United States
Environmental Protection
Agency
Office of Research and Office of Solid Waste and
Development • Risk Reduction Emergency Response
Engineering Laboratory
September 1989
Forum on Innovative Hazardous
Waste Treatment Technologies:
Domestic and International
Atlanta, Georgia
June 19-21,1989
Technical Papers
-------
FOREWORD
As a result of the high level of interest in innovative hazardous waste control
technologies, U.S. EPA's Office of Solid Waste and Emergency Response (OSWER) and
Risk Reduction Engineering Laboratory (RREL) jointly conducted this conference. The
conference consisted of presentations of technical papers and posters by international
and domestic vendors of technologies for the treatment of waste, sludge, and
contaminated soils at uncontrolled hazardous waste disposal sites.
The purpose of the 21/£-day conference was two-fold: to help introduce promising
international technologies through technical paper and poster displays; and to
showcase results of the U.S. EPA Superfund Innovative Technology Evaluation (SITE)
program technologies in addition to other domestic innovative technologies. Both
were aimed at increasing awareness of the user community in technologies ready for
application.
This compendium does not include all papers that were presented; only those
that were made available by authors and their institutions are included. A
publication containing one-page abstracts of each presented paper will be available
from EPA's Center for Research Information (CERI) in the Fall of 1989. Subsequent
inquiries may. be addressed to CERI, P.O. Box 12505, Cincinnati, OH 45212.
Although this document has been published by the U.S. Environmental
Protection Agency, it does not necessarily reflect the views of the
Agency, and no official endorsement should be inferred. Mention of
trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
TABLE OF CONTENTS
Page
Experience with the Harbauer PBS Soil Cleaning System 1
H.D. Sonnen, W. Groschel, M. Nels, Harbauer GmbH
Remediation and Treatment of RCRA Hazardous Wastes by Freeze
Crystal 1 i zati on 18
James Heist, Freeze Technologies Corporation
Purification by Froth Flotation 33
Cas Mosmans, Mosmans Mineraaltechniek
Reverse Osmosis: On-Site Treatability Study of Landfill Leachate 41
at the PAS Site in Oswego, NY
Charles Goulet, Septrotech Systems Incorporated
Electro-Reclamation in Theory and Practice 57
R. Lageman, Geokinetics
Vacuum Extraction Technology, SITE Program Demonstration at 77
Groveland Wells Superfund Site, Massachusetts
James J. Malot, Terra Vac
UV/Oxidation of Organic Contaminants in Ground, Waste, and 92
Leachate Waters
David B. Fletcher, Eriks Leitis and Due H. Nguyen, Ultrox
International
In Situ Steam/Air Stripping 112
Phillip LaMori and Jeff Guenther, Toxic Treatments Inc.
The Simplest Way to Clean Water 116
Ralf F. Piepho, The Piepho Corporation
Regional Biological Decontamination Centers for the Clean-up of
Contaminated Soil, Sludges and Industrial Wastewaters 124
Hein Kroos, Biodetox GmbH
Biological Remediation of Contaminated Groundwater and Soil -
Concepts of Remediation and their Technical Application 139
M. Kastner, Technical University of Hamburg-Harburg
K. Hoppenheidt and H.H. Hanert, Institute of Microbiology,
Technical University of Braunschwieg Biocenter
The Holzmann System for In-Situ Soil Purification 151
Hans G. Bathus, Philipp Holzmann AG
Biological Regeneration of Contaminated Soil 157
Volker Schulz-Berendt, Umweltschutz Nord GmbH & Co.
-------
TABLE OF CONTENTS (Continued)
Seven Years Experience in Thermal Soil Treatment 161
Rudolf C. Reintjes and Cees Schuler, Ecotechniek bv
Contaminated Soil Remediation by Circulating Bed Combustion 172
Robert G. Wilbourn and Brenda M. Anderson, Ogden Environmental
Services, Inc.
Residues from High Temperature Rotary Kilns and Their teachability.... 195
Ronald Schlegel, W&E Environmental Systems
Recycling of Contaminated River and Lake Sediments Demonstrated by
the Example of Neckar Sludge 231
H. NuBbaumer and E. Beitinger, Ed. Zublin AG
Oxygen Enhancement of Hazardous Waste Incineration with the
Pyretron Thermal Destruction System 241
Mark Zwecker, Fred Kuntz and Gregory Gitman, American
Combustion, Inc.
Process Description and Initial Test Results with the Plasma
Centrifugal Reactor 263
R.C. Eschenbach, R.A. Hill and J.W. Sears, Retech, Inc.
Superfund Innovative Technical Evaluation Findings and Conclusions.... 280
Timothy E. Smith, HAZCON Engineering, Inc.
SITE: Fixation of Organic and Inorganic Wastes/Intimate Mixing
Technique 289
Carl L. Brassow, J.T. Healy and R.A. Bruckdorfer, Soliditech,
Inc.
Advanced Chemical Fixation of Organics and Inorganics/In-Situ
Treatment 303
Jeffrey P. Newton, International Waste Technologies
-------
EXPERIENCE WITH THE HARBAUER PB3 SOIL CLEANING SYSTEM
Dr. H.D. Sonnen, W. Groschel, M. Nels
1. SUMMARY
The system to be described is an extractive soil washing
system, the HARBAUER PB3, which since July 1987 has been in
operation at the former Pintsch oil refining facility in
Berlin.
To date 20,000 tons of soil from the Pintsch site itself and
from other selected sites has been cleaned by the unit.
Experiences and results from these soil extraction operations
will be described as well as future plans for the application
and use of the technology.
2. INTRODUCTION
Parts of the Pintsch site are heavily polluted from residues of
former used oil refining activities.
The primary pollutant groups which were found in both soil and
ground water were: Mineral Oil, Halogenated Hydrocarbons,
Polycyclic Aromatic Hydrocarbons (PAHs), Polychlorinated
Biphenyls (PCBs), Aromatic Hydrocarbons and Phenols.
In addition Polychlorinated Dibenzodioxine and Dibenzofuran
were found at specific locations.
In order to control the immediate danger and limit the release
and spread of contamination through dust and air emissions as
-------
well as further contamination of the ground water, the Senate
of Berlin initiated a clean-up program in the fall of 1984
under the auspices of the Senator for City Planning and
Environment (2).
THe firm KEMMER/HARBAUER was responsible' for the majority of
the clean-up activities on the site. Major activities included
(4):
Demolition of contaminated buildings, stacks, and
equipment using maximum level protective clothing.
other
- Excavation of soil, removal of existing tanks, and equipment,
and digging out and removal from overflow trenches and
ditches, (work was carried out in maximum as well as in lower
levels of protection including the use of special earth
moving equipment which was equipped with a pressurized air
filtered cabin)
- Providing decontamination stations and protective clothing
for all employees and vehicles.
- Design, building and operation of a ground water treatment
plant.
- Design, building and operation of a soil cleaning plant.
The soil cleaning facility has been in operation since July
1987 and has, in the framework of a Demonstration project under
the auspices of the Berlin Senators for Building Construction,
successfully cleaned 20,000 tons of soil.
-------
EXCAVATION OF CONTAMINAITED SOIL
SITE CLEAN-UP IN PROTECTIVE CLOTHING
-------
3.0 THE HARBAUER EXTRACTIVE SOIL CLEANING SYSTEM
The soil washing system developed by HARBAUER is the result of
a scientific research program in combination with years of
practical experience on-site. The unit is capable of cleaning
contaminated soil and rubble so that the cleaned material can
be refilled on site and is suitable for further un-
restricted/normal use. The contaminant is washed out of the
polluted material and carried away in the liquid/extractant
phase. The process gases and effluent handling systems of the
soil washing facility prevent contaminant from being
transferred to water or air media.
Each soil washing technology has its own critical cut-off point
in the small particle size range where the process can no
longer effectively treat contaminated material. Those particle
sizes smaller than the cut-off point cannot be economically
cleaned and are found in the residual sludges from the washing
process. The advantage of the HARBAUER technology is that the
cut-off point is at 15 urn, which means that not only
scrap/metals, gravel and sand but also silt can be cleaned.
3.1 DESCRIPTION OF THE PROCESS
The principal process steps of the HARBAUER system can be seen
in the flow diagram of the plant (FIG. 1). The facility is
divided into four basic operations:
- Soil preparation and extraction or clean-up
- Clean-up of process waters
- Treatment/dewatering of remaining sludges
- Removal and cleaning of exhausted air emissions
-------
In the Soil preparation/extraction step (process unit 1) the
contaminated soil and rubble is screened and mixed with
extractant then subjected to intense vibrations or oscillations
produced by the use of mechanical energy which set the material
in motion and free the contaminant from the soil particles. It
can then be dissolved in the liquid phase thereby forming an
emulsion of contaminant with the extractant. The effect of the
energy can be enhanced by the addition of cleaning agents ( eg.
biologically degradable detergents).
Through multi-step rinsing, separation and dewatering opera-
tions the cleaned soil particles are recovered from the extrac-
tant medium and removed as clean product. The lower particle
size limit for separation is 15 urn and as such REPRESENTS THE
STATE OF THE ART FOR SOIL WASHING.
Operational costs and requirements for extraction, separation
and dewatering increase disproportionately with decreasing par-
ticle size. HARBAUER is now investigating, under a joint re-
search project with the Ministry for Research and Technology
and the City State of Berlin whether it is technically and
economically feasible to achieve an even lower particle size
cutoff.
Following the extraction step the dislodged pollutant is found
partly as particles and partly emulsified or dissolved in the
water phase. The dissolved/emulsified contaminants are carried
to the water treatment plant (FIG. 1, unit 2) where they are
concentrated out using the following four step process:
- Oil Separation
- Flotation
- Desorption
- Filtration and adsorption on activated carbon
-------
The contaminants removed during the water treatment process are
removed as oily sludges, flotation sludges and loaded activated
carbon.
The particulate contaminant is recovered together with the fine
particle sludge fraction (under 15 urn) and is dewatered with a
filter band press. (FIG 1, unit 3). The amount of residual
sludge is dependent upon the particle size distribution of the
input material and for the soils processed to date is between
5% and 10% of the input.
The degree of pollution of the residual sludge is ultimately
determined by the contaminants in question. Low solubility
contaminants such as heavy metals enrich the sludge phase
whereas high solubility contaminants ( organics eg., benzene )
result in moderate levels of sludge contamination.
The remaining sludge is currently landfilled but investigations
are underway to other more suitable treatment methods (eg.
chemical, thermal, solidification-).
Light materials in the soil such as tar, wood, roots and char-
coal particles will be separated by upward current classi-
fication/fluidized bed sorting and screening using a DSM
screen.
Because of environmental and worker safety reasons the first
three process steps, units 1-3, have an air removal system
which feeds the contaminated exhaust air directly into the air
cleaning unit (unit 4). Air emissions are cleaned in a two step
process using wet scrubbers followed by activated carbon.
In this step any separated volatile materials are recovered as
solvent mixtures with incineration as the indicated method of
disposal.
In the event that a planned crushing unit is added it will be
6
-------
necessary to have a dry air cleaning system (dust filter) to
collect emissions. These residues could then be treated by
solidification.
3.2 PROCESS FLOW CHART
The flow chart for the HARBAUER process is shown in FIGURE 2.
The contaminated soil is brought by a front end loader to the
receiving area after which it passes through an over-sized
particle screen where material larger than 60 mm is removed.
This step separates materials such as debris, wood, and other
miscellaneous large objects which may, depending upon the type
and level of contamination, require special disposal.
The debris free soil is then mixed with water in a blade washer
where lumpy material is broken up by the stirring motion. In
the same step light material (wood particles, charcoal and tree
roots) are separated out. This light particle fraction is
temporarily stored in containers.
The majority of the material from the blade washer goes into a
two step screening process. Material from the first screen
(greater than 5 mm) is rinsed, dewatered and removed as clean
material. The overflow from the second screen goes into the
extraction unit for further treatment. The washed-out fine ma-
terial greater than 15 urn is concentrated in hydroclyclone I
and is ultimately brought to the extraction unit.
The actual extraction of the soil occurs in a specially deve-
loped extraction unit. Here a forward screw is subjected to
axial vibration, created by an electronically steered mechani-
cal energy component. In this extraction unit the particles
between 15(im and 5mm are cleaned using mechanical energy. The
soil particles are subjected to an energy density sufficient to
-------
break the bonds between the contaminant and soil particle so
that the contaminant is freed and dissolved in the extractant
(water) phase. The frequency, and amplitude of the energy can
be varied to optimize conditions for the individual types of
pollutant and soil being treated. The extractant, which is
generally water ( sometimes with small amounts of biologically
degradable detergent), is added in a 1:1 ratio to the soil
(1,3).
In the next step the material is washed in a countercurrent
sandtrap. Here the fine fraction < 130 urn and its extractant is
freed of any light material like wood, charcoal, or tar using a
sieve and is carried to the fine particle wash.
The larger material (sand/fine gravel fraction) is dewatered in
sieve III and sent to the fluidized bed sorter. Here the
contaminated "light" materials are floated out and removed from
the washing process by a screen.
The clean-up of the fine particle fraction occurs in a five
step countercurrent hydrocyclone. Following the hydrocyclones
the dewatering of this material stream together with the
sand/fine gravel fraction which was the throughput of the
fluidized bed sorter occurs in sieve IV. Cleaned material is
carried out by a conveyor belt.
-------
EXTRACTION UNIT AND HYDRAULICS
MULTI-STEP HYDROCYCLONE UNIT
-------
The over flow of hydrocyclone unit I contains the fine particle
fraction under 15(jm, in which the contaminant may in some cases
be concentrated. The separation of these solids from the
process water stream is accomplished by thickening. The
sedimented sludge is then dewatered using a filter band press.
The solid free process water is then cleaned together with the
contaminated ground water. The ground water/process water
treatment plant is a 4 step process. First any oil layers are
removed by an oil separator. This step is followed by flotation
where heavy metals and and emulsified hydrocarbons are separa-
ted by addition of the appropriate chemical agents.
The primary pollutant group in ground water at the Pintsch site
is volatile chlorinated hydrocarbons. To achieve recovery of
volatiles an air stripping step was added which has a recovery
efficienly of approximately 99 %. The stripped volatiles are
recovered from the air stream by activated carbon. The carbon
is steam regenerated on-site.
The last step of the water treatment process consists of a
series of filters including three sand filters followed by six
activated carbon filters. Input to the filters can be
individually controlled for selective loading depending upon
the contaminant stream in question. Here the remaining trace
amounts of contaminant are removed. The cleaned process water
is then recirculated through the soil cleaning process.
The cleaned groundwater is fed into the adjacent waterway, and
must therefore be monitored to ensure that it meets the
stringent standards for drinking water purity.
10
-------
4.0 CLEAN-UP RESULTS
To date the unit has treated soils from the following sites;
Site of former chemical production/refining companies,
- Waste oil refining facility
- Tar chemistry facility
- Paint fabrication facility
The primary pollutants found in these facilities were:
Hydrocarbons
Chlorinated Hydrocarbons
Aromatic and Polyaromatic Hydrocarbons
PCBs
Phenols , • ,
Sites of former gas works,
- HKW Moabit, Berlin
- Eisstadion Wilmersdorf, Berlin
Primary Pollutants for these sites were:
Hydrocarbons
Polyaromatic hydrocarbons
Phenol
Cyanides
11
-------
Sites contaminated by Mercury
- Soil from the former chemical plant at Marktredwitz
The clean-up results for these various sites can be seen in
figures 3,4,5.
5.0. FURTHER DEVELOPMENT
Using the outlined systematic solution it is possible to clean
soil with complex pollutants and problematic particulate
composition in such a way that the cleaned soil may be refilled
and reused.
We now have experience and results from a relatively broad
range of sites ( former chemical/physical facilities, gasworks,
and heavy metal contaminated soil). Nevertheless there is still
a need for further research and development to evaluate the
total picture for abandoned site clean-up; in light of the
non-homogeneous nature of these sites and the unique character
of each clean-up problem.
The actual potential for development lies in the treatment of
the fine and medium clay fraction (material with particle sizes
under 15um) and in improving the separation, clarification,
dewatering and transport aspects.
Based on the positive results of the FBI, PB2 and PB3 one
should be able to assume that at the end of the .current
development phase ( the final development of the PB4 ) one
should have a system capable of handling the majority of
abandoned sites in a systematic way using an environmentally
safe and economically feasible technology.
12
-------
6.0 LITERATURE
(1) Sonnen, H.D. Vorstellung einer Bodenwaschanlage, in;
Abfallwirtschaft 18 (1987), S. 117 - 126.
(2) Woltmann, M
Die Sanierung des Pintsch-Gelandes, in:
Fortbildungszentrum Gesundheit- und Umwelt-
schutz Berlin e.V. (Hrsg.): Sanierung kon-
taminierter Standorte 1985, Seminar des FGU
Berlin am 23./24.09.1985, Wiesbaden, 1985.
(3) Sonnen, H.D. Erfahrungen mit einer Bodenreinigungsanlage
Klingebiel, S. in Berlin, in: Wolf, K-; van der Brink, J.;
Colon, F.J.; (Hrsg.): Altlastensanierung'88.
Zweiter internationaler TNO/BMFT-Kongrefi
viber Altlastensanierung vom 11.04.
15.04.88, Kluver Academic Publishers, Dord-
recht, Boston, London, 1988, S. 899 - 905.
(4) Sonnen, H.D.
Groschel, W
Erfahrungen mit einer Bodenreininigungsan-
lage, in: Thome-Kozmiensky, K.J. (Hrsg.):
Altlasten 2, Fachtagung Praxis der Altla-
stensanierung vom 01.11. - 04.11.198.8, EF-
Verlag fvir Energie und Umwelttechnik GmbH,
Berlin, 1988, S. 771 - 780.
13
-------
xi—.contaminated soil/ r-
NI—'demolition waste L
£—i biologically degradable
N/ ^detergents _
s—i contaminated qroundf
x/—' water I
oxidizing
. agents
(~J reducing
agents
-------
rubbtsh.wood
construction rubtjel
X^^x 5 " (
eaned medium- /large
gravel fraction
(
cont sod J
||
material
preparation
II
nun
sieve I
II
dissolving of
s*t
II
large particle
dean uo
II
>0 nun
„..
>
II
tiydrocyclone I
I !/•
e
[V
S mm 4. f
extraction
unit
jj
countercurrent
sanatrap
"?0 ufli^ y
sieve «
a.
S
\\
fluidized bed
sorrer
=->
i i
retrculafion of
cleaned^
processwater
<>
(jZH fresti water
cleaned aroundwafer N
> >-=xC^' - - 1
1 1
activated caroon
filter
11
sandf liter
li-
stnqpng step
" -"" ^ |
waterway
\ activated carbon \ ( separated A
* Of cleaning . v \^ solvents J
ft •
flotation
II
V" '^, i ^" ^ •/
\ ^ ~/
flotation sludge
> '[ — j
f
F '
1 h
thickener
11
sludge
dewatermg
• separated oi
well
0
K \ ^— ^ ^ -./
_> \^ ^ /
* V^^~— -" 7
dewatered sludge
dsm screen
N \- __^--' —-/
^ V-^CTIX^
I contaminated
\/ tiqnt oarrcle
fine particle
clean up
fraction
II
hydrocyclone 11
i i
III i
^jr JJ, • 15 -130 urn
/ -'^.N^
cleaned sitN/scna/
fine gravel fraction
n— n
nJjrnLr,
U-TLJLTU
Made wasfier
. 11
_ sieve IV
Figure 2"- Block Diagram System HARBAUER
Groundwater- and Extraktive Soil Claening
on the former Pintsch Site ,Berlin
15
-------
10000
[ppml
[clean-up success]
100%
0,01
Aliphat. Chlor. ' Aromat.
Hydrocarbon*
PAH
PCS
I Input
I Output
i Clean-Op Success
Figure 3: Clean-Up Results for HARBAUER Soil Washing System
(Origin of soil: former waste oil refinery and former paint
manufacturing plant, respectively)
100O-T-
[clean-up success]
--80%
--60%
100%
--40%
--20%
Petroleum Ether
Extract
PAH
Phenol
Total Cyanide
! Input
i Output I" I Clean-Up Success
Figure 4: Clean-Up Results for HARBAUER Soil Washing System
(Origin of soil: former gasworks site)
16
-------
1000 T
[mg Hg/kg dry matter]
[clean-up success]
...._._ 100%
2 ' 3
Number of Sample
Average
I Input
i Output
I Clean-up Success
Figure 5: Clean-Up Results for HARBAUER Soil Washing System
(Origin of soil: former chemical plant site A)
[mg/kg dry matter]
1000 ^r
100 i
[clean-up
T 100%
Aliphat. Aromatic
Hydrocarbons
PAH
PCS Phenol
0%
HI Input E
• Output
I I Glean-up Success
w
Figure 6: Clean-Up Results for HARBAUER Soil Washing System
{Origin of soil: former chemical plant site B)
17
-------
FREEZE TECHNOLOGIES CORPORATION
REMEDIATION AND TREATMENT OF
RCRA HAZARDOUS WASTES
BY FREEZE CRYSTALLIZATION
ABSTRACT
Freeze crystallization is a general separations process used to remove pure
components from solutions by crystallizing the materials to be removed.
This process has been used for applications as diverse as organic chemical
refining and fruit juice concentration, and is especially suited for treating
hazardous wastes. This paper will illustrate how the process can be used
in site remediation activities, including treating contaminated soils, where
it can be used to recover valuable by-products from RCRA and other
industrial waste streams, and the basis for its utility in mixed (hazardous
and radioactive) wastes.
Freeze Technologies Corp. has built a mobile site remediation prototype
commercial plant to demonstrate the field remediation aspects of this
technology. The capacity of the unit is nominally 10 gpm of ice production
from a leachate or groundwater, at 90% water recovery. It is contained in
two modules that are transported on standard low-boy trailers, and
requires less than 1 week to set up.
Freeze crystallization has several advantages for remediation and waste
recovery applications. First, it is a very efficient volume reduction process,
producing a concentrate that has no additional chemicals added to it - if
disposal in a hazardous waste landfill, or incinerator destruction is
required this will reduce these costs substantially. When a large fraction
of the solvent (usually water), is removed from a waste, the remaining
impurities often begin to crystallize as well - they are often sufficiently
pure to have by-product value for resale. Processing costs with freezing
are generally low, ranging from $.03 to $.15 for 40 and 5 gpm plants,
respectively.
18
-------
FREEZE TECHNOLOGIES CORPORATION
FREEZE PRO CESS DESCRIPTION
The basic operation involved in freeze crystallization is the
production of crystals by removing heat from a solution. Crystals
produced in this manner invariably have very high purities. Once small,
uniform crystals have been produced, they must be washed to remove
adhering brine. The brine is recycled to the crystallizer, so that as much
solvent as desired can be recovered. The pure crystals are usually melted
in a heat-pump cycle, which further improves the energy efficiency of the
process.
When one or more of the solutes exceeds its solubility additional
crystal forms are produced, but they are formed separately from each
other and from the solvent crystals. Since in most waste applications the
solvent is water, and ice is always less dense than the solution and the
solutes usually more dense, it is easy to separate these crystals by gravity.
A freeze crystallization process is then composed of the following
components, as illustrated in the process flow diagram of Figure 1:
- a CRYSTALLIZER, where heat is removed to lower the
temperature of the material to the freezing temperature of
the solution (usually crystallizing the solvent first);
- a EUTECTIC SEPARATOR to segregate the crystals of solvent
and solute into different streams, so that each can be
recoverd in pure forms;
- CRYSTAL SEPARATOR/WASHERS that function to remove the
crystals from the mother liquor in which they are slurried,
and to wash adhering brine to very low levels so that the
recovered crystals have high purity;
- a HEAT-PUMP REFRIGERATION CYCLE to remove refrigerant
vapor from the crystallizer, and compress it so that it will
condense and give up its heat to melt the purified crystals;
- HEAT EXCHANGERS are used to recover heat from the cold
effluent streams, improving the heat efficiency of the
process.
- DECANTERS and STRIPPERS are required in some processes
to remove volatile materials and/or refrigerant from the
effluent streams before discharge;
- appurtenant UTILITIES, CONTROLS, ELECTRICAL SWITCH
GEAR, PUMPS AND PIPING are required to implement the
freeze process in a continuous, closed system.
19
-------
FREEZE TECHNOLOGIES CORPORATION
The design, operating characteristics, capabilities and limitations are
determined largely by the type of crystallizer that is used, of which there
are three basic choices:
1) INDIRECT CONTACT, using a scraped surface or similar heat
exchanger that will crystallize by removing heat through the
heat transfer surface.
2) TRIPLE POINT crystallizers use the solvent as the refrigerant
at its triple point (where solid, liquid, and vapor phases are
all in equilibrium). For instance, with aqueous systems, the
triple point occurs at less than 3 mm Hg. absolute pressure
at 30 degrees F, or below.
3) SECONDARY REFRIGERANT freezing uses an immiscible
refrigerant that is injected directly into the process fluid,
and evaporates at several hundred to several thousand
times the vapor pressure of the solvent.
The third option offers a number of advantages in treating hazardous
wastes, resulting in a less expensive process that is inherently capable of
producing higher quality effluents and effecting a greater reduction of the
final volume.
10 GPM REMEDIATION PROTOTYPE
The freeze crystallization process that has been accepted into the
EPA's Superfund Innovative Technologies Evaluation (SITE) Program is a
secondary refrigerant process. A prototype commercial remediation plant
has been designed and built for this program, and is illustrated in the
photo of Figure 2. The capacity of the plant is a nominal 10 gpm of ice
produced from an aqueous waste stream with a freezing point of 20o F.
The plant contains all of the components described in a typical freeze
crystallization proce'ss above, including the crystallizer, eutectic separator,
crystal separator/washer, heat exchangers, heat pump (open cycle screw
compressor), decanters and strippers, and ancilliary utility related
systems.
The plant is designed for ultimate transportability, using modular
design concepts developed by Applied Engineering Co., Orangeburg, SC, the
acknowledged leader in this field. The plant is contained in two modules
designed for transport on the back of low-boy trailers. Each measures
approximately 50' 1 x 13' w x 11.5' h. They are picked off upon arrival at
the site by a standard road crane, and one placed on top of the other. The
20
-------
FREEZE TECHNOLOGIES CORPORATION
three electrical and thirty flanged-pip ing interconnections take about a
day to complete.
The plant is totally self-contained except for electrical supply,
containing instrument air, cooling water, electrical distribution, and
electrical heating components. It has a distributed digital processor that is
programmed to operate without attendance, using on-line quality sensors
to evaluate process efficiencies and discern any operating problems,
recycling effluent for re-processing if warranted.
S.I.T.E. DEMONSTRATION PROGRAM - STRINGFELLOW NPL
The FTC Direct Contact Secondary Refrigerant Freeze Crystallization
Process was accepted by EPA's Office of Research & Development, SITE
Program, into the third phase of their program, when first proposed by
Freeze Technologies. A program with the State of California, Department of
Health Services, Alternative Technologies Office, was already in place, and
the selection of the Stringfellow NPL site had been agreed to with the state.
The goals of the SITE program are, among other things, to
demonstrate applicability of a technology on as general a waste stream as
can be found, and to develop the best economic and performance criteria
that can be projected from test results. Freeze Technologies proposed that
testing could be performed with either a .5 gpm portable pilot plant or
with a 10 gpm prototype remediation plant, designed specifically for
Superfund-type work. The groups involved in the demonstration of this
technology concurred unanimously that the larger plant is a better vehicle
for accomplishing the goals of the program.
The current schedule will have the equipment at the Stringfellow
NPL site in late July and installed and ready to operate in early August.
The SITE tests and sampling program will occur over a two to three week
schedule in August, with a public visitors day on Sunday, August 16, 1989.
Further testing for reliability and longer-term performance confirmation
will continue into September, and the plant will be removed from the site
by the end of September, after appropriate decontamination.
This demonstration program is designed to have minimal impact on
the host site, a condition that is paramount in the planning at the EPA
Regional level and with local communities. In our case, we will fit the
freeze equipment in as a 'black box' in the pipeline that carries wastes
from the NPL site to the on-site pretreatment plant. We will intercept the
wastes between the collection wells and the pretreatment plant, process
21
-------
FREEZE TECHNOLOGIES CORPORATION
the wastes, then recombine them for transfer to the pretreatment plant.
The residence time in the freeze crystallization plant, including storage
before and after the actual freeze processing equipment, will be 72 hours.
Since the freeze process operates in closed vessels, with recycle of
refrigerants and wastes at temperatures well below ambient, there are no
emissions from the process. An option exists at Stringfellow for treating
the concentrated effluent for metals recovery, for organic by-product
extraction, or by solidifying the concentrate for subsequent landfilling.
The by-product recovery is an attractive option as it eliminates a large
volume that is now landfilled.
WASTE TREATMENT WITH BY-PRODUCT RECLAMATION
Remediation offers some opportunities for by-product recovery, but
more frequently these applications occur in on-going RCRA generation
facilities. Here we'll discuss, first, the generic conditions that favor freeze
crystallization treatment, and how other unit processes can be used with
freezing to offer a complete remediation or by-product recovery process
train. Then we'll cover a few recent applications we've reviewed and
tested using freeze crystallization.
GENERIC APPLICATION CONSIDERATIONS -
Freeze crystallization works by making pure crystals from water, or
other components, in a waste solution. The waste must be in liquid form,
so contaminated soils are treated by first washing the impurity out of the
soil into awash solvent, usually water. In the case of aqueous based
wastes the crystal that is produced is ice, and all impurities are excluded
and remain in the concentrated liquid portion. The process is effective in
removing water from these wastes, reducing the volume that must be
dealt with.
Since crystallization excludes all impurities from the ice, all
impurites are equally removed - the freeze process is capable of treating
wastes with heavy metals, all types of dissolved organics, and radioactive
materials. And all of the impurities are reduced in concentration
ia the effluent by a factor of about 10.000
Freeze crystallization is not the answer for all applications, and the
chart in Figure 3 gives an indication of the conditions that favor it's use
over alternative treatment technologies. The chart demonstrates that
22
-------
FREEZE TECHNOLOGIES CORPORATION
freezing becomes more economically and technically competitive as the
waste becomes more concentrated and more complex. For instance, wastes
with heavy metals require concentrations of 1,000 to 10,0000 mg/1 to be
economically recoverable with freezing. Aqueous streams require organic
concentrations of 3 to 7 wt-% before it is more economical to treat with
freeze crystallization than by conventional means. Yet, when the waste
contains both organics and heavy metals freeze becomes more economical
than multi-process treatment trains using conventional technologies at
between .5 and 1.5 wt-% total contaminants.
CASE STUDY 1 - Stringfellow Leachates
Treatability studies were performed on Stringfellow leachates with
the approximate composition shown in Table 1. These tests confirm that
freezing will recover over 90% of the water, concentrating the impurities
into a minimal volume for disposal and/or recovery operations. In the
treatability tests (methods and equipment described below), crystal clear
melt was produced with a reduction in conductivity from 39 mS/cm to less
than .02, approaching a ratio of 1000:1. Table 1 also shows anticipated
effluent qualities at 90% recovery of water from the leachate.
Currently the leachate from Stringfellow is treated in a conventional
system consisting of hydroxide precipitation of heavy metals and activated
carbon adsorption of the organics. The sludge is disposed of in a hazardous
waste landfill, and the carbon is returned for regeneration. Treatment
costs with this scheme are outlined in Table 2. Throughput at Stringfellow
has been slowly declining, with current production from the interception
wells at about 225,000 gallons per month. The annual cost for only the
variable items that can be displaced by treatment with an alternative
technology is about $1.2 million.
With freeze crystallization, the operating costs in this size plant are
about $.09 per gallon. Costs as a function of treatment plant size are
summarized in Table 4. The variable costs using freeze crystallization for
treating the Stringfellow strong leachate would total less than $.25 million
per year. Depending on amortization rates for the equipment, the annual
charge would range from $.2 to $.5 million. Freeze crystallization will also
leave a concentrated waste stream of about 1300 cubic yards. Incineration
of this might cost as much as $300,000 per year.
CASE STUDY 2 - Mixed Industrial Wastes
Industrial facilities often have wastes that are collected from a
variety of places and treated at the end of the pipe. In metal fabrication
-------
FREEZE TECHNOLOGIES CORPORATION
shops with electroplating, metal cleaning, pickling, stamping, machining,
and other operations, the waste coming out the end of the pipe might
resemble the analysis shown in Table 3. A typical treatment train might
include an oil-water separator, neutralization, hydroxide precipitation, and
discharge to a POTW for organics removal. However, municipalities are
looking much harder at toxic organic discharges, and often some of the
cleaners used in the plant are methylene chloride or other toxic chloro-
solvents.
Freeze Technologies has performed laboratory testing on a waste
such as this.,collected after an API separator. Over 90% of the water was
converted to ice. Alkalinity and sulfate salts were precipitated with heavy
metals, hardness cations, and probably to some degree with sodium ions.
After about 75% water removal a second organic phase began to form, and
a significant portion of the influent organics partitioned into this phase
that was composed of about 75% organics and 25% water. The water phase
contained less than 10% organics, but had most of the dissolved salts.
The organic phase that forms this way will have a high heating value
so it can be incinerated directly, or perhaps recovered for its fuel value.
The tests in the batch treatability lab showed that this organic layer had a
specific gravity much less than ice, and it collected in the top of the ice
drain column. This will allow its recovery from a freeze crystallization
process by making minor changes in the wash column design.
TREATABILITY TESTING
Within the context of the CERCLA/SARA legislation, and the programs
and regulations that the Agency has adopted to implement it's provisions,
it seems appropriate to address treatability testing and the impact on new
technology utilization in Superfund and other remedial actions. First, given
the complex waste matrices normally encountered in the vast majority of
remediation sites, Superfund or RCRA remedial actions, treatability studies
should always be conducted.
Most Remediation Project Managers realize that soil treatment
technologies are sufficiently new, and the wastes so diverse, that
treatment feasibility testing is needed, especially with new technologies.
But it is amazing how many of these managers labor under the
misconception that a wastewater analysis is sufficient to define the optimal
treatment train for cleaning up liquid waste streaims. At a major tertiary
treatment sewage reuse application in the early 1970's, where the effluent
24
-------
FREEZE TECHNOLOGIES CORPORATION
was thoroughly characterized, the reuse client wanted to demonstrate the
treatment train and reuse application. A 2 month, $150,000
demonstration program turned into a year effort and $1 million
expenditure, and resulted in several major equipment innovations and
rather drastic modifications to the conventional treatment design criteria.
From our experience with 8 years of development of the freeze
crystallization process, and another 10 years of development of
environmental process technology, there are a number of guidelines we
would propose for treatability testing in EPA Superfund and other .
industrial waste programs:
1) Treatability studies should always be done by the technology
developer, for several reasons. First, the developer is the one who
will give process guarantees, and this is one of the necessary steps
in giving warranties. Second, the best knowledge of how actual
waste characteristics impact process performance is resident in
the developers' staff, and expansion of this database is a needed
and beneficial part of the Agency's function in the Cleanup of
America. And third, the expenditure for treatability studies
means nothing if they don't result in better, more efficient, more
economical treatment at contaminated sites, and this won't be
done without the involvement of the developers.
2) Treatability studies should be incorporated into the RI/FS
activities, and replace the paper studies that are currently the
basis for selecting between alternative treatment technologies.
There are two reasons for getting the developers involved at this
stage: their input into technical solutions, and more accurate cost
information for the Agency to select between alternative
treatment options at a given site. The REMS and ARCS contractors
are at best 6 months behind on individual technology
development programs, and more frequently 18 to 30 months.
The RI/FS process is away for EPA to effectively help the
technology transfer from the developers to their contractors.
3) The EPA and its contractors should not attempt to define the
treatment technologies at such sites, but rather should conduct the
procurement activities in such away as to elicit complete
strategies from technology developers. Most remediation sites
have the need for a variety of treatment technologies. The wastes
occur in several media, and the compositon at most sites varies
significantly with the location and goegraphy. There is a great
deal of networking going on between technology developers, and
25
-------
FREEZE TECHNOLOGIES CORPORATION
the most efficient method for defining treatment trains is to
define a clean-up standard and let the individual developers
provide the train of technologies for providing that.
The technology developer should have an institutionalized place in
the remediation activities of Superfund and private clean-ups. The best
process technology is resident in this community, and this resource should
be used for its expertise in a way that ensures competition in procurement
and full assessment of alternatives. One clear way to provide this, if not
the only way, is to procure clean-up strategies on a competitive basis, by
writing performance specifications based on ARAR's. Very many ROD's are
issued with exceptions to ARAR's without a solid technical basis, which is
the only basis allowed by statute. If all bidders of new technologies take
exceptions to the ARAR's, then there is a technical basis which is
defensible.
Freeze Technologies performs treatability studies in the apparatus
shown in the photo of Figure 4. A simplified flow diagram for the unit is
shown in Figure 5. The unit operates in a batch mode. Feed is drawn
under vacuum into the crystallizer, and recirculation between there and
the drain column is started with the slurry pump. Refrigerant is pumped
into the crystallizer, where it bois.as it is heated by contact with the waste.
The refrigerant vapor passes into a refrigerated condenser, condenses, and
is pumped back to the freezer. When the drain column is packed full of ice
the process is stopped and the ice is washed and melted. The concentrate
drains out of the ice back to the crystallizer, and is saved for additional
testing. Since only about 50% of the water in any batch can be converted
to ice, simulation of 90% recovery requires 4 stages of concentration.
Twenty gallons of sample is required to allow 12 of these staged tests.
These tests accomplish the following:
- interactions between refrigerant and waste are observed, such
as foaming or emulsification, that must be chemically stopped.
- phase equilibria are confirmed, showing the operating
temperatures that will be required.
- the occurance of eutectic crystallization are observed.
- physical properties of the waste at the crystallizing conditions
are determined.
The impact of this information is that more detailed preliminary process
designs are possible, which in turn allows more accurate cost projections.
Problems that would be seen in the field often show up at this stage of
testing, and design adaptations made in the batch lab are the first stage in
modifying the full scale equipment for use with new applications.
26
-------
FREEZE TECHNOLOGIES CORPORATION
TABLE 1
STRINGFELLOW LEACHATE ANALYSES
Concentration, mg/1
Component Feed Melt
pH 3.5 6.5
Conductivity, mS/cm 39 .02
D issolved Organic Carbon 1125 < 1.
p-CBSA 1670 <1.
Volatile halocarbons 9.7 < .010
Volatile ketones H. < .010
Na 820 .
-------
FREEZE TECHNOLOGIES CORPORATION
1
TABLE 3
INTEGRATED MANUFACTURING END-OF-PIPE EFFLUENT
Waste Component
Total Dissolved Solids
Dissolved Organics
PH
Alkalinity
Sulfates
Chlorides
Heavy Metals
mg/1
Concentration
12,500
8,000
10.
5,000
4,700
1,500
3,000
TABLED
FREEZE CRYSTALLIZATION COST SUMMARY
COST COMPONENT
Amortization, 5 year SLD
Labor
Electricity
Supplies, Chem, Etc
Maintenance
TOTALS
COST AT PLANT SIZE, $/GAL
$ .08
.04
.008
.010
*MZ
.145
$ .05
.02
.0075
.0075
jmi
.09
$ .015
.005
.006
.005
.004
.035
28
-------
FREEZE TECHNOLOGIES CORPORATION
',• ; V ' " ' I ^^ '' nuf
' • . ,y
FIGURE 1
FREEZE PROCESS PROCESS FLOW SCHEMATIC
UJ
29
-------
FREEZE TECHNOLOGIES CORPORATION
FIGURE 2
10 GPM REMEDIATION PROTOTYPE PLANT
FIGURE 1
TECHNOLOGY COMPARISON CHART
TYPE OF CONTAMINANT
1
N
RC
E
A
5
E
D
C
0
N
C
E
N
T
R
A
i
KM
VOLATILE HEAVY
ORGANICS ORGANICS
C/SITC HEAVY
bALlb METALS
STRIPPING
CARBON ADSORPTION
BIOLOGICAL & CHEMICAL
OXIDATION
R
ION EXCHANGE
ELECTRODIALYSIS
EVERSE OSMOSIS
EVAPORATION
FREEZE CRYSTALLIZATION
30
-------
FREEZE TECHNOLOGIES CORPORATION
FIGURE 4
BATCH TREATABILITY TEST LABORATORY
31
-------
FREEZE TECHNOLOGIES CORPORATION
FIGURE
BATCH TREATABILITY STAND PROCESS SCHEMATIC
«"•
32
-------
Purification by froth flotation
ir. C. Mosmans
The purpose of this paper is to describe froth
flotation and the guiding principles in the field of
purification of contaminated soil and waste.
Purification by froth flotation.
The mosmans method has been used for the
decontamination of soil and waste in the Netherlands
since 1983. This method is effected by well chosen
mineral seperation techniques which together result in
•a complete purification process. The most important,
but also the most complex part of the mosmans method
is the froth flotation technique.
Mosmans Mineraaltechniek BV is an independent
organisation with a laboratory for research and
development of mineral seperation methods. The
laboratory spent several years to develop flotation
processes for the purification of polluted soils,
waste and waste effluents. It is fully equipped to
carry out tests on batches from grammes to many tons
by such processes as sampling, grinding, screening,
classification, gravity concentration, dense media
separation and several flotation methods.
As an extension of the laboratory efforts and to
obtain real field experience and well trained
operators a small processing plant has been in
operation in the past six years. After several years
of laboratory testwork the plant started in 1983 with
the froth flotation of 10,000 tons of sandy soil.
After that about 3,000 tons of copper containing muddy
soil and 10,000 tons of earthy soil with lead.
33
-------
A relatively difficult project was clayey soil from
gasworks with complex cyanide and pea. For the very
fine fraction (the slimes) special flotation
techniques have been used.
This year a conspicious soil cleaning operation will
be carried out in Zwitserland. This soil is
contaminated with pesticides, herbicides and mercury
during a fire in 1986.
Principles of froth flotation.
To achieve a separation between the
contaminatives and soil in a soil-water mixture, the
surfaces of the particles have to be adequately
manipulated. In such a way that the former will be
hydrophobic and the latter hydrophybic. The
manipulation is not related to changing the chemical
structure of the particles, but to modify the surfaces
by selective adsorption. The hydrophobic particles
glue themselves to air bubbles produced in the soil-
water mixture.
Flotation processes are determined by surface
phenomina. The available surface depends on the
grainsize. For example a cubic grain of mesh size of 1
cm covers an area of 6 crn2. By splitting this cube the
total area increases.
Mesh size
10,000 Urn
1,000 (Om
100 Jim
10 |im
1 jim
0.1 (am
total ar^a
6 cm2
60 cm2
600 cm2
6,000 cm2
60,000 cm2
600,000 cm2
remarks
grains are too heavy
for flotation
normal froth
flotation
special flotation
technology
It is obvious that clay or silt has much more
area/gramme than a sandy soil.
34
-------
roosmans
Potential
A particle in water obtains an electric surface
charge due to partly ionisation, adsorption,
orientation of dipoles or crystal lattice defects.
Around the charged particle a diffuse layer or a cloud
of counter ions with an opposite charge is formed,
reducing the potential to zero. This phenomenon is
called an electric double layer.
Figure I presents .this classic Stern - Grahame
model of the electric double layer. In the figure the
S plane represents the closest distance of approach of
the counter ions.
Stern layer - counter ions
\ayer - reduction of- potential te>
layey
t?is6ance -^
When the particles are forced to move in
relation to the liquid, the slipping plane in the
diffuse layer shows a potential, the Zeta potential.
This electrokinetic phenomenon is very practical to
study the manipulation of the surfaces. Two methods
are in use, electrophoresis and streaming potential
measurements. Figure 2 and 3 present these methods.
n»ic*-0sc0pa - w eaSHrement- of velocity
of jnoaU particles in suspension
of- water
35
-------
mosmans
There are particular ions that are free to pass
the electrical double layer. These ions are called
potential determining ions. By a titration procedure
the effect of these ions can be determinded. The
condition which the ions compensate the surface charge
is called zero point of charge or iso electric point.
For example, for a particle CaCO3 the potential
determining ions are Ca, CO3 and the pH, because of
the equilibria with HCO3 and 003.
Material
2PC
Controlled bv the concentration
CaCO3
CaF2
pH 9,2
pCa 3,5
pH 2,5
H
Ca
H
Coulowibfc. repulsion
Counter ions which adsorb only by means of
electrostatic attraction are called indifferent
electrolytes. They "compress" the diffuse layer.
Examples are A1-, Fe- and Ca-ions. These are the
anorganic flocculants in water purification.
The third type of important ions exhibit surface
activity and electrostatic attraction. They are called
surface active counter ions. To these belong the wide
range of flotation collectors,
The effects of an electrical double layer are-
illustrated by coagulation and a special flotation
method for very fine particles. For the coagulation of
two particles there are two forces, the coulombic
repulsion force and the London - v.d. Waals attraction
force, resulting in a barrier preventing coagulation.
Addition of Fe-ions compress the diffuse layer, the
barrier diminishes, and coagulation may occur, see
figure 4.
' -"—-•-- occurs
36
-------
water1
•fresh water
ity of surface active, re^tvit ->
Fine contaminatives, say less than 5 \im are
generally not floatable. But by the addition of coarse
"carrier", particles and under conditions in which the
unwanted fines adhere to the carrier, a purification
by normal froth flotation of the carriers is possible,
see figure 5.
In froth flotation adhesion is necessary between
the particles and the air bubbles rising through the
pulp. There are three interfaces, particle - water,
particle - air and water - air.
In fact, the thermodynamics of surfaces is valid
for all the three interfaces. But for the sake of.
simplicity the particle - water interface is conform
to the electrial double layer concept. The water - air
interface is conform to the alteration of surface
energy.
If a surface active or heteropolar reagent is added to
water the surface energy decreases, see figure 6.
This is a result of the heteropolar nature of
the compound. These molecules adsorp preferentially at
the water-air interface. They are arranged with the
hydrophobic tails into the air and the hydrophylic
heads immersed into the water, see figure 7.
Air
Water
negatively charged
® counter ions
These surface active reagents are called frothers.
A few examples;
CH3-CH-CH2-CH-CH3 013-(O-CsHg) n'-OH
CH3 OH , " '' ?'"%. -'•
methyl isobutyl carbinol polypropyiene^-glycol
37 •.-*:..•--.•
-------
mosfnans
water
particle A
weteabte"
water
air
L
particle &
water
In the soil-water or waste-water mixture bubble
contact must occur with the contaminatives. Figure 8
presents a cross section with the surface energy as
surface tension (= d).
Investigations
The investigator has to manage many flotation
variables. In addition he' has to cope with the
representativity of samples and with hardly to be
determinated contaminatives. A consequent approach
involves the following stages:
- Sampling of contaminated soil or waste.
The samples have to be representative, not only in
respect to chemical but also in respect to
mineralogical composition. If major differences exist
between parts of the contaminated area, separate
samples shall be obtained.
Sampling methods and sampling statistics as used in
the mining industry are very useful.
- Determination of soil and contaminatives.
Chemical analysis provide only part of the information
needed. Mineralogical analysis and grain size
distribution are necessary for developing a flotation
process. Techniques of mineralogical analysis are x-
ray analysis and the examination by a petrographic
microscope.
Micro scale tests.
Bubble pick up technique,. Hallimoand tube and a\r
release tube are employed in flotation research of
contaminated soils. Figure 9, 10 and 11, present these
simple but effective and valuable methods.
Batch testing
To investigate the flotation variables many tests with
quantities of about half a kilogramme have to be
carried out. Significant variables that are subject to
38
-------
rnosmans;
bubble pick-tip
-flotation
O
control are type of collector, frother, activator and
depressor, pH, conditioning time, flotation time, pulp
density and temperature.
In batch flotation a given mixture is fysico-
chemically prepared, aerated, whereby it becomes
separated into a froth and a residue.
The particles diminished into the froth
according to:
C - Co e~St t = time
Co = concentrations at t = 0
S = constante (slope)
Figure 12 presents this idealized equation with
data of sand, clay, water and complex cyanide. Each
being expressed as a percentage of the initial
quantity.
The research focuses on the alteration of slopes, to
make them steeper for contaminatives and flatter for
clay, silt, sand and (water).
- Laboratory testing may be followed by pilot
plant tests to provide continuous operating data for
design, and to ensure that laboratory tests are
reproducible on plant scale. Besides the possible
build up of bicarbonates, sulphates, calcium and
magnesium in the recirculating water should be
studied. Flotation cells are arranged subsequently.
Each cell separate only a part (the recovery) of the
contaminatives.
For example, 50% recovery in each cell results for 4
cells in: 50% + 25% + 12,5% + 6,25% = 93.75% recovery.
- Plant practice in soil and waste flotation.
Each separation flow sheet represents the best
solution at given conditions, such as volume of
contaminated soil or waste, the desired final grade,
da-ta from-laboratory work and" available equipment.
39
-------
mosmans
•fig 12.
Note
The above mentioned, mineral processing
techniques have been adopted from the mining
Industrie. Over the past 75 years froth flotation has
become the most important separation method in the
mining Industrie. Almost the entire world supply of
copper, lead, zinc, nickel and many others is first
collected in the froth of the flotation machine. But
also nonmetallics are caught in the froth such as huge
tonnages of phosphates, coal, fluorite, feldspar.
soluble potassium chloride etc.
