting Explosives/Organics Contaminated Soils,
Final Report, U.S. Army Toxic and Hazardous Materials Agency, c
Aberdeen Proving Ground, MD, Report No. AMXTH-TE-CR-86077.
10. Williams, R.T., Ziegenfuss, P.S. and Marks, P.J., 1988, Field
Demonstration-Composting of Explosives-Contaminated Sediments at
Louisiana Army Ammunition Plant (LAAP), Final Report, U.S. Army
Toxic and Hazardous Materials Agency, Aberdeen Proving Ground, MD,
Report No. AMXTH-IR-TE-88242.
11. Greist, W.H., Stewart, A.J., Tyndall, R.L., Ho, C.H., and
Tan, E., 1990, Characterization of Explosives Waste Decomposition
Due to Composting, Final Report, U.S. Army Medical Research and
Development Command, Ft. Dietrick, MD, Report No. DOE IAG
1016-8123-A1.
12. Lowe, W., Williams, R. and Marks, P., 1989, Composting of
.Explosive-Contaminated Soil Technology, Final Report, U.S. Army
Toxic and Hazardous Materials Agency, Aberdeen Proving Ground, MD,
Report No. CETHA-TE-CR-90027.
13. Williams, R.T., and Myler, C.A., Bioremediation using
Composting, BioCvcle, November 1990.
14. Myler, C.A, and Williams, R.T., Composting
Explosives-Contaminated Soil, Proc. National Research and
Development Conference on Control of Hazardous Materials, pp.
320-325, HMCRI, Anaheim, CA, February 1991.
491
-------
nor
SELECTION AND PERFORMANCE OF A GROUND WATER EXTRACTION SYSTEM
AT SACRAMENTO ARMY DEPOT WITH WATER TREATMENT
USING
ULTRAVIOLET RADIATION/HYDROGEN PEROXIDE OXIDATION
David O. Cook, P.E.
Anne Marie Collins1
ABSTRACT
Ground water beneath the Sacramento Army Depot (SAAD) is being extracted and
treated as part of a ground water remediation program. The extraction system is achieving
contaminant capture, and the ultraviolet radiation/hydrogen peroxide (UV/H2O2) unit is
achieving necessary treatment at less than design power dosage.
INTRODUCTION
The SAAD was established as the Sacramento Signal Depot at its present location in April,
1945 and was redesignated as the Sacramento Army Depot in August, 1962. The SAAD is
an electronics supply depot responsible for the storage, issue, maintenance and repair of
assigned electronics supplies and commodities.
The SAAD occupies approximately 485 acres in Sacramento, California. The SAAD is
located approximately seven miles southeast of the city's central business district. Figure 1
shows the SAAD location in relation to the center of Sacramento.
From 1965 to 1974, the SAAD was involved in various operations including night vision
device repair and automatic data processing. Current depot operations include shelter
repair, electro-optics, equipment repair, and metal coating and treatment. A variety of
hazardous and flammable materials are reported to have been used at the SAAD including
organic solvents, oils and grease, fuels, lubricants, caustic solvents and metal plating baths.
CONTAMINANTS OF CONCERN AND TREATMENT OBJECTIVES
As a result of the activities described above, and the disposal of associated waste streams,
contaminants have migrated into portions of the' shallow aquifers beneath the SAAD.
Ground water in the southwest area of the depot is contaminated with trichloroethene,
tetrachloroethene, dichloroethene, and eight other volatile chlorinated organics.
The hydrogeologic information available in the area of contamination indicates the
presence of a series of highly permeable channels beneath the site, intercalated with dense
silts and clays. These units have been grouped into four water-bearing zones in the upper
1 David O. Cook is a Senior Engineer, with Kleinfelder, Inc., 9795 Business Park Drive,
Sacramento, California 95827. Anne Marie Collins was, at the time this paper was
authored, an Associate Hydrogeologist with Kleinfelder.
492
-------
SACRAMENTO
BUSINESS \
DISTRICT V
FRUITRIDGE RD
SACRAMENTO
ARMY DEPOT
ELDER CREEK
APPROXIMATE SCALE
012345
KLEINFELDER
PROJECT NO.