40
-------
Reverse Osmosis: On-Site Treatability Study
Of
Landfill Leachate At The PAS Site In Oswego, NY
by
Charles Goulet
Seprotech Systems Incorporated
2378 Holly Lane
Ottawa, Ontario
K1V 7P1
and
Harry Whittaker
Environmental Emergencies Technology Division
Environment Canada
Ottawa, Ontario
Kl A OH3
and
Robert Evangelista
RoyF. Weston, Inc.
Response Engineering and Analytical Contract
GSA Depot, Building 209 Annex
Edison, NJ 08837
and
Tom Kady
Environmental Response Team
U.S. Environmental Protection Agency
GSA Depot, Building 18
Edison, NJ 08837
presented at the
Forum on Innovative Hazardous Waste
Technologies: Domestic and International
in Atlanta, GA
19 June, 1989
41
-------
Reverse Osmosis: On-Site Treatability Study
Of
Landfill Leachate At The PAS Site In Oswego, NY
1.0 Introduction
The Pollution Abatement Services (PAS) site in Oswego, New York, was
remediated in the mid-1980's when drums of toxic wastes were removed from the
site, and the contaminated soil was contained' by means of traditional methods: a
slurry wall along its perimeter, a leachate collection system, and a clay and high
density polyethylene cap. The continuous monitoring and maintenance of this
Superfund site has consisted of the collection and off-site treatment of approximately
65,000 gallons of leachate monthly. The leachate management methods has cost
government agencies money, manpower and time. To reduce these expenditures,
Region H of the United States Environmental Protection Agency (US EPA) requested
the engineering assistance of the US EPA Environmental Response Team (ERT) to
explore the feasibility of on-site treatment technologies.
In agreement with the 1986 Superfund Amendments Reauthorization Act (SARA),
alternative and innovative technologies were preferentially addressed.
Environmental engineers from the US EPA, Environment Canada, and the
Response Engineering and Analytical Contractor (REAC) analyzed previous
successful bench-scale work and selected reverse osmosis (RO) and enhanced
ultraviolet oxidation, also called UV photolysis/ozonation/hydr6gen peroxide, for
a field pilot-scale study.
This paper concerns the application of reverse osmosis for on-site treatment of
hazardous wastes. It reviews the basic principles that govern the process and reports
on the treatability study that was conducted at the PAS site in the summer of 1988.
Since four different brands of RO membranes were investigated, the evaluation of
their^ performance focuses on the quality of their respective permeate. The purpose
of this paper is to provide the preliminary results of this study.
42
-------
2.0 Reverse Osmosis: The Treatment Technology
2.1 Process Description
Reverse osmosis (RO) is a membrane technology which can serve a dual-purpose in
hazardous waste treatment. It is a means to reduce wastewater volumes, and in the
process, the membrane generally retains most of the contamination within one of
the two effluents, namely the concentrate. The permeate .stream is aimed to be
discharged in the environment, whereas the concentrate requires further treatment.
RO is accomplished by pumping an aqueous dilute solution at a high pressure over a
semi-permeable membrane. As the solution sweeps the surface of the membrane,
the solvent, water, is preferentially attracted. Its transport across the polymeric
layer is driven by the difference between the osmotic pressure of the solution at that
location on the membrane and the hydrostatic pressure of the solution.
The chemical substances present in a given feed solution differ in their affinity for
the membrane material. The interaction between the membrane, the solute
components, and the solvent governs the performance of any such system.
Different schools of thought have explained the mechanism by which solutes are
transported across the membrane. The most prevalent theories are : the preferential
sorption-capillary flow mechanism (Sourirajan and Matsuura, 1985) and the
solution-diffusion transport mechanism (Lonsdale et al. 1965).
Once the initial separation has been performed, the solution retained at the,
membrane surface is subjected to further separation as it flows along the polymeric
surface. Thus, there is not only a step-wise change in solute concentration between
the two sides of the membrane but also a gradual increase from the feed influent end
to the concentrate effluent end on the same side of the membrane.
2.2 Applications
Reverse osmosis has been used extensively for the past twenty-five years to remove
inorganic compounds for the production of potable water from brackish water and
seawater. Economic and environmental considerations have promoted its use for
concentrating electroplating rinse waters since the '70s, and its use has been
explored for other hazardous waste treatment since then. Beginning in 1984,
Environment Canada has been demonstrating reverse osmosis in the field for
treating dilute organic solutions such as landfill leachate, industrial wastewater,
and contaminated aqueous solutions generated by chemical spills. Some of the
solutions successfully investigated in Canada include:
43
-------
0 spilled wood protection solutions containing tetra and
pentachlorophenols (PCPs);
0 leachate contaminated by volatile organic compounds;
0 leachate contaminated by polychlorinated biphenyls (PCBs);
0 potato wash water;
0 fish blood water;
0 pesticide container wash water;
0 industrial wastewaters containing monoethylene glycol;
0 surface runoff contaminated by benzene, toluene, and xylenes (BTX);
0 solvent contaminated wastewater from an aircraft maintenance
facility.
2.3 Pollutant Concentration Considerations
The influent pollutant concentration for which a reverse osmosis system will
economically produce a permeate of sufficient quality for discharge in a sewer system
or mto surface waters varies from one chemical compound to another, and from
one application to another. For example, a treatability study of an industrial waste
effluent revealed that a reverse osmosis membrane would not withstand long-term
exposure to a.1% phenol solution. However, selection of another membrane and a
change in the chemical nature of the phenol allows the RO system to increase the
phenol concentration up to 8% without either affecting the physico-chemical or the
structural integrity of the RO membrane. As the permeate from the first stage is, at
1500 ppm phenol concentration, still too highly contaminated, a second stage has
been added and the permeate meets the effluent discharge criteria set by the
concerned jurisdiction. The phenol recovered, by the two-stage RO system is
expected to pay for the pollution control system within twelve months.
In this section, a review will be made of the two characteristics that play the most
significant roles in defining the ranges of pollutant concentrations in which RO will
be effective. These are the osmostic pressure and the chemical compatibility between
the pollutants and the membrane materials.
23.1 Osmotic Pressure
Osmosis is a natural phenomenon. The term defines the diffusion of a solvent
through a semi-permeable membrane which separates two solutions having
different molar concentrations of solute. The semi-permeability of the membrane
refers to this property to allow the solvent through but not the solute. An
equilibrium is reached when the solvent flow stops. At this point, the hydrostatic
pressure difference between the two liquid columns represents the osmotic pressure
difference between the two solutions. To generate a net water transport through the
membrane, the osmotic pressure difference must be overcome (Figure 1).
44
-------
The osmotic pressure of an aqueous solution is caused by the presence of chemical
species other than water. The type of compound and its concentration, the pH, and
the temperature of the solution have an influence on the osmotic pressure, and
thus, on the driving pressure necessary to induce reverse osmosis. The wastewater
should exert an osmotic pressure no greater than the membrane's operating
pressure limit to allow for water permeation.
2.3.2 Compatibility With Membrane Materials
The choice of a membrane is governed to a great extent by its chemical resistance.
The membrane surface material, the backing polymer, the adhesive agents, the
seals, and all related hardware must sustain the chemical abuse to which the
pollutants might subject them. Reverse osmosis systems are mostly sensitive to
long-term exposure to oxidants, and to low molecular weight halogenated organic
solvents. Also, field experience reveals that some RO membranes tend to swell
from long-term contact with BTX.
Membrane research and development continuously evolves as a result of customer
demands, the introduction of new synthetic materials, and the quest for better
performing membranes. Chemical resistance is constantly increased, in particular
that to free chlorine, and the allowable feed pH range has been widened.
Finally, compatibility includes the notion that any pollutant will be rejected. It is
undesirable that a membrane preferentially allows the permeation of contaminants.
2.4 Performance
In hazardous waste treatment, the ideal reverse osmosis membrane is one which:
retains all pollutants; possesses a very high water permeation rate; chemically
resists virtually all compounds; and needs very little maintenance (washing or
replacement). However, RO membranes do not separate 100% of the contaminants
and cannot recover 100% of the water from the contaminated solution. Thus,
performance assessment of RO membranes is primarily based on the qualitative
(percent rejection) and quantitative (water permeation) aspects of the separation.
2.4.1 Qualitative Aspects
Qualitative membrane performance is assessed by measuring the concentration of
given solutes in the feed solution and in the permeate stream. From this data, the
percent rejection is calculated by means of the following formula:
R, = (l - Cip/Cif)-x'100%
45
-------
where: Rj -is the rejection of solute "i" expressed as a percentage;
Cip represents the concentration of solute "i" in the permeate;
Cjj represents the concentration of solute "i" in the feed solution.
Rejection is considered from the perspective that a given solute is prevented from
entering the permeate stream by the semipermeable membrane. In this sense, one
should not confuse it with the removal of a given solute from the permeate.
Typically, a membrane retained for its rejection capability will cause a change in the
pollutant concentration by one order of magnitude or more (in other words, a 90%
rejection or more). As a result, the concentration of each pollutant in the feed
greatly affects the number of stages needed to meet the effluent discharge criteria for
the permeate.
2.4.2 Quantitative Aspects
RO systems are amenable to modular design. This characteristic allows for the
manufacturing of systems to serve specific purposes, and yet, the productivity of
each system can be increased by adding membrane modules, or replacing
membranes within each module by better, and newly developed membranes, or by
plumbing a second system to the first one.
All designs are based on the permeation rate of membranes. As each aqueous
solution interacts differently with a given membrane, a systematic approach to
determine the best membrane for each application must address the following
aspects: the osmotic pressure expected and the necessity of special cleaning methods.
In addition, the chemical compatibility of all materials should be examined.
Water recovery ratio is another quantitative criterion which is particularly
important for wastewater volume reduction applications. It is defined as the portion
of the feed solution which is recovered as permeate, and it conveniently
corresponds to the volume reduction achieved, the ratio of the permeate flowrate
to the sum of the concentrate and the permeate flowrates represents this quantity:
A = Qp / (Qp +QC)
where: A is the water recovery ratio (expressed in percent or as a fraction);
Qp is the permeate flowrate;
QC is the concentrate flowrate (units consistent to those used for Q ).
46
-------
Water recovery is limited by the rise in osmotic pressure that accompanies a volume
reduction, and the precipitation of salts at the surface of the membrane. Therefore,
a compromise must be reached between how much permeate should be recovered
and the membrane maintenance needed to keep the membrane permeating.
2.4.3 Mass Rejection Rate
The mass rejection rate is a practical means to compare membranes. Its concept
embraces the needs for high permeation rates, and for high contaminant rejections.
Its calculation is performed by multiplying the permeate flowrate by the
concentration difference existing between the feedwater and the permeate for a
given contaminant. It represents the mass of a given contaminant which is
prevented from entering the permeate or reciprocally that which is retained within
the concentrate. A mass balance performed around a membrane system yields:
where:
m, = (cif-qp).Qp =
nij is the mass rejection rate of solute "i";
Cic is the concentration of solute "i" in the concentrate;
all other quantities, Cif/ Cj_ Q , and Qc were defined above.
2.5 Limitations To Reverse Osmosis
t»
The following aspects may cause interferences to the RO membrane's performance:
the feedwater source, the chemical compatibility of the membrane with respect to
the feedwater composition, the solution osmotic pressure, fouling factors and
scaling agents. Silt or colloids can become entrapped and significantly reduce
productivity and rejection. The fouling potential is especially high for levels of iron
higher than 0.1 ppm as the divalent form (Fe"1"*") will likely form colloidal iron (Fe+++)
in the presence of oxygen. Those salts which are most susceptible to precipitate as a
scale on the membrane surface are calcium carbonate (limestone), calcium sulfate
(gypsum), and silica. Hence, pretreatment should be addressed carefully as it will
impact on the longevity of the RO membranes.
Various protective, monitoring and control components must be integrated to
allow for the safe and reliable field operation of any reverse osmosis system. On the
upstream end of the membranes, also called the feed end, filters act as a preventive
measure by removing suspended solids down to the micronic size range. Generally
it is recommended that particles having a size greater than one fifth of the smallest
dimension of the feed conveying channel be removed. As a rule of thumb, 5 um
filters represents a suitable size for reverse osmosis (Lombard, 1986). In addition,
pressure and temperature gauges and switches are normally incorporated. These
ensure that the membranes will be used under normal operating conditions.
47
-------
Currently, automation has yet to be developed to allow for feedback regulation of
key operation parameters such as pressure, temperature, recirculation rate, and
conductivity in the permeate.
3.0 Methodology
3.1 Leachate Characterization
Leachate characterization was carried out in December 1987. Samples collected at the
PAS site in Oswego, NY , were analyzed by REAC. The analyses focused on: the
total priority pollutants; total and suspended solids; titration curves; pH; TOC;
BOD; COD; and flashpoint. The TOC level was 870 mg/1. Furthermore, the volatile
organic compounds with the highest concentrations were: methylene chloride
(25900 ug/1); trans-l,2-dichloroethene (13300 ug/1); xylenes (10000 ug/1 total);
toluene (6300 ug/1); and ethylbenzene (4100 ug/1). Also, analyses showed phenol
(935 ug/1) and 2,4-dimethylphenol (435 ug/1) as the primary base/neutral/acid
extractable compounds. No organochlorinated pesticide, nor polychlorinated
biphenyl (PCB) was detected. Among all investigated metals, nickel (2150 ug/1),
arsenic (35.3 ug/1), chromium (22.8 ug/1), and cadmium (16.0 ug/1) displayed the
highest concentrations.
3.2 Bench Scale Treatability Study
As a result of the leachate characterization, phase I engineering studies were
undertaken. Another sampling effort was performed at the PAS site on
23 February, 1988. It involved the complete filling of two 55-gallon lined drums (to
eliminate headspace) and overpacking in 85-gallon drums. Subsequently,
bench-scale testing was conducted at the River Road Environmental Technology
Centre in Ottawa, Ontario, CANADA. Two leachate treatment schemes were
investigated from' 24 to 27 February, 1988. They were: powdered activated carbon
adsorption followed by microfiltration (PAC-MF) and reverse osmosis; and two-pass
reverse osmosis.
In both treatment tests, the landfill leachate was mildly acidified by hydrochloric
acid addition to pH 6. As the rate of oxidation by oxygen is inversely proportional to
the hydronium ion concentration, lowering the pH is meant to preserve iron as a
ferrous salt solution, thereby preventing colloidal fouling of the membrane. The
use of 5 um filters provided additional protection against debris and particulate
dogging. Analyses for the bench-scale studies covered: total priority pollutants plus
40 other pollutants except PCBs and pesticides; iron, calcium, and priority pollutant
metals, sulfate; cyanide; total suspended and dissolved solids; TOC; and COD.
48
-------
The rejection percentages of first pass RO alone, without PAC-MF, were
respectively: 60.0% for methylene chloride; 78.4% for trans-l,2-dichloroethene;
above 99.9% for xylenes (total); 98.9% for toluene; above 99.9% for ethyl benzene;
and above 99.9% for benzene. For the semi-volatile organics phenol was rejected at
79.8% whereas higher results were exhibited for 2,4-dimethylphenol at 96.8%. In the
case of inorganics, nickel (98.1%), arsenic (above 99.9%), lead (above 99.9%), and
chromium (above 99.9%) were easily removed by reverse osmosis. TOC level was
reduced by 88.8%. Second pass permeate did not significantly increase rejections
over the first pass permeate levels.
The success achieved in this evaluation demonstrated reverse osmosis to be a
technically feasible treatment for the PAS site. Also, potential savings identified in
a subsequent economic analysis substantiated the need to proceed with the phase II
engineering studies (Evangelista, 1988). On the other hand, powdered activated
carbon adsorption followed by microfiltration separation appeared to be an
unnecessary treatment step because of the economics and the generation of
additional solid wastes.
3.3 Reverse Osmosis System Configuration For The On-Site Tests
The raw leachate was pretreated to lower the concentrations of the ferrous ion, and
calcium carbonate, and solubilize residual metallic species. The following describes
the pretreatment system. Batches of raw leachate were pumped from the recovery
wells and piped to a concrete in-ground storage tank with a capacity of 200,000 litres.
A stainless steel submersible pump fed raw leachate to the pretreatment system. The
RO pretreatment steps included: basification; clarification; dead-end filtration;
acidification and dead-end filtration. Basification included the injection of a
25% sodium hydroxide aqueous solution into the raw leachate stream within an
on-line static mixer. Then, the basified leachate discharged into an 800-litre reaction
tank. The flocculating solution was transferred to a Lamella® clarifier. The clarified
effluent was then pumped through 5-micron and 0.2-micron cartridge filters in
series, acidified by means of a 50% hydrochloric acid on-line injection, statically
mixed, and finally discharged into a 2000-litre feed tank (Figure 2). Flows and pH
were monitored and adjusted as to maintain a pH of 10 in the reaction tank, and a
pH of 5 in the RO feed tank. The acidified leachate was then filtered through a
0.2-micron cartridge filter and introduced to two high pressure pumps before
contacting the RO membranes for separation.
Environment Canada's mobile RO system was used for the phase II engineering
studies. Its configuration was arranged to two pairs of 4"x40" spiral-wound
membranes in series in each bank. Four different manufacturers' membranes were
utilized in the course of this evaluation; therefore, each pair of RO membranes was
representative of one brand.
Lamella® is a registered trademark of Axel Johnson Engineering AB, SWEDEN
49
-------
Tests were carried out over a period of 10 days in August, and September 1988. The
distinction between each trial was established by a variation in the water recovery
ratio, and in the operating pressure. Sufficient time was allowed between each
sampling effort for the membranes' acclimatization.
Membrane separation generated concentrate and permeate effluents. The permeate
could, at any point in time, be discretized into four separate streams for sampling
purposes, each one being the product of one membrane type. The effluents were
either directed to the enhanced ultraviolet oxidation system for further treatment or
returned to an injection well in the landfill.
3.4 Sampling And Analysis Of Samples
Samples were taken at appropriate times during the trials from the RO feed solution,
also called basified/acidified leachate, from each of the four permeate ports, and
from the concentrate. The Lamella® clarifier sludge, and the raw leachate were
each sampled on one occasion. REAC Standard Operating Procedures (SOPs) #2001
to 2005 were observed for sampling, sample storage, and sample shipment. All
abovementioned SOPs are approved by the ERT (Evangelista, 1989). On-site
analyses were performed for volatile organic compounds (VOCs) as per US EPA
methods 601 and 602 using a Hewlett Packard (HP) 7675A purge and trap unit
followed by a HP 5830 gas chromatograph (GC) with a flame ionization detector.
VOC analyses were verified off-site for some samples by a modified version of
US EPA method 524.2 using a HP 5995C gas chromatograph/mass spectrophotometer
(GC/MS) equipped with a Tekmar LSC 2000 purge and trap concentrator. The use of
a reduced sample size, that is 5 ml, corresponded to the only adaptation to the
analytical method of US EPA method 524.2 (Evangelista, 1989).
Off-site analyses were performed for the semi-volatile organics, also known as base
neutral/acid extractables (BNAs), and priority pollutant metals. BNA analyses were
done in conformity with the separator extraction technique of US EPA method 625
by means of a HP 5995C GC/MS. Finally, priority pollutant metals were analyzed
according to the US EPA method 7000 series. Zinc and nickel were quantified by
flame atomic absorption spectrophotometry using a Varian SpectrAA-300. In the
case of arsenic and lead, graphite furnace atomic absorption spectrophotometry was
performed using either a Varian 400-Z or a Varian SpectrAA-20 both set up with a
GTA-95 graphite furnace unit (Evangelista, 1989).
50
-------
4.0 Results and Discussion
For each test> the percent rejection of each membrane was computed for each
pollutant. Then, the mean rejection was computed from the results of all tests.
Table 1 lists those latter values for each of the four membranes labelled A to D. The
selected pollutants were chosen because they represent various families of
contaminants and because of their relative occurrence in the PAS site leachate.
Therefore, the results tabled herein represent only a portion of those generated
during the phase II engineering studies at the PAS site. However, the selection of
representative pollutants is meant to provide a broad coverage of the quantification
process. All the data can be found elsewhere (Evangelista, 1989).
In general, the RO membranes tested in this study rejected the heavy metals with
the highest success. Among the organic compounds, the base/neutral/acid
extractable compounds were more easily rejected than the VOCs. Membrane A
rejected volatile organic compounds and semi-volatile organics at levels that did not
reach 85% in most cases. The only exception was 2-methylnaphtalene (85.0%). A
contaminant reduction of that order of magnitude was accomplished for all metals
except for lead (10.3%) and zinc (45.4%). Membrane B tended to retain organic
compounds at a higher percentage than membrane A. Among the VOCs,
bromoform (91.5%), 1,1,1-trichloroethane (85.2%), and
methyl isobutyl ketone (85.4%) were retained in the concentrate with the greatest
rejection efficiency. The results for membrane B also revealed with more clarity
than those for membrane A a trend of higher rejection levels with higher molecular
weight. Benzoic acid (85.5%) and 1,2-dichlorobenzene (91.7%) were the BNAs which
were separated with the greatest ease by membrane B. Rejections for metals
exceeded the 85% level except for lead (16.9%) and zinc (33.6%). For membrane C,
separation levels rarely exceeded 80%. The compounds that were separated above
that percentage were: meta- and para-xylene (coeluted 84.6%); ortho-xylene (83.4%);
1,1,1-trichloroethane (85.7%); and copper (84.4%). Membrane D demonstrated its
capacity to remove in excess of 90% of the contamination associated to six VOCs.
These were: 1,1,1-trichloroethane (93.3%); bromodichloromethane (91.6%);
bromoform (99.5%); ethylbenzene (94.9%); meta- and para-xylene (co-eluted 96.3%);
and ortho-xylene (96.7%). Five compounds among the BNAs were separated at
levels above 85%: 1,2 dichlorobenzene (87.9%); 4-methylphenol (89.8%);
2,4-dimethylphenol (88.5%); naphtalene (88.9%); 2-methylnaphtalene (92.6%). As
for all membranes, removal of metals reached one order of magnitude except for
lead (29.1%), and zinc (53.3%).
Overall, membrane D exhibited the best performance with the highest rejection
levels for the majority of the contaminants. Of all membranes, membrane D
achieved an order of magnitude change in the greatest number of pollutant
concentrations.
51
-------
The results indicate that the molecular size and polarity have an impact on the
performance of RO membranes. The RO membrane rejects the larger molecules
because the membrane acts as a molecular sieve. On the other hand, polar
molecules are less amenable to separation than non-polar molecules. For example,
benzene and toluene showed higher levels of rejection than the polar compounds
phenol and 4-methylphenol. It is believed that the phenolics, when dissociated,
can form a hydrogen bond with the membrane and diffuse through it more readily
than a non-dissociated molecule.
To complete our analysis of the performance from a quantitative standpoint, the
permeation rate of the single-stage RO system was examined for all twelve tests.
Between 0 and 21.3 hours, six different tests were performed on separate days of
operation where most of the fluctuations in the permeate flowrates are related to the
start-up conditions, and the adjustment of the operating pressure (Figure 3). At
10.3 hours, the "spike" illustrates the effect of pressurizing the system from 600 psi
to 800 psi. On the other hand, the increase in the combined permeate flowrate (12.5
to 27.0 litres per minute) that appears at 12.5 hours results from the use of a more
dilute solution, namely the enhanced ultraviolet oxidation effluent. In the
following tests, normal pretreatment of the leachate was resumed. The increase in
the osmotic pressure that accompanied the change in the feed solution at 15.2 hours
and a reduction of the operating pressure to 600 psi caused a decrease in the
permeation rate from 28 to 22 1pm. This test ended at 19.5 hours. As the solutions
left in the tanks lost during the following 20 hours of downtime part of their organic
contamination through volatilization, the permeate rates reflected these losses at
start-up. A pressure of 600 psi was maintained and flows gradually decreased to 10
litres per minute. This trial ended at 21.3 hours.
In the next sequence of tests, the variation in the permeate fluxes was mostly
accounted by the adjustment of the water recovery ratio and the fluctuation in the
chemical composition of the PAS site leachate. After 21.3 accumulated hours of
operating time, tests were initiated at 800 psi, and the system was not shutdown
until 63.8 hours. Varying the recovery ratio at a given pressure affected the
permeation rate. When more water was recovered, the osmotic pressure increased
at the membrane surface, therefore decreasing the permeation flux. On the other
hand, different osmotic pressures of the feed were associated to different chemical
compositions of the landfill leachate. At a fixed operating pressure and water
recovery ratio, the changes in the osmotic pressure translated into a variable net
driving pressure, and caused the instability in the permeate flowrates. The
concentration ranges listed in Table 1 illustrate that for some chemical species, the
concentration varied by as much as three orders of magnitude.
52
-------
5.0 Conclusion
Rejection of metals is generally easily accomplished. Caution should be exercized
when semi-volatile and volatile organic compounds are involved since the
rejection efficiency is somewhat lower than for inorganic species. Still, acceptable
levels of rejection can be achieved on some volatile organic compounds. For
instance, membrane D rejected several volatile compounds at levels exceeding 90%.
Therefore, membrane selection is an important step of any treatability study which
is to involve reverse osmosis as a hazardous treatment process.
The preliminary results of the phase II engineering, studies at the PAS site in
Oswego, NY, demonstrated that reverse osmosis can be used as an effective part in a
treatment process for landfill leachate. The economics of this separation process
depends to a great extent on the liquid wastes themselves. The pretreatment of the
feed water, the treatment/disposal of both the RO concentrate and the clarifier
sludge bear heavily on the economics of the process. Discharge limits ultimately
govern the costs and the benefits associated with this technology. The low
rejections of volatile organics suggest that other treatment methods be coupled to a
reverse osmosis system. Thus, RO should not be regarded as a panacea but as means
to resolve specific contamination problems within the context of a treatment train
where it can conveniently complement additional treatment processes.
6.0 References
Evangelista, R. 1989. Pilot-Scale Engineering Study at the Pollution Abatement
Services Site, Oswego, NY. Draft Report prepared for US EPA under EPA contract
No. 68-03-3482, Edison, New Jersey.
Evangelista, R. 1988. Preliminary Economic Analysis for the Proposed Treatment
Systems at the Pollution Treatment Systems at the Pollution Abatement Services
Site, Oswego, NY. Memorandum to US EPA ERT, Edison, New Jersey.
Lombard, G. 1986. Precedes de separation par membranes. Document de cours.
Ecole Polytechnique de Montreal, pp. 44-45.
Lonsdale,H.K., Merten, U., and Riley, R.L. 1965. Transport Properties of Cellulose
Acetate Osmotic Membranes. J. Appl. Poly. Sci., 9:1341.
Sourirajan, S., and Matsuura, T. 1985. Reverse Osmosis/Ultrafiltration Process
Principles. National Research Council Canada, Ottawa, pp. 1-77.
Whittaker, H., et al. 1989. Preliminary Results of Reverse Osmosis and Ultraviolet
Photolysis/Ozonation Testing at the PAS Site - Oswego, NY, In Proceedings of the
Technical Seminar on Chemical Spills. Calgary, Alberta. Environment Canada.
June 5-6.
53
-------
Concentrated
Solution "^
Dilute
Solution
, ^-
JU
T
t ^
Dilute
^x
Solution
»
"*"
>
si
1
/
J
Osmosis
Reverse Osmosis
Figure 1 - Osmosis: water transport through the semi-permeable membrane
is from the dilute solution to the concentrated solution. At
equilibrium, water flow stops, and the chemical potential is equal
on both sides. The head difference is the osmotic pressure.
Reverse osmosis: a pressure exceeding the osmotic pressure
forces the water transport across the semi-permeable membrane.
Solutes are for the most part retained on the side of the membrane
where the pressure is applied.
In-ground storage
tank
200,0001
Static mixer
Reaction tank
Cap. 8001
02. micron
cartridge
filters
5 micron
cartridge filters
RO feed-tank
Cap. 2,0001
RO system
0.2 micron
cartridge
filters
Figure 2 - Pretreatment System Utilized for the Phase II Engineering
Studies at the PAS Site in Oswego, NY.
54
-------
en
en
fl
-
«J
01
PU
Permeate Flowrate [1/min]
Recovery Ratio
1.00
-0.75
-------
Table 1 - Summary of the Results
for the Phase II Engineering Studies at the PAS Site, Oswego, NY.
Compound/Metal
Average Percent Rejection For All Tests
Concentration
Range
(in the feed) Membrane A Membrane B Membrane C Membrane D
(Mg/1]
or
01
Dichloromethane
Acetone
1,1-Dichloroethane
1,2-Dichloroethane
trans-1,2- Dichloroethene
Benzene
Bromoform
Toluene
Ethyibenzene
o-Xylene
Phenol
4-Methyl phenol
Benzoic Acid
Arsenic
Lead
Nickel
129
4912
178
635
319
60
226
75
1266
3339
121
67
46
23
4
1070
-99642
- 68768
-3612
- 6931
- 62577
- 2561
- 198966
- 26145
- 19211
- 52033
-7155
-4811
- 42331
-67
-59
- 2630
34.3
46.0
57.2
41.0
11.7
56.7
74.9
57.7
66.1
71.9
54.5
78.6
67.9
96.1
10.3
>99.9
51.5
49.5
64.7
52.3
37.3
58.0
91.5
68.4
74.3
83.2
55.7
57.4
85.5
85.5
16.9
93.8
42.5
31.6
64.3
53.7
37.0
55.8
69.1
53.3
74.6
83.4
37.8
59.8
51.7
46.1
20.0
53.6
50.5
73.8
88.2
75.3
47.4
83.9
99.5
83.7
94.9
96.6
72.2
89.8
83.3
98.6
29.1
88.2
Printed in Canada
-------
KINETIC
in theory and practice
ABSTRACT
During the last four years, Geokinetics has been developing a method to
remove heavy metals and other contaminants from soil and groundwater.
The method is based on the electrokinetical phenomena electro-osmosis,
electrophoresis and electrolysis, which occur when the soil is
electrically charged by means of one or several electrode-arrays.
The most important applications of these phenomena with respect to the
soil have been the dewatering of clays by electro-osmosis and experi-
ments to desalinize arable lands (USA, 1958 and USSR, 1966-1975).
Experiments on a very small scale to remove heavy metals from soils are
documented from UK (1980, 1981 and 1982). Though starting promisingly,
the authors reported problems around the electrodes (precipitates),
which influenced the process negatively.
Geokinetics has found a solution to these problems by developing an
electrokinetical installation, which monitors and controls the chemical
reaction environment around the electrodes. The core of such an
installation consists of the electrode-series and their housing, which
can be installed in principle at any depth, either horizontally or
vertically. The housings are interconnected and form two separate (one
for the cathode, one for the anode) circulation systems, filled with
different chemical solutions. In these solutions the contaminants are
captured and brought to a container-based water purification facility.
The energy is supplied by a generating set or taken from the main.
Electro-Reclamation can be applied both in situ (soils) and on- or off
site (excavated soil, scooped out river slush). The electrokinetical
phenomena can also be used to fence off hazardous waste sites or
potentially hazardous industrial sites.
The technique has been tested on the basis of numerous laboratory
experiments, using different types of soil (clay, peat, argillaceous
sand) and contaminants (As, Cd,.Co, Cr, Cu, Hg, Ni, Mn, Mo, Pb, Sb,
Zn). Besides, two in situ fieldexperiments (Cu, Pb and Zn) have been
finished and one "official" in situ remediation project (As) has been
succesfully terminated.
Reduction of individual heavy metal concentrations can be more than
90 %, depending on the energy supply and time duration.
Remediation costs are therefore directly related to these two factors.
Costs range from less than US $ 50 per ton, when relatively low energy
is supplied over long periods (several months), to more than US $ 400,
when the time period is reduced to several weeks and the enrergy supply
has to be increased accordingly. There is, however, a limit to the
current strenght which can be used. In practice therefore an optimum
is calculated for energy supply and time duration.
drs. R. Lageman
Geokinetics
57
-------
KINETICS
Forum on Innovative Hazardous Waste
Treatment Technologies
Atlanta, Georgia USA
19 - 21 June 1989
Theory and Practice of Electrc—P
drs. Reinout Lageman
Geokinetics
Rotterdam, Groningen
the Netherlands
co-authors :
drs. W.Pool
drs. G.A.Seffinga
Geokinetics
58
-------
KINETICS
Electro-Reclamation in theory and -practice
Electroklnetical phenomena
During the last 4 years Geokinetics has been developing a method to
remove heavy metals and other contaminants from soil and groundwater.
The method is based on electrokinetical phenomena, which in one way or
another have been made use of since the end of the last century. These
phenomena occur when the soil is electrically charged with DC by means
of one or several electrode arrays :
1. Electro-osmosis : Movement of soil moisture or groundwater from the
anode to the kathode.
2. Electrophoresis : Movement of soil particles within the soil
moisture or groundwater.
3. Electrolysis : Movement of ions and ioncomplexes within the soil
moisture or groundwater.
Electro-osmosis
the electro-osmotic transport depends on the following factors :
- the mobility of the ions and charged particles within the soil
moisture or groundwater;
- the hydratation of the ions and the charged particles;
- the charge and direction of the ions and charged particles, which
cause a net water movement;
- the ion-concentration;
- the viscosity of the pore solution, depending a.o. on the capillary
size;
- the dielectrical constant, depending a.o. on the amount of organic
and anorganic particles in the pore solution;
- the temperature.
From existing literature and own experiments the average electro-
osmotic mobility has been calculated to be in the order of 5.10"*
m2/U.s, where U = potential drop (V/m).
To drain 1 m3 of soil by electro-osmosis, the following parameters
should be known :
- the porosity;
- the moisture content of the soil to be treated;
- the conductivity of the pore solution;
Apart from these other factors like the desired time period, the use of
the soil after treatment and safety requirements regards maximum
voltage and current are also of importance.
Electrophores i s
Electrophoresis (kataphoresis) is the process of movement of particles
under the influence of an electrical field. With particles is meant all
electrically charged particles like colloids, clay particles floating
in the pore solution, organic particles, droplets etc.
59
-------
K I N
The mobility of these particles corresponds with that of ions. Within
the pore solution these particles transfer the electrical charges and
affect the electrical conductivity and the electro-osmotic current.
Clay minerals as such have 2 electrical polarity possibilities. One
consists of the stucture-based dipole moment, which depends on the
atomic masses and has an orientation parallel to the longest axis of
the clay particle. The second polarity stands at right angles to the
first and is caused by the external electrical field. It depends on the
way of polarization of the electrical double layer. The mobility of
clay particles is an interplay between these two moments and is,
therfore, less than the electro-osmotic mobility. It varies between
1.10-10 and 3.10" mVU.s.
Electrolysis
Analogous to electro-osmosis and electrophoresis, where one considers
only water transport or particle transport respectively, with electro-
lysis only the movement of ions and ioncomplexes is taken into
consideration. The average mobility of ions lies around 5.10-" mVU.s,
which is a factor 10 greater than that of the electro-osmotic mobility.
Therefore, the energy necessary to move all ions over an average
distance of 1 m through aim2 soil section its 10 times less than with
electro-osmosis.
To calculate the energy necessary to dispose of the contaminants within
1 m3 of soil, the following factors are of importance :
- the chemical form of the contaminants;
- the concentration of the contaminants;
- the required concentrations of the contaminamts;
- the behaviour of the contaminants at different pH-levels;
- pH-control around the electrodes within the soil;
- removal of the contaminants and particles at the respective
electrodes;
- supply of a conditioning solution to replace the removed contaminants
and other particles at the electrodes;
- processing of the contaminated solution removed at the electrodes.
The application of electrokinetical phenomena in practice
Up to a few years ago the most important application of electro-
kinetical phenomena with respect tot the soil had been the dewatering
of clays by electro-osmosis. Apart from that experiments and research
in the USA (Collopy, 1958) and the USSR (Vadyunina et al., 1966-1975)
were aimed at developing a technique to remove accumulated salts from
agricultural terrains.
Experiments on a very small scale both in the laboratory and in the
field are documented by Hamnet (1980), Agard (1981) and Warfield
(1982). These experiments were aimed at the removal of heavy metals
from the soil. Though the experiments started promisingly, the authors
reported problems around the electrodes (chemical precipitates) after a
certain time period, which influenced the process in an adversative
way. Especially the changing of the pH around both electrodes influen-
ces the mobility of the heavy (and also lighter) metals.
60
-------
KINETICS
'' 'contaminated
GENERATOR
(WHAM
circulation system
current supply
— — — - boundary of electrokinetical treatment
Fig. 1 : Schematic representation of ER-field unit and
electrokinetical transport in the soil
61
-------
KINETICS
Electrokinetical Installation
Geokinetics has found a solution to these problems. Both for the
laboratory and the field an electrokinetical installation has been
developed, which controls the chemical reaction environment around the
electrodes. The electrokinetical processes are monitored and managed
actively, thus averting adverse effects.
The core of an electrokinetical installation (fig. 1) consists of the
electrode-series and their housing. These can be installed in principle
at any depth, either vertically or horizontally. Both the cathode- and
anode housings are interconnected and form two seperate circulation
systems (one for the cathode, one for the anode), filled with different
chemical solutions. In these solutions the contaminants are captured
and brought to a water purification facility,, installed in a container
together with the solution tanks and measuring and monitoring devices.
The energy is supplied by a generating-set or taken from the main.
The consequence of this set-up is, that electro-reclamation can be
applied both for in situ remediation of contaminated soil (fig. 2) and
for on- or off site remediation of excavated polluted soil or scooped
out river slush.
Laboratory experiments
The method of electro-reclamtion has been tested on the basis of
numerous laboratory experiments. They focussed on important parameters
like kind of current, strength of current, voltage, moisture content,
chemical additives and the like. Besides, the effectlviness of the
method as regards certain soil- and heavy-metal types has been examined
with the help of several simulation experiments (clay, peat, fine
argillaceous sand polluted with As, Cd, Co, Cr, Cu, Hg, Mn, Mo, Ni, Pb,
Sb, Zn). They rendered also a good insight into the energy demand and
the time duration. Some results are presented in table 1 and figures 3a
and 3b.
Soil type
Peat
Pottery clay
Fine argil-
laceous sand
Clay
Metal
Pb
Cu
Cu
Cd
As
Cone, before
(ppm)
9000
500
1000
275
300
Table 1.
Cone, after
(ppm)
2400
200
100
40
30
Energy
(kWh/ton)
56
56
14
110
115
The next table shows the results before and after treatment of a soil
sample of fine argillaceous sand, contaminated with several metals. The
energy demand amounted to 30 kWh/ton.
62
-------
KINETICS
n
A) Remediation of residential areas
B) Remediation of industrial areas
0 Remediation/fencing of hazardous waste sites
D) Preventive electrokineHcal fence around potentially hazardous industrial complexes
Fig. 2 : Some applications of in situ Electro-Reclamation
63
-------
KINETICS
Soil type Metal Cone, before Cone, after
(ppm) (ppm)
Decrease
Fine argil-
laceous sand
Cd
Cr
Ni
Pb
Hg
Cu
Zn
319
221
227
638
334
570
937
< 1
20
34
230
110
50
180
99
91
85
64
67
91
81
Table 2.
Average :
83
Table 3 lists the results of a laboratory experiment with a sample of
mud, dredged from the river Weser in Germany. The energy demand for
this experiment amounted to 100 kWh/ton.
Soil type
River slush
Metal
Cd
Cu
Pb
Ni
Zn
Cr
Hg
As
Cone, before
(ppm)
10
143
173
56
901
72
0.5
13
Cone, after
(ppm)
5
41
80
5
54
26
0.2
4.4
Decrease
( % )
50
71
54
91
94
64
60
66
Table 3.
69
Fieldexperiments
Fieldexperiment 1
The first fieldexperiment took place alongside part of a waterbearing
ditch, on one side bordered by a former paint-factory and on the other
side by open grassland. The bank on the latter part was heightened by
sediment dredged from the ditch. This sediment was heavily polluted
with metals in the form of paint residuary. The raised sediment layer,
height 20 - 50 cm, length 70 m and width 3 m, contained Pb and Cu
concentrations up to 10,000 ppm and 5,000 ppm respectively. The
original peat soil underneath was contaminated by leaching of this
overlying layer with Pb-concentrations ranging from 300 ppm to more
than 5,000 ppm, while Cu-concentrations were in the order of 500 to
1,000 ppm.
A preceding laboratory test with a sample of the sediment reduced the
concentration of Pb from * 9,000 ppm to * 5,000 ppra and that of Cu from
« 4,500 ppm to « 1,600 ppm, all within a time period of 320 hours.
For the fieldexperiment one cathode- and one anode-series were
installed both with a length of 70 m and a mutual distance of 3 m.
64
-------
0.
a.
1000
800
I 600-j
ro
2 400-
c
o
5 200-j
KINETICS
0 20 40 60 80 100 120 140 160
time (hours)
Fig. 3a : Decrease of Copper during elecfrokinetical treatment
of contaminated pottery clay
Q.
O
'•^
ro
c
o
250
200-
150-
100-
50-
0
5 10
time (days)
15
Fig. 3b : Decrease of Cadmium during electrokineticat treatment
of contaminated fine argillaceous sand
65
-------
•-377?
K I N E T I C S
The cathode was installed horizontally, while the anodes were implaced
vertically into the soil about 2 m apart. On the basis of the energy
consumption during the laboratory test the fieldexperiment was confined
to 430 hours.
The changes in Pb- and Cu-concentrations were monitored at 26 sampling
locations, sampled at regular intervals at several depths (10,20,30,40
and 50 cm below groundsurface). The following table lists part of the
analysis-results for Cu and Pb within the peat at a depth of 30 to 40
cm below ground surface. The spatial distribution of the pollutants at
the beginning and at the end of the test are shown in figs. 4a and 4b).
Sample
(30-40
1
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
9
9
point Metal
cm b.gs)
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Pb
Cu
Cone, before
(ppm)
440
185
3900
540
> 5000
1150
> 5000
475
> 5000
1170
> 5000
580
3780
410
380
35
340
50
Cone, after
(ppm)
110
35
700
220
560
580
2450
250
610
230
300
45
285
30
180
15
90
15
Decrease
( % )
75
81
82
59
89
50
51
47
88
80
94
92
92
93
53
57
74
70
Table 4
Average :
74
Fieldexperiment 2
The second, fieldexperiment was carried out on the site of a galvanizing
plant. According to preceding investigations the soil (sandy clay)
around the plant was contaminated with Zn to a depth of 40 cm below
groundsurface. In the upper 10 cm Zn-concentrations were reported to
have a maximum of 3,000 ppm. At greater depths Zn-concentrations were
indicated as being in the order of 500 ppm.
For the experiment an area was selected with dimensions of 15 m x 6 m x
1 m. Two cathode-drains were installed at a depth of 50 cm below
groundsurface, while 33 anodes, divided along 3 rows were implaced in
holes of 1 m depth with a mutual distance of 1.5 m. The distance
between the cathode- and anode-series was also 1.5 m.
Energy was supplied by a 100 kVA generating set. The resistivity of the
soil was 5 iim. The installation was calculated for a DC-supply of
8 Araps/mz of soil, which should result in a potential drop of 40 V/m.
This potential drop could not be maintained during the whole period. As
a result of some material problems it was neither possible to maintain
a 24 hour energy-supply to the soil.
66
-------
KINETICS
- 3m -
cathode-series
anode-series
Cu>500ppm
1005000ppm
0 600a. Decrease of Cu
4b. Decrease of Pb
67
-------
KINETICS
Within 2 weeks temperature rose from 12 °C to an average of 40 °C. As a
consequence soil resistivity decreased to 2.5 Sim and the potential drop
to 20 V/m with an average strength of current of 8 Amps/m2. The
effective energy-supply per 1000 kg of soil amounted to 160 kWh during
the 8 week period.
Changes in Zn-concentration were monitored at 12 sampling locations,
which were sampled at 3 different depth intervals (10, 30 and 50 cm).
Changes In groundwater concentration were monitored in 2 observation
wells. In table 5 and fig. 5 the analysis-results are given for the
depth interval of 30 cm. .
Sample point Metal Cone, before
(30 cm b.g.s.) (ppm)
1 — , 5120
2
3
4
5
2030
1600
2320
2450
6 > Zn 4390
7
8
9
10
11
1960
3250
2400
70
150
12 — 1 120
Cone, after
(ppm)
4470
1960
800
2320
2450
2360
940
1960
2000
30
120
80
decrease
( % )
13
3
50
0
0
48
52
40
17
57
20
33
Table 5.
Average :
20
The energy demand for this test amounted to 160 kWh/ton. At the
beginning of the test the highest Zn-concentration amounted to 7,010
ppm with an average of 2,410 ppm over the whole area. At the end of the
test the highest Zn-concentration was 5,300 ppm and the average had
been decreased to 1,620. The concentrations of Zn, Pb and Cd in the
groundwater and the filtercake are presented in the tables 6 to 8.
A total of some 1000 kg of filtercake was produced with an average Zn-
content of 117 g/kg. This comes to a total removal of some 50 kg of
zinc, assuming an average moisture content of the filtercake of 60 % .
A rough mass-balans could be summarized as follows :
- treated volume of soil : 15 x 6 x 0.5 x 3/4 = 34 m3 (1/4 of the area
did not show increased Zn-concentrations).
- Weight : 34 m3 x 1.8 = 61 tons.
- Weight of filtercake : 1000 kg.
- average moisture content : 60 %
- total dry matter : 400 kg.
- average Zn-concentration : 117 g/kg.
- amount of zinc removed : 47 kg
- removed per 1000 kg of soil : 47 x 10V61 x 103 = 770 ppm.
The last value is in the same order of magnitude as the average
decrease in Zn-concentration (2410 ppm -1620 ppm = 790 ppm).
60
-------
K I N E
An important outcome of the test was the relatively high energy
consumption against a rather low Zn-mobility. This was the result of
the high buffering capacity of the soil, caused by the presence of NH3
and NH«C1 (as was established later), which is used during the
galvanizing process. During a following laboratory test with a soil
sample from the area it was found that the energy necessary to reduce
Zn-concentrations below the 200 ppm level would amount to 500 kWh/ton
of soil. With an unchanged power-suplly of 100 kVA this would mean
tripling the time period of 8 weeks to 24 weeks.
Metal : Zn
sample treatment : not
date
24-10-88
01-11-88
09-11-88
17-11-88
25-11-88
30-11-88
24-10-88
01-11-88
09-11-88
17-11-88
25-11-88
30-11-88
obs. well
1
1
1
1
1
1
2
2
2
2
2
2
Pb
Cd
acidified
Zn
Pb
Cd
acidified
ppm
200
120
130
172
130
120
10
1.5
2.5
6
2.8
4
0.09
0.07
0.06
0.07
0.17
0.13
0.17
0.09
0.06
0.03
0.03
0.09
0.06
0.00
0.02
0.03
0.03
0.03
0.02
0.09
0
0
0.01
0
270
140
160
198
180
150
40
2
3
8
5.8
5.5
1.4
0.09
0.17
0.11
0.22
0.16
0.34
0.15
0.15
0.07
0.18
0.14
0.07
0
0.02
0.04
0.03
0.03
0.02
0
0
0.01
0.01
0
Table 6. Zn-content (ppm) of the groundwater in obser-
vation wells 1 and 2.
Metal : Zn
date
24-10-88
09-11-88
17-11-88
25-11-88
30-11-88
136.9
199
99
89
61
Pb
g/kg
1.9
1.1
2
1.5
0.58
Cd moisture cnt
0.
0.
0.
0.
0.
34
18
12
16
11
in X
78
Table 7. Zn-content (R/kg) of the filtercake during
Electro-Reclamation.
Metal : Zn
date
30-11-88
Pb
Cd
ppm
30
0.6
' 0
Table 8. Zn-content (ppm) of the solution in
the anode-circulation system.
69
-------
KINETICS
E
ui
Zn-concentrations 30 cm below g.s.
(24/10/88)
Zn-concentrations 30 cm below
(16/12/88)
g.s.
Zn>tOOOppm ||| 2000 250ppm
Fig. 6 : Results of remedial action project 1
As-concentrations 1 m below g.s.
(28/^/89)
HJ30
-------
KINETICS
Remedial action Project 1
The first 'official' Electro-Reclamation project started at the end of
January 1989. It concerned an As-pollution on the site of a former
timber-impregnation plant. After a fire in 1984, which destroyed a
large part of the plant, it was decided not to rebuild the plant. After
dismantling the same a 'statement of unpolluted soil' was needed in
order to allocate the land to building plots. A following investigation
established the presence of As-concentrations up to several 100 ppm in
part of the heavy clay soil to a maximum depth of 2 m. Cause of the
pollution was attributed to 'Superwolmansalt D' (NaaHAsCU.THaO), used
for impregnation.
In April 1988 Geokinetics was requested to investigate the possibility
of remediating the soil by Electro-Reclamation. A following laboratory
test with a soil sample reduced the As-concentration of 300 ppm to 30
ppm against an energy consumption of 115 kWh/ton. An additional field
investigation delineated the pollution to an area of 10 m x 10 m,
contaminated to a depth of 2 m and an adjoining area of 10 m x 5 m,
contaminated to a depth of 1 m. Total volume of polluted soil : 250 ms
(= 450 tons).
The project started in January 1989. Along the length of the polluted
area 4x2 cathode-drains were installed : one at a depth of 1.5 m and
the other at 0.5 m. The cathode-series had a mutual distance of 3 m.
In between 36 anodes were implaced in the soil, divided along 2 rows of
14 and 1 row of 8 pieces. Within the area of 10 m x 10 m the anodes
were installed to a depth of 2 m below ground surface. In the other
area of 10 m x 5 m the depth of the anodes was limited to 1 m depth.
All anodes were placed at a mutual distance of 1.5 m.
On the basis of both the laboratory test and the field investigation
the duration of the Electro-Reclamation period was calculated to last
50 (24 hour) days, using an energy-supply of 200 kVA (= 44 kW effective
into the soil.
At the beginning the resistivity of the clay was 10 Qra and soil
temperature at a depth of 0.5 m was 7 °C. After 3 to 4 weeks temperatu-
re had risen to an average of 50 °C, while the resistivity decreased to
5 iim. The original potential drop of 40 V/m decreased accordingly to 20
V/m with an average current-strength of 4 Amps/m2 (total crosssectiornl
area being 110 mz).
Changes in As-concentrations were monitored at 10 fixed sampling
locations and numerous randomly distributed sampling-points. Of the
fixed locations, 2 were sampled at 0, 0.5, 1, 1.5 and 2 m depth. The
others at 0, 0.5 and 1 m depth. The analysis-results at a depth of 1 m
below ground surface are listed in table 9 and fig. 6. The last
sampling date being April 30th.
With an average As-concentration over the whole area of 110 ppm before
the process started, the total As-content amounted to some 50 kg. On
April 30, 3/4 of the area showed As-concentrations below the required
30 ppm boundary. One spot remained, however, with relatively high
As-concentrations. Calculations showed, that some 25 to 30 kg of
arsenic had been removed.
71
-------
KINETICS
150 m
anode-series (vertical)
O cathode-series (vertical)
flow P^hs of positively charged contaminants
Fig 7a. : Set-up of electrokinetical fence in soil with low permeability
L
-• anode-series (vertical)
-O cathode-series (vertical)
*- flow paths of positively charged contaminants
direction of groundwater flow
Fig. 7b : Set-up of elecfrokinetical fence in soil with moderate to high permeability
72
-------
KINETICS
Because of the time pressure it was decided tot continue the ER-process
for another two weeks, whereafter the soil, which at that time would
still show too high As-concentrations, would have to be excavated and
removed. When the process was stopped on May 12 and the excavation
began, many metal objects like tins, barrels, concrete-iron etc were
found. These objects were supposed to hav been removed by the owner
before the electro-reclamation started, but obviously, had failed to do
so.
The discrepancy, therefore, between estimated and real energy-demand
(and thus elapsed time) was caused by these metal objects, left behind
in the soil. These objects function as preferential flow-paths for the
electrical current and delay the movement of the pollutants in their
vicinity, as first all the iron objects will go into solution.
Sample-point
(1m b.g.s.)
Metal Concentration Concentration Energy
on 24/1/89 on 30/4/89 (kWh/ton)
1 -,
2
3
4
5
6
7
8
9
10 -
385
40
250
310
> As 50
75
40
175
40
60
250 -,
< 20
< 20
190
< 20
30
< 20
< 20
< 20
< 20 J
> 150
Table 9.
Other applications
Electrokinetical fencing
The electrokinetical phenomena occurring when the soil is electrically
charged, can also be used for fencing purposes. These so-called
electrokinetic fences can be installed either at refuse-sites/factory
complexes, where soil pollution has already been ascertained, or where
soil pollution is likely to occur. Depending on the local geohydrologi-
cal situation and the character of the soil, the elctrode configuration
can be such that :
- the elctrokinetical transport is directed towards the source of the
pollution (fig. 7a). The cathode-series is situated nearest to the
source of pollution. Such a set-up should be applied in less permable
soils without substantial groundwaterflow ( < 1 m/year).
- the contaminants, which are carried along with the groundwater flow
are diverted, collected around the electrodes and periodically removed.
In this case the cathode-series is farthest away from the source of
pollution and cathode- and anode-series are installed perpendicular to
the direction of groundwater flow. Such a set-up should be applied when
the soil and/or subsoil is relatively permeable (groundwaterflow
velocity > 1 m/year).
73
-------
KINETICS
Desalination, of arable land
Salination of arable land is a common problem in those countries, where
precipitation is generally low and evapotranspiration high. In
combination with relatively high groundwater levels (coastal areas and
river-lowlands), soils of low permeability and irrigation water with
high total dissolved solids, the accumulation of salts in the top
layers prohibits further agriculture.
The most common technique for landreclamation. consists of the lowering
of the groundwatertable and/or drainage of the soil by means of wells
and/or horizontal drains. The soil is then frequently irrigated with
relatively fresh water, thus leaching the salts from the soil. However
the low permeability of the soil hampers more often than not the
percolation of the leachate to the deeper layers.
By applying electrokinetical processes these problems can be overcome,
as clay and argillaceous soils are specifically suited for electro-
reclamation.
The As-remediation project mentioned afore showed, that after electro-
kinetical treatment, the permeability of the heavy clay-soil had
increased significantly. When furthermore gypsum is added at the anode-
solution, the soil structure will be improved even more and higher crop
production will be obtained (Collopy, 1958).
Cost estimates
Electro-Recalamation
Fig. 8 shows a set of graphs depicting remediation cost per ton of soil
as a function of remedial action time (fig. 8a) and as a function of
the measure of contamination (fig. 8b), assuming a polluted area with
dimensions of 500 m x 100 m x 1 m. From the graphs it is evident that
short remediation periods and highly polluted soil (low resistivity)
require a high amount of energy, having the greatest effect on the
costs.
In practice, however, there is a limit to the electrical current which
can be put into the soil. For every specific case, therefore, an
optimum must be calaculated for energy supply and time duration.
Electrokinetical fencing
In fig. 9 the energy costs per year are given as a function of
groundwater flow velocity (fig. 9a) and as a function of the rate of
pollution (fig. 9b), assuming an electrokinetical fence of 500 m length
and 10 m depth.
74
-------
I I I >
>or-K
< <» 3 c-
" ~° 3 3
energy costs per annum (us $ x 1000)
o o
ii.no
"
U3 10 il
•••
u> (/> n
-r -r
VI I/I o
(/I (A (V
C C n
»
a. c n
» 3 rr
a. n
•< 3 n :
" &£!
Ln £;
«=• ^ o ,
costs per ton (us ;)
> u
^ ^. 2. «
r. n I*
O o •«•
§' o
? "~
-i .
n
§
(A
O
I
s;
o '
Q-
§ ><
'?• -
s §
2. 3
en
5' !i
= 5
energy costs per annum (us t x 1000)
costs per ton (us
n
V)
a.
3
N
°
< t~
=V o
o
3-
3
-------
In areas of low groundwater flow velocity (clciy, argillaceous sand) and
low soil pollution, the yearly energy costs of an electrokinetical
fence are insignificant. This changes rather quickly, when the soil
becomes more permeable (sandy formations) and the groundwater flow
velocity Increases together with the concentration of the contaminants.
For relatively high groundwater flow velocities a combination of
hydrological measures and electrokinetical techniques will render the
most economic results.
Desalination of arable land
Preliminary cost estimations amount to US $ 1000 to 2000 per ha.
Geokinetics
Veerkade 7
3016 DE Rotterdam
the Netherlands
76
-------
VACUUM EXTRACTION TECHNOLOGY
SITE Program Demonstration at
Groveland Wells Superfund Site, Massachusetts
James J. Malot, P.E.
TERRA VAC
Princeton, New Jersey
Abstract
Vacuum extraction is an in-situ or ex-situ treatment process
for cleanup of soils and groundwater contaminated with volatile
'organic compounds (VOCs), liquid-phase hydrocarbons or semi-
volatile compounds. The process of removing VOCs from the vadose
zone using vacuum is a patented process. Demonstration of the
Vacuum Extraction Technology was conducted under the EPA Super-
fund Innovative Technology Evaluation (SITE) Program at the
Groveland Wells Superfund site in Groveland, Massachusetts. The
demonstration included an eight-week pilot test to removed VOCs
(mostly TCE) from the underlying soil and groundwater.
The subsurface conditions included multilayered glacial
deposits consisting of sands, silty sands and clays. Groundwater
was 27 feet deep with a perched water table at about 10 feet. A
multi-layered vacuum extraction and monitoring system was in-
stalled and operated sub-zero weather in northern Massachusetts.
Objectives of the pilot program included testing of soils
before, during and after implementation of the vacuum extraction
process. The effectiveness of the process was monitored by
measuring subsurface vacuum, rates of flow, rates of VOC extrac-
tion and adsorption on activated carbon. The system was designed
to operated on the fringe of the contaminant plume and to quanti-
fy the level of cleanup that could be achieved by the process by
using barrier wells to prevent the migration of contaminants from
the primary source area into the demonstration area.
Results demonstrated the effectiveness of the vacuum extrac-
tion process to clean up contaminated soils. Data from the pilot
test also demonstrated that contaminant distributions were dif-
ferent than originally suspected from the Remedial Investigation
(RI>. In the area at the fringe of the contaminant plume, soil
concentrations were reduced more than 95% to non-detectable
levels. Accordingly, evaluation of the process dynamics and
cleanup rates were adjusted to reflect the actual subsurface
conditions. Additional data from subsequent cleanup work at the
site is presented to further evaluate- the demonstration program.
The objective of evaluating "How clean is clean?" is ad-
dressed with respect to the vacuum extraction process. Results
from other sites where vacuum extraction has been applied is also
presented. Results indicate that vacuum extraction technology is
widely applicable for cleanup of soils and groundwater that are
contaminated with VOCs.
77
-------
Vacuum Extraction Technology
INTRODUCTION
Vacuum extraction technology effectively removes volatile
and semi-volatile compounds from soils and groundwater. Removal
of liquid-phase hydrocarbons floating on the water table using
vacuum extraction technology is faster and more effective than
traditional approaches. Vacuum extraction is typically imple-
mented in-situ, however, treatment of excavated soils on site
using vacuum extraction technology is also effective. Ground-
water can be removed simultaneously from vacuum extraction wells
to further enhance recovery of groundwater contaminants and
reduce the time frame for total cleanup.
Vacuum Extraction Technology was originally developed by
Terra Vac. Since its inception more than five years ago, the
technology has been widely used to cleanup soil and groundwater
contaminated with volatile organic compounds
-------
Vacuum Extraction Technology
soil and groundwater to nondetectable levels.
VACUUM EXTRACTION PROCESS
Under normal static conditions within the soil matrix, VOCs
are partitioned between four possible phases: 1) vapor, 2)
liquid, 3) dissolved in soil water, 4) adsorbed to solid parti-
cles. These four phases define the aggregate contaminant con-
centration in the subsoils.
The vapor phase partitioning is a complex function of water
content, organic content, .solubility, temperature and vapor
pressure. It is not necessary to define the exact relationship
between soil concentration and vapor concentration as a function
of time in order to understand that reductions in extracted vapor
concentrations are driven by continuous partitioning to the vapor
phase which corresponds to reductions in soil concentrations.
Furthermore, as concentrations in soils are rerf-.-.ced significant-
ly, vapor phase partitioning is generally controlled by Henry's
Law. '
As vast volumes of soil vapor are removed by the vacuum
process, fresh air naturally recharges the vadose zone from the
surface. Fresh air moves through the contaminated zone as VOCs
are partitioned from the soil matrix to the vapor phase and move
to the extraction wells. With vacuum induced volatilization and
air stripping of the soil matrix, cleanup occurs continuously.
As contaminant vapors are removed from the subsoils pore
volume, the three other phases (liquid, adsorbed and dissolved)
of VOCs vaporize in place, further reducing the aggregate soil
concentration. Since VOCs vaporiz'e readily, the vacuum extrac-
tion process continually drives the contaminants within the soil
matrix to the vapor state.
Progress of the vacuum extraction soil decontamination
system can be monitored by the concentration of the extracted
vapors. Since vaporization occurs rapidly within the soils, the
soil vapors beyond the immediate vicinity of the extraction well
are near equilibrium with respect to the majority of the contami-
nants contained in the soil matrix.
As VOCs are vacuum extracted from the subsoils, the removal
rate declines with time, indicating cleanup of the soils. During
the vacuum extraction process, vapors extracted at the wellhead
represent essentially an aggregate soil-gas concentration near
the screened interval. Under static conditions, the concentra-
tion of VOCs in the vapor phase is proportional to the aggregate
contaminant concentration in the soil.
The first step of vacuum extraction design is delineation of
the extent and magnitude of the soil contamination and liquid-
phase contaminants that may be floating on the water table.
Vacuum extraction wells are designed with a vacuum-tight seal
79
-------
Vacuum Extraction Technology
near the surface and an extraction zone (screen) corresponding to
the profile of subsurface contamination.
The spacing of vacuum extraction wells is critical to effi-
cient remediation. Depending on the depth to groundwater and
soil type, the radius of influence of an extraction well can
range from tens of feet to hundreds, of feet., Soil permeability,
porosity, moisture content, stratigraphy and depth to groundwater
are important factors in determination of the radius of influ-
ence.
Vacuum extraction technology is effective in treating soils
containing virtually any chemical with a volatile character. All
of the volatile priority pollutants and many of the semi-
volatiles have been successfully extracted with the vacuum proc-
ess. However, metals (except mercury), heavy oils and PCB's will
remain in place as the volatile compounds are extracted by the
process. . Dual vacuum extraction of groundwater and vapors has
been effective at restoring groundwater quality to drinking water
standards within short periods of time.
For those sites with numerous types of compounds (i.e.,
VOCs, PCS, pesticides and metals) a phased approach is often
required. In these cases, it is prudent t:o remove VOCs first
using vacuum extraction so that other technologies can then be
applied more cost effectively and safely. For example, chemical
treatment or incineration of soil, which require excavation, the
health risk of excavation is minimized if the majority of VOCs
are removed first, in-situ, by vacuum extraction. Many methods
used to chemically stabilize metals are more effective after
vacuum extraction has removed VOCs.
Vacuum extraction is an effective means of removing hydro-
carbons floating on the water table. Compared to typical
double-pump systems, skimmer pumps and air displacement total-
fluids product recovery systems, vacuum extraction is faster,
more effective and low cost per gallon of product removed.
Liquid-phase hydrocarbons are removed without pumping groundwater
so that separation and treatment of large volumes of contaminated
water is eliminated.
Where contaminants are within the saturated zone and ground-
water is relatively shallow (i.e., less than 30 feet deep) a
"dual extraction" approach is effective. Dual extraction is a
term used to describe the process of simultaneously extracting
groundwater and vapors under vacuum using the same well. In the
simplest form, operating a submersible pump within a vacuum
extraction well will lower the water table and increase the
effective unsaturated zone in which the vacuum extraction process
will vaporize contaminants.
Simultaneous extraction of groundwater and vapors under
vacuum has several benefits that enhance the rate of groundwater
cleanup. First, the rate of contaminant removal increases com-
pared to groundwater extraction alone since contaminants have two
80
-------
Vacuum Extraction Technology
pathways for removal: aqueous phase and vapor. Even in areas
where groundwater movement beneath the water table is the only
source of contamination, the dual extraction process often yields
the same mass flux (i.e., pounds/day) from the vapor phase as the
aqueous phase. This indicates that substantial partitioning is
occurring in the transition zone and capillary fringe. In medium
to low permeability aquifers the maximum rate at which ground-
water can be extracted from a given well increases two to three
fold using dual extraction. The net effect of these two phenome-
na can yield a six-fold increase in the overall contaminant
removal rate, and hence, a six-fold reduction in the time re-
quired to reach clean-up objectives.
Subsurface Conditions at the Demonstration Site
The Groveland site is underlain by glacial deposits. The
thickness of the unconsolidated deposits under the site is on the
order of 32 to SO feet. In the pilot test area the upper 10 to 16
feet consist of fill and glacial outwash sands. A dense clay
layer, 3 to 7 feet thick, is observed below the sands, but is
discontinuous throughout the site. Glacial till is present below
the clay. The water table is roughly 25 feet deep in the test
area.
A multi-layered vacuum extraction and monitoring system was.
installed, as shown in Figure 1, to segregate the upper sand zone
from the lower till. The vacuum extraction system consisted of
four extraction wells and four monitoring wells; each was capable
of extracting contaminants separately from above or below the
clay. This provided effective hydraulic separation of the two
more'permeable units (sand and till) and allowed differentiation
of the contaminants extracted from the two zones. Activated
carbon was used to control emissions from the site.
Volatile organic soil contamination is generally constrained
within soils above the clay lens, according to the Record of
Decision. The contamination begins just below land surface and
extends down to the top of the clay lens. Below the clay lens
contamination levels are generally less than 100 ppm. However,
one "hot spot" of soils containing 1500 ppm was observed distant
from the major source area and below the clay. The primary
source area was considered to be beneath the storage area of the
manufacturing building located at the site. Accordingly, the
SITE demonstration was designed to focus on the periphery of the
primary zone of contamination.
Pilot Test Objectives
The objectives of the pilot test was to
vacuum extraction technology would effectively:
- remove VOCs from the vadose zone,
demonstrate the
81
-------
Vacuum Extraction Technology
- remove VOCs from various soil type (sands, and clays),
- correlate declining recovery rates and cleanup levels,
- correlate VOC concentrations in soils with extracted
vapors.
Several constraints were imposed that impacted the design
and operation of the vacuum extraction system. One constraint
was to not clean up the site, although it was important to demon-
strate how clean the process could decontaminate the soils. In
addition, the time frame for cleanup had to be relatively short
in order to minimize costs.
The size of the soil volume to be treated was difficult to
contend with since the in-situ treatment process did not lend
itself to segregating the subsurface treatment process to a small
test area. Evaluating the effects of the subsurface dynamics and
the objective of judging "How clean is clean?" was a challenge
for the short term, in-situ~demonstration.
The location of the vacuum extraction pilot test was select-
ed to minimize interference with the on-site manufacturing opera-
tions, to be near the edge of the contamination, and to be acces-
sible during the mid-treatment and post-treatment sampling.
Since the contamination was wide-spread but predominantly located
beneath a building at the demonstration site, it seemed plausible
to install a line of vacuum extraction wells to act as a "barri-
er", isolating the high level contamination from an extraction
well installed in the less contaminated area, thus, focusing the
evaluation on the effectiveness achieved in the low level concen-
tration area. The presumption was that the contaminant distribu-
tion was sufficiently defined to predict the movement of contami-
nants in the subsurface during the demonstration.
The final objective in the SITE demonstration for the vacuum
extraction technology was to evaluate "How clean is clean?"
Several test methods were used including TCLP, soil concentra-
tions (headspace and Contract Laboratory methods), shallow soil
gas surveys and monitoring subsurface contaminant vapor concen-
trations with time.
RESULTS
The subsurface contaminant plume within the soils in the
demonstration area was erratic and somewhat different from the
data available from the Remedial Investigation. The highest
concentration measured during the baseline sampling of the demon-
stration was within the presumed "clean" zone. The extraction
well from the presumed "cleaner" area recovered more VOCs than a
barrier well located closer to the primary source area. This
critical difference between actual and presumed conditions re-
quired a refocusing of the results with respect to the barrier
design and actual results achieved.
82
-------
Vacuum Extraction Technology
Recovery rates from vacuum extraction wells varied signifi-
cantly, ranging initially from 9 to 77 pounds/day. Higher recov-
ery rates were obtained from the shallower zone where the highest
soil concentrations were observed. Recovery rates decline witn
time except where the contaminants were extracted from the wells
near the primary source area.
During the SITE demonstration, Terra Vac removed about 1300
pounds of VOGs from the subsoils during the eight week pilot
test. These data were derived from flow and concentration meas-
urements taken during the vacuum extraction test. The results
were confirmed by analysis of the activated carbon that was used
to collect extracted VOCs.
Results of the demonstration indicated contaminant levels in
soils on the order of 10 ppm were reduced to non-detectable
levels, as depicted in Figure 2. In areas with higher contamina-
tion (i.e. greater than 100 ppm in soil) contaminant concentra-
tions in soils were reduced about 967. during the 8 week pilot
test.
In general, reductions in VOC concentrations throughout the
shallow soils at the site were observed. Figure 3 shows that a
substantial reduction (to non-detectable in many areas) in soil
concentration was achieved in both the clays and the silty sands
within the short treatment period. The average reduction in VOC
concentrations in clay soils where significant contamination was
present was reduced about 977.. Similarly, in sands concentra-
tions were reduced 897..
Shallow soil gas surveys before, at mid-point and after
treatment indicated consistent reductions in concentrations.
Figures 4 and 5 shows the dramatic reduction in concentrations
observed by this method.
Steady declines in extracted vapor concentrations were
observed in both the extraction wells and the monitoring wells
during the pilot test. The wells which weren't affected by the
primary source area showed the greatest reduction in vapor con-
centrations both in the shallow extraction zone and the deep
extraction zone. In general, higher extraction rates were ob-
served in the shallow soils compared to the deep zone.
Correlation between the soil concentration and the extracted
vapor concentration was difficult due to the heterogeneities in
soil concentrations at the beginning of the treatment process.
At other sites where complete vacuum extraction systems are
implemented, better correlations have been achieved.
Overall the Terra Vac Vacuum Extraction Process was suc-
cessfully demonstrated in the EPA SITE Program. Cumulative
extraction of VOC was 1300 pounds as depicted in Figure 6. As a
result of this successful demonstration Terra Vac is currently
continuing the cleanup for the owner of the manufacturing facili-
83
-------
Vacuum Extraction Technology
ty on the site.
Objectives
Terra Vac has applied Vacuum Extraction Technology at numer"-
ous sites to achieve various cleanup levels. Since cleanup
objectives in soils are generally site specific, several examples
of cleanup objectives and results of the vacuum extraction proc-
ess achieving these objectives are presented below.
Based on the data from the SITE demonstration it is estimat-
ed that full scale cleanup can be achieved to 50 ug/kg using
vacuum extraction within about one year of operation. The clean-
up objectives for TCE in the soils have been designated by EPA in
the ROD at 6.3 ug/kg. Extrapolating the level of cleanup from
the data from the SITE demonstration, complete cleanup would
probably occur within about two years.
Florida Department of Environmental Regulacion (FDER) has
defined a excess soil contamination as 500 ppm of hydrocarbons in
headspace of a sealed container of soil. The concentration
extracted from the wellhead may be considered essentially an
aggregate "headspace" concentration of hydrocarbons in the soils
around the well screen so that the 500 ppm hydrocarbon response
would represent the upper limit of a cleanup goal. Lower values
may be required at certain sites in order to protect groundwater
resources. Application of the Terra Vac Process at a site in
Florida demonstrated the capability of the process to meet the
FDER cleanup objective in about 6 months of operation, removing
over 20,000 pounds of hydrocarbons during the process.
An alternative criteria for considering the cleanup goal
would be based on a drinking water standard for indicator parame-
ters and correlate to an extracted vapor concentration. For
example, the lowest Maximum Contaminant Level in Florida is 1
ug/1 for benzene. Goals for cleanup of the vadose zone may be
considered using the concentration of benzene in the soil water
or pellicular water just above the water table and Henry's Law to
calculate the vapor concentration.
Based on a measured Henry's Law Constant for benzene, the
soil gas concentration in equilibrium with water at 1 ug/1 would
be about 0.18 ug/1 or 0.05 ppm benzene in air. Assuming the
extracted vapors from the wellheads are sufficiently close to
equilibrium conditions, an area within the radius of- influence of
the well may be considered clean if the concentration of extract-
ed benzene vapors is less than 0.05 ppm. This objective was also
met during the FDER demonstration after about six months of
vacuum extraction by Terra Vac.
At an industrial site in South Carolina the same concept was
applied to lower TCE concentrations in soil to below levels that
would impact groundwater quality. Terra Vac utilized Vacuum
Extraction Technology with the dual vacuum extraction approach to
84
-------
Vacuum Extraction Technology
clean up soils and groundwater. Initial soil concentrations were
about 300 ppm and reduced to less than 1 ug/kg. At less than 1
ug/kg in soil, the TCE would be below drinking water standards in
the soil water. Hence, the source of groundwater contamination
was eliminated by the vacuum extraction process.
At another Superfund site in Puerto Rico, Vacuum Extraction
Technology was applied to reduce carbon tetrachloride concentra-
tions in silty clay soils to less than 10 ug/kg. Initial concen-
trations were above 2000 ppm. The full scale cleanup took about
three years of operations to complete, removing over 100,000
pounds of VOCs from the subsoils.
Conclusion
Vacuum Extraction Technology has been successfully demon-
strated in the EPA SITE Program to effectively remove VOCs from
soils. Concentrations of TCE in soil were reduced by 897. in
sands and about 967. in clays for the eight week demonstration
period. Data from this site and the more than 70 sites where
Terra Vac has applied this technology clearly demonstrates that
the Vacuum Extraction Process can completely treat soils and
groundwater contaminated with VOCs.
85
-------
Manufacturing Building
LEGEND
EW- I • SOIL VAPOR EXTRACTION WELL
MW-5 D SOIL VAPOR MONITORING WELL
86
TERRA VAC
EPA SITE DEMONSTRATION
SITE PLAN
FIGURE NO. I
VACUUM EXTRACTION TECHNOLOGY
-------
TOTAL VOC's vs. TIME
VE1-S
2500
2000
00
--4
1500
PPM
1000
500
0
0
500_ltjr_ ,. JOOO
TIME (hours)
TERRA VAC VACUUM EXTRACTION TECHNOLOGY
EPA SITE DEMONSTRATION
1500
FIGURE 2
-------
CLEANUP PERFORMANCE
TCE CONCENTRATIONS in CLAY and SAND
Avg Init Cone and Avg Final Cone
400
CO
Clay
MW -3 and EW - 4
Final
Initial
Sand
TERRA VAC VACUUM EXTRACTION TECHNOLOGY
EPA SITE DEMONSTRATION
FIGURE 3
-------
PRETREATMENT SHALLOW SOIL- GAS CONCENTRATION
89
-------
6 230 -=
a
a
0
vnu3
VMUI2
VMW4
POSTTREATMENT SHALLOW SOIL GAS CONCENTRATION
TERRA VAC VACUUM EXTRACTION TECHNOLOGY
EPA SITE DEMONSTRATION
FIGURE 5
90
-------
CUMULATIVE POUNDS EXTRACTED
Pounds vs Time
1500
1000
Lbs
500
0
0
20
40
60
All Extraction Wells
DAYS
TERRA VAC VACUUM EXTRACTION TECHNOLOGY
EPA SITE DEMONSTRATION
FIGURE 6
-------
ULTROX
INTEHIMATIONAL
3435 South Anne Street
Santa Ana. California 927O4
TEL: [714] 545-5557
FAX: [714] 557-5336
UV/OXIDATION OF ORGANIC CONTAMINANTS
IN GROUND, WASTE, AND LEACHATE WATERS
BY: David B. Fletcher
Eriks Leitis
Due H. Nguyen
ULTROX INTERNATIONAL
2435 S. Anne Street
Santa Ana, CA 92704
Presented at the
1989 EPA Superfund Symposium
Atlanta, Georgia
June 19 - 21
92
-------
TABLE OF CONTENTS
INTRODUCTION '-
DESCRIPTION OF THE UV/OXIDATION PROCESS 2
ULTROX® UV/OXIDATION EQUIPMENT 2
APPLICATION OF UV/OXIDATION .3
CASE STUDY: EPA SITE PROGRAM - LORENTZ BARREL
AND DRUM SITE, SAN JOSE, CA 4
CASE STUDY: AUTOMOTIVE PARTS MANUFACTURER,
MICHIGAN 6
UV/OXIDATION TREATMENT AND OPERATING COSTS 7
SUMMARY *7
Table 1 Oxidation of Methylene Chloride & Methanol 9
Table 2 Industrial Effluents Treated with
UV/Oxidation 9
Table 3 Groundwaters Treated with UV/Oxidation 9
Table 4 Treatability and Design Study Results
Using Pilot Plants On-Site 10
Table 5 Full-Scale ULTROX® Systems . ... 11
Table 6 Direct Operating and Maintenance costs for
UV/Oxidation at Industrial Installations 12
Table 7 Typical Direct Operating & Maintenance Costs
Using UV/Oxidation for Water Supplies 13
Table 8 Typical Capital Costs for UV/Oxidation
Systems
Figure 1 Isometric View of System 15
Figure 2 System Flow Diagram • 16
Figure 3 Ultrox SITE Demonstration - TCE Removal -17
Figure 4 Ultrox SITE Demonstration - DCA Removal 18
Figure 5 Ultrox SITE Demonstration - TCA Removal . . . . • • 18
93
-------
INTRODUCTION
The removal of low levels of organic contaminants from groundwaters
and industrial wastewaters presents a challenge to environmental
professionals. Well-known and commonly used treatment processes
such as granular activated carbon (GAG) and air-stripping transfer
pollutants from one medium to another. With increasing public and
regulatory concern over the final fate of pollutants, such
transference technologies are not optimal.
Conventional chemical oxidation has been used in the treatment of
various waters polluted by organic chemicals for a number of years.
Potassium permanganate, chlorine and chlorine dioxide have been
used for treating organics such as phenol and its homologs in
wastewaters. Hydrogen peroxide with a catalyst such as ferrous
sulfate (Fenton's Reagent) has been used for oxidizing phenol and
other^ benzene derivatives. Processes utilizing iron catalyzed
peroxides and chlorine compounds are attractive in that they
utilize relatively low-cost treatment equipment. The disadvantages
of these processes are that they can attack only a limited number
of refractory organics, and they produce iron sludges or
chlorinated organics. Ozone alone has been used to treat phenolic
wastes, cyanides and certain pesticides. Ozone treatment is a very
clean process but is limited in the number of compounds which can
be treated. These oxidation processes have been used and are
continuing to be used in a number of situations.
The use of ultraviolet light catalyzed ozone plus hydrogen peroxide
(UV/oxidation) as a water treatment technique is rapidly expanding.
It offers a means of solving many of the problems created by the
toxic water soluble organic chemicals that are found today in
groundwater, wastewater, leachate and drinking water supplies
without many of the disadvantages of more conventional treatment
techniques.
UV/oxidation, when used as a stand-alone treatment process, or in
tandem with some of the above mentioned processes, can cost-
effectively destroy or render non-toxic the organic chemicals found
on the EPA's priority pollutant list.
This paper describes the experience of Ultrox International in
developing and applying the ULTROX® UV/oxidation process to the
full-scale treatment of organic chemicals in wastewaters, drinking
waters, leachates and groundwaters. The oxidants used in these
applications are ozone and hydrogen peroxide. Ultrox International
was issued a process patent in 1988 covering the application of UV
light, ozone and hydrogen peroxide to the treatment of a broad
range of organic compounds in water.
94
-------
DESCRIPTION OP THE UV/OXIDATION PROCESS
The ULTROX® process was developed over a 15 year period.
Ultraviolet light, when combined with O3 and/or H2O2 produces a
highly oxidative environment significantly more destructive than
that created with O3 or H2O2 by themselves or in combination.
UV light significantly enhances ozone or H2O2 reactivity by:
I.
II,
Transformation of O3 or H2O2 to highly reactive
(OH)" radicals
Excitation of the target organic solute to a higher
energy level
III. Initial attack of the target organic by UV light
The importance of the conversion of the ozone or H2O2 to (OH) "can be
more easily understood after studying the relative oxidation power
of oxidizing species. Hydroxyl radicals have significantly higher
oxidation power than either hydrogen peroxide or ozone.
Species
Fluorine
Hydroxyl Radical (OH)
Atomic Oxygen
Ozone
Chlorine Dioxide
Hydrogen Peroxide
Perhydroxyl Radicals
Hypochlorous Acid
Chlorine
Oxidation
Potential
Volts
3.06
2.80
2.42
2.07
1.96
1.77
1.70
1.49
1.36
Relative
Oxidation
Power*
2.25
2.05
1.78
1.52
1,
1,
,44
,30
1.25
1.10
1.00
* based on chlorine as reference (= 1.00)
The effect of UV enhanced oxidation is illustrated in Table 1.
ULTROX® UV/OXIDATION EQUIPMENT
ULTROX® UV/oxidation equipment treatment systems: (1) have very
few moving parts, (2) operate at low pressure, (3) require a
minimum of maintenance, (4) operate full-time or intermittently in
either a continuous or batch treatment mode, (5) utilize efficient,
low temperature, long life UV lamps, and (6) can employ the use of
a micro-processor to control and automate the treatment process.
The ULTROX® UV/oxidation system consists of a UV/oxidation reactor
and an oxidation source — either an ozone generator with an air
preparation system and/or a hydrogen peroxide feed system. Figures
95
-------
1 and 2 show an isometric assembly view and drawing of a Model F-
650 system, which accommodates flow rates up to 60 gpm or batches
of 650 gallons.
The reactor is fabricated from stainless steel. The UV lamps are
enclosed within quartz tubes for easy replacement and are mounted
vertically within the reactor. Depending upon size of the reactor
and the type of water to be treated, the reactor can have 4 to 8
stages. Lamps are installed either in all stages or in designated
stages, depending upon the type of treatment specified. When
ozone is used as the oxidant, it is introduced at the base of the
stage. The ozone is dispersed through porous stainless steel
diffusers. The number of diffusers needed will depend upon the
type of organics being oxidized and the degree of removal required.
Ozone< is produced from either compressed air, dried to a -60°F
dewpoint by desiccant columns, or produced from cryogenic oxygen.
Up to 2% wt. ozone is generated from air, and up to 5% wt. ozone
can be produced economically from oxygen.
When hydrogen peroxide is used in the process, it is directly
metered into the influent line to the reactor.
Within_ the reactor, the water flows from stage to stage in a
sinusoidal path using gravity flow. When the reactor uses ozone,
the residual ozone in the off-gas is decomposed back to oxygen by
the use of a fixed-bed catalytic unit operating at 150°F. The air
is then vented to the atmosphere.
Ozone generators with varying capacities are used with the Model
F-650 reactor. The size of generator depends upon the ozone dosage
requirements. Present installations use 28 to 140 Ib. per day
capacities.
APPLICATION OP UV/OXIDATION
UV/oxidation equipment developed by Ultrox in recent years has been
used to treat a wide variety of waste streams. Tables 2 and 3 list
toxic compounds found in groundwaters and wastewaters that have
been successfully treated with the ULTROX® UV/oxidation System.
Specific case histories of treatability and design studies for
private industries and military installations are presented in
Table 4. In each of these cases, pilot treatment plants were
operated on site to develop treatment design and economic data.
Contaminates oxidized included: pesticides, petroleum compounds,
munitions, and chlorinated solvents.
Table 5 illustrates projects where the treatability and design
studies were converted into permanent on-site UV-oxidation
installations. The systems to date treat either industrial
groundwaters, wastewaters and process waters. Contaminants in
96
-------
these waters include phenols, chlorinated solvents, hydrazine,
dimethylnitrosamine, tetrahydrofuran and formaldehyde. Commercial
systems have been designed, built and installed to treat flows
varying from 10,000 gallons to 300,000 gallons per day. A system
to treat 1.3 million gallons per day is under construction.
Standard equipment designs are used in all of these installations.
Reactor size varies from 300 gallons to 4,300 gallons. Ozone
generators range from 21 Ib. to 140 Ib per day. In several cases
hydrogen peroxide is used in place of or with ozone.
Treatability studies are carried out first in the laboratory using
glassware equipment to determine the feasibility of treating the
water with UV/O3, UV/H2O2, or UV/O3/H2O2.
If the results are encouraging, the next step in the study involves
the installation of a skid-mounted pilot plant on site. Sufficient
design and economic data normally are collected within 2 weeks.
Specifications for the full-scale system are then prepared.
Standard reactors, ozone generators and hydrogen peroxide feed
systems are utilized. Systems are assembled and tested at our
facilities and then shipped to the job site. The systems are then
installed, checked out, and turned over to the customer. Full
service maintenance contracts are available.
Full-scale systems, in most cases, are automated using
microprocessor control. The system usually requires periodic
monitoring (once per shift or once per day). The systems are
designed to operate in a batch or continuous mode depending upon
treatment requirements.
In a number of cases, UV-oxidation is used as a part of a treatment
train. For example, at wood treating sites prior to the UV-
oxidation treatment, the wastewater or groundwater requires
breaking of oil/water emulsions and removal of suspended matter,
as well as adjustment of pH.
CASE STUDY:
EPA SITE PROGRAM - LORENTZ BARREL AND DRUM SITE,
SAN JOSE, CA
The EPA has established a formal program to accelerate the
development, demonstration, and use of new or innovative
technologies to be used in site cleanups. This program, called the
Superfund Innovative Technology Evaluation (SITE) program, has four
goals:
• To identify and, where possible, remove impediments to
the development and commercial use of alternative
technologies.
e To conduct a demonstration program of the more promising
innovative technologies for the purpose of establishing
97
-------
reliable performance and cost information for
characterization and cleanup decision-making.
site
• To develop procedures and policies that encourage
selection of available alternative treatment remedies at
Superfund sites.
• To structure a development program that nurtures emerging
technologies.
Each year, EPA solicits proposals to demonstrate innovative
technologies. To identify the best available technologies, an
extensive solicitation is necessary. A screening and selection
process follows, based on four factors: (1) the technology's
capability to treat Superfund wastes, (2) the technology's
performance and cost expectations, (3) the technology's readiness
and applicability to full-scale demonstrations, and (4) the
developer's capability and approach to testing.
In the third year of the SITE program, Ultrox was selected to
demonstrate their UV/oxidation technology. The Lorentz Barrel and
Drum Superfund site in San Jose, California was selected for the
demonstration.
The Lorentz site was used for drum recycling for nearly 40 years.
Over this period of time, the site received drums from over 800
private companies, military bases, research laboratories, and
county agencies in California and Nevada. Drums arrived at the
site containing residual aqueous wastes, organic solvents, acids,
metal oxides, and oils.
Since 1968, there have been several regulatory actions at the
Lorentz site. In 1987, the Lorentz facility ceased operation and
the EPA< assumed lead agency responsibility for the site
remediation. Investigations revealed that the groundwater beneath
the site was contaminated with a number of chlorinated solvents,
chlordane, toxaphene and PCBs.
An^ULTROX® P-150 pilot plant was moved in on February 21, 1989.
Thirteen (13) tests were conducted between February 24 and March
9, 1989, on extracted groundwater from the site. During the
treatability bench studies, TCE, TCA, and DCA were chosen to
monitor the progress of the pilot.
Figures 3, 4 and 5 illustrate analyses of the waste stream'influent
and the treated effluent for the targeted contaminants.
The final report has not yet been issued by the EPA. However,
based on the preliminary results, the ULTROX® UV/oxidation process
was successful in the reduction of all of the VOCs present in the
groundwater at the Lorentz site to below drinking water standards.
98
-------
The bicarbonate level of the groundwater was extremely high (1200
mg/1). Because of this, treatment costs are higher than what would
be experienced in more normal groundwater applications. Based on
the conditions tested at the site, treatment .costs were estimated
to be as follows:
Flow Rate:
Influent Concentration:
Effluent Concentration:
Treatment Costs:
Ozone (@ 0.06/KWH)
H202 (@ $0.75/lb)
UV (incl. power
and annual lamp
replacement
100 gpm
250-1000 MgA VOCs,
pesticides, PCBs
Mg/1
$/1000 gallons
$ 0.370
0.156
0.836
O & M Cost 1..36
Capital Amortization
(16%/year) 0.75
Total Treatment Cost: $ 2.11/1000 gallons
CASE STUDY: AUTOMOTIVE PARTS MANUFACTURER, MICHIGAN
Testing of water beneath a Michigan automotive parts manufacturer
revealed significant VOC contamination. TCE levels of 5,000 to
10,000 Mg/1 were recorded as well as trace levels of other
chlorinated solvents. The Michigan Department of Natural Resources
required that the manufacturer pump and -treat the groundwater.
The manufacturer investigated air stripping with GAG off-gas
treatment, aqueous phase GAG and ULTROX® UV/oxidation as possible
treatment alternatives. Bench scale studies were conducted at a
GAG supplier and at Ultrox's laboratory. While all treatment
techniques could provide the required removal levels, UV/oxidation
was the most economical. An ULTROX® P-75 pilot scale treatment
system was delivered to the site. Testing over a two week period
confirmed the data obtained in the laboratory. An ULTROX® F-3900
treatment system was ordered and installed in May, 1989. The
system is currently operating and achieving the following results,
which exceed Michigan requirements:
Flow Rate:
Influent Concentration:
Effluent Concentration:
210 gpm
5500 Mg/1 TCE
1 Mg/1 TCE
99
-------
Treatment Costs:
Ozone (@ 0.06/KWH)
H202 (@ $0.75/lb)
UV (incl. power
and annual lamp
replacement
$71000 gallons
$ 0.119
0.188
O & M Cost 0.44
Capital Amortization
(16%/year) 0.29
Total Treatment Cost: $ 0.73/1000 gallons
UV/OXIDATION TREATMENT AND OPERATING COSTS
Table 6 represents the direct operating and maintenance costs for
treatment of contaminants in groundwater at water utility sites.
The costs are based upon pilot plant studies at four different
sites in Southern California. At three of the sites,
perchloroethylene (PCE) and trichloroethylene (TCE) were the
contaminants with concentrations ranging from 20 ppb to 200 ppb.
Table 7 presents the actual costs of treating by UV/oxidation
wastewater and groundwaters at various permanent industrial
installations. Some of these costs are in the cents per thousand
gallon range and others in cents per gallon range.
In the case of the hydrazines, a small volume of water is treated
per day on a batch basis and a comparatively long reaction time is
needed. UV/oxidation was found to be the most cost-effective
method of destroying the three types of hydrazines and the
nitrosamine which is formed as a by-product by the oxidation. The
UV/oxidation system replaced a chlorination unit, which produced
chlorinated organic by-products.
The price range of UV/oxidation equipment at various installations
cited in Table 8 varies from $45,000 to $300,000 (uninstalled).
Pricing depends upon the oxidant requirement - whether ozone or
hydrogen peroxide is used, the chemical structure of the organic
compounds treated, the number of UV lamps required, and the
retention time required to achieve an acceptable discharge
standard.
SUMMARY
Over the last 15 years, UV/oxidation has progressed from research
and development to commercial operation. During these years,
Ultrox has advanced its design through applied bench testing, pilot
studies,^ and full-scale systems that remove contaminants from a
wide variety of wastewaters and groundwaters.
100
-------
UV/oxidation technology is not suitable for every organic
contamination problem. It can, however, effectively address a wide
range of clean-up needs. This form of on-site chemical oxidation
can offer real advantages over conventional treatment techniques
and should be considered when evaluating water treatment
alternatives.
101
-------
TIME
fMIN.l
0
15
25
TABLE 1
OXIDATION OF METHYLENE CHLORIDE
CONTROL
100
100
100
UV
UV/H202
100 100
59 46
42 17
O-3/H2°2 UV/O3
100
32
21
100
36-
16
100
19
7.6
TIME
(MIN.)
0
30
CONTROL
75
75
OXIDATION OF METHANOL
CONCENTRATIONS = mg/1
UV
75
75
UV/H202
75
75
UV/0,
75
31
UV/03/H202
75
1.2
TABLE 2
Industrial effluents containing:
Amines
Analine
Benzene
Chlorinated Solvents
Chlorobenzene
Complex Cyanides
Creosote
Hydrazine Compounds
Isopropanol
MEK
MIBK
Methylene Chloride
PCB' s
Pentachlorophenols
Pesticides
Phenol
RDX
-TNT
Toluene
Xylene
Polynitrophenols
TABLE 3
Groundwaters containing:
BTX
Creosote
1,2 DCA
DCEE
Dioxins
Dioxanes
Freon 113
MeCl2
MIBK
PCBs
PCE
Pentachlorophenol
bis (2-chloroethyl) ether
Pesticides
Polynuclear Aromatics
1,1,1 TCA
TCE
THF
Vinyl Chloride
Triglycol dichloride ether
102
-------
TABLE 4
TREATABILITY AND DESIGN STUDY RESULTS USING PILOT PLANTS ON SITE
o
CO
Customer
Bulk Chemical
Transfer Depot
Municipal Water
Producers
Aerospace Co.
Chemical Co.
Automotive Co.
Electronics Co.
Munition Plants
Army Bases
Petrochemical
Mfr.
Semiconductor
Mfr.
Application
Contaminated
groundwater
Contaminated drinking
water supply
Paint stripping
wastewater
Wastewater
Groundwater
Wastewater/runoff
water, groundwater
Wastewater
Contaminated
groundwater
Semiconductor Co. Wastewater
Wastewater
Groundwater
Contaminants
TCE, PCE, methylene
chloride
TCE, PCE, color
Methylene chloride
Misc. pesticides
(including DBCP)
TCE, MeCl2
PCBs, ViCl, DCA
+ other VOCs
TNT, RDX
DIMP, DBCD, VOCs
EDTA
Benzene
benzene, toluene,
xylene, ethyl benzene
Results
Water treated and.
reinjected into ground
VOCs and color reduced to
below state action levels
MeCl2 reduced from 4000 ppm
to less than 100 ppb
DBCP and other pesticides
reduced to less than 1 ppb
Reduced 10 ppm to 5.0 ppb
Reduced PCBs to less than
1 ppb; VOCs reduced to
below state action levels
TNT & RDX reduced from
100 ppm to less than 1 ppm
DIMP & DBCD reduced to
less than 10 ppb; VOCs
reduced to below state
action levels
Reduced EDTA from 6,000
ppm to 100 ppm (acceptable
discharge standard)
Reduced benzene from 10 ppm
50 ppb
Reduced contaminants from
14.0 ppm to 4.0 ppb
-------
TABLE 5
FULL-SCALE ULTROX® SYSTEMS
Customer
Wood Treating
Plants (2)
Closed Wood
Treating Plant
Chemical Plant
Automotive
Foundry
Aerospace Co.
Chemical Plant
Semiconductor Co.
Application
Wood treating
wastewater
Contaminated
groundwater
Fume scrubbing
water
Contaminated
groundwater
Contaminated
groundwater
Wastewater
Contaminated
groundwater
Contaminants
Phenol, pentachloro-
phenol
Phenol, pentachloro-
phenol
Hydra z ine, monomethy1
hydrazine, unsym-
metrical dimethyl
hydrazine
TCE, Trans 1,2-DCE
TCE, TCA, DCA, PCE
MeCl2, vinyl chloride
Phenol, formaldehyde
THF
Results
Water treated and
discharged to POTW
Water treated and
discharged to POTW
Destroyed parent
compounds to N.D.
levels and dimethyl
nitrosamine below
10 ppb
Water treated and
discharged to lake
Water treated and
discharged to POTW
Water treated and
discharged to POTW
Replaced a GAG system to
reduce THF from 1,000 ppb
to less than 5 ppb
-------
TABLE 6
TYPICAL DIRECT OPERATING & MAINTENANCE COSTS
USING UV-OXIDATION FOR WATER SUPPLIES*
o
en
Type of
Water
Contaminated
potable drinking
groundwater
Contaminated
potable drinking
groundwater
Contaminated
potable drinking
groundwater
Contaminants
TCE, PCE
Total Contaminant
Concentration
Treatment
Standards
TCE, PCE 200 ppb
Color
Direct 0 &. M
Cost Range
rS/1000
less than 20 ppb Drinking Water 0.10 to 0.20
Drinking Water 0.20 to 0.30
70 color units Drinking Water 0.10 to 0.15
* Assumes cost of electrical energy is $0.06/KWH
-------
TABLE 7
DIRECT OPERATING & MAINTENANCE COSTS FOR UV-OXIDATION
AT INDUSTRIAL INSTALLATIONS
Type of Water Contaminants
Wood Treating
Wastewater
Wood Treating
Groundwater
Fume Scrubber
Water
Contaminated
Groundwater
Contaminated
Groundwater
Contaminated
Groundwater
Pentachloro-
phenol and
phenol
Pentachloro-
phenol and
phenol
Hydrazene,
Monomethyl-
hydrazine
Unsymmetrical-
dimethyl-
hydrazine
TCE, trans DCE
MeCl2
TCE, TCA, DCA,
PCE, MeCl2 ViCl
THF
Volume
Contaminant Discharge Treated Direct O & M
to
Con centrat i on
150 ppm POTW
5 ppm
5,000 ppm
5 ppm
600 ppb
1 ppm
POTW
Biotreat- 600 -
ment Plant 1500
On-Site
Per Day Cost Range
30,000 $1.25-1.35/1000 gal
86,400 $0.90-$1.00/1000 gal
$0.086/gal
Surface 300,000 $0.47/1000 gal
Water
POTW 72,000 $0.33/1000 gal
Ground 216,000 $0.39/1000 gal
Wastewater
Phenol
90 ppm
POTW
4,:300 $6.48/1000 gal
-------
TABLE 8
TYPICAL CAPITAL COSTS FOR UV-OXIDATION SYSTEMS
Type of Water
Wood Treating
Wastewater
Wood Treating
Groundwater
Fume
Scrubber
Water
Contaminated
Groundwater
Contaminated
Groundwater
Contaminated
Groundwater
Wastewater
Contaminants
Pentachloro-
phenol
Phenol
Pentachloro-
phenol
Phenol
Hydrazines
TCE, trans DCE
MeCl2
TCE, TCA, DCA
PCE, MeCl,,
ViCl
THF
Phenol
Total
Contaminant
Concentration
150 ppm
5 ppm
5,000 ppm
5 ppm
600 ppb
1 ppm
90 ppm
Water
Flow Price Range
Rate - funinstalled)
(GPD) $
30,000 125,000-150,000
86,400 175,000-200,000
600-1500 125,000-150,000
300,000 225,000-275,000
72,000 130,000-150,000
216,000 250,000-300,000
4,300 45,000-55,000
-------
Cu
CotuS/Bc QxcfM D«cs
-------
Rutomiuf
topical)
HeedtaVoh*
tow)
FramShafln
Daund-WaUr
Vtb
Diluent
FIGURE 2
ULTROX SYSTEM
FLOW DIAGRAM
B&Altfl; it/yes IHEVBB; 02/02/93 | PiOtJ3ff
Hntogoa
FMTonk
16
-------
FIGURE 3
ULTROX SITE DEMONSTRATION
TCE REMOVAL
INF.