SITE LOCATION MAP
SACRAMENTO ARMY DEPOT
SACRAMENTO, CALIFORNIA
M -6
FIGURE
1
493
-------
200 feet: the A zone, from 89 to 107 feet; the B zone from 110 to 148 feet; the C zone from
156 to 188 feet; and the D zone from 195 to 230 feet. The primary areas of concern are the
A and B zones where a number of the priority pollutant volatile organic constituents
(VOCs) are present above EPA maximum contamination levels (MCLs).
The primary contaminants of concern found in the ground waters analyzed are
trichlorethene, (TCE), tetrachloroethene (PCE), trans 1,2-dichloroethene (DCE), 1,2-
dichloroethane (DCA), and chloroform. Ground water constituent concentrations used as
a design basis for the treatment system and the treatment objectives are shown in Table 1.
TABLE 1
SAAD GROUND WATER TREATMENT
SYSTEM DESIGN CONCENTRATIONS
Contaminant
-;.:.,' v
chloroform
carbon tetrachloride
trichloroethene
tetrachloroethene
1,1-dichloroethane
1,1-dichloroethene
1,1, 1-trichloroethane
trans-l,2-dichloroethene
1,2-dichloroethane
methylene chloride
bromodichloromehane
Average
ug/1
4.9
0.4
64.2
10.3
0.5
0.3
0.1
8.6
0.9
0.3
0.2
Design
ug/1
7.4
0.6
96.3
15.5
0.8
0.5
0.2
12.9
1.4
0.5
0.3
Treatmeii
Objective
._ .*.'.
ug/1 1
100
5
5
4
20
6
200
16
1
40
100
The average concentration is the average for all monitoring through 1988. The design
value is 50 percent higher than the average concentration and was used as an upper limit
for process evaluation. The treatment objective is the lower of the MCL or California
Department of Health Services (DHS) applied action level as existed at the time of the
design.
Treatment objectives were applied both as a goal for ground water remediation and a
quality objective for treating extracted water. Modeling indicated the need to extract
ground water at up to 360 gpm (which was used in the operable unit feasibility study).
SELECTION OF THE APPROPRIATE TECHNOLOGY
The remedial action selection process consisted of identification of appropriate remedial
response actions and technologies, initial screening of remedial technologies, development
and screening of preliminary remedial action alternatives, and detailed analysis of the
remedial alternatives. The ground water remedial alternatives considered were: air
stripping with and without the use of vapor-phase activated carbon; liquid phase carbon
494
-------
adsorption; and UV/H202 Oxidation. Based on the ability of UV/H202 to permanently
destroy the hazardous constituents, the UV/H202 process was chosen as the remedial
alternative of choice.
UV/H2O2 TECHNOLOGY
Unlike conventional treatment technologies such as air stripping or carbon adsorption
which transfer contaminants to air or activated carbon, which, in turn, require further
treatment or disposal, UV/H202 destroys the organic contaminants. The UV/H2O2
process uses UV light in conjunction with hydrogen peroxide (H2O2) to react with and
oxidize organic compounds to carbon dioxide (CO2), water, and inorganic salts.
UV/H2O2 destruction of organic compounds in water is a complex process involving the
synergistic interaction of UV radiation and chemical oxidizers. Destruction is believed to
proceed by each of the following interrelated mechanisms:
4 Direct photolysis using UV
* Chemical oxidation
Direct Photolysis using UV
Adsorption of light by organic molecules increases the internal energy of the molecule.
UV light at 180 to 400 nm can provide 72 to 155 kcal/mole of energy (Prengle, H.W. Jr.,
1983), which is sufficient to break bond of most organic molecules. Table 2 gives the
dissociation energy for several bond types.
TABLE 2
DISSOCIATION ENERGY FOR SELECTED ORGANIC MOLECULAR BONDS
Bond
C-C
c=c
c=c
C-F
C-CI
C-S
c=s
C-H
-O-O-
O-O(O2)
Dissociation Energy, kcal/gmol
82.6
145.8
199.6
116.0
81.0
65.0
166.0
98.7
47.0
119.00
495
-------
As shown, UV light can provide sufficient energy to break bonds encountered in organic
molecules. The rate of photolysis depends on the intensity of the UV light and the time a
given molecule is exposed, i.e., the UV dosage. With enough UV dosage, organic
molecules eventually are oxidized to CC>2, water, and inorganic salts. For UV, the UV
dosage is related to the electrical power fed to the lamps.