B
TEST NUMBER
-------
FIGURE 4
ULTROX SITE DEMONSTRATION
TCA REMOVAL
TEST NUMBER
FTflllRH 5
ULTROX SITE DEMONSTRATION
DC A REMOVAL
.5
TEST.NUMBER
-------
In Situ Steam/Air Stripping
Phillip La Mori and Jeff Guenther
Toxic Treatments (USA) Inc.
Toxic Treatments is currently remediating a site in San
Pedro, California, using technology developed and patented by
Frank Manchak (U.S. Patent No. 4,776,409.) This process removes
volatile organic compounds (VOC's) from contaminated soils by
injection of steam and air.
The steam and air are injected into the ground by means of a
pair of hollow kelly bars. The kellys distribute the air and
steam within the soil through rotating mixing blades five feet in
diameter. The volatiles are evaporated from the soil matrix,
into the remediation air. Effluent gases move up beside the
drilling shafts to the surface, where they are collected in a
metal shroud.
The shroud runs under a slight vacuum. A blower mounted on
a separate process chassis extracts the air and vapors, along
with a small amount of dust, from the shroud and directs them to
a process train where contaminants are removed and collected for
recycling or disposal. The treated air is recompressed and
reinjected into the soil.
The shroud effluent gas is analyzed continuously during
remediation. Instruments on board the process train include a
gas chromatograph, two flame ionization detectors, and pressure,
temperature, and humidity measuring equipment.
The apparatus is moved around the site by a heavily modified
Caterpillar 583 pipe layer. The drill stems will reach more than
thirty feet below the surface. Drilling rates of 3 feet per
minute are possible, depending on the nature of the soil and
contaminant type and concentration.
The San Pedro site must be remediated to remove residual
VOC's, including chlorinated compounds. There are also semi-
volatile hydrocarbons (SVH's) in this soil. These SVH's do not
present much, if any, environmental hazard because their
migration rate is negligible. You wouldn't want to eat this
soil, but the SVH's do not evaporate and were not expected to be
significantly affected by the remediation process because of
their low vapor pressures.
Initial VOC concentrations were between 824 and 1872 ppm.
The target remediation level was 100 ppm of VOC's.
Test results to date show that the VOC's are effectively
remediated. The range of reduction of VOC's is from 96 to 99
percent, using EPA 8240 analysis methods. There are about 15
volatile compounds in this soil. We were able to reduce their
total concentration to less than 55 ppm, well below the target
level of 100 ppm.
Silty soils remediate more easily than clay. Moist clay
112
-------
soils containing tetrachloroethylen were also remediated by
better than 95 percent, but the final levels were 53 to 203 ppm,
somewhat higher than target levels. This may be adequate, but
changes have been made in the process which improved the absolute
remediation level..
The process also removed most of the semivolatiles as listed
in EPA 8270. This was somewhat unexpected and we have spent
considerable time trying to determine the mechanism for their
removal. About 85 percent of the SVC's found on the site are
phthalate esters of various aliphatic radicals, e.g., ethylhexyl
phthalate. We believe that these esters are being hydrolyzed to
an intermediate acid salt and an alcohol. The alcohol appears to
dehydrate to an olefin. Soil and recovered fluid analyses
support these ideas. Some SVC's like isophorone and phenol may
be quantitatively removed by steam distillation or vaporization.
The recovered VOC's represent most of the starting
calculated mass. Field monitoring of gases outside the shroud
and analyses of adjacent blocks show little or no escape of VOC's
to the surroundings. We believe that the process is very
effective for volatile materials. The process also is somewhat
effective for semi-volatiles, but this depends on the species
present.
Treatment rates vary depending on the amount of contaminants
and type of soil, as indicated above, but the average is about 6
to 8 cubic yards per hour.
The Detoxifier (tra) apparatus may also be used for other
types of in situ treatment, including the addition of soil
stabilizers, injection of reagents, or for bioremediation. Two
more complete drilling units have been assembled, and process
trains for those units will be designed to match the needs of
particular applications.
113
-------
AVERAGE VOC CONCENTRATION
(ppm)
Area
A
B
D
Be fora
1,114
1,353
3,954
After
12
30
139
AVERAGE SVH CONCENTRATION
(ppm)
Area Before After
A 3,775 627
B 12,116 1,766
D 1,014 85
VOC CONCENTRATIONS (ppm)
Block
A-8-Q
A-9-g
A- 10-g
B-50-n
B-61-n
B-61-m
B-62-m
D-92-b
0-03 -b
D-94-b
Balflifl
1,149
824
1,368
1,123
1,500
1.872
917
2.305
3.720
6.838
Attar
18
7
11
23
13
65
29
63
163
203
98
99
99
98
99
97
97
98
96
97
SVH CONCENTRATIONS (ppm)
Block
A-8-g
A-9-g
A- 10-g
B-60-n
B-61-n
B-61-m
B-62-m
D-U2-b
D-90-b
D-94-b
Batata
1,794
2.510
7,020
22.829
14.924
10,040
669
707
1,486
869
Aftec
637
863
592
1.670
2,304
2,496
694
55
90
111
64
74
92
63
86
76
11
92
94
87
-------
FATE OF SEMI VOLATILE COMPOUNDS
phthalate salts + C8 to CIS olefins
- H20
phthalate salts + C8 to CIS alcohols
1032 Ibs. of bis-2-ethylhexyl phthalate
produce 296 Iba. of ~ "
VOC MASS BALANCE
JrflHtment Slocks
Pflmealated
6
10
Pounds Bounds * voc
Recovered Removed Recovered
124.5
264.8
142.5
296.6
87.4
89.3
FATE OF SEMI-VOLATILE COMPOUNDS
Pretreatment Psrcent Recovered Percent
Amount (Ibs)
bls-2-ethylhexyi
phthalate
butyl
cellosolve
Isophorone
2 ethylhexyl
adlpate
phthalate matrix
glycol ethers
C8-C15 HC
others
Total
424.4 38
13.1 1
3.8 <1
2.0 <1
607.8 55
11.8 1
34.1 3
7.5 <1
1104.5 100 t
10 Block
Amount (ibsl*
0
9
28
0
0
<1
393
42
473
Total
0
2
6
0
'•"" 0
0
83
9
100
Project
THC CONCENTRATIONS (ppm)
Slock
A-8-0
A-9-g
A-10-g
B-6O-n
B-51-n
B-51-m
8-52 -m
D-92-b
D-93-b
D-94-b
Befors
2,943
3.334
8,388
23.952
16,424
11,912
1,686
3,012
5.184
6,707
After
655
660
603
1.693
2,317
2,550
623
108
253
3',4
% Reduction
78
80
93
93 .
86
79
61
96
95
95
-------
THE PIEPHO CORPQRATTOTXT
9341 Cornwell Fanu Rd. Great Palls, Va. 22066, (703) 759-7074
INTRODUCTION
fn AbwassefTtechnik GMBH has been in the waste water
treatment business in West Germany for ten years. We are located in
Bredenbeck, a small village just outside of the city of Hannover
Over the years, our customers have ranged in size from extremely'
large concerns, such as the German Navy and VW to small, single
factory furniture and textile companies. Our latest project is
landfi11 for the government
Our firm is relatively small. We only have about 25 employees
RheineBrauTr *" J^S *?* ***** "* h*™ dev^°Ped a partnership with
Rhexn Braun, one of the largest energy firms in Germany. This
partnership, called Union Piepho, gives us the facilities and
resources to handle jobs of any size. We have also developed partners
in a number of other countries including Italy, Denmark, India, and
the United States.
PROCESS DESCRIPTION
We have developed a chemical-physical process for the treatment of
waste water. This process consists of Reaction Separation Agents NT75
and Piepho treatment machines.
Reaction Separation Agent NT75 ,
NT75 is a separating reactant on a clay basis which by virtue of its
extraordinarily high effectiveness can crack emulsions, adsorb
contaminants, flocculate and finally encapsulate, all in one process
stage.^Furthermore, once the contaminants have been encapsulated the
resulting sludge will no longer leach. Due to this characteristic, we
have been able to dispose of much of our sludge on standard household
dumps, thus reducing the total treatment costs dramatically.
The clays in Reaction Separation Agent NT 75 are minerals with an
expandable triple layer arrangement and a cation exchange capacity of
80 - 100 m/lOOg. The grain size distribution ranges from 0.5 - 1.5.
The use of bentonite as the main flocculant is unusual because the
particles are too small to give good adsorption. On the other hand
the negative charge on its surface causes it to be attracted to
cationic positive polymers. When the polymers have bonded all oil
molecules the polymer oil complexes created become positively charged
and coated by particles of clay. The bentonite then exchanges ions
for polymer oil complexes.
116
-------
THE PIEPHO CORPORATION
9341 Cornwell Farm Rd. Great Falls, Va. 22066, (703) 759-7074
The release of clay particles from this bond is virtually excluded. A
polymer oil complex coated by clay particles is practically cocooned
- it is fixed and the reverse reaction or breakdown is not possible,
even if the pH value changes, thus giving us our stable sludge. Also,
the stabilization and encapsulization process of the sludge doesn't
stop but continues even after dumping and as such increases the
resistivity to leaching.
One must bear in mind that the interaction between polymers and clay
particles starts immediately after input of the additive into the
waste water, and continues during the adsorption process. The
slightly reduced effectiveness of the polymers is balanced by an
increase in flocculaton, typical for clay additives.
This possibility, to control polymer oil complexes at the point of
their reaction and formation, is created by combining all chemicals
required in the water treatment into one formula. The prerequisite
for this is to prepare (isolate) those components capable of changing
the pH value such that they become effective at the right point in
time and in the right sequence - so raising or lowering the pH value.
The use of specially treated organic acids and alkalis with different
solubilities allows the pH value to be controlled and determined
throughout the process.
A condition of the treatment with reaction separation agent NT75 is
that the fine-grain powdered product is well mixed into the waste
water. During this mixing process the polymers are distributed, the
acids and alkalis dissolve, polymer oil complexes are formed - which
are immediately cocooned by the clay particles and so encapsulated.
This process is repeated continuously during the reaction until all
oil and other contaminants are fixed. The mixing may then be
terminated and conventional sedimentation and drying of the sludge
follows. The time period required in the treatment module of the
plant depends on the waste water - but it is usually about 5 minutes;
the same applies to the sedimentation phase.
Piepho Wastewater Treatment Machines
The Piepho treatment machines perform 5 basic functions.
1) Automatic addition of Reaction Separation Agent NT75
2) High speed mixing of the NT75 into the waste water
3) Sedimentation
4) Filtration
5) Drying,removal of the sludge
117
-------
THE PIEPHO CORPORATION
9341 Comwell Farm Rd. Great Falls, Va. 22066, (703) 759-7074
The machines have been designed to be fully automatic, but can also
be run manually. Many of our machines process in batches, while
?romrJo£a?/£1Ja%n°nnin?^Sly* *** capacities of the machines range
from 400 1/h to 30,000 1/h, and because they are fully automatic,
they can be run in excess of 20 hours a day, thus allowing our
largest machine to clean 600,000 liters in a single day. Also the
machines have been designed to fit into containers which al?ow for
easy transportation and single' day setup
TYPE OF CONTAMINANTS TREATED BY THE PIEPHO SYSTEM
In the past, using the Piepho process and other processes which can
be integrated into our system (reverse osmosis, biological treatment,
etc.), we have been able to handle all pollutants we have come 'across
in water. The Piepho process itself is most effective at removing
water-insoluble pollutants, both organic and inorganic. It has ^
been effective at removing a variety of dissolved organics and :
inorganics (through the addition of both our patented reaction •
separation agent NT 75 and active carbon), and heavy metals.
In Europe, we h'ave concentrated on the handling of industrial waste
water, polluted groundwater, and landfill leachate. Most frequently,
these applications lead us to a waste water contaminated with oil
bearing emulsions, chlorinated hydrocarbons, heavy metals, and paint
remainders.
We feel that our system has the capacity to clean any waste water.
Whether or not our process is economically efficient in its treatment
of a specific waste water must be tested in our laboratory. For some
industrial firms who generate small quantities of very complicated
waste water, it may still be more cost efficient to have their water
hauled away to a large waste water treatment plant.
•
QUALITY OF EFFLUENT ACHIEVED WITH THE PIEPHO SYSTEM
Our process is very flexible with respect to the achievement of
desired clean-up levels. The quality of the treated water can be
increased through the addition of chemicals, or through the choice of
processes to be used in the treatment system. The best way of
illustrating this is with examples.
118
-------
THE PIEPHO CORPORATION
9341 Cornwall Farm Rd. Great Falls, Va. 22066, (703) 759-7074
Example 1
For the continuous circulating process water in the -finishing
industry, it only makes sense to clean the water to the level where
it meets the quality demands for its continued use. In cases such as
these, single stage systems are usually all that is necessary.
Example 2
If the treated waste water will be removed to a community
biological treatment plant, then it must only be cleaned to the level
of the inlet values for the plant. In most cases, this can also be
handled by a single stage process.
Example 3 .
If higher quality is needed, for example, if the water will be
discharged directly into surface waters, or if total quality levels
or single parameters correspond to those of drinking water, our
process can also be used to meet these demands. The cost would
increase (depending, of course, on the quality of the waste water
before treatment), and many stages may be needed, including:
Pretreatment
(1) Prescreening to remove suspended solids
(2) Pretreatment to remove lower density materials (If the water
contains more than 2% oil,fats,etc., this pretreatment will
make the next stages more efficient)
Piepho Treatment
(1) First stage -.addition of NT 75, a fatty acid, bentonite
based coagulent/flocculent.
(2) Second stage - addition of active carbon
(3) Third stage - addition of Fuller's earth mixture
(4) Fourth stage - removal of 2,3 above by NT 75
Further treatment
If necessary, further stages can be integrated into the Piepho
system. These would include extra filters, reverse osmosis,
membrane filtration, etc.
To this date, we have been able to meet all quality demands that have
confronted us.
LIMITATIONS OF THE PROCESS
The limitations of our process are accounted for by our pretreatment
stages. Our process is not effective if the water contains solids and
coarse materials. Therefore, these must be removed as much as is
mechanically possible in the pretreatment. Also, it is not economical
119
-------
' CORPORATION
9341 Comwell Farm Ed. Great Falls, Va. 22066, (703) 759-7074
to use NT75 to treat low density material, such as oil and grease
Because of this, we have developed our own pretreatment'to re-over
these materials. Our process is especially good at dealing with high
concentrations of contaminants. However, if the concentrations are
too nigh (above 2%), then the water must be diluted. The limitation
here is mechanical. The machines are not designed to handle large
amounts of sludge. Mr. Piepho's suggestion in cases where the
concentration is too high is to use the water which has already been
treated to dilute the remaining waste water. In this way only a
minimal amount of fresh water must be used.
COST i
General statements about the economic side of our process are '••
difficult to make. Exact costs for the system will depend on the type
of contaminants in the waste water to be cleaned, the concentration
of these contaminants, and the quality level of cleaning which is
desired. The fixed costs, i.e. the cost of the machine(s) can range
from 21,000 DM for a single stage machine with a capacity of 300
liters per hour, to 750,000 DM for a multi-stage system with a
capacity of 25,000 1/h. The treatment costs are even harder to
generalize. They range from 0.3-3 cents per gallon if only a single
stage treatment is required, to 2 - lOc/g for water requiring
multi-stage treatment. However, most waste waters require in the
range of 2 - 5 c/g for cleaning. Examples may be helpful:
Single Stage Treatment \
Single stage treatment consists of a single Peipho machine, with
capacities ranging from 300 1/h to 25,000 1/h, and the appropriate
mixture of chemicals. This is researched by our laboratory for the
particular water to be cleaned. The water is pumped into the machine
where the chemicals are added and mixed with the water at high
speeds, allowing coagulation/flocculation to take place. This water
is then moved into the sedimentation tank where the sludge is removed
automatically by a filter band. The water must pass through a filter
before it is discharged. The result is a treated water and a sludge
which has encapsulated the contaminants and which, according to the
"Wehrwissenschaftliches Institut fur Material Untersuchung" (State
Scientific Institute for Material Study), was not susceptible to
leaching over the term of a one year study using distilled water and
a sludge contaminated with hydrocarbons. ',
120
-------
THE PIEPHO CORPORATION
9341 Cornwall Farm Rd. Great FaUs, Va. 22066, (703) 759-7074
Costs: The costs for a single stage system are as follows.
Machine: 21,000 - 300,000 DM
Chemicals - Generally 0.5 - 3.5 kg are used per 1000 liters.
1 kg costs 3 DM
Treatment costs are therefore 1.5 - 10.5 DM per
1000 liters or 0.3 to 2 cents per gallon.
Other costs such as electricity, labor, and
filter paper are insignificant.
Multi-stage Treatment
The full treatment offered would consist of the following.
1) Pretreatment - Skimming off of low density materials.
Screening of solids.
2) Stage 1 - Same as a one stage treatment. Just a single
machine and the addition of NT 75.
3) Stage 2 - A tank and a mixing turbine for the addition of
activated carbon.
4} Stage 3 - The same setup as stage 3, only here it is for the
addition of a special Fuller's earth mixture.
5) Stage 4 - A machine exactly like that in stage one. Here NT
75 is used to coagulate the pollutant loaded
absorbents from stages 2 and 3.
6) Stage 5 - In exceptional cases, a membrane process or a
biological reactor can be integrated into the
Piepho process.
Fixed Costs:
Pretreatment - oil skimmer 6000 DM
solid screening process - market price
Machines - 2 are needed (stages 1 and 4)
20,000 to 300,000 DM each
Mixing tanks - (2)
approx.
30,000 to 50,000 DM each
121
-------
THE PIEPHO CORPORATION
9341 Cornwell Farm Rd. Great Falls, Va. 22066, (703) 759-7074
Additional treatment - Membrane or biological reactor
market price ;
Total cost -
approx. 100,000 to 750,000 DM - this does
not include additional treatment
Treatment Costs: ( per 1000 liters)
Pretreatment - liquid separation agent (3 DM/1) = 3 DM
Stage 1 - NT 75 (0.5 - 3.5kg at 3 DM per kg) = 1.5 -!l0.5 DM
Stage 2 - Active carbon (0.5 - 3 kg at 5 DM/kg) = 2.5 - 15 DM
Stage 3 -'Fuller's earth mixture(0.5-3.5* 3DM/kg)= 1.5 - 10.5 DM
Stage 4 - NT 75 (0.5 -3.5 kg) = 1>5 _ jlo.5 DM
T°tal = 10-49.5 DM
or 2 to 10 cents per gallon
I
POSSIBLE USES OF THE PIEPHO PROCESS IN THE U.S.
We see three areas where our process could be used in the U.S.'
I
«i« °lean contaiminated water at its source of contamination; i.e.
sales to industry. This is the area where we have been most
JUS?!? J in ?urope- Our ran^e of Products allow us to treat a wide
variety of wastewater streams.
2) Cleaning contaminated groundwater and surface waters, under
Superfund and other similar state laws. We have had plenty of i
J«SaJi*nCf in t£ese areas in Eur°Pe and our products are well suited
ror this type of use. I
3) Using our technology as part of a larger process to clean '
contaminated soil. We have found two ways in which this can begone.
a) Using our process in conjunction with a mobile soil
washing plant. This plant uses specially pretreated water to
wash the contaminants from the soil. The contaminants end up
in the washing water and are removed using the Piepho
process. Plants of this type will soon be in production in
Europe.
122
-------
THE PIEPHO CORPORATION
9341 Cornwell Farm Rd. Great Falls, Va. 22066, (703) 759-7074
b) Using our technology as part of an in situ soil flushing
process. In this case we would force specially treated water
through the contaminated soil. The water would pick up the
contaminants as it flushed through the soil. This process
would be repeated until the required levels of contamination
had been reached or until the water could no longer remove
any contaminants from the soil. *
123
-------
REGIONAL BIOLOGICAL 'DECONTAMINATION CENTERS |
FOR THE CLEAN-UP OF '
CONTAMINATED SOIL, SLUDGES AND INDUSTRIAL WASTE-WATERS
A PRESENTATION FOR THE :
EPA-FORUM ON INNOVATIVE TREATMENT TECHNOLOGIES: DOMESTIC AND
INTERNATIONAL, ATLANTA, JUNE 20-22, 1989
BY
DR. HEIN KROOS, PRESIDENT AND GENERAL MANAGER,
BIODETOX GMBH, 3061 AHNSEN, WEST GERMANY :
124
-------
Introduction
Thank you very much for your kind invitation to
discuss the biodetox decontamination technology
with special regard to our concept for biological
clean-up activities in regional decontamination
centers.
There is sufficient scientific proof that oil con-
taminated soils and sludges can more or less easi-
ly be cleaned biologically once the preconditions
for biodegradation have been established. In West
Germany and surrounding countries virtually hun-
dreds of thousands of tons of oil contaminated
soils were cleaned that way during the last years,
with biodetox being one of a number of specialist
companies working since 1981 in this field and
offering a broad range of services including field
expertise, analytics, hazard evaluation and spe-
cial decontamination services based on various
patented biological clean-up processes which can
be used either in situ or on site.
As a result of these years of field experience and
discussions with the regulatory authorities the
"Bio-Pit"-process for biological clean-up activi-
ties, which is especially suited for temporary on
site or permanent decontamination centers, was
developed.
125
-------
2
3
4
foaming process
filling process
deep-ground
process
grouridwater
process
unsaturated zone
e.g. clay
foaming process
rotary cutting
irrigation
producing well
producing or
injection well;
bio reactor
contaminated
soil /qroundw.
uncon lamina ted
aquifer |
i
addition .of
microorqanisms
water back flow
126
-------
Special purpose of this presentation is to demon-
strate this proven process which has been equally
accepted by the German Federal and State Authori-
ties as well as by the public including the neigh-
bouring communities. This process is presently
running successfully in a number of German regio-
nal decontamination centers combining regional
private initiative with the central R & D-quality
control and marketing facilities of a nationwide
franchise system. The process is economically fea-
sible, without receiving government or state sub-
sidies whatsoever.
The biodetox Bio-Pit process consists of following
steps:
Contaminated soil is transported to high density
polyethylene treatment beds and receives on arri-
val a pre-inocculation with the appropriate micro-
organism cultures (biofoaming). Water is then
sprayed over the treatment bed, percolating down
to a drainage system which leads to a collection
sump and is then pumped to an aerobic bioreactor.
The bioreactor is filled with fixed, vertical
sheets or other filling material, on which biomass
either especially cultivated for this purpose or
as a standard mixed culture grows. The reactor is
aerated at a rate of 4 m3 per m3 reactor-volume
per hour and has a hydraulic rentention time of 6
- 12 hours. There the dissolved pollutants are
finally degraded should that not have happened in
the soil already. Therefore, we have clean-up ac-
tivities at two stages: First in the soil itself
(the problematical blackbox!),
127
-------
and then in the reactor with its more homogeneous
medium, especially suited for biodegradation; The
cleaned water passes into a transit tank tyhere
nutrients and additional biomass can be added as
required. The water is then pumped to the treat-
ment bed via the irrigation sprays establishing a
closed loop system. |
j
The following schematic illustration of the :Bio-
Pit process will demonstrate these details 'more
clearly. Initially, only soil contaminated ;with
mineral oil products such as gasoline, kerosene,
diesel fuel, motor oil, lubricants and heating oil
were treated in bio-pits. Meanwhile a broad range
of other pollutants can be treated successfully
including light aromatic compounds such as ben-
zene, toluene, xylene,' styrene, cresol, phenol and
phenolic compounds plus the so called coal' tar
constituents (PAH). ;
Degradation time for oil contaminated soil! has
been found to be between 6 to 12 weeks on the'ave-
rage. The following degradation diagrams will '• show
some average values. ,
Regarding other pollutants, biodegradation is lar-
gely depending on their chemical formula, concen-
tration and agglomaration, the soil character; and
i
structure and the presence of toxic or potentially
toxic metabolite forming agents. j
128
-------
BIO-PIT PROCESS-SCHEMATIC ILLUSTRATION
bioreactor
circulation
tank
irrigation
pump
irrigation
fa fa A
HOPE-foil
bio-pit
pump
-------
SOIL CONTAMINATION (DIESEL FUEL) mg/kg
co
o
CD
O
TJ
H
T)
7)
O
O
m
on
5
o
m
O
>
6
c
in
-------
BIODECRADATION RESULTS
BIO-PIT PROCESS
3.000
(mg/kg)
2.800 .
2.600
2.400 _
2.200
01
2.000
o>
_ 1.800
UJ
to
1.600
- 1.400
z
O
h-
<
I-
z
Q
O
o
t/v
1.200
1.000
800
600
400
200
BIO-PIT # 1/CH 5
range of value
average value
clean-up target
2 4
treatment time
10 (weeks)
131
-------
BIODEGRADABLE ORGANIC POLLUTANTS
(BIO-PIT PROCESS)
MINERAL OIL PRODUCTS
- GASOLINE
- DIESEL FUEL
- MOTOR OIL
- LUBRICANTS
- KEROSENE
- HEATING OIL
LIGHT AROMATICS
- BENZENE
- TOLUENE
- XYLENE
- STYRENE
- CRESOL
- PHENOLS
- PHENOLIC COMPOUNDS
- XYLENOLES
CO
ro
POLYCYCLIC AROMATIC HYDROCARBONS (PAH); e. g.
COAL TAR CONSTITUENTS
- FLUORENE
- PHENATHRENE
- ANTHRACENE
- PYRENE
- FLUORANTHENE
- BENZ(A)ANTHRACENE
._^_CHRYSENE --
- BENZ(A)PYRENE
- BENZ(B)FLUORANTHENE
- BENZ(K)FLUORANTHENE
- DIBENZO(A,H)ANTHRACENE
- INDENO(1,2,3-C,D)PYRENE
-------
Extensive research work has be done, e. g. regar-
ding biodegradation of polycyclic aromatic hydro-
carbons, found on former gasworks sites. They can
be actively biodegraded by a modified Bio-Pit pro-
cess, using especially cultivated microorganisms
either isolated from corresponding sites or di-
rectly taken from the incoming soil and being re-
injected after enrichment plus the use of other
biomechanical means such as bioemulgators and the
application of some special machinery, which was
developed for this purpose.
The following pictures will give you an idea of
the Bio-Pit operations in various Biological De-
contamination Centers, starting with a generalized
view of biodetox1 own Biodecontamination Center
which its pilot and R & D-facilities for other
centers.
The then following pictures will show additional
biodegradation aids, closing with a schematized
plant for the processing of polycyclic aromatic
hydrocarbons. It can only be shown as a presenta-
tion as it is just presently under construction.
These pictures will be followed by a process de-
scription showing the treatment of oil-/gasoline-
and sandtrap-residues (oil sludges) as an additio-
nal moneymaking proposition for these treatment
centers.
133
-------
Biological Drciiiilniiiiiinlinii OM|IM
for the purification of cmitnminntrcl
-------
Coming back to the initial schematized presenta-
tion of biodetox1 own Decontamination Center you
will have noticed that it encompasses a whole
package of activities, contaminated soil being
just one of them. The speciality of this concept
is that it can be expanded in modules, horizontal-
i
ly as well as vertically. A regional decontamina-
tion center can be e. g. started with one or more
open pits including the specific treatment equip-
ment.
Then a hall can be added, where especially gaseous
materials can be handled and the exhaust air can
be treated by a biofilter. Then e. g. equipment
for the treatment of oil-/gasoline- and sandtrap-
residues. (oil sludges) can be added. The process
was already explained.
The bioreactor farm where polluted waste water can
be treated is based on an original idea of our re-
gulatory authorities and can be seen as a third
step. Due to the restrictive fines for the uncon-
trolled delivery of highly polluted industrial
waste waters into municipal treatment plants, espe-
cially small companies, unable to finance expensi-
ve waste water pre-treatment equipment, would have
had to be shut down if there had not been an alter-
native. Therefore the regulatory authorities pro-
posed a bioreactor farm as an appendix to biode-
tox' biological decontamination center to serve as
a pre-cleaning facility for firms within our vicin-
ity including food and vegetable industry, slaugh-
ter-houses and others.
135
-------
This bioreactor farm will be used as well for the
treatment of cleaning fluids from a washing sta-
tion for multi product tank trucks, where deter-
gents, disinfectants and a high quantity of wash
water are producing waste water highly loaded with
BOD/COD. !
Presently, 3 Regional Decontamination Centers are
in operation, 2 are under construction and 5 ! in
various stages of planning. The economical operat-
ing radius is at roughly 150 miles each. Every qne
of these centers is strictly based 'on
non-subsidized private regional initiatives, the
operators being private companies, usually coming
from related fields of activity such jas
construction, hauling or waste removal. They are
cooperating with biodetox on the basis of a gene-
ral franchise agreement. j
The biodetox franchise package contains start-4up
assistance by a way of expertise, blue prints and
all red tape processing regarding the Government
and State permits, a complete legal and admini-
strative package regarding waste handling trans-
port and acceptance as well as protection against
product liability. Furthermore, biodetox manufac-
tures and ships the equipment and biochemicals to
the regional decontamination centers and monitors
the biodegradation processes, from time to time
improving on the specific formulation and the me-
chanical process engineering with regard to spe-
cial pollutants. !
13£
-------
At the same time, biodetox runs its own Regional
Decontamination Center at the company seat with
special facilities for an extensive research and
development. R & D is jointly conducted with a
number of universities and research institutes.
All R & D results are communicated to the fran-
chisees in order to further improve their biode-
gradation process and reduce the costs involved.
This way, biodetox ensures a consistently high
quality standard of operating methods nationwide.
The average clean-up charges including an adequate
profit- and risk margin for oil contaminated
soil, are at 380,00 DM per m3 or 140 US $/cu-yd.,
based on a soil kf-factor of 10~ and 3 % equal-
ing 30 000 mg (ppm) diesel fuel per kg of soil be-
fore and 500 ppm after a twelve week decontamina-
tion process.
The initial investment for the erection of a re-
gional decontamination center as described would
be between 500 000 and 1,5 Million US-Dollars with
an amortisation rate of ca., three years.
Last not least, you might be interested of what
happens to the decontaminated soil after clean-up.
Here an agreement with the regulatory authorities
has been reached to the extent that soil with a
threshold-level of equal or less than 500 mg of
mineraloil hydrocarbons per kg of soil can be used
for agricultural purposes. Other material with
threshold-levels at or above 500 mg up to 1000
mg/kg may be used as cover material for landfills,
as road building material or for noise protection
137
-------
walls. Another precondition being that an analysis
must prove that there are no soil pollutanjts at
the new site which would add to a further acicumu-
.Tation of pollutants. j
j
Obviously, biological clean-ups, such as conducted
in regional decontamination centers cannot be suc-
cessfully applied to all hazardous wastes. Never-
theless, applying it to all sorts of lower class
hazardous wastes as already described pre'cious
land fill space and treatment costs can be slaved.
Furthermore, biological clean-ups are lasting and
final. There is no problem transfer into the fu-
ture or to another place and there are no topical
residues that have to be taken care of. i"
With regard to these facts the German Federal
Government and the governments of various German
states are presently giving priority to biological
clean-up- and recycling methods. New laws and re-
gulations under way will stress this comittmeht to
assist nature to help itself. ;
I would now like to summarize and to add that! biode-
tcx as interested in licencees worldwide. !
Thank you very much for your attention.
I
138
-------
Biological remediation of contaminated groundwater and soil -
Concepts of remediation and their technical application
M. Kastner 1), K. Hoppenheidt 2), H. H. Hanert 2)
Department of Biotechnology; Technical University of Hamburg-
Harburg. Harburger SchloBstr. 37, D-2100 Hamburg 90, F.R.G.
Institute for Microbiology; Technical University of Braunschweig
Biocenter, Konstantin Uhde Str. 5, D-3300 Braunschweig, F.R.G.
Summary:
In this paper different ways to develop biological processes for the remediation
of contaminated soil and groundwater are discussed. With dichloromethane
as an example it is shown how elimination processes can be developed
which are designed for a specific organic pollutant and specific bacteria
(Such developments are often to be found in literature). The fixed-bed
reactor process presented here is able to achieve results of decomposition
of dichloromethane from groundwaters of 4,2 kg/m * d. These results
could be maintained for more than six months. As in many contaminated
sites bacteria adapted to the organic pollutants are already existing, the
example of a contamination with complex organic compounds will be used to
show the activation of the microflora from the site to degrade the organic
compounds and its technical application in remediation as another way to
develop processes. From these investigations a concept of action was de-
veloped which permits statements about the possibility of biological reme-
diation of a contaminated site and the processes which can be used, even
with relatively few experiments. This concept facilitated the application of
biological processes to different organic pollutants in all cases examined
until now.
Recent studies show that many microorganisms are not only able to degrade
natural organic substrates, but also synthetic organic compounds and compo-
nents of petroleum. This degradation was mostly to be found under special
laboratory conditions, where after suitable times of adaptation the parameter
of the respective metabolism were optimally adj.usted. Compounds, however,
which reached the environment as organic pollutants or hazardous compounds,
are persistant in most cases, or are only slightly transformed by the biological
activity. If natural compounds were introduced in concentrations not excessively
high, on the other hand, they were mineralized by processes of bacterial
self-purification, which made use of the whole spectrum of biological types
of metabolism. In the course of this the organic compounds may serve as
donator and acceptor of electrons, respectively. The different processes of
self-purification in soil and groundwaters are represented as biological redox-
processes in Figure 1 .
139
-------
OS
+0.5
REDU
<
KTION
<^02 •Hcduktion |
( Otnibilikalion |
^WIY)o«y«- Mntt
biydoLvonHnW)^
[Nj-NOj >
|0;-0ildur>9
>
-TO
-5
20 pe
Fig.1: Biological redox-processes as natural self-purification processes in
soil and groundwater (translated from H. Schwoerbel, 1984-) i
In technical microbiology these microbial self-purification processes are a ready
used to compost organic waste, for sewage treatment, for treatment of waste
gas and for the production of drinking water.
Own investigations at different contaminated sites in the Federal Republic of
Germany revealed the existence of a microflora which is already adapted to
the organic pollutants, but is evidently unable to develop its activity. Ini many
research works described in literature, dealing with the bacterial degradation
of specific organic pollutants in the .environment, the bacteria were isblated
from contaminated sites as well. From that two questions concerning the
remediation of contaminated soil and groundwaters follow:
Why don't the self-purification processes work on organic pollutants n the
environment, too? I
How can remediation-processes be developed from the results of the experiments
in laboratory ?
Principally it is to be noted that in contrast to chemico-physical processes
microbiological processes can be standardized neither in situ nor on site, as
nearly all contaminated sites are to be differentiated because of their different
140
-------
conditions. Technical instructions or standards thus are not yet in sight.
However, for the conception of a biological remediation and its necessary
microbiological investigations the way of proceeding can be explicity formulated,
as will be explained later on.
First of all, the two following methodological approaches are to be differen-
tiated:
1). Isolation of special bacteria, which are able to degrade a specific organic
pollutant (possibly the genetic optimization of the strains, too) followed
by the development of processes which are suited to the substance
which is to be degraded and the respective bacteria.
2). Stimulation of the microbial activity of the contaminated site by investi-
gations of microcosms, which are suited to the usable physiological type
of metabolism. The second method has the advantage of allowing direct
statements about the possibility of a remediation of the respective site
and the determination of suitable concepts of processes.
The first approach can be explained with the example of dichloromethane de-
gradation. From different contaminated sites caused by dichloromethane pure
cultures of Hyphomicrobium species were isolated. These strains are able to
grow aerobically with this substance in concentrations up to 1 g/l medium
as the sole source of carbon and energy. As these bacteria are known to in-
habitate rapid-sand filters used in water supply plants, and these technology
is generally used, the idea of using the same technology for the remediation
of contaminated groundwaters suggested itself, in order to keep problems of
the scale up to full technical scale as small as possible. For the continuous
degradation of dichloromethane in a modell groundwater a fixed-bed reactor
inoculated by one of the isolated strains was operated. The fixed-bed reactor
corresponds to the technology of rapid-sand filters and shall prove the principal
suitability of this technology. The process scheme of the reactor is represented
in Figure 2.
The degradation power reached by this reactor and the calculated values from
that for 1 m3 plant volume are represented in table 1. The degradation could
be kept stable for more than six months. During this time the used strain was
not overgrown by other bacteria. Produced biomass was removed by back-
washing of the fixed-bed material, at any one time when 1 00 g of dichloro-
methane were degraded. The degree of efficiency as well as the concentration
of the influent and effluent after a backwashing is shown in figure 3. As
these results show that this process can be used in technical applications.
Nevertheless for a concrete contaminated site further experiments with the
respective groundwater have to be carried out.
141
-------
b
c
d
e
f
g
Stock solution (water,1 50 I) h
Oxigenator I
Membrane pump (circulation of
gas) j
Stock solution (Mineral salts, DCM)
Peristaltic pump k
Sample port I
Backwashing drain
Fixed-bed reactor (2 l,r=|4cm)
Quartz sand (1 .7 I; i
pore space: 0.28 I) j
Sample port and !
backwashing inflow '•
Effluent '
Backwashing pump
Fig. 2: Process scheme of the fixed-bed reactor for the removal
dichloromethane
Table 1 : Dichloromethane removal of the fixed-bed reactor
velocity of influent;
detention time*)
ml/h ; min.
2623 6.6
524-6 3.3
104-92 1.7
vel. of infl.; load
det. time*) DCM
ml/h mg/h
2623 129.0
5246 281 .7
104-92 619.0
influent
DCM Cl~
mg/l
4-9.2 6.8
53.7 6.8
59.0 5.8
effluent
DCM Cf
mg/l
n.n 26.6
n.n 27.3
2.9 29.8
ADCM
from ACI
mg/l
23.5
24.2
28.4
DCM-removal
reactor
mg/h
61 .5
126.9
298.3
DCM-removal
3 I
per m plant v'p\.
g/m3xd j
869 j
1792 !
4212 i
effective detention time in pore space
_
142
-------
Efficiency
in %
Dlcnloromeihane
mg/l
100
80
60
40
20
50
40
30
20
10
Efficiency
Influent
Effluent
10
15
20
25
Fig. 3: Dichloromethane removal of the fixed-bed reactor with backwashing
(arrow)
The second methodological approach can be exemplified by a special contamina-
ted site in a solvent recycling factory. The contamination of groundwater and
soil consisted of a complex mixture of aliphatic and aromatic solvents and
volatile chlorinated hydrocarbons. In diethylether-extracts from soil and ground-
water over 70 single substances could be detected by gaschromatography. The
contaminated area had a size of 10000 m2, the area of the contamination
centre covered 1 200 m2. Until a depth of 25 m (level of groundwater: 4.5 m)
the contamination was proved. It is estimated that altogether 4-0-50 t of ali-
phatic and aromatic compounds and 4-5 t of volatile chlorinated-hydrocarbons
had seeped into the soil here. In the soil organic contamination (per kg dry
weight) up to 600 mg/kg of aliphatic and up to 700 mg/kg aromatic hydro-
carbons were to be found. At some points the concentration of chlorinated
hydrocarbons reached 4-60 mg/kg. The contamination of the groundwater
amounted up to 19200 mg COD/I as summary parameter for the organic
contamination and up to 150 mg/l for the volatile chlorinated hydrocarbons.
Because of the high contamination and the resulting costs for a chemico-rphysical
remediation it was decided to examine if the site could be remediated biologi-
cally. The compound consistence had the consequence that biological remediation
processes designed for specific organic pollutants and bacteria could not be
used here.
143
-------
As a result of this the microbiological investigations of the site were carried
out under different aspects. Besides the determination of the numbers of bac-
teria in groundwater and soil in dependance of the concentration of the organic
pollutants, parts of the investigations applied to the degradation of the highly
contaminated groundwater. After various pre-experiments a laboratory sevyage
treatment plant with a closed circulation of gas was developed for this purppse.
The process scheme is represented in figure 4-. The average COD-loading of
Q I
the plant amounted to 0.73 kg/m x d, the loading of volatile chlorinated hydro-
carbons amounted to 3.4 g/m3x d. With a suitable feeding of mineral nutrients
and electron acceptors (OJ, an activated sludge developed from the bactjeria
of the contaminated site reached average removals of 95 % COD (influent:
5490 mg/l; effluent: 250 mg/l) and 94-97 % of the chlorinated hydrocarbons
with a residence time of 7.5 days in the plant. The mineralization of the
toxic chlorinated hydrocarbons could almost completely be assessed by' the
a
g
m
Mineral nutrients
Groundwater !
Septum ports \
O2-stock (gas Ibag)
Reactor
Sedimentation
vessel
CO -absorption;
vessel !
Gas supply
Magnetic stirrer
Siphon |
Peristaltic pump
Membrane pump
!
Effluent I
}6.5I
Fig. 4: Process scheme of laboratory sewage treatment plant
144
-------
release of chlorine as chloride ions. Concentrations (influent and effluent) of
the chlorinated compounds as well as the concentrations of the chloride ions
are represented in figure 5. Table 2 summarizes the distribution of specific
chlorinated hydrocarbons and release of chloride ions made possible by this.
Batch experiments show that perchloroethylene could only be decomposed in
a small amount by the activated sludge; that is why at higher concentrations
a side-reaction of this substance has to be made use of. With a shortage
of oxygen the sludge dechlorinated tri- and perchloroethylene by way of re-
duction to cis-1,2-dichloroethylene very fast. By insertion of a unaerated col-
CKH ug/1 fCHC]
60000
50000
40000
30000
20000
10000
Zulauf .
Influent
Effluent
^ flblauf
0 10 20 30 40 50 60 70 80 90 100
d
Cfng/1
140
120
100
80
60
40
20
0
flblauf
EffluentJ
Zulauf
Influent
10
20 30
40
50
60
70
80
30
11
Fig. 5: Influent and effluent concentrations ( summarized volatile chlorinated
hydrocarbons and chloride ions) from the laboratory sewage treatment
plant
145
-------
Table 2: Distribution of volatile chlorinated hydrocarbons and chloride1 ions.
average concentration of influent and effluent from the laboratory
sewage treatment plant
Effluent
HO/1
Vinylidenechlorid
Dichloromethane
1,1 -Dichloroethane
cis-1,2-Dichloroethylene
Chloroform
1,1,1 -Trichloroethane
1,2-Dichloroethane
Trichloroethylene
Perchloroethylene
E CHC
Chloride
21.7
22558
662
21342
363
153
435
5471
40
51046
57500
Influent Removal
ug/i
0.5
28
183
1167
62
27
n.n
227
1.2
1695 A= 49350
(* 38800 CD
100300 A= 42800
/»
97.
99.8
72.3
94.5
82.9
82.3
(99.9)
95.9
97.0
96.
—
lection vessel for the return sludge (1 I; 16 h residence time) and the feuding
of the groundwater into this vessel the available amount of trichloroethylene
could be transformed into cis-1,2-dichloroethylene (Batch experiments \ with
perchloroethylene showed the same effect). The increase of cis-1 ,2-dichjoro-
ethylene concentration and the reduction of trichloroethylene concentration in
this experiment is represented -in figure 6. j
Tri pD/1
300
200
100
Cis
3B08
Tri
Cis
t4 T 8 T 12 T 16 Th
10
20
30
40
50
2000
1000
Fig. 6: Effluent concentrations of the laboratory sewage treatment plarjt at
different anaerobic residence times of the backflowing activated
sludge j
146
-------
In corresponding experiments with soil columns from the centre of the con-
tamination the degradation of the organic pollutants by the bacteria from the
soil could be activated with the addition of minerals (fertilizer) and electron
acceptors like O2 and nitrate. But a detailed degradation balance of the volatile
chlorinated hydrocarbons could not be carried out in the soil because of ana-
lytical problems.
The results represented here show that in contaminated sites with such a
complex composition of pollutants the biological power of self-purification of
the soil and groundwater can be activated if conditions are chosen adequately.
The persistance of the organic pollutants at this site is due to the limitation
of mineral nutrients and electron acceptor supply. It can be derived from
the results of the soil experiments that merely the addition of nitrate-con-
taining fertilizers in connection with atmospheric precipitates led to an elimi-
nation of the organic pollutants. Thus the soil of the site could be used as
an in situ fixed-bed reactor with a remediation time of approx. two years.
The seeping of dissolved fertilizer in injektion wells without hydraulic water
movement showed the same results for groundwater. The addition of fertilizer
was watched with special sampling systems in order to keep the supply of
nitrate as an electron acceptor constantly limited against the carbon source.
With this secondary contaminations with nitrate could be avoided safely. This
process of remediation manages nearly completely without the building of
plants above ground and causes only approx. 1 /20 of the expense of a chemico-
physical remediation. At the moment the remediation in a technical scale is
carried out and has led to a 60-70 % reduction of the organic pollutants at
the site after just 1 .5 years. The details of the remediation will be reported
in international journals when the project is concluded.
A summary of the pre-experiments at the site described above also reveals
a microbiological concept for the investigations of a site which has to be
remediated. This concept will be explained finally. The investigations mainly
have to answer the eight following questions, from which the possibility of a
biological remediation and the technical process to be applied can be derived.
1). How is the biological state of the contaminated site ?
(quantitative determination of the complete population of bacteria, fungi
and actinomycetes; comparison to the situation at the periphery of the
contamination; evaluation of the toxic effect of the pollutants upon the
microflora)
2). What are the rates of respiration activities of the microflora of the site ?
(physiological evaluation of the self-purification power of the site; re-
spiration of organic carbon sources with and without optimal supply of
minerals and electron acceptors)
147
-------
3).
4),
5).
6),
7).
Does a microflora adapted to the organic pollutants exist in the site ?
(site specific and pollutant specific evaluation of the self-purification
potential under optimal conditions) !
What are the reasons of the inhibition of the microbial self-purification
processes at the site ? :
(evaluation of the site specific inhibition potential: toxicity of pollutants,
electron acceptor or mineral limitations) j
How quickly is the biological self-purification carried out with an optimal
supply of nutrients ? I
(kinetics of the self-purification in batch experiments with discontinuous
addition of nutrients; determination of the duration of remedial acjions)
Which degree of purification can be achieved ? i
(the degree of purification should not fall below 90 %. at least under
laboratory conditions; determination via summation parameters; for areas
with contaminations of chlorinated hydrocarbons, polycyclic aromatic hydro-
carbons, chlorinated aromatic hydrocarbons, mineral and tar oils, according
to previous experiences corresponding powers of purification are to be
achieved even if the contaminations amount to the area of g/kg.}
Are organic/anorganic intermediary products with toxic effects on human
beeings produced during the purification and how much time does the
microflora needs to eliminate these substances, too ? |
(e. g. accumulation of vinyl- and vinylidenechloride or nitrite in the rrjieta-
bolism of volatile chlorinatd hydrocarbons) i
Which technical process should be applied to remove the pollutantjs on
the basis of the experimental results in connection to the situation at
the site (extent of pollution; who or what is endangered ?; which use is
intended for the site) and under aspects of economy ? ',
The design of experiments has to be developed in accordance to the specific
contaminated site; that means the physiological status quo has to be adapted
to the physiology of the degradation of the organic pollutants (see also fig;. 1).
In this not single the tests of bacteria or substances are concerned primarily,
but degradation tests of mixed populations under specific physiological conditions
which are possible. The answering of the questions in a phase of laboratory
tests of approx. 4 months and the testing of the resulting concepts of re|me-
diation in a half-scale technical phase of also approx. 4 to 6 months led to
the development of the remediation process for the site desribed above. This
concept of action in the meantime also led to the application of biological
remediation processes in some cases of contamination caused by mineral! oil,
tar oil, chlorinated organic compounds originated from the production of pe!sti-
cides and at sites of former gasworks. The investigations represented here .
8).
148
-------
allow to make predictions about the duration and success of a remediation
which is to be expected in every specific site. To this a decisive importance
is attached by the supervisory authorities in the Federal Republic of Germany.
In cases carried out until now the costs of the experiments for each site
amounted to 10-20 % of the total remediation costs, depending on the spec-
trum of pollutants.
Literature:
M. Kastner und H. H. Hanert: Biologische Elimination von Halogenkohlenwasser-
stoffen in belasteten Grundwasserleitern; in: Veroffentlichungen
des BMFT, Hrsg.: Sanierung kontaminierter Standorte - Dokumen-
tation einer Fachtagung 1985, Berlin 1986
(Biological elimination of halogenated hydrocarbons in contaminated
aquifers; in publications of BMFT [Federal ministerium of research
and technology], ed.: remediation of contaminated sites — documen-
tation of a special conference 1985, Berlin 1986)
H. Hanert: Mikrobiologische Bewertung von kontaminierten Standorten im 'Hin-
blick auf eine biologische insitu-Sanierung; S. 1 43-1 53
(Microbiological valuation of contaminated sites for biological in
situ remediation, p 1 43-1 53)
M. Kastner: Biologische Elimination von leichtfllichtigen Halogenkohlenwasser-
stoffen (Theoretische Grundlagen und Laborversuche); S. 155-166
(Biological elimination of volatile hydrocarbons [theoretical basics
and laboratory experiments]; p 155-166)
K. Hoppenheidt, H. H. Hanert: Untersuchungen zur biologischen Reinigung eines
organisch hoch kontaminierten Grundwassers (CKW, Aromaten,
Aliphaten) in einer Labor-Belebungsanlage; S. 167-175
(investigations for biological cleaning of groundwater highly conta-
minated with organic compounds [volatile chlorinated-, aromatic-,
aliphatic compounds] in laboratory sewage treatment plant; p
167-175)
P. Harborth, H. Rose: Biologische insitu-Sanierung kontaminierter B6den und
Grundwasser; S. 177-187
(Biological in situ remediation of contaminated soil and ground-
waters; p 177-187)
in: Veroffentlichungen des Zentrums fur Abfallforschung der TU
Braunschweig: Fachseminar Bodensanierung und Grundwasserreini-
gung -Wiedernutzung von Altstandorten- 24.725.9.1 986 in Braun-
schweig, 1 986
(In: publications of the center of waste research, Technical Uni-
versity of Braunschweig: special conference for soil remediation
149
-------
and groundwater reclamation — reusing of contaminated sites —
24.725.9.1986 in Braunschweig, 1986) |
i
M. KSstner, K. Hoppenheidt und H.H. Hanert: Bakterielle Umwandlungsreaktionen
in CKW belasteten Grundwasserleitern - Sanierungstechnike|n; in:
K. Wolf; W. J. van den Brink; F. J. Colon, Hrsg.: Altlastens|anie-
rung '88; Zweiter Internationaler TNO/BMFT-Kongress liber Alt-
lastensanierung, 11.- 15. April 1988, Hamburg, Bundesrej^ublik
Deutschland; Bd. II, S. 1 263-1 264; Kluyver Academic Publishers,
Dordrecht/Boston/London, i988 !
(Bacterial transformation reactions in groundwater aquifers c|onta-
minated with chlorinated hydrocarbons — purification techniques;
in: ..... eds.: contaminated soil '88; second international ITNO/
BMFT-congress in contaminated soil, 11.- 1 5. april 1 988, Hamburg,
Federal Republic of Germany; Vol. II, p 1263-1264; Kluyverj )
H. H. Hanert, P. Harborth, M. Lehmann, E. Windt, U. Rinkel, H. J. Scheibel,
K. Hoppenheidt, H. Rose, T. Niemeyer; Biologische Selbstreiriigung
in Boden und GrundwSssern und ihre technologische Nutzung in der
Bodensanierung und Grundwasserreinigung; Sonderdruck des Ver-
eins zur Reinhaltung der Gewasser e. V. Braunschweig, Hrsg.,
Claus-Druckerei und Verlag Braunschweig, April 1988
(Biological self-purification in soil and groundwater and their tech-
nical application for soil remediation and groundwater reclamation;
special print of the society for keeping waters clean. Braunschweig,
ed.. Glaus Printers and Publishers, Braunschweig, April 1988)
M. KSstner: Anreicherung und Isolierung von Chlorkohlenwasserstoffe abbauen-
den Mikroorganismen unter verschiedenen physiologischen Bedirjgun-
gen — Abbaukinetiken und Test auf technische Nutzbarkeit zur"
Sanierung kontaminierter GrundwSsser; Dissertation an der tech-
nischen Universitat Braunschweig, 1988
(Enrichment and isolation of chlorinated hydrocarbon degrading
microorganisms under different physiological conditions — kinetics
of degradation and test for technical utilizability for remedi'ation
of contaminated groundwaters; PH.D-thesis at the Technical Uni-
versity of Braunschweig, 1988) i
150
-------
The Holzaann Svstetn for In Situ Soil Purification
1. Introduction - Disused Gasworks Sites
During the past decade in West Germany, numerous abandoned or
disused gasworks sites and coke plants have been examined with
regards to environmental hazards.
By means o£ the chemical processes involved, and through the
coking process of fossile combustibles considerable amounts of
tar oils, containing numerous polynuclear(polycyclio) arom.il ic
hydrocarbons (e.g. naphtalene), cyanides, sulphureous sub-
stances, and phenol are produced as by-products. In the pant,
these by-products were separated, stored in basins, and in
more or less suitable containers, treated, and then largely
used for further chemical or technical purposes.
Through careless usage however, leaking tanks and basins,
spillages and in particular as a result of war activities, con-
tinuous and high rates of pollution have been discovered at
almost every gasworks and coke plant site.
Through the increased environmental awareness of the public
including the concern expressed by responsible authorities
during the last years, contaminated former gasworks sites have
aroused strong public interest.
Most investigations carried out on former sites have revealed
serious pollution hazardous for soil, groundwater, and detri-
mental to human health.
2. Former Gasworks in Bre»en-Holtmershausen
During initial soundings in mid-1987 on the site of the former
gasworks in Bremen-Woltmershausen, in which between 1900 and
1964 gas had been produced, extremely high contents of tar oil
(Polynuclear Aromatic Carbons, PAC) and cyanides were dis-
covered underneath a former tar basin, a gas tank, and a basin
for fire-extinguishing water.
Further detailed analyses revealed that below the tar basin
about 20000 m" of medium sand and a layer of clay containing
sea silt are contaminated with PAC and cyanides to a maximum
depth of approximately 10 m.
Soil specimens were characterised by intensive black coloration
and penetrant odor. Analysis of the soil-oil mixture revealed
up to 20000 mg PAC/kq and up to 600 rag cyanides/kg.
Fig. 1 shows the characteristic soil profile, grain size
distributions, and the qualitative distribution of PAC over
depth.
Soil Profile and Contamination
Soil Profile Contamination Typical Grain Size Distribution
PAC ind CyinidM
FID, S«nd
Medium m)
FkwSwd
CUywilh
SMSIIt
o
FkMSmd
RMind
MoftmSlnd
13f>
I.HIO
Fomw BIM
of tv Bum
Uuhun:
20000mg
MC/kg
600 mg
CpridM/kg
Lmr Migration
Until ol
Pollutants
00020000 002 000 02 08 2 6 SO 80
GfilnSlnkimm
•iCIiywithS«Sit L3 WdkmSind
Fig.l: Characteristic soil profile and grain size distribution.
Distribution of pollutants.
A layer of silt (clay containing sea silt) approximately one
meter thick lies in the quite uniform (U=2) medium sand strata
at the center of the contamination. The ground water level
lies approximately 2.7 m below the terrain surface.
The highest percentage of PAC was discovered above the silt in
the variation range of the groundwater level. The lowest
percentage occurs within the silt layer. The sand below the
silt layer shows surprisingly high PAC-contamination levels,
partly down to considerable depth.
The concentration of the cyanides is distributed less uni-
formly. A slight cyanide contamination can also be npted in
the silt.
Buildings are located directly adjacent to the contaminated
area, severely limiting the space available for the soil
purification work. Moreover, regular operations on the
grounds have to continue during the purification work.
Fig. 2 summarises the main characteristics of the contarai-
11,11. ion.
-------
r
Man Characteristics
of Contamination
Ongm
Sue
Contamination
Maximum
Concentrations
Soil Properties
former tutusinolaty gas
production punt,
results ol war acton
2Sx<5m'«Dd 15x32 m'
-8mrMep.le.~t6 000m"
next lo enstmoj structures
MycycJicAronatic Carbons (PAC)
Cyanides (CN)
HI so!) and groundwater
PACtloMfi)
20 000 mg/kg soil (well
CN tO:
600 mg/kg soil (well
- mainly medium lo tine,
homogeneous sand
-layer ol clay with sea sill
ol tow permeability
Treatment Methods
Considered
Method
EiS4uM>crobcal
Treatment
In Situ Microbiological
Treatment
OdSiIeSo.1V/am.na
Thermal Treatment
Placement on Hazardous
Waste Dump
Reasons lot Rcteebon
ocep ciotvnbon ncil to structure
• loo high ctmr.iirttr.it«ms
• duration
• soil profilrsancj tow pcrmc;u*Mics
• duralion
• deep excavation retiuired
• lujhcoMlixmcrm.il
treatment above 1000 C
• complex per mission requirements
• large volumes and high cost
• deep excavation required
Chosen Method: (n situ sod purification l>v Mali Pressure
Injection |HPl| with mechanical soil-fluid
separation and on site wall* Ircatrncnt and
off site microbiological sludge treatment
• Tlicraic purification
Te»[)oraturco between 1000'C and 1200*C would have been
required for purification of the pollutants prooent. This
would have resulted in complete sterilization of the entire
quantity of soil. A deep excavation and an appropriate
retaining structure would have increased the cost.
• UM site or off site soil washing
The high coat for the required deep excavation and the related
groundwater lowering ruled out t:.ese nethods. Groundwater
pumping would have produced tremendous anounts of contaminated
water.
Fig. 2 summarises the considered treatment methods and the
reasons for their rejection. Because established methods
proved unsuitable for the given task of soil purification,a
new method of in situ soil purification, developed by the
Philipp Holzmann AG, was suggested and laboratory tests
carried out to optimise the process parameters. The new in
cilu soil washing technique, for which several related pa-
tents are pending, wil] be described as follows.
4. The Holziann System for In Situ Soil Purification
01
IV)
Fig. 2: Main characteristics of contamination and treatment
methods considered
3. Selection of a Suitable Treatment Method
Disposal of the contaminated soil at a hazardous waste dump
was not even considered in detail. On the one hand, a dump
suitable for disposal of such large quantities as required was
not available. On the other hand, the main objective was to
reduce the pollution potential of the contaminated soil to
such an extent that the prescribed limits could be maintained
and the soil be re-used.
The following methods were examined for suitability:
• Microbiological treatment in situ
Too high pollutant concentrations in the most contaminated
soil portions and a nearly impermeable silt layer excluded
this approach as well as the long duration of treatment
to be expected.
• Microbiological treatment of.f site
Of f~silre- processes were also" eTimTna~tea~3ue~To the exces-
sively high space requirements for the compostinq piles and
the expense of transporting the soil. Moreover, unless the
soil was prewashed, the high concentration o£ tar oils would
result in intensive adhesion between the soil particles and
the pollutants resulting in problems for all niicrnhiojnni>:al
processes.
4.1 Basic Principle of the Holzmann In Situ Soil Purification
The first and main step of the Holzmann in situ soil washing
procedure is based on the meanwhile well-established con-
nlru<:tion procedure of high-pressure water injection (HPI).
Modern equipment makes it possible to inject water into the
soil with a special rotating lance at pressures up to 500
bar and flow rates of 300 liters/min.
For the treatment with high-pressure injection, the conta-
minated soil requiring purification is divided into individual
vertical borings, each jacketed by means of a casing. The
actual purification process is accomplished boring by boring,
eliminating the need for a large and deep excavation and
the related groundwater lowering. The borings overlap in
ouch a way that no soil remains unwashed.
The purification is a multistage process consisting of the
following parts:
- separation of pollutants from soil particles and washing of
soil in situ
extraction o£ soil-water mixture from the ground
reparation of clean-particles-f-ronr-water—and—s-ludge —
water treatment
microbiological sludqe treatment
-------
- return of purified soil and water into the ground
These steps of the purification process will now be described
in detail.
4.2 Stages of the Purification Process
en
OJ
Placement of Casings
individual cased borings are used for in situ Boil wash inn
to create a clearly defined "reaction volume" tor the
high-prosnure wachimj, avoiding mutual exchange of pollu
tants and the need for deep retaining structure;;.
Hoisted by a rope excavator W 180 an ICE 81b vibrator lowers
round steel casings (9.50 m long, 1.50 m in diameter, 14 mm
thick) 8.50 m into the ground for a treatment depth of
8.00 m (Fig. 3). The working platform lies 1.2 m below the
original ground surface.
The casings are positioned in such a way that they finally
form a regular pattern of overlapping borings across the
treatment area. The complete soil volume is covered thereby;
about 17% are treated twice. Exact measurements ensure
verticality of the casings.
Placement of Casings
and Pattern of
Overlapping Borings
Ground-Plan
[ twdiunt Stnd MB C—,
-nxsin -edxsm
F'ii:(!!>!i «l high-pressure washing
-------
Separator Unit
Cbart ol the «l»»tion process. The
ia puapad on to a vibratory »ieve,
°Ve ? ""' "ainly or9anic «*Pon«nt; and
?ravel are "Crated. The paoaing
la.p""ped lnto tw° hydrocyclones in parallel
n with a separation limit at about 60 - 80 un.
^»n«? 2d^?ludge and Hator Pa"»Bing the double cyclone are
S3 rS K ,Hlr;C«11? t0 th3mm
Bon-Hod
Fig. 5: Flow diagram of separator unit
Material botween 60 and 120 pa in particle «iz« i« passed
' upotraa'
theiltr canke? "" deHater1^ sieve where U
Fresh water and purified water fron the water treatment plant
is temporarily stored in a water tank fro» where three pumpa
regulate its usage as HPI-water (18 »»/h), additional flushing
water within the separator unit (17.7 n'/h), and support wate?
tor m.iterial extraction from the boring.
the separtor uni .
The whole separator unit, newly designed by the Philipp Holz-
mann AG, is a mobile unit, completely fitted into a single
container frame in an upright position, about 10 • in height
Separator Unit
Overview of
Mass Transport
Freshwater 17,7m3/h
r
fir
Pollutants
Contaminat
40,5 m3/h
13,0 t/h
sd
i
Mixture
•
StptntofWi
' V-S cm/mini
• Velocity ollowwnj
; HPI-Unce
i
'
Purified Sand
5,9 m'/h Pumpuivoiuii
9,0 t/h MassFtow
Fig. 0: Mass tlow through separator unit
-------
01
Ul
Now the water treatment, that can well be seen as a cnquciicc ot
standard steps known from sanitary engineering, will be de-
scribed in brief with emphasis on the special features tor the
Bremen gasworks job.
• Water Treatment Plant
The sludqe-water-pollutant mixture reaching the water treat
ment plant is first led into tlocculation tanks. A screw
conveyor transports the flocculated sludge into sieve tanks
lined with synthetic mats, where dewatering to a dry mibstnnce
content of 50% takes place.
In the next stage oily components are skimmed trom the sludge-
free water usinq buoyancy effects. A storage tank collects
the tar oils.
The oil-free water then passes an activated carbon filter,
where polar substances and remaining suspended particles are
adsorbed.
The last treatment stage is a cyanide flocculation.
Potable water leaves the water treatment plant into a storage
tank from where re-use is directed by pumps. Excess water is
returned to the ground via injection wells.
Our partner Umweltschutz Nord has designed and 'operates the
water treatment plant as well as the final sludge treatment
to be described next.
• Sludge Treatment
Sludge separated in the water treatment procedure and conta-
minated by-products of the separation are treated off site
microbiologically in composting piles where specially adapted
bacteria and substratum are added to optimise the decomposition
of the polluted sludge. Within six months the PAC-contents have
been reduced by 90 to 98%.
4.3 Overview of the Holzmann In Situ Soil Purification
The described complete cleaning process makes it pos.sible,
that no contaminated material has to be deposited nu hazar
dous waste.
For the Bremen gasworks site the described purification
process results in a plant arrangement shown in Fig. 7.
Fig. 8 summarises the purification and treatment steps and
shows their relationships in the whole process. •
Plan View of Site
j L
Supervision. *••'• ' Bfl H
Laboratory.' IHA .1L-A •
Site Cimp
Process
SludgeTank WileiTir*
' ' HPI-Pump
TnkfM
Contaminated
Wiler
Purified Sand
Fig.V: plant arrangement on site
Holzmann-System of In Situ Soil Purification
Placement ot Casing
First Purification Step by High-Pressure injection and Flushing In Silu
Pumping of Soil. Water, and Pollutants
Separation of Grain Size >3 mm and Organic Material:
.Vibratory Sieve
Separation ot Grain Size 60um -3 mm from Fluid Phase:
Hydrocyclones
Flushing of Sand with Freshwater:
Up-Stream Classifier
Dewatering ot Sand (0,1 -3.0 mm):
Dewatering Sieve
*
Exchange of Water within Casing
Replacement ot Sand
Removing of Casing
—
Floccuialion.
Separation,
and Dewatering
of Sludge
Skimming of Oil
Filtration with
Activated Carbon
Floccuialion
of Cyanides
)l flu? llolznvinn -Bv.'itoin of in r.ilu
purification
-------
b. Purification Kcnulto
In the pilot phase of the project a detailed liitiil and
laboratory tooting progranra van carried out by an
dent aupervinnr and the Holzmann Environnont.il l.ali.
Individual soil colimno (borinqs) were tcalcil Imliini ami
after purification treatment.
Typical purification results tor optimum action ol tho
pilot treatment plant lay between 98% and 991 reduction nl
the degree of contamination lor PAC. Planned variations ol
the test boundary condition!! and occasional raallunrtIOIIH
reduced the overall avcraqe reduction!: a:i nhown in Kitj. 9.
The relative degree.of purification is nearly independent
of the content of pollutants both for the accumulated
content of 16 PAC and a sun value for cyaniden.
Firnt result!; ot the Main purification contract, uiiii.-r
operation since Hay 1989, are very encouraging. Even tor
the most critically contaminated soil portions with up to
20000 rag PAC/kg dry soil it is most probable that the
guaranteed target value of less than 30 mg PAC/kq will be
reached. As average value of the first 30 borings in a
heavily contaminated soil portion approximately 5 to
10 rag/kg dry soil were measured after purification.
6.
Uuniiary and Aspects ot Application of Holzmann In Situ
Die soil purification plant in Bremen is designed for a
capacity of 6 m' soil per hour. Soil extraction from tho
tjround lien on the critical path. With the separator unit
about 9 n'/h soil can be treated if the input is continuous.
H.r the extraction and purification of 1 ton of soil about
J.,2 H* process water are used, most of which is partly
cleaned water circulating in an inner course. The specific
enerqy demand is about 50 kHh/»'. Up.to 80% of it are
rtiiiuircd by the high-pressure washing.
l-'or I he purification of the Bremen gasworks soil the Holz-
mami in situ purification comprises a number of advantageous
properties that were already expressed in the comparison of
different approaches in Fig. 2.
ll.iHcsvor the method aluo has leas obvious features that make
it most flexible tor a variety of contamination problems
.iricl Hiiperior to other methods where high degrees of complex
contaminations, narrow site conditions, and necessity for
restoration of original soil properties cone together. Fig.
10 summarises the genuine advantages of the Holznann In
Situ Soil Purification and gives hints towards future
supplementations even widening the outlined range of appli-
cation.
en
er>
Results and Test Data of In Situ Soil Purification
Average values of degree of purification for RAC and Cyanides
I PAC before
treatment
(mo/kg wen
60-200
200-2000
2000-14000
Average
TugMvtlwIor
puffictttonrfiurl:
Dutch refrain vilu* A:
Dutch ra(innc*nIu«B:
Dutch raliranc* nlu> C:
Av«f»88 hMay1989:
Pilot Phu* 1988
Degree ol
purification
1,
96.0
96,3
96,3
96,2
10-30
0.1
20
200
9.6
ICN Delate
treatment
Img/ttgwel)
40-200
200-600
Average
5
50
500
Di^woi
punhciitioti
"-,.
86.2
93.6
89.9
(rug/kg well
(mg/kg dry)
(mo/kg dryl
(moykgdryj
Img/kgdry)
Advantages of Holzmann In Situ Soil Purification
Trm In iitu todtnlqui (MpvaUon a Kit and polluUnt In atal)
Mum 01 porMtd origlnu toll In titu
SlmunmtomjrounAwiwpurtiicttton
Whofc Mpwuor unl| In • ttigl* conttliw franw
Nodlil«)t«lw^to^-^i I purification
r i
u.lO Advnntaqos of Holzmann in situ soil purification technique
-------
BIOLOGICAL REGENERATION OF CONTAMINATED SOIL
Volker Schulz-Berendt
Umweltschutz Nord GmbH & Co
West-Germany
1. Introduction
The clean-up of contaminated sites is a matter of topical interest
in the area of environmental conservation. Not only as a result of
the shut down of industrial plants, such as refineries, service
stations and terminals, but also due to accidental spills it is
becoming ever more important to find an environmentally- and econo-
mically reasonable method for the clean-up of contaminated soil.
The disposal of contaminated soil on dumps can be considered only
as temporary solution to the problem.
The biological regeneration is considered to be the better alterna-
tive to special waste land disposal for volumes from some hundreds
up to several thousand cubicmeters. This process should be pre-
ferred in cases where suitable areas in the immediate vicinity of
the contaminated site or in its neighbourhood are available and
hydraulic clean-up methods cannot be applied for technical reasons
e. g. very low permeability of the soil.
Since the biological degradation of organic compounds leads to a
valuable product which can fulfill its functions as soil again,
this is an ecological valuable method for the clean-up of contami-
nated sites.
The bioremediation of contaminated soil by microbiological degra-
dation depends on the ability of bacteria and fungi to utilize
contaminants as sources of energy and nutrients. It has been well
documented that nearly all organic toxins can be broken down to
harmless substances by microbes.
Large scale implementation of microbial cleaning techniques has
been extremely successful in the treatment of mineral oil spills.
Further successes.have been observed in the microbial degradation
of aromatic and chlorinated hydrocarbons and polycylic hydrocar-
bons.
157
Bergedorfer StraBe 49 • 2875 Ganderkesee 1 • Telefon (04222) 010 22-10 27 • Telefax (04222) 25 03
-------
2. Approach
The biological degradation-process depends on two requirements:
1. The presence of microbes with appropriate metabolic
potentials. i
2. Suitable conditions for microbial life and activity in
the soil.
On-site and off-site microbial soil regeneration by the TERRAFERM
intensified degradation method is designed to realize these assump-
tions.
Before the beginning of any soil reclamation the ground is analyzed
for contaminant contents, nutrient levels and soil structure. Next,
the enzymatic turnover potential, the actual microbial activity ariid
the microbial colonization are calculated into a microbiological
diagnosis. Based on these results, the most appropriate optimiza-
tion program for maximal contaminant degradation can be selected
and installed. Simultaneously, microbes specially adapted to conta-
minants are isolated from the soil, carefully examined for suita- <
bility, and used as appropriate in the optimization process. j
The sorted and classified soil is then subjected to extensive pre-'
paratory procedures. Large stones and cement blocks are crushed. !
Organic substrates are added to improve the soil structure. Mineral
nutrients and trace elements are added to support the microbial |
population. Finally, the soil is cultured with the adapted bacteria
and fungi under conditions of intensive oxygenation. !
Oxygen is introduced to the system through intensive soil aeration.
In special cases anoxygenic conditions are needed for maximum de- i
gradation or oxygen carriers like nitrate or hydrogen-peroxide
can be added.
The biological breakdown of toxins takes place in a totally en-
closed dynamic fermentation system, in which all parameters, such
as temperature, oxygen content, nutrient levels and microbial popu-
lations can be maintained at their optimum levels. Volatile pollu-
tants are contained under a specially designed air-discharge bio-
filter. Leaching water is avoided through careful controls and by
preventing rain water from entering the system. Thus, the conta-
minants do not escape to the environment.
158
Bergedorfer StraBe 49 • 2875 Ganderkesee 1 • Telefon (04222) 010 22 -10 27 • Telefax (04222) 25 03
-------
3. Results
Some of the degradation results are shown in the added charts.
Hydrocarbons are degraded within 4 to 8 months, depending on the
type and amount of mineral hydrocarbon pollutants. The residual
concentrations lie within naturally occuring ranges.
Other contaminants like Polychlorinated Biphenyls (PCB's)
need special conditions for optimum degradation. We found that com-
plete breakdown of chlorinated compounds only take place if anoxy-
genic steps are involved (see graph).
To guarantee these conditions a soil-fermenter with a filling-
volume of 200 cubicmeters was constructed to adjust the oxygen-
level of the soil exactly.
After regeneration the cleaned soil is testet vigorously chemically
and biologically. Aside from measurements of contamination, other
characteristics are measured, such as particle size, humus content,
water content potential, soil flora and fauna (as well as the
ability to support higher plant life), the absence of weeds, and
its hygienic suitability for an appropriate future application.
The entire process, from collection of the contaminated soil to
delivery of the cleaned soil is under constant biological and
chemical supervision. This assures that dangerous residues are not
forgotten, and that the prescribed limits are not surpassed.
159
Bergedorfer StraBe 49 • 2875 Ganderkesee 1 • Telefon (04222) 010 22 -10 27 • Telefax (04222) 25 03
-------
S8888-r
48888
35888 -F'
3888a|
ppa
hydrocarbons 25888 -'•
(dry basis)
28888 -f
large!
voluo'
peak 8
values
TERRAFERM BIOSYSTEM-SOIL
Degradation of hydrocarbons
old
contamination
PCB
Degradation Optimization
Sum
if.-
la-
s'
6-
4-
R-
£
s
\
, 4
. \
-* — "
x0--
. *•'
. L, 1 1 1
1
— +-"''
'"*..
~^—^
PCB-Hix
Standards Nr.
28,52,181,
138, 153, 188
j — i — i — i_i._| ...i
5 18
>v
*'*--X-*"""
V. *S_
v\x"
x^
[ !_ . | i r
15 2
-Oj.
»0p
_. . -<0i
5 ^1
<-.
^^^^_
•o*
8 25 38 3
\
tim. II !
I
|
tim III !
i
!
tim 10 '
tim. I
tim. 0 ']
Prscnp SCRD cuaa k cc
hi. :Rsp Germany ,
5 I
Time (days)
Bergedorfer StraBe 49 • 2875 Ganderkesee 1 • Telefon (04222) 010 22 -10 27 • Telefax (04222) 25 03
160 :
-------
Ecotechniek
SEVEN YEARS EXPERIENCE
IN THERMAL SOIL TREATMENT
Rudolf C. Reintjes
Gees Schuler
Ecotechniek bv,
Utrecht, The Netherlands
Forum on Innovative Hazardous Waste Treatment Technologies:
Domestic and International
June 1989
Atlanta/ Georgia
161
-------
Ecotechniek
SEVEN YEARS EXPERIENCE
IN THERMAL SOIL TREATMENT
1.
INTRODUCTION
Ten years ago, the word soil contamination was not yet known
in the Netherlands. There were of course some stories about
dump sites in the United States, but that was far away for
us. Suddenly, the existence of chemical wastes beneath
dwellings was reported. But it was thought to be a single
and exceptional case,, and with 100 million dollars the whole
problem could be disposed of. However, within a short time
the number of reported cases of contaminated soil in the
Netherlands, had already risen to one case every 6 km2.
i
As one of the main companies in Holland in civiil
constructions and particularly in road construction; Royal
Volker Stevin has been endeavouring since 1979 to solve this
problem not i
only in theory, but also in practice, and, in practice means
I
for us, amongst other things, in such a way that the costs
can be justified. For this purpose a special firm with the
name Ecotechniek was founded. In 1981 the first pla'nt for
soil cleaning was put into operation, and since 1986 a
second plant has been providing service. In the near future,
several of these plants will be used in Europe. ,
162
-------
Ecotechniek ^
&
2. REQUIREMENTS FOR A DECONTAMINATION PLANT
What conditions had to be fullfilled?
In most cases several contaminating substances are present
besides each other. Therefore a cleaning technique must be
universal.
The degree of cleaning must be such that the decontaminated
soil can be re-used without restrictions.
A cleaning operation should not take too much time.
An output of 25-50 tonnes an hour is therefore necessary.
No new problems must be caused. This means therefore that
residues are to be avoided.
We needed two years for the development of a process and the
construction of an appropriate plant. In ,the following seven
years we gathered a good deal of experience, and realised an
enormous number of improvements. During this period
decontamination of 600,000 tonnes of contaminated soil has
been carried out.
PROCES DESCRIPTION
Background and concept of the system can be explained with
reference to the schematic drawings of figure 1 and 2.
Contaminated soil as a rule has a low organic material
content, on avarage about 1 per cent. Therefore contaminated
soil cannot be cleaned economically in a normal plant for
incineration of chemical waste. A new concept for thermal
treatment had to be developed. We opted for a two-phase
system. During the first phase (figure 1), the contamination
is converted into gases by heating the soil, if necessary up
to 550°C.
163
-------
Ecotechniek
Phase I
soil+contaminants
evaporated
contaminants
decontaminated soil
Figure 1 :
The developed vapours are separated from the soil, so that
the soil left behind, is clean again. This soil i$ cooled
down with water and can then be re-used normally, for
example in the site where it came from.
Phase II
gaseous
contaminants
clean flue
7\ gases
\/
energy
Figure 2
During the second phase/ the gaseous contamination is
destroyed (figure 2). Therefore an afterburning kiln is
used. The gases and vapours are burnt with additional oxygen
at a temperature of 900°C to 1100°C, or at even higher
•temperatures if necessary. The destroyed vapours are freed
from dust and passed into the stack. In the kiln, of course,
a large amount of heat is generated. This heat is utilised
for the evaporation of moisture and contaminants
first phase.
164
in the
-------
Ecotechniek
&
The following scheme therefore arises.
oxygen energy
energy
contaminated
soil
cleaned soil
cleaned flue gases+
residual energy
Figure 3
It will be clear to you that the special technique lies in
the evaporator, which is designed like a rotary kiln partly
with direct firing and partly with indirect heating by means
of the flue gases of the afterburner, and, of course in the
combination of the various parts of the total plant.
4. COMPARISON WITH OTHER SYSTEMS
As stated, all poisonous, evaporable organic content of the
soil is disposed of, regardless of its chemical structure,
quantity and physical state. All types of soil can be
processed, although processing is not always equally easy.
The fact that the chemical composition, concentration and
type of soil don't need to be taken into consideration
distinguishes this process for example in comparison with
the biological process. The various washing methods are very
sensitive to fine fractions of the soil.
The thermal method, on the other hand, does not eliminate
heavy metals and other inorganic substances, although the
widespread inorganic cyanide can nevertheless be eliminated
very well.
165
-------
Ecotechniek
DIFFERENT CATEGORIES OF CONTAMINATES
heavy tetals
non volatile
chlorinated
organic
substances
volatile
chlorinated
organic
substances
cyanides
alifatic and
aroiatlc
hydrocarbons
polycycllc aroiatlc
hydrocarbons
Figure 4
So, a thermal cleaning method does not solve every problem,
but the majority of them as you can read from the diagram in
figure 4 which describes the distribution of substances in
polluted soils in the Netherlands. At least 70 % can be
treated thermally. With -respect to thermal methods other
than the one described in this paper, it must be stated that
most of them use a higher, or even much higher, soil
temperature. As a result, a lot of energy is consumed
unnecessarily, and the normal organic substances peculiar to
the soil, such as humic acids, are destructed too.
Hence the soil is not soil any more, but organic free sand
which is hardly of interest for re-use purposes.
Due to integrated technology, the plant is very small in
relation to its throughput. This allows the construction of
semi-mobile machines with a high capacity, an advantage over
other systems.
166
-------
Ecotechniek
5.
CLEANING RESULTS
It is clear that for low-boiling substances, only low
evaporation temperatures are required. Table I shows the
required temperatures.
In practice, slightly higher temperatures are used. In this
table, the required temperatures in the afterburner are
listed too.
Table
required temperature
contaminant ^^"">"""—— ^^^
petrol, diesel oil
benzene, toluene, xylene,
naphthalene ,
polycyclic aromatic
compounds
cyanides
evaporation
oC
200-300
200-300
450-500
450
destruction
oC
750
-S.
800, .
850
950
Indicative for the cleaning performance of the system are
the residual concentrations of PAH's because they are
difficult to eliminate and the required residual
concentration is low.
As can be seen from figure 5 residual concentrations could
be reduced in time due to growing experiences in time.
167
-------
10000
1000
1 100
jl 1008
t f 1IX>
10
1
t ,
H
1
DECONTAM
polyar
L_J_
INATIC
omatic
xnlamin
)NRE,
hydroc
oedtoa
SULT5
arbon
> OF ETTS
s
• fka up B 1 2000 me*fl
mlntnlot
ifler decontamination
Innft a itlutt
otXUnK)
I. t^:
! I
H
j
r j
1
1 8 ttindvd — — — —
f — K
•»• — Aiund»rt— —
fM, 1IW IMS 18W 1M6 1M9 1M9 1889
Figure 5
Residual concentrations can be reduced to more or less the
level of natural soil (the Dutch A-level), although
pollution degrees of processed soils are much higher than at
the beginning.
In figure 6 numbers are given for cyanides, a rather often
occuring pollutant in the Netherlands, an inorganic and very
stable substance as you know.
It seems that the realised resulting concentrations for
cyanides have grown through the years, but have a look at
the concentrations which are processed.
Konlnklljke
Voivei-stevln
MOO
1000
fi
If
100
soo
so
DECONTAMINATION RESULTS OF ETTS
cyanides
contaminated soil
after decontamination
yur
(RfKI
IMS
1IN
1M7
168
nng«o
-------
I must point out that the residual concentrations can be
reduced as much as you want, but that the price to be payed
for that rises more and more rapidly.
PCB' s and dioxins have not been mentioned up till now.
Contamination with these substances is not widely spread in
the Netherlands.
It is therefore only recently that experiments on the
decontamination of soils containing these substances have
been started. No problems are to be expected with PCB anyway.
The boiling points are low, and destruction is fairly easy.
The residual concentrations to be obtained are around 1 ppm,
and therefore fairly easy. It is just a matter of gathering
results of tests and take additional safety measures during
execution.
Dioxins have to be reduced to significantly lower residual
concentrations and the destruction temperatures are higher,
as you know. In four months time, tests will be finished
with which it will be proved that dioxin cleaning is
possible with this system.
6.
EMISSIONS
As already mentioned at the beginning, the process must not
lead to any pollution of the environment. The only things
which are released are flue gases. These must therefore be
freed from dust, SC>2, HC1 and HF (if any present).
In particular S02 and HC1 may be released if sulphur and
chlorine containing contaminants are present. Depending on
the requirements set with respect to the contaminants to be
eliminated and depending on the local regulations on
emissions, a flue gas cleaning system is selected.
This means therefore that the plants which are in operation
and those which are currently being built all have different
flue gas cleaning systems. The concentrations of substances
in the gascleaner effluent are that low that the effluent
can be used to cool the decontaminated soil. So the system
has vertually no residual waste streams.
&
169
-------
7.
PLANTS
Ecotechniek operates, as stated, two plants, and a> third one
is under .construction that will be in operation in Germany
.within a few months. The illustration below will give you an
impression of the process. '
THERMAL TREATMENT PLANT E.T.T.S
The soil is sieved and crushed in a preparatory line, to a
material size of less than 40 mm. Then it is partially
heated indirectly in a rotary kiln. This indirect heating is
done with the hot gases of the afterburner. The soil is
heated further directly by an oil or gas burner. The cleaned
soil is led into a mixer and cooled and moistened with
water. The gases from the kiln are passed to an afterburner,
in which fresh air and energy are supplied. Special
attention is given to energy recovery from the flue; gases of
the afterburner. The plant has three energy recovery
systems, including the rotary kiln. ;
Furthermore three dust collectors are incorporated, one of
which works by a wet process which regulates the pH value of
the gases before they leave the stack.
&
KV,o[rH!!Vr. 1 7H
1 / U ;
-------
8. COSTS
It is rather difficult to estimate the costs of
decontamination of soil with this process in the United
States of America.,
The costs are not only caused by the process itself but also
to a large extend by the moisture content and pollutant
content and by safety measurements and environmental
regulations.
Moreover we cannot transform our prices simple by exchange
rates.
So we can give only a very rough indication. We think that
the processing of contaminated soil in the United States
with this system will not cost more than $ 200/ton of wet
soil.
171
-------
AN OGOEN COMPANY
OGDEN
ENVIRONMENTAL SERVICES, INC,
CONTAMINATED SOIL REMEDIATION BY
CIRCULATING BED COMBUSTION
BY
ROBERT G. WILBOURN
BRENDA M. ANDERSON
This is a preprint of a paper to be
presented at the EPA Forum on Innovative
Hazardous Waste Treatment Technologies.
MAY 1989
172
-------
ABSTRACT
The Circulating Bed Combustor (CBC) is an advanced generation of
incinerator that utilizes high velocity air to entrain circulating solids in a
highly turbulent combustion loop. Because of its high thermal efficiency, the
CBC is ideally suited to treat organic wastes with low heat content, including
contaminated soil. This paper discusses the development of CBC contaminated
soils treatment technology and its application to site remediation. The CBC
process, pilot plant, and transportable field equipment units are described.
In March of 1989, a Superfund Innovative Technology Evaluation (SITE)
demonstration test burn of McColl Superfund site soil was conducted in Ogden
Environmental Services' (OES) Circulating Bed Combustion research facility.
The results of the successful test are presented herein. The paper also
reviews the on-going site remediation activities in Alaska and California using
OES designed, fabricated and deployed transportable CBC units.
173
-------
CBC PROCESS DESCRIPTION
The CBC is ideally suited to treat feeds with low heat content, including
contaminated soil. Figure 1 shows the major components of a CBC configured for
soil treatment. Soil is introduced into the combustor loop at the loop seal
where it immediately contacts hot recirculating soil from the hot cyclone.
Hazardous materials adhering to the introduced feed soil are rapidly heated and
continue to be exposed to high temperatures throughout their residence time in
the ceramic lined combustor loop. High velocity air (14 to 20 ft/s) entrains
the feed with circulating soil which travels upward through the combustor into
the cyclone. Retention times in the combustor range from 2 seconds for gases
to ~30 minutes for larger feed materials «1.0 in.). The cyclone separates the
combustion gases from the hot solids, which return to the combustion chamber
through a proprietary, non-mechanical seal. Hot flue gases and fly ash that
are separated at the cyclone pass through a convective gas cooler and on to a
baghouse filter which removes the fly ash. Filtered flue gas then exhausts to
the atmosphere. Heavier particles of purified soil remaining in the lower bed
of the combustor are removed at a controlled rate by an ash conveyor system.
As a consequence of the high turbulence in the combustion zone, temperatures
around the loop (combustion chamber, hot cyclone, return leg) are uniform to
within +50°F over the typical operating range of 1450 to 1800°F. The uniform
low temperatures and high solids turbulence also help avoid the ash flagging
that is generally encountered in other types of incinerators. i
i
Acid gases formed during destruction reactions are rapidly captured in the
combustor loop by limestone that is added directly into the combustor with the
feed. HC1 and S02 that are formed during the combustion of chlorine and sulfur
bearing wastes react with limestone to form dry calcium chloride and calcium
sulfate, (benign salts). Due to the high combustion efficiency attainable in a
CBC, an afterburner is not needed. In more than 90% of the cases studied to
date, post-combustor acid gas scrubbing is not required. Emissions of CO and
NOX are controlled to low levels by the excellent mixing resulting from the
turbulence, by the relatively low temperatures and by staged combustion that is
achieved by injecting secondary air at locations ascending the combustor.
174
-------
COMBUSTION
CHAMBER
LIMESTONE
FEED
SOLID
FEED
COOLING
WATER
FLUE GAS
(DUST)
FILTER
STACK
ASH CONVEYOR
SYSTEM
Fig. 1. Schematic Flow Diagram of Circulating Bed Combustor
for Soil Treatment
175
-------
Because of these design and operating features, the CBC can attain required
DREs for both hazardous wastes (ORE 99.99%) and toxic wastes (DRE 99.9999%) at
temperatures below those used in conventional incinerators which typically burn
at temperatures greater than 2000°F.
The Circulating Bed Combustion technology is well developed and is being
applied on two contaminated soil site remediation projects that will clean over
80,000 tons of contaminated soils. Ogden Environmental Services, Inc. (OES) and
its predecessors have pursued a systematic technology development and an
applications approach comprised of the following elements:
• Definition of treatable soil contaminant wastes types
• Fifteen years of fluidized and circulating bed pilot plant testing
• CBC performance demonstrations in private and governmental programs,
e.g., the Superfund Innovative Technology Evaluation (SITE) program
• Extensive permitting activities :
• Design, engineering, fabrication, deployment, and operation of
modular transportable CBCs for hazardous waste site cleanups
OES CIRCULATING BED COMBUSTOR UNITS
Research Facility
OES' research Circulating Bed Combustor (CBC) is the heart of an
integrated, highly flexible waste combustion demonstration facility located in
San Diego, CA. Initial CBC-soils treatment development and engineering studies
were carried out in the 16-in. (i.d.), 2 million Btu/hr CBC. The test data
obtained were used to design the larger, transportable CBCs. The configuration
of the research CBC is shown schematically in Figure 1. Figure 2 is a
photograph of the 16-in. CBC unit. The 16-in. CBC design features and
capabilities are summarized in Table 1. :
176
-------
Fig. 2. 16-in. CBC in the San Diego Research Facility
177
-------
TABLE 1
Combustor size
Inside diameter, in.
Height, ft
16" - DIAMETER CBC DESIGN FEATURES
AND EQUIPMENT CAPACITIES
16
27 '-'• •
Normal operating temperature, °F
Maximum outside surface temperature, °F
Auxiliary fuel maximum thermal output,
MMBtu/h
Soil feeder maximum throughput, Ib/h
Sorbent feeder maximum throughput, Ib/h
Bed ash removal system maximum throughput,
Ib/h
Liquid feeder maximum throughput, gpm
Continuous on-line flue gas analysis
Baghouse filters
Total surface area, ft2
Number of filters
Flow capacity, acfm
Maximum operating temperature, °F
Forced-draft fan
Maximum pressure, psig
Maximum flow, scfm
Induced-draft fan
Maximum vacuum, in. w.g.
1400 to 1800 ,
200 '
2
2000
200
2500
30
0 ico 25% oxygen
0 to 25% carbon dioxide
0 to 1000 ppm carbon monoxide
0 to 1000 ppm nitrogen oxides
0 to 1000 ppm unburned hydrocarbons
0 to 1000 ppm sulfur dioxide
550 .'...,..
50
800
375 '.
12
500
50 (at 800 acfm)
178
-------
Transportable CBC's ,
The transportable 36-in.; (i.d.), 10 million Btu/hr CBC consists of seven
structural steel modules that contain the process equipment and provide the
structural framework of the CBC. The modules do not exceed measurements of 8
feet 6 inches wide, 10 feet 4 inches high, and 35 feet long. As a result, the
modules can all be transported on single drop trailers that do not require
special highway transportation permits. The CBC cyclone and combustor are
mounted to the top of one of the structural modules. When erected, the
transportable CBC itself sits on a pad of 30 by 50 feet and is approximately 60
feet in height. In field operations, the transportable CBC's are incorporated
in a complete system layout which includes ancillary equipment units and
transportable buildings, e.g., a control room, a motor control center, an
analyzer room and a chemistry support laboratory (optional). Figure 3 is a
schematic drawing of the transportable CBC. Figure 4 is a photograph of a
field assembled transportable 36-in. (i.d.) CBC unit.
CBC RATE THROUGHPUTS
CBCs have been constructed and operated with thermal ratings that range
from 2 million Btu/hr to 50 million- Btu/hr. The soil throughput of these units
varies in response to the moisture content of the feed and to its energetic
value. Figure 5 compares the relative throughputs of several CBC sizes for
various feed streams.
MEDIA TREATED
Circulating Bed Combustion is widely applicable to many hazardous waste
forms. Solids, e.g., contaminated soils, liquids and sludges are treated with
equal facility by using the appropriate feeding systems. OES has conducted
extensive pilot plant and field unit testing on soils contaminated with
hydrocarbons and chlorinated hydrocarbons.
179
-------
Fig. 3. 36" Diameter Transportable CBC Schematic
180
-------
Fig. A. Field Assembled Transportable CBC Unit
181
-------
LEGEND
•
D
•
O
A
A
WASTE DESCRIPTION
PCB CONTAMINATED SOIL
PCB CONTAMINATED SOIL
CHLORINATED CHEM. SLUDGE
CHLORINATED CHEM. SLUDGE
CHLORINATED LIQUID WASTE
01 LAND SOLVENT WASTE
WATER
CONTENT
%
10
20
80
40
4
13
HEAT
CONTENT
Btu/LB
0
0
1,331
. 4,000
7,606
11,227
THROUGHPUT (LB/HR) VS. COMBUSTER ID.
16 IN.
1,260
930
440
340
210
130
24 IN.
2,840
2,080
1,000
770
470
280
36 IN.
6,380
4,680
2,250
1,740
1,060
740
48 IN.
11,340
8,320
4,000
3,100
1,880
1,140
60 IN.
17,720
13,010
6,250
4,840
2,940
1,780
104 IN.
53,240
39,130
18,770
14,530
8,840
5,340
100,000
10.000
=>
O
K
I
1,000
100
10
COMBUSTORIDdN.)
100
200
Fig. 5. CBC Unit Size Throughput Comparisons
182
-------
PROCESS WASTE STREAMS
The CBC process typically produces solids (i.e., bed and fly ash) and
stack gas, as shown in Figure 1. The composition of the stack-gas system
effluent must meet stringent EPA and other governmental requirements in
accordance with permitted conditions. All the CBC incineration tests of
contaminated .soils verify that the purified soil treated by the CBC is
non-hazardous with respect to organic residuals. Since most metals report to
the ash during combustion, the disposition of ash is specific to each waste
feed case and must be determined on an individual basis. For most
organic-contaminated soil sites, the ash produced by the CBC meets the criteria
for redeposition on site. Post-combustion fixation processes may occasionally
be required if the ash metals content or leachability exceeds permissible
levels.
RECENT OES TEST PROGRAMS WITH CONTAMINATED SOIL FEEDS
Su-perfund Innovative Technology Evaluation (SITE) Program
In 1986 OES's CBC technology was selected by EPA for a demonstration under
the SITE program. Contaminated soil from the McColl Superfund site in
Fullerton, California was selected as the waste feed for the demonstration.
Due to multiple delays encountered in the securing all of the required permits,
it was not possible to conduct the planned feasibility demonstration test until
this year.
The treatability study was conducted during the week of March 27, 1989 in
the combustion demonstration facility. The study covered approximately 31 hours
during a four-day period. The study was monitored by EPA, the California
Department of Health Services, and the San Diego County Air Pollution Control
District. A test profile is given in Figure 6. A total of 7,500 pounds of
contaminated soil was processed through the CBC, of which A,700 pounds were
actual McColl waste. The materials that were processed included: unblended
waste, waste blended with clean sand, and unblended waste "spiked" with carbon
tetrachloride (CCI^). The materials were processed without difficulty.
Samples of the waste feed, fly ash, bed ash, and stack gas were taken by an EPA
183
-------
STARTUP/
HEAT UP
THERMAL
EQUILI5RUIM
STABILIZE ON
H<=COLL FEED
SAMPLING
WINPOW
5Y5TEM
IDLE-
I
SYSTE-M
IPLE-
^u
&.
D
cfl
_j
<
h
^
8
(-
D
X
lO
PROTOCOL SAMPLING
\
\
\ 400 L6/HR
±j>o
12 24 36 46 60 72 34
MOKI
TUE
THU G.
Fig. 6. "McColl Treatability Study Test Sequence
184
-------
contractor for extensive analysis. The samples are being analyzed for organic
compounds, (including dioxins and furans), metals, criteria pollutants, and
physical properties.
EPA has officially released preliminary data which has been checked to
assure that it meets EFA standards and the complete demonstration test report
will be available in August 1989. The results show organic material was
effectively destroyed as exhibited by the destruction and removal efficiency
(DRE) value shown in Table 2. No significant levels of hazardous organic
compounds left the CBC system in the stack gas or remained in the bed and fly
ash material, as shown in Table 3, which presents information on feed and
residuals characteristics. The criteria pollutant and acid-gas release data
obtained meet federal, state and South Coast Air Quality Management District
(SCAQMD) requirements. Table 4 contains data on stack criteria pollutant and
acid-gas emissions. Table 5 presents stack gas particulate loading data that
are significantly below the federal requirement of 0.08 grains/dry standard
cubic foot (corrected to 7% oxygen). The average particulate loading, however,
is higher than the SCAQMD requirement of 0.002 grains/dscf. A Toxicity
Characteristic Leaching Procedure (TCLP) test was performed on the McColl CBC
ash. Arsenic, selenium, barium, cadmium, chromium, lead, mercury and silver
leachabilities were found to be low and meet or exceed the federal requirements
(40 CFR Part 268).
The EPA has concluded the test was successful based on the available
data.