Chemical Oxidation
Chemical oxidants, such as H^O^ destroy organic compounds by attacking bonds of the
molecule and eventually cleaving these bonds to form smaller compounds, given sufficient
time, this process continues until only CO?, Hy), and inorganic salts remain. t^O? reacts
with H^O and/or decomposes to form free radicals and atomic oxygen (Clarke, N.,
Knowles, G., 1982; Prengle, H.W. Jr., 1983), which account for a significant fraction of the
oxidation potential of each oxidant. Hydroxyl radicals preferentially attack organic
compounds by reacting with hydrogen atoms of the organic molecule, resulting in the
formation of organic radicals, which can initiate other reactions (Clarke, M., Knowles, G.,
1982). Free radicals reactions occur rapidly; therefore, the rate limiting factor in the
overall chemical oxidation mechanism is the oxidant decomposition.
A Combination of Mechanisms
In the presence of UV radiation, the rate of oxidant recomposition (such as H2C»2) is
accelerated, with a corresponding increase in the rate of hydroxyl radical formation.
Organic molecules that have adsorbed UV energy are in an excited state and are more
susceptible to attack. Therefore, the rate at which organic compounds are oxidized is
significantly higher than that attained by using UV radiation or chemical oxidants alone.
Efficiency of UV/H202
Previous studies of the UV/H2O2 process indicate that the overall reaction mechanism
displays first-order rate kinetics with respect to the contaminant concentration, oxidant
dosage, and UV intensity (Prengle, H.W. Jr, 1983). For constant oxidant dosage, the rate
of oxidation can be approximated.
C0e
-kt
Where C = constituent concentration, ug/1
k
I
t
= initial constituent concentration, ug/1
= rate constant for the constituent, gal/KW-min
= total UV intensity in reactor, KW/gal
= oxidation time, min
The required removal efficiency depends on the product of UV intensity and oxidation
time, i.e., the oxidation of a compound at constant oxidant dosage depends on the energy
delivered (UV dosage). As the UV intensity increases the oxidation time required to
496
-------
achieve a desired removal decreases because the organic constituent still receives the same
amount of energy. The UV dosage delivered by a particular reactor can be approximated
by:
D
= It = Pt/V
Where D = UV dosage, KW-min/gal
P = total UV lamp power, KW
V = reactor volume, gal.
The UV dosage is the critical parameter for UV/H2O2 systems. The UV dosage is
adjusted by installing more or fewer lamps in the reactor. Power consumption can be
optimized in the field by turning lamps on and off.
Organic carbon, soluble iron and manganese, and general turbidity can reduce the
efficiency of the UV/H^C^ process by reducing the amount of UV energy available for
adsorption by the organic contaminants and the chemical oxidant. Organic carbon will
compete with the constituent of concern by adsorbing UV energy and consuming oxidant.
Soluble iron and manganese will oxidize to their soluble form, thereby directly competing
with the contaminant for UV energy and oxidant. Highly turbid H20 will reduce UV
intensity in a similar manner.
BENCH TESTS
Bench scale tests of an UV/Chemical Oxidation process were performed to evaluate the
effectiveness of this technology for the treatment of the SAAD ground water. Results of
bench scale testing were used to assist in the design of full-scale equipment and in the
development of process operating parameters such as UV dosage and H202 feed rate.
Bench tests were conducted on ground water samples collected from the SAAD site. The
effect of UV and H2O2 dosage on removal of organic contaminants from the SAAD
ground water was assessed using the apparatus shown schematically in Figure 2. A high
intensity UV lamp enclosed in a quartz tube was mounted axially in a cylindrical stainless
steel reactor. A variable speed pump provides process flow from a glass feed reservoir. A
variable speed pulseless metering pump provides H2O2 injection. A heat exchanger is
provided to control process temperature, as necessary. The test apparatus is closed to the
atmosphere, and all wetted parts are either stainless steel, quartz, glass, or Teflon. Three
tests were performed each with different UV and H2O2 dosages to evaluate the effect on
oxidation efficiency.
EPA Method 601 results for the three UV/H2O2 process bench tests found that the
UV/H2O2 process reduced all contaminants of concern to nondetectable levels before the
first samples were collected. No intermediate organic compounds resulting from oxidation
of the SAAD ground water contaminants were detected using Methods 624 or 625.
TCE was the anticipated rate-limiting constituent for the UV/H2O2 oxidation process. A
TCE removal efficiency above 95 percent, which is higher than that needed for any other
contaminant in the SAAD ground water, must be achieved to meet the treatment objective
497
-------
sn~ji;pjia.-—
G7 ^
POWER
PANEL
uv
INDICATOR
STATIC MIXER IS
HEAT1
EXCHANGER
SAMPLE
-TO
DRAIN
KLEINFELDER
PROJECT NO.
M-6
UV/H202 TEST
APPARATUS FLOW SCHEMATIC
SACRAMENTO ARMY DEPOT
SACRAMENTO, CALIFORNIA
FIGURE
498
-------
of 5 ug/1 (see Table 1). To assess the power dosage needed to achieve the SAAD
treatment objectives, the rate constant for TCE oxidation was first calculated.
=-I/It In (Ce/Ci)
= TCE oxidation rate constant, gal/KW-min
= total UV intensity in the reactor, KW/gal
= oxidation time, min
= effluent TCE concentration, ug/1
= influent TCE concentration, ug/1
Where kTCE
I
Although oxidation reaction proceeded too rapidly to allow an accurate quantification of
the rate content, the minimum rate content applicable for these bench tests can be
estimated by assuming the effluent concentration for the first sample collected was at the
analytical detection limit for TCE of 0.12 ug/1. for the lowest H2C«2 dosage, the calculated
rated constant is 3.0 gal/KW-min. The influent concentration was 106 ug/1, the oxidation
time was three minutes, and the total UV reactor intensity was 0.75 KW/gal.
The required energy dosage for the full-scale design can be calculated from the bench test
data directly using the equation:
D -It =-l/k(Ce/Q)
Where D = OV dosage, KW-min/gal
Solving the above using the calculated rate constant of 3.0 gal/KW-min, the maximum UV
dosage needed to reduce the expected TCE concentration from 120 ug/1 to the treatment
objective of 5 ug/1 is 1.0 KW-min/gal with an influent John Boss concentration of 15 mg/1.
Due to inherent inefficiencies iri the bench test unit, it was estimated that a full scale unit
would necessarily be set a power dosage of .67 KW/gpm to achieve treatment objectives at
the design conditions.
REMEDIATION SYSTEM DESCRIPTION
The extraction system consists of seven stainless steel extraction wells. Hydrogeologic
information gathered during the remedial investigation indicated approximately 360
gallons-per-minute would have to be pumped from the zone A and B aquifers to intercept
the ground water plume and prevent further migration.
Computer-assisted modelling of the site using MODFLOW was used to select initial area
spacing. Spacing was adjusted during construction based on the field test data. The wells
as installed range from 150 to 300 feet apart with varying pumping rates, depending on the
production from the sand unit in which the particular well is screened.
One of the wells is located in the center of the contaminant plume, with the remaining six
located along the facility boundaries in a "fence" arrangement to act as a hydraulic barrier.
The wells were screened to intercept the discontinuous, high producing sand units targeted
for extraction. The extraction well pumping rates range from 25 to 100 gallons-per-minute,
depending on the capacity of the target extraction zones.
499
-------
The treatment system installed at the site is a skid-mounted, transportable, modular unit.
The components include the UV reactor, lamps and power transformers, and the hydrogen
peroxide feed systems. UV lamps in quartz sheaths are installed in a horizontal
arrangement and evenly distributed throughout the reactor. Figure 3 is a schematic of the
UV/H2O2 unit. The system is capable of up to 1.0 KW/gpm. Hydrogen peroxide dosage
can be varied as required.
SYSTEM OPERATION
The ground water extraction and treatment system installed at the SAAD began operating
in November, 1989. Early operation experienced random shutdowns. The cause of these
shutdowns was unknown. Troubleshooting checks revealed that the flow meter in the
UV/H2O2 unit was malfunctioning. It would register errant readings and then zero out
during apparently normal operation, causing the treatment system to shut down. The UV
interlock would open and de-energize a flow-switch which would, in turn, de-energize the
Master Control Relay (MCR) and open circuits to the pumps and the solenoid valve
controlling the 3-way diversion valve. The flow switch was found to be "sticky" and its
replacement was also recommended. These situations apparently were the cause of the
random shutdowns. In September 1990, an electrical fire occurred in the unit. Repairs
were effected by the manufacturer. Other than the above, operation and maintenance of
the system has been routine.