Transportable CBC Unit Field Demonstration Test
OES conducted a. PCB-contaminated soil demonstration test burn at our
Swanson River, Alaska remediation site project in September 1988, in accordance
with a test burn plan prepared by OES and approved by the EPA Office of Toxic
Substances. The test burn was conducted under witness of the EPA and the
Alaska Department of Environmental Conservation (ADEC) at the remote Swanson
River Alaska site on the Kenai peninsula. All required performance criteria
were met, and OES has been granted a nation-wide PCB Disposal Operating Permit
for its transportable 36-in. (i.d.) CBC unit.
185
-------
TABLE 2
SITE/McCOLL TEST
DESTRUCTION AND REMOVAL EFFICIENCY (ORE)
Feed Rate
Stack Emissions
ORE
0.203
0.000017
99.992Z
pounds per hour
pounds per hour
average values using Carbon Tetrachloride
as a performance indicator
186'
-------
TABLE 3
SITE/McCOLL TEST
FEED AND RESIDUALS CHARACTERISTICS
Waste Feed
Fly Ash
Bed Ash
Stack Gas
Emissions
ORGANICS (parts per million)
Benzene
Toluene
Xylene
Ethylbenzene
1,1,1, Trichloroethane
Naphthalene
2-Methyl Naphthalene
PHYSICAL PROPERTIES
4.9
35.5
165.0
23.0
not detected
30.2
34.5
not detected
not detected
•not detected
not detected
not detected
not detected
not detected
not detected
not detected
not detected
not detected
not detected
not detected
not detected
0.0008
0.0015
0.0015
0.0004
0.0002
0.0006
0.0004
Sulfur (%)
PH
Density
(pounds per cubic foot)
Heat Value
(Btu/ pound)
4.4
2.3
57.9
1387.0
3.6
12.6
76.9
—
0.9
12.1
88.4
- —
—
*™"~
— ~
'not detected" indicates a value below detection limits.
Organic feed and heat values are based on unblended waste averages. All other results are
based on blended and unblended waste averages. Waste feed, fly ash and bed ash values are
weight/weight. Stack gas emissions are volume/volume.
187
-------
TABLE 4
SITE/McCOLL TEST
AIR EMISSIONS
(average values in pounds per hour)
Sulfur Oxides
Nitrogen Oxides
Carbon Monoxide
Total Hydrocarbons
Hydrochloric Acid
(SOX)
(NOX)
(CO)
(THC)
(HC1)
<2.0*
0.34
0.12
0.007
<0.01
Further quantitation not possible due to S02 analyzer insensitivity in
lower range.
188
-------
TABLE 5
SITE/McCOLL TEST
PARTICULATE LOADINGS
(average values)
gr/dscf
mg/dscm
Ib/hr
Test Results
0.0038
8.72
Oo0242
Permit Limits
0.08
180.0
not specified
gr/dscf grains per dry standard cubic feet
corrected to 7Z oxygen
mg/dscm milligrams per dry standard cubic meter
189
-------
USE IN THE U.S.
The CBC technology has been recently applied in large scale at two site
remediation projects that will treat over 80,000 tons of contaminated soil.
OES has proven the effectiveness of transportable CBCs by locating and
operating them cost-effectively in demanding environments. Every regulatory
requirement for site operations has been met, and OES is consistently in
compliance at both sites. The transportable CBCs have been operated in weather
as cold as -40°F and as high as 110°F. The ruggedness of the units has been
demonstrated by mobilizing and operating successfully in a remote and
ecologically sensitive wildlife refuge. OES has maintained high levels of
availability through the use of careful logistics planning that includes design
factors, maintenance, and supply planning. A descriptive summary of OES' two
major projects is given below.
PCS SITE REMEDIATION PROJECT
The ARCO Alaska Swanson River site is located within the Kenai Wildlife
Refuge. PCS contamination was identified during site soil sampling conducted
by the U.S. Fish and Wildlife Service in 1984. The contamination was an '
indirect result of a compressor explosion which occurred in 1972. Remediation
activities were initiated by a voluntary "Order by Consent" signed by the site
operator. The remediation is being conducted under the direction of the U.S.
Fish and Wildlife Service, the Bureau of Land Management, and the Alaska
Department of Environmental Conservation, with the concurrence of the U.S.
Environmental Protection Agency Region X.
OES was selected for the remediation project because of demonstrated PCS
destruction capabilities and the CBC combustion technology. OES' comprehensive
site remediation workscope includes mobilization, on-site demonstration
testing, excavation, contaminated vegetation clearing, incineration,
contaminated water treatment, concrete and metal surface decontamination,
demobilization, and site restoration.
190
-------
The completion of the Swanson River project is scheduled for the end of
1991. Upon completion, over 70,000 tons of PCB-contaminated gravel/silt soil
will have been treated. Figure 7 is a photograph of the Swanson River site.
FUEL OIL SITE REMEDIATION
For more than 50 years, a leaking underground storage tank at a cannery in
Stockton, California contaminated surrounding clay soil with No. 6 fuel oil.
OES was contracted by the site operator to remediate the site using one of its
transportable 36"-in. (i.d.) CBCs. OES developed and is implementing a
remediation plan that encompasses site characterization, demolition of tanks
and buildings,, installation and operation of water intercept wells, water
treatment, soil excavation, stockpiling, CBC treatment, placement of slurried
purified soil; and site and building restoration.
The excavation and backfilling is complete and the CBC thermal treatment
of stockpiled soils is approximately 30Z complete (June, 1989). Upon
completion of the project later in 1989, over 11,000 tons of contaminated soil
will have been treated. Following restoration, the site will have its full
commercial value restored and it will be available for unrestricted use.
Figure 8 is an overview photograph of the Stockton project site.
TREATMENT COSTS
OES offers a comprehensive range of services for the disposal of hazardous
wastes, including pilot plant testing services, permit application and
responses to reviews and comments, economic and technical evaluation of waste
disposal alternatives, engineering design of CBC plants to meet customers'
needs and specific waste characteristics, supply of a mechanically complete
commercial CBC unit and supervision and staff for operation of the unit.
Site remediation costs are divided into three categories. The first
category includes the direct and indirect costs for engineering design, base
equipment cost, materials, foundation and installation labor to erect a
191
-------
Fig. 7. Swanson River Project Site, Aerial Photo
(Fall - 1988)
192
-------
WATER
TAbte
RESERVOIR LEVEL
I 101'T
DEPTH)
INTERIM
ASH STORAGE
AMP
PROJECT
TRAILERS
UNIT
C&C-TIS
FEED UNIT
EVCAVATIOKl
'SITE RECEPTION)
TRAILER
STOCKPILE
•SOIL
COPWER S-ITE
OF
HOU3&
WATER
FACILITIES
WELL
SYSTEM
STORAGE
5HEP
EQUIPMENT
DECONTAMIN AJIOKI
FACILITY
•STOCKPILE
6ERM ANP
LINER
\ / ^
V-'
CORNER
5UHP
PUMP5
r
\
\
PIPE TO V
HAKJOLISIG
Fig. 8. Stockton Leaking Underground Storage Tank
Remediation Site Photo
-------
mechanically complete CBC. The second category includes CBC costs consisting
of labor, materials, utilities, repair and maintenance, and indirect costs.
The third category includes material handling operations including excavation,
feed processing, and ash disposal.
Costs for CBC soil remediation typically range from $100-$300 per ton
depending primarily on soil moisture content and the quantity of wastes to be
processed.
SUMMARY
OES has developed CBC waste treatment technology and demonstrated its
applicability in private and government sponsored programs including the
Superfund Innovative Technology Evaluation (SITE) program. Based on this
systematic development and testing, modular CBC units have been designed,
fabricated and deployed. CBC treatment is being utilized in two large
remediation projects. Treating contaminated soil in a CBC is cost-effective,
highly efficient, and meets all performance and operation criteria established
by regulatory agencies. The Ogden Corporation, OES' parent company, has made a
major commitment to the site remediation business with four units soon to be
operating.
194
-------
Residues from ;:|
High Temperature
Rotary Kilns and
Their Leachability
*»*:• '.to
HELD AT THE ERA CONFERENCE INiArNlA
JUNg 19-21,1989 j
195
-------
Residues from High-Temperature Rotary Kilns
and Their Leachability
by
Ronald Schlegel
Held at the EPA Conference in Atlanta (June 19 to 21, 1909)
1.
Motives and Objectives
The characteristics of the residues generated by the
combustion of hazardous waste obviously vary depending
on where in the plant they were obtained. The largest
portion of solid residues (roughly 80 %) are obtained
immediately after the rotary kiln below the secondary
combustion chamber. In high temperature systems
slagging rotary kilns), these slags have the appearance
of black glass, and it seems that they also possess the
poperties of glass.
The remaining residues are discharged; in the form
either of boiler ash, filter cakes or filter dust,
depending on the individual plant design.
Since the largest portion of all residues are obtained
in the form of slags, we wanted to investigate under
normal industrial operating conditions whether the
composition of the slags allows for normal land disposal
or even recycling.
We have been repeatedly approached in this matter,
particularly .by American firms who are interested in
this question given the much tighter economic
environment and the stringent liability regulations in
196
-------
the U.S. The slag leachability tests were consequently
carried out by an EPA licensed laboratory.
The object of these tests was on the one hand to
investigate the composition and leachability of the
slags and on the other hand to obtain experimental
results on the material flow under different waste
composition conditions.
2.
The Plant at Rilnmond (Netherlands)
The W+E rotary kiln process line of AVR-Chemie's
hazardous waste incineration plant at Rijnmond took up
operation in 1987 and now handles about 50,000 mt of
waste a year. The plant consists of a high-temperature
rotary kiln, a secondary combustion chamber, a steam
boiler to recover the waste-generated heat and to cool
down the flue gases prior to their treatment in a wet
flue gas treatment system (Figs. 1 to 3, Table 1).
Since its purpose is to handle hazardous waste from all
over the Netherlands, the plant has to cope with the
entire spectrum of hazardous waste, ranging from
unpacked and packed waste (200,000 barrels a year) to
contaminated soil, liquid waste of any composition,
sludges and clinical wastes. Moreover, ARV-Chemie is
the only plant in the Netherlands licensed to incinerate
PCB-containing wastes.
The Tests
The tests were carried out in cooperation with AVR-
Chemie and our US licensee Combustion Engineering in
September 1988.
197
-------
Fig. 4 and Tables 2 to 5 show the sampling and
measuring points that were used during the tests and
whose data were recorded so that reliable energy and
mass balances could be established.
The slag and ash samples were extracted every 15 minutes
and mixed to representative samples) then immediately
packed and dispatched to the US for the leachability
tests.
The specified and analysed menu was fed four hours
before the tests so that the plant would have reached a
steady state condition by the time a series of
measurements begins. This precaution ensures moreover
that the slags and ashes are actually generated by the
analysed waste samples. <
Due to the large variety of waste—particularly solid
waste—the plant has to process, we were forced to run
the tests on a trace method basis. We ensured that the
input contained a sufficient amount of the particular
waste material in which we were interested. Otherwise,
the plant was operated under normal commercial
conditions.
The following test runs of different waste flows were
executed (Table 6):
Test # 1
Without barrels; high organic content:
Apart from the "regular" contaminants,
mixture contained the following:
the waste
Chlorine:
Fluorine:
Sulphur compounds:
335 kg/h
60 kg/h
18 kg/h
198
-------
Test # 3
High content of inert matter containing PCB
PCB
14 to 15 kg/h
Test # 4
High me,tal. content (large number of barrels):
Apart from the "usual" materials, the waste mixture
contained the following:
Heavy metals
Class I"••'•••
(Hg, Cd, Tl)
Class II
(Ag, Co, Ni, Se, Te)
Class III
(As, Pb, Cr, Cu, Mo,
Zn, Rh, Va, Sn)
Total amount:
1,887 g/h
2,352 g/h
16,364 g/h
20,603 g/h
During all tests, the plant was operated under normal
operating conditions and did not deviate from the usual,
commercial operations. The operating parameters of all
three test runs are given in Table 7.
The leachability tests were carried out according to the
following three methods :
o Toxicitv Test (EPA SW 846, Method 1310^
Water is added to the sample and a pH of 5 . 0 ± 0.2
is maintained. Extraction Procedure Toxicity
(EP-TOX) analytes are given in Table 8.
Toxicity Characteristics Leaching Procedure
In contrast to the EP-TOX test, a special extraction
fluid is used. The maximum allowable concentrations
are shown in Table 9 .
199
-------
Total Extractable Metals
Here, the metals are solved by acids and analysed by
means of argon plasma spectroscopy.
4. Results (Figs. 5.1 to 6.3)
4.1 Material flow
The consideration of all tests revealed that 90 % of the
solids input volume is discharged in the form of slags,
whereas only 10 % verge into other output flows.
Analyses of individual heavy metal flows in the plant
revealed a number of interesting results, though there
is obviously nothing new to be said about mercury; 92 %
was determined in the scrubber (scrubbing liquid) and
7.2 % in the flue gas.
Much more interesting are the cadmium analyses since
they revealed a total output that was about 3.7 times
higher than the measured input. These tests proved once
again that the major part of all cadmium in the waste is
not present in its pure form, but in plastics, paints,
and other materials.
The analysis results showed that the nickel, cobalt,
chromium and copper contained in the slag are mainly of
a form that does not allow leaching.
4.2
Leachabilitv Tests
The leachability tests according to the TCLP Method show
that all values fall short of the threshold limits by at
least a factor of 100.
200
-------
5.
Conclusion
The TCLP tests carried out the slag samples obtained
from a high temperature rotary kiln incineration system
show that the leachability of heavy metals of this slag
is considerably lower than the threshold limits for
landfill disposal that are in force in the U.S.
The fact that all these results were achieved with
regular residues and under normal commercial plant
operation conditions indicates that there is a
probability that this slag may be delisted. Although
any delisting would have to take place on a case to case
basis, there are enormous potential advantages. For
instance, the sanitary landfill disposal costs for the
largest portion of all residues—up to 80 %—could be
drastically reduced and, even more important, the
potential liability risk of this slag is not comparable
to that of hazardous residues.
These tests are obviously just a beginning; further
programs will be required to build a data base on the
disposal of slag obtained from these incineration
systems. They did, however, prove again that the high-
temperature rotary kiln technology as applied in Europe
is the best method of incinerating large amounts of the
entire range of hazardous wastes. In Europe, such
incinerators have to process hazardous waste of entire
countries, or at least vast regions. Therefore, they
typically have large capacities of up to 50,000 mt and
handle e.g. 55 gallon metal drums, sundry package types
and other solid waaste, sludges, slurries, liquids,
aqueous waste, etc.
It is of course essential that the-residues leaving the
plant be not as toxic as the input and that the volume
be reduced. The tests proved that these requirements
201
-------
6.
were met, at least for the largest portion of these
residues.
Bibliography
(1) TCLP Procedure, Part II, Environmental
Protection Agency, 40 CPR Part 260 et al.
(2) EP-TOX (Method C004), U.S. Environmental
Protection Agency/Office of Solid Waste,
Washington, D.C., "Test Methods for Evaluating
Solid Waste-Physical/Chemical Methods," SW-846
(1980), Section 7.
202
-------
RIJNMOND FACILITY, NETHERLANDS
Fig. 1
ANLAGE RIJNMOND, HOLLAND
203
-------
RIJNMOND FACILITY
Fig. I
ANLAGE RIJNMOND
204
-------
903
Glas fibre reinforced stack
Might 90m
Wet scrubber
Flue gas fan
Electrostatic precipitator (ESP)
82,000 NmVh
Fly ash transport system
Steam boiler
Fly ash transport system
Secondary combustion chamber (SCC)
Residue discharging
Secondary air system
Rotary kiln (RK)
Charging hopper for solid waste
Primary air system
Barrel charging
approx. 30 barrels/h
Bunker for solid waste
1400m3
-------
Table 1
Technical Data of the Plant
Throughput
Heat release
t/h
MW
Mio BTU/h
Steam : Production
Pressure
Temperature
Rotary kiln :
Diameter inside
outside
Length
Secondary combustion
chamber:
t/h
bar
°C
m
m
m
Height (Axis RK to top of SCC) m
Width
Length
m
m
Design
Continuous
load
6.2
23.8
81.2
23.3
30
370
4.2
4.9
12.0
11.6
6.6
6.1
Peak
load
6.8
30.2
103.0
30
30
370
W+E Environmental Systems
206
-------
FEEDWATER
(T,,
ESP
TESTS
Due to the large variety of
waste - particularly solid
waste - the plant has to
process, we were forced to
run the tests on a trace
method basis. We ensured
that the input contained a
sufficient amount of the
particular waste material in
which we were interested.
Otherwise, the plant was
operated under normal
commercial conditions.
PIO) BOILER
(CONVECTION PART)
0 ©
W+E Environmental Systems
The leachability tests were carried out according to the
following three methods:
O Toxicity Test (EPA SW 846, Method 1310)
OToxicity Characteristics Leaching Procedure (TCLP)
O Total Extractable Metals.
W+E Environmental Systems
-------
SAMPLING £KD MEASURING
Table 2.1
Date: August 1988
P = Plant Ocoputer / Screen Prints
I WRSTE EEEDINS T _ yest ccoputer - WfE
L = Local / Rack
8 = Screen (hand rec.)
Pt.
1
2
3
4
System
No.
A-251
A-252
A-201
A-202
System
solids feeding
barrels feeding
burner for oil
no. 6 and high
calorific waste
lance for oil
no. 6 and high
calorific waste
Required
Info
weight
number of lifts
heating value
main exposition
weight
number of barrels
feeding time
heating value
content, compo-
sition
flow
temperature
heating value
composition
flow
temperature
heating value
composition
Instrument
Check Point
crane scale
counter
sample analysis
sample analysis
barrel scale
plant computer
plant computer
sample analysis
sample analysis
flow meter no.6/
waste
thermometer
sample analysis
sample analysis
flow meter
thermometer
sample analysis
sample analysis
Instr.
No. /
Resp.
^
-
lab. AVR
lab. AVR
W8350
-
-
lab. AVR
lab. AVR
FI 8001/
8002
TT 8001/
8002
lab. AVR
lab. AVR
FT 8003
TI 8003
lab.
lab.
P
T
L
8
S
S
-
—
P
P
P
—
—
P
P
L
P
—
—
P
P
-
—
Recording /
Sampling
_
-
_
—
_
each diff . type
of barrels
u
_
cont.
cont.
15'
cont.
15 'or cont.
cont.
15'
15 'or cont.
—
Remarks
6: (Screen Prints)
Group no / channel
Crane on manual
operation, ca. 5,0 -
7,5 t/h
G 95/2 + tank level
G 95/3 + tank level
G 00
G 95/4 + tank level
G 00
IN3
O
CO
-------
SftMPLDG MJD MEASURING
Table 2.2
Date: August 1988
P = Plant Computer / Screen Prints
I waSTE EEEDINS T = Test Computer - WfE
L = Local / Rack
8 = Screen (hand rec.)
Pt.
5
6
7
8
System
NO.
A-203
A-204
A-207
A-206
System
burner for med.
calorifc waste
lance for
pasteous waste
lance for
polymerized
waste
burner SCO for
oil no. 6 and
high calorific
waste
Required
Info
flow
temperature
heating value
composition
flow
temperature
heating value
composition
flow
temperature
heating value
composition
flow
temperature
heating value
composition
Instrument
Check Point
flow meter
thermometer
sample analysis
sample analysis
level difference
thermometer
sample analysis
sample analysis
flow meter
thermometer
sample analysis
sample analysis
flow meter
thermometer
sample analysis
sample analysis
Instr.
No. /
Resp.
FI 8004
TI 8005
lab. AVR
lab. AVR
_
TI 8002
lab. AVR
lab. AVR
EC 8009
TI 8006
lab. AVR
lab. AVR
FE 8005/
8006
TI 8007/
8008
lab. AVR
lab. AVR
P
T
L
S
P
P
—
P
P
P
-
—
"
P
P
L
P
—
Recording /
Sampling
cont.
15'
15'
15'
cont.
15'or cont.
_
—
~
cont.
cont.
15'
cont.
15'or cont.
Remarks
Gt (Screen Prints)
Group no / channel
G 95/5
tank level measurem.
G 00
G 99/5, during test
not in operation
G 95/6 + tank level
G 95/7 + tank level
G 00
ro
o
10
-------
BaMPUNS
MEASURING
Table 2.3
Date: August 1988
ro
i— >
<->
_ ___ ,. p = Plant Cccpiter / screen Prints
J. WflSiE J* Kr'l YlNR rn m^.-. i. •• -«
L = Local / Rack
S = screen (hand rec.)
Pt.
9
10
System
No.
A-205A
A-205B
System
lance for low
calorific waste
(waste water)
lance for low
calorific waste
(waste water)
Required
Info
flow
temperature
heating value
composition
flow
temperature
heating value
composition
Instrument
Check Point
flow meter
thermometer
sample analysis
sample analysis
flow meter
thermometer
sample analysis
sample analysis
Instr.
No. /
Resp.
FI 8007
TI 8009
lab.
lab.
FI 8008
TT 8010
—
P
T
L
S
P
P
—
P
P
Recording /
Sampling
cont.
15'or cont.
15 'or cont.
15'
Remarks
G: (Screen Prints)
Group no / channel
G 94/2 + tank level
G 95/1 + tank level
-------
SAMPLING MID MEASURING
Table 3
Date: August 1988
ro.
— P - plant Computer / Screen Prints
wnw T = Test Computer - WfE
S - Screen (hand rec.)
Pt.
11
T
12
)
13
14
15
16
System
No.
K-251
K-251
K-255
K-
K-252
K-256
K-
System
primary air
primary air
cooling air
burner air
secondary air
oonding. air
burner air
Required
Info
flow
temperature
flow
flow
flow
flow
flow
Instrument
Check Point
annubar
PT-100
pitot or anemometer
pitot
annubar
pitot
pitot
Instr.
No. /
Resp.
F 8100
T 8102
W-E/AVR
WfE
F 8103
-
WfE
P
T
L
S
P/T
T
L
L
P/T
L
L
Recording /
Sampling
cont.
cont.
1 X
cont.
cont.
1 X
cont.
Remarks
G: (Screen Prints)
Group no / channel
G 94/3
travers measurement
(ca. 2000 m3/h)
G 94/4
travers measurement
(ca. 2000 m3/h)
-------
SAMPLING MJD MEaSURING
Table 4.1
Date: August 1988
L = Local / Rack
S = Screen (hand ree.l
Pt.
20
21
22
23
System
No.
A-266
X-250
X-251
X-252
System
slag discharger
boiler ash
radiation part
discharger
boiler ash
conv. part
discharger
ESP-ash
Required
Info
tot. weight
repr. sample
tot. weight
repr. sample
tot. weight
repr. sample
tot. weight
repr. sample
Instrument
Check Point
-
-
-
-
Instr.
No. /
Rfesp.
-
-
-
-
P
T
L
S
-
-
-
-
Recording /
Sampling
15'
cont.
cont.
-
cont.
Remarks
G: (Screen Prints)
Group no / channel
See procedure for slag
sampling
Sample analysis :CE
mixed - representative
sample
Sample analysis :CE
sampled through the
emergency opening with
a special installa-
tion.
- mixed - repr. sample
Sample analysis :CE
Sample analysis :CE
-------
SAMPLING AND MEASURING
Table 4.2
Date: August 1988
— — - ~~ ~ p = Plant Computer / Screen Prints
III SLAG, ASH AND FLUE GAS T = Jart 0-prtjr - W«
L = Local / Rack
S = Screen (hand rec.)
Pt.
24
(05)
System
No.
System
flue gas to gas
cleaning system
flue gas to
stack
Required
Info
flow
CD-content
HCL
HF
H2O
Dust-content
Heavy metals in
the gas
(^-content
CD-content
S02
CO
Instrument
Check Point
.
Sieroens-^Jltramat
_
_
—
Sick GM 21
Instr.
No. /
Resp.
_
QI 8101
AVR
AVR
AVR
AVR
AVR
AVR
AVR
QI 8205
QI 8208
P
T
L
S
L
P
L
L
L
L
L
L
L
P
Recording /
Sampling
cont.
cent.
cont.
cont.
cont.
cont.
cont.
cont.
cont.
cont.
Remarks
G: (Screen Prints)
; Group no / channel
G 96/2
Analysis: AVR
Analysis: AVR
Analysis: AVR
Analysis: CE
Analysis: CE
Analysis: AVR
Analysis: AVR
G 96/5
IN5
I—»
CO
-------
SAMPLING MJD MEASURING
Table 5.1
Date: august 1988
_T P = Plant Ocnpiter / Screen Prints
IV PROCESS AND CONTROL MEASUREMENTS T = Test Computer - Wf E
L = Local / Rack
B = screen (hand rec.)
Pt.
Tl
-T2
T3
T4a
T4b
P4
T5a
Q5
T6
T7
T8
System
No.
F-251
F-251
F-252
F-252
F-252
F-252
H-251
H-251
T-253
T-254
T-255
System
kiln shell
kiln
SCC wall
SCC outlet
SCC outlet
SCC outlet
boiler-radia-
tion-part
outlet
boiler-radia
tion-part
outlet
ash-hopper
ash hopper
ash hopper
Required
Info
wall temperature
slag temperature
wall tenperature
gas tenperature
gas tenperature
gas pressure
gas tenperature
CO Vol-%
gas tenperature
gas tenperature
gas tenperature
Instrument
Check Point
local thermometer
pyrometer
local thermometer
thermo couple
pyrometer
pressure transm.
thermo couple
CO-test
thermometer
thermometer
thermometer
Instr.
No. /
Resp.
T 8091/
94/97
TI 8106
T 8054/
55
TI 8107
TI 8108
PI 8110
TI 8109
QI 8101
T 8073
T -
T -
P
T
L
B
L
P/T
L
T/P
P
P/T
P/T
P/T
L
L
L
Recording /
Sampling
1 X
cont.
1 X
cont.
cont.
cont.
cont.
cont.
30'
30'
30'
Remarks
G: (Screen Prints)
Group no / channel
G 90/1
G 90/2
G 90/3
G 92/3 :
G 90/4
G 96/2
-------
SAMPLING AND MEASURING
Table 5.2
Date: August 1988
P = Plant Computer / Screen Prints
IV PROCESS AMD CONTROL MEASUREMENTS T = Test Computer - WfE
L = Local / Rack
S = Screen (hand rec.)
Pt.
T10
P10
P20
T20
F20
P23
T23
F2B
Til
-
-
System
No.
T-
T-
-
-
-
-
-
-
-
-
System
boiler outlet
boiler outlet
feed water
feed water
feed water
steam
steam
steam
fuel gas at
ESP outlet
cooling water
cooling air RK
Required
Info
gas temperature
gas pressure
pressure
temperature
flow
pressure
temperature
flow
temperature
flow
temp, difference
flow
Instrument
Check Point
PT-100
pressure transm.
pressure transm.
thermometer
annubar
pressure transm.
thermometer
annubar
PT-100
pump characteristic
thentoneter
anemometer
Instr.
No. /
Resp.
IT 8110
PI 8112
PI 8118
TI 8114
FI 8101
PI 8124
TI 8123
FI 8102
TI 8207
-
-
P
T
L
S
P/T
P
T
P/T
P
T
P
P
T
-
-
Recording /
Sampling
cont.
cont.
pont.
cont.
cont.
cont.
cont.
cont.
cont.
1 X
1 X
1 X
Remarks
G: (Screen Prints)
Group no / channel
G 90/5
G 94/6
G 101/4
G 94/5
ro
i—>
en
-------
ii £u
n
n
il
li
it
u
1! i—
ii ro
il ^£
II --
li
II
11
II
il t-
'! i-
11 -si
II 0
li
II
311
•Ml
1'!
il
II
ii ^~
:.- fO
u
ii
•i
li
: >
ii *•
ii PJ
ii
ii
!|
i|
;i
M
II
11 -si
11 I'-
ll
il
1!
II
1!
H
II *.
11 CD
II
II
M
li
ii cr.
n m
ii £
il CO
1! CO
-u to i-
0^sj£
O W 0^
cn -^ A
*— « i— *
•JJ « C3
O VD PJ
a a o
PJ
•JD a co
•- cn A
cn CD -ii
•si *. CD
cn co \0
CO i— CO
CD MJ -si
~N] CT> C"
D cn CD
SCO
r~i
a a a
P
*
n i-
CP 3
U3 O D C
^Of.
• X II
irt o
7T to rt-
UD in *n
X C
3- ft-
O
-f»
flj £
U! 1!
- 3- d-
iC 1
X
T .
a- x
rt 0
7T 3- «•
X .1'
1
to m
w in
JT 3" TJ
ID 1
X
3-
0 JO
3 O Si
•£l 3 £
x o in
*7| iJJ flj
3 3 M
CO T
t> C
2 i- O
5 O •-
> £ C
fit =1
x 5
3- 1
U!
0
r
H
n
w
y 1
il
li
il
r,
II
•i -.
II D
II ri-
ll In
II I'-
ll
j!
M
II
II
II
li
ii
11
il
II
II
II
II
II XI
II -b.
il •-•
it PJ
II
II
il
II
il PJ
ii co
u
ii --
n
ii
ii
II CD
II CO
1! •-
M
n
M
M
II
II 1
li
II
II
11
II
II
II
M
!! D
II "
ii cn i
ii PJ !
cn in
i— >-*
PJ .&.
u)
»-•
O
•a D en
010
.- c •-
lC 3 H.
s in a
IV PJ T: It
1
7T M
x -o
3- C
t— i
3
TJ
.n c
•i- rt-
w
r-i
0
Tl -!»
-D
H-
3
F
I/I
-1
n
10
H
ID
T
T
IB
1C
3
ft1
Hi
H
3
rt
u
n
il
II
it
ii ?
1! fi-
ll !a
1!
II
I)
U
It
II
II
ii
il
ii
ii
it
n cn
li CO
M CO
il -b-
II
li
II
il
il O
ii xi
II cn
II
il --
II
II
II
II
1! 0
u •-
i! cn
II
II
II
II
II
ii 1
11
11
II
II
II
H
II
i:
!! D
II —
II ~\i
cn cn
*— » *—
cc o..
ro ro
cn cn
T3 C3 CD
2-381
1C 3 1
i 3 W •'
b b T w
CO i—
PJ PJ
•JJ -fc. 4». CO i-«
PJ xi co cn a
a cn ii) cr. o
a — a a o
c coo o 5
o o a a o
a co o o o
^ *. 4>- f cn
i i i i i
D D ODD
o o o a o
• 2 — *
' C i*
l 3 3
' 3- T-
S
2 CD
C C
3 3
CTJ7-
jl -
II O
il
H
II
II
U 1
II
II
11
II
II
n
n
!! Q
II -
II CO
II ii!
11
0 '
r-t 1
" *~>" 1
•A 1
8£jwg;
CT f\J CO t
1
*
j
PJ >— i
cn U3 cn A i
co cn a co i
r j CD o -si s
i
t
-, rg !
•j -j .j ^ .
- CO i- xi- 1
a Q o Q i
1
• i
i
»- j
D O O i— 1
CD «- a f\j i
* CD A D i
1
1
<
i
1 1 I 1 1
1
1
1
1
1
1
1
n i~i n i— !
- - - • i
CO CO •- PJ 1
o a a o i
C til
3 3
3" O
IT- 'D
-i
-V *-<
"O
3^ C
rt-
3*
TJ
nc
•— rf
n
•%
, ,
O
TJ
CO
Hi
^
J
H-
0
0
3
ft
IS
3
rt
-------
Flmount •
Cl
120,29
1,50
201,14
12,22
335, 16
C kq/h 3
"F
54,54
0,60
5,32
0,23
60,70
, . p.
- •
_
S
5,84
1,50
8,87
1,75
17,96
Metal inpi
Hg
0,10
17,85
1,51
0,50
19,96
jt C g/h :
Cd
0,49
1,50
2,96
2,04
6,98
1
Tl
-
-
fis
-
-
!Rmount
C kg/h
Meta1 input
C g/h 3
Cl
2,10
0,71
0,48
44,06
0,92
48,27
F
1 , 05
0,04
0,19
8,42
0,09
9,79
P !
I
.-.
-
S
4,20
0,71
0,48
2,48
2,76
10,63
Hg !
0,63
0,06
0,06
2,65
2,39
5,80
Cd
2,10
0,36
0,48
2,48
0,92
6,33
Tl
,
. -
fls
-
-
flmount
Cl !
1
1,70
4,34
0,55
170,09
44,10
220,77
C kq/h 3
"F !
I
1
0,09
18,94
0,22
3,20
0,63
23,08
P !
--
•
S
1,70
4,34
0,55
17,26
14,70
38,54
Metal inpi.
Hg 1
0,81
1832,08
0,17
2,02
3,86
1838,94
jt. C g/h ]
Cd
0,85
32,54
0,55
2,47
11,55
47,95
Tl
'
~"
-
fls :
-
-
i
i
qas
_i
b
g/h
2,53
1,27
6,18
3,3
Gas-
hum i d i ty
vol-X
9,2
10,9
8,3
9,5
Slag
percent
from
solids
•/.
87,66
92,05
89,54
99,75
I
1
1
I
I
i
I
•
* " - i
i
i
1
I
I
HRZIFIRD GRSES
HC1 tr.
Rawgas-
conc.
calcul.
mg/NmA3
5509,50
823,94
3501,90
:HF tr.
! Rawqas-
! cane .
! ca 1 cu 1 .
i mg/NmA3
! 1021,59
! 171,01
! 374,84
1502 tr.
! Rawqas--
! cone .
! calcul .
! mg/Nn»A3
! 574,43
i 352,95
i 1189,26
I
I i ' i
f I I
217
-------
Co
15,10
Q2,50
283,97
39,58
421,14
Ni
40,42
222,00
760,21
81,48
1104,11
Sb
-
•
Pb
46,27
258,00
854,86
91,37
1250,50.
Cr
4,87
27,00
94,66
9,89
136,42
Cu
4,87
322,50
6321,25
72,17
6720,78
Sn
60,88
331,50
1990,73
122,22
2505,33
8a
-
-
Co
210,00
35,60
48,30
247,50
92,00
633,40
Ni
420,00
35,60
48,30
247,50
92,00
843,40
Sb
-
Pb
420,00
35,60
48,30
247, 50
276,00
1027,40
Cr
210,00
35,60
48,30
247,50
92,00
633,40
.• :•&&
1680,00
35,60
48,30
247,50
828,00
2839,40
Sn
630,00
35,60
48,30
247,50
184,00
1145,40
Ba
-
_
Co
85,00
144,60
55,10
246,50
210,00
741,20
Ni
85,00
144,60
110,20
246,50
210,00
796, 3O
Sb
Pb
85,00
2169,00
110,20
246,50
420,00
3030,70
Cr
85,00
144,60
55, 10
246,50
210,00
741,20
Cu
765,00
144,60
440,80
1232,50
210,00
2792,90
Sn
85,00
144,60
165,30
493,00
210,00
1097,90
Ba
218
-------
flg
65,75
360,00
236,44
132,11
794,30
Ge
4 , 87
15,00
29,58
5,82
55,27
Be
_
-.
-
?n
4,87
27,00
94,66
9,89
136,42
Mo
35,06 .
1 1 1 , 00
381,58
58,78
586,43
input rif
ana 1 ys.ed
heavy
metals
g/h
14737,65
Flg
210,00
35,60
48,30
247,50
92,00
633,40
Se
21,00
3,56
4,83
24,75
9,20
63 .,34
Be
_.
-
--
-
—
_
-------
Table 7
I Operating Data at Slag Tests
Measurable
Variable
Test 1 Test 3 Test 4
Temperature rotary kiln
Temperature secondary
combustion chamber
Boiler outlet
Steam temperature
Steam pressure
Steam production
Volume flow
Oxygen content
Gas humidity
°C
°C
bar
t/h
mVh (N, f)
% Vol.
% Vol.
1255 1290 1260
1076 ' 1250 1100
250
379
30
27.9
250
391
250
391
30 30
33.5 29.2
75,200 74,300 76,500
8.7
9.2
6.3
10.9
7.5
8.3
W+E Environmental Systems
220
-------
. 5.1
•£
0)
cB
_§ -O
_Q 3
1 "
1
WH^^H
r
1,8%
t
•>
i
i
«
(0
O)
0)
E
48%
0,6%
W+E Environmental Systems
221
-------
Secondary
combustion chamber
Slag
ro
ro
m
m
o
I
CO
CD
3
CO
Electrostatic precipitator
Fly ash
Wet scrubber
Scrubber effluent
Secondary
combustion chamber
Slag
Boiler
Fly ash
Electrostatic precipitator
Fly ash
Wet scrubber
Scrubber effluent
Flue gas
-------
W+E Environmental Systems
223
-------
ro
ro
m
rn
(D
Rotary kiln
Secondary
combustion chamber
O
I
c'
3
Electrostatic precipitator
Wet scrubber
Scrubber effluent &
"
p
Secondary
combustion chamber
Stack
Flue gas
UI
-------
0,2%
Copper (Cu)
100% §.
CO
«J
0)
0)
0,3%
W
(0
D)
(!)
SI
3,2%
18%
78%
0,1 %
W+E Environmental Systems
225
-------
Fig. 6.1
Comparison between the max. admissible heavy metal
concetrations according to TCPL and the concentrations
found in the slag Test 1
mg/l Hg Cd Se Pb Cr Ag As Ba
a
100
0.2 1.0 1.0 5.0 5.0 5.0 5.0 100.0
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.27
0.25
0.5
mg/l Hg Cd Se Pb Cr Ag As Ba
W+E Environmental Systems
226
-------
Fig. 6.2
Comparison between the max. admissible heavy metal
concetrations according to TCPL and the concentrations
found in the slag Tests
mg/l
Ag As Ba
T3
(0
X
ro
100
25
0.25
tl 0.5
0.2 1.0 1.0 5.0 5.0 5.0 5.0 100.0
0.00 0.01 0.00 0.06 0.00 O.QO 0.00 0-17
•!
• — •• —
•
I
"
t
mg/l Hg Cd Se Pb Cr Ag As Ba
W+E Environmental Systems
227
-------
Fig. 6.3
Comparison between the max. admissible heavy metal
concetrations according to TCPL and the concentrations
found in the leaching of the slag Test 4
mg/l Hg Cd
Se Pb
Cr
Ag
As Ba
.
1
•o
ro
x
1
100
0.25
Q>
0.5
0.2 1.0 1.0 5.0 5.0 5.0 5.0 100.0
0.00 0.01 0.00 0.06 0.00 0.00 0.00 0.16
mg/l Hg Cd Se Pb Cr Ag As Ba
W+E Environmental Systems
228
-------
TRBT3R 8 —
ANALYTE
MAXIMDM CONCENTRATION
Arsenic
Barium
Cadmium
Chrcsnium
Lead
Mercury
Selenium
Silver
Endrin
Lindane
Methoxychlor
Toxaphene
2, 4-D
2,4,5-TP (Silvex)
5.0
100.0
1.0
5.0
5.0
0.2
1.0
5.0
0.02
0.4
10.0
0.5
10.0
1.0
229
-------
TOBTJR 9 - IdJP MPVLYTES KND
UTTncnvBTJ!
ANALYTE
MAXIMUM GONOEOTRATION
irg/1
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Selenium
Silver
Endrin
Lindane
Methoxychlor
Toxaphene
Heptachlor
a-Chlordane
r-Chlordane
2, 4-D
2,4,5-TP
Phenol
bis (2-Chloroethyl)ether
1,4-Dichlorobenzene
1,2-Dichlorobenzene
2-Methylj*ienol
4HMethylphenol
Hexachloroethane
Nitrobenzene
Hexachlorobutadiene
2,4,6-Trichloropherioi
2,4,5-Trichlorophenol
3-^fethylphenol
Fyridine
2,3,4,6-Tetrachlorophenol
2,4-Dinitrotoluene
Hexachlorobenzene
Fentachlorojiienol
5.0
100.0
1.0
5.0
5.0
0.2
1.0
5.0
0.003
0.06
1.4
0.07
0.001
0.03
0.03
1.4
0.1
14.4
0.05
10.8
4.3
10.0
10.0
4.3
0.13
0.72
0.30
5.8
10.0
5.0
1.5
0.13
0.13
3.6
* Volatile organic conpounds were not analyzed.
230
-------
Recycling of contaminated river and
lake sediments demonstrated by the example of Neckar sludge
Dipl.-Ing. M. NuIJbaumer, M.Sc. * and Dipl.-Ing. E. Bellinger
Summary
The Neckar is an approximately 370 km long tributary of the Rhine, its confluence with the
same being not far downstream from Heidelberg. The river was made navigable over a length
of 202 km during the decades following the last World War, i.e. weirs with locks were
constructed at some 20 km intervals along the river to regulate the water-level. Primarily
fine-grained, suspended and sedimentary materials are deposited in the storage ponds and
lead to the necessity of river dredging.
The purposes of the dredging are
- to keep the navigation lane open,
- to maintain the required river cross-section for flood control and
- to maintain water quality.
The dredged material has become contaminated by heavy metals due to the expansion of in-
dustry in the area, cadmium being the prime example. The sometimes high cadmium cont-
amination precludes the use of the dredged materials for agricultural purposes. It was there-
fore initially proposed that the dredged sludge be dried and dumped at waste disposal sites.
Since sites for waste disposal are rare and expensive to put into operation in the Federal Re-
public, economic means of recycling had to be sought.
* Managing director, Ed. Ziiblin AG, Civil Engineering Contractors, Stuttgart, FRG
0 Senior engineer, Ed. Ziiblin AG, Stuttgart, FRG
231
-------
Toward this end one of the largest civil engineering contractors in the FRG, Ed. Zublin AG
of Stuttgart, has developed a process whereby dredged material is converted into spherical,
porous, lightweight aggregate for the production of masonry blocks and lightweight concrete.
The above-mentioned company has been awarded a contract to construct and operate a plant
for the thermal treatment of 500,000 m3 of sludge dredged out of the Neckar over a period
of 10 years. In order to enable the thermal procedure involving temperatures of up to
1150 °C to be successfully put into service in an environment-friendly manner, a new con-
cept for outlet-gas treatment had to be developed and tested.
Contaminated sediments
Dredged sludges from rivers and harbours can no longer be used for agricultural purposes
because of their high level of contamination. The quantity of sludge dredged by the water
and river navigation authorities to maintain navigation lanes is however considerable: in
1982 a total of 48 million cbm was removed from Federal waterways, of these 11 million
cbm were from the Elbe and 12 million cbm were from the Rhine alone.
During the last two decades some 2 million cbm of sediments have accumilated on the bed of
the Neckar between Plochingen and Heilbronn. Although extensive measures to improve ef-
fluent quality have reduced heavy metal pollution, the heavy metal content of the sediments
in such that they cannot be used for agricultural purposes.
One cbm of Neckar sediment contains 800 1 water and 550 kg solids. 5 % of the solids are
debris of organic or anthropogenic origin such as woods, bicycles, refrigerators and rubble.
The major part is, however, of mineral origin and has found its way into the river by way of
natural erosion in the river's catchment area. The material from the clay and silt fractions up
to 0.063 mm amounts to 40 % by weight of the sediment on average, while sand and gravel
between 0.063 and 60 mm also make up 40 % by weight of the sediment. The composition
varies, however, considerably.
232
-------
High levels of pollutants accumulate in the fine-grained sediments. A number of investig-
ations have demonstrated that a larger quantity of heavy metals accumulates in the fine-
grained fraction of the dredged sludges than in the sand and gravel fractions. The sand-and .
gravel fractions can therefore be disposed of relatively easily, while the fine-grained sludge
presents a considerable problem.
Of the pollutants, heavy metals, heavy metal compounds, salts and all organic materials
which enter a watercourse as a result of industrial production present a significant problem.
The focal point of the investigations of the dredged Neckar sludge was, however, the reten-
tion of cadmium, which was detected in concentrations of up to 240 mg/kg (dry weight).
The permissible concentration in soil used for agricultural purposes is 3 mg/kg (dry weight).
Up to ten years ago dredged Neckar sludge was used for agricultural purposes in large quan-
tities. A layer of sludge up to 1 m in thickness was spread over existing fields. This soil has
now had to be removed in part due to the high concentrations of heavy metals. Other pro-
cedures, such as drying in air, mechanical dewatering and hardening with alkaline binding
agents, require waste disposal sites for end storage. On the Neckar, the dredged material was
to be partially dewatered in three air-drying plants and subsequently transported to three
remote single-purpose waste disposal sites. The adoption of this approach met with conside-
rable resistance.
Procedure to recycle contaminated sediments
As a result of the above, it was decided to largely avoid dumping Neckar sludge. Instead it
was proposed that spherical, porous, lightweight aggregate be produced from the sludge by
way of adding expanding clay, the aggregate having characteristics similar to expanded clay
and being capable of use as a substitute building material in order to reduce the import of
expensive pumice-stone.
The procedure involves transporting the dredged sludge to a central treatment plant at which
an intermediate storage facility has been installed to even out seasonal variations in dredging
activity. The sludge is then partially dewatered in a screen belt press (mechanical process)
and mixed with clay and additives.
233
-------
The amount of clay and expanded clay additive amounts to between 15 % and 30 % by
weight depending on the required strength and heat insulation properties of the final pro-
duct. The mixture is subsequently made into pellets and then dried, burnt and cooled in the
actual rotary kiln plant. The temperature of combustion may rise up to 1150 °C. Heat is ex-
tracted from the outlet-gas to be used elsewhere in the system, thus ensuring a high degree
of heat reclamation.
Development for full-scale operation.
The procedure has been developed for full-scale operation with the support of the Bundes-
minister fur Forschung und Technologic (Federal Minister for Research and Technology) and
tested in two stages. The development work was based on the results of innumerable labora-
tory tests to find suitable mixtures for the differing sludge compositions, clay content etc.,
all of which have to ensure sufficiently good product characteristics.
3*?:- - /
dredging
treatment
filtration
water
discharge gas treatment
drying
mixing pelleting combustion grading
Fig. 1. [OPERATION SEQUENCE FOR RECYCLING SLUDGE
234
-------
The main technical and scientific objective of the research and development programme was
to prove that lightweight aggregate can be produced industrially from fine-grained river and
lake sediments, that the aggregate's physical properties permit a wide range of applications
and that the procedure can be operated economically. The production-of the lightweight ag-
gregate must be favourably priced, sure in operation, environment-friendly and adaptable to
variations in raw material quality.
Fig. 2. Pellets of dried Neckar sludge and expanded clay.
Neckar sludge served as the "raw material", the sludge having been dredged out of the up-
stream area of locks and demonstrating a wide spectrum with respect to both its grain-size
distribution and its contamination.
Pretreatment of the sludge
The sludge must be dried before it is burnt. The drying is firstly mechanical in screen belt
presses and then, in pellet-form, in the kiln plant drier. In order to achieve as high a solids
content as possible before the sludge is fed into the thermal drying plant, numerous tests
were caried out with screen belt presses and flocculation agents.
235
-------
Expanded Clay Production using Neckar Sediments
An industrially-used expanded clay production plant was modified and used to produce ex-
panded clay from Neckar sediments in the first and second test phases.
The mixture-ratios which had been selected during laboratory tests were shown to be suit-
able in the first test phase. The measured outlet-gas values were used as initial values to
design the outlet-gas treatment. The second test phase was spent optimizing and testing a pi-
lot outlet-gas treatment plant.
The end product is a lightweight aggregate of various sizes (0-16 mm). The expanded clay
can be used as a building material. It has a high compressive strength, excellent insulation
properties and is easy to handle.
The expanded clay was tested with reference to the requirements set out in DIN 4226, Part 2,
for lightweight aggregates. All the requirements were met. The heavy metal content lies in
the same region as is to be found in other building materials such as pumice, brick,
aerated concrete and conventional expanded clay. Numerous tests have proven that the ex-
panded clay is a suitable building material for walls.
Fig. 3. Hollow masonry blocks leaving the press.
236
-------
Hollow masonry and masonry blocks were produced in a cement works. These blocks also
met all the requirements for outer and/or inner insulatory masonry blocks.
Fig. 4. Hollow masonry blocks made from Neckar sludge replace valuable raw materials
and fulfil heat insulation requirements.
Outlet-Gas Treatment
There is no visible difference between the above-mentioned end product and expanded clay
blocks. Waste, in this case contaminated river sludge, has been used to produce a first-class
block and at the same time the use of valuable raw materials has been minimized.
Where have all the heavy metals gone? 100.000 m8 of river sediments which are dredged out
of the Neckar annually contain approximately 500 kg - 1000 kg cadmium. Cadmium has a
boiling-point of 700 °C which is below the combustion temperature. The outlet-gas treat-
ment has therefore to ensure that cadmium as well as other heavy metals and contaminants
present in the outlet-gas are not emitted into the atmosphere.
The outlet-gas values were measured during the first test phase, analysed and interpreted.
The areas concentrated on were heavy metals and the acidic gaseous components in the gas
such as SOX, N0x, HF and HC1.
237
-------
A treatment process capable of removing all these contaminants was not to be found so that a
plant had to be designed by the contractors. The Landesanstalt fur Umweltschutz in Karls-
ruhe (State Environmental Agency) took the outlet-gas measurements during the test phases
on behalf of the Regierungsprasidium Stuttgart (State Government).