EXTRACTION SYSTEM PERFORMANCE
The extraction wells and treatment plant were completed in November 1989. The influent
of the system was first sampled on November 30. The following TCE concentrations were
detected in the treatment plant influent samples:
* November 30, 1989
4 December 6,1989
* January 4,1990
* February 15,-1990
* March 1,1990
* May 14,1990
* June 18,1990
* October 19,1990
* November 16,1990
* December 19,1990
4 January 16,1990
4 February 19,1991
TCE 62 ug/1
TCE 55 ug/1
TCE 42 ug/1
TCE 49 ug/1
TCE 34 ug/1
TCE 28 ug/1
TCE 48 ug/1
TCE 21 ug/1
TCE 16 ug/1
TCE 0.6 ug/1
TCE 18 ug/1
TCE 23 ug/1
Since the treatment system started, there has been a steady decrease in the TCE
concentrations detected in the influent. Since the extraction was started in November of
1989, the lateral extent of TCE contaminant has been reduced.
From December 1989 through December 1990 this pumping activity was evident within the
A and B zones. The overall drop in ground water elevations in the vicinity of the extraction
system has ranged from approximately 2.5 to 5.0 feet. Pumping of the extraction wells has
increased the overall ground water drawdown near the southwest comer of the SAAD and
has influenced the flow direction and ground water gradient southwest of the SAAD. A
500
-------
Mound water elevation contour for the extraction zone in the winter of 1991 is provided as
gure 4. Historical trend charts for both of the upper water zones from our extraction
area monitoring well is shown as Tables 3 and 4.
TREATMENT SYSTEM PERFORMANCE
Power dosage in the field has generally varied from 0.3 to 1.0 KW/gpm and hydrogen
peroxide residuals (i.e., the measured H2O2 in the effluent of the reaction chamber) have
varied from about 8 ppm to over 20 ppm. Design treatment efficiency for TCE has been
achieved at approximately 0.35 KW/gpm as a residual hydrogen peroxide level of 8 - 10
ppm. Treatment efficiencies of TCE appear to drop off sharply at about .3 to .4 KW/gpm
and at hydrogen peroxide levels below about 6 ppm.
Current actual required treatment efficiency for TCE at this site is calculated to be about
80 to 90 percent; considerably lower than the design due to the lower than design influent
TCE concentrations. The unit has generally, been operating with effluent TCE levels at
ND (less than 0.4 ug/1) or slightly higher, rather than the 5 ug/1 assumed for design.
TableS is treatment data expressed as power dosage versus removal efficiency as a
percentage. In calculating efficiency, the detection limit was used for ND. Therefore, the
result may be artificially dependent on the influent concentrations. There is currently only
limited data at removal efficiencies below 95 percent on which to base further conclusions
regarding recommended power dosage and hydrogen peroxide residual for routine
operation.
CONCLUSIONS
The SAAD is successfully remediating ground water beneath the southwest corner of the
site with extraction and treatment. The treatment process employs a combination of
ultraviolet radiation and hydrogen peroxide oxidation. Treatment efficiency has been
higher than design at lower power dosages. The system was chosen due to it's ability to
destroy the contaminants completely with no "side effects".
LIMITATIONS
Only limited data has been gathered to date at this site. The operation may not be
optimized. Lower power dosages may be feasible. The manufacturer of the UV/H2O2
unit indicates the research and development effort are reducing power demands for newer
units. On the other hand, this site has not had scaling, which has been observed at two
other sites. The results should be considered site-specific and not extrapolated for other
situations.
501
-------
Ol
o
ro
CROUNDWATER TREATMENT SYSTEM
SACRAMENTO ARMY DEPOT
..*
FIGURE 3
-------
January 1991 Ground Water Elevations - Zone B
313500
. \. \- \.
309000
308500 •
308000 •
307!
2^9000 2170000
LEGEND
2171000
2172000
2173000
2174000
2175000
MW-1 MONITORING WELL LOCATION
GROUND WATER ELEVATION CONTOUR
(ELEVATION IN FEET REFERENCED TO
MEAN SEA LEVEL)
1000
I
2000
I
SCALE: r.ioar
HI KLEIN FELDER
PROJECT NO.
GROUND WATER ELEVATION CONTOURS
SACRAMENTO ARMY DEPOT
SACRAMENTO, CALIFORNIA
FIGURE
4
(£) Kleinfelder Inc., 1991
503
-------
frOS
Concentration (ug/L)
Elevation (ms\)
i
May
Fed
£» May
Apr
May
Aug
Sep
i s i * s a s
> o o o o o e
it.,,.
/
4
\
<
""^
/
J
./'
r
^
I
''
<.
I
i
_/
/
/
/
/"
\
J
^^^ — ~*~-r^--
__^^ — ^^"
S
V
•%.
^^^-~^^^
_^^^-^^'
*
\^
^/
^^
•*
c
I
&
i
o
gr
CD
GO
-------
90S
Concentration (ug/L)
Elevation (msl)
''
Oct87
New
Dec
JanU
Feb
, Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Doc
Jan89
Feb
Mar
Apr
§ Jun
5 Jul
Aug
Sep
Oct
Nov
Dec
JanSO
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Jan91
Feb
k * i k i s s
a b o "o o o o
t i . i ^.
T
I
\
\
\
|
?
j
. j
J
f
J
;
-
\
/
/
/
f
t/
\
I
:*=^^——~~=^
-------
SAAD Treatment System Operating Data
H2HO <
ppm
K202
8-10 ppm
01
H2O2
11-20
ppm
H2O2 >
21 ppm
DATA #
10
11
12
16
28
9
30
32
35
44
15
3
4
13
14
18
19
37
42
2
5
6
7
8
17
20
25
40
22
27
POWER
DOSAGE
1.0
1.0
1.0
0.3
0.3
1.0
0.3
0.4
0.4
0,6
0.3
1.0
0.8
0.3
0.3
0.3
0.3
0.5
0.6
0.6
0.7
0.5
0.3
0.2
0.4
0.5
0.3
0.6
0.5
0.2
POWER
KH
360.0
360.0
360.0
120.0
120.0
360.0
120.0
120.0
120.0
120.0 "
120.0
360.0
300.0
120.0
120.0
120.0
120.0
120.0
120.0
180.0
240.0
180.0
120.0
60.0
120.0
120.0
120.0
120.0
120.0
60.0
FLOW
gpm
360.0
360.0
360.0
360.0
370.0
360.0
370.0
290. Q
270.0
190.0
360.0
360.0
360.0
360.0
360.0
350.0
350.0
260.0
210.0
320.0
360.0
360.0
370.0
370.0
360.0
250.0
350.0
200.0
250.0
370.0
H202 %
PPM
6.0
3.0
1.0
2.0
0.1 .
10.0
10.0
10.0
10.0
10.0
8.0
15.0
20.0
15.0
12.0
18.0
17.0
12.0
15.0
25.0
25.0
25.0
25.0
25.0
25.0
22.0
23.0
25.0
30.0
28.0
removal%
TCE
99.6
97.3
81.6
75.5
86.6
99.6
93.1
96.1
96.4
96.7
91.0
99.6
99.6
96.2
95.3
95.2
96.2
95.8
96.7
99.4
99.6
99.1
97 .4
81.7
97.7
96.6
97.3
97.6
99.5
77.4
removal%
1,2-DCE
96.2
97.8
80.5
71.6
90
96.3
95.3
94.9
94.2
93.9
96.4
96.3
96.3
96.3
95.6
97.3
92.5
94.5
97.7
96.1
96.8
96.6
88.4
95.4
96.2
96
96.1
95.8
87.5
removal*
CHCL3
42.1
15.4
20.6
10
21.4
63.3
33.1
52.9
53.3
42.9
45
97.1
60.9
58.9
68
18.75
35.7
45
66.7
72.2
65.6
10.5
74.2
55.3
58.6
31
63.9
31.3
removal*
1,2-DCA
41
44.7
3.5
14.5
12.8
29.5
50
39
22.8
54.5
58.3
38.8
17.8
48
47.3
61.5
50
74.5
62.1
58.3
67
62.1
70
52
50
removal
PCE
77.8
50
81.1
66.1
84.6
72.9
84.6
85.7
87.5
75.3
70
60.8
73.7
72.6
66.7
78.5
90.5
77.8
55.6
63.6
58
66.7
67.2
76.7
84.6
87.5
89
68.2
Table 5
-------
-------
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Environmental Protection
Agency
Information
Cincinnati OH 45268-1072
Official Business
Penalty for Private Use, $300
EPA/540/2-91/015
-------