The outlet-gas treatment was designed in close cooperation with the Landesanstalt fur Um-
weltschutz in Karlsruhe. The treatment process is in four stages.
Stage I
Cyclone and sleeve filters seperate out larger particles. These are returned to the production
process at the mixing stage.
Stage II
A sleeve filter seperates out dust form the belt drier. This dust is also returned to the pro-
duction process at the mixing stage. The outlet-gas then undergoes secondary combustion at
temperatures of 800 - 900 °C and a residence time of more than one second. This waste heat
is then recovered in a heat-exchanger which has an effectiveness of 75-80 %. The recovered
heat is used to warm the belt drier.
Stage III
Dry absorption (lime dosage) with subsequent sleeve filter.
Stage IV
Double fines filter. The dust emitted into the atmosphere after these filters is less than
0.01 mg/Nms.
The outlet-gas system is hermetically sealed. Larger particles with a low contaminant con-
centration are seperated out of the outlet-gas in the cyclone. The sleeve filters then clean the
air further and prevent the heat exchanger and the secondary combustion chamber from
clogging up. Organic compounds are present in the gas released from the mixture whilst in
the combustion kiln. These are turned into carbon dioxide (CO2), water (H2O) and hydro-
fluoric or hydrochloric acid (HF, HC1) in the secondary combustion chamber.
238
-------
The acidic gaseous components present in the gas such as sulpher oxides (SOx), hydrofluoric
and hydrochloric acid (HF, HC1) react with lime (Ca(OH)2) during the dry absorption to
form solid, saline compounds which are retained in the sleeve filter.
The remaining heavy metal compounds are removed from the gas and are retained in the
double fines filter. The dust collected in Stages III and IV is packed into drums and deposi-
ted at a waste disposal site.
Of the outlet-gas stream 2000 Nm3/h was tested during pilot plant operation. Measurements
were taken before, after and in between the various stages so that the treatment efficiency of
the seperate stages and the treatment process as a whole could be ascertained.
All the outlet-gas values were well below those stipulated by the law at that time (TA-Luft
83) so that the values are still below those stipulated by the existing law (TA-Luft 86). In to-
tal 50,000 m3 of Neckar sediments with an assumed cadmium content of 500 kg produce a
dust emission of 1 - 3 kg and a yearly emission of less than 2 g cadmium into the atmos-
phere.
fresh water
clay
gas
Ca(OHb
Neckar - Sludge
50.000 m'/a
> sorting
dewalering
mixing
combustion
grading
coarse fraction 1.3 t/h
water 4,6 m'/h .Neckar
building
expanded \ material
clay *t/n / SOJDOOmVa
outlet-gasl2.000 NmYh .atmosphere
filter-dust 12 kg/h .land-fill
Fig. 5. [Expanded Clay Production using Neckar-Sludge - Throughput
IZOBLJNI
239
-------
interim store
Neckar-Sludge
Fig. 6. (Expanded Clay Production using Neckdr-Sludge -' Site plan
IZDBLINI
240
-------
FORUM
on
INNOVATIVE HAZARDOUS WASTE TREATMENT TECHNOLOGIES:
DOMESTIC AND INTERNATIONAL
Atlanta, Georgia
June 19, 20, and 21, 1989
OXYGEN ENHANCEMENT of HAZARDOUS WASTE
INCINERATION WITH THE
PYRETRON THERMAL DESTRUCTION SYSTEM
by Mark Zwecker, Fred Kuntz,
and Gregory Gitman
of AMERICAN COMBUSTION, INC.
NORCROSS, GEORGIA
241
-------
ABSTRACT
A SITE program demonstration of the PYRETRON® Thermal
Destruction System was conducted at the EPA's Combustion Research
Facility (CRF). The PYRETRON TDS, developed by American Combus-
tion, Inc. (ACI) of Norcross, Ga. , was installed on the pilot
scale rotary kiln incinerator. The demonstration tests were con-
ducted using waste material from the Stringfellow Superfund site
near Riverside, Ca. with a high heating value decanter tank tar
sludge waste from coking operations. The test objectives were to
evaluate ACI's claims that the PYRETRON TDS is capable of achiev-
ing:
Control of transient discharges of POHCs and PlCs
during operating upset conditions;
Higher waste feed rates than conventional incineration;
Economic system operation.
The demonstration test results showed that ACI's PYRETRON
TDS achieved the RCRA 99.99 percent POHC DRE at a waste feedrate
which was 100% .greater than the maximum rate achieved under con-
ventional incineration. Measured particulate emissions from the
PYRETRON testing were significantly less than the required 180 mg
per dscm corrected to 7 percent oxygen.
The PYRETRON TDS was also tested at the maximum conventional
system feedrate but with a 60% increased mass charge size.
During these tests the PYRETRON system was capable of handling
the increased charge mass without unacceptable levels of "puffs"
generation. The concentration of POHCs in the ash residue was
consistently below detection limits.
SITE Demonstration Test Program
American Combustion, Inc. was selected as one of the first
participants in the EPA's Superfund Innovative Technology Evalua-
tion (SITE) program. The SITE program aids in the commercializa-
tion of alternative and innovative hazwaste treatment tech-
nologies .
Within the SITE program, ACI performed a demonstration of
the PYRETRON Thermal Destruction System (PTDS). The PTDS is an
oxygen enhanced combustion system using ACI's patented technology
designed for waste incinerators. The PTDS also incorporates a
proprietary process control algorithm designed to anticipate,
detect, and respond to prefailure conditions within the system.
The testing and evaluation of the PTDS took place at the
EPA's Combustion Research Facility (CRF). The CRF is a fully
permitted incinerator established by the EPA to conduct incinera-
tion testing and research. The testing occurred at the end of
1987 and beginning of 1988.
ACI, the EPA, and Accurex, the operating contractor of CRF,
developed a test program to evaluate the PTDS versus a conven-
tional air based incineration system. The test program objec-
tives were: J
242
-------
1.
2.
3.
Evaluate the PTDS' ability to reduce the magnitude
and/or number of transient upsets ("puffs");
Achieve the RCRA 99.99% ORE at significantly higher
feedrates than that attainable in the conventional sys-
tem ;
Improve the overall system economics compared to the
conventional system.
For these tests the only change to the system involved
removing the existing conventional burners and installing ACI's
PYRETRON TDS. A schematic of the PYRETRON system used in the CRF
tests is shown in Figure 1.
The tests involved the incineration of waste from the
Stringfellow Superfund site spiked with decanter tank tar sludge
(listed K087 waste) at a 2/3 ratio respectively. K087 waste con-
tains high concentrations of several hazardous polynuclear
aromatic components. The K087 waste was added to increase the
level of volatile contaminants in the feed matrix. The resulting
volatility of the waste, therefore, provided a much more dif-
ficult test for the capabilities of the PTDS. Table 1 of the Ap-
pendix contains the chemical analysis of the waste mix used in
the CRF testing.
The test program included scoping tests of the conventional
system as well as the PTDS. These scoping tests were performed
to maximize the total mass feedrate of the waste mix in both sys-
tems. The maximum feedrate for each system was determined by the
system's ability to incinerate a charge mass without exceeding
regulatory CO limits in the afterburner exhaust. Additional
testing was done with the PTDS at the conventional system's maxi-
mum feedrate but with a 60% larger charge mass. This test was
used to provide data to confirm the PTDS' ability to control
transient upsets.
243
-------
FIGURE 1 - SCHEMATIC OF THE PYRETRON THERMAL
DESTRUCTION SYSTEM TESTED AT CRF
Measured
process
parameters
Valve train
(gas, oxygen, air) •
J
Gas, air, and oxygen
flows to the burners
Ash pit
PYRETRON THERMAL DESTRUCTION SYSTEM
PROCESS DIAGRAM
244
-------
Test Results
The optimum mass feedrate of the conventional system was es-
tablished at 105 pounds per hour. This rate was accomplished by
feeding mass charges of 21 pounds every 12 minutes. Data ob-
tained from the conventional system during this test is contained
in Figure 2.
Figure 3 shows data from the conventional system operating
at a charge mass and interval slightly above the optimum. The
operating data at this feedrate shows many failures due to the
depletion of available oxygen in the kiln atmosphere. As a
result, the system experienced flame failures, system pressure
excursions, and excessive process temperatures commonly as-
sociated with the development of "puffs." The conventional sys-
tem was incapable at the increased feedrate of simultaneously
maintaining sufficient excess oxygen and reducing' auxiliary fuel
input to control the upset.
The corresponding optimum mass feedrate for the PTDS was es-
tablished at 210 pounds per hour. This rate was accomplished by
feeding mass charges of 21 pounds every 6 minutes. Data for the
PTDS during this test is shown in Figure 4. The feedrate
demonstrated by the PTDS is twice the optimum feedrate of the
conventional system.
Stack test results showed that the PTDS was able to achieve
greater than 99.99% ORE at the increased feedrate. Stack samples
taken during the optimum feedrate test run did not detect any
POHC. There was also no detection of POHC within the ash residue
or scrubber effluent from this test. The PTDS also operated well
within the regulatory limit of 180 mg per dry scm for particulate
emissions. Table 2 located in the Appendix lists the results of
the stack particulate testing.
Figure 5 shows the data for the PYRETRON system at a charge
mass of 34 pounds and a charge interval of 19 .5 minutes. 'This
charge mass represents a 60% increase over the optimum mass which
the conventional system could handle. Under these operating con-
ditions, the PTDS provided enough oxygen to the system to control
CO discharge from the kiln. As a result, the afterburner system
was not overloaded at these conditions and failures were
prevented.
245
-------
FIGURE 2
C.R.F. Kiln .Data for 12-09-87
Conventional System—21 lbs/12 min
2200
in
800C
6000
<000
200C
V
v^s^yv^iv'-V^^^vV^JX^J/V^A^
30
-. 15
C
500
250
0
Propone
-.•0
100
ISO
Minutes
200
300
nr
y
i •/
Storting Time : 12:40
50
100
150
'vlinules
200
250
300
246
-------
FIGURE 2 (CONTINUED)
C.R.F. Afterburner Doto for 12--O9-37
Conventional System—21 lbs/12 min.
o
t)
2200
200O
ieoo -
f
Slorlinc Time : 12:40
' ^ . L.
1.6EX
I
3OOO.O
4000.0
0.0
Propone x 10
50
100
150
Minutes
200
250
300
f-l
z
CJ
o
c:
(J
~ STARTING TIME-12:40
10 —
50
100 ISO 200
MINUTES
250
300
C.R.F. Stack Data for 12-09-87
Conventional System —21 lbs/12 min.
55 r
O. 'jl
O.
0 v
o °-
Steeling Time : 12:40
Measured at Slock Exit
'-5 f
50
too
150
Minules
200
250
ioo
247
-------
FIGURE 3
C.R.r. Kiln Data for 12-08-87
Conventional System — 24 ibs/10 min.
2200 ,
V
f. Storting Time : 10:27
u. r. -
o J. •
.. 2000 f-
^ S
eooo
JC
y 6000
g <000
iZ
2000
0
100
Aif
Propone
200
Minutes
300
400
SlorUng Time : 10:27
100
200
Minules
300
-------
FIGURE 3 (CONTINUED)
C.R.F. Afterburner Doto for 12 — 08-87
Conventional System--24 !bs/10 min.
. 2COO -
o
p" i500
c
73
aooo.o
4OCO.O
o.o
Storting Time : 10:27
200
300
400
400
C.R.F. Stock Dato for 12-08-87
Conventional System —24 lbs/1 0 min.
95
70
E
O.
a.
45
O
-
-
-
r
20 L
0
;
Storting Tim*
Measured ol
Sleek Exit
.
1
10:27
0 1 00 200
,1
i
1 , ._.»
_
_
_
,
t
._ _.
300 ^00
Minules
249
-------
FIGURE 4
C.R.F. Kiln Dota For I -2 I -8b
Pyretron System —21 lbs/6 minutes
2200
2000
taoo
300O
6003
10
S 4000
o
2000
0
Slorling Time : 12:00
^v^
50
100
i/V
150
Minules
f^/
200
v- Oxygen
Propone
250
JCO
20
>0
SCO
c eco
o.
CL
R
-------
JIGURE .4 (CONTINUED)
C.R.F. Afterburner Data for 1 -21 -88
Pyretron System—21 lbs/6'min.
^200
jo MOO
D
O
Q. 180O
t.8H:'i -
Storting Time : 12:00
150
200
250
3OO
C-I
z
o
w
cu
25
20
15
10
STARTING TIME - 12:00
.END OF
0 III . i .1 I I I . I I II I 1 1 1 1 1 1 1 1 1 ! ! 1 1 1 ! ! ! 1 L.
0 50 100 150 200 250 300
95
75
a.
8 35
15
0
Stociing Time : 12:00
50
1OO
150
Minules
200
250
3OO
251
-------
FIGURE 5
C.R.F. Kiln .Date for I,-14-88
Pyretron Sys.tern— 34 !bs/19.5 minutes
2200
150
Minuies
300
20
_. 15
C
-------
FIGURE 5 (CONTINUED)
2200
C.R.F. Afterburner Data for 1-14-88
Pyretron System-34 lbs/19.5 minutes
2000
0.0
250
300
STARTING TIME - 13:10
I I I T I I 1 1 1
100
ISO 200
MINUTES
250
300
253
-------
FTPS Process Improvement Summary
The process improvements provided by'the PTDS result from a
more efficient burner design and a process control system
programmed to adjust the process parameters required to improve
destruction efficiencies for hazardous constituents. The burner
design incorporates a second more concentrated oxygen stream into
the flame pattern. Figure 6 shows an illustration of the
PYRETRON burner flame pattern.
Oxygen is fed at sub-stoichiometric rates into the center of
the flame. Even at sub-stoichiometric rates the heat release
within this core is sufficient to cause pyrolysis of the sur-
rounding fuel. The pyrolysis products of combustion in this zone
are more highly radiative at the temperatures developed. Conse-
quently, energy is rapidly transferred from this zone to the sur-
rounding atmosphere. Flame temperatures are, therefore, reduced
from those typically achieved by traditionally designed burners
at the same oxygen enrichment level.
Air is fed around the periphery of the fuel zone to create a
cooler combustion zone at the edges of the flame pattern. At
this point the fuel is burning from inside and outside of the
flame pattern. By directing the various process streams from the
burner, a third zone is created in which the combustion products
from the pyrolytic and excess oxygen zones are combined to com-
plete the combustion of the fuel. This results in a high combus-
tion efficiency even at very low stoichiometric ratios. In addi-
tion, highly reactive excess oxygen exits the stable flame en-
velope at elevated temperatures.
FTPS Control
The PTDS is preprogrammed to respond to changing process
conditions to maintain high destruction capability. The response
capability of the system involves an instantaneous change in the
level of participation of the two oxidizing streams. ACI' s
proprietary control algorithm provides a series of responses to
various instrument readings to maintain optimum process condi-
tions. Conditions such as system pressure, insufficient excess
oxygen levels, high CO concentrations, low afterburner tempera-
ture, as well as other potential "upsets" can be contained by the
changes initiated by the PTDS.
The PTDS designed for the CRF testing was preprogrammed for
three response actions. One response action involved adding
oxygen in anticipation of a batch charge. At a preset interval,
a signal from the ram feed mechanism alerts the PLC that the
waste charge would soon be introduced into the furnace. The PLC
then automatically responded by changing the oxygen participation
rate to change the furnace atmosphere to include a significant
amount of hot, highly reactive oxygen before the mass charge.
The response action to a batch charge signal is shown in
Figure 7.
254
-------
FIGURE 6
AIR
AUXILIARY FUEL
OXYGEN
z
AUXILIARY FUEL COMBUSTION ZONE
HAZARDOUS WASTE
INCINERATION ZONE
FINAL COMBUSTION ZONE
PYRETRON INCINERATION PROCESS
255
-------
FIGURE 7
D
cr
cr>
Batch
Charge
Signal
On
On
Air
c.
D
Z5
o
Oxygen
Time
PYRETRON
Oxygen Based
Control Response
to a
Batch Charge Signal
256
-------
As a result of the PTDS' response, the potential for deplet-
ing oxygen at the initiation of the charge into the furnace was
reduced. Therefore, the reaction of the volatiles in the batch
charge was completed thus preventing the generation of much of
the "puffs" that occurred in the conventional system. A water
quench spray, controlled by the PLC, was used to dissipate excess
heat released from the charge.
Another response action programmed into the PTDS is a
response to system pressure excursions. This response, repre-
sented in Figure 8, permits the quantity of oxygen fed into the
system to be maintained while the total gas volume is reduced.
By changing the oxygen participation rate, the volume of inerts
fed into the system is reduced. As a result, the total gas
volume does not exceed the maximum that the system can handle and
fugitive emissions from the kiln seals are prevented. This
response is also useful for systems which must maintain a minimum
gas residence time. The response sequence can also be used to
modulate low afterburner process temperatures.
The third response action, represented in Figure 9, involves
a response to kiln exhaust gas analyzer signals that show a
depletion of oxygen or an increase in CO exiting the primary
chamber. When the analyzer showed that adjustments in the excess
oxygen amount were required, the PTDS would again adjust the par-
ticipation rate to prevent a failure mode from occurring. This
response action is also useful for responding to a furnace oxygen
deficiency or an excessive combustibles feed rate to the system.
257
-------
FIGURE 8
3
W
CO
£
Du
System Pressure
AWFSO Point
Pressure
Air
3
a
Time
Oxygen
PTDS
Oxygen Based
Control Response
to
System Pressure
Also:
Low Gas Residence Time
Low Afterburner Process
Temperature
258
-------
FIGURE 9
cd
O
o
O
v
>^
u^
c
co
O
(Possible
CO Level
CO Level *' \ on Air onIY
AWFSO Point £ \ Control)
» *
Control: v/^^N^ \
CO Level _/ ^ ^xi----..
'Air
Oxygen
PTDS
Oxygen Based
Control Response
to
Carbon Monoxide
Also:
Deficient Oxygen
Excessive Combustibles
Feed Rate
Time
259
-------
Conclusions
The EPA's Site program tests showed that the PTDS is capable
of improving the environmental and economic performance of a con-
ventional incineration system. The PYRETRON system accomplishes
these improvements by a combination of innovative burner design
and dynamic process control. Specifically, the EPA tests
demonstrated that ACI's PYRETRON technology will:
1. Provide increased waste throughput capability by con-
trolling "puffs" and improving DRE;
2. Only require a modification to the existing
incinerator's combustion and combustion control equip-
ment to achieve the PYRETRON technology improvements;
3. Achieve destruction efficiencies well above required
levels;
4. Eliminate transient releases of POHCs and PICs to the
environment by the system's preprogrammed Transient Up-
set Control response action;
5. Handle as much as a 60% increased batch load of
volatile wastes without increasing hazardous emissions;
The PTDS is applicable to any waste treatment incinerator,
including medical waste, which requires efficient control for en-
vironmental compliance. The PTDS is being marketed commercially
by American Combustion, Inc.
American Combustion, Inc. has over ninety commercial ap-
plications which use the PYRETRON technology. These applications
include electric arc furnaces, copper smelting furnaces, glass
tanks, ladle heaters, and lead and aluminum refining furnaces.
All of these applications use ACI's PYRETRON combustion systems
as well as various aspects of ACI's patented dynamic control
technology.
260
-------
APPENDIX
TABLE 1 - CHEMICAL ANALYSIS OF WASTE FEED MIX
USED IN PYRETRON TDS DEMONSTRATION TESTS
Component
Naphthalene (C10H8)
Phenanthrene (CgH4CH)2
Acenaphthalene (Ci2Hg)
Fluoranthene (C16H10^
Pyrene (C16H10)
Anthracene (C6H4CH)2
Fluorene (C6H4 CH2 C6H4)
Dibenzofurane (CgH4)20
2-Methylnaphthalene (C10H? CH3)
Miscellaneous PolyNuclear Aromatics
Miscellaneous Unknowns
Concentration, ppm
49,772
22,569
11,722
11,529
11,036
6,764
6,397
3,926
3,457
5,017
1,031
261
-------
TABLE 2 -
PARTICULATE EMISSIONS DATA
Particulate
Concentration a/D
Particulate
Concentration c
Test
1
2
3
4
5
6
7
8
System
Conventional
Convent iona 1
Conventional
Conventional
PYRETRON TDS
PYRETRON TDS
PYRETRON TDS
PYRETRON TDS
ma/dscm
" 8
9
47
29
28
10
28
39
ma/dscm
4
4
28
19
24
7
17
20
aCorrected to 7% oxygen for the air-only tests.
kpYRETRON tests were corrected to 7% oxygen considering the
effect of oxygen enrichment.
GUncorrected
262
-------
T©st ResunlLts with
smm C emit riff MS a I
R.C. Eschenbach, R.A. Hill & J.W. Sears
HN€ UKIAH, CA
Forum on Innovative Hazardous Waste Treatment Technologies:
Domestic and International
June 19-22 1989
Atlanta, GA
263
-------
TITLE:
AUTHORS:
INTRODUCTION:
PROCESS DESCRIPTION AND INITIAL TEST RESULTS
WITH THE PLASMA CENTRIFUGAL REACTOR
R. C. Eschenbach, R. A. Hill, and J. W. Sears. Retech.Inc. Ukiah,
California
During the last few years Retech, Inc. has developed the Plasma Centrifugal Reactor (PCR)
to stabilize solid waste material while decomposing any toxic hydrocarbons into relatively
innocuous, simple molecules. The PCR uses heat from an arc to melt and vitrify solid
components, thereby accomplishing the decomposition and stabilization of the waste.
Plasmas can produce temperatures in excess of 10,000°C although the expected
temperature to be produced in the molten glass will be about 1600°C.
The development of this furnace has been achieved through a three phase program. The
first phase was directed at reducing the volume of radioactive mixed wastes. Initial tests
were performed at Ukiah in Retech's 100 kW lab-scale plasma furnace, followed by
demonstrations on surrogate materials in a titanium production furnace located at Oregon
Metallurgical Corporation, Albany, Oregon. In phase two of this program a quarter-scale
PCR was designed, built and tested at our Ukiah facility. During this phase the first patent
covering the plasma centrifugal furnace was issued (U.S. Patent # 4,770,109). The final
phase of this program will be to evaluate the performance of a full size furnace.
Preliminary tests were performed in Ukiah (March-May 1989), while further tests will be
run by the U.S. Department of Energy (D.O.E.) at their Magnetohydrodynamics facility in
Butte, Montana, as part of the EPA's Superfund Innovative Technology Evaluation (SITE)
program (summer 1989).
PHASE I (1985-1986)
The initial tests, conducted with a transferred-arc plasma on materials (metals, glass,
rubber, plastics, filter elements, etc.) like those which get radioactively contaminated,
demonstrated the effectiveness of volume reduction. The first tests were made in our lab-
scale furnace by melting down pint cans filled with plastic gloves, rubber, ceramics and
chunks of metal. Scale-up tests with an argon arc at about 1500 amperes and 130 volts
were conducted on 4 liter steel cans stuffed with simulated waste. The resulting volume
reduction was found to be a factor of 20 with the corresponding weight reduction at a factor
of 1.8.X1) Problems with arc stability were encountered; contributing factors were vision
264.
-------
problems caused by sooting shortly after each can was added and limited power supply
voltage. It was concluded that the addition of oxygen or air as an oxidant in the plasma gas
would be desirable in order to convert hydrocarbons to carbon dioxide and water instead of
soot, carbon monoxide and hydrogen.
water-cooled
copper electrode
plasma gas
injection -v
arc termination
nozzle
slag bath
FIGURE 1: Schematic of the quarter-scale reactor showing the orientation
of the plasma torch with respect to the rotating tub.
PHASE II (1986-1988):
The first plasma centrifugal reactor was developed with a 150 kW transferred-arc plasma
torch operating on air or oxygen-argon mixtures. The furnace design incorporates a
stationary plasma torch with a revolving tub (45.7 cm ID-18 inches) inside a sealed reaction
chamber. The centrifugal force imposed upon the molten material moves it to the outside of
the chamber, where mixing of new feed material is accomplished. Figure 1 shows the
orientation of the plasma torch with respect to the revolving tub.
Operating with an oxidizing environment greatly reduces the sooting associated with
straight pyrolysis. Tests were run with a number of feedstocks after completing assembly
and debug in July, 19872. Liquid feeds such as water, methanol, and fuel oil were fed at
uniform rates. A spoon feeder permitted adding dirt or dirt and hydrocarbon mixtures
265
-------
(polymers and such) in 0.5 to 1.0 kg quantities every 5 to 15 minutes. An effluent gas
treatment system which includes a quench spray followed by a rock tower was used to
cool, neutralize and separate particles from the acidic exit gas. An air-ejector located in the
exhaust duct was used to maintain the furnace pressure below atmospheric. Tests showed
that the non-volatile components retained in the vitrified mass are non-leachable as
determined by standard tests. This furnace was originally designed to 'tilt-pour1 but it was
found that the molten glass became too viscous to allow pouring of the material from the
furnace.
WATER-COOLED COPPER ELECTRODE
PLASMA GAS INJECTION
ARC
TERMINATION
SPINNING
REACTOR
WELL
EXIT GAS AND
SLAG REMOVAL
6
« o
*o
FIGURE 2: Schematic of the demonstration reactor showing the bottom
pour configuration for exit gas and molten glass.
Work with the quarter-scale reactor resulted in the EPA selecting the PCR for inclusion in
the second round of the SITE program. We decide to design a larger reactor for the
demonstration program (Phase IE), using the insights gained from the quarter-scale reactor
tests.
266
-------
WASTE
4
FEEDER
VACUUM
^ PUMP
^ -i
CENTRIFUGAL
REACTOR
^ PLASMA ^ POWER
' TORCH SUPPLY
^ 4
SECONDARY
CHAMBER
|
QUENCH
TANK
t
JET
SCRUBBER
t
PACKED
TOWER
OXDEING
* — GAS
k
SURGE
TANK
' t
BUTTERFLY
VALVE
f
GAS
ANALYSIS
L
VEN
ATMOS
i
TTO
PHERE
i
STACK
FAN
4k
CHARCOAL
ADSORBER
t
BUTTERFLY
VALVE
t
t
FIGURE 3: Block diagram of the PCR-6
PHASE HI ( 1988-PRESENT)
Design of a large PCR was started in late 1987. This reactor was dubbed the PCR-6 since
it had a 6 foot diameter reactor tub. The design incorporated a bottom port for both exit gas
and molten glass Melt chamber geometry is shown in figure 2. Figure 3 displays a block
diagram of the major components of the PCR-6. The reactor section consists of a feeder,
primary chamber, rotary throat, drive mechanism, secondary combustion chamber and the
glass collection chamber. The gas treatment system consists of a quench tank, jet scrubber,
packed bed scrubber, demister, activated charcoal adsorber and a stack blower. Figure 4
shows a schematic diagram of the system elements.
Hazardous waste is initially loaded into a feeder. Depending on waste type, it may be a
screw, Archimedes spiral or a conveyor. In the Preliminary Tests for the EPA, specially
prepared, spiked soil known as Synthetic Soil Matrix (SSM) was charged from 5 gallon
267
-------
containers into a screw feeder or an Archimedes spiral feeder. The latter feeder will be
used to accept the contents of 55 gallon drums containing contaminated soil from actual
superfund sites in Butte. The next generation reactor will be even larger (2.4 to 3 m in
diameter); this will accommodate feeding of whole drums.
ROTATING REACTOR WELL
FIGURE 4: Schematic of PCR-6 showing the feeder, primary and
secondary chambers and the gas treatment system.
During feeder loading operation, a plunger valve isolates the feeder from the hot reactor.,
Air is purged through the feeder into the reactor to insure that no toxic vapors are present
upon opening the feeder door. Once loading has been accomplished, the feeder is sealed
from atmosphere and the plunger valve is opened to access the waste material into the
reactor.
The reactor itself consists of three chambers: a primary reaction chamber in which the
plasma torch is located, a secondary chamber where incomplete combustion products react
with supplementary oxygen to form water and carbon dioxide and a chamber where the
molten glass is collected. Inside the upper chamber is a rotating tub about 1.8 m (6 ft.) ID,
which spins at about 40 rpm. The plasma torch provides one termination of a DC arc, with
the other termination being initially the copper throat of the rotating tub. The walls of the
tub are lined with refractory material and a "skull" consisting of solidified waste material
(vitrified soil) built up during previous runs. The skull begins to melt under the action of
268
-------
the torch and eventually becomes a conducting medium. When the waste material becomes
sufficiently molten to be able to conduct, the plasma torch can be adjusted to move away
from the copper throat and onto the molten glass. This process continues most of the
surface of the rotating tub has been treated and turned to glass. When all of the waste
material in the rotating tub has been treated, the feeder is started and new waste material is
fed into the reactor. During feeding, the reactor is rotated at sufficiently high speeds to
keep the new feed material from rolling down to the throat of the rotating tub.
The waste material that is fed into the reactor falls into the existing melt and is itself melted
by interaction with the plasma. Any organic contaminants that the waste contains will be
vaporized and burned in the oxidizing atmosphere of the upper and lower chambers. The
inner core of the plasma column can reach temperatures of over 10,OOQ°C which maintains
the ambient temperature in the upper chamber above 1000°C. At such temperature, organic
vapors that have been released from the waste will immediately burn when mixed with the
hot oxygen rich furnace atmosphere.
Feeding continues until material build up causes the glass to reach the point of flowing into
the copper throat The feeder is turned off and time is allowed to complete the removal of
any organic material. This is determined by the change in oxygen level in the exhaust and
the clearing of soot as observed in the upper chamber. Upon completion of vaporization
and melting of the waste, the speed of the revolving reactor tub is slowed. Rotating the tub
at speeds of 5 to 10 rpm will allow gravity to overcome the centrifugal force on the molten
glass and allow it to flow through the center hole, falling through the secondary chamber to
be collected in the chamber below. The molten glass is collected in a pig mold where it
later solidifies. After the molten glass has been tapped the rotational speed of the reactor is
increased to resume the process. Presently the glass is allowed to cool in the collection
chamber. In the future, an air lock will facilitate the replacement of the full glass mold with
a new one to allow uninterrupted operation. The process, treating wastes, reacting
organics, casting glass and then resuming treatment continues until a planned shutdown.
Shutdown consists of turning off the plasma torch, allowing time for the molten material in
the reactor tub to solidify to form the refractory layer for the next operation, and then
stopping the tub rotation.
During the melting process, gases move out of the primary chamber through the throat into
the secondary chamber. If combustion of the organic material is not complete in the
primary chamber, additional O2 is injected into the secondary chamber in the vicinity of the
throat to cause further combustion, which is completed before the gas flows around a baffle
and out into the gas treatment system. Gas temperature in the secondary chamber reaches
•269
-------
at least 500°C. Since the exit duct is refractory lined, exhaust gas arrives at the quench
tank at nearly the same temperature. Since air is used as the source gas for the torch, the
gas stream will be acidic due to the formation of NOX from the reaction of N2 with C>2 at the
high temperatures of the plasma. HC1 formed from the break down of chlorinated
hydrocarbons will also add to the gas acidity. Other gases in the exhaust will include the
combustion products COa and water vapor, plus excess 02- In the event that combustion
is incomplete, CO will be formed. If CO levels are considered to be excessive, the system
will be placed into a recirculation mode as discussed below.
There are several cooling circuits associated with the furnace. The torch circuits are cooled
by a high velocity distilled water system. This system is closed loop, consisting of a water
pump, heat exchanger, flow interlocks for each circuit and a pump discharge filter. All
parts of the rotating reactor well exposed to high temperatures are cooled by a glycol-water
system similar to the distilled water system described above. The remaining cooling
requirements for the furnace are met by circulating water from an open loop building
system. These circuits include the reactor lid, reactor chamber upper (non-rotating) wall,
secondary chamber walls, collection chamber and hydraulic system.
GAS TREATMENT SYSTEM
The gas treatment system consists of four parts: scrubber, sampler, recirculation and
exhaust. The scrubber is composed of four separators, two pumps and a tank. The
sampling system includes a pump and monitors for CO2 and O2. Recirculation consists of
a surge tank and a water ring vacuum pump. .Exhaust system is comprised of ducting, a
charcoal bed and an exhaust blower.
The gas stream exits the reactor via the secondary chamber exhaust duct. The gas then
enters the scrubber where it first passes into a quench tank encountering a fine spray of a
mild caustic solution (NaOH). This solution is intended to react with the acidic
components of the gas and neutralize them. The caustic solution is circulated through
scrubber elements by means of two high pressure pumps. The quenched gas is then
directed into the jet scrubber, which is designed to reduce the pahiculate load in the gas.
From the jet scrubber the gas enters the bottom of a packed bed tower. The packed bed
consists of one inch pall rings with a spray at the top of the column. The gas is then drawn
into the demister which is a dry packed bed.
After leaving the demister the gas enters a serpentine duct constructed for sampling
equipment. The distances between and location of the ports were determined by the
270
-------
requirements of the various analytical samples to be taken. One sample train is used for
control. A diaphragm pump delivers the gas sample to O2 and CO monitors. The O2 and
CO levels help determine the effectiveness of combustion. Excessive CO levels will
require activation of the emergency bypass valve allowing the system to go into
recirculation to prevent the release of toxic vapors. The initial set point for CO level will be
determined during further testing.
FIGURE 5: Photograph of the PCR-6 installed in the Ukiah lab.
The recirculation system includes two air operated ball valves which respond to a signal
from the control panel. Under normal conditions the valve to the exhaust stack is open and
the valve to the surge tank is closed. During operation the surge tank is kept evacuated by
the water ring vacuum pump. Upon initiation of recirculation, the valve to the stack closes
and the valve to the surge tank opens. There is enough volume in the surge tank to operate
for about 5 minutes, time enough to correct a problem. If the abnormal CO levels persist
after 5 minutes the system would be shut down.
Following the recirculation system the gas enters the exhaust stack blower which is capable
of producing 80 in. of HaO at 100 cfm. At the outlet of the blower is a charcoal bed
adsorber to remove any residual organics not detected by the sampling system. It was
271
-------
found during the preliminary tests that the charcoal adsorber became loaded with
condensate, rendering it useless as a gas adsorber. Therefore it has been removed from the
system. A stack blower raises the gas pressure to atmosphere for release.
FIGURE 6: Photograph of PCR-6 showing the feeder, plasma torch and
primary melt chamber during feeder reloading operation.
TEST RESULTS
A photograph of PCR-6 installed in the Ukiah lab is shown in figure 5. This picture shows
the layout of the facility. The closed cooling water systems can be seen in the lower right
hand corner. The plasma torch can be seen protruding from the top of the reactor. The
primary melt chamber and drive chamber are the main features of this photograph. Figure
6 gives a good view of the plasma torch and feeder. The operators are in the process of
placing material into the feeder. During the operations with the spiked SSM the feeder deck
was an exclusion area while the surrounding space in the reactor room was a control area.
The off-gas treatment system is shown in figure 7; the large tank to the left is the surge tank
while the components in the center of the picture comprise the gas treatment system.
Shakedown tests with the PCR-6 were completed using local soil and oil (15% by weight).
These tests were done before and after the preliminary tests for the SITE program. During
272
-------
these tests we were able to feed up to 200 Ibs/hr of oil and soil. Initial runs were
performed using argon as a torch gas. The argon arc proved to be unstable in the current
configuration because of the interaction of the plasma with the chamber atmosphere. It was
found that high moisture contents caused this instability. Further testing showed that air
was a much more stable plasma gas when interacting with the furnace atmosphere. The
plasma torch will operate satisfactorily with air as a plasma gas at arc currents up to 1200
amperes at 400 to 600 volts. In cases where the moisture content was low the plasma torch
was able to operate with argon gas at up to 1200 amperes and 400 volts. Use of Qz/argon
mixtures as plasma gas proved unsatisfactory due to shortened electrode life.
FIGURE 7: Photograph of PCR-6 exhaust gas treatment system.
Preliminary tests for the EPA SITE program were run using Synthetic Soil Matrix (SSM)
(both unspiked and spiked material) (See Table 1) which was prepared by Enviresponse,
Inc. Preliminary Test #1 was performed on 20 April, 1989 using argon as a torch gas on
550 Ibs of unspiked SSM. The second preliminary test was conducted on 25 April on 300
Ibs of spiked SSM using air as a torch gas. The third preliminary test was conducted on
April 26. An air plasma was used to treat 200 Ibs of spiked SSM. Continuous Emission
Monitoring (GEM) was performed during both shakedown and preliminary tests. Data
taken by the CEM included the levels of Oi, NOX, CO, and THC (Total Hydrocarbons).
273
-------
For the final preliminary test, samples were taken on the exhaust gas, which included
VolatUes/Semivolatiles, Paniculate matter, Velocity/Volumetric Flow Rate (Method 1&2),
Fixed Gases (Method 3), Moisture (Method 4), Flue Gas-Metals (Method 12), and HCL
Samples were also taken from the slag, the feed material and scrubber water.
SYNTHETIC SOIL MATRIX (SSM)
COMPONENT
Sand
Gravel No. 9
Silt
Topsoil
Clay
WEIGHT %
31
6
28
20
15
SPIKING MATERIALS
COMONENT
Tetrachloroethylene
Anthracene
Bis (2-ethylhexyl) phthalate
Chromium
Zinc
mp/kg - ppm
3,277
7,361
3,702
1,898
28,306
Table 1 SYNTHETIC SOIL MATRIX (SSM) composition both spiked
and unspiked.
Most important to the development of the PCR was the ability of the reactor to vitrify soil
and transfer that glass to a collection chamber. The bottom pour arrangement was proven
during the shakedown tests. As shown in figure 8, it was possible to produce a fairly large
area of molten glass. After an operating procedure was established it was possible to slow
the reactor speed and pour glass from the melt chamber into the collection chamber. A
representative sample of the glass is shown in figure 9. The slag produced during the
preliminary tests was similar to that shown. At the time of this writing the results of the
leach tests were not yet available.
CEM results during shakedown tests on local soil indicated high levels of NOX (up to
15,000 ppm) on runs when air was used as a torch gas. This level of NOX amounts to less
than 1 Ib/hr at the system gas flow rate (~70 cfm). However, it was found that the level of
NOX dropped significantly (to 5000 ppm) when processing soils with organics (oil) added.
This can be explained due to the competing reactions involved in the formation of
274
-------
combustion gases and NOX. During the shakedown tests the level of oxygen followed the
feeding patterns for soil with high organic loads as would be expected. Shortly after
commencing feed, ah indication of lowering Qj levels was observed. This is an indication
of a reaction (combustion) between the decomposed hydrocarbons and O2- Additional O2
was injected into the melt chamber during subsequent runs to maintain an excess of O2-
During tests with excess C>2 maintained, a green colored flame was observed protruding
from the secondary combustion chamber into the collection chamber. This was a positive
indication of combustion in the secondary chamber.
Tables 2,3 and 4 give the GEM results for the three preliminary tests. P-l was performed
with argon as a torch gas and unspiked SSM, therefore no additional Oa was needed. The
GEM results were consistent with this type of operation. The C)2, CQz, CO and THC were
low for the duration of the test as expected since no free O2 or hydrocarbons were
introduced. There were however some startup spikes as shown by the 1-min max results,
which at this time cannot be explained. Levels of NOX produced were attributed to purge
air from the feeder and collection chamber and small air leaks into the system.
PCR-6, P-l CEM RESULTS
MONITOR
O2 (%)
C02 (%)
NOx (ppm)
CO (ppm)
THC (ppm)
1-MIN MAX
6.3
18.1
4952
1003
115
1-MIN MIN
0.6
0.2
1045
0
4
1-MIN NORM
1.7
4.0
3500
4
6
Table 2 CEM results from preliminary test #1, performed on April 20,
1989 using argon as a torch gas.
Spiked SSM and air torch gas were used during P-2, in which 300 Ibs of material was
treated. No additional O2 was added during this test in order to determine a base level of
additions for P-3. Levels of O2, CO2, CO and THC were all higher than in P-l as
expected. Since air was used as a torch gas, the NOX levels were higher in this case but not
to the levels observed when using air and unspiked soil. The Oa level dropped to 6.2 %
during feeding operation. It was decided to try to maintain the level at ~ 13 % for P-3
during feeding to try to insure complete combustion. Unexplainable CO spikes still
occurred, even when not feeding. One explanation for the CO spikes might be a periodic
275
-------
leak of ethylene-gylcol (cooling medium) from the copper throat area. The throat area had
been a source of leakage during shakedown tests.
Near the end of P-2, it was determined that there was enough slag buildup to facilitate a
pour. The reactor speed was lowered and -100 Ibs of glass was tapped. Samples of the
slag were taken for analysis but the results were unknown at the time of this paper's
writing. A photograph of slag (from a later test) similar to that collected in P-2 is shown in
figure 9. The slag that has been collected to date appears to glassy in nature and well
mixed. There have been problems with non-treated material passing through the throat, but
this has been alleviated through a feeder modification.
MONITOR
02 (%)
CO2 (%)
NO* (ppm)
CO (ppm)
THC (ppm)
PCR-6, P-2 CEM RESULTS
1-MINMAX
15.5
18.0
7808
>10000
333
1-MINMIN
6.2
1.2
4315
176
48
1-MIN NORM
14.0
3.0
5000
230
60
Table 3
CEM results from preliminary test #2, performed on April 25,
1989 using air as a torch gas.
Supplemental 02 was added during feeding operation for the P-3 test. This resulted in
higher normal levels of ©2 and CO2 and lower normal levels of CO and THC, as compared
with P-2. NOx levels were about the same as in P-2, but higher than had been seen in
previous tests with higher levels of organics. The SSM used in P-2&3 contained only 1%
organics, not enough to have an impact on the NOX levels. This test consisted of feeding
200 Ibs of SSM, which is not a large amount of material considering that the system had
been drained the night before. It was because of this that it was difficult to obtain a
representative sample of slag from P-3; not enough material had been treated to obtain a
good pour.
Preliminary results by Radian indicate that DRE's of 0.9999 to 0.99999 were obtained.
High levels of organics were found in the slag from P-3 due to untreated material being
thrown down the throat by the feeder. This was only discovered as a problem at the time
of the P-3 test (the feeder was subsequently modified).
276
-------
PCR-6, P-3 CEM RESULTS
MONITOR
C>2 (%)
- CQz (%)
NOx (ppm)
CD (ppm)
THC (ppm)
1-MINMAX
18.4
9.8
6536
375
194
1-MINMIN
12.9
0.1
4652
17
16
1-MINNORM
17.5
0.2-4.0
5500
50
25
Table 4 CEM results from preliminary test #3, performed on April 26,
1989 using air as a torch gas.
CONCLUSION
Based on the information gathered to date, the PCR project has reached some degree of
success. The major objectives of the system have been accomplished, in that heavy metals
and organic wastes have been treated with favorable results. Hydrocarbons have been
broken down and reacted with Oi to produce CO2 and FfeO. Molten slag has been poured
from the reactor and collected. To date, all results indicate that these objectives have been
reached with a high degree of success.
Improvements will be made to this reactor (PCR-6) before the EPA SITE demonstration.
The seals around the copper throat will be improved. This will allow higher heat loads in
the area of the throat during feeding to insure total treatment of the toxic material.
Investigations are proceeding to increase the power input into the melt chamber.
Modifications are being planned to enable higher feed rates (up to 600 Ibs/hr). This will
require more advanced slag handling techniques to provide for increased amounts of molten
glass. Materials of construction for some components will be changed to resist the acidic
attack of the NOX and HC1. Modifications to the scrubber system will be required to handle
the higher gas loads.
277
-------
FIGURE 8: Photograph of PCR-6 melt chamber showing the extent of the
molten material.
FIGURE 9; Photograph of vitrified slag from local soil and 15% oil.
278
-------
At this time the equipment is being disassembled and sent to Butte, MT. for a SITE
demonstration scheduled for the summer, 1989. The feed for the demonstration tests will
be a mixture of mining wastes (soil containing heavy metals) and wood-treating waste oil
(containing pentachlorophenol and other components). Feed rates of 100 Ibs/hr for 6
hours are to be performed for the SITE test. It is planned to operated the plasma torch with
air at power levels of 600 volts DC and 1000 amps. O2 will be added to both the primary
melt chamber and the secondary chambers during feeding operations. It is planned at this
time to feed at rates of 200 Ibs/hr for 30 minutes and treat the material for 30 minutes while
reloading the feeder, then feed again during the 6 hour period. Six hundred pounds of
material will be enough to produce a reasonable slag pour. Results from the demonstration
will be available in report form about six months after the three planned replicate runs are
made.
After the EPA SITE test, it is planned to continued operations in Butte under the direction
of the DOE to further evaluate the PCR. A separate program has been established for this
work. Optimum feed rates and power levels will be investigated during these tests on a
variety of materials. The long term goals of the DOE are to process low level radioactive
wastes. Efforts to implement this technology will be provided by personnel at the Idaho
National Engineering Laboratory (INEL).
R. C. Eschenbach and K. D. Boomer, "Plasma Arc Stabilization of Hazardous
Mixed Wastes", American Nuclear Society Winter Meeting, Los Angeles, CA.
1987.
J.W. Sears, R.A. Hill & R.C. Eschenbach, "Stabilization and Decomposition
of Toxic and Radioactive Wastes by Transferred-Arc Plasma.", Incineration
Conference (8th International Conference on Thermal Destruction of
Hazardous, Radioactive, Infectious &Mixed Wastes), Knoxville, Tenn., May,
1989.
279
-------
HAZCON. INC.
SUPERFUND INNOVATIVE TECHNOLOGY EVALUATION
'FINDINGS AND CONCLUSIONS
by: Timothy E. Smith
HAZCON Engineering, Inc.
Brookshlre, Texas 77423
INTRODUCTION
In October, 1987, the U.S. Environmental Protection
Agency's.Offlees of Research and Development (ORD) and Solid
Waste and Energy Response (OSWER) Initiated their first test
of a sol Id'If Icatlon/stabI I Izatlon (s/s) process under the
Superfund Innovative Technology Evaluation (SITE) program.
The technology demonstration was developed by the .founders
of HAZCON, Inc., based In Texas.
The HAZCON solidification process was selected for Its
potential to effectively stabilize and solidify wastes
containing high concentrations of organic contaminants.
High organic wastes Inhibit the reaction mechanisms of most
s/s processes, and thus had rendered the technology non-
effective as a pretreatment alternative for land disposal of
these wastes.
The major objective of the demonstration was to develop
reliable cost and performance data on the HAZCON process so
that It could be adequately considered In Superfund decision
making. Evaluation criteria were established Jointly by the
EPA and HAZCON to evaluate the effectiveness of the process
In Immobilizing contaminants and In Increasing matrix
Integrity.
280
-------
The DouglassvlI Ie, Pennsylvania National Priority List
(NPL) Site, a former oil reprocess Ing fac11Ity, was selected
as the test site. The site was placed on the NPL In 1985
due to the presence of high levels of organic and Inorganic
contaminants. Contaminants Include PCB's, heavy metals,
volatile and semlvolatlle organlcs, base neutral acids and
other toxic materials. An estimated 250,000 cubic yards of
material may be contaminated.
PROGRAM OBJECTIVES
Five major objectives of the HAZCON SITE demonstration
were to determine the following:
1. Ability of the stabilization/solidification
technology to Immobilize and solidify the site
contaminants.
2. Effectiveness of the technology In treating
wastes with contaminant concentrations varying
over the range 1-25% by wt. oil and grease.
3. Performance and reliability of the process
system.
4. Long-term stability and Integrity of the
sol Id If led mass.
5. Costs for applying the technology to
commercial size or Superfund sites.
281
-------
METHODS AND MATERIALS
Contaminated soils from the site were processed by
HAZCON's Mobil Field Blending Unit (MFU). The system Is
truck mounted and typically used by HAZCON for small field
projects such as the SITE demonstration. The system holds,
meters and homogenizes the waste feed with Portland cement,
Chloranan (HAZCON's proprietary additive) and water. The
processed material was then placed In 1 cubic yard-sized
forms for curing.
Samples were collected both before and after treatment.
These samples were subjected to an extensive testing
protocol, to Include TCLP leachate analysis, permeability,
weathering and strength tests, and mlcrostructuraI analysis.
Materials from six plant areas were tested. Each area
contained varying levels and types of contaminants across a
variety of waste matrices. The following designations were
used to refer to the six plant areas:
LAN Lagoon Area North
LFA Landfarm Area
PFA Plant Processing Area
DSA Drum Storage Area
FSA Filter Cake Storage Area
LAS Lagoon Area South
Five cubic yards of material were processed from each
of the first five areas listed, twenty five cubic yards were
treated from Lagoon Area South (LAS).
In order to effectively gauge HAZCON's claim to
Immobilize volatile organlcs, the EPA requested permission
to Inject toluene at 125ppm Into the feed from DSA, PFA, and
LFA.
282
-------
Core samples were analyzed after 7 and 28-days of
curing. These analysis were performed to determine the
effects of time and weathering on the mass. This latter
series of tests provides additional evidence of the long
term stab 11Ity of the HAZCON treatment process.
FINDINGS AND DISCUSSION
The six test areas offered a wide diversity of waste.
The oil and grease ranged from \% by weight at the DSA to
25% at FSA. PolychI or Inated blphenols (PCBs) were detected
up to 80 ppm by weight with the maximum concentration at
LAS. Lead contamination concentration ranged up to 2.5% by
weight. Volatlles and base neutral acid extractable (BNAs -
semtvolat 11es) organlcs reached levels of about 100 ppm In
some areas.
Permeabilities of the treated soil were very low, In
—8 ~9 •• 10
the range 10 to 10 cm/sec. A value of 10 Is
generally considered an Indication of an Impermeable solid.
Unconflned compresslve strengths (UCS) ranged from about 200
psl for FSA to 1500 psl for PFA. These values are quite
satisfactory from a load bearing point of view, I.e.,
equipment traffic.
TCLP leaching tests, the results shown In Table 1 are
discussed below:
Metals - the leachates for the solidified soils
showed metal levels at or near the detection
limits. The results were a factor of 500 to 1000
times better than the untreated leachates.
233
-------
TABLE 1. RESULTS OF TCLP LEACHING TEST
Leachate
Leachate
ro
00
.£»
Plant
Area
DSA
LAN
FSA
LFA
PFA
LAS
Soil Concentrat
Lead* VOC+
3,230
9,250
22,600
13,670
7,930
14,830
ND1
2
150
ND
0.4
6
Ions, ppm
BNA°
12
21
534**
37
18
40
Untreated Soil,
Lead VOC
1.5
33.
18.
28.
22.
53.
0.9§
0.03
1.03
5.90
1.80
0.07
mg/l
BNA
ND
1.02
2.86
0.01
0.01
0.01
Treat Soil - 7 days, mg/l
Lead VOC BNA
0.015
0.002
0.07
0.04
0.01
0.014
0.40
0.02
0.74
0.23
0.40
0.06
0.055
1.340
3.910
0.040
0.090
0.470
* The great majority of the metals Is lead.
+ Primarily Toluene, trlchlorethene, tetrachloroethene, ethylbenezene, xylenes.
0 Phthalates, phenols, naphthalene.
H ND - not dlscernable
§ Toluene was Injected Into untreated soil samples for DSA, LFA, and PFA before
performing TCLP. Concentration Injection equal to level measured In 7 day solids.
** Very high In naphthalene and phenols.
-------
Volatlles - the primary compounds detected were
trJchloroethene, tetrachloroethene, toluene and
xylenes. Only the leachates for the untreated
soil and 7-day cores were analyzed. The levels of
leachable contaminants were approximately 2 to 20
times less In the treated materials than the
untreated leachates.
BNAs - the compounds detected In the leachate were
phthalates and phenols. The phthalates were
reduced to near their detection limits of .010
mg/l In both the treated and untreated soil
leachates. The total phenols In the leachates
were In the range of .030 to 3.85 milligrams per
liter, with the same concentration levels seen In
both the untreated and treated soil leachates.
PCBs - were nonquantlflab Ie In the leachates.
CONCLUSIONS
The following conclusions were drawn by the EPA and
their subcontractor Env Iresponse, Inc. upon reviewing the
data on the HAZCON process. They are:
o The process can solidify contaminated material high
In organlcs. Soils at the Doug IassvI I Ie Superfund site with
up to 25% organlcs were solidified. Other applications
showed successful solidification of petroleum refinery waste
streams, organlcs, water high In organlcs from a waste
storage pond, metal finishing sludge, and other less clearly
defined wastes.
285
-------
o Immobilization of heavy metals was observed, with
leachate Improvements for lead and zinc In excess of a
factor of 100.
o Organic contaminants, YOCs and BNAs, were hot
Immobilized for the most part. The extensive testing for
the Demonstration Test and other test programs showed no
Immobilization of the organlcs. However, there were two
Instances where Immobilization of organlcs occurred.
o The physical properties of the treated wastes are In
general quite satisfactory. High DCS, Iow permeabI I 11les,
and satisfactory results of weather tests were obtained.
However, large volume Increases In treated soils can be
expected, and the mlcrostructural analyses of the solidified
soil materials Indicates a potential for long-term
durability problems.
o Application for Immobilization of heavy metals (up
to 2.3? by wt.) In wastes containing high organlcs, up to at
least 25% by wt oils, has been shown. Successful
Immobilization of higher quantities of heavy metals at even
higher oil and grease levels would be anticipated.
o Immobilization of VOCs and BNAs did not occur In the
SITE Demonstration test on soils up to 25% by wt oil and
grease, and Immobilization of other organlcs, as reported by
other Investigators, was also unsuccessful. However,
Immobilization of some petroleum refinery wastes was
successfuI.
286
-------
o Therefore, applications for Immobilizing organic
contaminants, compared to a s Imp Ie sol Id IfI cat Ion process
with only cementltlous materials, may have to be tested on a
site-by-slte basis to prove applicability of the HAZCON
process. For high organic content wastes, solidification
may be very dIffIcuIt; the use of Chloranan will enhance
solidification of organlcs.
The estimated cost for commercial scale application of
the HAZCON process was not reported In the Demonstration
Final Report. Iristead a figure of $205 per ton of
contaminated soil was reported, representing the estimated
per ton cost of the 50-cubIc yard demonstration. This Is
much higher than the normal $50 to $65 per ton costs HAZCON
has charged clients In the past for commercial scale
applIcatIon.
One final observation made by this writer during the
demonstration and subsequent Interpretation of the data Is
that the use of EPA contracted support services In the
performance of this and other SITE evaluations may Introduce
an element of bias to the program. An element which, either
for or against the demonstrated technology, will distort the
true advantages and/or disadvantages of the technology being
evaluated, thereby corroding the credibility of the entire
SITE program. Inaccurate and misleading Interpretation of
the data from this test has been observed, and conclusions
have been pub I I shed wh Ich can not be supported. One EPA
contracted party Involved In this evaluation now operates a
solidification/stabilization firm.
287
-------
REFERENCES
Sawyer, S« and dePercIn, P. Volume 1, Technology
Evaluation Report SITE Program Test, HAZCON
Solidification Doug IassvlI Ie, Pennsylvania;
EPA/540/5-89-001a, U.S. Environmental Protection Agency,
Cincinnati, OH, 1989.
Final Draft Report: Sawyer, S. and dePercIn, P. SITE
Program, Applications Analysis: Assessment of the
HAZCON, Inc. Solidification Technology. U.S.
Environmental Protection Agency, Cincinnati, OH, 1989.
288
-------
SITE: Fixation of Organic and Inorganic Wastes/
Intimate Mixing Technique
by
Carl L. Brassow, J. T. Healy and R. A. Bruckdorfer
Soliditech, Inc.
Presented at
Forum on Innovative Hazardous Waste Treatment Technologies:
Domestic and International
Atlanta, Georgia
June 19-21, 1989
289
-------
ACKNOWLEDGMENT
Soliditech, Inc. would like to express its thanks and
acknowledge the help, cooperation, suggestions and critique of
the following individuals: Dr. Walter E. Grube, Jr., EPA SITE
program manager in the Risk Reduction Engineering Laboratory of
Cincinnati, Ohio; Dr. Kenneth Partymiller of PRC Environmental
Management, Inc; Mr. Robert Soboleski of the New Jersey
Department of Environmental Protection; Mr. George Kulick, Jr.
of the Imperial Oil Coimpany; Mr. Robin Somerville of Solidwaste
Technology, Inc. of Manhatten, Kansas; Dr. L. T. Fan, Chairman
of the Department of Chemical Engineering, Kansas State
University of Manhatten, Kansas; Dr. Danny Jackson and Ms. Debra
Bisson of Radian Corporation, Austin, Texas; Mr. Larry Malone
and Arthur Malone of Malone Service Company, Texas City, Texas;
and ENSR Corporation of Houston, Texas.
290
-------
SITE: Fixation of Organic and Inorganic Wastes/
Intimate Mixing Technique
Carl L. Brassow, J. T. Healy and R. A. Burkdorfer
Soliditech, Inc.
I. INTRODUCTION
Solidification/Stabilization of industrial solid waste,
particularly wastes classified as hazardous and toxic under
current U. S. EPA rules and regulations is a developing
technology. Physical/chemical processes -involving the use of
pozzolans or cements with the addition o£ various special
additives are the most prevalent. These processes may be
applied to the waste in-situ or the waste can be removed,
treated, mixed using an intimate mixing system and then replaced
or removed to another repository. The Soliditech process
described in this paper is an intimate mixing process based on
the use of pozzolans or cement and various additives, especially
URRICHEM, that enhances the ... ability of the mixture to
incorporate organic compounds into the matrix and reduce the
potential for these compounds to leach from the mass.
II. BACKGROUND INFORMATION
Soliditech, Inc., is a Houston, Texas based company established
to apply the solidification technology described herein to the
remediation and cleanup of abandoned disposal facilities and
sites and to the treatment of currently generated wastes.
Soliditech, Inc. is a wholly owned subsidiary of United Resource
Recovery, Inc., which is a new waste disposal company that will
incorporate the solidification process at a new facility to be
constructed in the near future.
Soliditech submitted the proposal to the U S. EPA to demonstrate
the technology under the Superfund Innovative Technology
Evaluation (SITE) program in 1987. The proposal was accepted
and ultimately a demonstration site selected which is located in
New Jersey. The site is the Imperial Oil Company site in
Morganville, New Jersey, some twenty or thirty miles south,
southeast of Newark. The site has been in existence since the
early 1900*s and has been used for several purposes from food
processing to recycling oil which is its current purpose. (See
Figure 1). Various waste streams are present at the site most
of which include some form of petroleum hydrocarbons in a matrix
of soil, filter media or as a tank bottom sludge.
The field demonstration was conducted the first week of
December, 1988. Three different waste streams were treated as
part of the demonstration which included a soil contaminated
with oily sludge, a filter media with a high percentage of
291
-------
hydrocarbons and an oily tank bottom sludge. The latter stream
was co-treated with the filter media during the demonstration.
A total of fifteen yards of treated waste was collected, sampled
and tested. The waste was placed in plywood forms which were
removed after 28 days. The waste blocks were stacked on a
protective sheet of HDPE and covered with additional sheets for
protection and storage long term. Sampling, testing and
observation of the blocks will continue for five years.
III. PROJECT DESIGN
The Soliditech process is designed to solidify waste materials
by mixing the wastes with a chemical reagent, various additives,
and pozzolanic materials. The following subsections describe
the Soliditech process chemistry, treatment process, and the
main components of the process.
Soliditech Technology
The Soliditech process solidifies wastes by use of UKRICHEM (a
proprietary chemical reagent, U.S. patent pending), additives,
pozzolanic solids, and water. The proportions of reagent,
additives, and pozzolan are optimized for each particular waste
requiring treatment. The solidified material displays
properties of excellent unconfined compressive strength, high
stability, and a rigid texture similar to that of concrete.
The waste material to be treated is first passed through a
screen with 4-by 4-inch openings to remove large rocks, debris,
or other materials. This screening step is performed to ensure
that laboratory samples do not include large rocks. During
normal operations, only very large material would be removed
prior to treatment. The oversized material is collected and
drummed for off-site disposal. Following screening, the waste
material to be treated is placed in the Soliditech mixer, where
the reagent and other additives are dispersed throughout the
waste material. The reagent and additives aid in the chemical
and physical immobilization of the hazardous constituents
contained in the waste. It is sometimes necessary to add water
to the waste material to achieve the proper moisture level and
to achieve uniform blending of the raw waste material.
The waste mixture is then blended with pozzolanic materials.
Pozzolanic materials are siliceous or siliceous and aluminous
materials that inherently possess oxides (lime) to form
compounds possessing cementitious properties. Pozzolanic
materials that may be used in the solidification process include
Class C or F flyash, cement kiln dust, slag cement, portland
cement, and steel baghouse dust. Depending on. the pozzolanic
material used, calcium oxide (or its hydrated form, calcium
hydroxide) may have to be added to assure a hard set. (Portland
cement was used for the demonstration).
292
-------
Hazardous compounds are immobilized by one or more of the
following processes: encapsulation, adsorption, and or
incorporation into the crystalline structure of the solidified
material. The waste may actually penetrate any porous
structures of the pozzolans as well as by bonding to the surface
of the pozzolans by either ionic attraction or electrostatic
forces.
Water contained in wastes that are treated by the Soliditech
process and water added during the processing are involved in at
least two types of reactions.
o Hydration reactions
o Equilibration with the environment through evaporation
The Paint Filter Liquids Test (SW-846, Method 9095, U.S. EPA
1986) which Soliditech routinely performs has shown that no free
liquid remains once the solidification processing is complete.
As each solidification mixture is optimized for a specific batch
of waste material and because considerable variability can exist
between types of pozzolans, each batch of treated waste material
potentially unique. All results from this demonstration
is
should be considered to reflect on the particular formulations
and pozzolans used for the demonstration.
Treatment Process
The treatment system of the Soliditech process is presented in
Figure 2. The treatment system, consists of a self-powered,
variable speed mixing unit mounted on a low boy trailer, a
control box containing all mixer controls, a hoist to raise and
lower the mixing unit, a vibrator to aid hopper clean out, a
liquid reagent storage tank with a metered feed system, a hopper
for storing and dispensing the pozzolanic solids, tanks for
storing process water, and a metered gate for discharging
treated waste material.
Three designated waste materials were treated during the
demonstration: Off-Site Area 1 (soil contaminated with oily
sludge, Waste Pile) (filter cake) and the Abandoned Tank (oily
sludge).
For Soliditech demonstration process, a batch of contaminated
soil from Off-Site Area One was collected, transported on-site
by a licensed solid waste transporter, sampled, and stored in a
lined and covered dump truck until it was to be treated. The
waste pile material was sampled immediately prior to its
treatment. Sludge from the abandoned storage tank was collected
and sampled prior to treatment and stored in 55-gallon drums.
293
-------
Immediately prior to treatment, each batch of waste material
from Off-Site Area One and the waste pile filter material was
weighed or its density and volume measured and transferred to
the mixing unit by a front-end loader. Prior to treatment of
the mixture of sludge and waste filter material, dry waste
filter material was collected in a front-end loader, weighed or
its density and volume measured, and placed in the mixing unit.
The weight or volume and density of sludge in the drums was
determined. The drums of sludge were lifted and poured into the
mixing unit. Adding the solid waste filter material first
minimized splashing of the liquid waste.
Once the contaminated waste material was in the mixing unit, a
measured volume of URRICHEM liquid reagent was pumped into the
mixing unit. Additives were measured and added to the mixer.
The mixing unit was turned on and the URRICHEM and additives
were thoroughly blended into the waste. When water was required
for the mix, the amount was measured and pumped into the mixing
unit.
After the waste material, reagent, additives, and water were
thoroughly blended, the volume or mass of pozzolanic material
was determined and, as required, was added to the mixing unit.
Mixing requires a minimum of 20 to 30 minutes until all of the
ingredients are thoroughly blended into a wet pasty mixture.
The mixing unit was lifted by its hydraulic legs and the slide
gate operated to dispense controlled quantities of the mix into
the larger sample containers and empty, plastic-lined, plywood
forms. As the forms were filled, samples of the treated
material were collected.
After each demonstration run was completed, the operation was
shut down and the process equipment was decontaminated using a
high pressure water and steam cleaner. This water was collected
and drummed for off-site treatment or disposal.
Test run data including information on material usage, weights,
volumes and all operating parameters, were recorded and
maintained in a field log by Soliditech personnel for each test
run of waste processed.
To confirm the Soliditech records, EPA kept similar records
containing similar information, as well as the time at which all
operations were performed. The EPA run records included the
results of the particle (sieve) size analysis for the treated
material and the results of slump tests.
IV. TEST RESULTS
Three types of waste material and one control mix were treated
as part of the demonstration. The control mix was a mixture of
294
-------
clean sand, concrete, pozzolan (portland cement), URRICHEM and
other additives that were part of the waste stream formulation.
The waste streams consisted of contaminated soil, waste filter
cake material and oily sludge.
Untreated waste samples were collected for each test parameter
from each of the three waste streams. These samples were
analyzed for total chemical constituents, physical
characteristics and the amount of solubles removed by
leaching/extractions. The results allow a direct comparison of
physical and chemical properties between the treated and
untreated waste and a determination of effectiveness of the
treatment process.
In addition to discharging the treated waste into large l -
cubic yard plywood forms, numerous cylindrical sample containers
or forms were filled with treated waste and allowed to cure 28
days. After curing, the small sample cylinders were shipped to
the laboratories for analysis. The final product was a
monolithic material with measurable structural strength. The
wooden forms were removed and the waste monoliths were placed in
an enclosed on-site storage area for long term monitoring. Long
term studies will include a six month leaching test and
performing other extraction procedures at various times up to
five years after treatment.
The analyses of the samples collected before, during, and after
the Soliditech demonstration are summarized in Tables 1 and 2,
and discussed below:
o Untreated Waste — Untreated waste from the site consisted
of contaminated soil, filter cake, and filter cake/oily
sludge. These wastes contained 2.8 to 17 percent oil and
grease, with relatively low levels of other organic
compounds. PCB (Aroclors 1242 and 1260) concentrations
ranged from 28 to 43 ing/g; arsenic concentrations from 14 to
94 mg/kg; lead concentrations ranged from 650 to 2,470
mg/kg; and zinc concentrations from 26 to 151 mg/kg.
o Treated Waste — The Soliditech stabilization process
produced solidified waste with high structural stability and
low permeability. UCS values ranged from 392 to 856 psi^
Permeability values ranged from 8.9 x 10~9 to 4.5 x 10~7
cm/sec. Because of the cementitious additives in the
Soliditech process, pH values of the solidified wastes
ranged from 11.7 to 12.0. Arsenic concentrations ranged
from 28 to 92 mg/kg; lead concentrations from 480 to 850
mg/kg; zinc concentrations from 23 to 95 mg/kg; and PCB
(Aroclors 1242 and 1260) concentrations from approximately
15 to 41 mg/kg. Low concentrations of phenol and p-cresol
295
-------
were found in solidified filter cake and filter cake/oily
waste samples. These compounds were not detected in the
untreated wastes.
o Control Mixture — The control mixture contained 20 mg/kg
lead. PCBs, phenols, and cresols were not detected in the
control mixture. The reagents used for the solidification
could not be analyzed for phenol, o-cresol, and p-cresol
because of the high alkalinity in the control samples. Low
levels (0.06 ug/L, total) of volatile organic compounds were
detected in the TCLP extract of the control mixture.
o Extract of Untreated Waste — Arsenic, lead, and zinc were
found in EP, TCLP, and BET extracts of the untreated wastes.
No PCBs were detected in the TCLP extracts of the untreated
wastes. Total concentrations of up to 1.3 mg/1 of volatile
organic compounds and up to 0.38 mg/1 of semivolatile
organic compounds were detected in the TCLP extract of the
untreated waste. Oil and grease concentrations of 1.4 to
1.9 mg/1 were detected in the TCLP extract of the untreated
waste. Untreated wastes could not be tested by ANS 16.1.
o Extract of Treated Waste — Significantly reduced amounts of
metals were detected in the TCLP, EP, BET and ANS 16.1
extracts of the treated waste. No PCBs or volatile organic
compounds were detected in the TCLP extract of the treated
waste. Phenol, p-cresol, o-cresol, and 2,4-dimethylphenol
were detected in the post-treatment TCLP waste extracts.
Oil and grease concentrations of 2.4 to 12.0 mg/1 were
detected in the TCLP extracts.
V. COSTS
The cost elements identified in the demonstration provide a
means to estimate the cost of solidification for a typical site
remediation. The actual costs of the demonstrations are biased
when reduced to a unit cost basis by the impact of mobilization,
demobilization, sampling and testing, decontamination time, low
utilization rates.
Cost elements identified in the demonstration include the
following:
Mobilization/demobilization of equipment
Capital cost of equipment
Cost of materials and additives
Cost of labor
Analytical expenses
Health and safety training
General overhead expenses
296
-------
There are various methods of applying these costs to any
particular project. The unit costs on the SITE demonstration
would be extraordinarily high if all the costs were to be
applied to the approximate 15 cubic yards of waste treated
during the 5-day demonstration period. As a point of
information, this cost would have been about $2700 per yard
assuming all pertinent costs were allocated to the treated waste
volume.
In an actual field cleanup or remediation situation, many of the
identified costs would become insignificant when spread over the
much larger volumes of waste treated. Assuming that most
factors would remain relatively constant, that the range of
pozzolan cost was between $25/ton and $70/ton and that the
throughput was about 15 cubic yards and 30 cubic yards of waste
per hour, then the relative cost of the process would be between
$46/yard and $75/yard for about 10,000 yards and $29/yard and
$58/yard for about 100,000 yards of waste respectively. These
numbers are for illustrative purposes only to reflect the impact
of waste quantity on unit cost. It is important to point out
also that these costs do not reflect waste handling costs prior
to treatment and do not reflect post-treatment handling and
disposal costs.
VI. CONCLUSIONS
Conclusions drawn from any study or effort are generally a
combination of objective results and subjective inferences based
in part on these results.
This SITE
following:
demonstration was conducted to determine the
The effectiveness of the technology to
stabilize waste materials found at the site.
solidify and
The ability of the solidified materials to maintain physical
properties and structural stability over a five-year period.
The change in volume and density of the solidified material
after adding pozzolans, water, reagent, and other additives.
Reliable capital and operating
Superfund decision making process.
costs for use in the
The results of this demonstration indicated that the process was
effective in solidifying and stabilizing the waste streams on
the site. The data indicate some noteworthy effects of using
additives with high pH values, specifically, 1) testing
procedures may need to be altered to accommodate pH range and 2)
there appears to be an effect on phenols which are apparently
selectively leached out of the matrix. Further study on these
effects is currently being conducted.
297
-------
The range of unconfined compressive strengths and low
permeabilities verify the solidification objective. The
objective of five years of testing hasn't been accomplished but
one can infer that concrete is not likely to revert to basic
constituents even though some cracking or spalling may likely
happen.
The change in volume ranged from 0 to 60 percent but the median
appears to be less than 30 percent. This is an important
parameter when estimating disposal volume of treated waste and
this level is probably an acceptable increase at this point in
time.
Capital and operating costs of an intimate mixing operation or
technique are also reasonable. Any treatment technology is
going to add a significant cost to waste management over the
no-treatment option. The objective is to optimize the treatment
costs consistent with the risk the untreated waste may pose in
the future. This should not be considered as an isolated
parameter but rather by a systems approach to reducing
environmental risks for the future.
This particular demonstration was successful in meeting the
demonstration objectives. However, the SITE program itself has
certain goals, four of which are listed below:
o To identify and remove impediments to the development and
commercial use of alternate technologies.
o To conduct a demonstration of the more promising innovative
technologies to establish reliable performance and cost
information for site characterization and cleanup
decision-making.
o To develop procedures and policies that encourage the
selection of available alternative treatment remedies at
Superfund sites as well as other waste sites and commercial
facilities.
To structure
technologies.
a development program that nurtures emerging
For the most part, these goals, as they were applied to this
project, we addressed successfully. Continued efforts must be
made within the consultant industry and the "regulated"
community to promote the use of these emerging technologies.
Satisfactory approaches will not evolve if the program goals are
not applied.
298
-------
DEMONSTRATION LOGISTICAL AREAS
AT THE IMPERIAL OIL FACILITY
ABANDONED STORAGE
LONG-TERM
MOMTOHMQAREA
RAW POZZO.AN STORAGE
UMT/WATER TRUCK
VISITOR VIEWMQ AREA
VISITOR ACCESS ROUTE
\XSTAGMQAHEA
*T
\ . PRC-EM TRAILER
IJ \ MP0VAL CH. GUARD TRAUER
VM w* eomim CWM
PftlVATC HOMKS
o
FIGURE 1
299
-------
WATER
FRONT BUCKET LOADER
WITH BUCKET ABOVE
MIXER IN DUMP POSITION
SOUCHTECH, INC,
MORSANVILLE, NEW JERSEY
I SOUD1TECH PROCESSING EQUIPMENT
OttATD: «/ll/li REVUHZDl 1/15/11 j JOJttCHJWO
PRO ENVIRONMENTAL MANAGEMENT. INC.
FIGURE 2
300
-------
TABLE 1
PHYSICAL PROPERTIES
Filter Cake/Oily
Sludge Mixture
Filter Cake Sludge Mixture Off-Site Area One
Untreated Treated3 Untreated Treated" Untreated Treated*
Bulk Density
(g/cm3)
Permeability
(cm/sec)
Unconfined
Compressive
Strength
(psi)
Loss of
Ignition
1.1
NA
53.7
1.4
392
41.3
1.2
1.7
NA
70.0
856
34.0
1.3
NA
36.3
1.6
NAb 4.5 x 10"7 NA 8.9 x 10"9 NA 3.4 x IO'8
677
33.7
Water Content 28.7
21.0
58.1
Notes:
a Treated waste sampled after a 28-day curing period.
b NA = Not analyzed.
14.7
23.5
12.6
301
-------
TABLE 2
CHEMICAL PROPERTIES
Filler Cake
Chemical Untreated
Analysis8 Waste
pH 3.4
VOCsd NDe
SVOCs9 ND
PCBsh 28
Oil and Grease 170,000
Arsenic 26.4
CO
ro Lead 2,200
Zinc 26
Notes:
Lcacnate Uachate Lcachate Leachate ' Leachate
from from from . from from
Treated Untreated Treated Untreated Treated Untreated Treated Untreated Treated Untreated
Waste.0 Waste0 Waste0 Waste Waste" Wade0 Waste0 Waste Waste6 Waste0
n-8 4.6 10.8 3.6 12.0 4.8 11.6 7.9 12.0 5.1
ND 0.27f ND 50f ND 1.3f ND 10 ND 0.87f
3fif ND 1.2 63f 17f 0.38 0.97f 79f 16f 0.12f
16 ND ND 43 15 ^ ND ND 43 40 ND
77.000 U 4.4 130,000 60.000 1.6 2.4 28,000 46,000 1.9
28 0.005 ND 14 40 0.01 ND 94 92 0.20
680 4.3 0.002 2,500 850 5.4 0.01 650 480 0.50
23 0-30 ND 150 54 1.3 ND 120 95 0.60
Leachate
from
Treated
Waste0
11.5
ND
0.32f
ND
12
0.02
0.01
ND
Analyte concentration units for the untreated and treated waste are mg/Kg. Analyte concentration units for the leachate from untreated and treated waste arc mg/L
Treated wastes were sampled after a 28-day curing period.
0 Lcachate values
d VOCs
e ND
ThftSP. unliti»c rm
refer to results from TCLP test.
volatile organic compounds.
not detected.
• i i
These values contain low levels of acetone, melhylene chloride, various phthalales, or other analytes which are commonly attributed to sampling or analytical
9 SVOCs =
contamination.
PCBs
scmivolatilc organic compounds.
polychlorinated biphenyls.
-------
ADVANCED CHEMICAL FIXATION
of
ORGANICS and INORGANICS/IN-SITU
TREATMENT
By
JEFFREY P. NEWTON
INTERNATIONAL WASTE TECHNOLOGIES
WICHITA, KANSAS
303
-------
INTERNATIONAL WASTE TECHNOLOGIES Inc.
Conceptual Basis of the Advanced Chemical Fixation (ACF)
Technology
Waste treatment techniques that are composed of Portland cement, fly-
ash, cement kiln dust, quick-lime, soil, clay, asphalt, sodium silicate,
slag, gypsum, etc., in various combinations for solidification/stabilization
(S/S) have been used for a number of years all over the world. Users
and suppliers of these technologies have in some cases also used the
term chemical fixation (CF) to define various compositions of these
materials, possibly with the addition of trace compounds to accentuate
their effectiveness with certain waste types. Whatever the terminology
used, S/S or CF, particular compositions of these materials are the basis
of the proprietary treatment products on the market. Certainly one of
the major appeals of this class of treatment technology is the relative
lower cost, if it is sufficiently effective for the waste types it is applied.
S/S has been viewed by many in the environmental field as a physical
or civil engineering process instead of a sophisticated chemical system.
Tests for the effectiveness of treatment revolved around certain levels
of physical changes in the "ante et post" treatment state."Successful"
treatments of some metals, radioactive and non- radioactive, in certain
concentrations in sludges, soils, and liquids as measured by particular
light acid and water leach tests gave additional marketing credence to
the notion that this was a viable and effective treatment for a wide
range of waste types. Very little in-depth research of the chemistry and
physics of S/S and CF as it applies to complex or mixed wastes was done
because "it worked", as determined by certain static and dynamic.
deionized water leach tests. Also the use of the CF term with various
associated unsubstantiated or stretched logic claims as to a given
product or compositions ability to treat a given waste effectively added
another level of apparent creditability. In parallel to the use of S/S over
the past few years there has been a somewhat erratic or wandering but
yet evolving regulatory structure, influenced by an increasingly
negative and dubious public opinion of the true effectiveness of S/S and
CF and it's users. The environmentally active and concerned public
304
-------
elements, and to a increasing degree the regulatory authorities, saw no
true distinction between S/S and CF except marketing hyperbola. In
many groups, S/S and CF is viewed as "low tech, no-tech, or pseudo-
tech" approaches to waste treatment. One important element needed to
solve the effectiveness issue are the questions of what comprises a
reasonable, realistic, or necessary test procedures of S/S and CF
technologies and the "how clean is clean?" standards. The treatment
evaluation methods and standards for S/S and CF are a point of
contention among the users and marketers of S/S and CF, competing
forms of treatment, the regulatory authorities, and the public.
In the midst of this complex scenario we have been doing research into
the basic chemistry of chemical fixation of organics and inorganic toxic
contaminated soils, sludges, and liquids for the past few years. We
believe the necessary reality of S/S or more appropriately defined CF,
involve exceedingly complex chemical mechanisms and phenomena.
And this class of technology should be evaluated on such terms as to its
true effectiveness. One of the major objectives of this article is to
develop a definitional separation between the nature of S/S and CF
using the HWT-20 Series ( Patent Pending. IWT) compositions as the
prototype of a new ACF class of treatment technology. In that regard we
will review GC/MS readings of acid leach and solvent extraction tests of
cases involving soils contaminated with PCBs and other high content
mixtures of organics treated with the International Waste Technologies
(IWT), HWT-20 Series Products. A recently discovered problem with
high content organic wastes treated with a typical S/S mixture using a
quick-lime.pozzo Ian base will toe discussed. We have also used infrared
adsorption (FTIR) and differential scanning calorimetry (DSC) to give
insights into the chemical bonding mechanisms of this particular type of
ACF technology with a range of organic compounds in a pure liquid or
contaminated soil form. It would be wise at this point to give some
background information relative to the chemical design and analytical
thinking that went into the HWT-20 ACF prototype. As was implied
earlier we believe that S/S and CF should be based on an accurate
paradigm of the chemical process rather than a adsorption/dilution
panacea judged effective by weak arguments of diffusion potential and
305
-------
end-state physical characteristics of questionable relevance and
validity. Our position is that the extent and strength of the chemical
bonding and alteration to innocuous forms within the treated matrix is a
truer measure on the short and long term effectiveness of this category
or form of treatment.
Chemistry Overview
This HWT ACF technology is based on three sets of interrelated
functional chemical groups. There is a cement matrix chemistry,
organophilic linking mechanisms, and a free radical and ion attack
mode. The underlying concepts have been discussed in some
intermediate level of detail by us in previous papers, so we will
summarize this time.
Matrix Cement Chemistry: The objective of the cement chemistry is not
primarily end-state physical properties but to facilitate the overall
objective of bonding the toxic molecules and ions within a given
contaminated material. In line with that point certain aspects of the
cement hydration reaction (CHR) are altered and stretched out in time,
the fibrils (sulpho-ferri-hydrates) that exist in the second stage of the
CHR are modified to be more chemically reactive, and caused to be more
dense. Certain admixtures are used to to cause a greater dispersion of
the cement particles in impure environments which in turn will
promote better development of the weak IPN (Interpenetrating
Polymer Network) bonding function. This function is the slowest
reaction of all three functional groups and its primary function is to be
the silicate anchor matrix to which all other reaction products attach.
Free Radical and Ion Attack Chemistry This is a parallel chemistry
that can be made up of a wide range of compounds that produce highly
reactive ions and complexes in the HWT-20 slurry. This activity does
not interfere with the functioning of the other major functional CF
groups. This chemical function should attack various toxic organic and
inorganic elements within the contaminated medium and reduce them
to relatively inert forms or reaction products that can subsequently
306
-------
react with one of the other functional groups in the HWT-20 ACF
material. A simple example of this capability is the use of transition
metal complexes, but even these must be thought out carefully for one
could cause certain counterproductive reactions. An example of this is
shown in Figure 2, Coordination Complexes, and an explanation of the
bonding is given in Figure 3.
Organoohiiic Linking Mechanisms These are intercalation compounds,
such as modified smectite clays, that interact with the organics present,
within certain ranges of predetermined selectivity, by a sorptive
process in either of two general modes. The strong, short range bonding
is based on a Bronsted and or Lewis acid or base reaction, see Figure 4
relative to such a reaction we have observed with triethanolamine. The
weak, long range forces are basically hydrogen bonds, see Figure 1
induced dipole or Vanderwaals forces. There can be an initial reaction
based on the weak force reaction and later a second strong or Lewis
basereaction.
These modified smectite clays have both organic and inorganic
properties due to the substitution in the normally inorganic clay
structure of the Group IA and IIA metal ions that present with
quarternary ammonium ions. This makes them ideal linking
mechanisms between the toxic organics in the waste and the cement
matrix.The introduction of the quarternary ammonium ions also opens
the basal spaces in a pillaring effect to allow like polarity organics into
the strong force reaction zone. Bonds can range from weak Vanderwaal
forces to strong co-ord inate covalent bonds.
The important point is that this is primarily a problem in
supramolecular chemistry and that multiple and secondary bonding
and positioning of the appropriate molecular structures is the key issue.
The individual FTIR shifts on functional groups seen are usually not
considered large and in some cases are slight, but the number of bonds,
the sum of the shifts, and the positioning of the bonds is a stronger
effect in most cases one would see in a primary bond. An indicationof
the strength of this multiple and secondary bonding phenomena can be
307
-------
seen in "Percentage Increase in Energy" of the DSC analysis of a given
waste. ( Figure 4 ). We have also done DSC studies on the treatment of
pure phenol, nitrobenzene, and trichloroethylene and achieved
percentage energy increases of 220.7, 275.9, and 52.8 respectively. An
significantadjunctcondition is that the behavior of a pure substance in
a laboratory gives one an idea of what is possible but in a real, complex
waste many other factors will interfere with the ACF material chemistry
in achieving the maximum desired effect. There are some positive
assisting factors but in most cases one must design ACF materials to
overcome a variety of chemical hurdles before the fixation reaction can
reach the desired level of efficiency. The use of advanced chemical
techniques and analysis provide invaluable insights into the waste
chemical mechanisms.
Fourier Transform Infra Red (FTIR) FTIR studies are used to understand
the extent of interaction between the toxic compounds and the HWT-20
ACF material. This analytic technique measures the changes in
vibrational motion in specific bond relationships. It also helps to
determine the functional elements present in a given molecule and
involved in a bonding process.
DifferentialScanningCalorimetrv(DSC) DSC measures the changes in
temperature and energies associated with various significant chemical
changes involved in bonding, such as the dH of melting, dH of
evaporation, dH of decomposition, dH of phase transition, etc. This
technique can indicate the strength of bonding.
OVERVIEW of IW-S1T1J APPLICATION
The application of the ACF technology can take place above ground level
using a variety of mixing systems or in the ground, in-situ, as was done
at the former General Electric transformer repair facility site in Miami,
Florida. The use of high and or low pressure, rotary shaft injection with
or without mechanical blade mixing has been done for more than ten to
fifteen years in the construction industry for creating injection piles and
308
-------
sub-surface barriers. It was a natural extension of this construction
method to be used in the treatment of contaminated soils and
sediments, if one had a cement chemistry that would prevent both
inorganics and organics from leaching at an unacceptable level. The low
pressure, rotary shaft injection and blade mixing equipment offered by
Geo-Con^ Inc. was chosen because it would give an even, homogeneous
blending of the HWT-20 ACF with the soil and subsurface porous
limestone strata encountered at the site and create accurately placed.
overlapping columns the entire length of the column. An additional
purpose of the in-situ method other than the fact it treats the
contaminated soil in-place is that GE was interested in using this
technique of application in sites where there was volatile and semi-
volatile organic contamination of the subsurface soils where it would
not be desirable to expose these soils to the air by a removal technique.
At the SITE'S Program Demonstration Geo-Con used a relatively small
diameter drill, one yard, for two reasons. The drilling for most of the
time was in rock and the objective was to prove the concept of in-situ
treatment. The mixing drill process could be operated in sand/clay soils
with a larger diameter drill ( 2 to 3 yards ) and or multiple.parallel
drills, up to four shafts quicker and more economically. In-situ
treatment costs can be as low as $20 to 30/yard, excluding treatment
chemicals, on a large project, up to $60 to70/yard in difficult situations.
PCB LEACHING and EXTRACTION STUDIES
The HWT-20 CF Series was successfully used in the treatment of PCBs at
the General Electric (GE) Miami Site under the U.S. Environmental
Protection Agency's Superfund Innovative Technology Evaluation (SITE)
Program. This application was done in-situ using mixing drills down to
6.15 meters from Geo-Con, Pittsburgh. The first 1.2 m was sand, the
second 2.5 m was a porous coral like limestone, and below that quartz
sand, with fresh water at 1.2 m. The level of addition of the HWT-20 CF
material was 15% by weight to soil. In other words for every metric ton
of contaminated soil 150 kg of dry HWT-20 was added in a slurry form.
The maximum concentration of PCB was 5700 ppm. The leaching
309
-------
procedure was the USEPA TCLP (Toxicity Characteristic Leaching
Procedure), 18 hour dynamic acetic acid test. ANS 16.1, and the MCC-1
leach procedure. Essentially no PCBs came out in the EPA testing and
only one sample was found to leach in GE's testing and that was 1.2 ppb
on a two week old sample. GE did methylene chloride extractions of drill
core samples as well and did not find in excess of 206 ppm PCBs in the
treated samp les us ing GC/ECD.
In some pre-project laboratory experiments carried out by Dr. R.
Soundararajan gave some insights into what was occurring in the
treated soil matrix. In a sample of similar type PCB contaminated soil,
eight-hour methylene chloride extractions were performed on the
untreated and crushed treated samples.
Analysis was done by GC/MS with the machine calibrated against all
221 position isomers of PCBs. In one sample of this experiment an
admixture was included that produced a sulfurous acid that would
totally disable the organophilie clay linking mechanisms between the
PCBs and the slower developing silicate based anchoring matrix. In the
second sample the admixture was removed so the primary linking
mechanisms could function.
The results are as follows, untreated soil released 28,800 ppm PCBs in a
methylene chloride extraction.
Sample 1: Admixture Sample 2: No Admixture
Treated Extraction 26,437 ppm 2,800 ppm
Treated TCLP 10.437 ppm 12.5 ppm
These tests were done only three days after treatment. A treatment
dilution factor of 20% is included in above numbers. TCLP numbers did
improve with the age of treated sample. This particular sample showed
only a 5-6,000 ppm value using standard GC/ECD analysis mainly
because the GC/MS could sort out all 221 PCB isomers. Also high
values(40-50,000 ppm) of chlorinated benzenes were found, most likely
the decomposition products of the PCBs, since no chlorinated benzenes
were ever used at the site. They were reduced in a similar proportion to
310
-------
the reduction in PCBs of Sample 2 from untreated to treated extraction
values. Almost all the existing PCB isomers after treatment were the low
value chlorine forms.
The conclusions of the above experiment were supported further in
another set of leach and extraction tests performed on a sample of a
clay/sand soil contaminated with low levels of PCBs(290 ppm). The
testing procedure was the same as above with the focus on which
isomers were bonded or retained in the treatment matrix after a
solvent extraction of crushed treated samples for eight hours. The TCLP
leach tests of the treated material were all non-detectable. The sample
cured for seven days. The treatment level of HWT-20 was 15% by
weight to the weight of soil. In the untreated soil there was some
chlorobenzenes and substituted phenols but none were found in the
treated. Only the lighter PCB isomers (tri.tetra, and penta ) were found
in the treated. The hexa and heptachlorophenols were not found in the
treated. The total PCB content extracted was 190 ppm or 65%.
Relative to PCBs the current HWT ACF treatment technology is able to
alter or bond to a high degree the heavier chlorinated PCB molecules
and a lessor degree the lighter chlorinated molecules. It is very
effective in preventing the leaching of PCBs against the TCLP of all
types. Also the HWT-20 ACF treatment sufficiently bonds and prevents
the leaching of the PCB decomposition products, substituted benzene
and phenols compounds. Newer formulations of a more advanced IWT
ACF product show significantly greater rates of bonding or chemical
alteration of organics including PCBs. An example of the treatment of a
PCB containing waste with our newer "Polyfunctional Reactive Silicates"
(PFRS) (TM) is shown in Figures 5 and 6. These ACF materials are based
on new inorganic carceplex structures and heretofore non-existent
organic trailers. As the results are shown it is are most effective PCB
reaction to date and the research is continuing to be positive. The first
of these new PFRS materials should be commercially available later this
year.
311
-------
Treatment bv ACF of High Content Organic Waste
An organic content waste with a high percentage of heavy hydrocarbons
with relative trace to small fractional loadings of volatile and semi-
volatile toxic compounds is normally a difficult material for the usual
S/S mixtures to effectively treat unless some integer multiple by weight
of the S/S material is added to the weight of the waste and end-state
physical properties are all that is being considered. The waste sample in
this case was a soil with a heavy concentration of long chain
hydrocarbons from a acid/clay process for recycling used oil. The major
organic toxic components of this waste as determined by a solvent
extraction and analyzed by a GC/MS were the following:
Untreated Extraction Treated Extraction TCLP
Bis(l-chloro iso propyl) ether 8.528 ppm ND ND
Naphthalene 18,060 ppm 1445 ppm ND
Phenanthrene 20,184 ppm ND ND
Benzo (A) anthracene 30,460 ppm ND ND
This contaminated sample was treated 50% by weight with a
experimental and more advanced CF material and allowed to cure for
only two days. This treatment level is too high for an actual project but
what we were trying to achieve is an accelerated effect that would
allow us to examine the bonding activity more quickly by the solvent
extraction (GC/ MS), FTIR, and DSC analysis. A dilution factor of 50% was
used in all quantitation. Standard methods of analysis and QA/QC were
used.
With such a complex mixture to analyze by FTIR and DSC the approach
has to be different then working with a pure known organic liquid. In
using FTIR we focused on functional groups that we knew were there in
relatively large concentrations and looked for shifts within those groups
to indicate a level of bonding activity. In reviewing the data we have
found that there were significant shifts in a number of FTIR frequency
regions, hydrogen bonding was occurring between the aliphatic amines.
hydroxy compounds, and the oxygens in A12O3 and Si02 in the CF
312
-------
material. There was a amine salt formation as the result of a Lewis base
reaction at 1580 cm-1. The most significant shift occurred at 1005 cm-1
where a hydrogen bond was formed with an Si02. This is explained by
the fact that the oxygens off the silica strongly interact with the
hydrogens of alkyl, hydroxyl, and ami no groups which results in a
reduction of the 0-Si-O bond order. (Table 1, Infrared Data)
The DSC data (Table 2, DSC Data) also confirms that relatively significant
bond ing activity is occurring. The dH of vaporization has increased by
54.9% from untreated to treated. The DSC analysis is a energy
summation function rather than a focus on a specific reaction.
Recent experimental work done by Dr. R. Soundararajan has indicated
that lime, lime/fly-ash, or pozzolan based S/S of organic content waste
of a sufficient level, what this threshold is is not known yet, but
certainly the oil recycling waste in this case applies, will generate
significant carbon monoxide and or acetylene gas when exposed to
water of even mild acidity. Also during the process of S/S or CF
treatment it is desirable to keep the heat generation by the reaction
with water as low as possible since the more heat generated the more
the volatilization of the organics. The IWT ACF technology does not
contain but trace amounts of lime in the cement used in a fraction of it's
composition and no pozzolans are used. The rate of addition of the IWT
ACF materials used to weight of waste is relatively low, 12 to 20%, and
has a lower heat of hydration than straight cement.
Summation
Chemical fixation technology is far more advanced and cost/effective
than most people realize, especially from those companies or groups
that have done in-depth chemical research of the basic fixation reaction
mechanisms. International Waste Technologies has and is doing this
basic research and believes that the HWT ACF Product Series and the
new, more advanced products, PFRS, that will come out in the near
future will compete favorably with thermal and biological methods in
the destruction, alteration, and or bonding of organics and the
313
-------
immobilization of inorganics. The effectiveness of ACF technology can be
verified analytically with relative quickness and function in a wide
range of contamination environments at relatively acceptable costs and
ease of use factors. An important key to the success of this category of
toxic waste treatment is for the regulators and users of this type of
technology insist on high standards of relevant chemical testing and
verification in the pre-project treatability analysis and careful QA/QC
during project application.
In-situ applications that are effective in terms of their homogeneity of
mixing of the ACF material with the contaminated soil to the required
level and do not leave any voids can be an effective, economical, and
necessary part of the waste treatment picture. The Geo-Con system
successfully demonstrated that objective.
Copyright
J. P. Newton - June 7. 1989
International Waste Technologies. Inc.
150 North Main-Suite 910
Wichita. Kansas 67202
316-269-2660. Fax-316-269-3865
314
-------
TABLE 1
INFRARED DATA
Waste
Extract
cm-1
3394
3382
3373
1597
1035
Waste
Extract + Binder
cm-1
3385
3372
3355
1580
1005
Infrared
Frequency
Shift
cm-1
-9
-10
-18
-17
-30
Infrared
Functional
Group
Assignments
OH Stretch O- - -H....O
NH Stretch N- - -H....O
Keto group C --"O....H
Hydrogen Bonding Si Si
\/
O
H
-------
TABLE 2
DSC DATA
Waste Extract
Waste Extract + Binder
DSC Endothermic Peak Values
Temperature
(Degrees C)
138.90
121.2
414.4
414.5
H (Vaporization)
Cal. per gram
18.63
6.15
2.32
2.06
10.53 Total
CO
I—>
CTl
Total H (Vaporization)
Total H (Vaporization)
corrected to 100%
Waste Extract
Cal/gram
18.63
Waste Extract
+ Binder
Cal/gram
10.53
28.65
TGA - Percent wt. loss at
given temperature range
Waste Extract + Binder
36.48
PERCENT INCRESE in H (Vaporization)
for Waste Extract + Binder
54.9
-------
FIGURE 1
Observed Bonding Phenomenon
in the HWT-23 Treatment Matrix
Hydrogen Bonding of Phenol Molecules
Oxygen
,o
H-
•o
\
o^,
o'
H
Oxygen
Aluminum
Oxide Layer
Phenol Molecules
Silicon Dioxide
Layer
Evidence:
Lowering of FTIR Stretching Frequency
Phenol
962 cm 1
3640 cm -1
Treated Phenol
934 cm 1
3632 cm1
317
Shift
-28cm1
-8 cm 1
Peak Assignment
C-O
H - bonded OH
-------
FIGURE 2
Coordination Complexes
3+
6 (C6 H5 OH) + M(Transition
(C6 H5 OH)
CfiHsOH
Octahedral Phenol
Metal Complex
C6HSOH Q^^
M+-\
C.H.OH
\
X
XC6H5OH
C6HSOH
Evidence:
UV - Visible Spectra - Drastic Color Change
318
-------
FIGURE 3
PIT to dor Bonding
TT Electron charge clouds
Metal dir orbitals (empty
or partially filled orbitals)
Evidence:
Shift in FTIR Frequency for Ring Breathing
319
-------
FIGURE 4
Lewis Acid Base Reactions:
Formation of Sigma Bonds (a)
Aluminum Oxide Layer
FTIR ANALYSIS
Triethanolamine
2104
1075
Treated
2297
1070
Shift
193
-5
Peak
Assignment
Amine Salt
Formation
H bonding
Silicon Dioxide Layer
X = Electron deficient species or H+which is
a Lewis Acid. Evidence: Shift in N - R (n)
frequency in positive direction (increase)
Edotherms
(°C)
150.97
337.30
DSC ANALYSIS - TRIETHANOLAMINE
Hof
vaporization Observed H of Percentage Boiling
(literature) vaporization Increase in point
Kcal/mol Kcal/mol energy (°C)
12.78
24.16
89.0
335.4
Highest
endotnermic
temperature
337.30
320
-------
FIGURE 5
Formation of Permanent O-+TT Bonding
(Covalent Bonding)
A = O + H2X
ACF Material Toxic
Waste
H,O
Irreversibly Bonded
End Product
321
-------
FIGURE 6
Supramolecular Chemistry/
Multiple and Secondary Bonding
Dichlorobiphenyl
AI2O3 Layer Q O
O O
Solvent Extraction and Leach Results:
PCB
Untreated
(ppm)
17,580
Solvent Extraction of Treated (ppm)
N.D.
TCLP
N.D.
Dichlorbiphenyi
1150cm-1
FTIR Study of PCB
Treated PCB
1118cm1
Shift
-32
*U,S,COVERNMENT PRINTING OFFICE: 1990-7*8- 1 5?0 0 ts 2
322
Peak Assignment
CjtoAl
Coordinate Bond
-------
-------
------- |